Close
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Thick concepts, reflection, and the loss of ethical knowledge
(USC Thesis Other)
Thick concepts, reflection, and the loss of ethical knowledge
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THICK CONCEPTS, REFLECTION, AND THE LOSS OF ETHICAL KNOWLEDGE by Francis J. Pryor V A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (PHILOSOPHY) December 2009 Copyright 2009 Francis J. Pryor V ii Epigraph: Every piece of philosophical writing is primarily addressed by the writer to himself. Its purpose is not to select from among his thoughts those of which he is certain and to express those, but the very opposite: to fasten upon the difficulties and obscurities in which he finds himself involved, and try, if not to solve or remove then, at least to understand them better. R.C. Collingwood, An Essay on Philosophical Method …moral notions do not evaluate the world of description; we evaluate the world by the help of descriptive notions. Moral notions describe the world of evaluation. If this sounds strange, then we have become aware of the framework within which contemporary moral philosophy moves. Julius Kovesi, Moral Notions iii Dedication: This dissertation is dedicated to my wife, Mindy, for her constant support and encouragement. iv Table of Contents: Epigraph: ............................................................................................................................. ii Dedication: ......................................................................................................................... iii Abstract: ............................................................................................................................ vii CHAPTER 1: Component Views of Thick Concepts ......................................................... 1 Section 1.0: Introduction ................................................................................................. 1 Section 1.1: CV Approaches to Thick Concepts ............................................................. 7 Section 1.2: Family Resemblances and a Thick Concept’s ‘Point’ .............................. 19 Section 1.3: Sophisticated-CV Approaches .................................................................. 25 Section 1.4: Sophisticated-CVs and Thick Psychological Descriptions ....................... 29 Section 1.5: Some Modest Suggestions on the Nature of Cruelty ................................ 35 Section 1.6: Thick Concepts and Quasi-Realism .......................................................... 43 Section 1.7: Concluding Remarks ................................................................................. 61 CHAPTER 2: Thick Concepts and the FV-View ............................................................. 63 Section 2.0: Introduction ............................................................................................... 63 Section 2.1: From Thick Concepts’ Shapelessness to their Justification ...................... 66 Section 2.2: Assessing McDowell’s Thought-Experiment ........................................... 72 Section 2.3: Wiggins and McDowell on Thick Concepts ............................................. 80 Section 2.4: Gibbard and the Warrant for Thick Feelings ............................................ 86 v Section 2.5: Assessing Sentimentalist Treatments of Thick Concepts ......................... 92 Section 2.6: Concluding Remarks ................................................................................. 98 CHAPTER 3: The Moral Psychology of Thick Concepts .............................................. 102 Section 3.0: Introduction ............................................................................................. 102 Section 3.1: The Basic Structure of Thick Concepts .................................................. 105 Section 3.2: Thick Concepts and Their Use ................................................................ 117 Section 3.3: Thick Concepts and Characteristic Reasons ........................................... 122 (a) Reasons Characteristic of (Some) Thick Concepts ...................................... 124 (b) Extending the Account: Some Negative Thick Ethical Concepts ................ 130 Section 3.4: Thick Concepts and Practical Reasoning ................................................ 137 Section 3.5: Concluding Remarks ............................................................................... 149 CHAPTER 4: Objectivity, Reflection, and the Loss of Ethical Knowledge .................. 154 Section 4.0: Introduction ............................................................................................. 154 Section 4.1: An Overview of Williams’ View of Thick Concepts .............................. 157 Section 4.2: Understanding Williams’ Nonobjectivism .............................................. 170 (a) Thick Concepts and Ethical Knowledge ...................................................... 170 (b) Objectivity and the Absolute Conception of the World ............................... 179 (c) Vindicatory Histories and the Ascription of Error ....................................... 194 (d) Recoiling from the Idea that there is Thick Ethical Knowledge .................. 200 vi (e) Nonobjectivism and the Hypertraditional Society ........................................ 204 Section 4.3: Objections to Williams Nonobjectivism ................................................ 211 Section 4.4: The Loss (or Destruction) of Ethical Knowledge ................................... 217 Section 4.5: Concluding Remarks ............................................................................... 235 CHAPTER 5: Lost Ethical Knowledge and the Desire for Transparency ...................... 239 Section 5.0: Introduction ............................................................................................. 239 Section 5.1: Criteria, Ethical Standards, and Disagreement ....................................... 246 Section 5.2: Ethical Rationales and Action-Guidance ................................................ 256 Section 5.3: Identifying with the Point of a Thick Concept ........................................ 264 Section 5.4: Reflection, Vindication, and the Loss of Ethical Knowledge ................. 270 Section 5.5: Lost Ethical Knowledge and the Need for Confidence ........................... 280 Section 5.6: Transparency and (Post-)Modern Ethical Authority ............................... 289 Section 5.7: Concluding Remarks ............................................................................... 304 Bibliography: .................................................................................................................. 311 vii Abstract: My dissertation defends two controversial and seemingly incompatible theses, both drawn from the moral philosophy of Bernard Williams. The first is that a type of ethical knowledge is possible. The second is that such knowledge can be undermined by certain forms of critical reflection. Both of these claims are tied closely to the possibility of there being (what are sometimes called) ‘thick’ ethical concepts: ethical concepts whose application is guided by users’ perceptions of acts and situations, and which can also guide their actions, by providing them with practical reasons of various sorts. The knowledge that comes with the disposition to actively use a thick concept is that certain defeasible practical reasons follow from the facts of a situation which determine whether such a concept should apply. In Chapters 1 and 2, I critically discuss several views as to how we should understand the nature of thick concepts, and consider, in particular, the question of whether the factual characteristics associated with a thick concept are logically independent of the practical and evaluative reasons that such a concept is thought to give rise to. I argue that they are not. In Chapter 3, I develop an account of why this is so, focusing on how such concepts come to be developed in the first place. Their development, I claim, is driven by antecedent interests that a group has in regulating social interactions. Certain characteristics of acts and individuals are selected by a group for special recognition, be it positive or negative, because encouraging or discouraging such characteristics accords with other interests and concerns that group has. These antecedent interests have an explanatory role to play both in the selection of the factual viii characteristic, and in the practical status granted it. Because of this, thick concepts cannot be properly understood without a grasp of the reasons why such characteristics were selected for recognition. In Chapter 4, I explain and defend Bernard Williams’ claim that, although thick concepts can provide a type of ethical knowledge, such knowledge is not of an objective subject matter. Here I defend Williams’ views on objectivity against the criticism that this position amounts to an objectionable form of metaphysical realism. In contrast, I argue that Williams’ idea of an objective subject matter should not be construed in realist terms at all, but rather as something importantly related to whether we can explain why certain judgments are correct or erroneous in a manner that is neither vacuous nor inherently question-begging. Finally, in Chapter 5, I return to the account of thick concepts started in Chapter 3, and defend Williams claim that the knowledge these concepts provide can indeed be undermined by critical reflection. The knowledge under a thick concept can be undermined when the capacity for believing the judgments involving that concept is lost. Reflective questions on whether a thick concept is the right or best one to use (as a means for regulating social interactions) cannot be answered affirmatively in a non-vacuous, non-question begging way, and an awareness of this can erode one’s confidence in the judgments which deploy that concept. I end the dissertation by considering the more positive role that critical reflection has in Williams’ thought, as something that serves a desire for transparency in the reasons why a group accepts or identifies with the use of a thick concept as an appropriate means for regulating social life. 1 CHAPTER 1: Component Views of Thick Concepts Section 1.0: Introduction In Ethics and the Limits of Philosophy, 1 Bernard Williams argued that a rigid (and altogether superficial) distinction between ‘fact’ and ‘value’ failed for certain class of ethical concepts. 2 Williams’ aim, in highlighting this fact, was to encourage moral philosophers to get past it, and to see that if there was a distinction to be made it should be found in the differences between ethical and scientific knowledge, and not in the idea that we should treat ethical concepts as merely contributing to evaluations. Some ethical concepts were, to put the point in a cursory manner, descriptive—they enabled claims about actions and individuals that could be straightforwardly true and false, but where those claims involved some general connection to practical reasons. In many ways, Williams claim was not an original one: what was unique to his view, were the consequences he drew from it—consequences that have yet to be fully appreciated in contemporary metaethics and moral philosophy. In this chapter, my aim is not to discuss Williams’ insights, those will be left for Chapter 4, but to uncover and elaborate the argument that some ethical concepts cannot be adequately accounted for by a rigid distinction between fact and value. This argument centers on the claim that there are thick concepts: concepts that involve a blending of the idioms of fact and value in such a way that they cannot be adequately separated from one another. 1 (Williams 1985). See also (Williams 1996b). 2 Williams cites Philippa Foot and Iris Murdoch as exposing him to thick ethical concepts in a seminar they led at Oxford in the 1950s. See (Williams 1985, p. 141n7). See also (Foot 1978a), (Foot 1978b), and (Murdoch 1970). 2 As Williams put it, such concepts are at once “world-guided” and “action- guiding.” 3 (That Williams chose to make his point in these terms, as opposed to saying that they are at once ‘factual’ and ‘evaluative’, is significant, and should not be read as merely a difference in wording, but explaining the significance of this will have to wait until we have explained, what I take to be, the argument for their being such concepts.) The idea that certain ethical concepts were thick was intended to highlight the idea that their world-guided and action-guiding nature could not be analyzed in a way that would reveal how these aspects could be separated from one another. Examples of such concepts are: CRUEL, 4 MURDER, LIE, PIOUS, GENEROUS, SNOBBISH, COURAGEOUS, LECHEROUS, JUVENILE, THOUGHTLESS, ADULTEROUS, RUDE, DISHONEST, LEWD, KIND, MEAN, UNFAITHFUL, COMPASSIONATE, LOYAL and INCONSIDERATE. Williams contrasted these sorts of concepts with other thinner ones, concepts like RIGHT, GOOD, and OUGHT. These latter concepts are considered ‘thin’ (or ‘thinner’) because they lack the sort of world-guided character that thick concepts possess—they are, in Williams’ terms, merely action-guiding. While concepts like RIGHT and GOOD apply to objects only in virtue of their being some right- or good-making property which the object possesses, this is understood as a purely formal constraint on their use. These are general (perhaps the most general) terms for noting correctness and agreement (in the case of RIGHT) and for offering commendation (in the case of GOOD), but there is no determinate range of properties that guides the application of these concepts. 5 Thin concepts have little or no 3 (Williams 1985, pp. 129; 140-41) 4 I will use of SMALLCAPS to indicate concepts in particular, as opposed to terms, features or properties. 3 restrictions on how they can be appropriately applied. By contrast, part of what seems to makes a thick concept thick is that its range of application is tightly constrained. The tempting thought is to conclude that such concepts apply to acts and individuals because of their possessing some determinate factual property, whereby the possession of such a property would be taken, by users of the concept, as action-guiding, that is, as providing those who use the concept with a reason to act or evaluate in some determinate way. As we will see however, although this picture is tempting, it is deeply misguided: and assumes a certain framework for explaining the nature of these concepts that Williams was intent on unseating. Williams, of course, was not the first to emphasize the importance of such concepts. 6 What is interesting about Williams’ view is that, while he agreed that such concepts genuinely involved an inextricable blending of factual and evaluative considerations, he denied that this claim supported the view that ethical judgments (involving such concepts) reflected any strong form of realism or objectivity. (Again, a substantial discussion of what Williams has in mind by the notions ‘realism’ and ‘objectivity’ will have to wait until later, but we might say, in a quite cursory fashion, that such judgments were not about the world as it is anyway, independent of any particular perspective on it. We will return to these topics in Chapter Four.) Judgments involving these concepts could be true and correct, and, in being so, supported a form of 5 Foot, however, argued that the concept of (moral) goodness was not completely open-ended, in that, someone could not be considered good simply because they clapped their hands every quarter hour or drank mud. See (Foot 1978b). Foot’s point, I take it, was to show that without some special background considerations that gave a point to these sorts of actions, the application of such a concept would seem to us absurd. 6 See Note 2 above as well as (Anscombe 1958). 4 ethical knowledge. This knowledge, however, was limited in scope, and this limitation blocked the sorts of aspirations he saw in other thinkers to move from the cognitive character of such judgments to some account of moral realism or moral objectivity. 7 These aspirations went well beyond what he thought our understanding of the concepts could support. Those who oppose the idea that thick concepts involve an inextricable blending of factual and evaluative considerations argue that such concepts can be factored into separate factual and evaluative components. 8 Such views tend to be advocated by noncognitivists (prescriptivists 9 or expressivists 10 )—those who feel that evaluative judgments are essentially expressions of commands, feelings, attitudes, or some other practical state, as opposed to the cognitivist view that evaluative judgments can be genuinely true and false 11 and capable of sustaining propositional knowledge. This, 7 See here (Platts 1979); (Lovibund 1983); (Lovibund 2002); (McNaughton 1988); (McDowell 1998c); (McDowell 1998f); (Wiggins 1998a); (Wiggins 1998b); (Hurley 1989), Ch.2; and (Dancy 1995). 8 Two central opponents are R.M. Hare and Simon Blackburn. For Hare, see (Hare 1952, pp. 118-20); (Hare 1963, pp. 188-90); (Hare 1981, pp. 70-1); and (Hare 1997, p. 61). For Blackburn, see (Blackburn 1981); (Blackburn 1984, pp. 148-50); (Blackburn 1992); and (Blackburn 1998), Ch. 4. See also (Bower 1988); (Burton 1992); and (Tappolet 2004). 9 Prescriptivism holds that evaluative judgments are to be explained in terms of their involving some appeal to a general and universalizable prescription. See (Hare 1963) as a central example. 10 Expressivism holds that evaluative judgments are to be explained in terms of some antecedent, though inherently practical, mental state that is expressed in the making of the judgment. See (Gibbard 2003b) and (Blackburn 1998) for contemporary explanations of this view. 11 Since expressivists hold that evaluative judgments are expressions of essentially practical states, such as attitudes, policies, or stances, it is typically thought that they must reject the view that such judgments are capable of being true or false. This has historically been seen as the central distinction between cognitivists and noncognitivists. Drawing it in these terms, however, is no longer as effective as it once was, for two reasons: (i) Few, if any, cognitivists hold that there are evaluative facts that are wholly independent of human evaluative sensibilities, and the attitudes, interests, and concerns that underlie them; and, perhaps more significantly, 5 however, is not exclusively the case—some argue that thick concepts can be factored into factual and evaluative components, even though this fact is neutral with respect to the larger debate between cognitivists and noncognitivists. 12 However, in siding with those who deny that thick concepts can be factored into factual and evaluative components, Williams agreement is only partial: both agree on the basic nature of such concepts, and on the idea that such concepts support a form of ethical knowledge, but Williams denies that these facts are best construed as yielding knowledge about an objective subject matter. Such concepts depend too heavily on an ethical form of life, and in particular on the interests and concerns that are partially constitutive of that form of life, to see their use as involving any strong claim to objectivity. Indeed, their dependence on such interests and concerns makes them, according to Williams, reflectively unstable: coming to a more reflective understanding of what these concepts involve has the potential to undermine a group’s ability to continue using them. This disconnect, between the thought that thick concepts yield a suitably credible notion of ethical knowledge and the thought that their use is not capable of achieving reflective endorsement, is what grounded Williams’ claim that critical reflection could (though not necessarily would) destroy the knowledge they provide. Such knowledge is destroyed by the fact that, in developing a reflective understanding of them, a group could come to find that they could no longer use them. (This too is a complicated claim, and will require significant (ii) The concept of truth is now taken by many to be best understood in minimalistic terms, that is that the notions of truth and correctness are not a metaphysically robust ones. On the former point, see (McDowell 1998d), (McDowell 1998e), and (Dworkin 1996). On the latter, see (Wright 1992), Chapters 1 and 2; as well as (Blackburn 1998), Ch. 9. 12 Burton and Tappolet are noteworthy here as they see their accounts as neutral with respect to the cognitivist/noncognitivist debate. 6 unpacking in Chapters 4 and 5.) It is in making this claim that Williams’ metaethical position is unique—for even though he agrees that there can be ethical facts and ethical knowledge under the use of thick concepts, he denies that this overcoming of fact-value distinction bolsters any serious claim to ethical objectivity. The goal of the dissertation is to elaborate and defend Williams’ unique position. Since Williams’ views on ethical knowledge (and how that knowledge relates to our ability to reflectively endorse the use of the concepts that provide it) depend crucially on his view of the nature of thick concepts, the first step in defending his position will be to critically review the some contemporary accounts that aim to deny that thick concepts have the sort of nature that Williams took them to have. This will occupy the discussion of the first two chapters. Chapter 1 will consider a family of views that treats thick concepts as separable into evaluative and non-evaluative components, what we might simply label as the Component-View (or CV for short). What I hope to motivate in this discussion is that CVs are indeed problematic. Their plausibility rests largely on assuming that the very distinctions that thick concepts are thought to overcome, namely those of ‘fact’ and ‘value’ and ‘description’ and ‘evaluation’, are clear and cogent. What will come out of this discussion, however, is that some of the more sophisticated variants of this approach fail to show their approach is a genuine threat to the claim that thick concepts involve an inextricable blending of factual and evaluative components. Chapter Two will then shift to the other side of the debate and look at some accounts that advocate the sort of view that Williams’ endorses (what we can call Fusion View, or just FV). FV theorists claim that thick concepts involve a fusion of fact and value, or 7 evaluative and nonevaluative considerations, which cannot be separated into completely independent components. Here I will hope to motivate the thought that these accounts seem committed to an overly close tie between thick concepts and our affective reactions that renders them problematic when we try to generalize the picture they have to offer. From these discussions, I will offer up my own view of thick concepts in Chapter 3. Though I do not make any strong claim to originality, my account will attempt to capture the motivations behind the FV view, without being saddled with the sorts of problems that are inherent to it. The second part of the dissertation, then, will be an articulation and defense of Williams’ views on ethical knowledge, objectivity, and the potential for instability that he saw in thick concepts when a group comes to critically reflect on their nature. This will occupy the discussion in Chapters 4 and 5. Chapter 4 will defend a version of Williams’ skepticism about being able to reflectively endorse the use of thick concepts, and argue that this view is less paradoxical than it has seemed to some. Finally, in Chapter 5, I will use the account of thick concepts developed in Chapter 3 to (i) explain how ethical knowledge can indeed be lost in the manner that Williams supposes, but (ii) show that this view also points to a positive role that critical reflection might have for the ongoing maintenance and stability of a group’s thick ethical concepts. Section 1.1: CV Approaches to Thick Concepts To begin, then, we need to articulate the basic differences between CV and FV views more carefully. The central difference between the two views centers on whether thick concepts involve isolable evaluative and nonevaluative (purely factual or purely 8 descriptive 13 ) components—CVs take there to be some interesting sense in which this is so, while FVs deny this. Even this, however, is somewhat quick as CVs, to be sure, constitute a family of views. We might begin, then, by distinguishing (what I will call) basic-CV, from more sophisticated-CV, approaches. On a basic-CV approach, the question of whether there are such components centers on whether the extension of a thick concept can be understood as completely determined by nonevaluative considerations. On this reading, thick concepts involve nonevaluative sufficient conditions. A basic-CV theorist claims that there is some purely descriptive property that a thick concept answers to, and that this property is intelligible classification apart from the ethical outlook or point of view of those who use the thick concept in question. So, for instance, a thick concept like KIND would, on this view, apply to acts and individuals on the basis of those acts and individuals having some property that would be understandable apart from the interests and concerns of the ethical outlook that uses that concept. More sophisticated-CVs developed in response to objections that were raised against the basic-CV approach, in particular the claim that the extension of a thick concept could be seen as constituting an intelligible apart from an understanding the ethical perspective of which that concept was a part. Their general strategy is to move away from thinking that thick concepts involve nonevaluative sufficient conditions to thinking that they only involve nonevaluative necessary conditions. This left room for 13 The distinction between evaluative and descriptive is admittedly a notoriously opaque feature of this debate. What seems generally assumed, however, is that the descriptive/evaluative distinction mirrors or is explained in terms of a representation/attitude distinction. That this is so seems clearest in Blackburn’s discussions of thick concepts (see (Blackburn 1991), (Blackburn 1992), and (Blackburn 1998)) and his general distinction between inputs (representation) and outputs (attitudes, stances, or other inherently practical mental states). 9 the thought that the extension of a thick concept could be controlled by both facts and values, but where this control could be explained in terms of a split in the criteria used for applying them (that is such concepts would involve both evaluative and nonevaluative criteria). And so, while sophisticated-CVs argue that thick concepts involve distinguishable evaluative and nonevaluative components, the relevant nonevaluative component is unable to explain why the concept applies in a given context. So, following the previous example, a sophisticated-CV theorist would argue that, while there is no neutral property that determines whether an act is kind or not, there are neutral constraints (say an act must be done in the right state of mind or for the right reasons), but that these constraints where independent of whatever evaluative constraints the concept of kindness might also involve (say that the act advances the well-being or good of the person being helped). This allowance, however, as I will suggest later, obscures the sense in which such nonevaluative considerations can be thought of as distinct components in a thick concept. The central argument that has galvanized the debate between CVs and FVs is one offered by John McDowell, in a series of papers. 14 The basic thought behind this argument is that, when it comes to thick concepts, the extension of the concept is something that is sensitive to the evaluations (or evaluative attitudes) of those who use the concept in a way that subverts a CV approach in principle. McDowell’s point, echoed again by Williams and Jonathan Dancy, 15 was that understanding a thick concept 14 (McDowell 1998b); (McDowell 1998c); and (McDowell 1998f). 15 (Williams 1985), pg. 142, and (Dancy 1995), pg. 276. 10 depended heavily on understanding the ‘form of life’ 16 or the ‘conception of how to live’ that informs the use of such a concept. The idea is that the ethical outlook (or point of view) of someone who uses a concept like KIND is not something that can be treated as an separable element, when considering how someone would deliberate over how to apply such a concept from one context to the next. Understanding such an outlook then is essential for understanding the concept, but the outlook itself cannot be represented as simply an evaluative-extra that was inessential to determining when and why the concept applied. McDowell notes that: We do not fully understand the virtuous person’s actions—we do not see the consistency in them—unless we can supplement the core explanations [of his behavior 17 ] with a grasp of his conception of how to live. (McDowell 1998f, p. 71) The thought here is that, unless we know the virtuous person’s conception of how to live, we will be unable to see how his various actions hang together as an expression of that conception (say as an expression of the virtue of kindness). This thought is given a complimentary treatment, when we consider that, from the standpoint of the virtuous person’s community, such a state will itself be one that is conceptualized by them in a certain way as a state that is admirable or praiseworthy, and which is generally worth modeling and inculcating. From the standpoint of such an outlook, kindness is generally taken to be an admirable quality of actions, though it is a quality of actions that can be outweighed in certain cases by an act’s other (negative) qualities. From such a standpoint, if an act is kind, this is always a reason to positively evaluate that action, even 16 (Wittgenstein 2001 [1951]), § 241. See also (Cavell 1969) and (Cavell 1979), especially Section 3. 17 Core explanations would be those that reveal what the agent took to be relevant or salient about the situation in question that was a determinate in his acting as he did. 11 though what counts as kind, from one context to the next, is not something that can be elucidated without appealing to the outlook of those who consider kindness a virtue. Given this, if there is a problem in seeing how the virtuous person’s actions hang together without importing a certain conception of how to live, it will be equally problematic when trying to make sense of the manner in which those actions are conceptualized from the standpoint of the relevant community, which, we are assuming, shares that conception (it is, after all, their virtue). If the deliberative state of the virtuous person 18 fails to reveal some neutral consistency in action and attitude without granting the related conception of how to live some explanatory role, then so too will be the manner in which that community thinks and talks about that state (in using the related thick concept). Yet, from the standpoint of a basic-CV this should not be the case: as it should be a neutral (i.e., independently intelligible) feature of various acts that prompts the use of the relevant thick concept. And so the importance of imputing a conception of how to live, in making sense of the virtuous person’s behavior, extends out to understanding certain of the community’s conceptual practices. As such McDowell writes (and it is worth quoting in full): Consider, for instance, a specific conception of some moral virtue: the conception current in a reasonably cohesive moral community. If the disentangling maneuver is always possible, that implies that the extension of the associated term, as it would be used by someone who belonged to the community, could be mastered independently of the special concerns that, in the community, would show themselves in admiration or emulation of actions seen as falling under the concept. That is: one could know which actions the term would be applied to, so that one would be able to predict applications and withholdings of it in new cases—not merely without oneself sharing the community’s admiration (there need be no difficulty about that), but without even embarking on an attempt to 18 That is, the manner in which a person, possessing the relevant virtue, deliberates about how he or she should act, given his or her understanding of a situation. 12 make sense of their admiration. That would be an attempt to comprehend their special perspective; whereas, according to the position I am considering, the genuine feature to which the term is applied should be graspable without the benefit of understanding the special perspective, since sensitivity to it is singled out as an independent ingredient in the purported explanation of why occupants of the perspective see things as they do. (McDowell 1998c, pp. 201-02) Bernard Williams makes a similar sort of argument in noting that: An insightful observer can indeed come to understand and anticipate the of the concept without actually sharing the values of the people who use it…But in imaginatively anticipating their use of the concept, the observer also has to grasp imaginatively its evaluative point. He cannot stand quite outside the evaluative interests of the community he is observing, and pick up the concept as a device for dividing up in a rather strange way certain neutral features of the world. (Williams 1985, pp. 141-42) What we are asked to consider, by both McDowell and Williams, is how someone who was an outsider to the community, of which some thick concept was a part, could come to understand that concept. One understands the concept if one is competent in using it, and one is competent in using it if one knows how to apply the concept to novel cases. To accomplish this, one must be able to grasp the pattern of features that is shared between past cases and this novel case. If things are as a CV-theorist takes them to be, then grasping this pattern should be possible without having to appreciate the interests and concerns that the community has in using the concept. Of course, there are concepts where this indeed seems possible: ethnic and racial pejoratives (e.g., n****r, k**e, etc.) reflecting the clearest examples. And it is no accident that CV-theorists have drawn upon these examples to bolster their argument. 19 Such terms function as linguistic devices of derogation, and their success in such a role typically presupposes an independent way of delineating the group that is marked for 19 Both Hare and Blackburn appeal to pejoratives as a model for understanding thick concepts. Cf. (Hare 1963), Section 2.7 as well as (Blackburn 1992) and (Blackburn 1998), Chapter 4. 13 contempt or derision. 20 (They could not be derogative unless these terms provided a ‘special’ way of referring to some group, and this entails that that there is some other way of classifying the derogated group apart from the use of this term. 21 ) In this way, pejoratives seem to reflect a clear distinction between fact and value, and the question is why think thick concepts are any different. 22 We can, however, quickly move away from such examples, to others that are not so easily accommodated, concepts like JERK and CAD, where there is no clear sense of what counts as falling under the concept without appealing to the interests that a group has in using them. Consider the concept CAD as an example. Such a concept was 23 one that could only be applied to men who treated women (or women of a certain class and standing) poorly: a cad would be someone, like Alec in Thomas Hardy’s Tess of the 20 See Simon Blackburn’s discussion of ‘Kraut’. (Blackburn 1984), pp. 148-9. 21 For a similar point, offered in the context of considering whether the attitudes expressed in pejoratives can be captured by Gricean conversational- or conventional-implicatures, see (Finlay 2005, p. 19). The idea that pejoratives have a special expressive function is captured in the idea of lexical choice: with pejoratives, speakers typically have a choice in the terminology they use to refer to a group, and the idea these terms function to express certain attitudes is the reason why the term is selected. Finlay takes this point to count in favor of the idea that the attitudes expressed by such terms are likely better modeled by the notion of a conventional-implicature, than a conversational one. Oddly, Finlay does not take this point to carry over to thick concepts, and thus argues that the attitudes expressed by thick concepts involve cancelable conversational implicatures, even though it seems that an action described with a thick term can be described in numerous other ways. Blackburn makes a similar argument in (Blackburn 1992). On the notions of conversational- and conventional-implicatures, see (Grice 1989). 22 For excellent discussions on the semantics of racial epithets and ethnic pejoratives, see (Hom 2008), (Williamson 2009), and (Hornsby 2001). Hom’s view is of particular interest, as he, unlike Williamson (and Blackburn), treats the derogatory nature of such terms as a function of their content, and not the pragmatics of the speech acts which involve them. Hom offers an externalist account of their semantics, arguing that their meaning, like that of natural kind terms, is not wholly determined by a speaker’s mental states. Rather, they gain their distinctive (and derogatory) content from their connection to certain racist practices and ideologies (what Hom calls ‘social institutions of racism’). 23 I am assuming that we no longer use this concept as a way of describing a certain form of inappropriate behavior. Not because we do not think that such behavior is in appropriate, but because our reasons for doing so have changed. 14 d’Urbervilles, 24 who would make inappropriate sexual advances or take advantage of a woman (either by seduction or force), causing her to lose her sense of self-worth. Of course, the behavior that would warrant the application of such a term to a man was sufficiently diverse (ranging from taking a private stroll through the park with a woman, who might perceive this as a precursor to a marriage proposal, to rape) to block the idea that there is any neutral or purely descriptive property associated with the behavior of cads. What would unify this behavior, as exemplary of some commonality, would be the interest that individuals had at that time for grouping men under the label ‘cad’, as meriting some special (negative) recognition. This sort of interest would provide the point to their having such a concept in the first place. Here we might think that the specific interest people had was for protecting the virtue (and sexual purity) of women. Women were seen as the special objects (or targets) of concern and respect, in virtue of the fact that they were women, and who had to be protected from the vicious and sexually predatory nature of men. That such an interest is at the bottom of the concept CAD helps to explain why such a concept came into being (namely, to discourage certain sorts of behaviors in men) but also how an otherwise disparate and gerrymandered class of behavior can seen as exemplifying a common feature of element. I will return to this point in the next section. What is likely to obscure this point is the fact that judgments involving the concept CAD, we might assume, were typically voiced in cases where the person making the judgment wished to express their contempt or disapproval towards a man’s treatment of some woman. However, it is important not to confuse the correlation of these things 24 (Hardy 2005 [1891]). 15 with one being the ground of the other. It does not seem correct (or wholly correct) to say that what drives the application of this concept from one context to the next is simply a function of whether the user of it thought such feelings and attitudes were appropriate. The problem with describing things in this way is that it fails to see that appropriately describing a person as a cad and finding certain feelings of contempt and disapproval to be warranted towards that person for this sort of behavior, both have a common justificatory source in the interests that group has for using the concept CAD: namely those interests in treating women (or women of a certain class or social standing) as deserving of special concern and respect (because they are women, or because they women from the relevant class or with the relevant social standing). Because of this, appealing to these feelings and attitudes as reasons for applying that term fails to capture how both the appropriateness of these feelings, as well as the description of someone as a cad, are both grounded in further ethical considerations. And it is by attending to these further considerations that one is able to apply the concept to novel cases. If we keep this sort of distinction in mind, then we might say that the point McDowell and Williams are stressing is that, although thick concepts undoubtedly classify acts and individuals, certain ethical interests influence this classification by providing a point to classifying in this way. And so, if one understands the concept only if they are competent in using it and one is competent in using it only if they know how to apply the concept to novel cases, it follows that an outsider cannot understand the concept without attending to the point in making such classifications with it, because it is only through an appreciation of 16 why such classifications are made that one is able to apply that concept to novel cases. Consider this McDowell’s Worry. Elijah Milgram and Simon Blackburn, 25 however, have argued that McDowell’s and Williams’ position here is really just a hypothesis without an argument. Both are concerned with why we should see the fact that an outsider might be unable to master the extension of a thick concept as a telling fact about why such concepts should be seen as bridging a distinction between fact and value. Both Milgram and Blackburn fasten on the fact that McDowell’s argument for his position rests heavily on an appeal to Wittgenstein’s rule-following considerations, and note that such considerations are entirely general and apply to all concepts. 26 As such, both suggest that McDowell’s appeal to them would seem to overshoot its intended target: instead of revealing something special about a certain group of ethical concepts, they reveal that all concepts are (in some sense) thick, as all concepts would involve a blending of factual and evaluative considerations. The general thought behind the rule-following considerations, as I understand them, is that grasp of a concept (or understanding a word) cannot be coherently represented as following a rule which determines independently and in advance of the practice of using that concept and what counts as a correct or incorrect application of it. The provoking metaphor is that understanding a term (concept, rule) is something like a gear engaging a preexisting set of rails. 27 The gear represents some mental grasp (both 25 See (Milgram 1995) and (Blackburn 2004). 26 This is explicit in both (McDowell 1998c) and (McDowell 1998f). See also (Wittgenstein 2001 [1951]). 17 literally and figuratively) and the rails are the rules for applying a concept. The trouble with this picture is that one’s understanding of a word or concept (and hence, with the rule that governs its application) involves only a finite number of applications of it, whereas rules can be interpreted in indefinitely many ways. This leaves open the possibility that past cases (or any supposed similarity between them) need not be seen as determining a single way forward in interpreting the rule (or applying the concept). If this is right, then, the picture of ‘rules as rails’ must be confused. What this picture implicitly assumes is that concepts discriminate between objects independently of the interests and practices we having in using them. But again, this is a general conclusion, and is one that does not involve any special appeal to thick concepts. Appealing to Wittgenstein’s rule-following considerations, then, seems to make the very notion of a thick concept superfluous and trivially true, as all concepts would be thick, in the sense of being related to various interests and purposes. Of course, Milgram and Blackburn are right in noting that an appeal to the rule- following considerations fails to say anything distinctive about thick ethical concepts in particular. What is unclear is whether such an appeal needs to: thick concepts might simply be the clearest examples in metaethics and moral philosophy of the sort of conclusion the rule-following considerations are advancing. Reminding us that there are such concepts is one way of showing how attention to the rule-following considerations in metaethics and moral philosophy pays dividends. In overlooking this possibility, however, the objection raised by Milgram and Blackburn starts to seem rather uncharitable. What might be distinctive about thick ethical concepts is how they 27 Cf. (Wittgenstein 2001 [1951], § 218) 18 overlook or undermine a distinction between fact and value. Even if all concepts involve underlying interests and purposes that an appeal to which is necessary for being able to apply them to novel cases, what might be distinctive about thick ethical concepts is that they involve an appeal to unique sorts of interests. Indeed, it would seem that matters are worse, not better, for the defender of the fact-value distinction if all concepts turn out to be thick, for in such a case it would turn out that the CV-theorist was not even partially right in making this distinction in the first place. In one sense, however, we might see Milgram and Blackburn’s objection as stemming from the suspicion that McDowell’s and Williams’ argument is under- explained. Admittedly, the argument is complex, and rests on the controversial views of a controversial philosopher. What is clear is that neither McDowell nor Williams thinks that an outsider to the practice of some thick concept could come to master the extension of that concept without grasping (or appreciating the relevance of) the evaluative perspective of which that concept is a part. 28 However, McDowell never gives us an example of a thick concept to work from, or explain why we should believe this, other than appealing in general to considerations about what is involved in following a rule. On this interpretation, the objection is reconstructed as a request for additional explanation. If we can work out such an explanation, even if it is (admittedly) schematic, 28 Milgram is perhaps clearest about this point, in relation to his discussion of Williams’ views on thick concepts in (Williams 1985, pp. 140-41). Milgram notes that Williams argument is “insufficiently explained” (Milgram 1995, p. 5). Milgram does not seem to connect Williams’ discussion with McDowell’s treatment of thick concepts, though I think the two are clearly linked. Williams provides no independent argument for the existence of thick concepts, but instead relies upon the argument that McDowell has made which he thinks is effective. 19 then we might be able to satisfy this worry, while at the same time substantiating the conclusions of thought experiment that both McDowell and Williams appeal to. Section 1.2: Family Resemblances and a Thick Concept’s ‘Point’ The place to start, I think, is not with the rule-following considerations themselves, but with another Wittgensteinian concept, namely that of family resemblance. 29 Appealing to this notion helps to put McDowell’s and Williams’ point in a clearer context. And here I want to start with a non-moral concept, TABLE, to ward off certain objections that might be raised about the tenability of concepts where the items that fall under the concept exhibit only a family resemblance to one another. 30 The comparison between the concept of a table and thick concepts is not intended to reveal that TABLE is a thick ethical concept, but to show how the objects picked out by such a concept do not exhibit a cohesive pattern, apart from the inclusion of certain interests that we have. In this way, the comparison helps motivate a ‘partners in crime’ argument. No one will (I assume) deny that we can be correct or incorrect in judging that certain things are tables, or that we can know that a certain object is a table. If that is right, then (I hope) it will be equally uncontroversial to say these things about thick concepts. The point of appealing to the notion of a family resemblance concept is to understand, in greater detail, how thick concepts achieve a certain identity or exemplify a certain pattern in their application. The idea of a family resemblance (or family resemblance concept) is to help make sense of how various objects can be grouped under a concept even though they do not 29 (Wittgenstein 2001 [1951], § 67) 30 Cf. (Kovesi 2004 [1967], pp. 6-7) 20 share any observable unifying feature between them. Objects A, B, C, and D, might all fall under the concept X, even though A and B only share features f 1 and f 2 , B and C only share features f 2 and f 3 , and C and D only share features f 3 and f 4 . Wittgenstein used the concept ‘game’ as an example of this. Some games involve a ball, others do not; some involve two or more players, others involve only one; some games are played using a television, others are not, and so on. There is simply no observable characteristic which unifies the various instances of things that we label as a ‘game’. That they lack such a characteristic, however, does not mean that they lack any unifying characteristic. It is rather that the unifying characteristic is not to be found in some feature that we appeal to in discriminating games from non-games, but in the reasons that we have in so discriminating, say to identity some activities as a cure for boredom, or as a way for two or more people to jointly entertain themselves, or as a way to compete with others or oneself. Tables are like games. They come in all sorts of shapes and sizes, materials and measurements: some are made out of wood, others steel, others plastic or fiberglass, and yet others out of a combination of these elements; some have four legs, some have more, some have less; some have circular tops, others are square or elliptical; some sit very low to the ground, others quite high; and so on. Although every table is constituted out of a certain combination of elements, for instance, having a certain surface shape, a certain number of legs, a certain height, and so on, there is no single combination that makes an object a table. Nor is there some additional observable property (call it tableness) that makes something a table over and above these elements being in the right combination 21 for being a table. The property of being a table we might say supervenes on an object’s other properties (or configurations of them) such that: (i) no two objects could be identical with respect to these other properties, and yet one count as a table and the other does not, and (ii) any change with respect to an object having the property of tableness entails that there has been some change with respect to its others properties. Since there is no unifying observable element that is shared among different objects that makes them tables, we might question what constitutes the identity of the concept TABLE. Since there is no unique property shared by all tables, how does one ‘go on’ to apply this concept to a novel object? Of course, the answer is somewhat intuitive: if the new object serves the needs that we have for tables, say that it sits at a certain height so that a person can comfortably eat, that it is stable so that food can be placed on it and not fall on the floor, and so on, then this new object could function as a table. What makes a combination of elements into a table is that such a combination serves certain needs and interests that we have for having objects of this sort. This is what determines the extension of the concept, and it is by attending to such interests that we are able to determine whether a new object, one that we have never come across before, could count as a table. What unifies the various instances of tables as things falling under one and the same concept is the fact that we have certain interests. Change the underlying interest, and you change the identity of the concept. 31 If some object comes to be constructed out of some as yet unknown polymer, that object could count as a table only if it served the 31 See here (Wittgenstein 2007 [1967], § 388): “From here life would run differently.—What interests us would not interest them. Here different concepts would no longer be unimaginable. In fact, this is the only way in which essentially different concepts are imaginable.” 22 interests we have in tables. If, however, a different group, with different interests, came to discriminate between objects that we call ‘tables’ but do so for different reasons, it would follow from this that they do not share our concept of a table (although our concept might easily be intelligible to them). They might flip such objects over and use their legs to hang objects on or, as Brian Morrison notes, they might use them to slide down hills as a form of entertainment. 32 Suppose that some group calls these objects ‘kjetils’, but that they flip these objects over, and use them as a means for carrying objects across frozen lakes and rivers, by setting the object on the flat surface and use its legs to push it along. Because they put such objects to different uses, it would be wrong to think that their concept of a kjetil and our concept of a table are the same. How one would ‘go on’ and apply these respective concepts to new cases would be driven an appeal to different interests and concerns. The same sorts of considerations that apply to the concept TABLE apply to thick ethical concepts. Consider the concept of murder. Murders, like games and tables, come in all sorts of shapes and sizes (metaphorically speaking): some involve objects that were specifically designed to kill people, some involve objects that were not, some do not even involve an object at all; some involve the production of holes in the body from which blood is released, some do not; some are done by lone individuals, others by a group; some are done for profit, others for personal reasons; some involve prolonged planning, others are done with no planning at all; and so on. As with the concept TABLE, the identity of the concept MURDER is not to be found in some observable feature that is shared by all instances of murder, but in the interests that we have in discriminating 32 (Morrison 2003, pp. 299-300) 23 between different sorts of killings. Some killings are intentional and deliberate, some unintentional. Of course, not all killings that are intentional and deliberate are murders: some are done in wartime; some in self-defense; some are to prevent a person from suffering unduly from a disease; some are done by the state; and so on. The interest we have in discriminating some intentional and deliberate killings from others stems from an interest we have in preventing or discouraging those sorts of killings that are morally unjustified, where the victim not only does not want to die, but is innocent and undeserving of being killed. It is with an eye to these sorts of interests (and no doubt others) that we are able to apply the concept MURDER to novel cases. And just as an outsider would be unable to grasp the concept TABLE, without seeing how that concept relates to certain interests we have in eating in certain ways, so too would an outsider be unable to grasp the concept MURDER without seeing how it relates to certain interests we have in protecting innocent individuals from an untimely death. This is precisely the point that McDowell and Williams are aiming at, and I think, assume in arguing as they do about thick concepts. There will be a temptation at this point to see something in the analogy between tables and murders that goes beyond what has actually been provided. In noting that a group might use the same configuration of elements differently, say as a table or as a kjetil, one might naturally think that there is some determinate underlying type of object that differentially serves both sets of interests. If we take this thought and extend it to the case of murder, one might naturally think that there is some determinate action-type that underlies what we consider to be a murder. This thought then could be deployed to argue 24 that we can indeed separate out the factual from the evaluative (or ethical) in this concept. The factual element is simply that sort of action which constitutes all and only cases of murder, call it murder*, 33 and the evaluative component is simply our disapproval of these sorts of acts. The trouble with this move is that it misses the point of the discussion just given. There is no type of action that we simply attach the label ‘murder’ to that is independent of the interests we have in distinguishing murders from other sorts of intentional and deliberate killings, just as there is no object that we simply attach the label ‘table’ to, or that others attach the label ‘kjetil’, too when considered apart from certain interests we (or they) have in classifying. The point of appealing to vast differences in the elements between the instances that fall under these concepts was to make this point clear. And so what makes this starring move “awfully fishy,” 34 is that it assumes that how we classify actions or objects is independent of the interests we have in doing so. Such a move gets off the ground, however, only by denying the sort of picture of concept formation that has just been offered, and, in order for it to be a credible objection, it would have to offer something in its place. Now the interests that underlie a thick ethical concept differ from those that underlie concepts like GAME and TABLE. They differ in that the interests involved are, on the one hand, those involved in providing a certain sort of social existence, one where individuals are able to live (relatively) peacefully and cooperatively with one another, and, on the other, are bound up with feelings such as a sense of respect, esteem, and 33 Milgram calls this move “starring the concept.” (Milgram 1995, p. 3) 34 (Milgram 1995, p. 4) Milgram admits that there is something suspicious about this move, but notes that it is up to the defender of thick concepts to explain what it is. 25 admiration, or conversely, feelings of contempt, disdain, and shame. In understanding some thick concept, one must grasp its role in sustaining a shared form of social existence, one that is bound up intimately with and partially constitutes a conception of what counts as a good person and a good life. In this way, although the identity of all concepts is constituted by certain interests that a group has, the interest involved in the case of thick ethical concepts are what we might summarily call ethical interests. Admittedly, this is a gloss of what is likely a complex set of notions, but this distinction is enough to show how thick ethical concepts can distinctively reflect a failure of the fact- value distinction (as it is sometimes found in metaethics and moral philosophy), even though the central arguments for them have a much wider appeal. Section 1.3: Sophisticated-CV Approaches One way of attempting to accommodate the sorts of worries that McDowell and Williams raise is to abandon the idea of thick concepts involving nonevaluative sufficient conditions. In their place, a CV theorist might argue that there still might be nonevaluative conditions that must be true in order for the concept to apply, even if such a condition never determines the concept’s application. Thus, thick concepts might still be seen as involving nonevaluative necessary conditions. In the next two sections we will cover two sophisticated-CV accounts that attempt this move. (The second of these approaches is much more interesting and deserves separate treatment in the next section.) In both cases there is an attempt to remain faithful to the general CV-intuition that thick concepts involve isolable and independent evaluative and nonevaluative components. What is abandoned here, however, is the thought that the distinction between components 26 be understood in the way that a basic-CV theorist understands it, and as such we might group together this family of views by calling them sophisticated-CV approaches. The guiding idea behind a sophisticated-CV is to see the criteria for a thick concept as involving distinct evaluative and nonevaluative components. Evaluative considerations, then, are admitted to have an influence on the extension of a thick concept, but this influence is isolable from that provided by the concept’s nonevaluative criteria. One approach to this has been advanced by Stephen Burton 35 but it has since been adopted by others. 36 On Burton’s view, thick concepts are seen as primarily evaluative concepts that have “descriptive qualifications.” 37 He distinguishes his account from the approach criticized by McDowell and Williams as follows: …instead of analyzing (positively valenced) thick concept ‘C’ as ‘X,Y,Z, etc.’…and therefore (pro tanto) good’ one might define it as ‘(pro tanto) good…in virtue of some particular instance of X,Y,Z, etc.,’ where again ‘good’ is purely prescriptive and ‘X,Y,Z, etc.’ are purely descriptive. (Burton 1992, p. 31) The proviso ‘some particular instance’ marks the important contrast here, for only some instances of ‘X,Y,Z’ will actually help explain why a thick concept applies. While it is necessary that an act be X,Y,Z in order for C to apply to it, this is not sufficient. There are two problems with this approach. The first is that it is not clear how the evaluative and nonevaluative criteria are to be distributed in our understanding of the concept. And the second is that such an account is trivially true. Grant that ‘X,Y,Z’ 35 (Burton 1992) 36 Cf. (Zangwill 1995) and (Tappolet 2004). 37 (Burton 1992, p. 30) 27 picks out the ‘facing of a dangerous or difficult situation’ 38 and that (C 1 ) is a necessary nonevaluative condition for an act to be courageous: 39 (C 1 ) It is an act that involves facing a dangerous or difficult situation. Of course, satisfying (C 1 ) does not make an act courageous. The trouble is that when we add the plausible evaluative necessary condition, (C 2 ), it is not clear what work (C 1 ) is doing: (C 2 ) It is an act that involves facing a dangerous or difficult situation that is worth facing. What Burton needs is for the explanatory contribution of (C 1 ) to be independent of the other evaluative conditions that might be involved. But, (C 1 ) is entailed by (C 2 ), so it is unclear why we should accept this. Is (C 1 ) necessary for an act to be courageous? Yes. Does it help to explain why certain acts are courageous? I am not so sure. Given (C 1 ) alone, we can certainly rule out certain acts as not courageous, but it is not clear, in telling us that certain acts are not courageous, how it explains why an act is courageous. Because of this it is unclear how we are to understand it as an independent constraint on the application of a thick concept. The worry here is that there is a rather obvious (but innocuous) sense in which thick concepts can be taken as having nonevaluative necessary conditions. Thick concepts, especially those that relate to virtues and vices, tend to have general domains of application—some area of human life and experience that provides a background for the 38 Burton here suggests “sticking to one’s guns despite great personal risk” (Burton 1992, p. 32). 39 I think that (C 1 ) is not a necessary condition for an act to be courageous, because there can be acts that are done in spite of intense, but irrational fears, that I think ought to count as courageous, but I will leave this point aside. 28 concept’s use and application. Courage, for instance, relates to those areas that involve danger, difficulty, or which otherwise cause people to be fearful. And, because of this, it only makes sense to talk about acts as being courageous or not only when they occur against such a background. But the fact that this concept only applies to acts that involve situations that are dangerous, difficult, or fear inspiring, does not yet tell us anything about the content of COURAGE, for our understanding of this concept is one that presupposes the agent handled himself well despite such difficulties. And this is contrasted with other concepts (i.e., rashness or foolhardiness) that reflect how an agent can handle himself poorly in such a domain. Jonathan Dancy brings out this point by noting: Locating a thick concept by specifying the domain of its operation is different from giving even a vague specification of its content. Telling us where to look for a concept is different from telling us what the concept is a concept of. To say that the concept of courage concerns fear and danger is not to say anything about its content. Being about fear and danger is not what an action has to be like to count as courageous. We are, of course, characterizing the concept in saying that it concerns fear and danger, but we are not doing so in a way that offers to generate even part of a descriptive meaning for the predicate. (Dancy 1995, p. 277) The idea, if I understand Dancy’s point here, is that specifying a general area of human life and experience, as something that constitutes a background against which a particular thick concept operates, is not enough to tell us anything about the nature of that concept. And so, if it counts as a necessary condition that such a background be present in order for the concept to apply, that condition is a trivial one, and fails to shed any light on how we should understand a particular thick concept. There is a more general worry here, one related to the concerns laid out in the previous section, in explaining why this feature is necessary. But, before addressing it, I would like to consider a second sophisticated-CV. 29 Section 1.4: Sophisticated-CVs and Thick Psychological Descriptions This second approach differs from the first in what it takes to be the content of the relevant nonevaluative component. A thick concept, on this view, involves a specific psychological description that must be true of an agent in order for that concept to correctly apply to what the agent has done. This approach has recently been advanced by Andrew Payne. 40 Roughly, if an agent acts in manner falling under a thick concept, then there is a true psychological description of the agent, at the time of acting, which helps explain why this is so. Again, as this is only a necessary condition, it can be true of agents when the concept does not apply. A complete explanation as to why a thick concept applies in a given case will depend on additional evaluative considerations. As with the previous sophisticated-CV, the relevant psychological description is intended to function as a constraint on the thick concept’s extension apart from an appeal to these other evaluative considerations. Payne borrows Gilbert Ryle’s notion of a “thick description” 41 to help explain his view. Ryle held that the correct characterization of some piece of overt behavior can vary according to the underlying beliefs, desires, and intentions an agent has in acting. Describing an eyelid contraction as a wink, Ryle noted, differs from describing it as a blink, but this difference is not captured by a characterization of the agent’s overt behavior (as each are indistinguishable from this standpoint). Rather, the difference between a wink and a blink is that, in the former case, the agent contracted his eyelid with, among other things, the intention to communicate something by a publically 40 (Payne 2005). 41 (Payne 2005, p. 94). See also (Ryle 1971). 30 recognizable code. The same point, Payne argues, holds for thick concepts. To be courageous, for instance, an act must “involve an awareness on the part of the agent of the danger or difficulty connected with the action, and the agent must intend to perform the action despite the danger in order to promote some end that is seen [by the agent] as valuable.” 42 As such, thick concepts apply to acts in part because a distinctive thick description is true of the agent who acts: “every instance of a thick concept will satisfy some thick description which mentions the beliefs, desires, and intentions necessary for falling under that thick concept.” 43 Similar to Burton, Payne holds that these thick descriptions are independent of whatever other evaluative considerations might be thought to help explain the application of a thick concept; they “[stand] alone without any accompanying moral evaluation, or with the normal evaluation reversed.” 44 The governing idea behind this seems to be that we can treat the relevant thick description as neutral fact about the agent because we can parse it away from whatever evaluations we depend upon in ascribing a thick concept to an agent’s act. For instance, that a person must act for the sake of an end they believe to be valuable, for their act to be courageous, is treated as a neutral fact about the agent, and distinct from the question of whether that end really was valuable. I think there is a good deal to admire in Payne’s account. Payne is surely right to claim that thick ethical concepts involve some connection to robust psychological ascriptions. Where I disagree with Payne is over his claim that, in applying a thick 42 (Payne 2005, p. 94) 43 (Payne 2005, p. 95) 44 (Payne 2005, p. 100) 31 concept to what some agent has done, the thick description is independent of ‘any moral evaluation’ of the agent’s act. The connection between the two is more difficult to separate than Payne realizes. Determining whether the relevant thick description applies to an agent, in applying a thick concept to that agent’s act, is not, I will argue, wholly separate from the evaluations one would make of the agent’s overt act, and hence, it is not clearly independent of the various evaluative considerations that are related to the application of a thick concept. There are two reasons for thinking this. First, when it comes to determining an agent’s reasons for acting, which, I assume, a thick description at least partly aims at capturing, this is typically something we arrive at in parallel to our describing an agent’s act in some way. It is not a starting point, or even prior to, our coming to describe what the agent has done. If this is right, our view of an agent’s psychology in acting (i.e., what sort of thick description correctly applies to the agent) is not fundamentally distinct from our coming to a view of how they in fact acted. Beliefs, desires, intentions, and other elements of the agent’s motivational structure, are not discernible apart from interpreting the agent’s overt behavior and the general context in which that behavior is situated. Certainly, in many cases, our ascription of some motivational structure to the agent turns out to be a rather straightforward process. But in such cases we tend to have some knowledge of a rich contextual background (e.g., some understanding of the participants involved, of the history leading up to the act, and so on). Take the following example: Jane, Jim, and Sally have all been colleagues together for some time. Jane knows that Jim wanted vary badly to get some special recognition from their department, and that, when this 32 recognition went to Sally, Jim made some deeply mean-spirited remarks to Sally that Jane judged were clearly cruel. To make this judgment, Jane must of course hold that Jim’s actions can be explained in a rather unique way. And such an explanation will depend greatly on Jane having knowledge of the sort of contextual background I just noted. But, even here, ascribing some thick description to Jim depends on Jane making certain evaluations: she would have to relate what Jim had said against the different possibilities she thinks were reasonably available to him in that context, but which were passed over. Such possibilities would include not only different types of actions, but different ways of carrying out one and the same act (e.g. offering a congratulatory remark). In order for Jane to conclude that what Jim had said was cruel, she would have to develop an understanding of how he might have behaved differently, and from this reverse-engineer the conclusion that Jim said what he did in order to gain some satisfaction from spoiling Sally’s achievement. The conclusion Jane would have to reach is that nothing else would explain why he acted in such a hurtful and cruel manner. If this is right, then reaching the conclusion that Jim had acted for this reason simply is reaching the conclusion that what Jim had said was cruel (she might not voice this opinion, but that is another matter). As such, ascribing the relevant thick description to an agent, in using a thick concept, is crucially linked to evaluating that agent’s conduct. There will be a temptation here to see this as merely an epistemic impediment to our getting at the truth of what some agent believed, desired, or intended, in acting as they did. This temptation motivates the thought that, even if we (as evaluators) do not clearly know what reasons the agent acted from, the agent surely will. To think that our 33 understanding of the agent’s psychology in acting functions as an independent constraint on the application of a thick concept, then, might be grounded in the idea that the agent’s own view as to why he acted holds some special authoritative position in how that agent’s behavior is explained. And this would provide some stock to thinking that the correct explanation of it is independent of how we evaluate that agent’s behavior. This leads to my second concern: namely, that we can rightfully hold that a thick concept applies to what an agent has done irrespective of what the agent took to be his reasons for acting. Intentional actions can arise from motivational structures, even though they and the decisions that arise from them are more or less unavailable to the agent’s conscious reflection. In such cases, the agent’s own interpretation of his behavior comes from no special vantage point in comparison to that of others. Important for our purposes, however, is the fact that it is always possible for an agent to think he is acting for reasons that are actually ill-suited to explaining how he behaved. The upshot is that there need not be any distinctive psychological facts that are wholly independent of the manner in which the agent’s overt behavior is explained and evaluated by others. Consider an agent who reveals to an acquaintance some piece of embarrassing personal information about a friend (say that the friend’s spouse was cheating on her) and grant that such information was deeply humiliating for this person. When asked, however, the agent says (sincerely) that he revealed it for the friend’s own good, to help her out, so that she could face it and get past it, and so on. In such a case, one could still rightfully claim that the agent’s act was cruel, even if the agent sincerely rationalizes his behavior to the contrary. Judging 34 this act cruel will certainly involve ascribing certain reasons to the agent for acting, but our doing so need not (and typically does not) afford a special place to what the agent takes his reasons to be, especially if those reasons conflict with the sort of evaluation that we think their behavior merits. In such cases, the evaluations we make and the reasons we ascribe to explain an agent’s behavior are too sensitive to one another to be treated as completely independent. We would like to think that, in this case, the agent’s rationalizations could be undone if (say) the information turned out to be so humiliating that, once brought to light, there was no getting past how obviously hurtful the revealing of it would be. But, individuals can be resilient in their opposition: there is not only the possibility of subconscious reasons, but denial about them as well. The general point is that we could rightfully hold that the agent acted cruelly in this case, regardless of how he might explain his own actions, and ascribe to him the relevant thick description that would go along with this, because how we evaluate his behavior fits with no other explanation. Since our ascribing a thick description to an agent, in applying a thick concept, is intimately connected to how we actually evaluate an agent’s behavior, I think Payne’s view fails to be compelling. We might put this point in terms that are congenial to how McDowell originally objected to early basic-CVs: understanding which thick descriptions correctly apply to an agent, in using a thick concept, is just as dependent on grasping a certain ethical perspective, and can be just as incomprehensible apart from it, as understanding how to apply the concept itself, because we have no independent access to the agent’s psychological state apart from interpreting it in conjunction with interpreting 35 the agent’s behavior. Because of this, I do not think, Payne has succeeded in showing how thick concepts involve (purely) descriptive components. Despite my reservations about his view, I think that there is a fundamental insight in Payne’s account that is worth preserving, namely that our ascription of a thick concept to something the agent has done depends heavily on our coming to a rather specific view about what that agent’s reasons for acting in fact were (even though, as I have argued, this is not independent of the evaluative concerns we have in assessing their behavior). In the next section, I will attempt to draw out this insight, and connect it with what seems to motivate taking a component view of thick concepts. What follows from Payne’s insight, I will argue, suggests strongly that thick concepts presuppose a constitutive relationship between a specific thick description, or something quite close to this, and certain ethical values, and the failure to appreciate this leads, in at least one case (cruelty), to a distortion in our understanding of that concept. Section 1.5: Some Modest Suggestions on the Nature of Cruelty To begin, it is worth noting that the example Payne chooses to center his discussion on (namely COURAGE) actually prevents him from making a much stronger point about the relationship between some specific thick description and the application of its related thick concept. For many thick concepts, their ascription to some act depends on the agent having acted for (what I will call) characteristic reasons. 45 This certainly seems true in the case of acts considered cruel, honest, conscientious, generous, kind, and loyal. In each of these cases, the concept can apply to what the agent has done 45 This point will be argued in greater detail in Chapter 3. 36 only if that agent has acted for certain reasons and not others. As such, what reasons the agent has for acting clearly makes a difference in how we are to describe that agent’s action. For instance, it would clearly be incorrect to describe an act as honest if the agent spoke truly only because he wanted to cause the hearer immense pain, and wrong to call an act generous if the agent gave to another simply to receive a greater benefit from that person (e.g., a promotion). Acting for these reasons is incompatible with ascribing such concepts to what these agents have done. This, I take it, is why Payne’s account reflects an important kernel of truth in how we should understand the structure of (many) thick concepts. The very notion of ascribing a thick description to an agent is itself something that aims (at least in part) to capture the agent’s distinctive reasons for acting. However, in all of this, the concept of courage is atypical, for it lacks any sort of characteristic reason: 46 while an agent needs to face a dangerous or difficult situation for the sake of a worthwhile end, in order for her act to be courageous, what those ends are is not determined by the content of courage. Because of this, any generalized description of the agent’s psychology will obviously underdetermine whether courage applies to what the agent has done (indeed, Payne himself provides a clear example of this 47 ). However, when a thick concept does have a characteristic reason associated with it, and an agent succeeds in acting for that reason, it becomes much harder to produce cases where the agent’s act does not thereby merit the corresponding evaluation. This point, I think, is overlooked in Payne’s discussion. In cases where a characteristic reason is involved, our 46 This is connected to the idea that courage is an “executive virtue.” It is exemplified in the execution of good and noble deeds but does not determine what these are. See (Williams 1985, p. 9). 47 (Payne 2005, pp. 98-99) 37 allowance that the agent has succeeded in acting for such a reason will typically be sufficient for ascribing the relevant value to what the agent has done. 48 What will appear troubling about this claim is that such reasons seem to be the very thing that a CV theorist would fix upon in claiming that thick concepts have logically independent nonevaluative components. After all, the reasons that an agent acts for seem to be nonevaluative facts about the agent. Indeed, such a view would appear bolstered by the thought that once we come to see that an agent acted for a reason characteristic of a thick concept, little else seems needed for that concept to apply. In the previous section, I provided reasons for thinking that this is perhaps an overly simple view of how we use these concepts, since, in ascribing a thick concept to an agent’s act, our determining what reasons the agent acted for is not independent of the manner in which we evaluate their behavior. But there is a further reason to be skeptical of this sort of view. Whether these characteristic reasons could provide a thick concept with some independent nonevaluative component depends a great deal on whether we can treat them as separate from a complete appreciation of the ethical value of the acts they give rise to. This sort of general worry stems from the sorts of concerns that I raised in asking about the interests that a group has in using a certain thick concept. And the problem that Payne faces is simply a particular case of a much larger concern for CV theorists: though they might be able to provide a list of nonevaluative necessary conditions that thick 48 The force of ‘typically’ needs clarification. What I have in mind is that, if the relevant evaluation is not manifested when the agent acts from some characteristic reason, this will generate a special need for explanation. In the normal run of things, the relevant evaluation simply follows from our concluding that the agent had acted for that reason. Cf. (Anscombe 1958, p. 14) Anscombe distinguishes the ‘normal’ number of teeth human beings have from the average number. What is implicit in the analogy is a special need for explanation in cases that deviate from what is ‘standard’. 38 concepts involve, they will have difficulty explaining why the features listed in those conditions are probative for applying the concept, without appealing to the interests and concerns that we have in discriminating between acts and individuals that have or fail to have those features. In this section, I shall argue that, in at least one case, the interests we have in the ethical value of certain reasons cannot be adequately distanced from the value we have in the sorts of acts those reasons are associated with. This, however, strongly favors the thought that, at least in this case, the thick concept in question does not have the sort of independent nonevaluative components that a CV theorist would suggest. For this discussion, I will focus on a particular thick concept—CRUEL—leaving it open as to how this argument might be generalized (though I think it clearly can). Grant that (CR) is a description of the characteristic reason associated with acts of cruelty: 49 (CR) The agent acts in order to delight in the harm or suffering caused in another. On the face of it (CR) seems to be a purely descriptive characterization of what drives our application of the concept cruel to various acts. 50 But I think that this conclusion is too quick, and to see this we need to appreciate more fully the fact that cruel is not a concept of a certain sort of reason, but the concept of an act done for a certain sort of reason. Although acts of cruelty are no doubt ones where an agent acts in accord with (CR), our 49 Here I am taking cruelty to involve an intentional causing of suffering. Some think that cruelty can simply involve an indifference to another’s suffering. This I think involves a different (but closely related) thick concept—namely, callousness—but I will not argue for this here. 50 Cf. (Skorupski 2007b, p. 265). Skorupski holds a view similar to this, and claims that the term ‘cruel’ is purely descriptive, because cruel acts are simply those that arise from a motivation that involves “a non- instrumental desire to inflict suffering.” Skorupski’s account, I think, is flawed because acts can involve such a non-instrumental desire and still fail to be cruel (e.g., when the agent gets it wrong as to what sort of act will satisfy such a desire). 39 understanding of what the agent has done and the fact that he has acted for such a reason are bound tightly to one another—so tight that one cannot be given a certain evaluation unless the other is. To think otherwise, I will claim, is to overlook an important aspect of what makes an act cruel. We are apt to overlook this connection because an appeal to characteristic reasons strongly suggests that the proper place to focus our attention is exclusively with the agent’s perspective, such that whatever it is that makes cruel acts worthy of a negative evaluation will rest solely on considerations drawn from that perspective. There is certainly an intuition here that is worth preserving, and I will come back later to what I think that is, but such an exclusive focus skews our understanding of cruelty in an important way. Acts of cruelty involve not only an agent, but a victim or sufferer. What the agent does, by acting in accord with (CR), is to cause someone to suffer (either physically or emotionally). And so, if we think our understanding of the nature of a cruel act is determined solely by appeal to the agent’s reasons, apart from the character of the acts those reasons give rise to, then we are liable to misrepresent such acts (or our understanding of them in using the concept CRUEL) by overlooking the fact that they also involve a victim’s state of suffering (what I will generally refer to as ‘the victim’s perspective’). Given this, we should conclude that an act of cruelty actually involves three basic elements: the agent’s reason for acting, the action, and the victim’s state of suffering. 40 The importance of the victim’s perspective here relates firstly to that person’s suffering (or, more precisely, to the fact that the victim has been caused to suffer). 51 Suffering, I take it, is an issue of basic significance for any evaluator. It is something that we feel a morally sensitive evaluator cannot omit from her account of a situation when it is present, even if its presence does not contribute to an evaluation of that situation. In the case we are considering, however, the importance of the victim’s state of suffering comes out in the fact that, with acts of cruelty, the harm or suffering caused is neither trivial nor minor. Causing a small embarrassment, a slight annoyance, or a trifling indignity, even if the agent derives great satisfaction from doing so, is not enough for that agent’s act to be cruel. Such acts might be considered nasty, mean, rude, or just plain inconsiderate, but they are not cruel. In order for an act to be cruel, the suffering caused must be severe, and bring serious harm or emotional trauma to the victim. And further, this must be something the agent seeks to accomplish in acting as he does. (We might explicitly amend (CR) to account for the seriousness or severity of the suffering caused, but I will take it as implied.) Acts of cruelty, then, are ones where an agent acts in accord with (CR), but in doing so causes his victim severe harm or suffering. Now, at this point, a CV-theorist is likely to claim that this actually supports his view of thick concepts: the agent’s perspective captures something purely descriptive that must be true in order for an act to be cruel, while the sufferer’s perspective captures what is relevant as an 51 Hilary Putnam argues that the victim need not actually suffer in order for an act to be cruel. See (Putnam 2002, p. 38). Putnam’s example is of an innocent person who is seduced in such a way so as to prevent her from realizing her talents. As a result, the seducer does something cruel to this person, even though she suffers in no clear way. I think that if the seducer has acted cruelly in such a case, this would still require the innocent person to be harmed in a significant way, such that, if this person had a clear view of what the seducer was doing this would (at least) cause them some severe emotional trauma. 41 evaluative aspect. These two perspectives, it will be thought, capture independent, and only contingently connected, conditions in our concept of cruelty. To think this, however, is to hold these two perspectives, and the two related aspects of a cruel act, at an artificial distance from one another. This is because the agent’s reason for acting makes the victim’s suffering not merely severe but excessive: the victim does not deserve the harm or suffering she is caused, or at least that much of it, and there is no adequate justification that the agent can appeal to in causing it. The excessiveness of the suffering caused, then, depends on the agent having acted in accord with (CR), or, at the very least, it is not wholly independent this. Certainly, had the agent’s reasons for acting been different, our view of whether his act was cruel would have been different, but this is because our view (and likely the victim’s view) of the suffering caused would have been different. This, I take it, is a full picture of what an act of cruelty amounts to: it involves considering not simply the agent’s reason for acting, but also how such reasons inform our understanding of what that agent has in fact done (i.e., that he has caused the victim to suffer excessively). These two considerations cannot be pulled apart in considering an act cruel. It is also important to note how this connects with the considerations raised above, regarding the sorts of interests we have in developing a concept like CRUEL. Of course, there are likely to be many interests involved in why we came to discriminate cruel acts from others, but certainly one interest that would likely be paramount in such a story would be the interests that we have discouraging people from causing others to suffer excessively simply for entertainment or enjoyment. The concept of cruelty, then, is one 42 that attempts to discriminate between acts along these lines. In this way, the point behind our having a concept like cruelty does not distinguish between the sort of reasons that it is concerned to discriminate actions on the basis of, and the sort of value that it attaches to those actions. Given this account, we can now see why a sophisticated-CV treatment of cruelty, inspired by Payne’s account, though perhaps prima facie compelling, is ultimately incorrect. While such a view rightfully looks to the agent’s reason for acting as the descriptive feature that determines whether an act is cruel or not, it fails to appreciate how this feature also determines the evaluative character of the victim’s suffering. And so, when Payne argues that any evaluation can be attached to acting for such a reason, this claim retains its plausibility only by disregarding how acting for such a reason makes the victim’s suffering excessive, and how it is this fact about the victim’s suffering that is at the root of our seeking to discriminate certain actions using the concept CRUEL. Once we move to consider the victim’s suffering in relation to the agent’s reason for causing it, such a reason fails to justify the causing of it. In this way, such a sophisticated-CV of cruelty provides only a partial account of what is involved in a cruel act, and to this extent it distorts our understanding of the concept. Why might a sophisticated-CV theorist think that the reasons characteristic of cruelty can sometimes be evaluated positively? The answer to this rests, I think, on a confusion over how even acts of cruelty (or their unintended consequences) can be viewed in a positive light. That they can, however, is not because the descriptive properties which make them cruel are sometimes reasons for a positive evaluation of 43 those acts. Acts can have a wide variety of properties associated with them, and so some acts of cruelty might also bring about unintended benefits. And while such benefits might give us a reason not to condemn or censure an agent for his having done something cruel, they are irrelevant to an evaluation of such an act for its cruelty. Consider the drill instructor who overlooks the fact that two soldiers are behaving quite cruelly to a weaker recruit, because he thinks their cruel behavior will help ‘toughen up’ the recruit to face the horrors of battle. 52 In such a case, it is not that cruelty is a positive quality of the soldiers’ actions, rather it is that the unintended benefit of their treatment of the recruit outweighs (if it indeed does) the disvalue of its being cruel. Cruelty, then, is always assessed negatively even though the negative evaluation of an act as cruel can be outweighed by evaluations carried by an act’s other properties. Section 1.6: Thick Concepts and Quasi-Realism The discussion of various CV approaches to thick concepts has to this point been focused on various sophisticated-CV approaches. And of the two examples listed, they are perhaps best interpreted under a cognitivist light, for they allow us the clearest sense as to how thick concepts can admit of correct instances while also allowing that their extensions are influenced by separate evaluative and nonevaluative considerations. In general contrast to this, there is an important and highly influential picture of thick concepts that must be considered, which arises from the noncognitivist tradition in 52 Cf. (Hare 1981, p. 71) from which my example is based. 44 metaethics. 53 This picture comes from Simon Blackburn’s quasirealism, and the idea that judgments involving thick concepts are simply “loaded descriptions.” 54 In general, Blackburn’s quasirealism is a meta-theoretical approach to handling various domains of discourse that seem, for one reason or another, metaphysically troubling—discourses such as mathematics, modality, aesthetics, and (in our case) ethics. In each of these, the discourse seems to require the positing of facts and properties that seem difficult to place (or locate) within a naturalistic worldview. The quasirealist’s methodological approach, then, starts from a theoretical vantage point which does not involve an appeal to such suspect facts and properties (e.g., mind-, attitude-, or practice- independent mathematical, modal, aesthetic, or ethical facts), and then works to develop an account of the discourse where the use of terms like ‘fact’ and ‘property’ can be legitimately used without lapsing into the sort of metaphysical extravagance that was taken to be originally problematic. In the case of ethics, quasirealism takes the form of what Blackburn has labeled “projectivism.” 55 Blackburn gives us the following characterization of projectivism’s basic view: “[v]alues are the children of our sentiments in the sense that the full explanation of what we do when we moralize cites only the natural properties of things and natural reactions to them.” 56 53 Early forbearers of this tradition include R.M. Hare’s prescriptivism in (Hare 1952) and (Hare 1963), as well as the emotivism of C.L. Stevenson (Stevenson 1937) and A.J. Ayer (Ayer 1946). 54 (Blackburn 2004) 55 (Blackburn 1984, p. 180). Blackburn now considers ‘expressivism’, ‘practical functionalism’, or ‘non- descriptive functionalism’ to be a better label for his position. See (Blackburn 1998, p. 77). 56 (Blackburn 1984, p. 219n21). See also (Blackburn 1991, p. 3). 45 Blackburn’s approach to understanding thick concepts is complicated and demanding. Blackburn argues against the view that thick concepts involve an inextricable blending or connection between evaluative and nonevaluative considerations, and, in contrast, holds that the connection between these elements is merely a contingent one. 57 He argues for this by claiming that thick concept advocates fail to adequately appreciate how any descriptive utterance can also express an evaluative attitude as a pragmatic feature of that utterance: the attitude in question is something that is expressed, not in what is said, but in how the statement is made, for example, in the tone of the utterance itself. In this sense, any descriptive utterance can ‘carry’ (more precisely: conversationally implicate) an evaluative attitude—one can express surprise, dismay, frustration, and so on, by how one says what is said. 58 Given that whether a purely descriptive utterance carries such an evaluation is itself a function of the conversational context in which that utterance is made, there is no need, according to Blackburn, to think that so-called thick concepts involve any special amalgamation of evaluative and factual considerations: the pragmatic feature can be removed leaving an unadulterated descriptive statement as the vehicle which carried it. Indeed, on Blackburn’s treatment, whatever positive or negative evaluation is conversationally implicated can be ‘cancelled without remainder’, in the sense of one’s being able to simply deny the assumed evaluation. As such, whatever attitudes are associated with a thick concept, are 57 This is clearest in (Blackburn 1992). 58 One can pragmatically convey evaluations, contrary to the typical evaluative meaning that is associated with their terms, by saying things like “That’s great” or “Good job” in a sarcastic tone as well. 46 contingently associated with it, as this association is too “complex, fluid, and mobile” 59 to be thought of helping to determine what that concept applies to. As noted above, Blackburn’s account of thick concepts is intended to be an explanatory picture of what the use of such concepts involves. It is to show how a quasirealist can describe the use of thick concepts without having to appeal to the thought that these concepts reflect an inextricable blend of descriptive and evaluative elements. The example he gives to make his point is extremely simplistic, but this is grist for Blackburn’s argument: it is up to the advocate of a thick concept to say what makes his case any different. Blackburn’s considers a group of people who despise people that are overweight. Members of this group register their contempt for such people by uttering the neutral term ‘fat’ with a negative tone (captured by ‘↓’): ‘fat↓’. 60 This reflects the sort of contingency that Blackburn takes there to be between the factual and evaluative aspects in a thick concept: these two aspects bear only a contingent connection to one another (as it is supposed that others might approve or admire overweight people, thus uttering ‘fat↑’ with an approving tone). Blackburn argues that such a model can capture a feature that thick concept advocates generally appeal to, namely how the evaluation can depend on the content of what was said: “[g]enerally speaking, saying something with an attitude in your voice licenses the hearer to suppose that the attitude is expressed because of what is said.” 61 And so, when a speaker of this group calls someone ‘fat↓’ this generates a conversational 59 (Blackburn 1992, p. 296) 60 (Blackburn 1992, p. 292) 61 (Blackburn 1992, p. 289) 47 implicature 62 to the effect that the contemptuous attitude is taken on account of their being overweight. 63 Strictly speaking, the content conveyed by the utterance is that the person in question is ‘fat’, while what is implicated in the manner of their saying it is a form of contempt, with the idea that the negative implication being something that can be cancelled or disavowed by the speaker. This feature can also, according to Blackburn, account for what is at issue between two speakers when they disagree over someone being ‘fat↓’: their disagreement is over whether that person is contemptible on account of their being overweight. Blackburn gives the following as a possible argumentative exchange: Amanda and Beryl may have been card-carrying fattists until Amanda met Clive. ‘Clive is so fat↓’ challenges Beryl. ‘No, not fat↓--stocky, well-built’ dreams Amanda. The dispute here need not be one about vagueness, as we can see if we play through with Pavarotti instead of Clive. Pavarotti in unquestionably fat, but many fattists would recoil from calling him fat↓. They want to overlook the fact that he is fat to an extent that would normally repel him, since he is so transcendentally uncontemptible in other ways. (Blackburn 1992, p. 290) Blackburn notes that, although simplistic, his example can include all of the features that advocates of thick concepts take to support their position. Fattists can be seen as ‘sharing routes of interest and feeling, senses of humour and of significance and of fulfillment’, 64 62 For the notion of conversational implicature, see (Grice 1989). For an interesting view that applies the notion of conversational implicature in defense of a descriptivist account of thin ethical terms like ‘good’, see (Finlay 2004). 63 For an interesting discussion of how some of the cases Blackburn appeals to in support of his view that thick terms can be used to express any evaluation, even those contrary to their typical use, see (Vayrynen 2008). Vayrynen argues that the cases which Blackburn appeals to can be explained by treating certain thick concepts as gradable adjectives subject to contextually sensitive standards. As such, Blackburn’s example can be shown to reflect a shift from standards for determining whether something is F to standards that determine whether something is too F or not F enough, and Vayrynen suggests that Blackburn has not shown why these latter cases should tell us anything distinctive about whether something is or is not F, let alone whether Fness is compatible with any evaluative attitude whatsoever. 48 as being involved in a common conceptual practice, as having an ability to detect certain nonevaluative features, and so on. Indeed, Blackburn allows that the sensibilities tied to the use of ‘fat↓’ can become quite refined: we can imagine that our hypothetical group comes to “only call some people ‘fat↓’—those they are disgusted by. Quite fine differences elicit different verdicts.” 65 That is, although the practice of calling people ‘fat↓’ might have started with a general derision towards those that are overweight, there is nothing preventing this practice from developing rather refined standards for who is and is not rightfully called ‘fat↓’. There are several worries for Blackburn’s account. The first is that the picture he gives might simply be too simplistic to do justice to the sort of ramifying connections that McDowell, Williams and others take thick concepts to have. For instance, certainly we can judge that an act is cruel, and think that its cruelty counts in favor of a negative evaluation of it, even if we feel no contempt or disapproval or abhorrence in what has been done (consider hearing about a case of torture that occurs in some unknown location, and where you are not privy to any of the details save that the torture occurred). This is not to say that there is no connection between thick concepts and our attitudes and emotions, only that such a connection will by no means be straightforward. There is no reason to suppose, as Blackburn’s example seems to suggest, that the connection between judgments using a thick concept must reveal some general or systematic connection to the emotional attitudes of those using the concept, in order to consider that concept a 64 This is paraphrasing some of what Stanley Cavell appealed as ‘the whirl of organism’ in order to describe Wittgenstein’s notion of sharing a form of life. See (Cavell 1969, p. 52). Cf. (Blackburn 1998, p. 96) 65 (Blackburn 1998, p. 95) 49 ‘thick’ concept. I will return to this worry in the next chapter, when I consider the sentimentalist treatment of thick concepts made available by McDowell and Wiggins. A second worry is whether the account can accommodate the full range of cases that thick concept advocates are likely to be concerned with. For instance, it is not clear how this model would explain how someone is able to doubt, worry, or simply be concerned about, the question of whether a person is fat↓, without having to treat these attitudes as being directed towards purely descriptive objects and considerations. This is because, on Blackburn’s account, thick concepts have a purely descriptive content, and so there is nothing to doubt, worry, or be concerned with, as to whether someone is fat↓ beyond their being just fat (i.e., overweight). Such a view would not do justice to the pre-theoretical intuition that, for instance, in doubting whether a person has acted cruelly towards another the content of this propositional attitude is descriptive. Since in these cases there is no utterance to carry an evaluative attitude, it is not clear why his view should treat such a doubt as one directed towards an evaluative question. Blackburn might respond here that a fattist could doubt, worry, or be concerned with, the question of whether they are contemptible on account of their weight, but this question, as Blackburn’s discussion notes above, is a question that has no specific bearing on an individual’s weight, but on the question of their contemptibility in general (i.e., ‘Pavarotti is unquestionably fat, but transcendentally uncontemptible in other ways’). This point leads to my third, and larger, worry. The third concern is that, in order to be convincing, Blackburn’s example would have to allow for the possibility that such judgments bear some connection to a concept 50 of how to live or ethical outlook. Blackburn wants his account to generalize, and to do this, it seems that he is hindered by appealing to a model that is more at home in accounting for certain pejorative terms like ‘n****r’ and ‘k**e’. Of course, the fattists’ use of this term might not bear such larger connections, but if it does not, then there is something that we should expect to be true about those who judge others to be ‘fat↓’, namely that there will remain a certain sense in which those judgments are opaque to those who make them. There will remain a sense in which the contempt they have for fatness will strike them as brute and to that degree inexplicable, something along the lines of finding a certain flavor of ice cream distasteful. If this is the case, then there is an important sense in which the example would different from the cases that advocates of thick concepts are interested in. If, however, their contempt for fatness is related in part to other interests and concerns they have for leading ‘a good life’, such as taking it to reflect a lack of healthy dietary habits, idleness, sloth, and so on, then their judgments would be explained by a certain world view that, however arbitrary it might be in its own right, blocks any simple division between attitude (value) and property (fact). And it is not clear how Blackburn can avoid building such a feature into his account without making the discussion decidedly question-begging. Blackburn’s account, then, as I understand it, fails to distinguish two (albeit) related questions. The first is whether there is some specific evaluative attitude that is associated with thick concepts, 66 while the second is whether thick concepts have associated with them specific evaluative valences, that is, whether a thick concept’s 66 Here, when Blackburn talks of specific evaluative attitudes and reactions, I take him to mean attitudes like contempt, disgust, shock horror, admiration, pride, esteem, respect, and so on. 51 applying to some object always provides a reason for a positive or negative evaluation (regardless of whether there is a conceptual tie to a particular sort of positive or negative reaction). Blackburn is clear about his denial on the former point. When discussing the concept of cruelty, he rejects the idea there is some special reaction associated with finding things cruel: There is no hint of a special reaction, finding something cruellish, perhaps, on the lines of finding it amusing, or seeing something as yellow. And this, of course, is only common sense, for there is no special reaction, distinct from horror, detestation, and abhorrence. This strongly suggests not the existence of a special property or even a concept in any interesting sense, but the existence of a term standing as the focus for discussion of which attitudes to have to particular actions and their amalgams of motive, intention, and upshot. (Blackburn 1998, pp. 99- 100) 67 Of course, the idea that a concept like CRUEL is simply a focal point in a debate over what attitudes to take to certain properties of particular actions is to some extent hyperbolic— as we do not take ourselves to be in a discussion with others in any strong sense in using the concept CRUEL. But, more importantly Blackburn assumes that if a concept lacks a conceptual tie to some specific reaction (as say YELLOW stands to certain visual image of yellowness, or, perhaps more closely, FEARFUL stands to feelings of fear), then it is only contingently connected to any attitude. And so, because we cannot locate a specific evaluative reaction that is associated with CRUEL (as reactions of horror, detestation, and abhorrence, can be rightly associated with any number of awful things), this concept lacks a determinate evaluative valence. Denying that there is some ‘special’ evaluative attitude associated with cruelty does not entail that cruelty can be associated with any attitude whatsoever. And such a point would come out clearest if we came to see fattists 67 The end of his remark is an allusion to (Wiggins 1998a, p. 240). 52 as having certain interests in discouraging people from being overweight (say interests related to healthy or slovenly lifestyles): from the standpoint of such interests, while it might be contingent whether a person our group has such interests, it is not contingent what sorts of attitudes would be consistent with their possession. In this sense, Blackburn fails to establish his point about the contingent connection between thick concepts and evaluative attitudes. Though the concept CRUEL is more complex than his example of ‘fat↓’, I think a similar problem arises with his example. As Blackburn has described his case, there is nothing more to the fattists’ use of ‘fat↓’ beyond their recognition of people being overweight and their contempt for this. But this seems too quick, as I have already suggested above, especially if we want to follow Blackburn’s suggestion that there is a certain artifice to tying thick concepts to specific affective reactions. If this example really is intended as a quasirealist description of the sort of thing McDowell and Williams are concerned to highlight with thick ethical concepts, then there is one point that is left unexplained in Blackburn’s treatment, namely the fact that these people hold that someone’s being fat is itself a (defeasible) reason to negatively evaluate them (even if, in certain cases, they wish to overlook the fact that some people do 68 ). We should say that these people would not be considered fattists at all, if it were not for this fact about them. But the reason why they treat fatness (or the right sort of fatness) as such a reason is left 68 This point I think is implicit when Jonathan Dancy writes, in discussing Blackburn’s view, that: “They may, as Blackburn says, wish to overlook the fact that he [Pavarotti] is fat but they are in no position to deny it” (Dancy 1995, p. 271). 53 unexplained. And, what I am suggesting is that Blackburn’s account is radically incomplete without this. 69 If, however, we import some concern or interest here, say one that is related to being healthy or one that is related to an opposition of slovenliness, it is no longer clear that Blackburn can hold his claim to contingency between property and attitude. While it might be a contingent fact that some group has these interests, a contingent relationship between these interests and a certain range of attitudes could no longer be upheld. The question here would be whether fatness, from the standpoint of a fattist who has some interest or other in discouraging people from being overweight, is only contingently related to some attitudinal reactions and not others. The answer to this would seem to be clearly: no. Now suppose that they had a term that answered to the sort of interest they had in fatness (e.g., instead of uttering ‘fat↓’ they adopt the term ‘fatty’ 70 ), such a term would also be improperly characterized as having only a contingent connection to some attitudinal reactions and not others. It is important to note also that, in building certain interests and concerns into our account, the picture that emerges comes to look a lot more like a case involving a thick concept. To an outsider the use of ‘fatty’ might fail to reflect a certain consistency or rationality in usage, unless that person was able to understand not only their standards as to what counts as overweight, but their reasons for treating being overweight as a reason 69 Blackburn notes that such a relation is indeed a part of calling someone ‘fat↓’ as this will imply that the negative attitude is due to the property being referred to: in other words that such-and-such a person is fat and on that account contemptible. See (Blackburn 1992, p. 289). However, he does not make much use of this point other than to show that the expressivist can rightfully allow for it. 70 Compare the term ‘racist’ as applying to an individual whose statements and views presuppose that an individual can be inferior (superior) solely because of their race or ethnicity, and that this fact grounds a negative evaluation of such a person. Cf. (Moore 2007) who considers RACIST a thick concept. 54 to negatively evaluate individuals. It would be unsurprising then to find that being a ‘fatty’ might, to borrow Blackburn’s felicitous phrase, be “shapeless” 71 with respect to the nonevaluative features of individuals. There is nothing in this picture that Blackburn need deny: there is certainly no argument for any robust form of moral realism, or the claim that users of the concept FATTY perceive or otherwise detect the property of fattiness. However, in making this connection to the interests and concerns that such group might have, in developing and codifying the term ‘fatty’, we are also able to defuse two related arguments that Blackburn has repeatedly raised against thick concepts. 72 The first is that, to think of thick concepts as involving an inextricable blending of fact and value, or description and evaluation, undermines certain avenues of ethical disagreement: My heresy is this: thickness does a disservice to ethics. It discourages critique. To take a deliberately uncomfortable example that I hoped made the point the point especially clear, imagine a group happy in the habit of appraising women as cute. We may want to say that there is something wrong with them, along the lines of this: they admire and respond excitedly (pr perhaps enviously, if they are women) to the nonthreatening, infantile, subservient self-presentations that some women consciously or unconsciously adopt. (Blackburn 2004, p. 3) This argument does not carry the sting that Blackburn takes it to. By it, unless we are able to disentangle the properties of things from the attitudes that a certain group has towards them, we lose the ability to say what is wrong with groups who evaluate in immature and superficial ways. In the case of our fattists (or those who use ‘cute’ in the way Blackburn describes), unless we were able to see what sort of property they take to 71 (Blackburn 1981, p. 167). The idea of a concept having a shapeless extension means that the concept fails to track a neutrally intelligible property. 72 (Blackburn 1992, p. 299); (Blackburn 1998, pp. 101-04); (Blackburn 2004); and (Blackburn 2009, pp. 19-20). 55 ground their attitudes of contempt (or admiration) we would not be able to say what was wrong with them. However, there is something question-begging in the complaint itself because, under the amplified example we are considering with fattists, there is no independent property of fatness that they are responding to, in calling people ‘fatties’. Blackburn has to allow this, if he is to hold, as I think he should, that the application of such a term might very well be shapeless with respect to the neutral properties of people. The same problem is made even more evident in the example that Blackburn takes from the above quote: as Blackburn himself describes the men who find certain women ‘cute’, there is no clear way of saying what it is these men are responding to other than to say that it amounts to certain ‘nonthreatening, infantile, and subservient self-presentations’ of women. But, this characterization cannot be one, within the ambit of quasi-realism, Blackburn could rightfully treat as a property or representation: we do not perceive nonthreateningness, or infantileness, or subservient self-presentations. Because of this, they cannot be treated as the properties that men are responding to in calling women ‘cute’, and so the move to disentangle founders. This is a general problem for Blackburn’s argument, for, in claiming that there is some property (or representation) that these groups are responding to, it is decidedly uncomfortable for his position that he cannot produce it. He cannot produce it, because there is no single set of properties or representations that ground the use of such terms: that fact is what makes them shapeless. Perhaps this objection is uncharitable. Grant, for the sake of argument, that we might be able to distinguish certain general tendencies that these groups have to ridicule people on account of their weight or to praise women who 56 behave in ways that they find agreeable. Does this provide us with the means of being able to show what is wrong with them? Hardly. What we would need to know is the point behind their ridicule or admiration: was it simply because (to take the case of the fattists) they do not like the way some people look, or is it because being overweight is tied with certain health problems, or is it because being overweight is (for them) connected with a failure to lead an active and productive life? 73 It would seem then that disentangling, even if it were possible, turns out to be a non sequitur: what we need to know is the point of classifying people in these ways. But we can know this, and criticize this, without knowing precisely what sort of weight grounds their contempt (or, in the case of women, what sort of self-presentations ground the ascription of ‘cute’). Because of this, I think Blackburn is simply wrong to assume that those who advocate thick concepts do a disservice to ethical criticism. In making his point, Blackburn has simply co-opted the notion of challenging the evaluative point behind a thick concept and transformed it into a claim of disentangling a thick concept’s evaluative and non- evaluative components. And, from the standpoint of understanding these concepts, these two things should be kept separate from one another: the former is available to all parties involved; the latter requires an argument for it to be substantiated. To be fair, some of Blackburn’s later work (Blackburn 1998) seems to have taken something like the above point to heart. Indeed, in discussing the views of John McDowell and David Wiggins, Blackburn insists that his view should be understood as a ‘notational variant’ of theirs only to the degree that they, and not him, want to pin some 73 As a parallel, we might ask: is the men’s admiration driven by the fact that these women behave in nonthreatening ways and the rest, or is it because this is how a ‘lady’ or a women of ‘mild temperament’ respectably acts. 57 form of moral realism on the use of these concepts. 74 In some sense, then, Blackburn seems willing to concede that users of a thick concept (which is now conceived simply as a term used by someone with a complicated ethical sensibility, involving a variegated mix of inputs and outputs) might very well take themselves to have correctly applied their concept in certain cases. Blackburn writes: Consider again the use of ‘thick’ terms, which carry a certain evaluative weight, but are also used quite closely to criteria of application that are relatively settled, at least in respect of being from a certain range of features. There may be nothing else to think than that Hengist is courageous, or that Horsa is chaste. This means that if we work with these terms at all, we will be bound to come to these judgments. Within communities that signal courage and chastity, this is how it is done. A dissident will need to work in other terms altogether, for he will not gain a hearing by denying that Hengist is courageous or that Horsa is chaste if they obviously are so. So there is only one thing to think, provided, of course, that we think in these terms at all. But the problem lies with the proviso. In so far as the words in question signal a distinct evaluation of courage and chastity, there may well be something else to think, for it is perfectly possible to dislike and disapprove of Hengist’s prowess in battle or Horsa’s avoidance of erotic engagements. 75 What Blackburn suggests here is that users of thick concept might very well find an application of their thick concept to be explained by the features of some particular case. 74 Consider Blackburn’s discussion of David Wiggins’ view (from (Wiggins 1998a)) which is worth quoting in full: A very similar theory [to John McDowell’s] is advanced by David Wiggins, with the exception that Wiggins is, plausibly, not concerned to classify the result as a kind of ‘realism’ about ethics, but is concessive to the idea that it represents a sophisticated subjectivism. Talking about the thick ethical concept of cruelty, Wiggins too gives us a story in which we classify actions as cruel because they affect us in some specific way. They do this in virtue of the ‘marks’ of cruelty that they possess, ‘what cruelty consists in at the level of motivation, intention, and outcome [see (Wiggins 1998a, p. 240)]. So far we have the familiar ingredients on the input side. But he too wants the ‘property’ (although he says that he would have preferred to work in terms of ‘concept’) of cruelty to play a critical role in the story; a role showing that it is a ‘sibling’ of the output, rather than in any sense its creation. I suspect the best way to read what is good in this story is as a notational variant of the theory I have presented. If it implies more, then I think we should see it as deconstructing itself in the telling. ( (Blackburn 1998, p. 99); emphasis added.) 75 (Blackburn 1998, p. 303); emphasis added. See also (Blackburn 1998, p. 314), and the last line of Blackburn’s response to the question of whether quasirealism is a form of relativism. 58 But this fact would have to be conditional the assumed legitimacy of using such concept in the first place, and that is something which can always be called into question. This latter point is something that need not be denied, as it is bound up with my above appeal to invest Blackburn’s account with certain interests and concerns that a group might have in classifying acts and individuals as they do. What is not settled by someone finding an application of a thick concept to be correct is (i) whether this has any implications for those who do not use the concept in the first place; and (ii) whether this blocks criticism of the overall use of the concept. 76 The second argument that Blackburn raises is similar to the first, but focuses on our inability to see certain changes in terminology as resulting from ideological conflict, instead of on the ability to criticize others who use a certain thick concept. Blackburn notes that explaining when and where the change from one thick concept to another occurs is a difficult sticking point for those who advocate that such concepts involve an inextricable connection between fact and value, or description and evaluation: So if the change [in the application of a thick concept from one set of behaviors to another]…was simply to be described as a gradual replacement of one set of thick concepts by another, there would be no reason to see the change as involving conflict: it would be in principle open to someone to conjoin both descriptions without there being any necessary reason to see a tension between them. But if that is what remained, we would lose sight of the essential historical fact, which is that the change was indeed one of ideological conflict. The displacement was not unmotivated; it was not one of random conceptual drift, but the deliberate deployment of persuasive speech to foment a revaluation of different activities and qualities that they demanded. You simply cannot say that, without noticing that we are in the domains of both description and evaluation, the same things differently valued and differently compared to other paradigms of virtue and vice. (Blackburn 2009, p. 20) 76 Consider here Blackburn’s interesting note about how much dissent there actually was to the British abolition of slavery (Blackburn 1998, p. 303n25) which are to be taken in contrast to Wiggin’s suggestion that there might be nothing else to think but at slavery is cruel and unsupportable. 59 My objection to this second argument is related to my response to the first. What matters, in being able to locate how a change in terminology or a change in the application of a single term arises from ideological conflict, depends, not on their being some steady background of representation against which we can see how different attitudes are being applied. Rather, it depends on our being able to work out the different points that are behind the classifications being made. Consider an example. The concept of USURY use to be one that described the practices of lending money, and it used to be a negative description of such practices, tied to the idea that lenders, in allowing people to borrow money from them at certain (often high) interest rates, were behaving in ways that were not charitable or compassionate. The general thought behind the use of such lending practices was that they were inherently tied to the motives of greed and manipulation. Of course, we no longer think of the practice of lending money (even at high interest rates) as essentially a product of greedy and manipulative behavior. And so we no longer use the term to express such feelings. Given this, however, it would seem that we have a test case for Blackburn’s point: what we have is a common practice underlying certain changes in attitude. But, this fails to capture what the ideological shift was over: our shift in attitudes was not towards a common practice, but towards our understanding of how that practice was integrated into the larger social world. Prior to the onset of modern notions of investment, the practice of borrowing money was tied to concerns over basic subsistence: those who were borrowing money, were not borrowing it to make more money, but to survive. And this was something that the lenders were aware of, hence the thought that high interest rates (or charging interest at all) was tied to 60 their being greedy and manipulative, and that the called for response would not be to lend money at all in such circumstances but to be charitable and compassionate. 77 But, because the lending of money has now become integrated with a practice of investment, we no longer see the charging of interest (even high interest) as an expression of vice. Rather, we see it as the ‘sharing of profits’ that a lender is reasonably entitled to. What matters in this case, for seeing the shift in terminology as something prompted by ideological conflict, is not that there is a common practice or representation underlying the shift, but that there is a different point to classifying this practice using certain terms. In this way, it is quite misleading to represent the shift in attitudes as one that occurs against a background that has remained fixed, for, as the example of usury shows, the change in attitudes has occurred because a change in the background has occurred: the significance of the practice of lending money is different from what it once was. As such, the advocate of thick concepts is not committed to the thought that ideologically motivated changes must simply be treated as inexplicable (a sort of ‘random conceptual drift’, as Blackburn puts it). Again, where Blackburn assumes that our awareness of such ideological conflicts is made available only because we are able to hold fixed a certain descriptive background, the advocate of thick concepts could see the same conflict as a change in the interests or points in making certain thick classifications. 77 See (Duncker 1939, p. 41) and his observation that in the medieval world loans were taken out when needs arose for basic survival. In the modern world, loans are always for investment and profit, and so the idea of charging interest is tied to ‘getting a share of the profit’, which is, at the very least an ethically neutral consideration. 61 Section 1.7: Concluding Remarks A thick concept, as I have suggested above, is best seen as a concept that is developed to serve some ethical or evaluative interest. This interest not only drives the development of the concept, but it is what is appealed to in determining whether the concept applies to novel cases. Because of this, understanding what the relevant descriptive features of acts and individuals are that drive the application of such a concept is, as McDowell rightfully argued, 78 impossible to appreciate (or find intelligible) as a grouping or classification without also coming to appreciate the ethical outlook of those who use that concept. This is due to the fact that the classification a thick concept makes is one that is driven by a certain antecedent interest or concern (or set of interests and concerns): the classification does not exist in a void independent of a group’s conception of how to live. And so, understanding such concepts cannot be parsed away from understanding what is important to those who use it, and, in general, the sorts of action- guiding and emotional connections that are taken to follow from an application of the concept. Component views (whether cognitivist or quasirealist) hold that the factual and evaluative aspects of a thick concept are only contingently related to one another. The general error in these views, however, centers on the fact that, even if there is some determinate factual aspect to the concept, this aspect helps constitute a reason to evaluate acts and individuals in some determinate way. And because of this, we need to have some grip on why they treat it as a reason in order to understand when and how they use 78 Cf. (McDowell 1998b), (McDowell 1998c), and (McDowell 1998f). 62 such a concept. To bring this point out, we might be aided by a quote A.W. Moore offers, in discussing Williams’ view of thick concepts: The concept of being intrinsically wrong is not itself a thick concept. Its applicability is not “world-guided” in the way that say the concept of being a racist is. My conviction that racial discrimination is intrinsically wrong is not an item of knowledge. But—and this is the point—it does enable me to know such things as that Wagner was a racist. The clumsy appeal to the fact/value distinction obscures this. (Moore 2007, p. 43) Although Moore is undoubtedly right that, at the bottom of our using a concept like RACIST, there is a conviction (or disposition) to assess items that are racially discriminatory in a certain way (namely negatively), I would add that coming to an understanding of the conviction itself would require understanding its connection to the other (related) ethical considerations (such as the belief that race is largely an irrelevant measure for judging the merits of an individual or their acts). What is obscured by attempting to treat thick concepts as having factual and evaluative aspects that are only contingently related to one another, is simply that certain factual properties can be treated as intrinsically wrong, 79 they can be treated as grounds for negatively evaluating objects that possess them. Facts can be evaluative by being reasons for determinate evaluations. To think that the connection between the evaluation and its ground is merely contingent, then, would be to simply overlook it is indeed a ground for such an evaluation. 79 Here I am assuming, as I think Moore does, that the fact that something is taken as intrinsically wrong does not preclude our being able to justify this by means of other evaluative considerations. 63 CHAPTER 2: Thick Concepts and the FV-View Section 2.0: Introduction This is the picture that Chapter One left us with. Of the various CV approaches we canvassed, be they of a general cognitivist or expressivist bent, none provided a suitable account for holding that thick concepts involve distinct and merely contingently connected factual (descriptive) and evaluative components. This is because such views have to discount one feature of thick concepts that is taken to be crucial to them, namely that they presuppose certain concerns and interests which provide a point to their application. Such interests and concerns are what drive the application of a thick concept from one context to the next and which also determine how that concept is to be applied in novel cases. And the appeal to such interests and concerns, I argued was needed to explain the sort of reason-giving relation that even CVs were willing to grant in thick concepts, namely that the descriptive component provides a reason for the evaluation the thick concept reflects. Without an appeal to certain interests and concerns, as I suggested in the last chapter, this connection is left opaque. When we moved to discuss Simon Blackburn’s quasirealism in particular, it was important to note that it was Blackburn’s commitment to a CV approach to thick concepts, and not his quasirealism that was problematic. If it is conceded that thick concepts lay no strong claim to ethical realism, a point that (in support of Williams’ general view I will defend in Chapter 4), then dispute centers merely on whether a quasirealist can accept that, when it comes to a given thick concept, the factual component provides a reason for some determinate evaluation. I think the quasirealist can and should be open to allowing this, since doing so would 64 neither stifle criticism regarding the use of the concept, nor have any great implications for those who did not use it. In this way, even the quasirealist can agree that evaluation influences the extension of a thick concept, that thick concepts presuppose certain ethical interests and concerns, and that, because of this, they can provide users of them with certain practical reasons, under the proviso that they are actually committed to using the concept. 1 In this chapter we will consider a contemporary FV approach to thick concepts (what I am calling the fact-value fusion view). For FVs, it is not possible to separate out distinct factual and evaluative components when it comes to thick concepts. The reason for this is that what explains whether the thick concept applies in a given case is something that typically cannot be settled without addressing certain substantive evaluative considerations. On both of these points, I have clear affinities. Where I find contemporary FV approaches problematic, however, is in their attempt to make thick concepts seem more akin to other affective concepts (concepts such as FEARFUL or SHAMEFUL) that I think they really are. The central focus of this chapter will be to critically discuss this popular FV approach. The goal of this chapter will be to make a start towards developing a new account of thick concepts, an account that will itself fall in the FV camp, but which will attempt to sidestep some problems that I will raise for this contemporary approach. The beginnings of this picture, however, will not even start to emerge until we are near the end of this chapter, and will not be spelled out, in any great detail, until the next. 1 Here I think it is important to recall some of Blackburn’s own concessions regarding thick concepts in his later work. See (Blackburn 1998, p. 303 and 314). 65 Recall, then, that the central (and indeed initiating) argument for the FV view came from John McDowell. The thought was that understanding how to ‘go on’ in using a thick concept (how to apply the concept to a novel case) will involve an essential appeal to the evaluative outlook of which that concept is a part. The rationality in using such a concept cannot be rendered simply in terms of some sensitivity to non-evaluative features of the world—it must also involve the ability to appreciate the importance of certain ethical interests. The first half of this chapter will be directed towards expanding and defending this aspect of McDowell’s argument. The central idea I will appeal to is that, at least for thick concepts, understanding how to go on using them in novel contexts requires understanding the sorts of considerations (including evaluative considerations) that would justify its application in a given context. 2 The second half of this chapter will then turn to look at a popular approach to defending these central FV-intuitions, offered by Allan Gibbard. This approach aims at seeing the applicability of a thick concept as itself answerable to some particular affective reaction that is unique to that concept, and which is warranted or made appropriate by the circumstances in which the concept applies. This approach is problematic because it does not readily generalize, and is (at best) a picture of how we should understand affective concepts where there is some determinate emotional reaction associated with them. 2 Cf. (Cavell 1979, p. 62): “We know that what may be incomplete in a claim to truth is not its correspondence with the facts but the claimer’s right to the claim…Knowing how to make serious assertions is knowing how to justify them, and also knowing how to excuse them…in case they come to grief.” 66 Section 2.1: From Thick Concepts’ Shapelessness to their Justification Let us return for a moment to a point I raised in the previous chapter, namely the assumption with basic-CVs that what explains whether a thick concepts applies (in a given case) can be rendered in a completely non-evaluative manner. What is ultimately at issue is not the thought that the relevance of any non-evaluative feature (or set of features) is underdetermined but rather it is undetermined from a neutral standpoint. The relevance of a non-evaluative feature will depend on its standing in the proper relationship to the relevant ethical outlook, to the concerns and interests that give a shape to the various contexts in which a thick concept rightfully applies. For instance, whether the danger and risk (assuming for the moment that these are non-evaluative features) of rushing into a burning building count toward whether a certain act is courageous will depend on whether facing the danger and taking on such risk in that context is related to a proper goal (i.e., furthering some greater good). Failing this, it is not as though the danger and risk count in favor of thinking the act is courageous, but simply did not count enough. Rather, the danger and risk are better seen as likely to count in favor of another assessment, namely that the act was rash, careless, or fearless, or just plain stupid. In projecting the application of the concept forward, one is guided by both evaluative and non-evaluative considerations alike. Though related, the issue of what constitutes a projectable pattern is different from what, in any particular case, might be thought to explain the application of a thick concept. To see this, consider the following list of actions (characterized in a fairly neutral manner): 67 (1) X tells her husband that she is going to leave him. (2) X refuses to move from her seat after being told to do so by another individual. (3) X steps directly in front of the path of a moving bus. (4) X crosses a street, facing jeers and protests from individuals from one side and applause from individuals on the other. (5) X remains at the podium after her speech has concluded and refuses to sit down. (6) X applauds at the end of Y’s speech. (7) X does not applaud at the end of Y’s speech. (1 – 7) each picks out a very different actions, neutrally described. We might consider some (or all) of (1-7) to depict courageous acts, and consider each description to count as a, somewhat truncated, explanation of why such acts are indeed courageous. Equally importantly, however, we also might think that (given the proper context) none of these acts are courageous, and continue in this denial no matter how much additional (neutral) information was provided. Assume, however, that (1 – 7) are each a basic characterization of some courageous act. Given this, there is still room to worry that one would lack the ability to project the use of the concept into novel cases. Moreover, it would add very little to a person’s understanding of COURAGE if they were told that all of (1 – 7) were good pro tanto either in virtue of falling under these individualized characterizations or in virtue being courageous (assuming that all courageous acts indeed 68 are good pro tanto). That piece of information will not help one understand what makes an act courageous. Consider (8): (8) Y applauds at the end of W’s speech. Although (8) bears a basic similarity to (6), what we do not yet know is whether these two cases are similar in the appropriate or relevant way, that is, we do not yet know whether (8) is similar in a way that supports its being considered courageous. There is a sense in which what is relevant toward ascribing COURAGE has not yet been captured by (1 – 7), and so whether (8) is appropriately related to them has also not been captured. What is still absent is the point of grouping (1 – 7) together. In looking for a ‘point’ to the grouping, we are looking for the interests and concerns that are internal the concept COURAGE, the interests, concerns, and values that a group has which provides them with a reason for classifying actions as courageous in the first place (or better: the reason they had for developing their concept COURAGE). If understanding this evaluative point is central to understanding why these characterizations should all be grouped together as instances of COURAGE, assuming they should be, then any purely non-evaluative characterization of the concept (however embellished) will fail: at best such neutral characterizations will be highly disjunctive and reflect a gerrymandered set of, non- evaluatively described, instances. Why would one think that the various applications of a thick concept constitute a projectable pattern in solely nonevaluative (or purely descriptive) terms? One reason stems from the idea that the evaluative (in general) supervenes on the non-evaluative 69 (purely descriptive). 3 Supervenience is a philosophical term of art for the relation of necessary asymmetrical covariance between two differing classes of entities, such that some variance at the supervening level must co-occur with some variance at the subvening level. Claims to supervenience are often made when we have two disparate classes, vocabularies for, or levels of, entities, and the supervenience relation reflects our feeling that there is some sort of connection between them. 4 For instance, many are apt to think that mental properties supervene on neural or physical properties of the brain, that color properties supervene on reflectance properties of surfaces, or that the constitution of mid-sized objects supervenes on the organization of the object’s constituent atomic particles. The supervenience relation is typically characterized by the two following theses: (S 1 ) If object o has feature f, and f supervenes on a class of properties p 1 – p n , then any o* with p 1 – p n will also have f. And: (S 2 ) If any o has f, and f supervenes on p 1 – p n , then the loss of (or change in) f entails a change with respect to p 1 – p n . 5 In the evaluative case, two points are being appealed to: (i) that there can be no evaluative change without a non-evaluative one; and (ii) that non-evaluatively similar cases must be treated evaluatively similar. What supervenience leaves open, however, is 3 See (McDowell 1998c, p. 202). 4 Supervenience is often, however, thought to underwrite the view that evaluative properties depend on natural ones. It is important to note that a claim of dependence does not follow (or at least does not do so in any straightforward manner) from (S 1 ) and (S 2 ) below. One reason it does not is that dependence seems to be explanatory, and while supervenience is transitive, explanatory relationships are not. 5 My characterization follows Jonathan Dancy’s. See his (Dancy 1993, p. 77). 70 the possibility that an intelligible pattern might only be found at the supervening (or higher) level: at the subvenient (or lower) level there might be nothing that could count as an intelligible kind or grouping. McDowell exploits this gap to argue against the basic-CV approach. For McDowell, all thick concepts are (to borrow Blackburn’s expression again) non- evaluatively shapeless, 6 in the sense that the grouping a thick concept reflects has no clear or trackable pattern at the non-evaluative level—the groupings that are made are nonevaluatively heterogeneous. Recall that McDowell’s claim was that in order to conceive of thick concepts in the manner that the basic-CV argues for, it should be possible to characterize the relevant subvening property or set of properties independently of any evaluative considerations. A competent user of the concept could then be seen as having a sensitivity to the presence of certain non-evaluative features quite apart from any evaluative attitudes or reactions that they might have. If such a division between evaluative and non-evaluative elements could indeed be brought off, then it should be possible that an outsider (say an astute anthropologist or ethnographer) could be able to come to understand the concept, without having to appeal to any of these evaluative considerations. 7 What McDowell was skeptical about is whether there is anything in the CV arsenal that could make this picture 6 (Blackburn 1981, p. 167) 7 Cf. (Wiggins 1998a, p. 193): “Our subjective reactions to objects and events will often impose groupings upon them that have no naturalistic rationale.” 71 compelling. And in this regard he simply dismisses the idea that supervenience will have any impact on the problem. 8 Note that dismissing the importance of supervenience is not the same as denying it. Understanding a thick concept (like COURAGE) requires (at a minimum) understanding how to justify the ascription of it. It requires (at a minimum) understanding the sorts of contexts and considerations that count in favor of ascribing the concept, understanding the sorts of considerations that might defeat such an ascription, and understanding how one might argue for (or against) its ascription. To admit that thick concepts are embedded in a moral perspective and that this is something one must make sense of in order to make sense of the thick concepts that perspective deploys, is simply to recognize the sort of intimate connection between how users of the concept go on in applying it to novel cases, and how one must go about justifying the ascription of such a concept (if one is to be said to be justifying that concept). Saying, in addition to this, that two cases in which a thick concept applies must be similar in the relevant non-evaluative sense adds nothing, since an account of the supervenience relation cannot tells us what counts as relevant. 8 See (McDowell 1998c, p. 202). Admittedly, McDowell does not specify whether his understanding of the supervenience relation is to be read as weak, strong, or global, but it is also not clear whether this would affect his argument. Weak supervenience is the claim that if A-features weakly supervene on B-features, then, if, in any logically possible world, two individuals x and y are indistinguishable with respect to their B-features, they will also be indistinguishable with respect to their A-features (this has also been called ‘intra-world’ supervenience). Strong supervenience claims that no matter whether x and y are in the same logically possible world, or in different logically possible worlds, if they are both B-feature indistinguishable, then they will also be A-feature indistinguishable (this has also been classed ‘inter- world’ supervenience). Lastly global supervenience holds that two worlds that are B-feature indistinguishable are also A-feature indistinguishable. For more on supervenience in general, see (Kim 1993). Whether McDowell intends his conception of supervenience to be read weakly, strongly, or globally, will not affect my discussion. 72 To concede this, however, is simply to concede that there is nothing independent of originating evaluative practice that a thick concept answers to: understanding the concept requires nothing more than understanding how to engage with others who use it. Reaching this point, however, does not entail that the outsider must agree with the attitudes and reactions they in fact take towards these classifications (as McDowell says “there need be no difficulty in that” 9 ). Understanding the way of life 10 of a community is one thing, accepting it, or being a part of it, is another. An outsider’s ability to share a life with some moral community (to go native, as it were) will itself require more than her being able to appreciate the relationship between the alien community’s thick concepts and certain values and concerns: it will require her to accept a substantial portion of the values, beliefs, and standards that go into determining the identity of that outlook. Getting to this point, however, will not be a function of such an individual simply changing her attitudes towards some preexisting set of features that were independently accessible to her. Section 2.2: Assessing McDowell’s Thought-Experiment McDowell’s central insight comes from his pressing for the idea that the intelligibility of a thick concept cannot be wholly prized apart from understanding the ethical outlook of those who possess it. McDowell’s actual defense of this claim, however, centers on the idea that an outsider would be unable to “predict applications and withholdings of [the concept] in new cases” or master the “extension of the 9 (McDowell 1998c, p. 202) 10 Here I am alluding to Wittgensteinian notion of a form of life. See (Wittgenstein 2001 [1951]), §241. The general idea is that a group’s language is not something that can be successfully understood if it is separated from the practices, activities and general everyday life in that group. 73 associated term” 11 without having an appreciation of this perspective. These are, I think, rather unhappy ways of stating the spirit of McDowell’s own point, and they give rise to two potential problems for his view that I think are worth noting. The first problem relates to his placing an emphasis on prediction rather than more directly, as I have suggested, justification. For McDowell, an outsider would be unable to predict novel uses of the term, unless she was able to make sense of the framework from which the term is deployed. This, however, seems to assume that understanding the framework would make predicting the use of such a term possible. The second problem also relates to a measure of such success; the idea of mastering the term’s extension. Some have argued that the possibility of co-extensive, yet non-synonymous, terms and concepts (e.g., ‘water’ and ‘H 2 0’) opens up the possibility of being able to meet this standard of success while still failing to understand the relevant term or concept. And so, McDowell’s point cannot be secured on extensional grounds, since one might be capable of grasping the extension of a thick concept, in virtue of grasping a different, but co- extensive, non-evaluative term. 12 Let’s consider these points in turn. As I am apt read him, McDowell’s appeal to prediction is really a way of cashing out (or explicating) what is involved in understanding a thick concept. Clearly, whether a person has come to master the relevant term is tied to whether such a person understands how the term should be applied in the future. An appeal to the notion of prediction, 11 Both quotes are from (McDowell 1998c, p. 201); emphasis mine. 12 See (Miller 2003, p. 251): The non-cognitivist can agree that a competent user of evaluative language will not be able to isolate, merely by a priori conceptual reflection, the non-evaluative kind into which fall all and only items to which the evaluative predicate can correctly be applied. But he will point out that this precludes the possibility of the evaluative term standing for a non-evaluative kind only on the assumption that co-referring expressions must be synonymous. 74 however, can be given both a strong and a weak reading. Under a strong reading, there could be one who was granted mastery of the concept, even though his predictions diverged radically from the actual usage in the community. Under a weak reading, one’s mastery of the concept would have to disallow this, but retain the thought that divergences in usage would still be socially recognizable as divergences in one and the same term (i.e., the grounds for disagreeing with the rest of the community on various instances would have to be based on reasons recognizable to the community for making such a judgment). Whether we should read McDowell’s appeal to prediction here as consistent with the stronger reading, I think, relates to whether we are apt to see his position as ultimately consistent with the idea that grounds for applying the concept are codifiable, that there could be rules for using the term that the community as a whole could get wrong in interpreting. Such a reading would place McDowell at odds with the sorts of appeals he makes to Wittgenstein’s rule-following considerations as blocking the possibility of such codification. Gerald Lang, for instance, raises McDowell’s use of ‘prediction’ as problematic for this view, but does so by assuming that what McDowell means by ‘prediction’ should be read in the above strong sense. 13 According to the rule-following considerations, no one has the ability to predict the application of a concept, since any finite sequence of prior applications is consistent with an infinite number projected patterns of application. Predicting applications and withholdings seem to be the very sort of thing that the rule-following considerations forbid, since this would assume that there was some practice-independent standard that 13 Gerald Lang raises precisely this point. See (Lang 2001, p. 203). Lang, however, seems too quick to conclude that McDowell’s point should be read in a manner inconsistent with the rule-following considerations. 75 the concept was answerable to, apart from mere participation in the relevant evaluative practice. Because of this we should see McDowell’s appeal to prediction as a rather ill- phrased way of explicating the thought that understanding how to go on, in using a thick ethical concept, is understanding how to justify its ascription in novel cases. The issue here is that of understanding the community’s practice of justifying an application the concept, not (strictly speaking) being able to predict when it would apply that concept. If an outsider is engaged in an attempt to understand some alien thick concept, then she is engaged in trying to understand a certain aspect of the life of the community from which that concept is actually deployed—an understanding that includes grasping not only what the concept has been applied to, but also why it did (when it did). If, however, we shift the emphasis away from prediction and towards justification, we can still retain much of what McDowell wanted to argue for (namely the necessary connection to understanding the underlying ethical outlook), without requiring a suspicious appeal to some practice-independent account of a thick concept. The relationship between justification and an evaluative perspective connects closely with another point, raised independently, and in different ways, by both Simon Blackburn and Frank Jackson. This is the thought that an outsider’s inability to understand a thick concept without grasping the related evaluative perspective is really a contingent human (or finitely rational) limitation. The need to appeal to an evaluative perspective (on this view) stems solely from the outsider’s inability to adequately grasp how the concept applies by appealing strictly to the non-evaluative information available. The implication being that, if this limitation were removed, then the outsider would be 76 able to discern a projectable pattern in the variously different cases the concept applies to. Blackburn expresses this in the following passage: Let us suppose for a moment that some group of human beings does share a genuine tendency to some reaction in the face of some perceived properties or kinds of thing. Surely it need not surprise us at all that they should know of no description of what unifies the class of objects eliciting that reaction, except of course the fact that it does so. We are complicated beings, and understand our own reactions only poorly. Now suppose the outsider, who fails either to share or to understand the reactive tendency, cannot perceive any such unifying feature either. Then he will be at a loss to extend the associated term to new cases, and there will be no method of teaching him how to do so. 14 Frank Jackson expresses a similar thought, in seeing our inability to grasp the non- evaluative groupings of an evaluative concept as an expression of our limited cognitive capacities: …ethical language may be needed in practice to capture the similarities among the various descriptive ways that [global supervenience] tells us constitute ethical nature, but ethical properties are, nevertheless, possibly infinite descriptive properties. 15 Both of these views, in different ways, assume that an ethical outlook is only of heuristic or pedagogic value to determining a thick concept’s application. Such outlooks are simply crutches that must be relied upon to help us overcome our cognitive short- comings, as if we suffered from something like a learning disability in being able to determine the right non-evaluative pattern associated with a thick concept. Thus, such evaluative perspectives (at best) simply help us to identify or discover the patterns that underlie our ascription of a given thick concept, but these perspectives play no role beyond this (or at least no role that is of philosophical interest). 14 (Blackburn 1981, p. 167); emphasis mine. 15 (Jackson 1998, p. 124) It is not clear whether Jackson is using the term ‘descriptive’ in a manner roughly synonymous with ‘non-evaluative’. Though I will not pursue the matter, I assume that he is. 77 This picture, however, fails to appreciate (or take at all seriously) the constitutive role that an evaluative perspective plays in the context of justification. There is little reason to suppose that, if there is a nonevaluative pattern that mirrors the application of a thick concept, this would matter to the justification of that concept, since we might think that such a pattern matters little to how the concept’s application is justified in a particular case. And as I have suggested, what matters to understanding the concept is tied to understanding what would justify its application: one would still need to appreciate the connection that such a concept has to the values, interests and concerns that help make up its users’ ethical outlook. The second problem I mentioned above relates to the possibility of coextensive, yet non-synonymous terms, opening the door for their being coextensive non-evaluative terms to thick concepts, and hence, for it being possible that an outsider would be able to grasp the extension of a thick concept without understanding the concept itself. Certainly there are coextensive, yet non-synonymous, terms that, in other discourses, we could see an individual as grasping one and yet still fail to grasp the other (e.g., one might understand the concept WATER while lacking any understanding of chemistry, or there might be shut-in chemists who never learned that H 2 0 is what everyday folk call the stuff that falls from the sky, flows in rivers, or is sold in bottles of Evian). Given this, one might wonder why we should be so reluctant in thinking that there might also be coextensive, yet non-synonymous, analogues to thick concepts. Assume, then, for the sake of argument that we could construct a well-formed, if massively complicated, open sentence that specifies, for each application of a given thick concept, some non- 78 evaluative characterization of that instance. Now, to offer such a construction, we would specify every naturally possible situation (i.e., every possible world that is consistent, say, with the laws of physics), identify in those various situations the instances of the thick concept we are interested in, and then characterize those instances in non-evaluative terms. Given this, there could be a non-evaluative (though massively complex) disjunctive predicate which covers all the instances to which the term has applied, and hence would be (for our purposes) coextensive with it. 16 Conceding the possibility of such a massively complicated non-evaluative term, and specifying (in a somewhat schematic way) how we would go about constructing such a coextensive surrogate, however, glosses over some serious concerns in how we would actually be able to carry such a procedure off. It would require at least two things. First, it would require a complete non-evaluative specification of all possible worlds. This, to say the least, would be daunting, but I will not pursue this point. The more pressing concern is that such a procedure would itself presuppose an adequate grasp of the concept that we are seeking to supplant. 17 McDowell’s point need not be read as denying that such a term could be constructed, but to deny that it could be constructed from a position independent of an appreciation of the outlook it is inherently related to. And so, McDowell need not be concerned with this possibility, for at best, if they developed such 16 This is Frank Jackson’s approach in (Jackson 1998), but it is also adopted by Allan Gibbard in (Gibbard 2003b). 17 See here Brad Major in (Major 2005, p. 483) and his ‘kind constraint’, which is that “if a property is of kind K, then it must be possible to pick out that property using only the descriptive and explanatory resources of level K.” The upshot is that if someone (such as Gibbard or Jackson) wishes to claim that ethical concepts denote purely descriptive properties, then it must be possible to pick out that property using purely descriptive means. Major takes the fact that both Gibbard and Jackson see the descriptive properties denoted by thick concepts to be infinite disjunctions as counting against their view for this very reason. 79 a term, we would have to consider it a sort of accidental concordance. What such a person would still lack would be an understanding of why the concept had the extension it does rather than a different one—she would lack an understanding of why those instances fall under one and the same term, other than the fact that they do. Without an understanding of the evaluative considerations that are intimately related to a given thick concept, one would still fail to understand this. Pulling off a non-accidental reconstruction, though, would be parasitic on an understanding of the original concept. The entire point, however, was that one could get away with doing so without having to import certain evaluative considerations as essential for determining that concept’s extension. If we have to import the concept (or at least our understanding of the values and concerns it relates to) in order to carry off the reconstruction, then the process of doing so is simply a redundant move. Thus, granting the possibility of a non-evaluative surrogate need be nothing but an empty concession. What might tend to mislead here is the thought that the concept’s extension (i.e., what the items are that fall under the concept) is the central or important issue that needs to be accounted for, over why those items would be thought to fall under that concept. In my view, this simply seems to get things back to front. Given this, I think McDowell’s argument is on a rather stable footing. The central issue, in understanding how to go on in using a thick concept, relates to understanding how to justify its application, and not knowing how to predict novel cases or grasp its extension (assuming these latter points can be read as related to something other than justification). How McDowell actually understands the detailed nature of thick concepts, 80 however, goes beyond this picture in ways that are significant (and, to a certain degree, disappointing). In the next section, I will consider these positive views, and, in the subsequent sections, I will argue that (at best) they are capable of explicating only a small portion of thick concepts. Section 2.3: Wiggins and McDowell on Thick Concepts John McDowell 18 and David Wiggins, 19 though differing in detail, each offer up quite similar FV accounts of how we should understand the nature of thick concepts. 20 Their respective approaches both start from a common rejection of the (assumed) exhaustive distinction between Platonistic (i.e., primary-quality, mind-independent) realism and straightforward antirealism in metaethics. The choice between these two options reflects, for McDowell and Wiggins, something of a false dichotomy, 21 which can be expressed in terms of a Euthyphronic contrast between two orders of explanation. Consider (OE 1 ) and (OE 2 ): (OE 1 ) Actions seem cruel to us in virtue of their being cruel. (Platonism) (OE 2 ) Actions are cruel in virtue of their seeming cruel to us. (Antirealism) The crux of the debate between Platonism and Antirealism over thick concepts would be whether the application of a thick concept is something determined by some mind- independent feature of the world (OE 1 ) or simply by the attitudes and responses we 18 (McDowell 1998d), (McDowell 1998e), and (McDowell 1998f). 19 (Wiggins 1998a), (Wiggins 1998b), and (Wiggins 1991). 20 Positions similar to that of McDowell and Wiggins can be seen in the works of Jonathan Dancy and Charles Taylor. See (Dancy 1995) and (Taylor 1989, pp. 6-8). I will return to critically address Dancy’s own emendation to the McDowell-Wiggins position below. 21 Cf. (McDowell 1998d, p. 157) and (Wiggins 1998a, p. 106). 81 actually have (OE 2 ). What McDowell and Wiggins both stress is that this represents a false dichotomy, and we should not feel forced to choose between the two. Rather, there is an intermediate position, one that takes there to be a kernel of truth reflected in both (OE 1 ) and (OE 2 ). The Antirealist rightly thinks that the discriminations made by a thick concept cannot be properly understood in abstraction from how human beings are constituted, including their constitution as emotional and social beings. To concede this, however, is simply to grant that the properties denoted by thick concepts as anthropocentric in nature. Admitting such anthropocentric dependence need not hinder our thinking that judgments ascribing such properties can nevertheless be candidates for plain truth and factuality: some acts that we take to be cruel actions are genuinely cruel, just as some signs that we take to be red really are red. Neither the property nor the response need be seen as having a legitimate claim to explanatory priority. The particular approach favored by both McDowell and Wiggins, 22 in characterizing thick concepts, draws upon an analogy with secondary property realism. 23 22 McDowell aims to secure a form of evaluative realism, and in this sense, should be seen as having greater ambitions than Wiggins. Wiggins rejects the realist / antirealist distinction altogether, and opts instead to consider his position a form of ‘cognitivism’ or ‘objectivism’. See (Wiggins 1991) and (Wiggins 2005). 23 It is important not press the analogy too far. There is an exegetical point to the analogy (namely to highlight the anthropocentric nature of value properties), and beyond this it serves little purpose for either thinker. For instance, Wiggins notes that: …values are non-natural things and that may make them seem queer. But on a careful understanding of ‘non-natural’, this only means that values do not have to pull their weight in one of the experimental sciences. That does not make them inaccessible to thought, experience or feeling….But let me add that one need not take the view that the way in which values are accessible to thought it to be understood on the model of the way in which colours are accessible to perception. The true purpose of the analogy that has sometimes been made between colours and values is not that comparison. The purpose of the analogy was only to draw attention to an abstract point about the part played by a sensibility… (Wiggins 1991, p. 80) McDowell similarly wishes to limit what he thinks the secondary-quality analogy is good for: By means of the secondary quality analogy I intended no more than to remove one obstacle to that thesis [that moral judgements can be candidates for plain truth and factuality] an obstacle that 82 What is central to this analogy is admitting that there is a circular relationship between the relevant evaluative property denoted by a thick concept and human sensibility, in much the way that color properties are circularly related to human perceptual capacities, neither of these can be completely abstracted away from certain anthropocentric considerations. Because of this McDowell and Wiggins view their own positions as elucidations, not reductive analyses. The analogy provides us with a picture for furthering our understanding of the relationship between a thick concept, its related property, and various affective sensibilities, even if that analogy does not hold in all respects. McDowell characterizes a secondary-quality in the following manner: A secondary quality is a property the ascription of which to an object is not adequately understood except as true, if it is true, in virtue of the object’s disposition to present a certain sort of perceptual appearance: specifically, an appearance characterisable by using a word for the property itself to say how the object perceptually appears. Thus an object’s being red is understood as something that obtains in virtue of the object’s being such as (in certain circumstances) to look, precisely, red. 24 The intelligibility of a secondary-quality (e.g., redness) is related to our understanding certain objects as having a disposition to produce a certain perceptual experience (e.g., a perceptual experience of an object as red, in suitable perceivers under normal conditions). Such properties, then, are not comprehensible apart from certain paradigmatic experiences. These experiences are such, however, that a certain color property must be made available to explanations of them. This marks a contrast between secondary- and could be formulated by saying that truth is objective whereas moral judgements are subjective…I want to stress that that is all I meant the analogy to show. (McDowell 2000, p. 112) 24 (McDowell 1998e, p. 133). For a clear account of such dispositional views of secondary-qualities see (McGinn 1984), Chs. 1 and 2. 83 primary-qualities: in the latter case, the intelligibility of the property is not tied to any perceptual experiences whatsoever, and as such is understandable completely apart from them (or our explanations of them). 25 Of course color properties do not figure in the fundamental explanations offered by natural science, and in this sense we might think of such properties as having a narrower cosmological role than say primary-qualities. 26 But this should not be taken as a strike against their objects being genuinely (say) red—only as a statement of their inability toward being assimilated into the explanatory schemes of science. Color properties are real, then, in the sense that they are indispensable to certain explanations of color experience—in this sense, they can be seen as pulling their own weight in such explanations. And yet, while color judgments are essentially tied to human capacities, these judgments can be legitimately true and factual. As such, secondary-qualities can be taken to provide a model of comparison for a type of anthropocentric objectivity (i.e., as a model for how a property can be both genuinely real and yet constitutively connected to human sensibilities). 27 25 The properties of mass and squareness can be intelligible without any connection to our experiences of them—they can be made intelligible (say) by simply appealing to the explanation of certain physical or geometric relations. 26 The notion that properties can have wider or narrower cosmological roles is Crispin Wright, see his (Wright 1992, pp. 196-97). Wright enters the distinction as a determining factor for whether we should be realists about a certain domain of discourse. 27 A central concern here (both in the color case as well as the case of evaluative properties) is that the account is openly circular, in the sense that neither property nor response can be adequately characterized apart from one another. The worry is that without an independent characterization of either the sensibility or property, different sensibilities (or properties) cannot be distinguished from one another. For instance, it is easy to understand a peg fitting into a round hole in virtue of its shape, but “what shape in particular do an otherwise unspecified peg and hole have thanks to the fact that they fit one another?” (Darwall, Gibbard & Railton 1992, p. 158). I will return to this point below. 84 In applying this model, however, to thick ethical concepts there is a crucial difference between them and concepts of secondary-qualities. While color properties are understood as properties that elicit certain characteristics perceptual experiences, the properties denoted by thick concepts do not simply “elicit the appropriate ‘attitude’ (as a colour is merely such as to cause the appropriate experiences), but rather are such as to merit it.” 28 If concepts like COURAGE or CRUEL apply to some situation, this is not simply because the situation elicits (or tends to elicit) such a judgment from those who are aware of its features, but because the situation itself provides those who use the concept with a reason for so ascribing it. The actual response that we might have to a particular case is of little relevance in contrast here to the idea that the application of such concepts can be considered fitting or appropriate in light of the situation they apply to. 29 This shift of emphasis, however, reflects a change in what such an account is supposed capture: the explication of a thick concept is to capture, not the causal triggers of a thick judgment, but the way in which users of such a concept can understand certain situations, actions, and individuals as worthy or deserving of their ascription. 30 28 (McDowell 1998e, p. 143) 29 While ‘meriting’ is McDowell’s favored idiom, ‘makes appropriate’ is Wiggins’ (Wiggins 1998a, p. 187). The locution of ‘fitting’ or ‘fittingness’ that I have borrowed here can be traced back to (Ewing 1939, p. 4) and to (Broad 1930, p. 164). See also (D'Arms & Jacobson 2000). 30 These different evaluative locutions (meriting, fitting, deserving, worthy, appropriate) all amount to roughly the same thought: that a judgment ascribing a thick concept to some act or situation finds there to be something normatively at stake in the act or situation it is applied to. The act or situation provides one with a reason for ascribing such a concept, and is not something that is merely elicited by our awareness of the situations features. 85 Wiggins considers the relationship here to fit what he calls a “<property, response> pairing,” 31 where neither the property, nor the attitude (i.e., response), are taken to have a more basic explanatory role, since the property can only be understood as one ‘calling for’ or ‘making appropriate’ a certain response, and the response is only understood as one ‘called for’ or ‘made appropriate’ by the property. The <property, attitude> involves a type of symmetrical normative dependence. For instance, in discussing the concept GOOD, Wiggins notes that there is no possibility of finding something GOOD (in some respect) without finding it as something that merits (or makes appropriate) approval, nor is there any possibility of finding something FEARFUL 32 without finding it as something that merits (or makes appropriate) a fear response. Since debate and criticism can focus on the applicable situations themselves as being either worthy or ill-suited for such a response this marks an important change of emphasis from an expressivist view (e.g., Blackburn’s quasirealism) which takes such an evaluative judgment to be the expression of an attitude directed towards a world devoid of features for which such expressions can be understood genuinely answerable. For Blackburn, there is a clear explanatory priority given to our responses, whereas on the McDowell-Wiggins view the order of explanation requirement is not reversed but simply abandoned. And, given this, we can also see how the case against a CV theorist unfolds. If the property associated with a thick concept is such as to merit a some attitude (or response), while the attitude (or response) is something that itself is made appropriate by the property, then there is no hope of singling out some non-evaluative property without 31 (Wiggins 1998a, p. 196) 32 McDowell’s example, see (McDowell 1998e, p. 144). 86 appealing to the reciprocal, patently evaluative, relationship that the property and attitude are taken to stand in. The idea of independently intelligible elements within a thick concept, then, cannot even get off the ground. Though I have deep affinities to the view put forward by McDowell and Wiggins, I also think that there are some serious problems facing such an approach (the central one being Blackburn’s concern that, when it comes to thick concepts, there does not seem to be a reaction uniquely merited by their application). But, before moving to discuss these concerns, I think it will be helpful to first compare it with an approach to thick concepts that attempts to make sense of them from an avowedly expressivist standpoint—that is, a standpoint which chooses to understand these concepts in terms of their expressing certain specific emotional responses (or being bound up with judgments to the effect of finding such responses warranted). This is the view of thick concepts offered by Allan Gibbard. Setting up such a comparison will aid us I think in determining what is worth salvaging from the McDowell-Wiggins view. Section 2.4: Gibbard and the Warrant for Thick Feelings Allan Gibbard has developed an account of thick concepts that occupies an interesting place in the debate between CVs and FVs. 33 Though his general metaethical orientation (like that of Simon Blackburn) is expressivist, Gibbard’s position on how we should understand thick concepts bears some striking similarities to the McDowell- Wiggins view. While Gibbard objects to a basic-CV approach to thick concepts, as 33 Here I will draw mostly from (Gibbard 1992). See also (Gibbard 2003a) and (Gibbard 2003b). 87 Gibbard argues that there is simply “not enough” 34 descriptive meaning to form a component capable of determining a thick concept’s application from context to context, he is unwilling to side with those who think that such concepts are incapable of being given an expressivist treatment. Before getting into Gibbard’s account of thick concepts, some background is needed on his larger metaethical position. The theory presented in his book Wise Choices, Apt Feelings 35 provides the best overview of his position. This text was directed towards how we should understand normativity broadly conceived, and offered an account of the meaning of the term “rational,” as when we judge action, belief, or emotion to be so. The evaluative force of judging something to be rational is explicated in terms of the mental state of norm acceptance, a type of mental state that is taken to have essentially practical connections. When one judges some act, for instance, to be ‘rational’, one’s judgment is itself an expression of their acceptance of a body of norms that permit acting in such a way. As such, attitudes of norm-acceptance explain the normative aspect of evaluative judgments, as opposed to the positing of some property of rationality that the action in question has independently of our attitudes towards it. Gibbard’s discussion of thick concepts centers on seeing them as involved with quite specific emotional attitudes that are warranted by certain sorts of actions or situations, although what these actions or situations are is not something that can be fully specified in purely descriptive terms. 36 For Gibbard, there is simply not enough on (what 34 (Gibbard 1992, p. 278) 35 (Gibbard 1990) 88 we might take to be) the non-evaluative side of thick concepts to think that there is any straightforward independent (and non-evaluative) property that the concept is picking out. Gibbard notes that there are indeed non-evaluative properties associated with thick concepts (what he calls “descriptive constraints” 37 ), but that these ‘constraints’ alone are not able to determine whether the concept applies in a given context. The example that Gibbard focuses on is the concept LEWD, where an act is roughly understood to be lewd if it is “objectionably sexually open or provocative.” 38 To be sure, whether an act is lewd is not simply a matter of whether it is objectionable, for non-lewd acts can be objectionable, nor is an act lewd simply if the act was objectionable for its display of sexuality, for sexual displays can also be objectionable on grounds other than their lewdness. Rather, the objectionable nature of lewdness involves a further related component. There is a specific emotional reaction that is typically elicited in those who find certain acts to be LEWD, such that judging a certain act to be lewd involves thinking that this specific reaction is warranted. The specific reaction that is typically elicited by LEWD acts is (what Gibbard calls) L-censoriousness (i.e., lewd- 36 Gibbard rejects three models that he thinks a CV-theorist might try to avail himself of in understanding thick concepts (as involving a combination of descriptive and evaluative components): conjunctive (a judgment deploying a thick concept is really the conjunction of a straight description and an evaluation), licensing (a judgment deploying a thick concept is really stating only a descriptive statement, but the speaker is licensed by the language to use the term in question only if the speaker endorses the evaluative valence of the concept), and presuppositional (a judgment deploying a thick concept presupposes an evaluation in concordance with the concept’s assumed evaluative valence, but the judgment itself is simply conveying a descriptive statement). Gibbard argues that all three models fails because there is “too little descriptive meaning” to make sense of there being a separable and independently intelligible non- evaluative property. See (Gibbard 1992, p. 274). 37 (Gibbard 1992, p. 277). Cf. (Burton 1992) 38 (Gibbard 1992, p. 279) 89 censoriousness). Given this, Gibbard offers up the following description of the central features of involved in making a judgment that some act is lewd: ‘Act X is lewd’ means this: L-censoriousness toward the agent is warranted, for passing beyond those limits on sexual display such that (i) in general, passing beyond those limits warrants feelings of L-censoriousness toward the person doing so, and (ii) this holds either on no further grounds or on grounds that apply specially to sexual displays as sexual displays. (Gibbard 1992, pp. 280-81) The idea behind Gibbard’s view seems to be this. Individuals who learn a term like ‘lewd’ come to accept certain presuppositions to the effect that certain sorts of sexual behavior should be limited, and, in the process, they come to develop a sensitivity for having feelings of L-censoriousness evoked when sexual displays ‘pass beyond’ certain implicit limits. (Part of their training, then, involves developing a sensitivity to what sorts of sexual displays indeed go beyond the relevant limits.) When an individual becomes a mature (or fully competent) user of the term ‘lewd’, they might no longer have feelings of L-censoriousness each time they judge an act to be ‘lewd’, but their so judging nonetheless involves taking such a reaction to be warranted towards the violator. There is a great deal to be sympathetic with here. What seems right is the thought that a thick concept involves certain (perhaps implicit) standards—ones that are tied to some general sphere of social activity (in this case: sexual displays). Another move that seems correct is with the presupposition of certain evaluative concerns (namely, that a user of the concept accepts the presupposition that certain sorts of sexual acts should be limited). Indeed, in the next chapter, I will use these elements in advancing my own account of a thick concept. What I find troubling with Gibbard’s account, however, is his appeal to the idea that thick concepts have specific affective reactions associated with 90 them (e.g., L-censoriousness). Since it seems to play an explanatory role in whether the concept LEWD applies to some individual, it would seem crucial to his account that such a feeling be at least intelligible, independently from that concept. Indeed, Alasdair MacIntyre gives voice to this sort of worry in the following passage (though it is directed towards the emotivist predecessors of Gibbard’s norm-expressivism 39 ): …if the theory is to elucidate the meaning of a certain class of sentences by referring to their function, when uttered, of expressing feelings or attitudes, an essential part of the theory will have to consist in an identification and characterization of the feeling or attitudes in question. On this subject proponents of the emotive theory are in general silent, and perhaps wisely. For all attempts so far to identify the relevant types of feelings or attitudes have found it impossible to avoid an empty circularity. ‘Moral judgments express feelings or attitudes’, it is said. ‘What kind of feelings or attitudes?’ we ask. ‘Feelings or attitudes of approval’, is the reply. ‘What kind of approval?’ we ask, perhaps remarking that approval is of many kinds. (MacIntyre 1984, pp. 12-13) I think there is a similar concern that plagues Gibbard’s approach here. For Gibbard’s account to succeed, it must be possible for us to understand what the feeling is that is at stake in the application of a given thick concept without resorting to an understanding of the concept itself. This is because these specific feelings are taken to have an explanatorily basic role in accounting for the character of this concept. Telling us that the specific feeling associated with attributions of lewdness is a type of censure that is typically felt when confronted with lewd acts renders the account circular and potentially vacuous. This is a quite serious problem for Gibbard’s view. But it is also due to a rather superfluous part of his treatment of thick concepts. The appeal to the warranting of a specific feeling actually makes his account of thick concepts (if the reader will allow me to put it in such terms) doubly evaluative, given that an account of the concept must 39 For example, (Stevenson 1937) and (Ayer 1946). 91 already invoke the idea that there are certain limits (i.e., standards) to be imposed on sexual displays qua sexual displays in order for an act to count as LEWD or for it to warrant L-censoriousness. I will return to discuss this point in detail in the next chapter. But, circularity is not the only concern here. There is also a sense in which the account is ad hoc in construction. The need to come up with a unique affective attitude for each thick concept seems necessary only to preserve a certain metaethical picture of how we should understand ethical judgments in general, namely expressivism. To preserve this picture, each thick concept would require a unique emotional reaction that is paired with it. When we consider the wide plurality of thick concepts there are, it would be natural to feel as though we are being asked to postulate a body of reactions simply because the view on offer does not have the resources to explain these concepts by any other means. What is ironic, however, is that Gibbard’s treatment could avoid this problem if it were only to drop this required connection to specific attitudes, and retained the notion and role of an evaluative or ethical standard. But (again) this point will have to wait until the next chapter (and Chapter 5) for fuller consideration. In the interim, however, there is a much more general concern that I want to pull from this discussion. We can formulate it by asking the following questions: Why should we think that the McDowell-Wiggins view is, with respect to these problems, in any better shape? Should the fact that a situation merits the appropriate attitude, as opposed to one’s judging that a situation warrants it (where the concept WARRANT is given an expressivist gloss), make a crucial difference here with respect to whether we think the account is ad hoc or not? To be sure, there are important differences between the 92 McDowell-Wiggins view and Gibbard’s own position, but it is unclear whether these differences alter the problems faced by this general sort of strategy. In the next section we will consider these worries in detail. Section 2.5: Assessing Sentimentalist Treatments of Thick Concepts The short response to this worry is, I think, that these problems (namely the problem of circularity and the sense of the accounts being ad hoc in construction) are not only problems for Gibbard’s view, but also that of Wiggins and McDowell. Each thinker, in some form, requires a specific emotional reaction to be at stake in our understanding of the concept: indeed the meriting or warranting of such an emotion is a key factor in determining whether the concept applies (or relevant evaluative property obtains), as both the application of the concept and the meriting or warranting of the attitude are dependent on the character of the situation (as something that ‘calls for’ both). Of course, one might think that these problems are only a concern for Gibbard’s account, given his clear acceptance of an order of explanation that would require a unique, independent, and intelligible emotion standing behind each thick concept. 40 For Gibbard, the attitude must relate both to the individuation of thick concepts (i.e., tell us something about what the difference is between say LEWD and COURAGE) and also give us some insight into what is at stake in the concept applying (i.e., tells us something about the subject matter that could be the object of disagreement). By contrast, one might think 40 Allan Thomas, in discussing Gibbard’s view of thick concepts (to which he is rather uncharitable) accuses it of being viciously circular. Yet, his only ground for seeing Wiggins and McDowell as avoiding this problem seems to be that they cognitivists and he never justifies why the appeal to an evaluative property aids Wiggins and McDowell in avoiding this problem. See (Thomas 2006, pp. 127-28). 93 that Wiggins and McDowell escape the problems of circularity and ad hoc construction because they both reject having to choose between different orders of explanation. I do not think this makes for a strong defense, because, regardless of their general stance towards an antirealist or expressivist order of explanation, as each invokes (or depends on the invocation of) such specific attitudes as a way of elucidating the sense in which thick concepts are evaluative. In this, Wiggins and McDowell seem to face the very same worries as Gibbard. The reason for these problems being common to each thinker is that they all share a common theoretical commitment to sentimentalism in their discussion of specific evaluative concepts. Sentimentalism is the (broadly Humean 41 ) view that evaluative judgments involve a rather tight and intimate connection to human sentiments and emotional responses. 42 In this regard, each of thinker shares a common methodological approach to understanding thick concepts—they are to be understood or explained (or elucidated) in terms of the concept in question standing in some normative relationship to a feeling or emotion. 43 Though I think that a generally sentimentalist treatment of thick concepts places too heavy a burden on finding a finely-tailored connection between such 41 (Hume 1978, p. 470) 42 For an excellent and concise history (from Franz Brentano to Gibbard) of the sentimentalist approach, see (Rabinowicz & Ronnov-Rasmussen 2004, pp. 394-400). 43 This normative idiom (whether it be warranting, meriting, making appropriate, etc.) must itself be carefully specified, as there are cases in which the warranting of the relevant attitude and the proper application of the concept can come apart. See (D'Arms & Jacobson 2000). This sort of conceptual separability leads to what D’Arms and Jacobson call the conflation problem. It might be inappropriate to be amused at a genuinely funny joke, say because the joke is will be offensive on moral grounds, or because showing amusement by it will hurt your chances of getting a promotion. In this sense, moral and pragmatic considerations can make it inappropriate to be amused by the joke, but these considerations need not be relevant to whether the joke was funny or not. Sentimentalism has yet to show how there is a sense of appropriateness or fittingness that is uniquely relevant to the concept in question, without having to appeal to evaluative properties. 94 concepts and our emotional nature, there is an important intuition about the use of such concepts and our nature emotional that I think should be defended, and that is the essential role that our emotional natures have in the genesis and development of thick concepts and their related evaluative practices. Had our emotional constitutions been different, then there is no doubt that the sorts of thick concepts that would have developed would have been different. Where I think sentimentalist treatments err is in seeing this role in the very determination of the concept’s content (i.e., as needing to have a role provided for certain emotional reactions in our very understanding of such concepts). We need not see the connection to our emotional natures as having a direct bearing on any particular emotion or affective reaction in order to appreciate how the identity of those concepts (in terms of their genesis and development) depends on our emotional nature. Wiggins discusses directly the issue of circularity and dismisses it as a concern on the grounds that his account is not giving a reductive definition of these concepts, but simply an explication (or elucidation) of them, and because of this we might think that this concern is indeed less troubling than it appears to be. The charge of circularity is problematic, according to Wiggins, only if we are tempted to think that what is being done here involves the giving of a complete and reductive conceptual analysis, as opposed to a “commentary” on a concept we are already familiar with. 44 Yet, even if we grant that circularity is not necessarily a problem here, we will still be left with the worry that the explication on offer is somewhat vacuous, because the circle we are being invited 44 (Wiggins 1998a, pp. 188-89) 95 to accept is a rather tight one between property, response, and concept. 45 Indeed, to borrow a critique from Stephen Darwall, Peter Railton, and Allan Gibbard, while we understand how a round peg fits into a round hole, and a square peg fits into a square hole, how are we to understand “what shape in particular…an otherwise unspecified peg and hole have thanks to the fact that they fit each other.” 46 It is fine to admit that (at the end of the day) a complete analysis might be unattainable, but there is something worrisome in an account that seems to leave us not understanding what even the different parts of the elucidation are supposed to be, apart from the fact that they are parts of an elucidations of something we are concerned to understand. Might the sentimentalist approach limit itself to reactions that (at least prima facie) seem to reflect an emotional complex that is independently understood (i.e., understood apart from the values and interests of some particular group)? If what counts as the relevant affective reaction were limited to those emotional responses that we might already consider independently intelligible (e.g., emotional responses like: fear, joy, envy, abhorrence, shock, disgust, contempt, etc.), then it would seem that there are three immediate problems that would need to be dealt with or explained away. First, there are 45 This sort of tight connection is reflected in Wiggins’ view of the different sorts of acts that thick concepts will have to cover. Wiggins writes, reflecting on Harmon’s example of hoodlums setting fire to a cat (Harman 1977): To measure up to what seems to need to be explained here, a general theory of some sort will be required, and this will have to place our particular reaction in the face of what the hoodlums did in some relation to our responses to a whole heterogeneous collection of events whose only common property is that they are cruel in some degree. (Or so one is tempted to put it. What else after all do the following have in common?—A man snubs a child; the courtiers steal Rigoletto’s daughter; the hoodlums set fire to the cat….) (Wiggins 1998b, p. 157); emphasis mine. What is worrisome is that, if broader connections cannot be elucidated, then saying that these different cases all have something in common seems to make the connection between them deeply inexplicable. 46 (Darwall, Gibbard & Railton 1992, p. 158). The same charge, however, applies (ironically) to Gibbard’s own view. 96 simply too few emotional responses and too many thick concepts on offer, to see them as uniquely paired with one another. Consider this the problem of the paucity of unique emotional responses. 47 Second, it is plausible to say that, in some cases, the application of different thick concepts could ground (or be a reason for feeling) one and the same emotional reaction. Consider this the problem of emotional overlap. And finally, in some cases one concept could merit multiple reactions (e.g., different acts that are disloyal and unjust could be seen as both meriting contempt, while a cruel act could merit multiple reactions: shock, dismay, abhorrence, even feelings of forgiveness). 48 Consider this the multiplicity problem. Of course, these are similar to the problems Blackburn appealed to in thinking that thick concepts must simply be concepts with a strict nonevaluative content: the trouble, however, is that there is no convincing argument for accepting this, so long as such concepts are taken to have an clear connection to some ethical outlook. These concerns are likely what compel the positing of a unique and special emotional response for a given thick concept (e.g., Gibbard’s L-censoriousness for LEWD), one that was related specifically to that concept over and above what we would consider to be more commonplace emotional reactions. Such a move directly addresses the above three problems, and also frees the responses that we are more familiar with 47 Along these lines, Justin D’Arms argues that the sentimentalist project should be limited to those concepts (and properties) that can be paired with independently understandable emotional responses. See (D'Arms 2005). 48 A similar point is made by Jeremy Koons against response-dependence views of ethical concepts, but he (inexplicably) takes the McDowell-Wiggins view to be excluded from his criticism. See (Koons 2003, pp. 278-79). 97 from being artificially tied down to any particular thick concept. 49 However, it is because such emotional responses must be unique that they also seem tailor-made for each individual concept. Recall, l-censoriousness was the sort of feeling warranted when one judges that a certain action LEWD, but it is not the feeling warranted in judging some sexual display as harassing or simply objectionable. That this emotional response seems tailor-made for the concept LEWD was the basis for my criticizing such an approach as simply ad hoc. Might not the McDowell-Wiggins view simply disown the assumption that there needs to be a single emotional reaction involved in their view of thick concepts? After all, unlike Gibbard, they allow for there to be some sort of evaluative property paired with the concept as well. Why not simply see cruel acts as making appropriate a whole host of more garden-variety attitudes as opposed to essentially just one unique and specific one? Jonathan Dancy takes just this approach in attempting to defend the McDowell-Wiggins view from these problems. 50 In light of Blackburn’s criticisms, he concedes that such a position should be able to see a single thick concept as meriting a wide variety of emotional responses, and so an account of them should not be saddled with holding a single emotional response as uniquely at stake when the concept applies in a given case. Of course, the McDowell-Wiggins view could follow Dancy here, but they can do so only at a rather high price. Denying that there is a unique emotional reaction at stake in the application of thick concept would require such a view to give us an account 49 This, of course, is not to say that such responses will not be tied down to at least one specific evaluative concept: fear and FEARFUL will bear this sort of treatment, for instance. 50 (Dancy 1995) 98 how the concept’s extension could nonetheless be affected by evaluation, but do so without appealing to the sorts of emotional resources that the spirit of such an approach was originally framed around. In his attempt at being more realistic in how we should treat the relationship between various emotional responses and thick concepts, Dancy loses the ability to clearly state what it is that makes cruel acts cruel, while retaining the idea that certain affective responses are central to that determination. At the very least, Dancy’s proposal would require more work before it could be seen as a reasonable workaround for the McDowell-Wiggins view. Section 2.6: Concluding Remarks Even if we are apt to see past the problem of circularity, and accept that, at bottom, there will not be an complete, reductive analysis, but rather an explication of thick concepts that some group is already familiar with and lives by, and even if we were to see past the problem of there being no straightforward or systematic connection between the thick concepts that an individual possesses and their emotional lives without making such a connection seem simply ad hoc in construction, there is still the question of how we should understand the connection between thick concepts and such feelings, attitudes, and emotions. Here there is an assumption shared by Allan Gibbard and Simon Blackburn (and to a lesser extent perhaps by John McDowell and David Wiggins) that is worth nothing. While Gibbard and Blackburn differ in whether this connection is logically separable from the content of the relevant thick concept, both accept the idea that how we should account for the judgments involving thick concepts is by relating them in a rather straightforward way to the expression of some attitude or emotional 99 reaction. Gibbard accepts it, and agrees that there are thick concepts; Blackburn accepts it as a condition for their being thick concepts, and argues from it that there are none. I think this assumption is a mistake. It seems clear that there can be particular cases where a person judges some act to be cruel or courageous, and yet has no attendant feelings towards that act, nor thinks that any particular emotional response is merited or warranted on his part. This, however, does not mean that such a judgment is devoid of evaluative content, as we would be hard pressed to make sense of such a judgment unless we were willing to see it as an expression of an evaluation, as an expression of this person’s moral outlook. The connections that are to be found between a given thick concept and the emotional life of the person who possesses it, are, to borrow Bernard Williams felicitous phrase, “below the level of the speech act,” 51 or below the level of this or that judgment which deploys the concept. If there is a connection to be found it will be in the general emotional and affective dispositions that one has, and how the use of thick concepts helps to both express those dispositions from time to time, and also to make sense of them by providing reasons for feeling certain things on certain occasions. And so, while the above sort of sentimentalist treatment might be well suited for more straightforwardly affective concepts, it would seem that, for a large number of thick concepts, this sort of approach misplaces the role and emphasis that emotional responses have in determining the content of a thick concept. None of this is to say, however, that, in giving some developmental story, about how a certain group came to have certain thick concepts, one could leave out the emotional life of that group as unimportant or tertiary to the development of those 51 (Williams 1973, pp. 216-17) 100 concepts. Rather, it is only to say that, if we start with the thought that these concepts must bear some clear and perspicuous connection to the emotional lives of those who use them, then we are likely to either think that the connection is wholly contingent (e.g., Blackburn), or else come to make it a requirement of our account as opposed to a discovery (e.g., Gibbard). With most thick concepts (e.g., CRUEL, KIND, GENEROUS, HONEST, SELFISH, SNOBBISH, SHABBY, and so on), we will be better served by seeing their connection to our emotional lives in a more general (i.e., non-specific, non-systematic) way, rather than either of these unsatisfying alternatives. 52 The sentimentalist treatment of thick concepts was generally defined as holding that what was at stake in the application of a given thick concept was a specific emotional response (or as such a response being warranted or merited). In its place, we might see some thick concepts as less directly tied to particular emotions, but whose influence is felt in one’s coming to understand why certain thick concepts have the evaluative standards that they have, and what is at stake in accepting those standards. Certainly there will be a range of emotional responses that will be considered the normative upshot in the application of a given thick concept (consider the various sorts of emotional responses that would be appropriate in coming to find a certain act to be cruel), but we can hold onto this thought without having to consider there to be any attitude within this range that is taken to be warranted or merited as a constitutive part in determining 52 In all fairness to Wiggins, his approach diverges from McDowell’s in an important way here. Wiggins gives clearer credence to the idea that, for the larger body of thick concepts an evaluative community might have, there must be some genealogical story to be told as to how their emotional responses, practices, and interests, ultimately come to shape these concepts. See (Wiggins 1998a, pp. 196-97). Such a story starts with certain basic responses which evolve in a progressive developmental loop that refines both the responses as well as the features make those responses appropriate. 101 whether the concept applies. In the next chapter, I will develop an account of thick concepts, starting from a schematic view of Allan Gibbard’s approach, which shows how this is possible. 102 CHAPTER 3: The Moral Psychology of Thick Concepts Section 3.0: Introduction In Chapter 1, I brought up the notion of a family resemblance concept as a way of motivating a ‘partners in crime’ argument about thick concepts. Tables and games can be variously constituted. The elements that comprise items that fall under these concepts can differ dramatically from one another, making it difficult (if not impossible) to provide a concise and intelligible understanding of the concepts TABLE and GAME as denoting some unique property. In many ways, the sorts of acts that thick concepts pick out are no different: murders and acts of generosity can be diverse and fail to share any (or any observably recognizable) features between them. And so, just as there is no distinct observable feature that all and only tables share with one another, above their having the right configuration of features to fall under the concept TABLE, there is no distinct observable feature that all and only acts of murder share with one another, above their having the right configuration of features to fall under the concept MURDER. Where component views err is in thinking that we are responsive to some distinct property in grouping things together under a thick concept, when in fact the process of grouping is complex and driven by the interests that underlie that concept. The point of invoking the comparison was to show that, in this way, thick concepts are no different from other sorts of concepts, where the organization of features needed for falling under the concept is something determined by certain needs and interests we have. In this way, the sort of objects that the concept TABLE picks out is no less (and no more) “shapeless” 1 with 1 (Blackburn 1981, p. 167) 103 respect to other properties than concepts like MURDER, GENEROUS and CRUEL. And so, if shapelessness is taken to be a barrier to the claim that one can be correct or incorrect in the application of concepts like MURDER, GENEROSITY and CRUEL, and further, a barrier to the claim that one can (when circumstances are favorable) know whether such concepts apply, then it must also be seen as a barrier in cases that no one would reasonably deny, namely that we can be correct or incorrect that some object is a table, or that we can know whether certain objects are tables. The idea that thick concepts are shapeless with respect to the purely descriptive or nonevaluative features of the world motivated some (notably McDowell, Wiggins, and Gibbard) to close what was seen as an explanatory gap by appealing to certain specific emotional reactions. This idea was given its clearest expression in Gibbard’s account of LEWD. There is no neutral property that can be associated with an act of lewdness; lewdness lacks a purely descriptive shape; it fails to involve a clearly nonevaluative pattern. At best, such a concept applies to agents only if their sexual display is of a certain sort. What closes the gap, and explains whether the concept applies, then, is the fact that some emotional reaction, unique to this concept, is warranted. It is by appeal to the cohesion of such a reaction that we can see thick concepts as having an intelligible use from context to context. On such a view, we are committed to the idea that there are emotional reactions unique to each thick concept, where the concept now plays a regulatory role with respect to that reaction. While I argued that such an approach might work for some concepts, namely affective concepts where we have a clear (and, to some degree at least, conceptually prior) understanding of the emotional reaction at stake, this 104 approach seemed difficult to generalize to most other thick concepts. With no other reason for making such a postulation other than the need to render the position consistent with a sentimentalist view of such concepts, these accounts come across as ad hoc. I will not rehearse the problems I noted for this view in the last chapter. However, it is worth noting that such an approach seems prompted by the concern that, in saying that the groupings made by a thick concept are neutrally or nonevaluatively shapeless, what is needed is an appeal to some additional evaluative consideration (in this case a warranted emotional response) to close an explanatory gap. Since the neutral features or properties associated with a thick concept cannot explain why it applies from context to context, there must be some further (evaluative) feature which does the job. In this chapter, I want to critically examine this approach, find fault with it, and attempt to move past it. I want to do this by looking again at the interests and concerns that underlie the development of such concepts as a way of showing how such concepts come to exhibit an intelligible pattern of use. Of course, the discussions in Chapters 1 and 2 were largely critical. This is not to say, however, that the views considered were without important insights. In this chapter, I will attempt to draw these insights out. As I will argue below, understanding why a given thick ethical concept applies in a paradigmatic case requires understanding how the action in question relates to certain independent concerns and interests. Such concerns and interests not only help to explain why there is such a concept for a community, but they provide guiding insights into how such concepts are to be applied in novel cases. In many cases, the concerns and interests that underwrite a thick concept are directed 105 towards encouraging agents to and discouraging agents from acting for certain sorts of reasons. It is because of this that many thick concepts have, what I called in Chapter 1, characteristic reasons associated with them. Part of the goal in this chapter is to substantiate that claim. It is important, however, to note up front one caveat to this: since we do not confront an agent’s reasons for acting directly, that is, since we infer what an agent’s reasons for acting are from the observable character of the agent’s act, along with the context in which it occurs, we cannot treat such reasons as functioning in a thick concept as independent neutral constraints on its application. They are too closely wedded to how we interpret an agent’s behavior for us to see them as simply neutral features that we respond to in applying a thick concept to something the agent has done. I hope to make this point plain as the discussion progresses. Before getting to this, however, let us return first to some of the critical remarks made in the first two chapters. This will help us to recall some of the relevant terrain that has already been covered, and also frame the discussion for the view I will advance. To begin, then, I would like to retrace some central points by using Allan Gibbard’s model of thick concepts as a foil. Section 3.1: The Basic Structure of Thick Concepts Recall that the starting point in Chapter 1 was McDowell’s Worry: namely that an outsider, to a community that uses some thick ethical concept, could not come to understand that concept without appreciating how the use of that concept was connected to that community’s ethical outlook. Voicing a concern that he takes to be allied to McDowell’s worry, Allan Gibbard notes that, when it comes to thick ethical concepts, there is simply “not enough” descriptive meaning available so as to understand them in 106 the way that a basic-CV theorist wants to envision them. 2 Of course, it is quite obscure what is meant by ‘descriptive meaning’, and Gibbard does not explain what he means by this, but the thought is that, I think, there is no neutral property that can be associated with a thick concept, such that the concept’s application to an act or individual is driven exclusively by the presence of that property. Whatever neutral, nonevaluative features one takes to be associated with a thick concept, these alone do not explain why that concept applies in a given context. Some, however, fail to see the problem that Gibbard is pointing to. Dancy, for one, thinks that any amount of ‘descriptive meaning’ is enough to count as a separate component in the analysis of a thick term: I just don’t understand the notion of ‘not enough descriptive meaning.’ Any amount of descriptive meaning is enough descriptive meaning to be the whole meaning of some term, even if it does not make for a very interesting or useful one. If a term has descriptive meaning, its descriptive ‘aspect’ is capable of standing alone as a neutrally descriptive concept. (Dancy 1995, p. 275) 3 Though Dancy balks at the suggestion, I think Gibbard is right to highlight it, as this is simply an expressivist (or noncognitivist) take on McDowell’s Worry. If the content of a thick concept is radically underdetermined by those non-evaluative features that are associated with it, whatever they turn out to be, and further, if we accept that (in at least certain cases) the application of the concept can be explained (i.e., it can rightfully or correctly apply to certain actions, individuals, and situations), then there simply is no determinate set of such features to construct a neutral component out of. In this sense, Gibbard’s descriptive lacuna fits nicely with McDowell’s Worry: it is because thick 2 (Gibbard 1992, p. 278) 3 To be fair, it is not clear what Dancy is disagreeing with here. See (Dancy 1995, p. 276), where Dancy admits that a thick concept “will have less descriptive meaning, somehow.” Dancy never explains what he means by ‘descriptive meaning’ either. 107 concepts are nonevaluatively shapeless that their application in a certain context seems to leave a hole in the resulting explanatory picture that cannot be filled in by just adding on more non-evaluative detail. Gibbard’s own positive account seeks to overcome this problem by treating the descriptive component relatively indeterminate, and taking the warranting of a specific emotional reaction to fill in the resulting gap. In the previous chapter, I offered several considerations as to why such an approach is problematic. However, the manner in which he tries to accomplish this leads, I think, to a fresh insight about how we should understand thick concepts. This relates to the need for involving additional ethical considerations into our account of the concept, in particular ethical standards (or criteria) and certain related ethical interests which give a point to adopting such standards. This insight, I think, is detachable from any commitment to Gibbard’s particular metaethical position, expressivism. To be sure, my goal is not to argue against expressivism. Rather, it is to show how the particular expressive element Gibbard adds is simply redundant and hence eliminable, given the other aspects of his view. Isolating the core insights of his account, and separating out those aspects that are inessential, will help us to appreciate the basic structure of a thick concept. To begin, then, let’s revisit the details of Gibbard’s account as a starting point for reverse-engineering (what I believe is) the correct view. What plays a crucial explanatory role in Gibbard’s account, and explains why a thick concept applies to some cases and not others, is the warranting of some specific emotional reaction. This emotional reaction is one unique to the thick concept in question, as something that is 108 generally felt when confronting acts and situations which would fall under the concept. A judgment of warrant for such this specific reaction, then, closes the explanatory gap left open by whatever general, non-evaluative features are associated with the concept: when the attitude is warranted, the concept applies. Recall that the specifics for Gibbard’s account centered on the concept LEWD: φ is lewd IFF L-censoriousness towards the agent is warranted, for passing beyond those limits on sexual displays, such that: (i) in general, passing beyond those limits warrants feelings of L-censoriousness towards the person doing so, and (ii) this holds either on no further grounds or on grounds that apply specially to sexual displays as sexual displays. 4 The concept LEWD applies only to sexual displays (the non-evaluative component), but not all sexual displays are lewd (which is why this component is indeterminate). Some sexual displays can be harassing or illegal or humiliating and still not count as lewd. 5 Touching someone in a certain way might count as a form of sexual harassment, even though it does not also count as a lewd act. What is specific to those sexual displays that are lewd is that the emotional response of L- (or lewd-) censoriousness is warranted towards them. 6 Let us first focus on how Gibbard incorporates these evaluative considerations into his account of the concept. Consider the role of L-censoriousness. Not only does 4 Gibbard’s analysis actually deploys the intentional notion ‘means this’. See (Gibbard 1992, p. 280). I am leaving it out as it adds nothing to the substance of his account, and (in my opinion) only undermines the plausibility of his view. I think Gibbard’s position is better read as an account of the concept of lewdness rather than the meaning of the term ‘lewd’ as it is used by various individuals. 5 Indeed some sexual displays need not be taken as negative at all. Setting a limit on sexual displays does not rule out there being sexual displays that are perfectly appropriate (e.g., between a committed partners). 6 And here it is perhaps important to note that for Gibbard judgments about warrant are not properly considered true or false, but rather express one’s acceptance of certain norms. 109 this reaction close an explanatory gap with respect to when LEWD applies, its role also reflects how, for Gibbard, judgments of lewdness are actually specific evaluations: they are specific evaluations because they involve the implicit judgment that L-censoriousness is warranted towards the agent for her sexual display. What is interesting for our purposes, however, is that the considerations that could warrant such feelings of L- censoriousness are fairly specific. L-censoriousness cannot be warranted for just any reason. It cannot even be warranted for reasons that pertain only to sexual displays. In order for L-censoriousness to fulfill its explanatory role, the reaction must be warranted on the grounds that the sexual display in question passed beyond certain limits imposed on sexual displays. Given this, there are two key points to consider. First, the imposition of limits presupposes that there are criteria which determine what those limits are, even if they are themselves fuzzy and imprecise. Imagine that there really were no such criteria for these limits, but that the application of the concept LEWD depended on an act still ‘passing beyond’ them. In such a case, then, a competent speaker would be allowed to make a sincere judgment that some sexual display was lewd, but be unable to articulate what it was about this particular sexual display that constituted its passing beyond such limits. Of course, Gibbard might be right to suggest that what is actually expressed in uttering the statement ‘that is lewd’ only presupposes that there are such limits, but if we are attempting to understand the concept LEWD itself, then certainly coherent and competent use of the concept would require a sense (however vague or imprecise) of what its criteria amounted to. If this is so, then the intelligibility of the concept would rest not simply on their being limits, but also on some understanding as to 110 what those limits are. But if the presupposition of criteria is necessary to make sense of coherent and competent use of the concept, then I think we have a good reason to simply drop the reference to specific evaluative reactions, like those of L-censoriousness, and the judgments warranting them, from the account altogether. The reason for this is that such criteria have a certain role in the community at large: the point of classifying certain acts as lewd is to discourage people from acting in lewd ways. If this is right, then we might generalize it to say that: An act falls under a positive (negative) thick concept just in case that act merits or warrants a positive (negative) evaluation in virtue of its falling under that thick concept. In short, the fact that such a concept applies provides its users with a reason to evaluate the act it applies to in some determinate fashion, in virtue of its applying. In other words, there is a certain ethical (or evaluative) point in developing a concept that has the very sort of criteria it has. If this is right, then there is no need to posit some evaluative reaction that is specific to that very concept. In dropping emotional responses like L-censoriousness from the account, however, nothing is actually lost in the process. We simply replace the idea of there being a specific emotional response warranted towards sexual displays, as explanatory, with the idea that some sexual displays meet the criteria for being lewd, and, in virtue of this, transgress certain ethical standards. Some sexual displays go beyond these standards, and some do not, and the difference between them is an evaluative difference that is connected to the fact that not every sexual display counts as lewd. It is connected to this point in virtue of the fact that the criteria developed are developed for a specific regulatory function. Indeed, with 111 this correction in mind, I think Gibbard’s analysis of LEWD is largely correct. This is the general structure of thick concepts that I want to put forward. This gives us the freedom to accept an important insight—namely, the imposition of certain independent ethical considerations as a means of characterizing the content of a thick concept—while rejecting that part of his analysis that seems straightforwardly motivated by a particular metaethical commitment. How, then, should we understand this notion? I will come back to this question in detail in Chapter 5, but one way to give a gloss of the idea would be to see the use of a thick concept as bound up with further ethical interests, interests that are geared towards encouraging or discouraging certain forms of behavior. Indeed, this seems to coincide with how concepts like LEWD (as well as concepts like CHASTITY and BAWDY) are straightforwardly ascribed to sexually displays. While lewdness might involve a certain sort of sexual display, what is distinctive of lewd acts is not something distinctive in the sexual display itself, but in how that display can reflect a failure on the agent’s part to exhibit a proper degree of sexual modesty. Lewd acts, we might say, are taken to defy certain standards of decency that are closely related to certain conceptions of self-respect and personal dignity, insofar as these notions are particularly related to the areas of sexuality and sexual expression. And so, what unifies various forms of sexual display as falling under the concept LEWD can be found in the interests that one has in upholding such conceptions of self-respect and personal dignity. To this extent judging that a piece of behavior is lewd is connected with additional thoughts, such as that such behavior is shameful, dishonorable, or disgraceful: as such considerations all relate to the idea that the individual has failed to 112 live up to certain standards of respectable behavior (insofar as these standards govern sexual displays). What gives content to the connection between lewdness and these others ethical notions is that their tie to a conception of sexuality and sexual expression is mediated a possession of a certain conception of self-respect or dignity. In terms of the deliberations that an agent goes through, in reaching the conclusion that a certain piece of behavior (of another or oneself) is lewd, we cannot make sense of the concept unless such conceptions are taken to figure in the background of those deliberations. To the degree, then, that what lewdness is relates to a flouting of standards of decency, self-respect and personal dignity, the question of whether a given act is lewd will depend on the content of these further considerations. In making this claim it is of course important to highlight the ethical (or evaluative) point of having a concept like lewd. Such a concept is not developed simply to discriminate some types of sexual displays from others: it is not, as Bernard Williams noted, “a device for dividing up in a rather strange way certain neutral features of the world.” 7 Classifying acts as lewd has an ethical role for the group that uses it. Its ethical role in the community is to discourage such acts. But noting that this concept has such role is likely to be misunderstood (or superficially characterized) unless we see this role as itself motivated by certain ethical considerations: namely those that relate to self- respect and personal dignity. These considerations provide the concept with its point: they are the reason why such a concept was developed and they serve as an explanation as to why it has the ethical role that it does. Identifying the ethical point of a thick concept intimately involves an appreciation of the ethical interests that underlie not only 7 (Williams 1985, p. 142) 113 its having the regulatory role that it does in the community, but also the manner in which it groups certain acts together as falling under one and the same concept. This conclusion, however, is not limited to an account of lewdness. It is a general mark, in my view, of what makes a thick concept thick: we cannot make sense of the content of the deliberations of those who use such concepts, both why they focus on the situational features that they do, but also how they focus on them, in arriving at their conclusion, unless we are able to see their use of the concept as informed by their having certain ethical interests that provide them with a reason for using that concept in that context. 8 In order to fill out the content of one’s deliberations for judging a certain act courageous, we must appreciate the judger’s view of the situation as one informed by a view that certain things are worth risking one’s safety and livelihood over. The point of using the concept COURAGEOUS is to classify actions on the basis of such an interest. With judgments of kindness, we must appreciate how acts are classified as a means of encouraging direct and unmediated responses to help and aid those in need, and so on. The point of having a concept like KIND is to group actions together that manifest such unmediated responses. But, in neither case, is the point of making such groupings bound up with simply identifying acts that possess these qualities: rather, it is bound up with wanting to encourage people to behave in these ways. Thick concepts, then, are nested in complex web of ethical standards and ethical interests, and because of this cannot be 8 This is not to say, however, that every judgment using a thick concept will involve a conscious deliberation. Some judgments might very well be phenomenologically spontaneous or intuitive. Even in such cases, however, I am supposing that there will be a need for implying a certain form of deliberation in the making of the judgment. 114 pried apart from them, on pain of either (a) misrepresenting the deliberations that one goes through in order to apply these concepts or else (b) making them entirely opaque. A further example might be helpful here. Consider again the concept of USURY. In making sense of a judgment using this concept, one would have to see the relevant deliberations in applying this concept as involving a standard for determining when certain interest rates for lending money were excessive, and how such practices (at the time in which this concept was actually used) were essentially taken to be a product of greed and manipulation. 9 Given such a view of these practices, it is no wonder that a term was coined to describe such practices in this way. We, however, no longer think of lending money as essentially a product of greedy and manipulative behavior. This is not because we have simply changed our attitudes towards one and the same fiscal practice. Rather, it is because the practice of lending money is now a quite different one than it once was. Prior to the onset of modern fiscal notions like investment and capital development, the borrowing of money was aimed at providing for the basics of human subsistence. Now, however, the aim of lending money is exclusively directed towards the goal of wealth creation. In short, the entire background behind the practice of lending money has, therefore, dramatically changed. Because of this, it would be wrong to think that USURY is simply a different concept for these more contemporary lending practices. Without the addition of the ethical considerations that drove the application of the 9 The idea that the lending of money with interest per se always counts as a form of usury is one that we can scarcely make intelligible today. But for a rather convincing gloss of the change in perspective on this practice from the middle ages to the modern world, see (Duncker 1939, p. 41) and his observation that in the medieval world loans were taken out when needs arose for basic survival. In the modern world, loans are always for investment and profit, and so the idea of charging interest is tied to ‘getting a share of the profit’. 115 concept of USURY in past communities, including the ethical view of what the lending of money involved (in particular the considerations surrounding the reasons why people, at these times, borrowed, money), we would simply fail to understand the point and role of such a concept. 10 If this is correct, then to make sense of a thick concept, we have to see how that such a concept is shaped by certain ethical interests and standards that are intimately connected to it. Such interests and concerns constitute a thick concept’s evaluative or ethical rationale. 11 They provide a rationale by constituting reasons for using such a concept as a means of classifying acts and individuals. The point of using a particular thick concept at all is that the concept classifies (indeed as been developed as a means to classify) in such a way so as to serve those interests. Understanding how such a concept relates to these interests is essential for understanding ‘how to go on’ in applying it. The interests that underlie the concept help users to structure situations in ways that enable them to determine whether the concept applies or not: these interests help to frame situations in such a way so as to make certain features of those situations relevant to the application of the concept. Given this, making sense of how such interests relate to the concept is essential for understanding the concept itself. 12 That a thick concept involves 10 For a view that seems to presuppose such a naïve view of the concept of usury, see (Copp 2001). I suspect that it is a failure to see this concept as one answering to different interests from those that drive our contemporary descriptions of the practice of lending money that motivates Copp’s view that such a concept is capable of being split into descriptive and evaluative components. 11 Gibbard also uses the notion of a rationale, in offering his account of thick concepts, but this is to highlight how the use of such concepts provides a rationale for certain feelings. (Gibbard 1992, p. 280). For Gibbard, the thick concept is the rationale behind feeling certain things. On my view, the notion of a rationale is bound up with applying the concept in the first place. 116 an ethical rationale helps to see how its use is connected with (as McDowell put it) a group’s “conception of how to live.” 13 McDowell took this to be an essential element in how an outsider could come to understand an alien thick concept (or some number of them). And now we can see why. Without an appreciation of a thick concept’s ethical rationale one would be unable to understand the point of using such a concept, that is, she would be unable to grasp how to frame situations in a way that would bring out reasons as to why classifying agents and behavior in that way was thought to be appropriate. The picture of thick concepts that I have been developing focuses on the idea that thick concepts classify. This I take to be an uncontroversial point about such concepts. Because of this, we might also say that thick concepts are ‘descriptive’, as long as this notion is not taken to imply anything further beyond the point that these concepts classify. However, although they classify acts and individuals, the manner in which they do this is rests heavily on certain ethical interests. Such interests seek to encourage or discourage certain forms of behavior by focusing on certain features of actions that are to be encouraged or discouraged. Because these interests guide how a thick concept classifies acts and individuals, understanding such a concept is impossible without a grasp of what these interests are: even if one were able to have a complete extensional grasp of the concept. Even if another concept exhibited the very same pattern as did the thick concept under investigation, one would still be at a loss as to what the different reasons were for making such classifications if one did not grasp the interests that 12 This also connects with what I think is a fundamental misreading of McDowell’s position, and how the criterion for understanding should be on how to justify the application of a thick concept, as opposed to being able to predict future usage. On this, see Chapter 2, Section 2.2. 13 (McDowell 1998f, p. 71) 117 provided a point to classifying things in that way. Since such classifications occur from the standpoint of certain interests, we might say that, when it comes to thick concepts, judgments deploying them constitute descriptions, but that these are descriptions from a certain ethical point of view. 14 And since the interests that drive the manner in which a thick concept classifies are ethical interests, there is no further step or component needed in the concept in order to see such a manner of classification as expressive of a certain group’s values. In this sense, thick concepts describe the world as it is understood from the standpoint of those very interests. Section 3.2: Thick Concepts and Their Use Thus far we have been discussing how thick concepts relate to certain ethical interests that guide how these concepts classify or group together various acts and individuals. Understanding how such a concept relates to these interests is essential for understanding the concept itself. But understanding this is not enough for one to be able to go on and use the concept for oneself. The reason for this is that one might not identify with or accept the interests that give a thick concept its point. In this way, accepting (or identifying with) a thick concept’s rationale can be seen as a criterion for being able to sincerely use that concept. It is, however, critical to see such a notion of ‘use’ as involving more than the mere ability to make certain statements, as this is something that one might gain by coming to understand the concept and its related rationale. For those who accept this rationale the use of the concept must be seen as having ramifying consequences that extend beyond the possession of a certain linguistic 14 Here I am inclined to agree with Julius Kovesi, when he notes that “moral notions do not evaluate the world of description; we evaluate the world by help of descriptive notions. Moral notions describe the world of evaluation.” (Kovesi 2004 [1967], p. 115) 118 capacity. An outsider, one who did not identify with the concept’s rationale, might still come to understand such a concept: she can develop a grasp when and why it applies, and, in doing so, she can appreciate the interconnections that this concept has to other evaluative concepts and ethical interests. This alone might enable her to grasp how such a concept could be applied in novel contexts, and enable her to enter imaginatively into disagreements with native users over whether a particular act fell under the concept or not. In all of these ways, it makes sense to say that a person could ‘use’ this concept. In another sense, however, she cannot ‘use’ such concept. Because she does not accept the interests that underlie the concept, she cannot use the concept to give voice to ethical judgments that she would endorse. Consider as an example, the concept CAD, mentioned in Chapter 1. The concept CAD is one that we no longer use to admonish men with. This concept is connected with a value system and a way of viewing women, as special objects of concern and respect in virtue of their being women, that we no longer accept or share. When this concept was used, however, a man was considered a cad just in case he treated a woman (or a woman of a certain class or social standing) in a manner that was at cross-purposes with their being special objects of concern and respect: more specifically if he acted in ways that would endanger her respectability and sense of self- worth. The interest that such a concept is bound up with is that of discouraging such behavior: the reason why certain pieces of behavior and certain men were classified as cads is to highlight them and their behavior as something that is counter to the values that are already held by the community. Now, the reason why I cannot use this concept is not that I do not think women are deserving of concern and respect; they are. What I do not 119 accept is that they are deserving of such concern and respect because they are women (or women of a certain class or social standing); rather I believe that they are deserving of such respect because they are people or human beings or rational creatures or individuals capable of feeling embarrassment or suffering. That I feel they are deserving of concern and respect for these reasons does not make me a licensed user of the concept CAD, for in order to use that concept (with right) I would have to accept that the reason why they are deserving of concern and respect has something essentially to do with the fact that they are women. Since the concept CAD is intimately tied to these sorts of interests, and, because of this, it is a concept that I am not entitled to use. I am barred from sincerely using it in just this sense: my judgment would implicate me as having (or sharing) interests that I do not have, and, because the concept is one that is designed to meet such interests, I cannot (with right) use the concept to issue in a judgment that is (in a basic sense) my own. Because of this, there remains a sense in which such a concept is not mine, even if I develop a reliable ability to spot correct instances in which the concept applies. My lack of identification with the concept CAD’s underlying rationale entails an inability for me to issue judgments using it in propria persona—that is, to actually judge acts and situations by it in a way that would be an expression of my values, concerns, interests, character, and so on. This idea of speaking for oneself (speaking in propria persona) involves the idea of personal accountability, but it also involves something more. In making a serious and sincere evaluative judgment a person not only holds others accountable for what they have said, such a person is prepared to hold themselves accountable to such judgments (both by themselves and others). In speaking for yourself, 120 then, you are prepared not simply to defend, justify, or excuse, what you have said to others, but also, you are prepared to treat responses to what you have said as precursors to changes in the relationships you have with the respondents. In this way, it is not so much that the outsider cannot assert certain things using the relevant thick concept; she can, at least to the rather minimal extent that she understands what doing so would require. Rather, it is that, if she did go on and assert a statement using an alien thick concept, there is a deeper sense in which she could not avow what it is that she asserted. 15 It is, in this way, that I cannot use the concept CAD with right: it is not the sort of concept whose use I would hold myself or others accountable for. The basic structure of a thick concept is roughly, then, the following. A thick concept is one that presupposes a specific evaluative rationale, where this rationale consists of certain interests and concerns which provide a point to classifying things in the manner that a thick concept does. Because of this, one can only possess a thick concept (and use it to judge things in propria persona) if the interests and concerns that underlie the concept are their interests and concerns. As such, I agree with Williams, when he notes that being able to use a thick concept is largely (though not exclusively) a matter of belonging to a certain culture. 16 Williams’ idea I take it is that membership in a culture is simply one way of denoting the possession of different ethical interests: possessing such interests is largely a function of being raised up into the community, as 15 For an excellent discussion on avowals, how avowals differ from mere assertions, and how they bear constitutive connections to both first-person authority and aspects of rational agency, see (Moran 2001, pp. 89-94). 16 (Williams 1985, pp. 144-45) 121 the process of raising a child to accept certain interests is one of the natural ways in which members of a community propagate their culture from one generation to the next. This helps to explain McDowell’s Worry by locating the outsider’s ability to make sense of a thick concept in her ability to understand how that concept relates to certain ethical interests endorsed by the community that uses the concept. This does not entail, as McDowell rightly noted, that one must share those interests in order to make sense of the concept. As already noted above, we are capable of understanding thick concepts without accepting or endorsing the interests that underlie them. 17 (Cf. the example of CAD above.) What is needed is for the outsider to understand how the native’s interests and concerns render (or can render) their use of such concepts reasonable in certain situations (or towards certain actions). Grasp of the concept, then, involves putting oneself imaginatively in the position of the native user, and be able to generate an appreciation for the different ways in which the ascription of such a concept would be driven by its evaluative rationale. The basic underlying structure to thick ethical concepts, then, depends on seeing the actions or individuals that fall under them as already standing in some relation to various independent values, interests or concerns. The disvalue associated with cads or caddish behavior depends upon the antecedent value in finding women to be special objects of concern and respect. The value associated with acts of courage depends upon the antecedent value in holding that certain things are worth risking one’s life and safety 17 For a position that seems to advocate the idea that, in order to make sense of a social practice, one had to actually join it, see (Winch 1958), see esp. Ch.III, Sections 5 and 6. Winch writes that the relation of the observer “to the performers of religious activity cannot be just that of observer to observed. It must be analogous to the participation of the natural scientist with his fellow-workers in the activities of scientific investigation.” (Winch 1958, pp. 87-88) 122 over. The disvalue we attach to cruelty depends upon the antecedent value we attach to people being innocent of wrong doing, the disvalue we attach to undeserved suffering, and so on. Acts and individuals that fall under a thick concept are already conceived as standing in certain relationships to such independent interests and concerns. Given this brief articulation, something more needs to be said about how we should understand the nature of the actions which are thought to further or fail to further such independent interests and concerns. The crucial connection that I will argue for is that thick concepts (or a large number of them) apply when they do because an agent has acted for certain characteristic reasons. Such reasons make a clear difference in how we can describe certain acts as falling under a thick concept, but they are also the sorts of considerations that ground our view that such acts either further or undermine the concept’s related interest. It is this point that the next sections will attempt to develop. This will yield some important insights not only in how we should understand various thick concepts, but also in how we should understand their capacity to provide users with a certain sort of ethical knowledge. Section 3.3: Thick Concepts and Characteristic Reasons We can begin with the uncontroversial thought that an action can be described in a variety of ways. One and the same piece of behavior can be described in numerous ways. 18 When it comes to thick concepts, however, or at least a good number of them, the application of the concept to some act is sensitive to what the agent’s reasons for 18 (Anscombe 1957) 123 acting in fact were. 19 Consider two brief examples. First, suppose A donates to B a sum of money, and that B does not deserve to receive this sum from A. Whether A’s donation amounts to an act of generosity, depends greatly on A’s reasons for the donation. If A gave the money to B simply because B had a need for it, then this would be compatible with considering A’s act a generous one. But if A gave the money to B, simply because doing so would impress C, such a reason would not be compatible with considering A’s act generous. Secondly, suppose A kills B, but A’s reason for doing so what because he believed this was the only way of protecting himself from B. Such a reason would be incompatible with considering A to have murdered B. Had A killed B simply to be rid of him, and not because he was a threat to A’s life, then, ceteris paribus, this would be compatible describing what A had done as murdering B. Because of this, we can say that thick concepts 20 have characteristic reasons associated with them: whether a thick concept applies or not to an agent’s action depends on what the agent’s reasons for acting were. Such reasons, in virtue of providing a suitably strong constraint on the application of a thick concept, constitute a clear difference in how an action can, and indeed sometimes must, be described. For instance, if A’s reason for killing B was to remove B as a business competitor, describing A’s act 19 There is a good deal of disagreement over the nature of practical reasons, but for this discussion I will assume, without argument, that the correct understanding of them is what is often called ‘internalism about reasons’, or the view that a reason is a type of explanation that can rationalize an agent’s behavior, given his aims, interests, and motives. See (Williams 1981a). For Williams, A has a reason R to φ just in case there is a sound deliberative route between R and some motivational element (conceived broadly) that A has, which would be furthered by φ-ing. On this view, R is some consideration that an agent could act from, such that, in doing so, R would be an explanation for A as to why he acted. As I understand Williams, one’s belief that they have a reason to φ can count as a motivational element (i.e., an element in what Williams calls their ‘subjective motivational set’), and so one could act from such a belief without the need of some further motivational element. For an excellent discussion of Williams’ views on internal reasons, see (Skorupski 2007a). 20 From here on, I will drop the qualification “for many,” but it should be taken as implied. 124 as say ‘a move to increase his market share’ would be a moral or ethical misdescription from our ethical point of view. It would be a description that attempted to omit or neglect facts which, from our point of view, license only one description of what has happened, namely that A murdered B. Increasing a share of some market is not a morally justified reason for killing someone, and because of this, when someone kills for such a reason, there is only one ethically appropriate description. Similarly, such misdescriptions can occur by applying a thick concept to something the agent has done when the appropriate reasons for acting are absent. Of course, as I argued in Chapter 1, our ability to ascribe reasons to an agent is not wholly independent of how we evaluate their overt behavior. In general, the reasons an agent acts from cannot be determined in complete abstraction from a wide range of facts dealing with the context in which the action occurred (including general facts about the agent’s psychology). That said, I want to highlight a contrast between different ways of describing an act, and how certain acts would be misdescribed in applying a thick concept to them, if they failed to be done for the right sorts of reasons. What will be important for our purposes, however, is to show how such reasons connect with the ethical rationale that underlies the thick concept in question. In each case, the reasons characteristically associated with a thick concept are those that a group seeks to either encourage or discourage agents from acting on. (a) Reasons Characteristic of (Some) Thick Concepts Acts falling under a thick ethical concept are identified in part by the reasons an agent acts from. A strong reason for thinking this is that our coming to find that an agent 125 had acted for reasons different from what we initially thought is typically a decisive ground for retracting a judgment using a thick concept to evaluate that agent’s action (e.g., an action cannot be kind if the agent acted because he thought it was his duty to do so). It is crucial to note, however, that not all ways of characterizing an act are like this. Not every concept that can be used to describe a type of act is constrained by considerations as to whether the agent indeed acted for the right sort of reason. Indeed, for many such concepts, the concept can correctly apply to what the agent did irrespective of what reasons the agent had for acting. Consider the following scenario: Smike trips and falls in the lobby. Nicholas sees this and immediately reaches over, gives him his hand, and pulls Smike up. He then dusts him off, pats him on the back and asks if he is alright. This description leaves out any appeal to Nicholas’ reasons for doing what he did (although we can assume that Nicholas’ behavior was intentional, and to this extent, he had some reason for so acting). Yet, despite this lack of information, we can still describe Nicholas as having aided or assisted or helped Smike by reaching over and pulling him up. Helping, assisting, aiding, and other sorts of descriptions, can be understood to apply to an agent’s act irrespective of what reasons the agent had for acting. What Nicholas did, in helping Smike, is consistent with Nicholas’ having a whole range of different reasons for acting, some laudable, some rather unsavory. For instance, Nicholas might have pulled Smike up in order to help curry favor from Mr. Squeers, or to assuage his own guilt (because Nicholas was actually the cause of Smike’s having tripped and fallen), or to impress Madeline who is watching from a distance, or to gain a foothold on 126 Mulberry Hawk’s largesse (whom he knows is looking for a kind and decent person to become heir to his fortune). None of these are particularly laudable reasons for helping Smike. There are, however, genuinely noble reasons that would work just as well: to fulfill his Christian duty, to make good on a promise made to Smike earlier to help if he tripped, or even to maximize utility. 21 It is not that some acts wear the appearance of being helping acts, but are really not because the agent’s reasons are incompatible: Smike really was helped by what Nicholas had done, regardless of the reasons for doing so. Rather, the point is rather that for some characterizations of an agent’s act, such as helping or assisting, the agent’s reasons for acting are immaterial to whether such a characterization is correct. 22 In contrast to this, whether an act as falls under a thick ethical concept depends crucially on whether the agent acted for some reasons and not others. This is perhaps least controversial when we think of kindness (or kind acts). While Nicholas might have aided Smike with the aim of impressing Madeline, his aiding, given this reason, would not have been kind. In order for Nicholas’ act to have been genuinely kind, his reason for acting would have to have been connected more directly to Smike’s well-being. What sorts of reasons might these be? It is difficult to state in specifics what these might be, but in general, to count as a kind action, Nicholas’ reason for helping would have to be related to a concern or sensitivity to Smike’s needs in this situation: only if it was because 21 Michael Stocker gives an excellent example of a person visiting a friend in the hospital because he thought it a moral duty to do so. See (Stocker 1976, p. 462). 22 See (Stocker 1973, p. 51): “That a person does an act with good consequences is conceptually consistent with his doing that act with most any moral sort of belief, intention, motive, character trait, and so on: good, bad, or indifferent.” 127 Smike’s need for help constituted a basic or primitive 23 reason for helping, could Nicholas’ acting count as kind. Because of this, we can say that such a reason is constitutively related to or characteristic of kind acts: characterizing an act as kind cannot be prized apart from seeing the agent as acting for a reason characteristic of kindness. Aiming to relieve discomfort, pain and sadness, trying to comfort grief or heartache, to calm distress, or even to bolster confidence—where it is someone’s discomfort that provides a basic reason to relieve it, or someone’s grief that is the basic reason to be comforting, or someone’s distress that is a basic reason to be soothing—are all further cases of reasons that are constitutive of kindness (or kind acts). Though it is perhaps clearest with thick concepts like KIND and GENEROUS, I think this point holds for others as well. Reasons characteristically related to acts of COURAGE are ones that involving an aim to save, protect, or defend some independent, but threatened value, where it is the threat to such a value that provides a reason to save, protect, or defend. The content of the concept of COURAGE of course does not determine any specific reasons that an agent must act from, in order for his act to be considered courageous. Rather, the content of the concept constrains the agent’s reasons to those that would be expressive of some project, relationship, or goal that, for the agent, was both extremely valuable and threatened. Part of what made Darnay’s act, in a Tale of Two Cities, courageous was that he was trying to save Gabelle from the guillotine. It was the danger that Gabelle faced that provided Darnay with a reason for risking his own 23 This appeal to basic-ness or primitiveness is to rule out cases where one might treat Smike’s need as instrumentally related to something else. What makes Nicholas’ act kind is that Smike’s need provided Nicholas with a basic reason for acting, and not a reason for acting because this would lead to something further (such as a favorable assessment of him by Madeleine). 128 safety to save him. 24 Acting from some reasons, however, can disqualify an act from being courageous: self-interest, personal advancement, a concern for station, and similar others. Had Darnay’s reason for saving Gabelle been to gain access to Gabelle’s good graces and wealth, his act would have failed to have been genuinely courageous. This disqualification stems from the fact that his reason for acting has little to do with the value of what was in jeopardy (i.e., Gabelle) and a great deal to do with what saving, protecting, or preserving Gabelle would do for him. Just as one can help without one’s helping be an act of kindness, one can protect another at great risk without having it count as courageous. In such a case, there are a range of terms we might use to characterize Darnay’s act, terms which place no restriction on what the reasons one acted for. Saving, protecting and preserving something that is threatened can make what one does bold, daring, or even opportunistic, but not courageous. Similarly, there are reasons constitutively related to TRUTHFUL acts, namely those where the agent’s aim is, depending on the context, to be accurate and/or sincere in what she says. 25 Simply telling the truth does not make your doing so truthful, if the reason for your statement is to gain personally or hurt someone by making it. 26 Depending on what 24 This is not to say that one has to have a particularly other-regarding motive in order for their act to count as courageous. Michael Stocker gives a great example of the courageous of a drug-addict attempting to “escape from a degenerate, ‘subhuman’ state and lift himself up into a desirable, human state.” See (Stocker 1970, p. 139). 25 This, of course, follows closely with Bernard Williams’ view that the virtues underlying truthfulness are that of sincerity and accuracy. See (Williams 2002) Chapters 5 and 6. Accuracy is a virtue relating to things like describing states of affairs or recounting events, while sincerity is a virtue relating to the expression of one’s beliefs and feelings. Both, for Williams, are domains where the concept TRUTHFUL has application. 26 Consider how the statement ‘I am only being honest’ rarely, if ever, excuses what one has said. This is it is often made in contexts where the ‘true’ or ‘sincere’ statement uttered is done so in order to hurt, embarrass or offend. 129 different reasons one might have for telling the truth, one’s doing so can be self-serving, malicious, or even cowardly, but not truthful. In order to be truthful, the speaker’s aim must be (at the very least) to have said something that they took to be correct or at least genuinely believed to be so, and not simply to avoid peril, embarrass, harass, or pursue some gain. Another example can be found with the concept of a LOYAL act. Reasons characteristic of loyal acts are those where the agent’s aim is to support, preserve, or protect an important, but threatened, personal relationship. 27 One acts loyally, for example, when one’s aim is to defend a good friend, simply because they are a good friend. Friends might be supported or protected for other reasons (e.g., because it would impress some third party, or because you have great affection for them), but in acting for these other reasons such support or protection would not count as loyal. This also helps to mark a difference between kind and loyal acts—in doing something kind my reason for acting is provided by another person’s well-being, even if that person stands in some important personal relationship to me, but in doing something loyal my reason for acting is provided by the relationship itself. I can be loyal towards a friend, but in being loyal my actions (and although my actions might help the friend, advance her well-being, and so on) are not necessarily kind, because my reason for acting is stems not from a concern for their well-being, but from a concern for that relationship. Of course, there is nothing 27 Here I am in agreement with Thomas Hurka, who argues that loyalty is (in a certain sense) backward looking. It is concerned with facts about a shared history, but not just any history of a personal relationship will ground demands for loyalty. Rather, those demands depend greatly on the moral quality of that history (e.g., whether they were a good friend). See his (Hurka 2001, pp. 204-06). 130 in this difference that precludes an act from being both kind and loyal (one’s reasons might over-determine how one acts). (b) Extending the Account: Some Negative Thick Ethical Concepts The idea that thick ethical concepts have characteristic reasons is by no means limited only to positive thick ethical concepts. Negative thick ethical concepts, concepts such as CRUELTY, DISHONESTY, and BETRAYAL, as they apply to particular acts, have characteristic reasons as well. The reasons characteristic of CRUELTY (or CRUEL acts) are those which involve an aim to seriously harm or cause to suffer, debase, humiliate, degrade another, in order to delight in the infliction caused. In Chapter 1, we considered how such reasons related to the victim’s suffering, namely how an awareness of such reasons grounds the judgment that the victim’s suffering was needlessly brought about. With the concept CRUEL and its cognates (MEAN, NASTY, SPITEFUL, and so on) one thing that differentiates these concepts from one another is that there is a certain threshold in the consequences brought about: acts can be mean without being cruel, though when an act is cruel we are likely to also think it was mean. The threshold has to do with the degree of suffering, humiliation, or degradation caused, whether the degree of it was deserved, the way in which one acted, and so on. While this is perhaps not a particularly novel view of CRUEL, there are some who are apt to focus exclusively on the consequences of an agent’s act, as determinative of its being cruel, in exclusion of the reasons that an agent had in acting. Philip Hallie, for instance, holds something similar to such a view, and thinks that: …any such understanding of cruelty should leave out the phrase ‘intention to hurt’—the intention might not be there, but the maiming may be as substantial as 131 if it were there…the main thing is that the maiming is done, not the accident of whether the person who does the maiming is doing it for the sake of maiming itself. Cruelty is for us the infliction of ruin, whatever the ‘motives’. 28 Hallie, to his credit, wishes to emphasize the victim’s suffering and humiliation in cases of cruelty, and to highlight how the very language of victim/victimizer seems necessary to make sense of the parties involved in a cruel act, an act in which the victimizer must (at least from the position of acting) be in a position of power. I am quite sympathetic to this line of thinking (especially as it takes there to be an important conceptual relation between our notion cruelty and that of victimhood). But, in placing his focus exclusively on the consequences of the agent’s act, Hallie actually excludes a large part of what gives cruel action the evaluative character it has. The cost in doing so for his account of cruelty is obvious: (i) many acts, acts which fail to maim and ruin, can still count as cruel, and (ii) many acts which do maim and ruin might not be cruel. There is no reason to think that, in order to be cruel, the effects of the cruelty must be so long lasting as to permanently damage the victim, and, equally, there is no reason to think that all acts which do such damage a victim are cruel—such damage can be the result of say injustice or dishonesty, indeed even acts done for laudable reasons can have such effects. Hallie’s intuition, that the character of a cruel act relates to the amount of suffering produced in the victim, is not so much wrong as incomplete. A joke can be deeply cruel, even though it does not have lasting detrimental effects on the person to whom it was directed. An amputation can cause great pain and suffering, and indeed maim and ruin, and yet be an act of compassion, and hence not cruel. 28 (Hallie 1969, p. 14). Hallie’s have remained relatively unchanged; see (Hallie 1992, p. 229). 132 As I alluded to in Chapter 1, it is the moral justifiability of the suffering that matters to whether an act is cruel (as I put it there, whether there is a need for it). Prisoners suffer degradations on a daily basis, medical patients might have to endure excruciating levels of pain and agony, and people can be deeply humiliated by being the subject of joke gone wrong. But there is a difference, a moral difference, between the degradation faced by a prisoner and that of someone in similar conditions placed there simply to debase them for one’s own enjoyment. There is a moral difference between the field surgeon performing amputations without anesthesia and the individual who acts similarly in order to delight in the power he has to induce agony in his victim. And there is a moral difference between a joke made because someone thought the victim would get a good laugh out of it, and being utterly mistaken about what the actual effects would be. Whether any particular case of degradation, agony, and humiliation counts as cruel depends on whether the agent acted to cause one harm or suffering in order to delight in that suffering. The relevance of the agent’s reasons play a crucial role not only in whether we are right to characterize an act as cruel. These reasons are also of moral import because we have an interest in discouraging acts that arise from them. The explanation as to why we have a concept like CRUELTY is that we have an interest in distinguishing certain forms behavior from others, namely those that arise from certain reasons, so as to discourage agents from acting on those reasons. It is because of this that an agent’s acting from a reason characteristic of cruelty is a decisive ground for negatively evaluating the agent for what she has done. I will return to this point in a moment. 133 Acts of DISHONESTY are deceptions 29 that also arise from certain characteristic reasons. Two general reasons come to mind here: (i) that of aiming to escape responsibility for something (e.g., when one deceives to escape punishment for some act), and (ii) that of aiming to gain some advantage (e.g., when one deceives to secure something valuable they would not have received without the lie). Deceptions that arise from these reasons are what make those deceptions instances of dishonesty; they are also part of what makes them condemnable. This, of course, is going to seem controversial, especially because many are likely to think of dishonesty as straightforwardly related to deception. Let me try to motivate this point in the following way. These two types of reasons are not exhaustive of the reasons that are characteristic of dishonesty. But they are enough to help motivate a distinction between dishonest acts and those deceptive acts that would fall under the heading of innocent deceptions (a straightforward case being the notion of a white lie). Some will be apt to take a rather hard line here to making this sort of distinction, and consider even white lies to be instances of dishonesty (again in the technical sense of treating dishonesty and deception as basically one and the same thing), but hopefully the following example will help prime some intuitions for upholding a distinction between the two. Examples of this sort are legion, and we tend to lose site of the fact that, although innocent, they still count as deceptions. Consider one aptly understated by Charles Dickens: 29 Importantly, the focus here is with the contextually rich notion of deception and not that of simply lying (i.e., the uttering of a believed falsehood with the intent to deceive). One can deceive without actually uttering anything false at all, and so there can be forms of deception that are unrelated to lying. See here (Williams 2002, p. 98). For a view that takes the intent to deceive as an inessential component to the concept of lying, see (Carson 2006). 134 …Caleb knew that with his own hands he had brought the little rose-tree home for her, so carefully; and that with his own lips he had forged the innocent deception which should help to keep her from suspecting how much, how very much, he every day denied himself, so that she might be the happier. 30 That the deception of Caleb’s sacrifice was necessary so that his gift would not be ruined for his daughter Bertha makes a difference in whether we are apt to say that Caleb was being dishonest. It seems at the very least an oddity, perhaps even an overly moralistic reaction to what Caleb is trying to accomplish, to consider his deception a dishonest act (or perhaps worse: a series of them). Central to the difference here, if we accept that there is one, would be that Caleb’s reason for his deceptive behavior stemmed from a desire was to preserve his daughter’s joy in and enjoyment of the rose-tree upon receiving it for Christmas, both of which would have otherwise been ruined had the truth of how hard it had been for Caleb to get it, how much he had to give up, been revealed to Bertha. Beyond cases of deception that are in the service of quite noble aims, there are cases of other rather banal pro-social deceptions, deceptions that are not necessarily motivated by morally upright concerns, but are commonplace nonetheless. They are sometimes necessary in order for one to remain polite (where telling the truth would be rude or indecent), prevent unnecessary stress and anxiety (where the truth would cause undo confusion), and even to just make life simpler (where the truth is simply not that important and would unnecessarily complicate matters). The central difference that separates dishonesty from cases of both innocent deception and other more banal forms of pro-social deception, even though it can be agreed that they are all instances of deception, is one that rests on the agent’s reasons for the deception. 30 (Dickens 1911, p. 191) 135 Finally, as a last example, we can consider acts of BETRAYAL. Acts of betrayal are difficult to describe because they can only occur in complex interpersonal contexts. Such acts typically involve an egregious breach of trust, but it would be too quick a move to see the contexts where an act can count as betrayal as simply those that involve some unspecified sense of trusting another person. Trust involves more than mere attitudes of reliance on another’s behavior. As Annette Baier 31 notes the concept of trust at least involves: (i) The granting (or entrusting) to the trustee a degree of discretionary control and responsibility to care for something of value to the trustor; but, (ii) That (i) is granted on the condition that the care would be conducted in a manner that accords with the way the trustor values it; (iii) The trustee accepts this responsibility (either explicitly or implicitly); and, (iv) The trustor has a reasonable expectation that the trustee will care for the valued object in just that way. Given this, we can say that two people who satisfy (i - iv) are in a trust relationship. Many times the valued thing turns out to be persons involved in the relationship, such as when friends, spouses, siblings, and so on, are thought to inhabit trust relationships. Because of this trust relationships are often symmetrical in character (i.e., both parties to the relationship can be thought to occupy the role trustee and trustor). Because a trust relationship itself must involve a reasonable expectation that the concerns and goals of the entrusted will accord with the concerns and goals of the trustor, over the way the valued thing will be cared for, acts of betrayal often presuppose deep and intimate relationships existing between the betrayer and betrayed. 31 See (Baier 1986). 136 This, of course, is a very brief characterization of what a trust relationship amounts to. Hopefully it will be somewhat intuitive. Saying something about the nature of trust and trust relationships, however, is necessary for understanding betrayal, as it is against the background of trusting others, and having an interest in that trust being upheld as intrinsically valuable, that acts of betrayal are possible. This is because the reasons characteristic of such acts are those that involve aims to exploit the relationship for personal advantage and gain. These sorts of aims reflect the fact that, from the perspective of the betrayer, trust relationship has come to be viewed in fundamentally instrumental terms. The manipulation of the trust relationship must occur in a context where manipulation is reasonably expected not to occur, namely where the betrayer is expected to provide due care for the betrayed and has accepted (again, either openly or implicitly) the responsibility of doing so. Yet, not all trust violations count as acts of betrayal, even when one feels deeply that she is being abused and manipulated by someone she trusted deeply (one can abuse another’s trust without treating it as instrumental, say, by being careless). Whether the violation counts as a betrayal depends greatly on the agent acting for a reason that would involve treating the relationship as simply instrumental for some personal gain. 32 32 There is one group of thick concepts that I have not discussed, but is worth mentioning. These deal with actions that do not have any characteristic reason associated with them, but are applied to actions because some standard of due care has been violated. Here I have in mind concepts like RUDE, TACTLESS, INCONSIDERATE, THOUGHTLESS, CARELESS, and so on. What is characteristic of these concepts is not that an agent acts for some specific reason or other, but that an agent fails to be adequately sensitive to certain facts of a situation or certain feelings of others, where such sensitivity is taken to be called for in the circumstances. For instance, following Philippa Foot’s characterization of a rude act as one that “causes a certain kind of offence” (Foot 1978a, pp. 104-05), we should say that, in order for an act to be rude the offence caused must be one related to some amount of respect was owed to the offending party. In this sense, it is not the fact that a certain kind of offence was caused, but because the offense was, relative to the appropriate standard of respect, unjustified. 137 Section 3.4: Thick Concepts and Practical Reasoning One strand to the picture of thick concepts developed in the last section is that whether such a concept applies to some agent’s action depends on whether that agent acted for reasons that are constitutively related to that concept. If this is right, then how should we understand how thick ethical concepts figure in a judger’s own reasoning, about what she herself should do, will have to be articulated carefully. We might say (as a rough gloss) that the former issue concerned how we should understand the application of a thick concept to someone else’s behavior, and that the current issue concerns how such concepts should function in the judger’s own deliberations about what they themselves should do. And, with respect to the former issue, I noted that thick concepts involve evaluative rationales which relate that concept to other evaluative considerations. Because of this, I think it is a mistake to translate the manner in which one uses thick concepts to describe the actions of others into a first-personal deliberative scheme whereby thick concepts come to denote features of the various action alternatives that an agent can coherently rely upon in deciding what to do. This is because the reasons that are characteristically related to a thick concept are rarely characterized in terms of that concept, and to see them as more directly relatable to questions of what one should do is to obscure the sort of ethical attention one must have in order for that concept to correctly apply to what one has done. For instance, the thought that one’s act would be kind is not one that is typically featured in an agent’s deliberations where the resulting action turns out to thereby be kind. To make sense of such a picture would require a transformation in the agent’s deliberations from the thought that doing such a thing would be kind to 138 doing it for a reason that is characteristic of kindness (e.g., because she needs help, or she was in distress, and so on). This is precisely because whether an action is to count as kind at all, depends on the agent acting for a decidedly different reason than the thought that what one would be doing was kind. There is, however, a rather pervasive view, typically offered by moral particularists. Moral particularism is the doctrine that the practical relevance of any situational feature can only be determined holistically, and so a feature that might count as a reason for acting in one context might count against acting in another (or not count at all). 33 For example, the fact that a certain act would be pleasurable might, in certain contexts, be a reason to do it, but there are certainly contexts in which the fact that doing something would be pleasurable actually makes the doing of it worse. Whether the pleasure that would be derived from acting in a certain way actually counts in favor of doing it, then, can only be determined on a case by case basis. In this vein, thick concepts are treated as having a place in an agent’s deliberations as a clear means for deciding on what to do: not in the sense that possessing the concept shapes ones dispositions but in the sense that such concepts denote features of action alternatives in the same way that the pleasure which might be derived from certain actions denotes a certain feature of them. Such a picture takes for granted that we can make sense of thick concepts applying to particular prospective courses of action apart from the fact that acting for such a reason is not the sort of considerations that would make such a prospective course fall under that description. Such a view assumes, at bottom, that whether a thick concept applies to an agent’s own behavior is something independent of 33 On this, see (Dancy 2006) and (Dancy 2005). 139 the manner in which she actually deliberates in deciding what to do. This I think is incorrect. If the previous sections are right, this view is deeply at odds with the idea that thick concepts have characteristic reasons at all, not to mention the thought that the reasons characteristic of a thick concept are rarely if ever described in terms of such a concept. 34 To begin, then, let’s consider some representatives of this view. The first is provided by Simon Kirchin: It is assumed that judges have to make judgements about situations or possible courses of action. Such situations or actions comprise features. For example, if one could act so as to save a child from falling to its death over a cliff, this action might be composed of a number of features: the bravery of one’s action, the danger that one exposes oneself to, the (probably) happiness of the child and her parents afterwards, the swiftness with which one could have to act, the colour of the child’s shoes and so on…It is assumed that features can generate reasons for or against acting…For example, the fact that one’s act would be brave gives one reason to act. 35 Kirchin lays out, quite clearly, the thought that BRAVERY (or COURAGE) denotes a feature of potential courses of action and that competent moral agents are capable of being sensitive to the presence of this feature and can choose to act because of it (or treat such a feature as counting in favor of so acting). A second example is provided by Jonathan Dancy. Dancy’s remarks, however, are presented in contrast to the thought that thick concepts denote features of actions that can be treated as having invariant practical relevance. Dancy holds a rather strong version of moral particularism. Dancy’s view is a 34 One concept which often figures directly in the deliberations behind deciding to do something that would fall under that concept is the concept of JUSTICE or FAIRNESS. See (Williams 1985, p. 9). There are, however, questions about whether these count as genuinely thick concepts, or whether they are thinner concepts, but not as thin perhaps as GOOD and RIGHT. See here (Scheffler 1987, pp. 417-18) and (Williams 1996b). 35 (Krichin 2007, pp. 8-9). See also, (Krichin 2003, pp. 57-8). For criticism similar to the one that I am offering here, see (Norman 2007). 140 particularly strong form of moral particularism, because he takes the idea that the practical relevance of a thick concept can only be contextually determined. As such, whether the fact that an act is courageous or cruel counts in favor of doing that act is something that cannot be determined independently of the contexts in which such actions might occur. Dancy writes: …an action can be considerate without necessarily being the better for it. It may be considerate to wipe the torturer's brow, but this fact hardly functions as a reason to wipe, or makes his sweat a reason for us to wipe it off. The torturer's other activities prevent what would ordinarily give us a reason from doing so here. Similarly, it may be that a cruel response is exactly the one called for in the circumstances; cruelty, according to particularists, need not be an invariant reason. (Dancy 2005) 36 The core assumption behind both these quotes is that thick concepts denote features of actions that are intelligible quite independently of the agent’s actual deliberative processes. The agent is represented as not simply being able to consider her possible courses of action as either kind or courageous, but as being able to treat such features as considerations that can typically have a weight in their deciding what to do. Since this is a deeply pervasive view of the relationship between thick concepts and practical deliberation, I will refer to it simply as the received view. To isolate what is problematic about the received view, let’s consider the following alternative accounts of Mary’s deliberations on the question of whether she should visit her grandfather in the hospital: (R 1 ) Mary thinks she should visit because doing so would cheer him up. (R 2 ) Mary thinks she should visit because doing so would be kind. 36 See also (Dancy 1993), Ch. 4, and (Dancy 2006, p. 107). 141 Certainly, other things being equal, Mary’s acting in accordance with (R 1 ) is consistent with her doing something kind: as acting in accord with (R 1 ) is consistent with her acting for a reason that is characteristic of kindness. The central question for us, however, is how we are to understand the sense in which the reference to kindness in (R 2 ) could function as a reason. Certainly, in this case, Mary is concerned with doing something kind, but having such an aim is not the sort of reason that is characteristic of kindness. This, however, reflects something suspect about (R 2 ) as it stands, at least to the extent that we are not clear about the content of Mary’s deliberations yet, without further clarification. Mary’s reasoning, on the face of it, assumes something about the quality of her proposed action that she is not entitled to assume: namely that her visiting her grandfather would be kind, if done for that reason. 37 She is attempting to choose a certain course of action from a reason that, in principle, cannot be a feature of that course of action if she acts for that reason. There are, of course, various ways in which we might reinterpret (R 2 ), so as to avoid this sort of incoherence. In doing so we would be building in content to Mary’s deliberations that would extend beyond her simply seeing a proposed course of action as being kind, and therefore a reason to do it. For instance, consider (what we might call) back-stories (B 1 ) and (B 2 ): 37 I do not make any great claim to originality here. Indeed, Hume made a similar point in arguing that: To suppose, that the mere regard to the virtue of that action, may be the first motive, which prodc’d the action, and render’d it virtuous, is to reason in a circle. Before we can have such a regard, the action must be really virtuous; and this virtue must be deriv’d from some virtuous motive. (Hume 1978, p. 478) From this Hume concluded: …that no action can be virtuous…unless there be in human nature some motive to produce it, distinct from the sense of its morality. (Hume 1978, p. 479) 142 (B 1 ) Mary has just ended a rather hard and frank discussion with her sister about how she (Mary) has always been very cold and distant towards her friends and family, and how she tends to act as if she only cares for herself. As a result of the discussion, Mary resolves to change. She makes a start here by trying to deliberate from the standpoint of a kinder person, and by trying to consider what such a person would do in the circumstances that she will face. (B 2 ) Mary, as of late, has been really depressed and not her typical caring self. She used to be quite concerned about the well-being of her friends and family but recently she has been focused only on her own plights. Realizing this, however, she resolves to change her ways and return to her old self. As a result, Mary commits herself to making an effort to do more kind things for her friends and family. (B 1 ) provides us with a back-story to (R 2 ) that centers on trying to inculcate certain aims, reasons and attitudes by trying to habitually behave as a kind person would. The hope being that by acting and trying to deliberate like someone with certain concerns and interests, this will eventually entail that one comes to acquire the relevant dispositions to respond for the reason a kind person would respond. If this is how we are to understand (R 2 ), however, then her appeal to the kindness of her action is really to be seen as an appeal to an act that would be brought about by a kind person, and this would be consistent with the general aim of attempting to do what a kinder person would do. There is nothing incoherent about having such an aim, or acting for such a reason. Yet, her acting from such an aim would still not make her act kind, unless, as noted above, aiming to do such an act transformed her deliberations in such a way that she acted for reasons that are characteristic of kindness. By comparison, (B 2 ) offers a back-story that is somewhat ambiguous between the incoherence originally cited with (R 2 ) and something entirely consistent with Mary acting kindly. The reason for the ambiguity is that, on top of reading this as (strictly speaking) a confused piece of practical 143 deliberation, we might also see Mary, in committing herself to do more kind things for her friends and family, as committing herself to considering the interests and concerns of her friends and family more often and as having a weight in her deliberations, as considerations that she should be sensitive to in deciding how to act. Having this sort of aim, however, is entirely consistent with counting her act as kind. The reference to kindness in (R 2 ), then, would simply be a restatement of the relevant interests that would be furthered by visiting her grandfather. The trouble with either of these approaches provided by (B 1 ) or (B 2 ) is that they require re-interpreting how we should understand the reference to kindness in (R 2 ). In neither case can that reference be taken at face value: that is, as providing us with a sense of how kindness, as a feature of a course of action one is deliberating over, is actually capable of functioning as a reason that would make what she had done kind. At best, (B 1 ) and (B 2 ) point towards thinking that the term ‘kindness’ in (R 2 ) is really a placeholder for a more complicated body of considerations and deliberative processes. And so, while there might be ways of interpreting statements like (R 2 ) that reflect a coherent piece of practical reasoning, such interpretations could hardly be marshaled to support the received view. What makes a more straightforward reading of (R 2 ) seem so odd is that it requires Mary to approach her own proposed course of action from a decidedly third-person perspective, and this, I suspect, is what garners our suspicions about it. Typically, when an agent does something that is kind, and is asked why she acted as she did, the considerations given are often quite sparse: ‘It would cheer him up’ or ‘She needed it’. 144 Sometimes it is simply a restatement of the circumstances: ‘the child did not see the toy she was crying for’. From the perspective of first-personal deliberation, these were the considerations that, for the agent, counted in favor of acting. In most cases, however, getting a statement of the agent’s reasons is not enough to ascribe a thick concept to what the agent did. We need to know not only what was salient for the agent, but also how it was salient, in order to determine whether what the agent did was kind or otherwise manipulative or opportunistic. By asking ‘how it was salient’ we are concerned with how certain features of the agent’s situation were treated in her deliberations. This is what I take to be captured by the notion of there being reasons (aims, goals, objectives) that are constitutively related to a thick concept: the agent’s aim marks a midpoint between their recognition of the circumstances they face, and their response to those circumstances in action. What marks (R 2 ) as an object of suspicion is that it is as if Mary had: (i) taken an external (or commentary) position on what she was planning to do, (ii) rendered a verdict on whether her proposed course of action, if realized, would be kind or not, and then (iii) brought that verdict back to the perspective of first-personal deliberation, as a reason for deciding to do that very thing. 38 The paradoxical air of this, however, stems from an assumption implicit in the received view, namely that Mary’s act can count as kind independently of how she deliberates about her situation (i.e., what aims she comes to 38 Bernard Williams writes: Thinking about your possible states in terms of the virtues is not so much to think about your actions, and it is not distinctively to think about the terms in which you could or should think about your actions: it is rather to think about the way in which others might describe or comment on the way in which you think about your actions, and if that represents the essential content of your deliberations, it really does seem a misdirection of the ethical attention. (Williams 1985, p. 11) Williams relates this to the idea that in such cases ones deliberations are not ‘first-personal’ enough, meaning that there is a sense in which one is, in taking such a third-person perspective, trying to avoid the sort of responsibility that is inherent in the fact that this is the agent’s decision. 145 develop in response to the circumstances she is facing). As it stands, the most direct interpretation we can give of (R2) is that it reflects a concern for appearing kind, it reflects a concern in the wrong place, or directed towards the wrong sort of object, for it to issue in a kind act. 39 There are, however, cases where taking up a third-personal perspective on one’s prospective courses of action is far from problematic. Yet, these cases involve a narrower role in an agent’s deliberations than the received view would suggest. There are, for instance, cases where the reflexive application of a negative thick ethical concept exhibits a certain corrective function: its occurrence in an agent’s deliberation is typically to rule out a certain course of action, or at least provide a very strong reason for not doing it. One might, for instance, consider a certain way of assessing a student as really being mean, or come to see a failure to speak out as cowardly, or a discussion with a co-worker as a betrayal. These sorts of judgments are all-too-common, and are often not only the precursors to deliberating differently about how one should behave, but also the grounds for feelings of guilt, shame, and regret. In going through such a process of deliberation, an agent might come to find that certain considerations weigh too heavily in her deliberations, while others do not weigh enough. Regardless, it would be a mischaracterization of this sort of corrective process to think that it involves getting an agent to do certain acts because they are (say) kind or courageous (or fall under such a description). The role that thick concepts play here, however, is limited to being a 39 The idea that (R 2 ) reflects an unhealthy concern with the appearance of kindness, however, should be contrasted with a different sense in which an agent might come to be overly concerned with the ethical appearance of her action: namely, when she wants to appear as if she were kind. In such a case, the agent is concerned with appearing to others as kind, perhaps as an object of their admiration. This sort of case is not my primary concern. 146 corrective, as opposed to marking out certain features that make certain action alternatives choiceworthy. Certainly, however, there is a sense in which having a thick concept (in the sense of identifying with its evaluative rationale) involves an agent treating certain considerations and not others as reasons for acting. But this connection is less direct than the received view would suppose. Because of this the view of practical deliberation implicit in the received view for one who possesses a thick concept is deeply at odds with how agents, whose actions are rightfully described in their terms use to ascribe these concepts to the actions of others. As I have already noted, a thick concept is not simply the concept of an act that is brought about independent of the manner in which an actually agent deliberates, it is the concept of an act done for certain reasons (or it is an act that is the product of a certain deliberative path). If this is right, then the received view is wrong in how it attempts to model the action-guiding character of thick concepts. To be sure, thick concepts are action-guiding, but, if the above remarks are correct in prompting suspicion for the received view, then how they guide action is not in terms of their being labels for certain alternative courses in an agent’s deliberations. Rather, they guide action in the sense that the deliberations of a person who possess the concept will be different from one who does not. Not in the sense that there is some property or feature that a concept denotes which a person lacking the concept would fail to pick up on, but rather, in the sense that, in possessing such a concept, such a person’s deliberations will typically have a different structure from that of a person who lacks such a concept. 147 A person who possesses the concept COURAGE, who accepts or identifies with the evaluative rationale that underlies this concept, will (at least sometimes) find the fact that some important thing has been put in jeopardy or threatened, as constituting a reason to risk one’s own safety in coming to its aid. Possessing the concept of courage, then, entails the possession of a disposition to not simply rank the value of certain endangered things as more important than one’s own safety, but to treat their being endangered as a reason for risking that safety. In this way, possessing the concept does make a difference to the reasons one acts from, but not in the sense of identifying those reasons in its terms. One who lacked such a concept altogether, would lack a corresponding disposition to treat the value of such threatened things in this way, and accordingly their deliberations would be quite different. Similarly, a person who possesses the concept LEWD, will be one who views questions of sexuality and sexual expression in such a way that they call for monitoring and sometimes restriction because certain limits to sexual expression are firmly tied to conceptions of modesty and personal dignity. Here, however, we might see the concept as figuring more directly in an agent’s deliberations, namely as a corrective, and one does not possess the concept unless one possesses a corresponding corrective disposition. Surely, as many writers have been apt to highlight, there might be circumstances where lewd behavior is ‘called for’, 40 but this is something that one who possessed the concept, and the corrective disposition, should have no problem allowing. Saying that there are contexts in which lewd behavior might be best thing to do is different from saying that it is best because it is lewd. A modest woman might find the need to put on a display that 40 See here (Blackburn 1992) and (Dancy 1995) who consider similar cases. 148 is indeed lewd, so as to distract the prison guards and allow her fellow partisans a chance to escape, but it would be a mistake to think that in this case lewdness has been transformed into a beneficial quality of acts. Rather, it was the price that one had to pay to obtain a greater good. In short, if there is to be a connection between thick concepts and practical reasoning it must be one that is traced out in terms of the structure of the practical deliberations of those who possess the concept, and not, as the received view assumes, as labeling courses of action in advance of what an agent’s particular deliberations in fact are. That an agent who possesses a thick concept will deliberate differently from someone who lacks that concept is explained by the prior fact that such an agent possesses, through a process of upbringing and acculturation, the ethical interests which underlie that thick concept, and, in virtue of this, naturally deliberates about what to do in a manner that reveals a disposition to act either for reasons that are characteristic of a positive thick concept (like those related to concepts like KINDNESS, GENEROSITY, and LOYALTY) or to reject courses of action because they involve acting for reasons that are characteristic of some negative thick concept (like those related to concepts like CRUELTY, DISHONESTY, and DISLOYALTY). It is because the agent treats such considerations as relevant in deciding what to do, and not because she has the ability to see how certain courses of action might be described by others, that possession of a thick concept guides action. 149 Section 3.5: Concluding Remarks The account of thick concepts developed in this chapter rests on three related points. First, thick concepts have certain ethical rationales. The argument behind this is as follows. Thick concepts, like all concepts, classify. They discriminate some acts from others, and group acts together on the basis of their exemplifying certain features. Yet, the various acts that are grouped together under a thick concept are heterogeneous from the standpoint of their observable features. That this is so does not mean that we have the ability to perceive or detect some unobservable feature: rather it means that an individual’s ability to infer whether a certain course of action falls under a thick concept depends on their possessing certain ethical interests which structure and give shape to situations. These ethical interests provide a rationale for classifying acts in a certain way. The point of classifying acts with a thick concept is to mark them out for special recognition for having certain ethical important features, such as whether a certain instance of intentionally caused suffering is unjustified, as in the case of cruelty, or whether someone helps another simply for the sake of furthering that other person’s well- being, as in the case of kindness. We do not detect or perceive that certain instances of suffering are unjustified or that someone else’s well-being has been treated as an intrinsic good: we infer these things from our understanding of the context in which the action occurs and our understanding of the agent’s reasons for acting. Such inferences are aided by the fact that, in possessing such a concept, an individual possess certain ethical interests which determine which features, in which contexts, are relevant for making such a determination. And so, while thick concepts classify, the manner in which they do 150 cannot be prized apart from these interests, and so it is not something that can be made intelligible apart from them. Because the various instances falling under a thick concept lack any observable commonality, we can see why philosophers like Gibbard have thought of them as ‘lacking in descriptive meaning’. If by ‘descriptive meaning’ Gibbard intends something strongly connected with the observable features of an individual’s behavior, then thick concepts indeed are lacking in descriptive meaning. Gibbard’s mistake, however, was to think that this lacking required some extra evaluative component which provided a common element across the various instances in which a thick concept applies. 41 Because of this, he was motivated to see thick concepts as involving a concept-specific emotional response which would unify these various instances as all sharing a similar (affective) element. As I noted in the previous chapter this move is problematic in its own right, but it is also important to see that it is unmotivated: while the various applications of some thick concept might lack any observable commonality between them, so too might the various applications of concepts like TABLE and CHAIR. In short, Gibbard mistook the fact that such concepts involve family resemblances as revealing some insight into their ethical nature. What I have argued in this chapter is that the pursuit for such a common element is misguided. What unifies the various applications of a thick concept is a way of interpreting actions that is guided by certain ethical interests. Here I think this general point might be aided by an appeal to (and contrast with) John Kekes’ discussion of moral idioms: 41 The same mistake, in my view, can be found in (Hurka 2001) and (Hurka 2007) where he takes the ‘evaluative extra’ to be provided by an embedded evaluation using a thin ethical concept. 151 Moral idioms are interpretive rather than descriptive. This is a notoriously difficult contrast to draw. I am not supposing that a sharp distinction can be defended. Yet there is a kind of argument which is different from descriptive arguments and it is one involved in disagreements about the application of moral idioms. What I have in mind are arguments where what is at stake is not the fact, but how to think about the significance of agreed upon facts. Such arguments typically occur when there is agreement about the purposes of the interpretation, but there is disagreement about the arrangement of the elements upon which the interpretation is based. The arrangement hinges on what is taken to be central rather than marginal, novel rather than banal, surprising rather than routine, revealing rather than obscure. (Kekes 1984, p. 11) Moral idioms, for Kekes, are specific moral concepts, and in this way differ only in title from what have come to be thought of as thick concepts. Kekes’ point in considering such concepts to be interpretive rather than descriptive is to highlight the fact that their application depends on an action (or an action in a context) coming to be seen as having a certain structure. Whether this structure is present is not a matter of detecting some mysterious hidden element, rather it is a matter of whether the observable features of an action are thought to be arranged in the right way. I think that something like this is correct: the application of many thick concepts depends wholly on our coming to see the action as exemplifying a certain characteristic reason, and whether some act does this depends (at a general level) on a certain arrangement of features constituting a ground for inferring that an agent acted for such a reason. Where I would disagree with Kekes is in his denial that thick concepts (or moral idioms) should be seen as descriptive. They are descriptive in virtue of the fact that they classify or group actions together: what makes it appropriate to also consider them interpretive is the fact that such a grouping depends on a complex inferential process. 152 From Keke’s description of moral idioms, we can already see the second point outlined in this chapter: namely that many thick concepts classify actions on the basis of an agent’s reasons. Because of this I said that thick concepts have characteristic reasons associated with them. There are two subsidiary points to be made here. First, reasons are (obviously) not observable features of actions: as I just noted, to ascribe a reason to an agent’s act is to engage in a complex process of interpreting that person’s observable behavior within the social context that such behavior occurred. And secondly, the fact that many thick concepts are sensitive to an agent’s reasons helps to explain why certain forms of behavior are selected for special recognition: they are selected because a group has an ethical interest in encouraging or discouraging agents from acting on those reasons and certain courses of action are taken to exemplify such reasons. Of course, the interest in focusing on certain reasons as ethically important cannot be adequately explained without an appeal to the general sorts of consequences that are taken to follow from acting on them. We would, for instance, be at a loss to explain why acting simply to delight in the suffering caused in another should be discouraged if we could not appeal to the fact that suffering is something that we are generally (even if not categorically 42 ) opposed to. What drives our discriminating cruel acts from other acts are the various interests we have in discouraging people say from intentionally causing suffering in others, except in very unique circumstances, or from treating other people as simply objects to do with as one will. The complex pattern of behaviors that exemplify the 42 I say this because I think we are opposed to needless or senseless suffering. Some forms of suffering might very well be thought of as necessary for certain things to have a positive value, and in this sense a certain positive value is attached to the suffering itself. Consider, for instance, certain athletic endeavors, such as the Iron Man Triathlon: the valuing of competing in and completing such a race is tied importantly to how extremely grueling an undertaking it is. 153 reasons characteristic of the concept CRUEL cannot be adequately understood without a grasp of when and where certain situations involving the intentional causing of suffering and humiliation are needless and unjustified. But, because of this, there is also no further step to be taken in assessing whether there is something negative about what the agent has done once we have concluded that this agent has acted cruelly: for in describing their behavior with this concept we are already framing it in such a way so as to make this point apparent. Finally, although thick concepts are indeed action-guiding, I have argued that how we understand this point should relate more to how we should characterize the deliberations of those who possess the concept, and less with their ability to treat the actions whose undertaking they are considering as falling under the concepts themselves. On this view, what makes a thick concept like KIND action-guiding is that a person who possesses this has an interest in acting for the reasons that are characteristic of kindness, while a person who possesses the concept CRUEL has an interest in not acting on the reasons that are characteristic of cruelty. What I have argued against is the view that such concepts are action-guiding by directly providing an agent with a reason for action, as if such concepts denote features of actions independently of the manner in which an agent deliberates. Such a view is deeply at odds with the thought that thick concepts involve characteristic reasons, and because of this, their application to what an agent has done cannot be prized apart from the manner in which the agent deliberates in deciding how to act. 154 CHAPTER 4: Objectivity, Reflection, and the Loss of Ethical Knowledge Section 4.0: Introduction In this chapter, we will consider the notion of thick concepts in relation to the broader range of questions. In particular, whether they provide some form of ethical knowledge and how they relate to the issue of objectivity. These issues will bring us into direct contact with the work of Bernard Williams on thick concepts and his claim that the ethical knowledge provided by a thick concept can be lost or undermined when a group comes to develop a great reflective understanding of that concept. What is meant by ‘losing ethical knowledge’ and a ‘reflective understanding of a concept’ will require a good deal of unpacking. Because of this, the chapter’s discussion will be largely exegetical. It will, however, also be a critical discussion, at least to the extent that I will try to clarify and defend Williams’ views from a number of different detractors. Some have found his views patently objectionable (and bordering on incoherence), while others have them to amount to nothing more than a rather radical way of stating something that is altogether uninteresting. I think that Williams’ position is neither objectionable nor banal. Rather, it provides a picture of ethical thought that is worth defending because it helps us to appreciate some of the deeper worries that are endemic to modern consciousness (especially in relation to considerations about ethical diversity). Central to this discussion will be the picture of thick concepts that has developed in the discussion of the previous chapters, and especially the thought that thick concepts involve ethical rationales which structure the deliberations of those who possess them. Williams’ view is, without a doubt, disquieting, for it suggests that thick concepts are decidedly less than 155 what others might have hoped for, but this disquiet, as I will suggest in the next chapter, should be tempered with certain positive suggestions that are implicit is his views on reflective understanding and explanation as they relates to the use of thick concepts. My own view is quite sympathetic to Williams’ position, and because of this much of the discussion will be defensive in nature. Many of the criticisms made against Williams’ here, I think, have been directed towards a caricature of his position. This is not the result of philosophers not genuinely attempting to understand Williams: the position he is arguing for is a complex one and involves several interconnected elements. However, I also think that at least some of these criticisms result from Williams’ own style of exegesis. A good deal of his argument is presented in a highly compressed form, and he sometimes does not take care enough to highlight how his view is attempting to break from traditional forms of skepticism about the objectivity of ethical thought. As such, I will attempt to both lay clear some misunderstandings and also respond to objections that I think clearly fail. In Section 4.1, I give a brief overview of Williams’ picture of thick concepts. Much of this ground has already been laid in previous chapters, but hopefully this work can help to clarify some of Williams’ remarks. Section 4.2 will then turn to discuss Williams’ nonobjectivism. Williams, in wanting to side neither with the position that judgments involving thick concepts were merely expressions of preference, feeling, or emotion, nor with the view that such judgments had any hope of being objective, opted for nonobjectivism as a sort of middle ground. My discussion of his nonobjectivism will have three subcomponents. First, we need to understand the sort of ethical knowledge that thick concepts provide. Second, we need to 156 consider Williams’ general picture of objectivity. This is a daunting task in its own right, as Williams’ view here is somewhat idiosyncratic and breaks with more traditional ways of understanding what objectivity involves (i.e., that a judgment is objective just in case such a judgment can be true or correct or state a fact; Williams intends something different from this). One part to this will be Williams’ notion of an absolute conception of reality, and here I will offer a way of seeing how this notion fits into Williams’ larger argument that differs from what other commentators have thought. The notion of an absolute conception of the world (i.e., a conception of the world as it is anyway, independent of any particular perspective on it) is something that has been largely misunderstood, and has garnered the majority of criticism directed against his position on the nature of ethical thought. And third, we will discuss Williams’ notion of (what he calls) the hypertraditional society. The hypertraditional society is a hypothetical community that is homogeneous to a large extent in its ethical outlook (i.e., members of such a group subscribe by and large to one and the same set of thick concepts) but which is also minimally reflective (i.e., they are not concerned with reflective questions about how they came to have thick concepts they possess). In Section 4.3, I will attempt to put these pieces together in a way so as to clarify Williams’ view that reflection can unseat the ethical knowledge provided by a thick concept by undermining the group’s ability to actually use it. Finally, in Section 4.4, I offer up comments on what we should think of Williams’ overarching position. Much of the discussion in this final section will be to set up the question of how we are to describe the sort of loss that Williams is gesturing at, and to locate the sort of tension that Williams took there to be between the practice of 157 using certain thick concepts, and developing a reflective explanation of how that practice came about. These points will then be discussed, in more detail, in the next chapter. Section 4.1: An Overview of Williams’ View of Thick Concepts It is essential to understanding Williams’ view of thick concepts, that we place such a view against his general concern about the focus that moral philosophers, for at least the first half of the 20 th Century, had on moral language. Williams thought that many philosophers, caught up in the ‘linguistic turn’, had focused exclusively on a rather narrow range of terms: ‘good’, ‘right’, and ‘ought’ being exemplars. 1 But it would be a mistake to think that there was some alternative range of terms that Williams had in mind when he claimed that the linguistic focus of moral philosophy would be better served if it was “more concerned with what we say.” 2 Williams invitation, I take it, was in part to motivate the thought that the endeavor to understand ethical thought would be better served by the recognition that the ethical vocabulary of different cultures, and at different times within the history of one and the same culture, was not as homogeneous as the selection of (or emphasis on) certain general moral terms, by many previous moral philosophers, would suggest. But Williams’ point was also an invitation to see certain ramifications that were tied to the idea that ethical communities were diverse not only in their views, but in the concepts which they used to express those views. To possess a concept, not in the minimal sense of being competent with some subset of a language, is to have one’s life and experiences shaped by it. Williams notes: 1 See here not only (Williams 1985, pp. 128-31), but also (Williams 1973). 2 (Williams 1985, p. 127); emphasis mine. 158 How we ‘go on’ from one application of a concept to another is a function of the kind of interest that the concept represents, and we should not assume that we could see how people ‘go on’ if we did not share the evaluative perspective in which this kind of concept has its point. (Williams 1985, p. 141) How one thinks ethically about the situations they face, or themselves, or the relation they stand in to others, or about what to do, is in part a function of the concepts they bring to bear on those questions. But the concepts themselves have a point within a community because they answer to ethical interests that are a constitutive element of such a community’s ethical practices. Without a grasp of the underlying ethical interests, one would be unable to grasp how different applications of a thick ethical concept constituted a way of ‘going on’ and doing the same thing. Consider again an example I raised in Chapter 1. I take it that no one uses the concept CAD anymore. As it was once used, say three-quarters of a century ago, the concept applied to men who behaved in dishonorable or disrespectful ways towards women (or women of a certain social standing, class, or upbringing). The fact that this term is no longer used to structure our understanding of how men might behave poorly towards women is simply to register Williams’ point that ethical communities, even different stages in the same ethical tradition, can differ in the concepts they use to structure and interpret actions and individuals. While the concept CAD did genuine ethical work for past generations, it does not, and indeed cannot, provide the same sort of structure for contemporary Western (or Anglo-American) cultures. 3 The reason why such a concept has no place in our contemporary social world is that its sincere use and adoption cannot be squared with the ethical interests and concerns that we currently prize. 3 I am assuming that this is because the concept CAD involves something like the recognition that women in particular are especially important objects of respect because they are women. 159 The interests that give such a concept its point are those that relate to treating women (or women of a certain class) as special objects of concern and respect in virtue of their being women (or women of a certain class). A contemporary commitment to equality among the sexes is incompatible with holding such interests, and, because of this incompatibility, such a concept is not one that a person from our social world could sincerely use. Put in terms of the last chapter’s discussion, the concept CAD involves (or presupposes) an ethical rationale that we no longer accept or identify with. To say, however, that such a judgment cannot be sincere is not (or not yet) to say that a person from our social world making it is thereby insincere in their judgment. It is not that, if such a person were to judge (say) that Greg is a cad, such a person would be speaking insincerely (as when a person offers someone congratulations but does not think well of what they have done). Rather, it is that such a person would be making a judgment that (in a certain sense) they had no right to make. This concept presupposes certain interests and concerns that this person does not share, in virtue of being a person from our contemporary social world, and, because of this, their use of the concept cannot be one that was determined by the interests and concerns that give a point to using such a concept. Such a judgment could not be an expression of what they find valuable or important; it cannot be one that is deployed to give voice to their values. 4 It is in this sense that when such person makes a judgment with this concept, he is not speaking or 4 Importantly, this point extends to the diversity of ethical concepts can be found currently operative in other cultures. Consider the Chinese concept XIAO, or filial piety. Bound up with this concept, I take it, is the notion that one’s parents are special individuals deserving of respect and admiration, not because a child’s parents are assumed to be good parents or somehow wiser due to their age and experience, but because one’s parents were, regardless of temperament and wisdom, the providers of the gift of life: in short because one’s parents were one’s parents. To have a right to use this concept would require identifying with such an interest. For an excellent overview on stories covering paradigmatic instances of filial piety in ancient China, see (Holzman 1998). See also (Li 1997). 160 judging insincerely, but rather is doing so without right. Put another way, he lacks the epistemological license to use the concept, because such a judgment would not be one shaped by concerns he actually has. All of this is, of course, quite schematic. The goal of this chapter and the next will be to fill in the details and to flesh out the ramifications that this point has with respect to how we should understand thick concepts and the knowledge they provide. Beyond these schematic points, Williams’ own view of thick concepts, in terms of its relation to broader metaethical views, is unique and worth highlighting. It is perhaps on account of the uniqueness of this position that many commentators have objected to it on grounds that fail to connect with much of what Williams has actually said. Williams certainly disagrees with those who think that thick ethical concepts can be divided into evaluative and nonevaluative aspects, but he also disagrees with those who see such concepts as supporting some form of ethical objectivism. Each suffers, in his view, from a different, albeit related, error. Those who wish to treat thick concepts as constructions of independently understood evaluative and nonevaluative components err in not properly appreciating that such constructions could not be simply substituted into the ethical viewpoint from which the original thick concept originated. 5 Certainly, many thick concepts will be constrained by various conditions, some of them being ‘descriptive’ (that is, characterizable in patently neutral terms) while others being ‘evaluative’, and to this extent it might seem an inescapable temptation to divide the content of these concepts in such a way that would show clear which aspects fell on which sides. (The temptation is so great, I believe, that Williams opted out of using these terms all together 5 Cf. (Williams 1995d, p. 187) 161 in discussing thick concepts. But we will return to this point in a moment.) This error, however, would be mischaracterized if it were thought to be purely semantic, or purely a matter of definitions or criteria. It is an error tied to thinking that we can subtract from their understanding of these concepts that which is important or interesting or valuable for such a group, leaving some neutral characterization of what it is that they are reacting to when they issue judgments using such a concept, and then graft back onto that characterization some evaluative concepts that are generally shared between us and them (such as ‘good’ or ‘wrong’). The error is in thinking that we are capable of understanding such concepts in any other way than the way that they understand them. Those, however, that agree with Williams that such ethical outlooks or perspectives are central to the understanding of a thick concept nonetheless fail to appreciate that there is more than just one such outlook; they tend to assume that everyone shares the same set of thick concepts. This, Williams thought, had a bearing on questions of objectivity regarding the judgments made in their terms. The fact that we do not share a single set of them (as is amply witnessed by an appreciation of the diversity in such vocabularies throughout history 6 ), coupled with the fact that we cannot ably organize the diversity of sets of thick concepts we find into a cohesive, and livable, unity, forces upon us the justificatory demand of determining which thick concepts we should use to conceptualize our ethical experience. And to this demand, Williams thought, was one that we lacked any adequate response to, and because of this are in no position to claim that one set is objectively grounded over another. This denial, however, did not mean that we must recant the thought that judgments using thick concepts could be true 6 Cf. (MacIntyre 1984) 162 or correct, or that people could be mistaken in the judgments they make using such concepts, or that there could be a form of knowledge tied to their use. Williams’ firmly believed that these were all possibilities for ethical thought under thick concepts. It was rather that the knowledge provided (and the conceptual issues related to this) would have to be understood as local or perspectival: it would not be knowledge that was available to all. Such a metaethical position is caught, then, in between the idea that ethical judgments using thick concepts are merely subjective (i.e., involving simply the expression of the ethical views of one’s group) and objective (i.e., conforming to an independent order of moral truths and facts). It is for this reason that Williams labeled his position nonobjectivism. Yet, despite being both a cognitivist (insofar as he held that certain ethical judgments could be true and known) and an antirealist (insofar as he denied their objectivity), Williams certainly did not think (like J.L. Mackie 7 ) that thick ethical thought and discourse suffered from some global error: namely that, while aspiring to be true, all judgments using thick concepts were in fact false, on account of the fact that there was nothing for them to be true of. To think this would be to convict the use of these concepts of the very objectivist pretences that Williams denied outright as needing to be a part of our understanding of the practices that use them. And so, while Williams is indeed a cognitivist and an antirealist, his intermixing of these metaethical positions is unique from what others took their connection to involve. According to Williams, thick concepts are both world-guided and action-guiding. Thick concepts are world-guided is the sense that their application is (in some sense) 7 (Mackie 1977), Ch. 1; (Schiffer 1990) and (Joyce 2001) also provide error-theories. 163 controlled the facts of a situation. These concepts, for instance, can be straightforwardly applied to an act on the basis of how a user perceives that act. Since Williams’ aim is to provide an account of these concepts that does not support some form of realism, the idea that such concepts are controlled by a user’s perception of some act should be read liberally as something that is interpretation-laden. A thick concept might be controlled by a user’s perception of an act even though, strictly speaking, the user does not perceive some thick property. Rather, in a more relaxed sense, how one perceives some act can be structured by possession of a relevant thick concept such that it is seen as fitting or failing to fit the relevant pattern for the concept to apply. And, in virtue of this: [a] concept of this sort may be rightly or wrongly applied, and people who have acquired it can agree that it applies or fails to apply in some new situation. In many cases the agreement will be spontaneous, while in other cases there is room for judgment and comparison. Some disagreement at the margin may be irresoluble, but this does not mean that the use of the concept is not controlled by the facts or by the users’ perception of the world. (Williams 1985, p. 141) While the application of these concepts is guided by how the world is like (say by an understanding of how a person might have acted in some context), thick concepts also guide action: They are characteristically related to reasons for action. If a concept of this kind applies, this often provides someone with a reason for action, though that reason need not be a decisive one and may be outweighed by other reasons….Of course, exactly what reason for action is provided, and to whom, depends on the situation, in ways that may well be governed by this and other ethical concepts, but some general connection to action is clear enough. (Williams 1985, p. 140) Williams’ aim is to show that, however complex it might be, there is a general connection between thick concepts and action. What reasons are provided to whom and why might be an irreducibly contextual matter: in one an agent might be provided a reason for not 164 acting in some way by the fact that he comes to see such behavior as mean, in another, an evaluator might be provided a reason for censuring an agent because what she has done is mean, while in a third case, the reasons might be provided to counsel someone so that they avoid contact with a person (or avoid certain sorts of contact with them) because they are known to respond in mean ways. While the relationship between the application of a thick concept and various reasons for action might be complex, we can understand how there is such a general relationship by considering the reasons behind having such a concept: namely that the group has an interest in discouraging people from acting in mean ways towards others. Such an interest can be seen as sitting behind the agent’s rejection of a certain course of action, the evaluator’s negative evaluation of some other agent’s act, and the counselor’s advice on why a certain person should be avoided. Williams treats the notion of an “evaluative point” 8 as central to understanding a thick concept, but he never explains what he means by this or why it is important. I am apt to read Williams’ view in line with the suggestions I have made in previous chapters on this, namely that thick concepts involve rationales which involve certain ethical interests and which provide a certain ethical role for the concept. To say that thick concepts have an ‘evaluative point’ means that there are certain reasons why the group uses this concept to classify acts and individuals, and that these reasons not only help explain the role that such a concept has in the community’s ethical economy but they help structure how the concept is to be applied from case to case. A failure to grasp a thick concept’s rationale, then, would entail a failure in competence, for without an understanding of the point of using the concept one would be unable to see how it relates 8 (Williams 1985, p. 141) 165 to new and novel cases. For instance, one might be given a wide range of examples of acts that constitute behaving like a cad, but without an understanding of the point to having such a concept (or the point of classifying things in this way), that is, the reasons why various pieces of behavior are grouped together for discouragement, 9 one would be unable to see how certain novel actions, undertaken by a man, would lead to a women’s loss of respectability. In this way, understanding the interests that underlie such a concept comes to be a pre-condition for understanding that very concept. Despite this, though, it is important to note that, for Williams, a thick concept can be intelligible to those who do not accept or identify with these interests. While an outsider must appreciate how such a concept fits into the ethical economy of the community that uses this, such appreciation does not force an acceptance of that outlook upon the outsider. Use of the concept, after all, is not forced upon the outsider either. Some might take this to capitulate an intuition in support of a component view of thick concepts, but it does not: although thick concepts can be intelligible to an outsider, they cannot be used by an outsider unless she comes to share the ethical perspective that underlies the concept (e.g., in order for one to judge that someone is filially pious, and have that judgment count as a moral judgment, one would have to, for instance, believe 9 Especially strong moral particularists, such as Jonathan Dancy, are likely to disagree about this, since they hold that whether any consideration provides a reason to do or feel something can only be determined by the place of that consideration in the whole context in which it is found. And thus, that ‘being an instance of caddishness’ or ‘acting cruelly’ might actually count in favor of acting. See (Dancy 1995). I think such a view is mistake concepts because it undermines the very thought that there are thick concepts at all. By treating their reason-giving capacity as inherently tied to contextual features, this makes them no different from straightforward descriptive concepts. On this criticism of Dancy’s form of particularism, see (McNaughton & Rawling 2000, p. 273). I do not follow McNaughton and Rawling (or Dancy for that matter; see (Dancy 2006, pp. 121-22)) in supposing that thick concepts necessarily mediate between the non-moral world and judgments made in thin concepts, but I do think that there is something suspect in treating them as moral concepts, and their having a central connection to an evaluative point of view (as even Dancy does), only to treat that connection as somehow fickle with respect to the importance of the features it denotes. 166 that one was under a special obligation to respect and revere one’s parents, simply because they were one’s parents). Apart from identifying with the interests which underlay a thick concept, there will remain a genuine sense in which such a concept is not that the outsider’s. She might come to understand it, and understand when and why it applies, but such a person would still be unable to manifest that understanding in a way that could issue in (what was for them) a genuine moral judgment. While I already have intimated above why this is so, a full defense of the point will have to wait until the next chapter. There is one final point worth noting about Williams’ view of thick concepts, and that is with the contrast he draws between them and what he calls ‘thinner’ ethical concepts, concepts like GOOD, RIGHT and OUGHT. And, again, it will be worthwhile to remove a certain misconception about thin concepts. Since it seems appropriate to treat thick concepts as involving a greater amount of descriptive or empirical content (leaving open here as to what that precisely means) than thin concepts, the temptation will be to think that, if there is a difference at all between thick and thin concepts, it is a matter of degree and not kind. I think that this is a mistake. First, if thin concepts have any descriptive content associated with them at all, the association is either stipulative or formal. It is stipulative if we simply define ‘good’ as say the hedonist would by saying that all and only pleasure is good. If pleasure is the only good, this is only because of the hedonist’s has insisted to define it in this way: such a definition does not fall out of an analysis of the concept. What might be thought to fall out of an analysis thin concepts, however, is the requirement that if something is ‘good’, ‘right’ or ‘ought to be done’, then 167 there will be some reason as to why this is so. As such, descriptive content would seem to get in via existential quantification: (∀x) (∃y) (x is right in virtue of y). R.M. Hare, 10 for instance, took something like this to follow from the universalizability of prescriptive judgments. But it is important to see that this is a formal requirement, and does not count as world-guided in the sense intended by Williams. Thin terms are not guided by the world because there is no determinate way of interpreting acts and individuals that is associated with their use. What must be the case here, in order for one to make a sincere and intelligible judgment using these concepts, is simply that there are some features which a person takes to ground their judgment. Some have also argued that the distinction between thick and thin concepts is not as clear as Williams considers it to be, as there seem to be other concepts, such as JUSTICE 11 or the concept GOOD PERSON, 12 which Williams’ distinction fails to adequately account for. I will not venture to answer this question, save to raise a point that Williams was keen to highlight about thin concepts. Williams’ decision to focus on concepts like GOOD, BAD, RIGHT, WRONG, and OUGHT as exemplary of thin concepts was due (at least in part) to the fact that such concepts are the most general and abstract in evaluative thought. He thought many moral philosophers were naturally inclined to focus on them, for this reason, as those concepts that are most amenable to ethical theorizing. Because 10 (Hare 1952), Ch.6. 11 For this criticism, see (Scheffler 1987), esp. pgs. 417-18. Williams accepts this criticism insofar as there might be intermediate concepts between thick and thin ones. See (Williams 1996b, p. 26). 12 For this, see (Quinn 1987, p. 203) and (Hursthouse 1995, p. 203). 168 of this, thin concepts come to be seen as more or less conceptually fundamental. 13 Yet, these concepts, in ordinary language, seem to gain their clearest sense by the roles that they play in various sorts of judgments: ‘good’ has clear role in expressing preferences, ‘right’ has one related to reaching certain deliberative conclusions or in agreeing with others, ‘ought’ has a role in the offering of advice or making recommendations. Williams certainly did not take any these terms to be reducible to their role, but he did point to these roles as a ground for being suspicious of the thought that they have anything but the barest content. Williams, in a later discussion, highlights this point by noting: …although I am sure there has to be a univocal sense of ‘right’ and ‘wrong’ closely related to practical reason such that [two disagreeing parties] can come to understand that what one thinks is right the other thinks is wrong, I’m not the least clear that this is identical with the terms ‘right’ and ‘wrong’ that turn up so often in ethics. I don’t know what these terms, in these ways they are so often used in ethics, means (I’m not sure that those who use these terms in ethics know what they mean either). But at some level very near to practical decision there has to be a term that is used univocally in different cultures. (Williams 1996b, p. 32) What I take Williams to be gesturing towards is the thought that concepts like RIGHT and WRONG, as ones that could be seen as univocal across cultures, lack any distinctive content associated with them. Our best understanding of these concepts seems to come from their respective roles. It is because of this that I think Williams doubted whether evaluative concepts like RIGHT, WRONG, and OUGHT, could be sensibly thought of as action-guiding at all, 14 as our very understanding of them depends importantly on how they fit into the expression of practical decisions (or ones agreeing or disagreeing with it) 13 Susan Hurley labels this view “centralism.” Centralism is the view that there is some privileged (i.e., central) class of evaluative concepts that is both independently intelligible and theoretically prior to the rest. As such centralism is a reductionist view, which sees all evaluative concepts as constructions out of this privileged class. See (Hurley 1989, pp. 10-12). Non-centralism is simply a denial of the centralism, and hence a denial that there is such a privileged class of ethical concepts. 14 (Williams 1985, p. 141) 169 as opposed to guiding them. This also fits with a point that Williams raises about ethical knowledge, and how a convergence in the use of thin concepts will not be enough to sustain a form of such knowledge, 15 and how a social world in which thin concepts predominated would be very different from one in which thick concepts did. If what has been said is a correct reading of Williams’ position, then even a universal convergence in a claim like ‘abortion is wrong’ will not amount to a piece of ethical knowledge: it will (at best) be a convergence in a conviction about the moral status of abortion. What would be required for knowledge would be a convergence in the grounds for this conviction, which would be unsettled by our simply converging on the view that abortion is wrong (and the fact that there must be some ground for such a conviction does not entail that it will be the same ground). Much of Williams’ discussion on thick concepts, ethical knowledge, and reflection, especially in Ethics and the Limits of Philosophy, rests on arguments that are quite complicated and at times highly compressed. 16 Because of this, a significant part of the remaining sections in this chapter will be dedicated to clarifying and defending Williams’ nonobjectivism. The focus will be on coming to a defensible interpretation of Williams’ position. Defending such a picture will help to remove some controversy in understanding Williams’ conclusion that, for thick concepts, “reflection can destroy [ethical] knowledge.” 17 Before getting to this, however, we need to consider in greater detail both Williams’ views (or a plausible extension of them) on how thick concepts can 15 (Williams 1995c, pp. 206-07) 16 Williams himself admits as much about his writing. See (Williams 1986, p. 204). 17 (Williams 1985, p. 148) 170 be thought to provide a form of ethical knowledge, as well as, his views on how we are to understand this in relation to questions of objectivity. Section 4.2: Understanding Williams’ Nonobjectivism In order to understand Williams’ nonobjectivism, we will have to proceed in a stepwise fashion. His account is complex and nuanced, with several stages in the argument requiring independent treatment and defense. There are three interconnected issues that we must consider. The first is how we should understand the notion of ethical knowledge as it relates to thick concepts. This, in itself, is a complicated question, but I will provide a gloss of a description in subsection (a). The second point concerns Williams’ views on objectivity, and how we might understand certain concepts (and conceptual practices) to be in the business of attempting to describe the world as it is anyway (indeed what this latter notion amounts to for Williams requires unpacking as well). This will occupy the discussion in subsection (b). Finally, in subsection (c), we will consider how certain concepts might simply lack any objective aspirations. This point is crucial to Williams’ view that the use of thick concepts is best seen as nonobjective. (a) Thick Concepts and Ethical Knowledge For Williams, thick ethical knowledge (knowledge involving the use of a thick concept) is a type of social knowledge: possession of a thick concept provides its possessor with a way of navigating the social world of which that concept is a constitutive part. 18 Williams, however, is not entirely clear as to what he takes the 18 (Williams 1985, p. 151) 171 connection is between thick concepts and navigating a social world to be, nor on how we should unpack the metaphorical notion of navigation. One way of unpacking this notion would be to see thick concepts as laden with or connected to a vast array of social expectations, and that coming to appreciate these expectations is a part of competency with the concept. 19 To see what I mean by this, consider first example. Sally and her friends have always been a tight-knit group: they not only shares a rich history together, but each is quite familiar with the various intimate details of the others’ lives. On Sally’s birthday, she and her friends get together to celebrate. As the evening proceeds, the group starts teasing Sally about the various relationships she has been in. The teasing starts out as a playful way of making her the center of attention, but soon some of the teasing (by one member of the group in particular, Mark) crosses a line. Sally becomes clearly uncomfortable with Mark’s joke and starts to get upset. Mark is oblivious to this, but everyone else in the group comes to think that his comments were inconsiderate, even downright mean. No one, however, voices their opinion of Mark’s teasing outright. Some in the group move to change the subject of the conversation; some try to console Sally without making her discomfort the object of further attention; and some quietly pull Mark aside and try to get him to see how distressed his joke has made Sally. Mark initially thought that his teasing was just a little harmless fun, but after seeing Sally upset realized that his joke had indeed gone too far. As a result, Mark goes over to Sally and apologizes for his behavior. 19 On this point, I am indebted to the discussions I have had with Dallas Willard. His suggestions and way of looking at this issue have been invaluable. 172 In this scenario, certain thick concepts (namely concepts like INCONSIDERATE and MEAN) were taken to be operative in and between the members of Sally’s group. Even though no one explicitly voiced a judgment to the effect that Mark’s joke was inconsiderate and mean, these concepts played a role in guiding the group’s actions. The fact that Mark’s joke was thought by the group to be inconsiderate provided some of them with certain practical reasons: some were directed towards easing Sally’s distress, either directly or indirectly, while others were directed towards getting Mark to see his behavior in a clearer light, while still others (namely with Mark) were directed towards making amends for a thoughtless remark. The thick concepts that were at play here not only provided a medium for identifying the problem within the situation itself, but they also provided a means for navigating through that problem by way of various courses of action. Which courses of action were recommended (and to whom) no doubt depended on many factors, but central to whatever reasons were presented was the thought that Mark’s teasing was rude and inconsiderate. It is in this way that we might say thick concepts are thick with certain social expectations, as their application in a particular case can be seen as meriting (i.e., providing reasons for) a variety of responses. Williams is clear, however, that the knowledge provided by thick concepts is not to be understood as simply practical knowledge. It is rather a type of knowledge that clearly and straightforwardly meets the general philosophical conditions for propositional knowledge: (1) that an individual believes a judgment made using a thick concept (which, I take it, involves a certain degree of conviction in that judgment’s truth), (2) that the 173 judgment is true, and (3) that conditions (1) and (2) are “non-accidentally linked.” 20 In our above example, a member of Sally’s group: (1) comes to believe that Mark’s joke was inconsiderate and mean: (2), given the criteria for such concepts, and given the nature of Mark’s joke, it was indeed inconsiderate and mean; and, (3) the fact that a member of Sally’s group came to believe this was no accident: she witnessed the joke being made, and came to believe that the joke was inconsiderate and rude on the basis of what she witnessed. Put another way, had the situation been different, then this person would have witnessed something other than what she actually witnessed, and because of this would have formed different beliefs (assuming that in this counterfactual what she witnesses is not Mark making a different inconsiderate and mean joke to Sally). Given that such differential responses are possible with respect to changing circumstances, there seems to be no reason to deny that members of Sally’s group are capable of ‘tracking the truth’ when it comes to things like inconsiderate and mean behavior. The idea that thick concepts provide a form of ethical knowledge is tied to the fact that the members of Sally’s group all identify with the interests that are relevant to their possession of the concepts INCONSIDERATE and MEAN: they believe that due attention should be paid to the feelings of others, that people should not be harassed or be intentionally made to feel uncomfortable or distressed, and so on. These interests made certain features of the situation relevant towards acting in certain ways. However, being identified with the interests underlying such concepts does not make one infallible in 20 See (Williams 1985, pp. 142-43). This latter condition Williams spells out in terms of a differential sensitivity to changing circumstances, which can be expressed through a variety of subjunctive conditionals, whereby, in believing the proposition that p, one would have believed q, if q, or r, if r, and so on. This captures Robert Nozick’s notion of “tracking the truth” in (Nozick 1981). 174 understanding when something is INCONSIDERATE and MEAN, as Mark’s situation reflects. Yet, even in Mark’s case, the ethical rationale and interests underlying these concepts can be seen as guiding his actions, for, in coming to see his behavior as indeed inconsiderate and mean, this provided him with a reason to apologize to Sally. That such reasons flow from the correct application of these thick concepts, then, is what their ethical knowledge consists in. It is a form of social knowledge that crucially depends on certain propositions being true (or at least held true by those who use the relevant concepts). Beyond this, there is one additional connection that thick concepts bear to the notion ethical knowledge, and this is to be found in the fact that some people might be more adept at using these concepts than others. 21 Consider here Mark’s inability to see that his joke was inconsiderate and mean to Sally without the counsel of another. Williams’ thinks this point helps us to understand how thick concepts play a role in the offering of advice: 22 “the idea of a reliable ethical informant should be construed rather in terms of practical experience and judgment….The valuable informant can be seen as one who, so to speak, has been down this road before.” (Williams 1995e, p. 206) The good advisor is not simply someone who is better positioned (presumably through life experience, or through keen sensitivity and powers of imagination) to know that some of the actions contemplated by an advisee might be, say, INCONSIDERATE, MEAN, DISLOYAL, 21 Here I have in mind (Williams 1995e), (Williams 1996b), and (Williams 2002), Ch. 2. In all three cases Williams acknowledge his debt to (Craig 1990). Craig’s book is a pragmatic account of the concept KNOWLEDGE, which considers how a group of investigators, given certain assumptions about human nature, human powers of perception and recognition, and our need for cooperation, would come to develop such a concept. 22 I take it that anyone can offer their opinion as to what someone should do, but that advice always aims to be good advice, that is, it is something that is always geared towards bettering the receiver’s position, and is thus always attuned to their cares and concerns. 175 INSENSITIVE, GENEROUS, and so on. She is also someone who can translate this into how the advisee’s choice of action might go better or worse for her, because the advisee’s proposed actions would have or lack these qualities. She would, that is, be adept at tracing out the social expectations which would follow from its being true that her act was (say) inconsiderate or disloyal. It would be an understanding of the social cost or benefit to her life from acting one way or another: the personal relationships that would be irreparably hurt or lost by an act of infidelity, the respect with colleagues that would be earned by being honest even if the truth is embarrassing, the loss of trust that one would face from coworkers who would view one’s crossing the picket line as deeply disloyal, or, at an extreme, the complete ostracism that one would have to endure from the entire community by acting in such a cruel manner. The idea, then, is that good advice is an opinion grounded in a concrete experience of a social world as to what sorts of choices will make a person’s life better or worse in that world. Of course, the notion of a good advisor does not require that the advice be given in thick concepts; and we can certainly make sense of being a good advisor without the aid of thick concepts. What thick concepts provide is a shared conceptual apparatus between the advisor and advisee for tracking the social implications of the advisee’s actions. This brings us back to question of what sense is to be given to the idea that thick concepts help those who possess them to navigate or ‘find their way around’ 23 the particular social world they are a part of. The idea that thick concepts provide such a form of ethical knowledge is tied to the fact that their possession enables one to lead a good life within that particular social world. Such concepts not only demarcate various 23 (Williams 1985, p. 151) 176 ways in which one can behave poorly, they also demarcate ways in which a person can be recognized as honorable or principled, the fit object of admiration and respect, a good friend or colleague, and so on. As such, possession of a battery of thick concepts (both positive and negative) will enable a person to lead a life that is seen as worthwhile from the standpoint of that community. Although the knowledge provided by thick concepts will be decidedly local, this is not a barrier for someone outside the community being able to understand these concepts or the ethical outlook which is presupposed in their use. Nor is it a barrier to an outsider coming to understand what it is that they know in using them. Coming to understand what it is that they know, however, is not the same thing as knowing what they know. A.W. Moore offers an analogy relating to historical knowledge would be helpful here. 24 There are certain things that were once known but cannot now be known: for instance, that no one has not yet flown faster than the speed of sound, or ran a mile in under four minutes. Someone, say in the early 20 th century, could be credited with knowing these things. These things are not, however, judgments that anyone can be credited with knowing now. But, even though I cannot know that no one has flown faster than the speed of sound or ran a mile in under four minutes, I can still understand what it is for someone to know this. I can understand how it could be a piece of knowledge, even though it is not a piece of knowledge that I can possess. Similarly, while an outsider is debarred from knowing what the user of a thick concept knows, this does not entail that such a person is thereby debarred from seeing judgments that use such a concept as (sometimes) amounting to knowledge. This is 24 (Moore 2003, p. 349) 177 central to what Williams called the ethnographic stance, 25 and the idea that there can be detached, yet sympathetic outsiders, who are capable of developing a finely tuned understanding of a group’s thick concepts. Moore renders this point nicely. While someone might actually be unable to share in the ethical knowledge of an alien culture, they might still be able indirectly articulate what it is that they indeed know (that is, describe and explain that knowledge without using the thick concept that they use to express it): We can learn enough, about them [the members of some alien culture], about their culture, about the history of their culture, and perhaps also about Homo sapiens, to be able to see how their use of various thick ethical concepts enables each of them to live, with others, in a particular social world; and we can learn enough about the particular circumstances in which they find themselves to be able to see that these circumstances warrant their being these judgments, using these concepts. But although what we thereby learn will in some (non-trivial) sense entail the truth of their judgments, we need not use any of the relevant concepts either in learning it or in subsequently recounting what we have learned. (Moore 2003, pp. 348-49) Just as one could describe the situation of someone prior to say October 1947 (for the sound barrier) or May 1954 (for the four minute mile), understand the epistemic position of someone living in those times, and see that claims like ‘no one has broken the sound barrier/ran a mile in less than four minutes’ would be true when made at the appropriate point in time, and thereby understand that such claims amounted (at that time) to knowledge, one could make a similar attempt at understanding how certain claims couched in thick concepts amounted to knowledge (from the perspective where the use of such concepts is assumed). The idea is that it is open for an outsider to develop an understanding of some alien social world (and its thick concepts) that shows (1) how the 25 (Williams 1986, p. 204) 178 use of those concepts enables them to navigate their social world, and (2) how, from the perspective of those concepts, circumstances within that social world might make some applications of them correct. This would be enough, as in the temporal case, to show that the natives (or insiders) knew certain things in applying them, and that this knowledge counted as a form of ethical knowledge (i.e., a form of knowledge that helped to guide their deliberations in that social world). Such indirect vindication of their knowledge as knowledge, however, would be consistent with (3) being debarred from using their thick concepts, for (ex hypothesi) they are not the outsider’s concepts, or (4) developing some neutral surrogates for them that would allow the outsider to know what the insider knows (for, as Williams notes, thick concepts are not devices for “dividing up in a rather strange way neutral features of the world” 26 ). In this sense, the claim to knowledge can be vindicated in a quite limited way—it can be represented as knowledge from a standpoint that is external to those who actually possess it. Such knowledge can thereby be shown to be reasonable and appropriate for the sort of social world that the insiders actually occupy. 27 Given that we have provided some substance to the idea that thick concepts can provide their users with a form of ethical knowledge, insofar as these concepts are tied to specific criteria and are related to ethical rationale and certain ethical interests, the question that remains is whether we should think of this as sufficient for constituting a 26 (Williams 1985, p. 142). See also (Williams 1995d, p. 187). 27 Moore labels this strategy “indirect endorsement,” where the knowledge in question is endorsed even though it is not knowledge that can be shared from the position of endorsement, and distinguishes this strategy from that of direct endorsement, where the endorsement would involve an explanation that justified one’s use of a certain body of thick concepts. Moore, I think, is rightly skeptical of direct endorsement. See (Moore 2003, pp. 346-47). For skepticism about the indirect approach see (Thomas 2006), Ch. 6 and (Thomas 2007, pp. 54-56). 179 form of objective knowledge. Williams’ did not think that it was, and, because of this, such knowledge was best considered nonobjective, but much of his skepticism about this hinges on what he took objectivity to amount to. In the next subsection will attempt to explicate Williams’ position, and defend it from some mischaracterizations. (b) Objectivity and the Absolute Conception of the World We might understand one aspect of the previous discussion in the following way. On Williams’ view, the concepts of TRUTH and KNOWLEDGE, as they might be understood to apply to something as uncontroversial as scientific beliefs, can also apply to ethical beliefs, insofar as they involve the use of thick ethical concepts. Yet, although both truth and knowledge are equally available in both science and ethics, what this amounts to in these areas turns out, for Williams, to be very different. Thick ethical knowledge is knowledge relating to a particular social perspective (or social world), with a particular history, and grounded in a specific set of concerns and interests that are constitutively related to that perspective. Scientific knowledge, by contrast, is knowledge of the world as it is anyway, independent of any particular perspective on it, and as such is not grounded to one perspective on it. In other words, scientific knowledge, for Williams, is objective knowledge, or knowledge of an objective subject matter. Knowledge under thick ethical concepts is not; it is not knowledge of the world as it is anyway. In drawing a distinction between science and ethics, Williams wants to preserve the idea that there is something distinctive and important about scientific knowledge. In claiming there can indeed be ethical knowledge, one might think that this requires lowering the bar as to what constitutes knowledge, and as a result be forced to give up on 180 the idea that there is any distinction to be had at all between scientific and ethical truths: on such a view, the gains made in granting that there is ethical knowledge entail a loss in seeing something distinctive about scientific knowledge. The argument that Williams provides aims to uphold some distinction between the two, and this argument centers on what he calls ‘the absolute conception of the world’ 28 : a conception of the world as it is anyway, independent of the peculiarities of human experience. What Williams means by this is complicated, and I feel, generally misunderstood, largely because he is trying to capture certain realist intuitions by refocusing the debate over what realism in science amounts to. But, before we get to this, there needs to be some ground clearing. Williams’ argument is connected importantly with the idea that representations 29 can be more or less perspectival. To get a handle on this idea, Williams argues that it is best to “concentrate not in the first instance on what our beliefs are about, but on how they represent what they are about.” 30 Williams’ focus on ‘how’ representations 28 (Williams 1985, p. 139). See also (Williams 2005 [1978]a, pp. 196-97), where the notion of an absolute conception of the world is first introduced. Importantly, Williams’ discussion of knowledge in (Williams 2005 [1978]a) concerns knowledge of the world as it is anyway and not knowledge per se. He writes: “if knowledge is what it claims to be, then it is knowledge of a reality which exists independently of that knowledge, and indeed (except for the special case where the reality known happens itself to be some psychological item) independently of any thought or experience. Knowledge is of what is there anyway.” (Williams 2005 [1978]a, p. 48) Interpreting this quote as intending that knowledge can only be of an objective subject matter would be deeply uncharitable, given that he sought to defend the possibility of nonobjective knowledge in (Williams 1985). For a reading that opts to see a tension here in Williams’ views see (Moore 1991). Moore has since changed his position on this, see (Moore 2003) and (Moore 2007). 29 Williams treats the notion of a representation in the following way: Suppose A and B each claims to have knowledge of the world. Each has some beliefs, and moreover has experiences of the world, and ways of conceptualizing it, which have given rise to these beliefs and are expressed in them: let us call all of this together his representation of the world (or part of the world). (Williams 2005 [1978]a, p. 49) 30 (Williams 1985, p. 138). In terms of a distinction between sense and reference, Williams is encouraging us here to drop concerns about reference and focus exclusively on sense. 181 represent is to highlight the idea that a representation always depends upon a certain mode of conceptualization. The manner in which a representation represents something is determined by (or structured by) the concepts that are employed in that representation. One representation is more perspectival than another, then, from the fact that the concepts deployed in one are more perspectival than the concepts deployed in the other. For instance, our secondary-quality concepts (e.g., RED, HOT, SAVORY) depend crucially on the particular sensory-modalities that we have, and the characteristic experiences associated with those modalities. Because of these dependencies, such concepts (and the representations that they figure in) will be opaque to rational investigators which completely lacked the relevant sensory-modality. Such creatures would be unable to have these sorts of representations. By contrast, concepts like MASS, FORCE, and ACCELERATION involve no distinctive connection to any particular sensory modality or specific type of experience. As such, the representations in which they figure would be intelligible to a wider group of rational investigators: what a creature would have to be like in order to have these representations is not constrained (or as constrained) by their being able to perceive or experience things in a distinctive way. What will be important for Williams’ argument is that these latter sorts of concepts, or something similar to them, might be intelligible to the widest group of rational investigators possible. Simon Blackburn offers an interesting example to capture this difference. While the scientific concept of heat is that of molecular kinetic energy (i.e., molecular motion), it is not the concept that figures in our everyday representations of such things as drinking a cup of fresh coffee or sitting in a sauna: “someone feeling heat usually does 182 not feel it as motion.” 31 These everyday representations relate to a concept that has certain deeply engrained phenomenological aspects. Because of this, finding such a concept (or the representations that employ it) intelligible requires being able to understand these sorts of characteristic experiences, which, in short, means being able to have them. In this sense, the everyday concept HEAT E differs in important ways from the scientific concept HEAT S : HEAT S might be intelligible to some rational investigators, say some extra-terrestrial, even though, for this creature, HEAT E will remain opaque, because this sort of creature lacks the ability to have the requisite sorts of experiences. In this sense, HEAT E is a more deeply perspectival concept that HEAT S . Generalizing this point, the idea that a concept can deeply perspectival, then, is tied to how much the possession conditions of that concept constrain the range of rational creatures that are able to find it intelligible. 32 The more perspectival a concept is, the more its possession depends upon such a creature possessing certain distinctive (i.e., non- universally shared) characteristics. Such characteristics can range from the possession of certain sensory modalities to certain affective and emotional sensibilities or the ability to have certain concerns and interests. 33 Thick ethical concepts can be understood as perspectival in these latter senses: finding a thick ethical concept (and the representations that employ it) intelligible depends upon the ability to have certain affective capacities, such as the capacity to experience certain emotions or the ability to identify with certain ethical concerns and interests (though not necessarily the ethical concerns and interests 31 (Blackburn 2009, p. 12) 32 For the notion of a “possession condition” see (Peacocke 1992). 33 For the latter, see Williams’ discussion of Wittgenstein in (Williams 1981d, p. 157). 183 that are directly related to a given thick concept). For instance, our above extra-terrestrial might find concepts like MASS and HEAT S , as well as concepts like HEAT E , RED and SAVORY, intelligible, and yet find concepts like CRUEL and KIND opaque, on account of being unable to recognize discomfort and suffering in others (say because they have a particularly acute form of, what we would call, mind-blindness, or the inability to naturally ‘read’ the thoughts and feelings of others). Because thick ethical concepts are deeply perspectival, Williams was suspicious about the idea that they hold any strong claim to objectivity. Importantly, however, this suspicion centers on how we should understand the sort of knowledge that these concepts could be thought to provide, and not on whether such concepts are capable of providing a type of knowledge at all. 34 Rather, it centered on how we would have to explain the fact that they had knowledge: in the case of thick concepts, a group’s convergence on a certain thick ethical belief will have to be explained (at least in part) by reference to certain distinctive features of that group (e.g., their possession of certain distinctive ethical concerns and interests). When it comes to scientific judgments, however, there exists the possibility of explaining a group’s convergence on a specific belief that does not have to make reference to any distinctive features of that group. 35 In such a case, their convergence on that belief would be explained by the fact that such belief represents the world as it is anyway, independent of the more distinctive ways that such a group 34 See here (Williams 1985, p. 148): “various members of the society will have knowledge, when they deploy their [thick] concepts carefully, use the appropriate criteria, and so on.” See also (Williams 1985, p. 139): “We can know things whose content is perspectival.” 35 Importantly, Williams argument does not turn on whether there will actually turn out to be such convergence in either ethics or science. See (Williams 1985, p. 136). The issue turns on how such convergence will have to be explained if it does occur. 184 might represent that world or any other group’s distinctive (or more peculiar) take on it. That such a possibility exists with scientific knowledge, but not thick ethical knowledge, marks a contrast between science and ethics (and has important consequences for the latter). And so, while some might rest content with the thought that establishing applicable notions of TRUTH and KNOWLEDGE in ethics is sufficient for defending a type of ethical objectivism, 36 Williams’ aim is to defend the idea that there is a sense in which claims of objectivity go beyond these notions: namely that an objective subject matter is one that will be open or available to any rational investigator. Williams’ critique of the idea that ethical thought is objective is, of course, highly controversial, and has been criticized by a number of philosophers. Unfortunately, most of these criticisms rest on something of a caricature of his views, as they are liable to interpret him as attempting to develop a type of metaphysical realism: an account that identifies what exists or what is really ‘out there’ in the world. Some argue that his position depends on (as John Tasioulas claims) “a metaphysically incredible account of truth and of the relation of mind to world” 37 or as resting on a conception of the world as it is anyway that is (as Hilary Putnam puts it) “incoherent,” 38 or as ultimately involving the idea (as John McDowell supposes) that such a conception “is simply dictated by the world itself.” 39 Since these objectors, in one way or another, assume that Williams’ aim is to develop a form of metaphysical realism through his appeal to the absolute 36 As an example of someone who takes such an approach, see (Wiggins 1991). 37 (Tasioulas 1998, p. 396) 38 (Putnam 1990, p. 171) 39 (McDowell 1986, p. 380) 185 conception, they reject his skepticism about the objectivity of ethical knowledge out of turn. Most of Williams’ commentators, however, fail to appreciate the fact that he is not interested in providing any version of metaphysical realism (though he is certainly interested in trying to make sense of certain realist intuitions). His argument centers on how we should understand different forms of knowledge, and is not directed to the question of what exists or is ‘really’ a part of the world. 40 What I think has thrown many of Williams’ commentators are his uses of expressions like ‘absolute’ and ‘the world as it is anyway’, in saying that it is possible to develop an ‘absolute conception of the world’ or a ‘conception of the world as it is anyway’. The former expressions certainly have a metaphysically realist feel, at least insofar as they are expressions one would take such a theorist to trade in. But such a reading of Williams’ position is deeply at odds with its explanatory ambitions. To get to this point, however, more needs to be said about what Williams takes the absolute conception to amount to. Let’s begin with the idea that we have already developed, namely that some representations of the world can be distinctively perspectival. The example that Williams gives is of two people who have different representations of the world, and that they both claim to constitute knowledge: say two people looking at an object from different spatial positions. 41 If that is right, then there must be some conception of the world (some further way of representing the world) that is able to explain how these people have the 40 In making this point, I find Williams refusal to engage a debate over the issue of realism in his (Williams 1996b, p. 21) interesting. 41 (Williams 2005 [1978]a, pp. 49-50). See also (Moore 1997, pp. 68-74) and (Moore 2007) for a take on Williams’ views that places him more closely to concerns about realism than I do. 186 representations that they do, and also how those representations constitute knowledge. The sort of perspectival ascent relates their distinctive representations to a single world, but in doing so it must employ concepts that would represent the world devoid of what made those particular representations distinctive. In the case of the two people looking at the same object from different spatial positions, the conception that would include those representations would have to be one that (at least) included both of those positions within a larger coordinate field. Where the previous representations represent the world in a certain distinctive way—namely one arising from their particular perspectives on it— the conception which comes to explain those representations is one that represents the world in a way that can be taken as shared between those perspectives. Such a conception is of a world that is shared between those more distinctive perspectives in at least this way: the representations that make up that conception are not constrained by whatever features which make the perspectival representations contained in it. If this process is continued, then we can see how any particular perspectival representation, which claims to be knowledge of the world, must be capable of being coordinated with any other perspectival representation, which also claimed to be knowledge of the world, if those claims to knowledge are indeed valid. Such a conception of the world would be one that was able to coordinate all claims to knowledge of that world in a way that would not invalidate those claims to knowledge: that is, such a conception would be able to explain (or at least make some explanation possible of 42 ) how each of these representations indeed constituted the sort of knowledge they purport to constitute: knowledge of the world. Such a maximal conception, one which was able 42 (Williams 2005 [1978]a, p. 230) 187 to carry out this sort of explanatory work, is what Williams means by “the absolute conception of the world.” 43 It is a conception of the world that would be able to “nonvacuously explain how it itself, and other various perspectival views of the world, are possible.” 44 Consider the following. One person drops a sugar cube into a class of water and watches it dissolve, another person tastes the same water and experiences a sensation of sweetness. Here we have two representations of the world: one that the sugar dissolved in the water, the other that the resulting water tasted sweet. In order to coordinate these representations together, as representations of the same world, we would need a single conception of the objects and individuals in question which would show how these representations (or experiences) were possible. In the first case, the explanation would involve the molecular properties of the sugar cube and the water, and the fact that, upon coming into contact with the water, the molecules that make up the sugar cube bind with various water molecules and thus break apart from one another, thus dissolving the cube into the water. In the second case, the explanation again involves the molecular properties of the sugar cube and the water, and how their coming into contact with the second person’s taste buds causes the release of certain chemicals in this person’s brain which account for the sensation of sweetness. If I understand Williams correctly, this is a gloss of the sort of non-vacuous explanation that an absolute conception would enable: it would provide a conception of the world as one involving certain molecules and chemical reactions, and in virtue of a world so conceived it would be able to show (or enable an 43 (Williams 1985, p. 139) 44 (Williams 1985, p. 139) 188 explanation that would show) how other representations indeed constituted pieces of knowledge. The absolute conception’s ability to enable these sorts of explanations would be tied to the fact that the manner in which it represents things would be minimally perspectival: it would constitute (what we might say is) a maximally impartial view of the world. The manner in which it represented the world would be devoid of any of the peculiar (or distinctive) features of those perspectives that it was aiming to coordinate and explain. Because of this, although the absolute conception will be one that is expressed using various concepts, the idea that such a conception itself constitutes a certain (peculiar) perspective on the world drops out completely. 45 That this is so helps to explain why Williams shifts from talking of it as consisting of representations that are minimally perspectival to its being nonperspectival: “It [the absolute conception] will be a conception consisting of nonperspectival materials available to any adequate investigator, of whatever constitution.” 46 Since the world will be represented in such a way that the conception of it will be available to any adequate investigator, there is no explanatory room left for applying the term ‘perspective’ to that conception. If this concept could do explanatory work at all here, it would have to be intelligible to say that 45 In this, I think Williams’ discussion of the work of the Later-Wittgenstein is instructive. See here (Williams 1981d, p. 161), and his skepticism about the idea that a linguistic form transcendental idealism— the sort of transcendental idealism that would be attributable not to particular subjects or to the entire group of creatures capable of language—works with a general conception of language which cannot intelligibly be thought of as constituting a limit or constraint on our world. (cf. (Wittgenstein 1974 [1922] 5.6 and 5.64). Williams’ view, I take it, is that this general view of language cannot constitute ‘a limit on the world’ because there would be nothing to contrast it with. In a similar fashion, a minimally perspectival conception of the world cannot be treated as perspectival, for it cannot be thought of as one perspective among others. 46 (Williams 1985, p. 140) 189 this conception was one take on the world among others, and that is precisely what is being denied in saying that such a conception is minimally perspectival. And so, while this conception of the world, as one that is maximally “independent of our perspective and its peculiarities,” 47 will, of course, be a conception, and thus will depend on concepts and language for its expression, this fact, does not bring any substance back to the idea that such a conception is itself a perspectival way of representing the world—for there is no alternative that we could point to which would make this notion distinctive. Since the notion of an absolute conception is (and has been) liable to invoke certain misleading comparisons to metaphysical realism, it would be helpful to contrast it with Hilary Putnam’s distinction between external and internal realism. 48 This will help us to set aside a misunderstanding of how the absolute conception should be understood. At the outset, it is worth mentioning that Williams openly adopts a minimalist view of truth, 49 whereas Putnam prefers to consider the difference between internal and external realism in terms of different understandings of the truth predicate. External realism is the view that the world’s structure is wholly independent of our minds, whereby our various concepts and representations are taken to connect to or correspond with an unconceptualized reality. As such, external realism exemplifies a traditional understanding of truth as correspondence with reality: judgments are true because they correspond to certain mind- and language-independent facts. Putnam’s central criticism of this view is that it amounts to nothing more than an empty picture: 47 (Williams 1985, p. 139) 48 See (Putnam 1987) 49 See here (Williams 1996b) and (Williams 2002, pp. 63-66) 190 “[t]he only access we have to the world is to the world as it is represented in thought and language.” 50 The external realist looks to treat the world as it is anyway as something that is independent of thought and language, yet structurally mirroring a certain order of thought and language. What makes this an empty picture, according to Putnam, is that it asks for the world to contribute a structured set of objects and states of affairs, whose structure could only be provided by the aid of concepts and language. (This sort of unintelligible demand is what, I think, prompts Putnam to criticism Williams’ view as “incoherent” 51 ; Putnam, by my lights, interprets Williams as simply offering an updated version of external realism.) By contrast, the internal realist (Puntam’s preferred position) attempts to avoid the sort of incoherent disconnect between our linguistic practices and some unconceptualized reality exhibited by external realism by denying that there is any gap at all between the world as it is anyway and how we represent that world. The internal realist holds that, although the world is causally independent of our concepts and representations, its structure is (or is at least partly) a function of them. On this view, there is no attempt at making the problematic move of thinking that a judgment somehow steps outside our linguistic practices and connects with some part of an unconceptualized reality in order to treat that judgment as objectively true. Rather, all our true representations are placed on equal footing as objectively true until some practical worry or problem causes us to doubt that status (i.e., the thought that our concepts might fail to map onto the world is seen, from the standpoint of an internal realist, as an illusory concern, held over from thinking 50 (Putnam 1994, p. 294) 51 (Putnam 1990, p. 171) 191 about realism as the external realist does). Truth, on this view, turns out to be something like idealized rational acceptability, 52 where a statement is true if it is warranted under epistemically ideal conditions. In this way, a conception of the world as it is anyway is one that is provided by the best conceptual practices we have, although it remains open for us to revise what counts as ‘best’, and continually improve upon. Given this brief summary, it is best, I think, to see the absolute conception as amounting to neither external nor internal realism, but as an explanatory alternative to the dichotomy itself. Certainly, Williams’ absolute conception is intended to capture at least some of the intuitions that drive various realist views, but, as I noted above, Williams is not interested in providing a version of realism. Both internal and external realism are interested in providing, what we might call, an inventory view of reality: they are aimed at explaining how mind, language, and the world come together so that we can not only understand how there are things like trees, planets, carbon molecules, and mitosis, but also how there are things like tables, cars, basketballs, elections, and murders. To this, each aims to provide an account of how various representations relate to the world: in the external realist’s case, they relate by corresponding to the world, in the internal realist’s case, they relate by partially constituting that world. Williams’ explanatory aims are different. He is not interested in providing a way of taking stock of what exists or what is ‘really’ a part of the world. Rather, Williams is interested in providing an account that explains how it is that we have the various pieces of knowledge that we think we have. To do this, such an account will have to take into consideration the fact that some of the representations that constitute knowledge depend peculiarly on our distinctive nature, 52 (Putnam 1981, p. 55) 192 whereas others do not. In the case of representations that purport to represent the world as it is anyway, then, if such representations do indeed provide us with knowledge, then there will be a conception of the world which is able to explain how this is possible. Such a conception will be one of the world that is devoid of our particular (or peculiar) take on it: it will be a conception of the world that will be available to any other rational investigator of that world, even if we or they never actually develop it. Given this, we can set aside some of the misconceptions that have been placed against Williams’ position. First, there is no claim that such a conception is to be seen as validated by something external to it: it is a conception that would be developed by certain rational investigators in order to meet certain interests in understanding how they have knowledge of the world as it is anyway. If we were ever to actually develop one in a completed form, such a conception would “itself admittedly [be] a human artifact.” 53 It is also not the incoherent notion of being a conception of a reality that is somehow brought off without the use of concepts, 54 nor is it a conception drawn from some mythical perspectiveless position. Conceptions always invoke concepts, and concepts are always some group’s concepts: on this point, the absolute conception is no different. 55 53 (Stout 2001, p. 55) 54 Putnam and Rorty both qualify as individuals who understand Williams’ position in this way. See (Putnam 1990, pp. 170-74), esp. 173; and (Rorty 1994), esp. 22. Williams, however, is quite clear about his not being seen in this way, when he writes: My idea was not that you could conceptualize the world without concepts. The idea was that when we reflect on our conceptualisation of the world, we might be able to recognize from inside that some of our concepts and ways of representing the world are more dependent than others on our own perspective, our peculiar and local ways of apprehending things. In contrast, we might be able to identify some concepts and styles of representation which are minimally dependent on our own or any other creature’s peculiar ways of apprehending the world. (Williams 2006a, p. 185) 55 See (Williams 2006a, p. 185): “…my aim in introducing the notion of the absolute conception was precisely to get round the point that one cannot describe the world without describing it, and to 193 That the absolute conception will require the use of certain concepts, however, does not undercut its claim to absoluteness, so long as the concepts deployed in developing it do not depend for their intelligibility on possession of certain the distinctive (i.e., non- universally shared) characteristics. 56 The absolute conception, as I noted at the outset of this section, was intended to help draw a contrast between science and ethics. Here we can lay out one half of the contrast. In the case of science, there is the possibility that we might arrive at a conception of the world that is in no way wedded to our distinctive nature as human beings. How it would conceptualize the world would be intelligible to any rational investigator of it, in virtue of the fact that the concepts it would employ could be intelligible to them without their having to have a distinctive (i.e., peculiar to them) nature or experience the world in any particular way. Such a scientific conception of the world would be from no perspective in particular: it would be, as it were, an impartial view of the world. As such, when we aim to explain how it is that we converged on a certain scientific judgment as true about the world, our explanation of this would be able to invoke the very scientific conception of which that judgment was a part. And so, the best explanation that we would have as to why that convergence took place would be that accommodate the fundamentally Kantian insight that there simply is no conception of the world which is not conceptualized in some way or other.” Cf. (Williams 2005 [1978]a, p. 244). 56 Here we might compare Williams’ view of the absolute conception with that of Thomas Nagel’s view of objectivity in (Nagel 1986, p. 5): “A view or form of thought is more objective than another if it relies less on the specifics of the individual’s makeup and position in the world, or on the character of the particular creature he is.” Note that for Nagel the notion of objectivity comes in degrees (i.e., one form of thought can be more objective than another). Williams’ absolute conception, then, would be something like a limit case of Nagel’s view of objectivity. 194 such a judgment represented the world as it is anyway. 57 The contrast that is to be found, then, between scientific knowledge and ethical knowledge, is that the former, but not the latter, is knowledge of the world as it is anyway: a single world that is shareable by all those who investigate it. (c) Vindicatory Histories and the Ascription of Error Closely related to Williams’ view that science is capable of developing an absolute conception of the world is the idea that science can also be seen as involving a certain sort of historical progress. This provides us with another contrast between science and ethics: scientific thought can be represented as having a vindicatory history, ethical thought, at least as far as that thought involves thick concepts, cannot. In this way, Williams’ view of scientific development might be compared to C.S. Peirce’s view of objective truth as what is finally arrived at through an idealized process of inquiry: 58 All human thought and opinion contains an arbitrary, accidental element, dependent on the limitations in circumstances, power, and bent of the individual; an element of error, in short. But human opinion universally tends in the long run to a definite form, which is the truth. Let any human being have enough information and exert enough thought upon any question, and the result will be that he will arrive at a certain definite conclusion, which is the same that any other mind will reach under sufficiently favorable circumstances. (Peirce 1871, p. 455) On a Peircean view, objective truth is the opinion that a group of investigators would inevitably come to agree upon given that they had unlimited time, resources, and so on. 57 Cf. A.W. Moore’s comment on how explanations of scientific knowledge can be vindicatory: A good reflective explanation for someone’s having a given item of scientific knowledge can make use of the very concepts exercised in the knowledge, and so can straightforwardly vindicate the knowledge, by revealing that the person has come by the knowledge as a result of being suitably sensitive to how things are. (Moore 2007, p. 26) 58 See (Peirce 1931, pp. 406-07): “The opinion which is fated to be ultimately agreed to by all who investigate is what we mean by the truth, and the object represented in this opinion is the real.” See (Hookway 1985) for a general discussion of Peirce’s views, and (Hookway 2004) for a discussion of the comparison of Williams’ absolute conception to Pierce’s views on objective truth. 195 This view was importantly connected, not to some mysterious notion of judgments corresponding to an unconceptualized reality, but to the idea of removing the various natural errors and idiosyncrasies that are an inevitable part of any human investigation. In this way, the image that Peirce presents us with, of objective truth as something like the ‘final opinion’ or ‘last word’, can be taken as a characterization of the process at which we might come to develop an absolute conception: later stages in the development towards an absolute conception can be represented as correcting for various errors, idiosyncrasies, and distortions that were present at earlier stages. 59 Central to this picture would be the possibility of later stages being able to provide a non-vacuous, non-question begging, theory of error directed towards earlier stages. Such a theory of error would not be geared towards explaining how particular judgments were wrong, but rather it would be concerned with revealing something missed or misunderstood at that earlier stage. As such, it would be a theory of error that would be directed towards the very manner in which that earlier-stage conceptualized the world; it would be a theory of error that revealed something erroneous or problematic in the concepts the earlier stage was using. Such a theory of error would only be possible, however, because both earlier and later stages could be characterized as both offering attempts at providing an absolute conception of the world: having such a conception as a common goal provides the conceptual space for finding one stage in the development of such a conception defective in relation to another. The possibility of providing such a non-vacuous, non-question begging, theory of error marks a distinction, then, between 59 See (Williams 2005 [1978]a, p. 229). 196 objective and nonobjective subject matters: an objective subject matter is one where such a theory of error is genuinely possible, with nonobjective subject matters it is not. Here it will be helpful to focus our attention on an example. Take the difference between Aristotelian and Galilean conceptions of motion. 60 On the Aristotelian conception, the continuous movement of an object was explained by the continuous application of some force on that object. According to this view, objects moved only as long as a force was being applied to them, and so, the continuous movement of an object required there to be some continuous force which was propelling it along. Now consider an arrow shot from a bow. The arrow continues to travel in a certain trajectory after leaving the bow. On the Aristotelian conception, the arrow’s continuous movement meant that there must have been some force behind the arrow propelling it along its trajectory. This proved to be a problem for such an account, as there was no readily obvious force that could be observed moving the arrow along. To account for this, various explanations were offered, for instance, that the arrow was being propelled forward by a ‘vacuum’ which was created behind the arrow when it was released from the bow. Only after this ‘vacuum’ had dissipated would the arrow cease to move and fall to ground. Such an explanation made it difficult, if not impossible, to predict the duration and trajectory of things like arrow shots. By contrast, the Galilean conception of motion was able to explain why arrows continued to travel after they had left the bow, without having to appeal to mysterious ‘vacuums’, and these explanations proved to be better at predicting such behavior than 60 I appreciate John Dreher for pointing out this particular example. For an excellent discussion the Galilean conception of motion, see (Damerow et al. 2004), Ch. 3. 197 the Aristotelian conception. What made the Aristotelian conception problematic (or defective) was that it was a conception ignorant of the law of inertia, the idea that an object would maintain its velocity and direction as long as no other force (such as friction) was applied to that object. The same phenomena could be explained, described, and predicted, in a clearer, more precise fashion using the Galilean conception of motion than the Aristotelian one. Beyond being able to characterize the same sorts of phenomena more accurately, the Galilean conception also provided the means for developing a theory of error by locating something that was missed (or misunderstood) on the Aristotelian conception. By being able to provide such a theory of error, the Galilean conception of motion can be understood as an advance over the Aristotelian conception. What is important for us here is coming to see how the ability to provide an error- theory relates to the development of an absolute conception of the world. The absolute conception would be the final product of a process of enquiry, and it would be one that was able to provide an error theory with respect to prior stages of that enquiry, but it would also be able to produce error theories for alternative conceptions that were not a part of the same process of enquiry. It is capable of doing this because the same regulative goal (of arriving at an absolute conception of the world) can be something which not only different historical periods of one tradition of enquiry can be seen as aiming to produce, it can also be seen as something that different, independent traditions of enquiry were aiming to produce. In this way, certain human scientific conceptions might be capable of providing theories of error for various alien conceptions of the same 198 phenomena (or vice versa) provided we were able to understand those theories. The idea that certain conceptual practices from diverse conceptual perspectives can be seen as aiming towards one and the same goal, then, provides a point of comparison, and thus opens up the possibility of finding one defective relative to the other, or else, finding both to be “incompatible, but equally legitimate conceptualizations” 61 of the same phenomena. In reaching the former conclusion, we are simply stating that a theory of error could be offered for one of the two accounts. In reaching the latter, we are saying that, given the goals implicit in each conceptual practice, and the explanations, descriptions, and predictions made by each conceptualization, there is nothing that would impugn one account over the other. Here, such alternative conceptions would be, as Quine put it “empirically equivalent” 62 : whatever observations counted for (or against) one would count for (or against) the other, but the concepts used in each would be irreducible to one another, or the concepts of some further conception. 63 The upshot to this is that objective subject matters are those where alternative conceptions of those subject matters are capable (at least in principle) of being found problematic, where a mistake or misunderstanding in that conception could be revealed by a theory of error. The ability for one conception to ascribe this sort of conceptual error to another conception, however, depends upon both conceptions having certain general implications follow from what they are conceptions of. In the example given 61 (Moore 2007, p. 27) Moore notes that although Williams’ label of ‘the absolute conception’ implies that there would be only one, his account does not rule out the possibility of there being more than one. 62 (Quine 1990, pp. 96-7) 63 For an excellent discussion of the comparison (and contrast) of Williams’ views on the absolute conception to Quine’s, see (Moore 2007, pp. 29-32). 199 above, of the Aristotelian and Galilean conceptions of motion, the general implication shared by each account was that it was aiming at providing (something like) an absolute conception of motion. Each was offering an account of motion that was thought to be the right or correct account of this concept. Because of this, these two differing conceptions could be brought into conceptual contact with one another, thereby allowing for the possibility that one could be in error relative to the other. Conceptual practices, no matter how distant from one another in space and time, which both aim at describing the world as it is anyway, always involve such a general implication: each account, in virtue of what it is attempting to do, implies that all other accounts which differ from it 64 are thereby incorrect. To anticipate, this point helps to explain why Williams’ thinks viewing ethical thought, at least insofar as it involves the use of certain thick concepts, as nonobjective is appropriate. A nonobjective view of some particular thick conceptual practice would represent that practice that as being concerned with some other aim than that of providing an absolute conception of ethics. It would rather be seen as simply a practice geared towards sustaining that particular form of social life that it was a constitutive part of. Because of this, it would be inappropriate to think that some non-vacuous, non-question begging theory of error could be provided against the use of the thick concepts that such a practice deployed, and so changes in a social world, constituted by one thick conceptual practice being replaced by another or a thick conceptual practice simply ceasing altogether, could not be taken as involving the same sort of vindicatory history as is the 64 Or, all other accounts that are not empirically equivalent to it and differ from it. 200 case in science. 65 With various scientific conceptions there is at least the possibility of offering an explanation of how a group arrived at their concepts that will justify the continued use of those concepts (i.e., there will be something like a vindicatory conceptual history). With ethical conceptions, at least to the extent that they involve thick ethical concepts, there is no such hope of their involving such a vindicatory history. (d) Recoiling from the Idea that there is Thick Ethical Knowledge Coming to this conclusion, might lead one to recoil from the idea that thick concepts are capable of providing a type of knowledge at all. We should pause here, however, and consider that, in the midst of our discussion of the absolute conception of the world and how science presents us with chance for objective knowledge of a world that is there anyway, that the concepts of TRUTH or KNOWLEDGE play a rather narrow role in the argument for these points. Williams is not claiming, nor should not be read as claiming, that only judgments that are apt to figure in the absolute conception are thereby the only (or even superior) candidates for being true or being known. Williams is clear that “[w]e can know things whose content is perspectival,” 66 and in saying this he should not be read as qualifying the status of that knowledge (as something second-rate, perhaps). We can know that certain things are green or savory and we can know that objects are tables and chairs. In neither case, is the knowledge involved knowledge of 65 Examples one could consider here might drawn from the different catalogue of virtues offered in (MacIntyre 1984), (Confucius 1979), or (Aristotle 1998). For instance, although Aristotle’s general views on virtue are directed towards the idea that they enable a good life for human beings, the particular list of qualities that Aristotle gives could not be seen as anything but an account of what would enable an Athenian male to flourish. If this is right, then it would be inappropriate to read into Aristotle’s list of virtues the implication that these are virtues that all human beings should have. 66 (Williams 1985, p. 139) 201 the world as it is anyway—this is a world where there are no tables and chairs, but things like molecules, atoms, and quarks. The idea that we can know things whose content is perspectival is not intended to make some claim that what is thereby known is somehow limited or incomplete. Rather it is to point us in the direction of looking at the sort of subject matter that such knowledge claims are about: a human world, where we have certain needs and interests, and certain (perhaps unique) modalities for experiencing the world as it is anyway. Williams is not denying that there really are tables and chairs, or green and savory objects; he is trying relocate the idea of who the ‘really’ applies to. It applies to human beings (or to a larger set of beings that are related to their lives, and others, and the world as it is anyway in manner similar to human beings). The only difference between saying that ‘there really are tables’ and (say) ‘there really are molecules with a certain atomic structure’ is that ‘really’ in each case has a different scope: and this is the distinction that is lost to views that adopt some form of metaphysical realism. Accordingly, the possibility of arriving at an absolute conception of the world should not be thought of as getting closer to an account of what really exists or what can genuinely be known. 67 Williams is also not claiming that the absolute conception is any sense better or more important than the more localized (or parochial) conceptions. To think this would be to place a value on the absolute conception that Williams’ argument does not (and could not) support. Such a view motivates the thought that perspectival judgments are inherently distorted or merely the product of certain prejudicial influences; as something 67 Williams offers an excellent discussion, and clarification, of his views regarding the absolute conception, its possibility, and its relative importance in (Williams 2006a, pp. 184-88). 202 to be corrected through proper reflection (and perhaps the aid of science). 68 And this would stand in tension to Williams’ claim that we can genuinely know things which are perspectival. Here it is important to see the absolute conception as something that only aids in our understanding of one area of human life, and that there are other areas that will require the use of more perspectival conceptions and outlooks. Williams writes: Even if it were possible to give an account of the world that was minimally perspectival, it would not be particularly serviceable to us for many of our purposes, such as making sense of our intellectual or other activities, or indeed getting the most out of those activities. For those purposes—in particular, in seeking to understand ourselves—we need concepts and explanations which are rooted in our more local practices, our culture, and our history, and these cannot be replaced by concepts which we might share with very different investigators of the world. (Williams 2006a, p. 187) Conceptions are (at least in part) responses by certain investigators to their environment, and come to bear on what they do as a way of coping with some issue, answering some question, or making sense of or serving some interest. The absolute conception is no different in this regard: if developed it would be developed because a group of investigators had an interest in coming to understand the world as it is anyway. But this need not be (and likely is not) the only interest that such investigators would have: we are an example of such a group, and our interests are far and away more diverse than simply desiring to understand a world that was not of our making. We should not think that our arriving at an absolute conception would do anything (or much of anything) for us with 68 For instance, see (Crary 2002, p. 380): The idea [behind the absolute conception] is that, just as an accurate determination of the shape of an object must corrected for the way things appear from this or that literal perspective, an accurate description of how things really are in the world must be corrected for the way things appear from ‘perspectives’ afforded by our subjective peculiarities. This quote implies that the fact that perspectival concepts depend on certain ‘subjective peculiarities’ this amounts to those ‘peculiarities’ having a skewing or distorting effect. This would be an odd implication for Williams’ view, given that he is clear we can know things which are perspectival (e.g., that stop signs are red or that lava is hot). 203 respect to the other (more local) interests we have—like those of leading a sentient life, or leading a human life, or leading a human life at some particular point in time, in some particular location, within some particular community that has some particular history. There is, however, a remaining challenge to the idea that we can genuinely have ethical knowledge under thick concepts, and this comes from a contrast that emerges between thick concepts and those concepts whose use can be directly related to the absolute conception. One key difference between concepts that relate to our sensory modalities and thick concepts, is that we can provide an explanation of how we came to have the former that would justify our continued use of those concepts. The representations that employ these concepts put themselves forward as being related to the world in various ways: we do, after all, perceive certain objects, in a rather straightforward way, as having a color we call ‘red’ or a taste that we call ‘savory’. An explanation of why this is might rely on something like evolutionary biology, and the claim that creatures who developed reliable ways of experiencing the world gained an increase in their survival fitness. This is a non-vacuous explanation of why we have come to experience the world as we do, and it is also an explanation which justifies these concepts in what they purport to be: concepts related to various perceptual experiences we indeed have. In the case of thick concepts, however, our explanation as to how we came to have the one’s we have will not look like this. Instead, it will have to make heavy an appeal to a notion that does not figure in the absolute conception, namely culture: and as such, the explanation as to why a group has a certain set of thick concepts could be no more ambitious than an explanation as to how a certain culture, or certain 204 cultural stage, came into being. This sort of explanation, relying heavily on psychology, sociology, anthropology, and history, would not be the sort of explanation that would vindicate the use of those concepts. Such an explanation would not provide us with a reason to continue using them. But, importantly, this fact does not undermine the claim that these concepts provide knowledge. Or at least, it does not do so yet. In order for this to be a challenge to their claim to provide a form of knowledge, we first need to consider whether that form of knowledge is something that we take to be knowledge of a world as it is anyway. And there is nothing in the mere claim that these concepts purport to offer a type of knowledge that forces this conclusion upon us. In saying this, however, there is still the problem that an explanation of how a group came to have some set of thick concepts will fail to justify their use of them. Once this conclusion is reached, the thought that such judgments are still capable of providing us with knowledge might now appear to not be enough. And this is at the very heart of Williams’ skeptical worries about the unstable relationship that exists between critical reflection and the sort of knowledge that thick concepts provide. (e) Nonobjectivism and the Hypertraditional Society Williams relates his concerns about the reflective stability of thick ethical knowledge through the thought experiment of (what he calls) the hypertraditional society: “a society that is maximally homogenous and minimally given to reflection; its members simply, all of them, use certain ethical concepts of this [thick] sort.” 69 The 69 (Williams 1985, p. 142) 205 hypertraditionalists are assumed to meet all of the requisite standards for having ethical knowledge: they conceptualize their social environment in terms of the shared body of thick concepts, they (at least sometimes) apply those concepts carefully, their judgments are differentially sensitive to changing situations, and (at least sometimes) their concepts correctly apply to the situations they apply them. In short, the hypertraditionalists, in using their thick concepts, sometimes make judgments that they know to be true 70 ; that they indeed have knowledge under these concepts is key to Williams’ argument. What makes this society hypertraditional is that its members are minimally disposed to ask certain reflective questions about their thick concepts. The sense of reflectiveness that they lack here is one that relates to general questions about the adequacy or correctness in using them (or some set of them). As such, we are being asked to consider their practice at a point where they are simply not inclined to ask questions, such as: Are our thick concepts the best/right/correct ones to be using? The fact that they are not inclined to ask such questions, however, does not mean that there application of them is necessarily done without reflection: we can assume that they (at least sometimes) take care in applying these concepts, and perhaps reflect a great deal, in some cases, on the question of whether their application is appropriate. What they lack, in being minimally given to general reflection, is any sense of asking questions about the 70 One objection that Williams considers to the idea that the hypertraditionalists have knowledge under their thick concepts is that the concept itself might be defective, thus making the entire range of judgments which employ it systemically false. (Williams 1985, p. 145) Williams is keen not to rule out the possibility, but to show what such a possibility would have to involve. To claim that one of their concepts was defective, we would have to see their use of it as bound up with something like a rudimentary scientific conception of the world. Williams gives the example of magic, as a concept that can indeed be seen as defective in this way, because their practice of using it is to provide certain rudimentary causal explanations: which places what they are doing in line with our scientific conception of causation. Central to Williams’ response to this objection is the idea that some thick concepts might not be engaged in such an endeavor. 206 practice of using the concept itself. Of course, the hypertraditionalists’ collective body of thick concepts and conceptual practices certainly constitutes a point of view on what is and what is not proper (i.e., right) conduct, but we must be careful in how we cash this thought out. While they do not explicitly have the thought that it is a perspective about what constitutes proper conduct for them (that is, they do not consciously recognize the perspectival character of the ethical concepts they are using), we must avoid thinking that their use of these concepts necessarily involves broader and highly reflective answers to the question of whether these concepts are the right or best ones to be using. Williams suggests that, in considering the hypertraditional society, there are two ways that we might interpret the hypertraditionalists’ practice of using these concepts, and here he distinguishes an objectivist and nonobjectivist model for interpreting them (and their thick conceptual practices). 71 On the objectivist model: …we shall see the members of the society as trying, in their local way, to find out the truth about values, an activity in which we and other human beings, and perhaps creatures who are not human beings are all engaged. We shall then see their judgments as having these general implications, rather as we see primitive statements about the stars as having implications that can be contradicted by more sophisticated statements about the stars. (Williams 1985, p. 147) Importantly, what makes the application of this model appropriate, is not related to whether a group typically tries to determine whether some of their judgments using a thick concept are correct or not. This alone does not raise the sort of general reflective implications that Williams takes the objectivist model to be concerned with. What is important is whether their conceptual practices imply, even though they have never explicitly raised these questions, certain views about whether their thick concepts are the 71 Williams notes that these are “different models of ethical practice.” (Williams 1985, p. 147) 207 right or best ones for anyone to have. (Compare this to practices that are engaged in providing an absolute conception of the world: such practices always involve such implications, as they always involve the aim of giving an account of the world as it is anyway.) If the practice of using certain thick concepts has this sort of reflective implication implicit built into it, then this will place their thick conceptual practices in direct conflict with all other ethical outlooks (and their thick concepts). Their use of these concepts will be seen as involving an implicit attempt at providing (what we might call) an absolute conception of ethics, and it would involve the implicit belief that the use of some other set would be wrong or incorrect. This, on Williams’ view, is an implication that we can see in their conceptual practice, even if they have never considered what such an alternative set would look like or involve. It is, of course, indisputable, that differing ethical outlooks are in tension with one another, as different sets of thick concepts can give rise to different practical pressures directed towards one and the same type of act neutrally described. But this tension does not bring with it the claim that these other groups are thereby mistaken, or that their concepts are based on some misunderstanding, precisely because such tensions can be present without their conceptual practices involving such general aims. On the nonobjectivist model, however, a group’s thick conceptual practice will not be seen as having these sorts of reflective implications, and they will not be seen as implicitly trying to develop an absolute ethical conception. On the contrary, with such a view: …we shall see their judgments as part of their way of living, a cultural artifact they have come to inhabit (though they have not consciously built it). On this, 208 nonobjectivist, model, we shall take a different view of the relations between practice and critical reflection. We shall not be disposed to see the level of reflection as implicitly already there, and shall not want to say that their judgments have, just as these stand, these implications. (Williams 1985, p. 147) It is important to note that it is open for us to apply either model to the hyper- traditionalists and their thick conceptual practices. That is to say, there is nothing about their day-to-day use of these concepts per se that would itself determine which model was appropriate. Such a determination will depend on the character of the concept and conceptual practice itself. What is central for Williams is the idea that whichever model we choose to apply, this will have extension implications for whether we should think that they indeed possess a form of ethical knowledge through their use of their thick concepts. If the objectivist model applies, then it will turn out that they do not have knowledge under their thick concepts. Williams’ draws an analogy to other sort of perspectival concept—MAGIC—where their use of it can indeed be seen as having very general implications. 72 Such a concept is, at the very least, involved in offering certain causal explanations (e.g., ‘The reason you are sick is because Old Lady Withers put a curse on you. And that is the reason why I am staying clear of her’). That this is so, however, places such a concept in line with our own scientific conceptual framework, from which we are able to provide a clear error-theory as to why such a conception is based on a misunderstanding; and hence that the entire region of discourse involving this 72 (Williams 1985, p. 146). Gibbard, in (Gibbard 1992, p. 271), puzzles over this notion in saying that there is no room for ‘truth and falsehood’. If I am right in understanding Williams, the whole point of applying a nonobjectivist model to the hypertraditionalists would stem from the fact that their thick conceptual practices are not governed by an implicit goal of providing an absolute conception of ethics, a question that is independent of whether any particular thick judgment is deemed true or false. 209 concept is systematically false. The basis for claiming the concept defective, in this case, is that it can be seen as involved in a primitive attempt at providing an absolute conception of the world. In the case of thick ethical concepts, the application of the objectivist model to their practices would also reveal a systematic error, one which would show that they in fact do not possess ethical knowledge under these concepts. This failure would be that their practices implicitly hold that their thick concepts the right or correct ones to be using. But, as Williams has aptly noted, there is nothing outside these conceptual practices which they could be “true of.” 73 This might seem like an unargued assumption, and a question-begging one at that, if we do not include two additional (but related) considerations. The first is that, with concepts that relate to various forms of objectivist thought, there is the possibility of providing a non-vacuous, non-question begging explanation which would justify the continued practice of using those concepts. The second is that, when it comes to their thick concepts, there is no possibility of providing such an explanation. What would be implied in their use of these concepts, then, would be the requirement of offering an error-theory that they would be unable to provide. 74/75 73 (Williams 1981c, p. 143) 74 To be sure there are (or have been) explanations which attempt to vindicate a scheme of ethical concepts as best, notably Aristotle’s. But, like Williams, I agree that these invoke considerations (like a normative view of biology) that we no longer accept. For a view which attempts to bridge this gap by treating the very concept of human nature as something constitutively related to the social world in which members of a certain group belong, see (Lawrence 1993). The failure of such a project, I think, is revealed in the fact that it fails to explain the untendentious assumption that there are different sets of thick ethical concepts and different social worlds that they apply to. Are we to accept on this basis that there are different types of human beings, each with a nature distinctive of the social world they are a part of? Here I am in agreement with Williams, when he notes that such “neo-Aristotelians are rather too Aristotelian.” (Williams 1996b, p. 33) While we might be able to understand how Aristotle held such a view, our understanding of the world, human beings as a biological species, and the ethical diversity found among groups of this species, is not so innocent. And the fact that it is not makes such views implausible to say the least. 210 Under the nonobjectivist model, however, such general implications are taken to be lacking, and because of this there is no reason to deny that, given the other conditions listed above hold, they have knowledge under their thick ethical concepts. We might also think of Williams’ argument here in a different way. Start with the idea that we are intent on denying that a certain group’s judgments using thick ethical concepts amounted to knowledge. If we are to make good on this denial, then we would have to be able to explain how it is that their use of these concepts is systematically in error; we would have to reveal some defect in their conceptual practice that would license our denial that none of the judgments couched in those concepts ever (in principle) amounted to knowledge. We could only do this if we were able to see their thick conceptual practices as presupposing something which they cannot succeed in substantiating (i.e., the view that their thick concepts are the best/right/correct ones to be using 76 ). But, ex hypothesi, their practices do not presuppose anything of the sort, and so there is no room for us to accuse those practices of error or mistake. In short, our denial that their judgments ever amount to knowledge depends on our ability to ascribe systematic error to their practices, and our ability to do this depends on our being able to (non-vacuously) explain what that error 75 Denying the possibility of a theory of error, does not mean that one group cannot criticize another. Far from it. But what it does mean is that such criticisms cannot go as far as to accuse them of having made a mistake or were missing something in using the concepts they were using, or that their social world was a product of some misunderstanding. There are complications here, of course, in relation to how ethical views and views of the natural world might influence one another, such as the view that Africans were an inferior race, or less evolved form of humanity, which was (at least part of) the justificatory base for slavery. The error here, however, stems from false views about Africans as human beings, and is thus an error attributable to primitive scientific conception of human beings as a species. 76 Resting on false empirical presuppositions would be another reason for discounting their judgments as knowledge. But this sort of scenario, I take it, is not the one of central interest here. 211 is. 77 In lacking any explanation of what that error is, we lack the grounds for denying that their thick judgments can indeed amount to knowledge (“when they deploy their concepts carefully, use the appropriate criteria, and so on” 78 ). And so, if we cannot provide a solid footing for an objectivist interpretation of some group’s thick conceptual practices, a nonobjectivist interpretation of them is all that we have left. Section 4.3: Objections to Williams Nonobjectivism Given this overview of Williams’ nonobjectivism, we might consider some objections that have been raised against it. The first, offered by Alan Thomas, is that Williams’ view of objectivity involves a suspect assumption. 79 The assumption is that, in order for an ethical outlook (or a thick conceptual practice) to be considered objectivist 80 at all, thick judgments will have to be seen as being bound up with these general reflective implications, about their concepts being the right or best ones to be using. Thomas writes: Is this not, in the present case, too controversial an assumption? On a more realistic picture of moral cognitivism, people take themselves to be finding out the truth about ethical questions from within established traditions of enquiry. As in other forms of enquiry, propositions that come into question are not doubted until some concrete concern is raised that problematizes them against a set of relevant alternatives. (Thomas 2007, p. 66) 77 In characteristic fashion, Williams notes that the “insistence that a given person is wrong, disconnected from any possible understanding of how it comes about that he is wrong, tends to leave the commentator entirely outside that person, preaching at him.” (Williams 1985, p. 241n16) 78 (Williams 1985, p. 148) 79 (Thomas 2007, pp. 66-67). See also (Thomas 2006), Ch. 6. 80 This is the model of ethical thought which Thomas seeks to defend, though he prefers to call it ‘moral cognitivism’. 212 There are several things to say about Thomas’ objection. My first three comments will be intended defuse some of the tension that Thomas takes there to be between his ‘moral cognitivism’ and Williams’ nonobjectivism. Much of Thomas’ ‘realistic picture’ of moral cognitivism is something that Williams already accepts. First, as I noted above, and as is acknowledged by Thomas himself, Williams does think that judgments involving thick concepts can be both true and known to be true. Second, consider that, for Williams, the only plausible picture of ethical thought is one where, as he notes, “a concretely experienced form of life can be extended,” 81 where a community can adapt and change over time by appealing at different points to its own values and other ethical considerations from within its perspective. Perhaps Thomas wishes to invest something more than this into his notion of a ‘tradition of enquiry’, maybe along the lines suggested by Alasdair MacIntyre, 82 but if this is so, then he is conspicuously silent about the details. These two points lead to a third. The view that Thomas is espousing is one where ‘established traditions of enquiry’ are concerned about their own ethical problems, or worried about ethical questions that are internal to that tradition. But, because of this, they are engaged (perhaps only implicitly) in a debate that need not be seen as involving concerns about what other traditions of enquiry are doing. If this is right, then it is no longer clear what difference there is at all between Thomas’ moral cognitivism and Williams’ nonobjectivism, for the pursuit that such a tradition would be engaged in would be one that attempted to ‘extend a concrete form of life’ as opposed to discovering the objective truth about values. 81 (Williams 1985, p. 104) 82 See, for instance, (MacIntyre 1984) and (MacIntyre 1988). 213 My fourth point is a query as to what Thomas actually takes objectivity to amount to, given these remarks. Certainly the notion of what counts as an objective subject matter is a complicated one—as Williams attempt to explicate this notion in terms of a pursuit for an absolute conception makes apparent. And perhaps it is too complicated to give any satisfying definition here that would also be both definitive and substantial. But we do know this. It is not a question of truth or knowledge; this is something we have already established, as both Thomas and Williams agree that thick judgments can be both true and known. Nor is it a question about metaphysics, about their being a world of values, independent of an ethical outlook, which it is attempting to mirror accurately (whatever that may mean), as both Thomas and Williams agree that ethical enquiry is something that must be prosecuted from within ‘traditions of enquiry’ or a ‘concrete form of life’. What is suspect about Thomas’ objection, then, is that we would seem to lack a clear sense of what it is for a subject matter to be objective beyond Williams’ nonobjectivism. Williams, however, does have an answer to this (or at least a suggestion). And it is something that, from what I can tell, Thomas does not seem to fully appreciate. For Williams, to say that a conceptual practice is engaged in accounting for some objective subject matter, this has implications (namely negative ones) for those who think differently (namely that they are mistaken in doing so). And, perhaps most importantly, these implications are ones that demand substantiation in a non-vacuous, non-question begging fashion. One would not, for instance, support their claim to objectively know something if it was based on premises which other reasonably informed people would not 214 (or worse could not) assent to. In short, a claim that one has knowledge of an objective subject matter cannot be parsed away from the implication that someone who thinks differently on the matter might be just plain wrong. But such a general implication is precisely the one which Thomas wishes to deny in claiming that ethical thought is objective. If I am right, Thomas’ objection is vacuous because everything that he clearly wants to establish for his moral cognitivism is already granted by Williams’ nonobjectivism. The second objection I want to consider is related to Thomas’, but is concerned more about substantiating the possibility of their being some form of objective vindication for a group’s thick ethical concepts. The objection begins with the claim that Williams foists an unreasonable requirement on our being able to see some thick concepts as capable of objective vindication: Williams unreasonably assumes that objective vindication must occur at the level of an entire social world, as opposed to more piecemeal vindications that would comparatively look at the individual thick concepts within different social worlds. John Tasioulas voices just such an objection when he writes: …it is questionable whether successful explanations of convergence commit us to addressing such mind-bogglingly general questions as that of identifying the best kind of social world….There is no obvious justification for Williams’ unargued supposition that ethical thought must be conceived in terms of the social worlds (or traditions) in which it is rooted and that these have to be assessed and ranked as monolithic entities. Instead, inter-traditional critique is best conceived as piecemeal in character, focused on comparing the merits and demerits of rival social worlds in nuanced and concrete terms rather than on the ambition of constructing a ‘grand narrative’ that issues global vindications or denunciations of entire world views. (Tasioulas 1998, p. 394) 215 The heart of this objection I take to be the following: Williams assumes that, at the appropriate level of reflection, the questions we will be concerned with are those relating to entire social worlds, as opposed to particular concepts that occur in and between social worlds. Indeed, as I have laid out the problem above, I have moved without qualification from talk of sets of thick concepts to social worlds and back again. Making this assumption, however, leads rather obviously to reflective questions that, in ethics, seem so general and abstract that we lack a clear sense as to how we could even seriously formulate them, let alone answer them: What would the best social world be like? Are we even sure what this question is asking? This ‘monolithic’ assumption, however, can be rejected. In doing so, we can make way for a sense of reflective vindication geared towards the comparison and evaluation of particular thick concepts. And so, we can see that particular thick concepts might be objectively vindicated (as better than others), even if there is no hope of this when it comes to entire social worlds. This objection, however, severely misunderstands the force of Williams’ position, and as a result fails to see that Williams would agree that such ‘inter-traditional’ critiques are possible (cf. my above discussion of Thomas’ objection). What Williams would seek to limit (and I think he is right in doing so) is the sense that we are employing ethical resources in these critiques that somehow place us in an objectively better position than those that deploy different resources. Certainly it is open to us to provide explanations, make various arguments, and otherwise make use of any ethical resources from within our own outlook that we thought would help sustain that outlook, and in doing this show how certain thick concepts are ones that we should be using (for a narrower, contrastive 216 ‘we’): because they help to make sense of our interests or our social situation or the sorts of lives that we think are worth living. We might also be able to look out to alternative outlooks that are roughly contemporaneous with our own, find within them the very same ethical resources and be able to justify the belief that our concepts (or some of them) are superior to theirs (and, indeed, also have the ability to justify this to them, from the very same considerations). There is nothing in Williams’ view that would deny these possibilities. Yet, such ‘inter-traditional critique’ would not entitle us to think that the vindication or vitiation of a particular concept was an objective matter: such a critique would be, in a certain sense, preaching to the (partially) converted. To accomplish something stronger, something that had the hope of yielding genuinely objective results, we would need to embark on a program to justify the resources upon which such a critique would depend, and be able to justify them in the face of ethical schemes that use very different considerations. Such a program would have to explain what it is that they are getting wrong, and do so in a non-question begging fashion. When put in these terms, we quickly come back to the question of whether we are entitled (in some objective sense) to deploy the ethical resources from our tradition (or social world) in mounting the vindication or vitiation of particular ethical concepts. 83 As with Thomas’ objection, it is not clear that what Tasioulas seeks to preserve is not something that is already allowed for within the ambit of Williams’ nonobjectivism. If there is something more, if the question of whether we are using the right thick concepts is intended as something stronger than ‘right for us’ (using a contrastive rather 83 If this is not what we are led to, and Tasioulas seems to suggest this at one point, see (Tasioulas 1998, pp. 394-95), there is a question of what work a claim to objectivity is doing here. 217 than inclusive sense of ‘we’ and ‘us’), then there is the inescapable question of whether the social world that we occupy is something objectively better in comparison to some other, distinctive, us (and here it might be a helpful reminder to note that this us might not even be human). Section 4.4: The Loss (or Destruction) of Ethical Knowledge Williams’ skeptical remarks about the possibility of thick ethical knowledge started by considering whether a group’s conceptual practices were best understood along the lines of the objectivist model. If we find this interpretive model attractive, it would be because we see their practices as presupposing answers to general reflective questions that they are incapable of answering, even if those questions have never been explicitly raised by that group. But, because there is no satisfying answer to those questions (i.e., there is nothing for their conceptual practices to be true of; or, there is nothing outside those conceptual practices which they could be seen as genuinely approximating) we will be forced to see their use of those concepts as suffering from some inherent defect. As a result, we will have to conclude that their concepts do not in fact provide them with a form of ethical knowledge. Such a position is similar to the sort of anti-realism advocated by J.L. Mackie, 84 which assumes that all moral thought includes certain objective pretensions. However, Williams is clear that the objectivist model is not something that is forced on us. Absent such general implications, their practices are represented more accurately by the non-objectivist model. As such, their use of these concepts will not be seen (even implicitly) as being in competition with other ethical 84 (Mackie 1977). 218 schemes. Rather, these concepts will be seen merely as devices which provide a certain form of practical guidance to those who possess them, where this practical guidance helps them to regulate their interactions with one another in the social world that they (and these concepts) are a part of. Here we have no reason to deny that such claims amount to knowledge. This ethical knowledge, however, is still fragile to the effects of reflection, and in this sense is epistemically fragile in a way that scientific knowledge is not. Since it is not knowledge that is grounded in anything beyond the social world that sustains those concepts necessary, coming to reflect on their adequacy can have a destabilizing effect. Williams writes: If we accept that there can be knowledge at the hypertraditional or unreflective level; if we accept the obvious truth that reflection characteristically disturbs, unseats, or replaces those traditional concepts; and if we agree that, at least as things are, the reflective level is not in a position to give us knowledge that we did not have before—then we reach the notably un-Socratic conclusion that, in ethics, reflection can destroy knowledge. (Williams 1985, p. 148) The tension that Williams sees, in ascribing ethical knowledge on the nonobjectivist model, is a tension between the practice of using thick concepts and what can be said about that practice at a general reflective level. When a group comes to reflect in this way about the adequacy of how they are conceptualizing their social world, such a general question is not whether they have reasons, from within that world, for doing so. Rather, such questioning is taken to have wider or more global import. It is the product of a general reflective stance that is attempting to reconcile its use of certain ethical concepts with two important considerations. 219 The first is the fact that their social world is not the only social world there is. There are, or at least could be, other ethical outlooks, deploying different schemes of thick ethical concepts. Such a thought might never occur to a group until it has come in contact with an ethical form of life that is sufficiently distinct from its own. Or such a considerations might occur to them, with the requisite amount of force to present itself as something requiring explanation, merely through their considering potential alternatives merely on the basis of applying their own imaginative capacities. Regardless, it is a fact that we (in the modern world) must contend with, and it is a fact that others have had to contend with in the past. The second consideration concerns an understanding of that outlook’s own historical development, and how that history is a radically contingent one. Such a history (if it is anywhere near correct) will likely reveal how their outlook could have been different, by revealing how the structure of that outlook was shaped by social forces less directly relatable to notions like conscious choice or deliberation. But it will also reveal how many of the historical changes were brought on by factors which were completely beyond social control (and hence incapable of being seen as the expression of internal pressures to choose one set of concepts over another)—wars, disease, and natural disasters can radically alter the shape of a community, but they do so in a way that deepens the sense of contingency attached to its current state. In both cases, reflection would reveal how there are epistemically unfavorable considerations that have to be accounted for in understanding how the group came to possess the thick concepts it did: 220 considerations which play a role in determining the identity of certain conceptual practices, but which could not be mustered as grounds for them. Given this, we can give the following rough sketch of the sort of phenomena that Williams is trying to bring to our attention. It will be a rough sketch, because some of the details in it will require additional discussion. This task will be taken up in the next chapter. When a group comes to ask reflective questions about the adequacy of their current thick concepts, and in doing so realizes that whatever answer is given will have to be of sufficient generality to make sense of the sort of ethical diversity that is no doubt possible, as well as the contingency of its own development, these reflective questions can be seen as requiring answers for which none can be given. To provide the appropriate sort of response to these general reflective questions, the answers would have to enable two things: (1) an explanation of how they came to use a particular set of thick concepts which justified their use (cf. the case of perspectival concepts related to our sensory modalities), and (2) a (non-vacuous, non-question-begging) theory of error which would explain why, even though there are a diverse number of ethical outlooks, those outlooks were in error. But this is something that they will be unable to do, for the very possibility of such an enterprise is at odds with the facts that they are now aware of, namely that there are other outlooks and that their own history is a radically contingent one, and as noted above, how much of that history involves elements that are epistemically unfavorable. This realization that no adequate answers to these reflective questions will be forthcoming, this can have a destabilizing effect on their use of the thick 221 concepts to which those questions were directed. They might, as a result, find that they can no longer use those concepts anymore, because where they once seemed to make important distinctions, now they and the distinctions made by them seem arbitrary. But, in finding themselves barred from the use of these concepts, they are also barred from the knowledge that they provide, given the uncontroversial assumptions that one cannot know that P unless one can believe that P, and one cannot believe that P unless one can seriously judge that P is true. Hence, this sort of general reflection can cause a loss of ethical knowledge by precipitating a process that leads to the abandonment of certain ethical concepts and provides nothing to replace them with. While certainly disquieting, Williams’ position on the relationship between this sort of general reflection and the knowledge provided by thick concepts can be overstated. First, Williams notes that reflection can destroy ethical knowledge, not that it must or even will. 85 And so, it is possible that some thick concepts might indeed survive such reflection. Surviving reflection, however, does not mean that they are thereby objectively vindicated: ‘survival’ simply implies that an awareness of the concepts’ contingent historical development and the diversity of alternatives to it are not things that the community takes to be a (sufficiently strong) reason for abandoning the concept. 86 Lacking an objective grounding does not entail that there are no considerations which 85 Note Williams’ emphasis on this in (Williams 1995c, pp. 207-08) and (Williams 1996b, pp. 29-30). 86 Williams writes: I did not mean [in Ethics and the Limits of Philosophy] that the knowledge which is available under such a [thick] concept when the concept is properly deployed is turned by reflection into reflective knowledge: it will not have been validated, as under the Aristotelian aspiration our ethical concepts would be integrated into a dense and comprehensive understanding of nature, or at least of human nature. It survives reflection just in the sense that we would not have encountered any considerations that led us to give it up, lose hold on it, or simply drift away from it. (Williams 1995c, p. 207) 222 they might muster as reasons for continuing to use those concepts: at best it means that they lack reasons which will be universally recognized as such. 87 Secondly, Williams is, at least as I read him, attempting to leave room for a model of social and moral change that is (at least internally) progressive in character. The idea that reflection can destroy ethical knowledge allows us to see how moral change can reflect something similar to a paradigm shift in science, 88 as it would be a type of moral change that could not be adequately described simply as a matter of changing our particular opinions or feelings towards some neutrally understood social environment. Such moral change would have to be seen as a type of conceptual change. A social world can change radically by the fact that certain judgments cease to be available to the community, because the concepts that would make those judgments possible can no longer be used. The notions of ‘availability’ and ‘use’ here have to will be explained, as well as a related notion of incapacity. These topics will be discussed at length in the next chapter, but for now we can give them a brief gloss by saying that their use of those thick concepts, which became destabilized by reflection, would (at best) be something that was ironic or inauthentic. 89/90 The concept, as a member of the group would now see 87 In this I am in agreement with Thomas Scanlon, in (Scanlon 1995, p. 353), when he notes that the fact that one does not have reasons that would convince anyone does not mean that one does not have “good and sufficient reasons” for one’s position. 88 See (Kuhn 1996 [1962]). For an excellent discussion of the various meanings of ‘paradigm’ for Kuhn, see (Godfrey-Smith 2003), Ch. 5. 89 (Williams 1996a, p. 212). Cf. (Nagel 1971). In a line of thought that parallels Williams quite nicely, Nagel writes the following on pg. 724 of that article: Philosophical skepticism does not cause us to abandon our ordinary beliefs, but it lends them a peculiar flavor. After acknowledging that their truth is incompatible with possibilities that we have no grounds for believing do not obtain—apart from grounds in those very beliefs which we have called into question—we return to our familiar convictions with a certain irony and resignation. 223 understand it, would no longer enable judgments that would give voice to what that person cared about or thought important. This fact, however, is compatible with their still retaining an understanding of the concept, and an ability to make judgments using it. Such judgments though would not be expressive of their interests and concerns, or worse would be expressive of interests and concerns that they are now opposed to, as things now stand (in their post-reflective state). Consider this in comparison to the thick concept I gave, at the beginning of this chapter, as an example of one we no longer use. Certainly someone who has made a judgment like ‘John is a cad’, in the post-reflective state where the person making this judgment no longer has the sort of specific interests that underlie the concept CAD, has not thereby formed something like an incomplete thought. Nor is it to say that such a judgment could not very well be true, for ‘John is a cad’ just in case John is a cad— which is to say—just in case how John is behaving towards women is something that And then following this: What sustains us, in belief as in action, is not reason or justification but something more basic than these—for we go on in the same way even after we are convinced that the reasons have given out. If we tried to rely entirely on reason, and pressed it hard, our lives and beliefs would collapse—a form of madness that may actually occur if the inertial force of taking the world and life for granted is somehow lost. If we lose our grip on that, reason will not give it back to us. 90 Stephen Finlay has pressed me as to why we should not see this as lapsing into a form of error theory, like (Mackie 1977). The question is: Where is there error? Part of what I take to be an upshot to Williams discussion is that, in coming to find that they now can no longer use certain of their thick concepts, this does not force upon them to deny that they had knowledge under those concepts prior to the lapse. All that is lost, in the case that Williams is envisaging, is a loss of a sense of importance in those concepts. If we take seriously the thought that this loss of importance in using such a concept amounts to an error, then it will have to be explained what that error is. And I am not clear how this can be done in a non-vacuous, non-question begging way. Of course, their reflection might lead them to see that their use of these concepts is predicated on false factual assumptions, but then those assumptions would have been there before they recognized them as such, and there would be no room for talking that they knew something in using their thick concepts at all. But here there would be an explanation of what that error was, and not a blank pronouncement that they had previously gotten something wrong in attaching so much importance to the use of that concept (or concepts). 224 falls under the criteria for considering a man a cad. Rather, the sense in which such a judgment would be ironic or inauthentic is more along the lines that that thought is not one that the (post-reflective) person genuinely believes. While beliefs are credited as cognitive states, representing the world in a certain way, believing is a matter (to some degree or other) of conviction, and it is having this sort of attitude towards the judgment ‘John is a cad’ that is undercut by reflection. One lacks a requisite sense of (or appropriate form of) conviction in the standards for applying the concept CAD, and because of this any attempt to make a judgment along those lines will fail to count as genuinely believing that judgment. 91 Of course, in some sense, such a person may be able to think ‘John is a cad’, but we would not be able to describe this thought as amounting to her actually believing it, for implicit in that judgment would be the inability to passively accept a concept which reflection has now called into question. Her stance towards the judgment would be similar to that of someone outside the community who was attempting to understand the community’s use of this term: 92 there would be a genuine sense in which such a judgment did not (and could not) count as an avowal of her own views. Given this, we might say that what is lost, in losing a thick concept to reflection, is (at least initially) the ability to form a genuine moral judgment with that concept. 91 This is what I think Williams means when he says that people in this post-reflective state “can no longer form beliefs of that kind.” (Williams 1985, p. 167) Of course, the point I have raised here abstracts away from concerns relating to an understanding of conviction in terms of degrees of belief. I take it that not all levels of epistemic conviction towards some proposition counts as believing that proposition. 92 Williams uses this analogy, see (Williams 1985, p. 167). 225 What grounds are there for saying that such knowledge is lost? There is a temptation here to think that this loss has something to do with the epistemic status of that knowledge, but this is not what Williams had in mind. First, it is not lost because some set of claims which were actually true in the past are now no longer true. This would be to convict the issue here with a very strong form of incoherence. Secondly, it is not because the body of claims which was once believed to be true is now known to be (systematically) false. This would amount to a denial that there was ever knowledge to begin with, and would require the very same appeal to a theory of error that we are (in this case) assuming is unavailable. The key to seeing such knowledge as lost or destroyed is in the realization that there is nothing in the above description that prevents our post-reflective thinker from seeing their past judgments as amounting to knowledge. In this way, reflection does not alter the status of certain claims to be true or knowable, rather it alters a subject’s ability to know them. Prior to the onset of reflection such a person could agree that they were disposed to genuinely believe things like ‘John is a cad’, that they offered such judgments as expressive of their concerns and interests, and treated such judgments as reasons for (among other things) censuring John. What makes this knowledge lost for this person is that she does not have the ability to do these things now, because she lacks the requisite conviction in this sort of judgment. In this sense, there indeed is a type of ethical knowledge which is knowledge such a person cannot now possess. Such a position might be quite disturbing for the person who comes to this position. They might feel as though reflection has robbed them of something, and that 226 they would have been better off had they never raised those concerns at all (though I doubt that their doing so would ever be a purely voluntary issue). But, as Williams rightfully notes, “there is no route back from reflectiveness.” 93 Once these reflective concerns are seriously raised, they are likely to remain a hindrance to using the concept, until some explanation is provided that silences them. But, equally, such reflection might be taken as liberating (this is a point that we will return to towards the end of the next chapter), and the loss of ethical knowledge might be offset by gains in having a better understanding of ethical diversity and the development of their own ethical outlook. 94 I think that the sort of phenomena Williams is highlighting can also be brought out by considering another example, one drawn from Mark Twain’s Huckleberry Finn. 95 Admittedly, the example does not conform to all the details of Williams’ discussion, as the example itself has little to do with the sort of reflective concerns that Williams thinks are endemic to modern life (concerns relating to ethical diversity and an outlook’s radically contingent history). That said, I think that there are enough parallels in the example to help make sense of the idea that reflection can cause a loss of ethical knowledge. There are two points that I want to draw out in the comparison. The first relates the idea that ethical knowledge can be lost through reflection at a more personal level, what we might call losing one’s faith in an ethical outlook. 96 The second relates 93 (Williams 1985, p. 163) 94 Williams raises this point, see (Williams 1985, p. 168). 95 (Twain 1994, pp. 291-93) . Another example might be drawn from William James’ The Ambassadors, and how, in traveling from one culture to another, the ethical scheme that one had grown up with can come to be seen as unenlightened and petty. For an excellent discussion of this story, see (Kekes 1984). 96 My use of ‘faith’ here is borrowed from A.W. Moore. He writes: 227 this idea to how we (as individuals looking back on the social world of the antebellum South) are able to relate to Huck’s situation. Let me start, by laying out the particulars of the example. Towards the end of the novel, Huck returns to the raft that he and Jim (a slave) had shared, and comes to find that one of the two con-men that he and Jim had been traveling with had in fact turn Jim in for a reward. Huck, worried about it being discovered that he had helped Jim escape, considers writing Miss Watson (Jim’s owner) to inform her of where Jim is. But then he decides against this, for even if Jim was returned to Miss Watson there was no way that this would make Jim’s life any better. From this, however, Huck’s thoughts turn to his role in Jim’s escape, and how he had done something disgraceful: he had stolen something from Miss Watson, a woman who had never done him any harm (and who was the sister of the Widow Douglas, a woman who had taken Huck in and cared for him). There was no way of getting around the fact that Jim was Miss Watson’s property, and that his hand in helping Jim escape amounted to theft. Huck tries to pray, as a means of setting himself straight and atoning for these ‘wrongs’, but finds that he cannot because this was not something he really believed: “You can’t pray a lie,” Huck thinks. So, instead, he writes the letter to Miss Watson, to see if the very writing of the letter would ease his conscience. After writing the letter, however, Huck’s thoughts turn to Jim, and the time they had spent together on the raft, and to how gentle and kind Jim had been towards him…to the fact that they were friends. Practical reasoning…includes a pure element: keeping faith with concepts. Theoretical reasoning also includes keeping faith with concepts. What makes it possible for keeping faith with concepts to have a practical dimension as well as its more theoretical dimension is, ultimately, the fact that some concepts—thick ethical concepts—equip those who embrace them with certain reasons for doing things. (Moore 2006, p. 143) 228 From these thoughts, Huck comes to see that he cannot carry through with sending the letter, so he tears it up. What makes Huck’s crisis of conscience compelling is that he cannot bring himself to act for an ethical system that he does not genuinely believe it. To be sure, Huck never reaches the point of actually criticizing that system on ethical grounds. But, the lie that Huck could not pray was that he really believed that all things considered he should turn Jim in. And this conclusion, we might think, is based on the fact that, in reflecting on his and Jim’s time together, Huck had come to take a very different view of Jim. It is here, perhaps for the first time, that Huck sees Jim as a friend, or perhaps in more basic terms, as a person, and that this view was fundamentally incompatible with the thought that Jim was someone’s property: a thing which could just as easily be bought or sold as a chair or horse. That thought (that Jim is, or could be, someone’s property) is something that no longer holds sway with him: though he might certainly realize that others believe this. Huck’s very conception of property has changed, and this change means that there are certain judgments that, although he might have made in the past and in doing so knew them to be true, they are judgments that he is now simply incapable of making. The idea that Jim is someone’s property is now something Huck cannot believe. But, at the time in which Huck is deliberating about this, Jim indeed was Miss Watson’s property, and what he had done indeed amounted to a form of theft (just as it would be if he had stolen her chair or horse). These facts follow from the social institutions that helped constitute slavery in the antebellum South. So it would not be that 229 Huck thought it was false that Jim was Miss Watson’s property, it was rather that he could not think of Jim in these terms any more. 97 We might say, then, that what is lost for Huck, by way of his reflections on what to do, are certain conceptions of theft and property. To be sure, Huck does not lose these concepts per se. It is not that he ceases to believe that someone can ever own something and have that thing stolen from them (this is one of the ways in which the example might not fit precisely with the sort of case Williams is envisaging). But Huck does lose something, and we might very well think of this as his losing the capacity to deploy in judgment the antebellum South’s conceptions of property and theft. And so, we might say that, if he had ever had it to begin with, Huck had lost faith in the ethical system that supported those conceptions, and because of this he is now no longer able to believe certain judgments, judgments that, prior to his crisis of conscience, he very made and (when the circumstances were right) very well knew. In this way, there is a certain form of ethical knowledge that is lost for Huck. At a less personal level, we can reflect on Huck’s situation and appreciate how very different things are for us. Focusing, for instance, on Huck’s consternation in trying to decide what he should do, and not the resolution which resulted (as I am suggesting) in his coming to view his situation fundamentally differently than he had before, we will fail to grasp the import of Huck’s deliberations if we do not appreciate the fact that judgments like ‘Jim is Miss Watson’s property’ were both true and known to be true by Huck, and that they provided him with a very strong reason to turn Jim in. Knowledge of 97 For a view sympathetic to this reading of Huck’s situation see (Freedman 1997). For a contrasting view on the moral psychology of Huck Finn, see (Bennett 1974). 230 this was what drove Huck’s moral dilemma to begin with: one which pitted his feelings for Jim against the thought that Jim was someone else’s property. And yet, our describing his situation in this way also constrains how we must explain our own relation to Huck’s plight, largely because we are now incapable of treating this situation as a dilemma at all. What is of interest here is that Huck’s understanding of his own situation, prior to his resolution to break from that outlook, was bound up with certain ethical conceptions of THEFT and PROPERTY. Such conceptions were a part of the social background of the antebellum South. 98 These were concepts that structured much of Southern life prior to the Civil War, and were ones that almost everyone came to acquire in virtue of being raised up in that world. Even though they are (in a certain sense) an artifact of this social world, they nonetheless underwrote pieces of knowledge that provided people at that time with a type of practical guidance. However, while being central to the antebellum South, they are now ones that are not simply alien, but abhorrent, to us. But, part of what also separates us from that social world is the fact that our understanding of these concepts differ significantly from what was prevalent to that world: there were things that members of the South could know, that we now cannot (e.g., that Jim was Miss Watson’s property). Such knowledge provided them with ethical reasons that we cannot share. In 98 A compelling picture of this conception of property is provided by Frederick Douglas, in his description of the antebellum South’s institution of slavery: I will state, as well as I can, the legal and social relation of master and slave. A master is one—to speak in the vocabulary of the southern states—who claims and exercises a right of property in the person of a fellow-man. This he does with the force of the law and the sanction of southern religion…The slave is a human being, divested of all rights—reduced to the level of a brute—a mere “chattel” in the eye of the law—placed beyond the circle of human brotherhood….From this monstrous relation there springs an unceasing stream of most revolting cruelties. (Douglas 1855, pp. 429-30) 231 this way, the knowledge had by members of the antebellum South—namely that being someone’s slave provides one with certain reasons for acting (e.g., a reason not to help them escape)—was a piece of ethical knowledge. What is destroyed, then, when ethical knowledge is destroyed, in this larger, social sense, is the possibility of counting the fact that someone is another’s property as a reason to do certain things, because the very possibility of seriously entertaining the concepts needed for doing so is undermined. Given this brief discussion, we can confront a central objection to Williams’ view that reflection can destroy ethical knowledge. The objection (shared by McDowell, Blackburn, Putnam, and others) amounts to the charge that Williams’ position here is seriously confused, or worse, incoherent. Reflection cannot destroy knowledge, McDowell objects; rather “when reflection dislodges one from a practice, it does so by persuading one that judgments which seemed to embody knowledge did not embody knowledge after all, since they were not true.” 99 Reflection might be capable of revealing to a group like the hypertraditionalists that their thick judgments were not really true, but this would entail that they were mistaken in thinking they knew anything to begin with by them. What reflection cannot do is both show how there is a genuine body of ethical knowledge, but then undermine one’s ability to share in that knowledge. Reflection, so the objection goes, could not take from some group the ability to know certain things, if it is granted that they knew those things in the past. The substance to this objection centers on what it is that reflection undermines. The hypertraditionalists do not lose their truths—as I noted above, it is not that certain 99 (McDowell 1986, p. 383). Similar objections are raised, in one form or another in (Blackburn 1986, p. 200); (Putnam 1992, pp. 104-06) and (Putnam 1990, p. 169); (Quinn 1987) and (Moore 1991). 232 true claims are now not true. What they lose are their concepts (or, in Huck’s case, a certain conception). But, even this, has to be qualified: what they lose is a certain conviction in the adequacy of those concepts, and this loss of conviction undermines their ability to be genuinely convinced by the truth of the judgments they would make using those concepts. 100 The worry at the heart of this objection comes from the assumption that, if we are to claim that a certain group has a piece of ethical knowledge, this entails that we must endorse or affirm that knowledge. This assumption, I think, is tied in different ways to (i) the thought that truth is always objective truth, 101 or (ii) certain performative aspects of calling a claim true: aspects which are highlighted by the fact that ‘is true’ is often used to express an endorsement of what is said or judged. 102 But we need not accept either of these points. To say that a certain group has ethical knowledge, in using their thick concepts, this can be understood in descriptive or explanatory terms which does not force on us any direct claim to endorse (or agree with) their judgments. As such, we can simply ask: (1) Are they convinced that certain thick ethical judgments are true? (2) Are those judgments, relative to the criteria implicit in the concepts used to form them, true? 100 Williams writes: What I had in mind [in Ethics and the Limits of Philosophy] was the situation in which they longer have the concept with which they used to express a certain class of beliefs. They lose a concept, and so cease to have a disposition that expresses itself in categorizing the world in those terms. (Williams 1996b, p. 30) 101 Something like this is implicit in the writings of Donald Davidson, though I think that Davidson’s view of truth are more complex than many are apt to treat them. See (Davidson 1984), Chs.2, 9, 12, and 13 and (Davidson 2005), Chs. 1 and 2. McDowell certainly seems to favor this reading. See (McDowell 1996). 102 This is implicit in Blackburn’s objection. See (Blackburn 1986, p. 200). 233 (3) Is (1) non-accidentally connected with (2), or conversely, is the fact that they are convinced that some thick ethical judgment is true simply an accident and in no way connected with the truth of that judgment? and (4) Are their conceptual practices free from implications which would entitle us to think that they suffer from some inherent defect? I take it that, given Williams discussion, if we can answer affirmatively to (1 – 4), this would entitle us to say that a group had knowledge of those judgments. The condition that will be seen as suspect will be (2), for the reasons already mentioned: but we need not accept the thought that an affirmative answer to (2) entails that we endorse the judgment in question. It can simply be understood as a descriptive question relating to whether the criteria of the thick concept involved has indeed been met by the situation the judgment employing it is directed towards. But, importantly, reflection undermines (1), insofar as it is no longer true that they are convinced that certain thick ethical judgments are true, even when they are, and when the truth of those judgments is non-accidentally related to the circumstances towards which they are applied. What reflection does is to alienate some group from the thick ethical concepts which are necessary for expressing certain knowledge claims. If they could only bring themselves back to the point of serious engagement with the concept (and if they could only forget why they lost that sense of seriousness to begin with), then they would be able to continue on using them as they had before. But as long as the reflection persists serious, non-alienated engagement is impossible: they will be unable to passively accept the thick concept upon which those knowledge claims depend. Given this, however, we might agree that talk of knowledge being ‘destroyed’ is, in one sense, unnecessarily provocative. It brings to mind the seemingly counterintuitive thought that reflection is in 234 the service of some rational reversal: that the hypertraditionalists are (say) in a worse spot for having less ethical knowledge. Yet, there is no reason to assume this (as Huck’s case reveals). There is nothing in Williams’ account which prevents the reflective hypertraditionalists from viewing their pre-lapsarian state as one which reflected a deep prejudice or which was the product of an unhealthy ideology. Reflection might uncover that their use and acceptance of such concepts was largely the product of misunderstanding and manipulation. Such points might bolster the thought that there was something that they should have been attuned to earlier in their practices, but failed to appreciate. In this sense, they might have a genuine reason for seeing themselves as better off with less ethical knowledge. But the important point is that they need not: indeed no fixed response is dictated by their coming to this more reflective state (and this is why the loss is just that, a retrograde movement in the amount of ethical knowledge they possess). If we return then to the plight of the hypertraditionalists, we can now see that, although, their loss consists in their (post-reflection) now being unable to use some thick concept, it still remains open for them to consider (at least some) of their previous judgments to be true and to have expressed knowledge. Consider, again, our reaction to Huck’s plight, and how describing it as a moral dilemma demands that we see Huck as having certain pieces of knowledge that we do not share. There is nothing in the hypertraditionalists reflective predicament that forces upon them the thought that their previous practices were necessarily defective. The lack of conviction in a thick judgment is tied here to an inability to treat the terms in which that judgment is couched as 235 authoritative, but this is completely consistent with the recognition that absent this they would have been able to know certain things through that judgment. Certainly they might come to think their previous practices were defective in some way, the point here, however, is that there is nothing in the very description of knowledge being lost or destroyed which forces this on them. Section 4.5: Concluding Remarks It is important not to put too much stock in Williams’ example of the hypertraditionalist society. Nowhere does Williams pretend to think that anything like the hypertraditional society has ever existed, 103 or that we (as existing in a society that is not hypertraditional) are able to see ourselves in precisely these terms. 104 It is a hypothetical case that is intended to reveal something about the nature of traditional ethical thought, and how the practices which enable that thought to amount to a form of knowledge are themselves sensitive to reflection. But, if Williams’ view is to say anything at all relevant to our modern predicament, then we must be prepared to see hints of this model in real world examples that are only an imperfect fit. As I noted above, the example of Huck’s plight is not a perfect fit for the sort of picture presented by the hypertraditional society. This, however, does not mean that we should not look to Williams’ model as framing the right sorts of elements to help us understand it, and, in doing so, our relationship to part the antebellum South. 103 Cf. (Williams 1985, pp. 164-65). 104 For a view that takes Williams’ hypertraditional society to speak more directly to our own plight as to whether we have knowledge under certain thick concepts, see (Thomas 2006) and (Thomas 2007). 236 Certainly there are important differences between us and the hypertraditionalists. While the hypertraditionalists could be seen as subscribing to a single conception the good life, modern societies are typified by a plurality of such conceptions. We are not as pre-reflectively innocent (if at all) as the hypertraditionalists are. Because of this, there is good reason to think that these sorts of reflective questions might not be seen as serious challenges. The continued use of some thick concepts will likely be unaffected, then, largely because our unselfconscious use of these concepts has adapted itself to an openly pluralistic environment. For others, however, their use might as yet become impossible (such as the concept CAD). And depending on the degree to which such a concept is centrally important, there might be the real threat of a social chasm developing: two groups which had previously seen themselves as a part of the same social world might now come to see each other as engaged in two quite different ways of life. There are various ways in which such complications might arise, and there are various directions in which the stability of ethical thought can shade off into issues regarding the very idea and identity of a singular ethical community. In this sense, a realistic approach to the character of ethical thought in modern society must appreciate the tension that exists between attitudes towards pluralistic conceptions of the good life, and the debilitating effects that this can have on the importance we want to attach to our own conception. It goes without saying that function of this chapter has been largely exegetical. I have tried (at certain points) to clarify Williams’ position in ways that are (I hope) helpful, and to unpack what are certainly compressed arguments. Admittedly, I have 237 objected to nothing of substance in Williams’ picture. This is because I think much of his position, though perhaps unsettling, is correct and worth defending. We might in general say that, for Williams, the importance of truth in ethics comes not from simply stating the truths themselves, but from our dispositions to believe those truths. With that said, there are three issues that I take to be largely unresolved. We might see them as constituting an explanatory lacuna in Williams’ account. First, how is it that an individual or community can come to be unable to use such concepts, when they were certainly able to do so prior to reflection? The discussion thus far, and the examples given, I take it explain why it is plausible to think this is true, but it has not explained how this occurs. Such an explanation, I suspect, requires a deeper understanding of the special sense in which they are incapable of forming moral judgments using these concepts. Second, what is the relationship between the authoritative use of a thick concept and one’s ability to use it at all? Much of this discussion has traded on a distinction between one understanding of a thick concept (and hence, one’s having in some sense the ability to use it) and one’s ability to make ethical judgments using that concept (and hence, the possibility that one might be able to understand a thick concept and still be incapable of making such judgments). What we require is a proper sense of how a person can be unable to use such concepts. And finally, what is it about thick concepts in particular that makes the central elements in the first two questions possible? This latter question is asking for an explanation as to why reflection is able to destroy knowledge under thick concepts, but in the scientific case it would uncover, as McDowell puts it, “that judgments which seemed to embody 238 knowledge did not embody knowledge after all, because they were not true.” 105 These three questions will be a central part of the discussion of the next chapter, when we revisit how it is that reflection can destroy ethical knowledge in precisely the manner that Williams intends here. My aim will be to show how the elements we discussed in chapter three (namely the idea that thick concepts involve both ethical standards and rationales) provide an answer to these questions, and help us to understand what is distinctive about thick concepts. 105 (McDowell 1986, p. 383) 239 CHAPTER 5: Lost Ethical Knowledge and the Desire for Transparency Section 5.0: Introduction Central to Williams’ view that a certain sort of ethical knowledge is possible is the idea (or perhaps: the reminder) that not all knowledge is knowledge about a world that is independent of our making: We can know things whose content is perspectival: we can know that grass is green, for instance, though green, for certain, and probably grass are concepts that would not be available to every competent observer of the world and would not figure in the absolute conception. (Williams 1985, p. 139) Not all knowledge is knowledge pertaining to the world as it is anyway, independent of humanity. Some knowledge is knowledge pertaining to (what we might call) a human world—a world where there are certain object, experiences, and activities only because there are human beings. The knowledge that some things are tables, beds, and diet cokes, or that certain activities count as playing a game, holding an election, or even doing physics, are all pieces of knowledge pertaining to (or relating to) a world that exists only because human beings have certain needs, interests and concerns. And, to this extent, it is a world that is parochial in character: an extra-terrestrial who failed to find human beings and their various activities intelligible would fail to be in a position to know that something was a table or that a certain person was elected to office. They would fail to know these things because such objects and activities are intelligible only against a background constituted by a certain human form of life. In this case, there is a clear sense in which the knowledge possessed by various human beings could not be shared with such a being: there are things that these people would know (through the use of 240 certain concepts and modes of conceptualization) that this extra-terrestrial would be incapable of knowing. The idea that certain groups can have knowledge under their thick ethical concepts is simply a special case of this. The concepts in question, whose knowledge is at issue, are ones that are directed towards sustaining a certain form of social organization. Certain acts and character traits come to be classified together by some group so that they may be (ceteris paribus) encouraged or discouraged. And so, if we narrow our perspective further, from that of a general or abstractly considered human world, we might say that there is a type of knowledge that is related to living in (say) 19 th Century England (or a certain part of it)—what we might call knowledge of or pertaining to life in the Victorian (social) world. This is a body of knowledge that, among other things, involves the idea that some men behave as gentlemen, while others like cads and bounders, 1 and where some women are chaste, while others are trollops or strumpets. 2 These concepts came into being as a means of regulating the interactions between people in that social world: acts and character traits came to be classified in these and related ways because there were interests and concerns that members of this world shared in encouraging or discouraging various modes of social interaction. The manner in which these classifications developed, and the interests and concerns that underwrote their development, cannot be pulled apart: since the interests 1 The difference between a cad and a bounder was that the disvalue attached to the former was on account of how they treated women while the disvalue attached to the latter was on account of their social standing and class. Cads behaved dishonorably towards women; bounders were ill-bred and socially inferior by comparison. 2 While a strumpet was a woman who engaged in sexual intercourse for money (i.e., was a prostitute), a trollop was a woman who was sexually promiscuous, but not necessarily a strumpet. See, for instance, (Egan 1832). 241 involved in classifying acts and traits in these ways are first interests in guiding the group’s behavior and attitudes, such concepts were action-guiding from the start. And yet, as an appreciation of the history of any human culture makes apparent, the interests and concerns that a group has in regulating its behavior and attitudes is something that is both fluid and fragile: fluid, because these interests and concerns can, for a myriad of reasons, change over time, and fragile, because they can come to be seen by the very group which held them as simply arbitrary. It is the latter sort of possibility that interested Williams, and that, he thought, involved certain interesting philosophical implications. If the interests and concerns underlying a thick concept come to be seen as arbitrary, so too will the thick concept itself, and this will ultimately lead to an abandonment of that concept as a way of regulating the group’s behavior and attitudes. There are perhaps no clear historical examples of the sort of possibility that Williams is envisaging. This is largely because Williams was interested in relating a specific sort of concern, one unique to modernity, to the question of ethical knowledge under traditional, thick concepts. Modernity differs from previous periods in at least one respect: it is a period in which human beings and groups have never been more aware, not only of the existence of other groups, but also of their own development as a group. Both of these points underwrote a strong trend towards reflectiveness. Modern groups are more liable than previous groups to question the authority of traditional classifications, and this reflectiveness could, in some cases, make those very classifications impossible. And so, while Williams point might lack a specific historical 242 example which fits its details perfectly, we can indeed draw important parallels to it from others cases. Consider again the concept CAD (i.e., someone who acts in dishonorable or disrespectful ways towards a woman, or whose behavior towards a woman causes her to lose her sense of self-worth or self-respect). This concept, I take it, is one that we no longer use: calling someone a ‘cad’ would be seen today, perhaps, as at best anachronistic. It is a concept that is a part of a social world that we do not identify with, even though, as an ancestor of our own, contemporary social world, its values are not completely foreign. What interests us here, however, is the sense in which such a concept is different (the sense in which its use is say ‘old-fashioned’), and to capture this, we can consider what it would be like to look at this concept from a standpoint completely outside the interests and concerns that shape it. From such a standpoint, what we are interested in understanding is the point of having such a concept. We need to ask: Why did some group come to coin this term? Why did they single out certain men as deserving of some special (negative) recognition? The gloss of an answer that I proposed, in previous chapters, was one that centered on a certain view of women: men came to be called ‘cads’ in part because women were seen as special objects (or targets) of concerns and respect, in virtue of the fact that they were women. Their virtue (and, perhaps more importantly, their sexual purity) was something that required protection from the wanton interests of some men, who would use their charm and (often) aristocratic up-bringing as a means of manipulating women for their own selfish gratification. This view was no doubt driven by the interests and concerns that this 243 Victorian-era group had in women, and it provided the concept CAD with a rationale: the point of classifying certain men as cads was that doing so would discourage the sorts of behavior that threatened the virtue (and sexual purity) of women. Such a concept, no doubt, enabled its users to ‘find their way around this social world’. 3 It provided them with a certain sort of ethical knowledge. Acting like a cad was something to be avoided, and such behavior was not only deeply shameful, but merited strong contempt from others. Being the victim of a cad also made certain reactions appropriate, ranging from reactions that we might think, for different reasons, are appropriate, such as sympathy, to those that we might think were harsh and unfair, such as the social ostracism that women sometimes faced. But, if it is right to say that they had ethical knowledge under this concept, it is also right to say that this is knowledge that ‘we’ can no longer share. 4 As historian Angus McLaren notes: The triumph in the late twentieth century of the belief that women are not passive beings who have to be ‘protected’ from men presumably explains why it is now hard to think of any polite word commonly employed to designate the man who takes sexual advantage of a woman. 5 Our view of women has changed, as has our view of the responsibilities the community has in protecting a women’s virtue (and sexual purity); no doubt these changes are due to social events such as the equal rights movement for women, the ‘sexual revolution’, and the resulting onset of feminism. These events are all products of certain forms of critical reflection being brought to bear on a certain view of a woman’s place in society. As a 3 (Williams 1985, p. 151) 4 The ‘we’, I take it, relates to members of contemporary Anglo-American culture. Perhaps, this class is too wide; if it is we might consider a subgroup that considers the relevance of certain ‘women’s movements’ of the last century to be a part of its cultural history. 5 (McLaren 1997, p. 59) 244 result, the social background against which this Victorian-era concept flourished was transformed: the interest in seeing women as special objects (or targets) of concern and respect, was replaced with a more egalitarian view of the sexes. Because of this change, the concept CAD is one that we no longer see a point in using. The concept classifies men in a somewhat arbitrary fashion, since women are no longer seen as especially deserving of concern, respect, and protection, in virtue of the fact that they are women. That we see the distinction drawn by this concept as, at bottom, arbitrary, undercuts our ability to ascribe the proper sort of conviction to the judgment that someone is a cad: we lack the ability to treat such a judgment as genuinely expressive of something that a (contemporary) speaker believes. 6 And so, if there was ever a type of ethical knowledge involved in the use of this concept, and I think that there was, it is knowledge that we cannot share. Coming, in this way, to reflectively view a thick concept as arbitrary, however, is only a possibility—a group that reflects on how and why they classifying acts and traits as they do, and thereby comes to see how such modes of classification depend on certain interests and concerns they have, might very well continue carrying on with those concepts. 7 What interested Williams were the logical ramifications that followed from a group coming to see (some of) their thick concepts as arbitrary. The goal, in this final 6 Of course, as I noted above, this point is an idealization of sorts: our interests lie at a relative distance to those of the Victorian-era. I am assuming that the distance is sufficient to undermine the claim that we can sincerely use such a concept. If one thinks differently, then perhaps the example is a poor one. In which case, one might substitute it for something more exotic, such as the Yanomamo notion of certain men being ‘fierce’ or Aristotle’s concept of the ‘great-souled man’. For the former, see (Chagnon 1968); for the latter see (Aristotle 1998, pp. 85-89). 7 See, for instance, Williams’ remarks on in (Williams 2006a, pp. 193-4). I will return to this point towards the end of the chapter. 245 chapter, is to fill out the details of this picture, and, in the process, hopefully make Williams’ view a more plausible position to hold. In doing this, the account of thick concepts that I developed in Chapter 3 will be of central importance. There I attempted to elucidate a structure to thick concepts that would help to make sense of some of the independent problems that such concepts were thought to face: especially in relation to the idea that such concepts were shapeless with respect to the neutral or nonevaluative features of actions and agents, and whether they involved or presupposed specific emotional or affective reactions. Recall that, in Chapter 3, I argued that thick concepts should be seen as involving two different components: a set of standards or criteria and an ethical rationale. In this Chapter, we will begin by considering these two central elements in more detail. Sections 5.1 to 5.3 will look more closely at the related notions of a set of ethical standards (or criteria) and that of an ethical rationale. Central to this will be developing an account of how a sympathetic outsider might be incapable of using an alien thick concept, even though she might clearly understand what that use involves. Section 5.4 will then return to the ending discussion of Chapter 4 and reformulate the argument that leads Williams’ to the conclusion that critical reflection might destroy ethical knowledge. Section 5.5 will then develop an explanation as to how this sort of loss is possible. The crux of the explanation will depend heavily on the plausibility of the picture, provided in sections 5.2 and 5.3, of the sympathetic, yet un-indentified, outsider. At its essence, ethical knowledge is destroyed for a community when reflection places members of the community in an epistemically similar position to that of the sympathetic, yet unidentified outsider. Finally, in Section 5.6, I will explain and defend Williams’ 246 view that confidence might be the best ethical state, as an attempt to provide a model of ethical authority that can be applied consistently and transparently to the use of those thick concepts that withstand critical reflection. Section 5.1: Criteria, Ethical Standards, and Disagreement In Chapter 3, I introduced the idea that thick ethical concepts are best understood as presupposing specific ethical or evaluative standards. This was something that I took to be nascent the fitting-attitude approaches that we considered in Chapter 2 (e.g., McDowell, Wiggins, Dancy, and especially Gibbard). These approaches to understanding the nature of a thick concept took concept-specific attitudes (or emotional responses) to play a role in determining when the concept applied: a thick concept would apply to some act, just in case, some emotional response, unique to acts falling under that concept, was warranted or merited towards the agent who performed that act. Such judgments necessarily presupposed certain standards, where the relevant standards would have a hand in both (i) explaining the application of the concept, and (ii) warranting the concept-specific attitude. The idea I suggested there was that an act fell under a positive (negative) thick concept just in case that act merited a positive (negative) evaluation in virtue of its falling under that concept. That such a standard is presupposed is supported by the idea that the very point of having the concept is one that is related to a certain form of behavioral regulation: the concept came to be as a means of encouraging (or discouraging) a certain form of behavior. I argued in Chapter 3 that, if this is right, then the explanatory work of the concept-specific attitude simply falls away, and that explanatory work actually rests with the presupposed standards. 247 This discussion also involved an additional element, namely that the sincere use of such a concept presupposed certain interests, namely those that would underwrite the agent’s acceptance of the underlying standard as action-guiding. Such interests provided the ethical rationale for having the concept in the first place, and helped us to understand how the concept was to be applied from context to context. Yet and still, as it stood there, the notions of an ethical standard and an ethical rationale were both fairly unspecific. My goal in this section (and the next) is to clarify what I take each notion to involve. This discussion will be a transitional step to a larger one relating thick concepts to the idea that their use can be undermined by critical reflection. To begin, though, we should start by elucidating the idea of an ethical (or evaluative) standard, and how this requires us to see the sense in which thick concepts are (using Bernard Williams’ phrase) ‘world-guided’ perhaps a bit differently than other commentators have understood it. So, how should we understand the notion of an ethical standard? Here, I think we will be aided by starting from some remarks Stanley Cavell has made on the notion of criteria (or at least as he is apt to find this notion used in Wittgenstein’s Philosophical Investigations 8 ). Cavell writes: …criteria are specifications a given person or group sets up on the basis of which (by means of, in terms of which) to judge (assess, settle) whether something has a particular status or value. 9 And, again that: Criteria are “criteria for something’s being so,” not in the sense that they tell us of a thing’s existence, but of something like its identity, not of its being so, but of its 8 (Wittgenstein 2001 [1951]) 9 (Cavell 1979, p. 9) 248 being so. Criteria do not determine the certainty of statements, but the application of the concepts employed in statements. 10 Because criteria determine what an object must be like in order for a related concept to apply to is (i.e., they determine the object’s identity in relation to some concept), but do not settle (substantive) questions as to whether any particular case meets them with certainty, the very notion of criteria leaves itself inherently open to the question of when and whether it is appropriate to doubt the application of some concept. Given this, the very presence of criteria, on Cavell’s view, never constitutes a(n adequate) response to disagreement and skepticism. 11 This is because the role of criteria is to determine what counts as evidence for a concept to apply, and is not to determine whether, when that evidence is found to be present, this thereby settles definitively that the concept applies. In this way, any distinction between having criteria for a concept and having independent standards for applying that criteria is likely to be blurred (as Cavell admits): the criteria for a given concept determines what counts as a reason (or evidence) for applying that concept, but there is no further question about how that criteria is to be applied, no further set of specifications for applying it, even though there can be a question as to whether the criteria applies or not, and if so how well. 12 10 (Cavell 1979, p. 45) 11 This (I take it) is one of the central themes Cavell defends in The Claim of Reason. This can be contrasted with the view offered by (say) Rogers Albritton, who takes the notion of criteria to be a basis for certainty in a claim and thus a response to skepticism, in his (Albritton 1959). Cf. (Foot 1978a) and her use of thick concepts to show that ethical disagreements are not always liable to breakdown in the ways that early emotivists and prescriptivists (such (Stevenson 1937), (Ayer 1946), and (Hare 1952)) have held. Foot, I think, was right to deny that such emotivists and prescriptivists captured the sense in which such disagreements might breakdown by appealing to a fact-value distinction, but wrong to conclude that such interminable disagreements were impossible with thick concepts. 12 (Cavell 1979, p. 13). In the sense then that I am using the notion there is no hard and fast difference between having a criteria and having a shared practice that uses that criteria. Shared practices are not 249 Of course, these remarks are meant to apply quite globally—that is, to all concepts—and my interest in them is narrower than this. We are concerned with this view of criteria as a means of helping to elucidate the idea that thick concepts presuppose certain ethical standards. What I have in mind is the idea that thick concepts presuppose ethical standards insofar as the classification made by a thick concept in question are connected to the aim of regulating behavior and social interactions. For instance, the concept CAD presupposes the ethical standard that cads are to be evaluated negatively (this is the ‘status’ or ‘value’ that is conferred upon individuals who fall under this concept). Such a standard is importantly connected to the interests a group has: namely interests aimed at discouraging men from behaving as cads. And so, where the criteria of this concept determines the sort of qualities a man must possess in order for the concept to apply, the ethical standard which is presupposed by the concept is one where these qualities are viewed negatively. Depending on what the ethical standard is that underlies a thick concept, the acts falling under that concept will be taken to do so because they are understood as having or exemplifying certain ethical features. None of this, however, entails that simply because one thinks that a thick concept’s criteria are satisfied (and that the act or individual has or exemplifies the relevant ethical features) that they really are: in other words there is ample room for individual error. The idea of ethical standard, as I am attempting to elucidate it here, is closely related to the sense in which we might grade or assess something. If grading is anything like what it purports to be, or what people who grade and assess generally take these simply public, but they are logically antecedent to any particular individual becoming a part of, or understanding that practice: it must be present to be learned or understood. 250 activities to be, then to grade something is not to (or not simply to) express one’s attitudes towards the object graded, nor is it simply to describe the object graded. Rather, to grade is, in this sense, to apply a standard: we might equally say that to grade something is to apply a description to it for a certain purpose. In grading an essay, for example, the question of whether the essay is deserving of a certain letter grade (or whether it is appropriate to describe it as, say, a ‘B’ essay) depends on the prior question of whether it has certain features. But these features are grouped together as exemplifying the grade in question because we attach a certain value to these features, and to this extent the application of such a grade presupposes a standard that attaches such a value or import to them. Moreover, the importance assigned to the grade will be internally related to the point of having the standards to begin with: in this case to encourage the development of a certain level of competence with the subject matter in question. In this way, ethical standards are internally related to there being rationales behind the use of those standards: interests that give a practical point to the standard being what it is. To clarify the sense of ethical standards that I have in mind here, 13 it might be helpful to consider two examples: 13 For an excellent discussion of the concept of a standard, which has greatly influenced my views here, see (Copp 2001, pp. 19-22). Copp takes the notion of a standard to be something like a norm or a rule (following Allan Gibbard’s discussion of a norm in (Gibbard 1990, p. 46), something that is expressible by an imperative. Because of this Copp thinks that standards are not truth-assessable. I agree that standards are not truth-assessable, but this has to do with their status, not with how they might be expressed semantically. Indeed, in the examples I give below, none of the standards are expressed with imperatives, and it would be quite difficult to see how they could. On my view, being a standard has nothing to do with their being expressible in terms of an imperative (or set of them). Rather, it has more to do with their status in the very practices of evaluation and assessment of which they are a part. On this see (Wittgenstein 2001 [1951]) §50: 251 University Policy Concerning Hazing: 11. 14 Psychological Hazing, which is defined as any act or peer pressure which is likely to: (a) compromise the dignity of any student with the organization, (b) cause embarrassment or shame to any student affiliated with the organization, (c) cause any student affiliated with the organization to be the object of malicious amusement or ridicule, or (d) cause psychological harm or substantial emotional strain. Section H, Article 1: (1) 15 For the purposes of this Declaration, torture means any act by which severe pain and suffering, whether physical or mental, is intentionally inflicted by or at the instigation of a public official on a person of the purposes as obtaining from him or her or a third person information or confession, punishing him for an act he committed or is suspected of having committed, or intimidating him or other persons…. In both cases, the criteria elicited are an attempt to describe (in more or less specific terms) the features that an act must possess in order for certain thick concepts to apply (e.g., PSYCHOLOGICAL HAZING and TORTURE, respectively). Yet, the criteria themselves presuppose certain ethical standards that are in line with the point of having these concepts to begin with: which is (in both cases) to discourage certain forms of behavior. These standards attach a certain disvalue to the features that make a certain act an instance of psychological hazing or torture. And so, while the question of whether these concepts apply (and how well) are largely questions of whether the relevant criteria are met (and how well), the question of whether (and how well) these criteria are met are also There is one thing of which one can say neither that it is one metre long, not that it is not one mertre long, and that is the standard merte in Paris.—But this is not to ascribe any extraordinary property to it, but only to mark its peculiar role in the language-game of measuring with a metre- rule. The role that the standard metre in Paris has is that of being the standard for a certain form of measurement. But, in being the standard for a certain form of measurement, there is no sense or point to reflexively applying it to itself, in order to claim that it is also a metre-long. This would involve something like a confusion over the idea that there is no difference between a measuring device and what is being measured. 14 (University of Southern California 2008, p. 147) 15 (United Nations 2004) 252 related to the question of whether (and how much) a certain sort of act is to be discouraged. In Cavell’s terminology, then, while the concept’s criteria determine the identity of the things that these terms correctly apply to, the concept’s evaluative standard governs the practical status of the objects which fall under that concept. Of course, why such an act has the practical status that it does is not something settled by the standard itself. Rather, the answer to this question is settled by the rationale which went into establishing the criteria and ethical standards to begin with (e.g., those interests that aim to prohibit psychological hazing or torture). Importantly, however, that thick concepts possess specific criteria and ethical standards does not foreclose on the possibility of substantive disagreement. Disputes can still arise with regard to whether those standards apply in a particular case. This, however, might lead to another worry, namely that the possibility of interminable disagreement undermines the rationality of their use. A similar worry to this is voiced by Alasdair MacIntyre: …if we possess no unassailable criteria, no set of compelling reasons by means of which we may convince our opponents, it follows that in the process of making up our own minds we can have made no appeal to such criteria or reasons. If I lack any good reasons to invoke against you, it must seem that I lack any good reasons. Hence it seems that underlying my own position there must be some non-rational decision to adopt that position. (MacIntyre 1984, p. 8) MacIntyre’s concern rests on the assumption that, unless we have established criteria that would provide everyone with reasons which would either compel or guarantee their eventual agreement, then we must reflexively conclude that we really lack good reasons for the positions we stake out in applying them. If we cannot deploy criteria and ethical standards that command universal authority, then our appeals to there being such criteria 253 and standards at all must ultimately be taken as a façade for some (ultimately) non- rational process of deciding the matter. What MacIntrye seems to rule out is the possibility of two speakers taking up reasonably different positions with the question of whether something is cruel or generous: the positions are ‘reasonable’ in the sense that both are allowed by the communally accepted criteria for these concepts. There is nothing in having such criteria that need preclude two individuals from reasonably disagreeing over whether a certain act was indeed cruel or generous. What is precluded, however, is whether to disagree about the value accorded to the application of such concepts. To disagree about this would be to question the point of classifying acts in terms of this concept at all, and this, while indeed a type of ethical disagreement, is not one that can be understood in terms of whether the concept itself applies. The disagreement is over whether such classifications should be made at all. Failing to foreclose on the possibility of reasonable and substantive disagreement in the application of a thick concept, however, does nothing to support the idea that such disagreements are (i) manifestly irresolvable, (ii) unavoidable, or (iii) even probable. The possibility of such substantive evaluative disagreement is simply a possibility: it does not stand in the way of their being agreement on when and how such concepts apply. And the fact that such disagreements do not regularly occur (as a matter, perhaps, of sociological fact) can only be taken as a testament to the idea that, in sharing a set of substantive evaluative standards, a group also largely shares a sense as to how those standards should be applied. 16 But the empirical fact that there is such agreement between users of the same thick concept should not be read back into the content of the 16 Cf. (Cavell 1969) and the notion of a ‘whirl of organism’. 254 concept as somehow providing an explanation as to why this is so, and in doing so, foregoing or precluding this sort of disagreement. This hopefully brings to relief the sense of ethical standards that I have been assuming for much of the discussion in this dissertation. This discussion, however, also helps to clarify the emendation I made to McDowell’s views on thick concepts, in Chapter 2, and what is required for someone to understand a given thick concept. There I argued that such understanding should be seen as involving how to argue for and justify the application of such a concept. If thick concepts presuppose specific ethical standards, then, in grasping the underlying standard, one would understand how to argue for or against that standard having been met in a given case. 17 Understanding an alien thick concept, then, is not a matter of ‘predicting its extension’ (as McDowell puts it), or being able to generate accurate hypotheses about future applications and withholdings, but rather a matter of being able to argue for its application in a socially recognizable way (i.e., a way that can be understood by other, native, users of the concept). Given this relationship between thick concepts, criteria, and ethical standards, we should take care with how we consider such concepts to be, as Williams put it, world- guided. Some commentators tend to treat this idea as roughly equivalent to some unspecified notion of description, 18 or as an application of the idea that the application of a thick concept is empirically verifiable, 19 or that “when [thick concepts] are competently 17 Cf. (Raz 1999, p. 130): “…part of understanding any concept is having some grasp of what counts as relevant evidence showing that it applies or does not apply.” 18 See (Burton 1992); (Tappolet 2004); see also (Milgram 1995). 19 See (Altham 1995, p. 162). 255 applied the reliability of the resultant judgments can be explained in terms of their causal relation to the relevant states of affairs ‘out there in the world’.” 20 Each of these characterizations is tempted by the thought that what is indicative of the world-guided nature of thick concepts is that they are (in some sense consistent with Williams’ other remarks on this matter) answerable to the world as it would be conceived non- evaluatively (i.e., independently of ethical concerns and interests). This is a mistake. The sense of ‘world’ in world-guided is better understood as relating to a specific social world: an indigenous constellation of values, interests, concerns, relationships, roles, institutions and other ethical concepts, that are together constitutive of a group’s way of life. 21 (Cf. the idea of a Victorian social world noted in the introduction to this chapter.) It is because of their connection to this rich body of ethical considerations that thick concepts can be seen as guided in their application at all. Attempting to explain the reliability of those judgments in terms of their causal relationship to states of affairs ‘out there in the world’, as Nicholas Jardine puts it, 22 unhelpfully encourages the thought that there is something outside the social world which contributes to their governance. It is also doubtful that Williams himself would agree to such a characterization. If such a description of his views were correct, this would motivate the idea that there is something severable and recoverable in the application of such concepts, apart from the interests and concerns the community has in applying them. This, however, places us squarely back in 20 (Jardine 1995, p. 32). Jardine attempts to argue that ‘world-guidedness’ is a poor standard for objectivity. This criticism however would seem at odds with the thought that (for Williams) thick concepts are both world-guided and non-objective, and so it could hardly be a standard for objectivity. 21 For an illuminating discussion by Williams as to how we might understand the difference senses in which ethical claims are related to ‘the world’, see his (Williams 1995a, pp. 177-78). 22 (Jardine 1995, p. 32) 256 line with the sort of views of thick concepts (as involving distinct factual and evaluative components) that Williams clearly, and rightfully, rejects. 23 Section 5.2: Ethical Rationales and Action-Guidance Beyond presupposing specific criteria and ethical standards, I noted in Chapter 3 that thick concepts also presuppose an ethical rationale. The role of an ethical rationale is needed, I think, to help explain two points that are typically overlooked (or underappreciated) in discussions of thick concepts: one relating to (what I will call) the ethical identity of a thick concept, and the other relating to the fact that using a particular thick concept is not something that is necessarily forced on someone as a means of describing various acts and individuals. However, in the course of defending this notion, it is first important to distinguish what I have in mind from McDowell’s more general thesis that thick concepts have to be understood in relation to a certain “conception of how to live.” 24 What I intend by the notion of an ethical rationale is certainly consistent with McDowell’s remarks, but, is narrower in focus. An ethical rationale involves a constellation of interests and concerns which relate to a sphere of human activity within the sort of broader ethical conception McDowell appeals to. Both the specific conception of this sphere of human activity and the interests themselves are determined by this 23 Williams notes that “there is no reason at all to think that people could substitute for a linguistic practice the term in which that practice was psychologically or sociologically explained.” (Williams 1995d, p. 187) Williams arguing against centralism (the idea that thin concepts like RIGHT and GOOD are conceptually and explanatorily prior to thick concepts—cf. (Hurley 1989, pp. 9-11)) but I think his point has a broader application: what is lost on centralism is the idea that a group’s concerns and interests cannot be taken out of the story in how we understand their thick concepts, such that they are represented as simply approving or disapproving of certain neutral described acts and character traits. Such interests and concerns provide the very point of making such classifications to begin with. This point, I take it, also applies to how we should understand the very notion of world-guidance. 24 (McDowell 1998f, p. 71) 257 broader ethical conception. For instance, in discussing Gibbard’s example of the concept LEWD, I noted that it was a concept that was connected to, or presupposed, certain views on sexual modesty and personal dignity. Those who use this concept have an interest in upholding such views, and register lewd behavior as a sort of failure in precisely this regard. Having such an interest explains why certain sexual displays are grouped together as the sorts of things that need to be discouraged. Now to the first point noted above. How does such a rationale shape the ethical identity of a thick concept? It does so by helping us to understand the point of having the concept in the first place. Here I think it is instructive to highlight a remark made by Michael Dummett, which, although directed towards what an adequate account of the concept of truth must involve, nonetheless provides us some insight into what it is that we are trying to understand when it comes to thick concepts: We cannot in general suppose that we give a proper account of a concept by describing those circumstances in which we do, and those in which we do not, make use of the relevant word, by describing the usage of that word; we must also give an account of the point of the concept, explain what we use the word for. Classifications do not exist in the void, but are always connected with some interest which we have, so that to assign something to one class or another will have consequences connected with this interest. (Dummett 1978, p. 3) In connection to thick concepts, we might say that, while it is true that thick concepts classify acts and individuals, it would be a mistake to equate the point of such concepts simply with an interest in making certain sorts of identifications. We mischaracterize their role in a group’s conceptual practices, if we simply consider them to be, as Williams stated, devices “for dividing up in a rather strange way certain neutral features of the 258 world.” 25 In order to understand the concept we must understand the interests that are behind classifying acts and traits in that way, and merely appealing to the idea that we have an interest in such classifications fails to explain this. Julius Kovesi offers a good example that will help explicate this point. A teacher might draw certain shapes on the chalkboard for his students and coin the term ‘tak’ by pointing to those examples that have a pointed projection. After going through several examples, we might assume that the students are able to determine which novel drawings a teacher makes on the chalkboard count as taks or not. Kovesi denies, however, that, simply in virtue of in being able to do this, the students have thereby learned the meaning of the word ‘tak’: There is…something strange in assuming that the pupils can leave the classroom and say: “Now we know the meaning of ‘tak’.” This is so not only because they may never see any taks outside the classroom; after all they might, on occasion, see figures that exhibit the characteristics of taks. If however they see one, what are they to do with them? Should they stop and say ‘tak’ each time, or point it out to someone saying: “Look here is another tak,” or should they perhaps make a record of the number of taks they see each week? What is the point of the word ‘tak’? (Kovesi 2004 [1967], p. 31) The reason why the students will not know the meaning of ‘tak’ is that, given the structure of the example, this word has yet to be given a meaning. Kovesi continues: Without the need for ‘tak’ in a way of life we will not start forming the notion [concept] or using the word, and the word will not acquire meaning. The need for ‘tak’ might arise in a factory where tak-shaped figures for some reason have to be sorted out, or in a new game where tak-shaped counters are used. Possibly the word might play a part in a larger system, of crystallography say, or meteorology, where the recognition of tak-shaped crystals or cloud-formations may help us in manipulating, controlling or predicting systems. (Kovesi 2004 [1967], p. 32) 25 (Williams 1985, p. 142) 259 According to Kovesi, understanding how to go on in using a certain term, is tied to the interest related to developing it, and such an interest is never merely connected with identification. The same point applies to thick concepts: it is not that we are simply interested in identifying certain acts as (say) cruel. Our interest in discriminating between acts with the concept CRUEL rather stems from our interest in discouraging people from acting on certain reasons, namely those that involve causing suffering in others so as to delight in that suffering. The mistake is to think that, because we classify certain acts and character traits with thick concepts, our interest in doing so is tied merely to identifying those acts and traits. 26 Rather, we classify certain acts and character traits using thick concepts because we have other interests in such classifications, and these interests drive the very manner in which the concept classifies. To fully understand a thick concept, then, requires that one be able to understand why such a classification is important. The second point to note here is that there is nothing inherent in the notion of leading a good life or living in a social world which forces or demands the use of any 26 This assumption, I think, is behind certain remarks that Hilary Putnam makes in the process of arguing that thick concepts reflect clear counter-examples to a basic fact-value distinction. Putnam writes: …“cruel” can also be used purely descriptively, as when a historian writes that a certain monarch was exceptionally cruel, or that the cruelties of the regime provoked a number of rebellions. “Cruel” simply ignores the supposed fact/value dichotomy and cheerfully allows itself to be used sometimes for a normative purpose and sometimes as a descriptive term. (Putnam 2002, pp. 34-5) Putnam oddly seems to think that by ‘ignoring’ the fact/value dichotomy, thick concepts are to be taken as at home on either side of that very dichotomy. The trouble with this way of stating things is that it is something a component theorist (or someone who wished to uphold such a dichotomy) would wish to uphold: in some cases the evaluative aspect of cruelty is present, in others it is not. Cf. David Wiggins’ insightful critique of Putnam’s position here (Wiggins 2006, pp. 378-79), and the idea that the very idea that there are such ‘purely descriptive’ uses is a part of what needs to be denied if someone, like Putnam, wishes to hold that such concepts ignore the fact/value distinction. 260 particular thick concept. 27 We know that there exists a plurality of social worlds, as well as a plurality of visions of the good life, that are partly constituted by the use of certain thick concepts. The plurality of (what we might call) thick vocabularies is a fact that must figure into our explanation of their nature. Because there is nothing compulsory in the use of any particular thick concept, why it is that some are used in one group and not another must be accounted for. Of course, a good deal of such an explanation will rests on how that social world came to develop as it did, but if the explanation is to help make sense of how such concepts can differ between groups, then it will have to also involve a reference to those groups’ different interests and concerns. Those who use a thick concept must, in some sense, identify with its rationale; they not only must want to use the concept but have a higher-order positive attitude towards that want. And, what is important for us, is the fact that such identification raises the cost of being able to sincerely use a thick concept to express a moral belief. Expressing a moral belief which has a thick concept as a constituent element in that belief requires an acceptance of the concept’s underlying ethical standard, and such acceptance presupposes that one identifies with the point of using the concept in the first place. While the sincere use of a thick concept is tied to identifying with its ethical rationale, coming to understand the concept is not. This is what makes the possibility of there being a stance that is sympathetic, yet unidentified, so central to our understanding of the nature thick concepts. Such a person is typified, as Williams highlights, by the 27 Cf. (Williams 2006b, p. 158), and Williams note that what has been overlooked by many moral theorists is “that there is no compulsion to using a given value concept at all.” Williams comment is in the context of discussing thick concepts, and so what I take Williams point to be is that there is no compulsion to using a given thick concept at all. 261 anthropologist or ethnographer. 28 What this opens up, however, is the possibility of there being certain incapacities that are rooted in the dispositions and distinctive interests of such an individual with respect to the thick concept in question, but which are incapacities wholly consistent with her understanding what that use entails. We will come back to this in the next section. First, we should consider an alternative way of explaining the ethnographer’s position with respect to the sincere use of an alien thick concept. This explanation is ultimately unsuccessful, but reviewing it will help to draw a contrast with what I think is the proper way of explaining this point. The explanation is one that Williams himself offers, or it is at least one implicit in a rather unhelpful example he offers to describe the ethnographer’s position in relation to some alien thick concept. The example turns on there being certain conventions governing the use of a term, where these conventions restrict its use to only a certain group of people. Williams writes: For a naïve example, we may imagine a certain school slang that uses special names for various objects, places, and institutions in the school. It is a rule that these words are appropriately used only by someone who is a member of the school, and this rule is accepted and understood by a group outside the school (it would have to be, if it were to be that rule at all). People know that if they use these terms in their own person they will be taken for members of the school, or else criticized, and so forth. Suppose that in this slang “Weeds” were the name of some school building. Under the imagined rules, an observer could not, entirely in his own person and not playing a role, properly say “Robertson is at Weeds.” 29 Williams’ aim here is to show how it might be possible for an outsider to agree that an insider spoke truthfully in making a judgment with a thick concept, but was nonetheless 28 See here (Williams 1986, p. 203); (Williams 1995c, p. 206); and (Williams 1996b, p. 32). 29 See (Williams 1985, pp. 143-44). To be fair, it is worth emphasizing that Williams considers the example ‘naïve’. 262 incapable of making the same judgment himself. This is intended to support to idea that there can be judgments involving thick concepts that constitute ethical knowledge, and that the outsider can understand them as indeed constituting such knowledge, even though he is unable to share that knowledge because he is incapable of having the beliefs that would be necessary for it. I agree that defending this sort of possibility is necessary for making sense of what is distinctive about the sort of ethical knowledge these concepts provide, and for understanding Williams’ claim that reflection can destroy or undermine thick ethical knowledge. The trouble is that the example he provides ultimately works against these claims, rather than providing support for them. There are two problems with the example. First, the force of it rests on the idea that outsiders accept that only students (i.e., insiders) can use this term. In this way, part (if not all) of the explanation as to why an outsider would take himself to be incapable of using that term would have to rest on the rather implausible ground that this person has already accepted a rule stipulating that only students at the school could use it. This would be an incapacity premised on a prior agreement not to use the term: which would hardly amount to an incapacity at all (it could be overturned simply by deciding to go back on the previous agreement). Of course, the ethnographer’s use of the term might imply that they are a part of the group, or earn the criticism from ‘insiders’, but neither of these considerations is suitably strong enough (on their own or jointly) to explain how the outsider would not be able to use that concept. If there is to be something distinctive about the idea that a person might understand such a term, but nonetheless be an ‘outsider’ to its sincere use, it will have to involve more than a conventional restriction 263 that depends on the antecedent fact that the outsider is an outsider. What we are looking for is an explanation of what constitutes his being an outsider in the first place, and Williams’ example presupposes this fact rather than explains it. The second problem with this example is that what is taken to be known within the example is something that is nonetheless available to the outsider, thus undercutting the effectiveness of what the example could show: namely that there was something the insiders knew which the outsider could not. Insofar as they are the only ones licensed to use the term ‘Weeds’, the insiders gain nothing in terms of knowledge. The fact that only certain people can call (say) the Administration Building “Weeds” does not mean that, when X makes the judgment “Robertson is at Weeds” he knows anything different from what the outsider knows in knowing that Robertson is at the Administration building. 30 Indeed, this point is suggestive of the very sort of picture of thick concepts that Williams was deeply opposed to: one that takes their content to be amenable to neutral description. Because of these two points, Williams’ example actually works against the case he is trying to make: namely to show that there is a sense in which the outsider cannot use a thick concept to make judgments in her own person, and, because of this, be unable to know certain things that would require the use of that concept, even though the concept itself can be intelligible to this person and be understood (in epistemically favorable circumstances) as one deployed in judgments constituting knowledge. All of this, however, rests on their being a sense in which the ethnographer can be incapable of using these concepts that is richer than mere conventional stipulation. And yet, whatever 30 Cf. (Moore 2003, pp. 345-46). This is what I take Moore to mean in saying that “the message and the medium are easily separated” in Williams’ schoolyard analogy. 264 explains this, will have to be something that does not press us back onto the faulted view of thick concepts as having a content comprised of separable evaluative and non- evaluative components. Section 5.3: Identifying with the Point of a Thick Concept The picture I will advance, of how an outsider might genuinely be unable to use a thick concept, will be explained in terms of the extended costs that its sincere use would involve. Consider again the concept CAD. Our inability to use this concept, to form an ethical belief, is explained by the fact that the point of using such a concept is connected with interests and concerns that we no longer share. Such a concept is bound up with discouraging men from treating women in certain ways, simply in virtue of the fact that they are women; women are deserving of special protection from the wanton interests of men. Identifying with these interests is a necessary condition for being able to use CAD to form a moral judgment, and this provides us with a model for understanding how the use of the concept to distinguish some men from others can involve a cost that extends beyond the mere interest in being able to identify cads. We must tread carefully, however, in how we explain this as a genuine form of incapacity. Certainly, it cannot be a purely semantic or linguistic issue, that is, in saying that an outsider can be incapable of using such a concept we are not making a claim on the outsider’s linguistic competence in using the term. 31 This would scuttle the necessary thought that it would be on account of understanding what the concept meant (i.e., what its point was) that the outsider took 31 Joseph Raz treats the notion of parochiality in this way: a concept is parochial just in case it is a concept that not every thinker (i.e., a creature capable of possessing knowledge) is able to master. See (Raz 1999, p. 132). My discussion is an attempt to preserve at least one sense of being parochial that is not tied to semantic mastery. 265 herself to be incapable of using it. So, insofar as semantic competency is concerned, we might say, the outsider is capable of using the term: she can know what would count as correct and incorrect ascriptions of the term; she can know how the application of such a term is to be justified; and so on. The account rather must be one where it is in virtue of her knowing these things that she can find herself incapable of using it. 32 The sincere use of the term must require something that the outsider (in such a case) is simply unable to give. 33 Also, there are senses in which such a person can use an alien thick ethical concept, even though she does not accept the underlying standards and interests that govern its use. For instance, the anthropologist’s goal of describing the social world she is studying might require a rather refined understanding of their ethical concepts. In order to gain this, she must engage in certain sorts of hypothetical or imaginative exercises: namely ones where she would attempt to occupy the social and psychological position of one who actually uses that concept to conceptualize her social relationships and exchanges. 34 She might, then, try to think about these social relationships and exchanges from the point of view of someone who identifies with the rationale behind 32 Here I am in agreement with Dancy, when he notes that: “To reject a concept one has to do more than fail to see its point; one has to know what its point is and reject it for that reason.” (Dancy 1995, p. 269) 33 This issue is what I take to be at the heart of a disagreement between Gibbard and Blackburn in their symposium papers on thick concepts. Gibbard claims that he is not a “licensed user of the term [‘lewd’].” (Gibbard 1992, p. 281) Blackburn, however, denies this and thinks Gibbard indeed is a licensed user of the term. See (Blackburn 1992, p. 296). On my reading, Gibbard is trying to leave room for the sense in which he feels incapable of being able to make sincere judgments of lewdness (‘lewd’, to borrow from Oscar Wilde, is not one of his words). Blackburn, however, seems only concerned with the conditions that need to be satisfied in order for one to be considered a competent user of the term. I think, for obvious reasons, that Gibbard’s intuition here is correct, but that his use of a quasi-semantic term (e.g., ‘license’) invokes the wrong sorts of connotations with its relationship to semantic competence. 34 (Gibbard 2003a) offers an account of this in terms of hypothetical plans: to understand a thick concept is to plan on how to use it, under the wild hypothetical case of being an insider. 266 such a concept. Alternatively, she might find it necessary to play the role of an insider, and this might include issuing judgments using certain thick concepts that are constitutive of their way of life. What will be suspect about these judgments is not to be found in the judgments themselves, but in the subject’s relation to them: in both cases, even if the subject uses the concept correctly, there can still be a barrier to treating such judgments as genuine ethical judgments (i.e., as judgments that express that person’s ethical views). Imaginatively entering into or mimicking these judgments is different from making them in propria persona. What is missing is the sort of conviction that one must have in the relevance or importance of classifying with that concept, and because of this there is still a sense in which the outsider does not take themselves to be accountable to the truth of the claims they are making (even if the claims they make are, in fact, true). What is interesting about the outsider’s position, if we are apt to see it as manifesting a genuine incapacity to form certain judgments, is that this is perhaps best explained as a moral or ethical incapacity. 35 Seeing the cultural outsider’s inability to use a thick concept in such a way leaves us free to admit the obvious: that she may be nonetheless semantically or linguistically competent with the term. Here we will be helped by briefly discussing the notion of a moral incapacity. A moral incapacity, according to Williams, is an incapacity to do something knowingly. In this way, moral incapacities differ from say straight physical incapacities. If A is morally incapable of φ-ing (e.g., betraying a friend), there is still a sense in which A can φ, whereas if A is physically incapable of φ-ing (e.g., lifting a large car) there is 35 See here (Williams 1995b). My focus here will be on the central case of moral incapacity that Williams covers. There are other sorts of moral incapacities that he allows for, see, for instance, (Williams 1995b, p. 46 and 51). 267 not. In the case of straight physical incapacities, the incapacity is not conditioned on whether one knows what one is doing. Other incapacities, however, depend on how the agent views the action she is considering. If one is psychologically incapable of eating roast rat, for instance, because one finds doing so thoroughly disgusting, whether one is incapable of eating something (that turns out to be roast rat) depends on her knowing that that is what she is doing. Moral incapacities have this sort of dependency as well: Oedipus, we might assume, was morally incapable of killing his father, but this did not prevent him from killing a person who (unbeknownst to him) was in fact his father. The principle difference between moral and psychological incapacities is the fact that, if a person is say psychologically incapable of φ-ing, then it would follow that, if that person tried to φ, and knew that φ-ing was what she was trying to do, she would fail. With moral incapacities, however, this is not the case: someone who is morally incapable of φ-ing might very well succeed in knowingly φ-ing, if she tried. What is distinctive of a moral incapacity is the fact that she will not try. The reason, according to Williams, is that moral incapacities bear a special relation to an agent’s reasons and deliberations, and this marks a difference between them and other (psychological) incapacities to knowingly do something. When it comes to moral incapacities the deliberative conclusion an agent reaches (‘I cannot…’) is partly constitutive of that very incapacity: the incapacity depends on pre-existing dispositions that the agent has, but those dispositions are brought into focus by the very deliberations that are aimed at applying them to the case at hand. In this way, an agent’s moral incapacity is best represented as the output of a certain deliberative process. By contrast, psychological incapacities are always “inputs into 268 decision” 36 ; they mark off the boundaries of what is and is not decidable for the agent. The explanation, then, as to why an agent, who is morally incapable of φ-ing, will not even try to φ is that such an agent takes there to be decisive reasons against φ-ing. Given this brief description, we can use it to explain the sense (or senses) in which someone can be incapable of using a thick concept. In the typical case of a moral incapacity, namely those where the incapacity centers on the performance of some overt act, the root of the incapacity is a deliberative conclusion based on reasons that the agent takes to count decisively against acting in that way. In the case of being incapable of using a thick concept, however, there are two cases to consider. In the first, one is incapable of using a thick concept because one finds there to be decisive reasons against identifying with the concept’s ethical rationale. In the second, one is so incapable because one does not find there to be good reasons for such identification. The former reflects the sense in which someone might reject a thick concept’s ethical rationale; the latter reflects the sense in which someone might simply fail to accept it. The parallel to more straightforward cases of moral incapacity comes from the fact that, in each case, the incapacity centers on how the agent’s pre-existing evaluative dispositions (i.e., interests and concerns) have come to be focused on the question of whether to use the thick concept in question. There is an important difference between more straightforward moral incapacities and those that apply to the use of certain thick ethical concepts, and this difference might give one pause in thinking that the idea of a moral incapacity can do explanatory work here. In applying this notion to thick concepts, there is little sense to the idea that one 36 (Williams 1995b, p. 51) 269 could unknowingly do the thing that one has the relevant incapacity to do. Such a difference, however, does not undermine our drawing a relevant parallel. This comes from understanding the role that this possibility (i.e., of unknowingly doing what one is knowingly incapable of doing) has in marking off the incapacity as a moral incapacity. It helps sustain our understanding of the incapacity as one that depends on there being the right sorts of connections to the agent’s reasons and values. The role that this alternative plays, then, is one that marks out a way in which an agent is still capable of doing the relevant thing, even though (from a moral standpoint) she is not. When it comes to relating this notion to thick concepts, a quite similar way of expressing this underlying role is ready to hand. The same sort of possibility is explained, not by the speaker being able to unknowingly make a judgment using the relevant thick concept (assuming that such a notion is even coherent, which I doubt). Rather, it is explained in terms of the speaker understanding what the relevant criteria and standards for using the concept are; it is explained in terms of her knowing what would go into using it: even though she is morally incapable of using the concept, she still might have the semantic or linguistic ability to do so. This (I submit) parallels the way in which one might unknowingly act in a way that one is otherwise morally incapable of doing. And so it need not be taken as marking out a relevant difference between the two cases. Both can be seen as genuine incapacities because (i) there is a sense in which the agent is capable of doing that which she is morally incapable of doing, and (ii) the ground of the incapacity itself centers on the agent’s reasons and other evaluative dispositions. 270 If we are inclined to see the outsider as someone who can have a moral incapacity to use some alien thick concept, then there is indeed a sense in which the outsider is unable to know what it is that the insider knows. Since she cannot use such a concept to form judgments in propria persona, there is a sense in which she does not believe the proposition that would result from such a judgment. However, if she does not believe the resulting proposition, then she certainly does not know it either. This possibility will (in the section after next) provide us with a model for understanding Williams’ claim that reflection can destroy ethical knowledge, as it leaves open the possibility of her still understanding how others might genuinely believe such a thing. Before getting to this, however, we should briefly review the sort of reflective situation that Williams took to lead to such a loss of knowledge. Section 5.4: Reflection, Vindication, and the Loss of Ethical Knowledge Williams’ view that reflection can destroy ethical knowledge arises out of the thought that a convergence in the belief of some proposition can be explained in two very different ways: one supports the view that the belief in question expresses objective knowledge (or knowledge of an objective subject matter), while the other supports the view that the belief in question express a form of nonobjective knowledge (or knowledge of a nonobjective subject matter). Williams offers two powerful insights regarding objective knowledge. One connects objective knowledge with the possibility of vindicating the beliefs that express that knowledge. And the other connects such vindication with that of an explanation as to how we came to hold that belief. As 271 Williams puts it, for certain beliefs there is the possibility that what explains justifies. 37 The trouble, however, is that, left as it stands, this slogan fails to reveal a deep ambiguity in the relationship between explaining and justifying a belief. Understanding this ambiguity is central to understanding Williams’ general position on the relationship between certain bodies of knowledge and reflection, and so it would be helpful to carefully make these differences plain. Let’s start by noting that the connection between explanation and justification can operate on two different levels. Consider the first level to be that of a particular belief or a judgment made on a particular occasion: call this the judgment-level. At this level, there are certain operative conceptions of appearance, reality, and error. They are conceptions that apply specifically to our understanding of the formation of particular beliefs and the making of particular judgments. For instance, think of someone who believes that p because there is some situational distortion that is preventing him from seeing that really not-p is the case. We can understand this counterfactually: if the context had not provided her with this distortion, then she would not have held this belief. Alternatively, however, we could relate her circumstances to the possibility of revision given the right explanation of what those circumstances are. If she were given a correct explanation as to why she formed that belief in the first place, then she would no longer hold that belief. What appeared to her to be correct (accurate, really the case) was not, and as a result her forming the belief was in error (even though it might not have been an error she was responsible for or could have avoided making). 37 For Williams this is the telling difference between secondary-quality concepts and thick concepts, not (as with McDowell and Wiggins) their inherently normative aspects (i.e., their meriting certain affective responses). See (Williams 1985, pp. 149-50). 272 The conceptions of appearance, reality and error, offered at the judgment-level, cut across a distinction between objective and nonobjective knowledge, and have a clear application when we think about particular judgments using thick concepts. Take an extremely simple case: someone walking along a busy street might come to think that the behavior of some stranger to what seems to be his companion is incredibly cruel and mean-spirited. She might think this, until she finds out that this was a part of some street performance, and the players intentionally mixed themselves into the crowd for dramatic effect. Here an explanation of this person’s belief would reveal that her judgment was incorrect by revealing those distortions that stood in the way of her forming an accurate picture of what was the case, given the concepts she would naturally deploy (in this case CRUEL and MEAN-SPIRITED). There are, however, cases where such distortions are absent, and because of this, explanations as to how a thinker came to hold the belief will justify her holding it. At the judgment level, such explanations are part and parcel with the everyday practice of correcting particular judgments because the thinker was (say) mistaken about certain pieces of information. Such explanations are a part of our view that we can track the truth and reliably believe that p when that p. 38 And as I noted above, this assumption is no less a part of our use of thick concepts, as it is with other concepts. However, in order to objectively vindicate a belief, something more is required than simply (i) having the thinker apply her concepts carefully, (ii) it’s not being the case 38 Cf. (Jardine 1995, p. 32). Jardine, I think, fails to see that the reliability of a type judgment is not, in itself, a measure of objectivity (I can be reliable in judging in accordance with my preferences, because I clearly know what I want). By my lights, reliability and the idea of being able to ‘track the truth’ are issues formulated at the wrong level of reflection to understand Williams’ point. 273 that the situation presents the thinker with various obstacles for doing so, and (iii) it is indeed a situation that the concept correctly applies to. 39 The additional factor that we need to consider (and call into question), in order to sustain a claim to objective vindication, is something that can be found in what was assumed at the judgment-level. In considering the validity of beliefs at the judgment-level, the relationship between explanation and justification depended crucially on assuming the use of certain concepts, namely those thick ethical concepts in which the thinker’s beliefs were actually formulated (CRUEL and MEAN-SPIRITED). 40 At a conceptual-level, however, this assumption itself must be drawn into question, and because of this we will need to 39 In line with this, Susan Wolf writes: A person [according to Williams] may have ethical knowledge, for example, that such and such an institution is corrupt, or that such and such an individual is pious, or even, I presume, that such and such an act is the right thing to do. But his knowledge of these judgments does not imply knowledge that the concepts and values on which these judgments really are good concepts and values for a society to have. (Wolf 1987, p. 828) And yet, if ethical thought were to be objective it would certainly have to be something more than its simply being good concepts and values for the society to have, for, presumably, they might very well think this. Rather, they would have to be able to formulate an explanation grounding the thought that their concepts and underlying values were the best ones to have (i.e., that having any others would be an inferior ethical position). 40 Consider here how David Wiggins’ formulation of a vindicatory explanation for a belief that p overlooks this issue completely: A subject matter is objective or relates to an objective reality if and only if there are questions about it that admit of answers that are substantially true. It is sufficient for some judgment that p to be substantially true that one could come to know that p. One can come to know that p only if one can come to believe that p precisely because p. And one comes to believe that p precisely because p only if the best full explanation of one’s coming to believe that p requires the giver of the explanation to adduce in his explanation the very fact that p. What follows from this is that his explanation will conform to the following schema: for this, that and the other reason (here the explainer specifies these), there is really nothing else to thing but that p. (Wiggins 1991, p. 66) It is worth noting that for Wiggins ‘a full explanation of one’s coming to believe that p’ is understood as the providing of reasons in favor of believing that p. The point here is not that there is anything wrong with providing this sort of vindicatory explanation for a given belief. It is not to deny that, from the assumed use of certain concepts, there really might be nothing else to think but that a certain act or institution is cruel. Rather, the point is that there are additional (and decidedly different) explanations that would need to be provided once we moved from the vindication of a given belief to the vindication of the terms in which that belief is cast. And, as I understand Williams, these questions can come apart under reflection. Cf. (Williams 1996a, p. 213). 274 consider and evaluate what sorts of alternative concepts might have been deployed to form opinions in that context. The idea is that, if a belief is to be objectively vindicated then the explanation of our coming to use these concepts will have to speak to whether there were equally legitimate conceptual alternatives that could have been deployed. Put differently, the explanations offered at this level will be ones directed towards the practice of using a certain concept as a whole. Conceptual-level reflection is concerned with providing an explanation as to how it is that we came to have the very concept that our judgment is framed in, and whether we would have been in error to have used some other concept. The idea is that some explanations at this level have the potential for being vindicatory, while others do not. However, not being ‘apt’ for vindication at the conceptual-level does not thereby entail that the use of such concepts is illegitimate or that their use of them is automatically disqualified from expressing knowledge. How we should view legitimacy in a thick conceptual practice, however, is a complicated question, and one that we will have to put off until near the end of the chapter. Just as we did at the judgment-level, we can make room at the conceptual-level for operative conceptions of appearance, reality, and error. These notions, however, must be interpreted differently. At the conceptual-level, the distinction between appearance and reality must relate to how we can respond to alternative concepts and conceptions, and whether there is room, within certain forms of inquiry, for the thought that some concepts are simply better suited for that inquiry than others. It is at the conceptual-level, then, that we can make sense of objective vindication: that is where we can make sense of a certain concept (or conception) being better than any competitor, relative to some 275 domain of inquiry. 41 Vindicating our use of certain concepts (or conceptions) depends, however, on our being able to provide (non-vacuous, non-question begging) explanations as to why our use of some alternative concept would put us in a worse conceptual position. It would be tied to our providing an error-theory directed against various potential competitors that reveals them to be inferior or the product of some misunderstanding. 42 Such an error-theory would show how an alternative concept (or conception) suffered from problems which prevented its use in various judgments and claims from being straightforwardly true. Such judgments would thereby come to be seen as constitutionally defective. The idea is that if such an error-theory can be provided then this would open up the possibility seeing the use of that concept as strongly vindicated: to use some alternative would be to adopt a conceptually inferior view of the subject matter in question. Consider again how this plays out in the case of Aristotelian and Galilean conceptions of motion, as noted in Chapter 4. On an Aristotelian conception of motion the continued movement of an object depended on there being a continuous application of force, and so, an object would continue to move only as long as some force was being applied to it. Such a conception of motion is clearly inferior to the Galilean conception, informed by the law of inertia, at explaining the behavior of things like the movement of 41 I am leaving out here the thorny issue of their being equally legitimate, though incompatible, absolute conceptions. In such a case, one view would be incapable of developing a non-question begging, non- vacuous, explanation as to what the alternative conception was missing or mistaken about. Cf. (Moore 2007, pp. 27-30). It is perhaps enough to ground the sort of vindication involved here if we can say that some alternative conception can be shown to be defective or the product of misunderstanding. 42 In this way Williams’ absolute conception of the world comes out as a regulative goal that we can assign to any practice attempting to formulate concepts that describe the world as it is anyway. Our having such a regulative goal in science, and lacking on in ethics, makes a great deal of difference in whether we can genuinely ascribe a suitable sense of error in the latter case. 276 cannon balls, javelins, and arrows. Not only is there something which could feature in a theory of error (something which the Aristotelian clearly missed, namely the concept of inertia), there is an explanation as to how the Galilean conception of motion was arrived at that (at that time) vindicates that conception over the Aristotelian one. Importantly, however, both conceptions could be seen as attempting to provide (or enable) an accurate account of one and the same set of states of affairs. It is because we can see both as being engaged in such a similar enterprise, that we can locate both conceptions against a common goal, that there is the conceptual space for finding one better than the other relative to that goal. At the conceptual-level, however, the prospects for vindicating one particular set of thick ethical concepts simply falls apart. The difference between thick concepts and scientific concepts is at least this: there is a coherent sense of their being a vindicatory conceptual history in science, one where different attempts at developing concepts which characterize the world as it is anyway are all aimed at achieving that very goal, whereas, in ethics, insofar as this involves the use of certain thick concepts there is no such hope. This is because our response to diversity and potential alternatives will be decidedly different here than it will be with scientific concepts, or even secondary-quality concepts. When it comes to even secondary-quality concepts, what diversity might be found in such concepts (e.g., our seeing a certain object as red, and the alien hearing a certain pitch when in the same object’s presence) is no threat to a vindication of their use, as a proper way to conceptualize a mode of access to the world as it is anyway, because we can coordinate variously different groups of such concepts as each reflecting a mode of 277 access to a unified picture of the world as it is anyway. There is no hope, however, of being able explain the diversity of thick concepts that there undoubtedly is which would allow us to retain the thought that they provide a mode of access to some single objective reality. There are two reasons for thinking this. The first is that, unlike differences in secondary-quality concepts, alternative thick concepts (or alternative conceptions of the same thick concept) can be in tension with one another—they can ground opposing demands, be a part of verdicts with differing practical consequences, be the basis for warranting different feelings, and so on—even though they are directed towards one and the same act, practice, individual or group, as it might be represented in more neutral (non-evaluative) terms. 43 The second, and perhaps more important, point is that we lack the means to ascribe error or mistake to these alternatives (that is, a sense of error or mistake that is not based on question-begging assumptions). 44 This means that when we look to explain how it is that we arrived at some set of thick concepts, our explanation will have to leave room for this tension and diversity. The best explanations available, then, as to why some group came to use some set of thick ethical concepts, will be one that relates their social world to others (or to the possibility of others by tracing its own unique historical development). But this explanation will not be vindicatory. Rather, it will be the likely product of paleontology, 43 Note that the fact that we can register practical differences with respect to a neutral description of an act, individual, practice, group or institution, does not mean that this neutral description explains the application of either sides thick concepts, even if it is true that either’s thick concepts might rightfully apply to that act, individual, practice, group or institution. 44 The idea of being unable to convict alternatives of error can be expressed perhaps in the following manner, although we ourselves might be incapable of using their concepts, we also might be incapable of providing them with a non-question begging argument that they ought to be doing something else. Cf. (Lear 1983, pp. 57-58). 278 anthropology, sociology and history, and evolutionary, developmental and social psychology. It will all lack the sort of vindicatory appeal that say an explanation of two competing groups, taken from the history of physical science, will have: as two groups engaged in meeting some common conceptual goal. What, then, makes a group’s reflection on the use of their thick concepts potentially disturbing to their ongoing practice of using them? The answer, in short, is that, by coming to see (through reflection) that their use of these concepts rests on interests that are not universally shared, they can come to see that, in virtue of being just theirs, such interests are simply arbitrary, or worse: the brute imposition of power by some advantaged group over the disadvantaged. 45 As such, the concepts which they underwrite can come to be seen as making classifications that are (given their current reflective state) either poorly grounded or else grounded in interests that were decidedly different from what they took them to be. As a result, they now might see some of their concepts as imposing a social structure that they can indeed live without. Coming to this view could be deeply unsettling, and would present a challenge to their being able to see a point in continuing to use those concepts as a way of conceptualizing (and regulating) social interactions. They would become “shaken realists,” to borrow from the Wallace Stevens’ quote which Williams cites in the opening of Ethics and the Limits of Philosophy. 46 Faced with the realization that there is nothing holding them to the use of these concepts save themselves and their own concerns and interests, along with the 45 Williams has an excellent discussion of this, and how it relates to certain conceptions of liberalism (Williams 2002), Ch. 9, esp. Section 5. 46 Wallace Stevens, “Esthétique du Mal.” 279 shaken awareness that these interests are somewhat arbitrary in nature, the community can simply abandon those concepts, and opt to no longer classify things in their terms. When this occurs, the knowledge provided by these concepts is lost. And yet, even though the group can no longer use these concepts, there is nothing in this process of reflection that need prevent them from seeing their past use of them as genuinely providing a sort of knowledge. This is a consequence of the fact that the very same inability to ascribe error laterally to alternative sets of thick concepts can now be seen as operating historically, as applied to their past usage. To deny that they previously had knowledge under these concepts would require something more from their critical reflection than is actually given: it would require uncovering an error or mistake in their previous use, such that it was on account of this error or mistake that their previous use of the concept was constitutionally defective. But this is decidedly more than coming to the awareness that one’s set of thick concepts is not the uniquely correct set of thick concepts. And so, it would remain open for them to realize (and make light of that realization) that, if they could now make such judgments, some of those judgments (i.e., the correct ones) could count as knowledge. What we are lacking at this point is an explanation of what the post reflective state amounts to (i.e., the state in which this knowledge is now inaccessible). What we need is an explanation of the sense in which, in virtue of their arriving at this reflective position, a group can become incapable of using some of their thick concepts. There is also, however, an additional point that needs to be explained, namely Williams’ rather 280 enigmatic statement that ethical knowledge might not be the best ethical state. 47 It is to these points that I will now turn. Section 5.5: Lost Ethical Knowledge and the Need for Confidence Explaining what it is for a group to lose some body of ethical knowledge in the manner that Williams’ suggests requires explaining how a rather complex set of epistemic circumstances is possible. First, the explanation given cannot undermine the claim that their past use of a thick concept indeed constituted knowledge. If this was the actual result of reflection, then we should conclude that the group never had knowledge under it to begin with, and not that the knowledge available under it was now lost (or inaccessible) to them. Sustaining this first point, however, requires showing how a second and related point is possible: namely that these concepts are still capable of providing this group with a certain form of ethical knowledge. Odd as this may sound, we have to remain clear about what has changed as a result of the group becoming more reflective: it is the group’s view as to what the use of this concept amounts to that has changed, not the structure of the concept itself. In this sense, all that has changed is the group’s ability to access or utilize the ethical knowledge in question. 48 Third, this group has to retain some understanding of how it is that their past practices in using the concept actually operated. Of course, community wide amnesia could explain their inability to us these concepts, but this would involve a concomitant loss of their competence in the concept. So their inability to judge with this concept must be something consistent with 47 (Williams 1985, p. 168). See also (Williams 1995c, pp. 207-08). 48 Any reference to “access” here is, of course, a highly metaphorical way of putting things, and one goal for this section will be to discharge this metaphor (as much as possible) in favor of an explanation couched in terms of changing ethical interests. 281 their retaining (or at least being capable of retaining) some level of semantic competence with it. The loss must be consistent with their still knowing how to deploy the concept, even though they are unable to sincerely do so, given their new reflective state. Fourth and finally, whatever explanation is offered has to be an explanation that is specifically related to the nature of thick ethical concepts. This is due largely to the fact that this phenomenon cannot be replicated with scientific concepts. In their case, knowledge is never lost. What is revealed by reflection is that they never had knowledge under these concepts to begin with. The explanation that I will advance centers on the two components that I have proposed as key to our understanding of thick concepts: (i) that they presuppose specific criteria and ethical standards, and (ii) that the sincere use of these concepts presupposes identification with the concept’s underlying ethical rationale. The explanation itself, then, must center on showing how it is that a group is capable of understanding how a certain concept functions, and how it is able to provide those who actually use it with a distinctive sort of knowledge, but that it is simply not open for them to use this concept. Of course, this is a picture we are already familiar with in not only describing how the ethnographer might be related to certain alien concepts she is studying, but in also describing (at the beginning of this chapter) what is distinctive about our position in relation to a thick concept like CAD (as prevalent in Victorian society) or (in the previous chapter) with conceptions of property and theft (as they were prevalent in the antebellum South). Both we and the ethnographer are capable of understanding precisely what is involved in making judgments using these concepts, 282 and to this extent we are capable of understanding that such judgments are capable of being known. And yet, despite this understanding we are both incapable of using them to voice judgments that we could hold ourselves accountable to. Our group, post-reflection, stands in a similar position to these other thick concepts, as does the hypothetical group we are considering. Though previously capable of issuing such judgments, now they are not, even though they are still in a position to understand what is involved in making them, and to this extent they are capable of understanding that such judgments are capable of being known. If, however, part of the explanation here relies on the distinctive nature of thick ethical concepts, when we turn to consider how conceptual-level reflection in particular can destroy ethical knowledge, another part will depend essentially on changes that the group has in the interests for using such a concept. This is due to the fact that, if we are to meet the complex set of epistemic circumstances that are needed in order to sustain Williams’ idea that conceptual-level reflection can destroy ethical knowledge, what must change for the group is not their concepts, but their relation to them. The possibility of such reflection destroying ethical knowledge comes (in part) from the fact that their unselfconscious understanding of what they are doing and their reflective explanation of what they are doing can come apart. The possibility of this should be unsurprising since most of use come to use a repertoire of thick concepts before we are ever able to occupy some sort of reflection position on the type of reasoning that this involves. Being raised up into some particular social world is rarely (if ever) accompanied with the self- conscious proviso that maintaining a form of social regulation is what we are (in part) 283 doing. As such, there is a good deal of room for there to be a difference in how ethical thought under these concepts seems to us prior to reflection and after, especially after our coming to accept that our inheritance of these concepts is a product of a deeply contingent historical process. 49 Because the picture behind a group’s unselfconscious view of their conceptual practice can come apart from the picture that arises of that practice from a reflective awareness of its contingency, there is room for disappointment. The group’s awareness that the use of such concepts rests squarely in their having certain ethical interests can be enough to unseat the use of them if those interests come to be seen as arbitrary. There is certainly something provocative in saying here that ethical knowledge is destroyed, but the idea that reflection can do this with thick concepts is not just a controversial way of stating some more obvious fact (as some have claimed). We commonly accept that reflection can reveal to us that something which we had previously held to be a piece of knowledge was not so. In such a case, however, there was never really any knowledge to begin with—only false belief—and so, if something was destroyed, it was only our belief that we ever really had knowledge in this area. In the case of thick concepts, what is destroyed here is a capacity for knowing certain things. This is what makes the metaphor of access (as I mentioned above) so compelling, because it suggests a picture that what is lost is not some body of claims that were once true but our ability to sincerely believe them. Consider this in comparison to the example drawn from Huckleberry Finn. There are things that Huck knew, prior to his seriously reflecting on whether to turn Jim in, that 49 Cf. (Williams 1985, p. 135) 284 he was incapable of knowing after that reflection. Of course, the role of reflection here is not precisely the same as what Williams intended—Huck never considered exactly whether the conceptions of property and theft were good ethical conceptions to have— but the results of his reflection are largely similar. Huck’s reflection, on the question of whether to turn Jim in, altered his dispositions in such a way that he was no longer capable of seeing Jim as property, or of treating his helping Jim to escape as something that amounted to theft (these would now simply be things that other people thought and said). As a result, Huck’s post-reflective state is one where he is incapable of seeing Jim’s being Miss Watson’s property as a reason for turning him in. It makes sense to say, then, that Huck suffers a loss of ethical knowledge. (Of course, it goes without saying, that we think this loss is a good thing. And we will return to this point in a moment.) Huck’s case is simply a micro-level representation of what Williams saw as a macro-level, social phenomenon: once a community’s interests and concerns have been sufficiently altered, due to their reflecting the nature of their thick concepts, such concepts can become obsolete. The community no longer sees a point to thinking about acts, traits, and attitudes in this way; and as a result the concept is ultimately driven altogether from active use. At the social level, what reflection ultimately undermines is a group’s confidence in using such a concept as a way of regulating social interactions. In some cases this confidence is undermined by other ethical considerations, in others it is undermined by the sheer awareness that allegiance to such a concept is held up only by interests and concerns that the group now sees as arbitrary. What reflection on the nature of a thick concept inherently creates is a need for finding ways to bolster confidence in 285 thinking that its use is indeed a good way to regulate their social world. Post-reflective survival for a thick concept, then, depends on being able to muster the social and argumentative resources to foster a continued identification with that concept’s ethical rationale. In this sense, confidence (as it relates to the use of thick concepts) turns out to be a deeply important socio-ethical attitude. 50 And yet, it is something especially difficult to sustain in modern communities, precisely because they are more prone to an awareness of the plurality of ethical conceptions and the historical contingency that underlies them, as Williams noted. Such an awareness is what makes these communities inherently more reflective than their predecessors. Having confidence in the use of a thick concept, as constituting a good way for regulating social interactions, in a reflective and pluralistic society must be done in light of the fact that there are different ways in which such a regulation might be conducted. The recognition of a need for confidence, then, is a recognition that is available to a group only after they have attained some sort of post- reflective stance towards their thick concepts. That this is so helps to explain why the need for confidence was not something that the hypertraditionalists could appreciate. For them, the very idea of regulating their social world by conceptualizing things differently had not yet arisen, and so, the confidence they had in using their thick concepts was simply an implicit part of their unselfconscious use of them. This point also helps to explain why some commentators on Williams’ views have failed to adequately understand the role of confidence in the use of thick concepts: confidence in the use of a 50 Cf. (Williams 1985, p. 170): confidence for Williams is “basically a social phenomena.” See also Moore’s comments about how the use of any concept (be it practical or theoretical) involves a “pure element of faith.” (Moore 2006, p. 5) 286 thick concept is itself a good in a post-reflective community that can be lost (even when it was never seen as necessary prior to reflection) and it is something that can be fostered in unhealthy ways. 51 Contrast this view of confidence in Williams’ thought about thick concepts, with the views offered by J.E.J. Altham 52 and Allan Thomas. 53 Both take Williams’ invocation of confidence to be out of place and confused, given his remarks about the possibility of ethical knowledge. Altham argues that if certain thick concepts are able to remain usable in a post-reflective state, the need for confidence becomes superfluous: Confidence is a post-reflective state. We may be confident of what we know, but this notion of confidence would have little point unless, in favorable circumstances, we could also be confident in ethical matters where knowledge is not available. 54 Thomas, perhaps more succinctly, writes that “if we have knowledge, we have no need to bolster it with confidence.” 55 Both Altham and Thomas treat the role of confidence in Williams’ thought too much like an epistemic attitude that is directed towards the content of particular thick judgments, one that can be a replacement for, or offered in support of, the thick ethical knowledge that we (by all accounts) are taken to have. Of course, we can have confidence in judgments that we do not (or cannot yet claim to) know: I might, 51 Cf. (Williams 1985, pp. 163-64) and especially (Williams 1995e, p. 213) where he talks about the need for reasonable confidence and not the confidence of bigotry or prejudice. 52 (Altham 1995) 53 (Thomas 2006, pp. 154-55) 54 (Altham 1995, p. 164). Admittedly, part of Altham’s criticism relates to Williams’ remarks about our having confidence with certain claims that are made with the use of thin concepts. Although I will not discuss this point in detail, I think a good deal of the confusion centers on the fact that here Williams is indeed talking about particular judgments. Since I cannot have knowledge with the use of judgment using a thin concept, like ‘later term abortion is wrong’, it is still open for us to have confidence in this judgment. 55 (Thomas 2006, p. 155) 287 for instance, feel confident that Goldbach’s conjecture is true, even though I know there is no proof for this, or I might feel confident that my car is undamaged even though the parking garage has been flooded because of where I recall my were my car was parked. In these cases, confidence is an attitude that one typically has when they are not in a position to know certain things, and so, in cases where it is admitted that knowledge is possible, confidence seems redundant. Confidence, as I am apt to understand Williams here, is to be seen neither a replacement epistemic attitude to that of knowledge nor as one that is able to support the knowledge claims that we have under thick concepts. It is rather, what I would like to call, a pre-epistemic attitude. That this is so also helps to explain why Williams sees it as “basically a social phenomena.” 56 It is something that is exhibited by a community as a whole, and not by particular judgments made by particular agents within that community. Such confidence is both a pre-epistemic, and inherently social, attitude due to the fact that it relates to the general use of certain thick concepts, and not to any particular judgment that might be voiced with them. It is a basic attitude that is related to a community’s view that their use of some thick concepts is a good means for regulating their social interactions and is thus an appropriate way to conceptualize various acts and traits. The role of this sort of confidence can remain implicit in the community’s use of their concepts when they are unaware of there being any possible alternative, as the need for it goes unquestioned in that unselfconscious use. It is the awareness of possible alternatives that calls into question whether these concepts indeed constitute a worthwhile form of social regulation. The need to foster social confidence in a post-reflective state, then, 56 (Williams 1985, p. 170) 288 becomes paramount, because the possession of an adequate level of confidence in the use of their thick concepts is a central pre-condition that must be satisfied if they are to have knowledge under them. 57 What both Altham and Thomas fail to see is that this sort of social confidence is related to a conceptual practice as a whole, and not any particular judgment made within it. Because of this, it is simply a mistake to see it as something that could replace knowledge under a thick concept or that would be made redundant by the admission that we have such knowledge. Given this, what is the argumentative upshot? What does the possibility that reflection can destroy ethical knowledge tell us about ethical thought in general? Part of it, I think, is this: the idea we have ethical knowledge (and, to this extent, a sense of their being ethical facts, or something to be known) is of secondary importance once we have come to appreciate the historical contingencies that lie behind that knowledge (and those facts), contingencies that are bound up with the very practice of using a certain thick concept. What turns out to be of primary importance, in light of this awareness, is how we might set about supporting our commitment to that concept, given that there are no facts that could tell us whether we should be using it. The question that is left to be considered, then, is how we might understand a relationship between critical reflection, social confidence, and an awareness of this historical contingency that is not necessarily corrosive, but that is in some sense progressive (or at least anti-conservative). Considering the links between these notions will help us to see how conceptual-level reflection can itself be in the service of a desire for transparency, and how such a desire 57 Williams writes, in response to Altham’s criticism, that “[t]hick concepts under which we can have some pieces of ethical knowledge are not themselves sustained by knowledge, but by confidence.” (Williams 1995c, p. 208) 289 can be understood as an aspiration to lay clear the conditions for identifying with the ethical rationales that underlie our thick concepts. By revealing the historical contingency behind our inheritance of some thick concept, reflection can be seen as prompting a greater understanding of them as things whose use we can still assent to. But, equally, it might reveal them to be the products of bias or the result of unjust application of power. In this way, conceptual-level reflection can enter in as part of a process for unmasking prejudice or the role of power in the community as things that are especially connected with the use of certain thick concepts. That this is so will also help us to see the very idea of fostering social confidence as itself something of a social ideal. The promotion of clear-headed, non-prejudicial, social confidence, then, can itself be understood as type of regulatory goal that the practices of ethical reflection in a community can aspire to promote. Taking up these points will be the goal of the next section. Section 5.6: Transparency and (Post-)Modern Ethical Authority Williams’ argument that reflection can unseat thick ethical knowledge was intended (in part) to show that there was something unique about the modern (or post- modern) world and its relationship to ethical thought. Much of this is shown in precisely how different things are for us in contrast with the hyper-traditionalists. What is distinctive about the modern world in contrast to the hypothetical one that Williams presented are the dual facts that we are aware of both a variety of social alternatives that exist contemporaneously to us and the variety of social worlds that are our historical predecessors. This awareness is, if not driven by, is largely helped along by, the findings 290 of the social sciences (e.g., psychology, sociology, and anthropology), the historical- sciences (e.g., archaeology, paleontology, and paleoanthropology), and history, since history itself is very often, as Williams has put it, a “history of alternatives.” 58 Each of these investigatory practices, in its own way, puts forward pictures of human social life that can diverge from our own in deep and incommensurable ways. More importantly, however, the pictures provided all suggest that, if anything, the nature of such social life is itself deeply contingent: wherever we find human social life, we can see that its character at any given point is the product of choices and changes that could have just as easily been different from what they in fact were. We might represent this conglomeration of points with (CSD): (CSD) Any social world is a temporally extended notion of a certain form of behavioral regulation whose history is decidedly non- vindicatory. As such, the particular structures and methods of conceptualization manifest at one point in its history, including its thick ethical concepts and underlying ethical standards, are themselves radically contingent artifacts of that way of life. Call this the contingency of social development thesis. As we have seen, in the previous two sections, coming to accept (CSD) can have a massively destabilizing effect on a community’s ethical practices. At some points in history, we might think that an appreciation of (CSD), or something like it, was simply not possible. What is distinctive about the modern world, however, is that such a picture of human life must find a way of being integrated into an image of human social life as one of an ongoing practical concern for its participants. The social environments within the modern world have no choice but to reconcile the self-image they have of their conceptual practices and 58 (Williams 2006a, p. 196) 291 underlying standards with the truth of (CSD), since, as a matter of empirical assumption, these social worlds are, at some level, already aware of the point expressed by (CSD) and that awareness, in the future, is only likely to grow. This, however, is liable to seem overly controversial, and beg certain questions: Why should we be concerned with (CSD) at all? And why should we think that integrating it into our understanding of our own ethical practices is something that we should do (or perhaps, more strongly, must do)? These are, of course, valid questions. And although I doubt I will be able to do justice to them here, there are several responses that are worth offering. I will begin with two smaller responses, and then return to discuss a third a little later on in this section. The first point to consider is simply that (CSD) is true, or at least, it constitutes the best reflective/explanatory picture we currently have regarding the development and diversity of human social life. (CSD) is of course a general statement, but it draws from a scientifically informed understanding of the development of human society, as it has occurred in vastly diverse times, places and ways. And so, in response to the claim that we might get away with not being concerned with (CSD) at all, one might respond simply by stating that (given the reflectiveness of the modern world) we already have an inherent interest in trying to make our own understanding of what we are up to (in using our thick conceptual practices and standards to regulate our social world) consistent with the other truths (or explanations) that we hold. As such, we have a reason for having our own self- image of this social world be one that is not at odds with (CSD). 59 We might add to this, a second, and more important, practical point. The development and maintenance of our 59 Jonathan Lear takes such bifurcated, un-integrated perspectives on our ethical practices (one that is ‘inside’ a reflective stance, and one that is ‘outside’ it) to be inherently unstable, and perhaps even incoherent, as it would require individuals to lead a “fragmented life.” (Lear 1983, p. 58) 292 various thick conceptual practices has proceeded by and large without the sort of self- conscious image of what those practices amount to, as provided by (CSD). This is because (CSD) is largely the product of scientific and historical discoveries that have taken place (relatively speaking) quite recently. There is the risk, then, that aspects of our social world might rest upon assumptions that are deeply at odds with (CSD). There is the risk that our thick conceptual practices involve, as Miranda Fricker aptly puts it, “bad kinds of conservatism.” 60 And this provides us with a pragmatic reason for trying to integrate our understanding of what we are doing with (CSD): we have a practical interest in having some sort of reflective stability (even if it amounts to something less than objective vindication) in our social practices. Recall that it is precisely because (CSD) can diverge radically from our unselfconscious use of thick concepts that attempts at integration at later stages in a community’s development can go so poorly. As a community gains a clearer understanding of how human cultures and societies develop and sustain themselves, there is room to lose a certain “pre-reflective innocence” 61 in thinking (or better: assuming) that their own way of conceptualizing and organizing their social environment was the only way of doing so. With this loss of innocence, a community can become cut off from viewing themselves in certain ways. Such a break can be so severe to what they are doing, and to the concepts they are using, that the thought of continuing on as before is just no longer an option, or, if it is an option, it is no longer an option for them. Indeed, this very possibility is what allows for reflection to destroy ethical knowledge. 60 (Fricker 2001, p. 108) 61 For an excellent discussion of this notion, see (Kekes 1993, pp. 180-84). 293 Williams’ view of the corrosive effects that reflection can have is indeed disquieting, but contained within this picture is also the thought that a loss of pre- reflective innocence can be empowering. Indeed, it can be transformed into a very modest foundation for ethical authority. Making this transition, however, will require that we give conceptual-level reflection a more positive role in the maintenance of our social world. Indeed, conceptual-level reflection can be seen as playing a central role in the development of social confidence. Coming to accept this model of ethical authority will also reveal how difficult and complex a matter it is for a reflective social world (such as ours) to sustain confidence in its use of certain thick concepts. Yet, it is only by understanding (and appreciating) the role that conceptual-level reflection can have here that we can come to grips with the place that thick concepts will have in a modern social world. When we first considered conceptual-level reflection our interest was in the prospect of reaching a certain reflective goal in the use of thick concepts, namely that of objectively vindicating some set as the right or best ones to have. Vindication, however, need not be the only epistemic goal that conceptual-level reflection is directed towards. It can also be directed towards the more modest one of not wanting to have one’s acceptance of (or identification with) certain thick concepts, and their underlying ethical rationales, rest upon misunderstanding or manipulation. Here we might consider the entire range of interests, concerns, values, and standards, on the one hand, as well as various social goods, meanings and institutions, on the other, which are bound up with the acceptance and use of a certain thick concept as the conditions for acceptance of that 294 concept. Perhaps, as a gloss of the role that conceptual-level reflection has, we might begin by saying that it is not directed towards providing the users of the concept with a reason for accepting some one of their thick concepts. They already have reasons for accepting them. After all, they are their thick concepts: they are concepts that they have been raised up to adopt in becoming a member of that social world; they are concepts that such a group already has certain interests in using, and, as of yet, remain identified with those interests. Rather, what we should say here is that conceptual-level reflection is directed towards discovering whether they have good reason to abandon those concepts. It is directed towards determining whether they should consider their concepts and related standards to be reasonable by considering how they came to have those concepts and standards, under the proviso that such an explanation will be one consistent with (CSD). Such reflection, then, can be directed towards determining whether our acceptance of these conceptual practices has been conditioned by bias, prejudice, or the influence of previously unseen ideological forces. This brings us to what is a third reason for taking seriously the need to integrate our self-image of our conceptual practices with (CSD). It is based on an ethical argument from drawn from our concern that any ethical agent should be (in some minimal sense) autonomous. We value this sort of autonomy in at least the following way: that we think an agent should not have her beliefs influenced by factors that she could not (in turn) reflectively endorse, and when this is the case, it is typically because her endorsement is the product of someone other agent’s power. When her beliefs are the product of elements that she cannot reflectively endorse then there is a real sense that her holding 295 them at all is simply a product of either a misunderstanding of the nature of her situation or worse: that her acceptance of the concept is a product of having been misled by others. Because of this interest in being able to reflectively endorse one’s own beliefs, as a condition of autonomously holding them, this agent has a vested interest in making sure that her beliefs can be endorsed when she has a clear picture of how it is she came to possess the concepts which give those beliefs their distinctive content. At the very least, she should be able to accept (or remain subscribed to) those conceptual practices while having a clear-headed understanding of their origins and development. But such a clear- headed understanding involves accepting (CSD). Thus, she has an interest in integrating her picture of the thick conceptual practices see currently accepts with the picture of their development provided by (CSD). It might be difficult to see how a community’s thick conceptual practices can depend upon considerations and attitudes that a self-conscious and accurate explanation of would make unusable, or would at least put into question. One reason for this is that we tend to forget (or fail to appreciate) how such conceptual practices can be bound up with various attempts at legitimation or with certain legitimating pictures. Such legitimations can form a part of the community’s self-image of their thick conceptual practices. They often form part of a story that helps orient members of the community to what they are doing conceptually, and in the process helps motivate (or worse, coerce through fear of punishment) acceptance of these concepts and standards. To be clear, though, having a legitimating picture is not, in and of itself, grounds for objective to a thick conceptual practice. It is at least not such a ground when we are considering the use 296 of conceptual-level reflection as a means of getting clear about the origins and development of such a practice. There is no reason why some legitimating pictures might be completely innocuous in this regard, such as an attempt to get members of a community to subscribe to certain thick concepts by suggesting that attunement (or balance, or sustainability) is the proper way to think about their social world’s relationship with the natural world. 62 We need not think that there is anything at odds between this thought and the sort of self-conscious awareness that members of the practice would have of what they were doing in coming to accept something like (CSD). There is also nothing wrong with the fact that legitimating pictures attempt to persuade one to accept or identify certain thick concepts that they might not have done so without. At bottom, this is simply the function of legitimation as such—to persuade by offering what are to be taken as good reasons for doing so. Since we should be willing, as I just noted, to allow some legitimations to be consistent with accepting (CSD), the fact that they constitute a method of persuasion should not (in itself) be seen as an objectionable feature of them. The concern we have is not with a legitimating picture’s power to help prompt or motivate acceptance. Rather, it is over whether the legitimating picture’s power to do so is deeply at odds with the acceptance of (CSD), and because of this cannot be sustained once members of the community are brought to fully reflect on and accept (CSD). 62 Here I am thinking of Charles Taylor’s excellent discussion of how certain practices, such as rain dances in certain cultures, are better interpreted, not as involving crude attempts at a scientific understanding of the world, such as providing the community with a means to actually alter the weather, but as forms of coping with life events, such as droughts. (Taylor 1981, pp. 207-09) Taylor’s point, I take it, is entirely consistent with Williams’ view that whether we interpret a certain conceptual practice in objectivist or nonobjectivist terms will depend largely on whether we see their practices as (perhaps only implicitly) making claims as to how other people should do things. 297 One reason we might be concerned about the presence and power of such legitimating pictures is that we (as external observers on some thick conceptual practice) might not be moved to accept the relevant thick concepts on the grounds they provide. The presence of such legitimating pictures highlights the idea that they might have reasons to remain committed to their thick conceptual practices, that, despite being consistent with (CSD), we do not accept. And this in itself might seem to be objectionable. There are several things to note here. First, we should recall that the goal of conceptual-level reflection here is not to vindicate their set of thick concepts over some other set. And, since the goal here is not vindication, there is one challenge to their practices that is simply not probative: namely that of questioning the reasons that are offered for remaining committed to those conceptual practices simply on the grounds that they are their reasons (i.e., considerations that they cite to legitimate their continued acceptance of some thick concept). 63 Put another way, since objective vindication is not the goal of critical reflection here, it is not an objection to the output of such reflections if they turn out to not be acceptable or reasonable to everyone. To think that this is a defect in their argumentative position is to assume that there is something stronger that could be obtained. A denial of this assumption, however, is latent in accepting (CSD) and explicit in the fact that there is no way of ascribing some (non-vacuous, non-question begging) notion of error to people who are conceptualizing things differently. (Of course, our inability to ascribe error to alternative ways of life means that at a certain level we will be unable to say that such groups are (say) conducting themselves irrationally or contrary to human nature, does not mean that we have to be tolerant of such perspectives. They 63 Cf. (Scanlon 1995, p. 353) 298 might after all be, from our lights, cruel and unfair.) The very fact that the reasons and arguments given speak to interests and concerns that might not be shared outside the community is not a reason for their finding fault with those reasons and arguments. The offering of such argumentative resources is directed towards the lesser goal than that of trying to show those who are already committed to such practices that they have some reason to remain committed: namely that, given that they already are so committed, there is no good reason for them not to be. In other words, it is to get them to remain identified with a set of thick concepts that they already have a history of identification with, but to do so now simply because they are their thick concepts, because they are a firm and reasonable part, or at least not an unreasonable and capricious part, of their ethical way of life. Given the concerns and interests they already have, and given that they have no reason to drop these concepts (assuming they do not), critical reflection can show then that there is still a point in their continuing to use certain of their thick concepts. 64 This, however, leads us to another important point, implicit in this discussion: From what point of view should we consider the conceptual-level reflection being done? As I have approached this point, namely as one of a community attempting to come to grips with how they arrived at their own thick conceptual practices, it is clear that this sort of critical reflection is not to be applied from some standpoint wholly external to the community in question. What is certainly not of interest here is how a theorist, as someone who lacks any appreciation of the relevant concepts, interests, and underlying 64 It is important at this point not to confuse the idea that there are limits to the sorts of errors we can ascribe to groups using different thick concepts, from our being limited in applying our thick concepts to them. We might be right in thinking that a different group’s practices are cruel and barbaric, even if they lack such concepts. 299 standards, might reason as to what should or should not be accepted, rejected or reformed. Rather, what we are considering here is how the possibility of engaging in conceptual-level reflection and accepting (CSD) might be put into service by the participants for themselves as a way of guiding the continued development of their social world. In doing so, we are considering how their cognitive position might be improved in ways that they would accept if they were able to seriously engage in such reflection. Assuming that they can, what is of interest is how they would react to such a reflective awareness. And this will be expressed by the social possibilities they see as available to themselves after accepting (CSD). After all, these reflections and decisions are about how their social life could continue. They are not exercises in academic (or merely academic) reflection. As such, it is only from within a committed participant’s perspective (or the perspective of one who clearly appreciates the demands of living within that perspective) that there is a genuine sense of being able to reconcile their practices with a reflective explanation of them. Once (CSD) is accepted throughout the community, however, there is also the possibility that a multitude of variously overlapping social worlds might be revealed when the larger community had previously seen just one. Indeed, such overlapping worlds might be revealed by a community’s uneven acceptance of (CSD). Either way, we should not assume that the effects of conceptual-level reflection are limited only to changing the stock of thick concepts available to the community—it might also effectively alter the identity of the community itself, in virtue of changing what sorts of thick conceptual practices are accepted by different aspects of the larger community. 300 Such disintegration and fracture can arise from the realization that the only thing holding the larger community together is now something that neither sub-group can accept. There might simply not be enough common interest left after their reflection to hold the larger social world together. We need not think of this as the only social reaction available to them. There is also the possibility of forming a political confederacy or agreeing to certain non-aggression pacts. Both of these options might be seen as beneficial to each part, even if nothing stronger is, given the sheer fact that their previous coexistence reveals them to be possibilities each could live with. The point, however, is that how these groups are able to carry on together after such fracture is more a political question than an ethical one. Fracture need not be the only, or even the most likely, result from a society coming to accept and integrate (CSD) with their thick conceptual practices. At another end, there is the possibility of reform, and thus progress. Reflection at the conceptual- level can yield substantive criticisms directed towards some of the community’s thick conceptual practices by uncovering how their coming to have these concepts is rooted in certain factual errors, errors that were perhaps accepted as correct due to fear, ignorance, resistance to change, and so on. This would undercut their ability to sincerely use them, since such concepts would now likely be seen as having developed out of an overreaction to some situation (a reaction that, given what they know now, was not even supported by the facts when it happened then). It is, of course, an open question as to what reactions count as overreactions in this sense, but there is certainly critical space for a community to look back at their conceptual practices and ask whether such a reaction was the only 301 reaction available to them, given the evidence at that time. Michelle Moody-Adams argues that communities can engage in forms of “affected ignorance” which amounts to “choosing not to know what [they] can and should know.” 65 The fact that individuals and communities can be caught up in this sort of ignorance is reflected in such things as their use of euphemisms to mask clear debate (e.g., “enhanced interrogation technique” as opposed to “torture”), their deployment of practical strategies to allow for ‘plausible deniability’, their being resistant to debate or criticism of any form, and so on. The fact that conceptual practices can arise in such fragile and fallible environments is a reason that conceptual-level reflection can be deployed as a powerful tool for reform: it can aid a community by getting them to see how they came to have a certain thick concept was itself affected by processes that were evidentially and ethically suspect. As a result, they might come to see those concepts as being suspect as well. A much more interesting possibility here, however, is that the acceptance of (CSD) might actually encourage the realization that their ethical outlook is not the cohesive whole they had assumed it to be. In getting them to see appreciate precisely how radically contingent its development was, it might be difficult for them to continue on thinking that each element within that outlook operates without conflict with all the other elements. A denial of this is reflected in Aristotle’s unity of the virtues thesis (i.e., that one cannot have any virtue without having all of them, and hence that the possession of any virtue is wholly compatible with the possession of any other), 66 but here it would 65 (Moody-Adams 1994, p. 301). Moody-Adams lists several types of affected ignorance. 66 (Aristotle 1998), 1144b30-35. For an interesting, and more contemporary, discussion of coherence of the unity of the virtues, see (Wolf 2007). 302 be applied to the entire thick conceptual repertoire of a community, and not to the moral psychology of a particular agent. Coming to such a realization of potential internal tension would likely involve abandoning a certain idealization about the nature of ethical thought itself: namely that it constitutes (or can constitute) a seamless structure that never needs to be reformed or re-calibrated to the nature of social living it is a part of and which can be purged of all incommensurable elements. Williams highlights this thought, in noting that: Once we regard the ethical life we now have as a genuinely historical and logical structure, one that is peculiarly self-conscious about its own origins and potentialities, we shall have less temptation to assume that it is a satisfactorily functioning whole, and we will be more likely to recognize that some of the widely accepted parts of it may stand condemned in the light of perfectly plausible extrapolations of other parts. 67 Reflecting on how a community came to have the thick concepts that it does, then, can not only uncover certain distortions (misunderstandings and manipulations) that would have prevented the community from actually taking up some other thick conceptual practice, it can also remove the pretense that the entire set of thick conceptual practices is (or even could be) a well-integrated structure. And, in removing this pretense, they might also come to see that structure as something that requires (through critical reflection) continual adjustment. Whatever role that thick concepts might occupy in a modern social world, then, that place will be one of continuous critical engagement with those concepts. And, because of this, the community must come to live with the ongoing possibility of having to disown certain aspects of that world because they constitute unreasonable ethical demands. 67 (Williams 2005b, pp. 36-37) 303 So, in the end, how should be understand the relationship between conceptual- level reflection and social confidence? Social confidence, as I have described it, is something achieved through continuous and reflexive critical engagement between a community’s understanding of its own contingent development and its thick conceptual practices. It is, however, also a normative state, as the very notion of social confidence can be seen as falling somewhere between two extremes: those of hyper-traditionalism and hyper-reflection. A community is in this state when its critical practices are somewhere between accepting everything between simply because it is a part of their social world and questioning everything for the same reason. We might say, then, that a community is in this state if it recognizes the following two conclusions about its critical relationship to its thick conceptual practices: (1) That it is not an objection to the ongoing use of a thick concept to cite that the concept is merely theirs (i.e., to cite the diversity of thick concepts and social worlds); but that (2) In being theirs at all, the continued acceptance of such a concept is not above critical reproach (i.e., self-criticism). Given this, there are two ways in which a community can fail to occupy this state. First, there is an unreasonable gain of confidence that runs concomitantly with their ceasing to remain reflectively engaged with their thick conceptual practices. Here a community might be seen as once again accepting a legitimating picture of their practices that is at odds with accepting (CSD). This would be a move towards hyper-traditionalism. And yet, such confidence can be undermined when the community begins to unreasonably question the reasons and arguments offered in support of their thick conceptual practices simply because they think they must have greater uptake with others who not a part of 304 their community. Thus, they might begin to question the legitimations they offer because they are merely local legitimations. This would be something similar to thinking that such an ethical perspective was biased or prejudicial simply because it could not attain universal assent. And this would be a shift towards hyper-reflectivism. The fact that a community can fail to occupy a state of social confidence in these very different ways is enough to sustain the thought that social confidence is itself a normative state: it is something like a reflective mean (to borrow from Aristotle) “between two vices, one involving excess, the other deficiency.” 68 Social confidence, then, is that state of critical reflection that allows a community to remain committed to as many of its thick conceptual practices as possible, in full light of the awareness that these practices rest on a radically contingent developmental history, while the rest pass into its history. Section 5.7: Concluding Remarks Williams’ general metaethical aim, in advancing the notion that there are thick ethical concepts, the idea that that such concepts can rightfully be seen as providing a type of ethical knowledge, and the further thought that such knowledge, as it is rooted firmly in certain socially developed interests and concerns, is, because of this, fragile in important and unique ways, can be seen as a part of a larger project that was attempting to meet head on the general question of “what needs to be, and what can be, restructured in light of a reflective and nonmythical understanding of our ethical practices.” 69 When we take a reflective and impartial eye to the ethical concepts that help us to navigate our social world, in an attempt to understand how it is that we came to have them, there exists 68 (Aristotle 1998, 1109a20-22) 69 (Williams 1985, p. 194) 305 the possibility of our coming to see those concepts, and the knowledge that they make possible, as simply arbitrary. The interests and concerns which provide a point to our thinking about social life in these ways can come unhinged, and appear to us as simply the product of a generally uncontrolled developmental drift: the result of a historical process that cannot lay claim to some clear and definitive navigational heading. To be sure, this was a possibility, and not an eventuality, that Williams was intent to highlight, and for good reason. Much of everyday ethical thought and practice has proceeded without a wider understanding of the historical development and diversity of humanity, and in some cases this has led to the development of certain ‘mythical understandings’ as what our ethical practices amount to, especially in the direction of providing a foundation for those practices, which cannot be squared with that wider understanding. Williams’ metaethical views, however, were not directed primarily towards uncovering or debunking any particular mythical view that might have found its way into the unselfconscious practice of regulating a social world through the use of thick concepts. It was rather directed towards how such mythical views have come to be (in one way or another) reproduced in moral philosophy itself. One way in which this has happened can be found in the development of something which has not been touched upon at all in this dissertation, namely normative ethical theory. The impulse towards developing an ethical theory is one which aims at the implicit goal of systematizing ethical thought and practice into a reflectively endorsable system which could parallel the consistency and completeness which (at the very least) seems possible in natural science. At a deeper level, however, such mythical understandings have found their way into the 306 very manner in which we describe the nature of ethical thought, and it was towards unseating this that Williams discussion of thick concepts can be seen as primarily directed. Here we find Williams concern over the ideas of objectivism and realism in ethics: metaphysical views which attempt to show ethical thought and practice can be seen as having credible parallels to the thought and practice of science. Ethical thought and practice cannot be coherently represented as directed towards the achievement of some absolute ethical conception: the development of some set of ethical concepts and ideas which would capture a form of social regulation that was the best one for anyone to have. Indeed, when it is represented in this way it leads to (or invites) the distorted view that ethical thought, and the use of thick ethical concepts, rests on something else (or something in addition to) human dispositions. We might see this sort of distortion as present even in the earnest attempts of thinkers like David Wiggins and John McDowell, who, despite serious and thoughtful attempts to show that ethical thought depended in crucial ways on human dispositions, attempted to go outside them in the thought that thick concepts refer to non-natural properties, properties which, although their existence depended on certain forms of human life, could nonetheless be rendered as something that the developed moral agent was in ‘contact’ with. This, in Williams’ view, is something of a consoling myth; it is something that naturalism in ethics should avoid, as it is very difficult to reconcile this picture of ethical thought with the wider view of human and social development. What this theoretical account overlooks (or under-appreciates) is the fact that the efficacy of the very concepts we use to regulate our social world rest upon our identification with certain 307 concerns and interests, and that this identification can be undone by the slenderest of means. This would not be as disturbing a conclusion as it is, if it were not for the fact that, in recognizing this as a possibility, we are limited in how we can rightfully respond to such a change: we are disturbed by the inability to see it as necessarily the product of mistake or misunderstanding, an option that any form of ethical realism or objectivism should be capable of substantiating. One product of this consoling myth is that it tacitly covers over our inability to do this, as well as the fact that this inability requires explanation. Williams’ attempt at uncovering ethical realism or objectivism as a consoling myth, however, should not be confused with other theoretical reactions that it has motivated (such as contemporary expressivism). These theoretical moves have straightforwardly denied ethical realism or objectivism, but have done so thinking that ethical thought nonetheless has such aspirations, if only implicitly. As such, these moves have tended towards the development of positions that, while aiming to provide everything that an ethical realist or objectivist might want, succeed only in neutering ethical thought of any cognitive significance by denying that there is a straightforward sense in which ethical knowledge is possible. Such knowledge consists in certain (defeasible) practical reasons following from the application of certain thick concepts, concepts whose very classifications depend on interests and concerns that are directed towards providing a certain form of social regulation. The underlying problem with these positions is that they assent to the legitimacy of certain objectivist aspirations in ethical thought, while denying that such aspirations can ever fully be met. 308 In this way, we can see how ethical objectivists and expressivists share an underlying assumption about ethical thought that Williams has sought to deny: ethical thought, insofar as it involves thick ethical concepts has no (or at least need not be seen as having) objectivist or realist aspirations. Once freed from the ‘mythical understanding’ that ethical thought is (or should be) at all like this, we are free, in Williams’ view, to develop an understanding of it that is at home with our understanding of human beings as biological creatures who are capable of non-genetic learning, who develop and propagate certain cultures and forms of life, and who have a rich historical and scientific understanding of themselves. Ethical thought and practice, on this view, is simply a natural part of human social life, understood as it is in these terms. Coming to this view, however, is not without its costs. The central cost of such a ‘reflective and nonmythical understanding’ of ethical thought is the clear awareness of the fragile nature of thick concepts and their ability to provide ethical knowledge. While our interests in regulating social life run deep enough to affect the very manner in which we conceptualize our social interactions, their being a point to such modes of conceptualization is itself something that is held hostage to our retaining those very interests. And reflection is something that is always potentially at odds with our doing so, as it allows us to ask (simply and frankly) whether we should be doing something else. This leads to another (perhaps equally) consoling myth, that Williams thought aims to unseat, namely that we are, in some sense, free to walk away from these concepts at a whim. Here the myth is that our interests, simply in being our interests, means that they are things we can simply abandon at will. It is this assumption that perhaps drives the 309 starkest resistance to Williams’ view that ethical knowledge rests on a deep, radical contingency. In having a radically contingent foundation, our thick concepts and the knowledge that they provide, are no less ours, and things we are no less identified with in using (to the extent that we are), even if those concepts and the knowledge they provide amount to something less than we had unselfconsciously took them to amount to. What all of this might seem to recommend is something similar to Richard Rorty’s view that irony is what we are left with; 70 that the inevitable conclusion we will reach, in reflecting on how we came to have our ethical concepts and standards, is one that will be (in some sense) a letdown. As participants in some ethical scheme or other, we cannot make our participatory stance in that scheme wholly harmonious with the reflective understanding we have of it, as our personal engagement in that scheme calls for something more that this reflective understanding could be thought to provide. Of course, this sort of reaction is central to Williams’ view that reflection can destroy ethical knowledge, but, importantly, it is not a reaction that Williams thinks we need be ineluctably drawn to. Indeed, Williams holds out hope for a more tempered response to nonobjectivism in ethics: one that does not see it as a loss, as Rorty must, but as one which simply accepts it as a piece of our own self-understanding. On this view, the radical contingency that underlies ethical thought is not met with disappointment, as though it should have involved something more, but, as a point of actual fact, does not. Rather, it should be accepted as simply a basic fact about the nature and evolution of human beings as social creatures. It is perhaps best to end our discussion by way of a quote where Williams tries to capture this very sentiment: 70 See here (Rorty 1989), especially chapter 4. 310 In fact, as it seems to me, once one goes far enough in recognizing contingency, the problem to which irony is supposed to provide an answer does not arise at all. What we have here is [something like] the phenomenon of counterfactual scientism. The supposed problem comes from the idea that a vindicatory history of our outlook is what we would really like to have, and the discovery that liberalism, in particular (but the same is true of any outlook), has the kind of contingent history that it does have is a disappointment, which leaves us with at best a second best. But, once again, why should we think that? Precisely because we are not unencumbered intelligences selecting in principle among all possible outlooks, we can accept that this outlook is ours just because of the history that has made it ours; or, more precisely, has both made us, and made the outlook as something that is ours. We are no less contingently formed than the outlook is, and the formation is significantly the same. We and our outlook are not simply in the same place at the same time. If we really understand this, deeply understand it, we can be free of what is another scientistic illusion, that it is our job as rational agents to search for, or at least move as best we can towards, a system of political and ethical ideas which would be the best from an absolute point of view, a point of view that was free of contingent historical perspective. (Williams 2006a, pp. 193-94) To be sure, Williams is not encouraging complacency or the idea that we should simply accept our ethical outlook as it is. The point is not that we should ease our critical approach to our own ethical practices because they are ours, no more than we should blindly reject some other set of ethical practices because they are not. What Williams is encouraging here is a change in view as to what we think critical reflection should seek to achieve. It is enough if it aids us in re-calibrating our own perspective, when the need arises. And so, it need not be something that speaks to everyone in how they should conceptualize their social engagements. It is enough if it says something to us. 311 Bibliography: Albritton, R 1959, 'On Wittgenstein's Use of the Term 'Criterion'', Journal of Philosophy, vol 56, no. 22, pp. 845-57. Altham, J 1995, 'Reflection and Confidence', in J Altham, R Harrison (eds.), Mind, World and Ethics, Cambridge University Press, Cambridge. Anscombe, G 1957, Intention, Blackwell Publishing, Oxford. Anscombe, G 1958, 'Modern Moral Philosophy', Philosophy, vol 33, pp. 1-19. Anscombe, G 1972, 'Contraception and Chastity', The Human World, vol 7, pp. 9- 30. Aristotle 1998, Nicomachean Ethics, Oxford University Press, Oxford. Ayer, A 1946, Language, Truth and Logic, Dover Publications, New York. Baier, A 1986, 'Trust and Antitrust', Ethics, vol 96, no. 2, pp. 231-260. Bennett, J 1974, 'The Conscious of Huckleberry Finn', Philosophy, vol 49, no. 188, pp. 123-34. Blackburn, S 1981, 'Reply: Rule-Following and Moral Realism', in SHAC Leich (ed.), Wittgenstein: To Follow a Rule, Routledge, London. Blackburn, S 1984, Spreading the Word, Clarendon Press, Oxford. Blackburn, S 1986, 'Making Ends Meet', Philosophical Books, vol 27, no. 4. Blackburn, S 1991, 'Just Causes', Philosophical Studies, vol 61, pp. 2-17. Blackburn, S 1992, 'Morality Through Thick and Thin', Proceedings of the Aristotelian Society, vol 66, pp. 285-99. Blackburn, S 1998, Ruling Passions, Clarendon Press, Oxford. Blackburn, S 2004, 'Disentangling, Disentangling', Unpublished Manuscript. Blackburn, S 2009, 'The Absolute Conception: Putnam vs. Williams', in D Calcutt (ed.), Reading Bernard Williams, Routledge Press, New York. Bower, B 1988, 'Virtue Concepts and Ethical Realism', Journal of Philosophy, vol 85, no. 12, pp. 675-93. 312 Broad, C 1930, Five Types of Ethical Theory, Routledge and Kegan Paul, Oxford. Burton, S 1992, 'Thick Concepts Revised', Analysis, vol 52, no. 2, pp. 28-32. Carson, TL 2006, 'The Definition of Lying', Nous, vol 40, no. 2, pp. 284-306. Cavell, S 1969, 'On the Availability of Later-Wittgenstein', in Must We Mean What We Say, Scribner's Press, New York. Cavell, S 1979, The Claim of Reason, Oxford University Press, New York. Chagnon, N 1968, Yanomamo: The Fierce People, Holt, Rinehart, and Winston, New York. Confucius 1979, The Analects, Penguin Books, New York. Copp, D 2001, Morality, Normativity and Society, Oxford University Press, Oxford. Craig, E 1990, Knowledge and the State of Nature, Clarendon Press, Oxford. Crary, A 2002, 'Why Can't Moral Thought Be Everything It Seems?', The Philosophical Forum, vol 23, no. 4, pp. 373-91. Damerow, P, Freudenthal, G, Mclaughlin, P & Renn, J 2004, Exploring the Limits of Preclassical Mechanics, 2nd ed., Springer-Verlag, New York. Dancy, J 1993, Moral Reasons, Blackwell Publishing, Oxford. Dancy, J 1995, 'In Defense of Thick Concepts', Midwest Studies in Philosophy, vol 20, pp. 263-79. Dancy, J 2005, Moral Particularism, viewed 3 December 2007, http://plato.stanford.edu/entries/moral-particularism. Dancy, J 2006, Ethics Without Principles, Oxford University Press, Oxford. D'Arms, J 2005, 'Two Arguments for Sentimentalism', Philosophical Issues, vol 15, pp. 1-21. D'Arms, J & Jacobson, D 2000, 'Value and Sentiment', Ethics, vol 110, no. 1. 313 Darwall, S, Gibbard, A & Railton, P 1992, 'Toward Fin de Siecle Ethics: Some Trends', Philosophical Review, vol 101, no. 1. Davidson, D 1984, Inquiries into Truth and Interpretation, Clarendon Press, Oxford. Davidson, D 2005, Truth, Language, and History, Clarendon Press, Oxford. Dickens, C 1911, 'The Cricket on the Hearth', in The Collected Worrks of Charles Dickens, P.F. Collier & Sons, New York. Douglas, F 1855, My Bondage and Freedom, Miller, Orton and Mulligan, Auburn, NY. Dummett, M 1978, 'Truth', in Truth and Other Enigmas, Harvard University Press, Cambridge, Mass. Duncker, K 1939, 'Ethical Relativity? (An Inquiry into the Psychology of Ethics)', Mind, vol 48, no. 189, pp. 39-57. Dworkin, R 1996, 'Truth and Objectivity: You'd Better Believe It', Philosophy and Public Affairs, vol 25, pp. 87-139. Egan, P 1832, Grose's Classical Dictionary of the Vulgar Tongue, Paternoster-Row, London. Ewing, A 1939, 'A Suggested Non-Naturalistic Analysis of Good', Mind, vol 48, no. 189. Finlay, S 2004, 'The Conversational Practicality of Value Judgment', The Journal of Ethics, vol 8, pp. 205-33. Finlay, S 2005, 'Value and Implicature', Philosopher's Imprint, vol 5, no. 4. Foot, P 1978a, 'Moral Arguments', in Virtues and Vices, University of California Press, Berkeley, CA. Foot, P 1978b, 'Moral Beliefs', in Virtues and Vices, University of California Press, Berkeley, CA. Freedman, C 1997, 'The Morality of Huck Finn', Philosophy and Literature, vol 21, no. 1, pp. 102-113. Fricker, M 2001, 'Confidence and Irony', in E Harcourt (ed.), Morality, Reflection and Ideology, Oxford, Oxford University Press. 314 Gibbard, A 1990, Wise Choices, Apt Feelings, Harvard University Press, Cambridge, Mass. Gibbard, A 1992, 'Thick Concepts and Warrant for Feelings', The Proceedings of the Aristotelian Society, vol 66. Gibbard, A 2003a, 'Reasons: Thick and Thin', Journal of Philosophy, vol 100, pp. 288- 304. Gibbard, A 2003b, Thinking How to Live, Harvard University Press, Cambridge, Mass. Godfrey-Smith, P 2003, Theory and Reality, University of Chicago Press, Chicago. Grice, H 1989, Studies in the Ways of Words, Harvard University Press, Cambridge, Mass. Hallie, P 1969, The Paradox of Cruelty, Middleton, CN, Wesleyan University Press. Hallie, P 1992, 'Cruelty', in LE Becker (ed.), Encyclopedia of Ethics, Garland, New York. Hardy, T 2005 [1891], Tess of the d'Urbervilles, Barnes and Noble, New York. Hare, R 1952, The Language of Morals, Clarendon Press, Oxford. Hare, R 1963, Freedom and Reason, Oxford University Press, Oxford. Hare, R 1981, Moral Thinking, Oxford University Press, New York. Hare, R 1997, Sorting Out Ethics, Oxford University Press, Oxford. Harman, G 1977, The Nature of Morality, Oxford Unversity Press, New York. Holzman, D 1998, 'The Place of Filial Piety in Ancient China', Journal of American Oriental Society, vol 118, no. 2, pp. 185-99. Hom, C 2008, 'The Semantics of Racial Epithets', The Journal of Philosophy, vol 105, pp. 416-40. Hookway, C 1985, Peirce, Routledge and Kegan Paul, London. Hookway, C 2004, 'Truth, Reality, and Convergence', in C Misak (ed.), The Cambridge Companion to Peirce, Cambridge University Press, Cambridge. 315 Hornsby, J 2001, 'Meaning and Uselessness: How to Think About Derogatory Words', in P French, H Wettstein (eds.), Midwest Studies in Philosophy, Blackwell Publishers, Oxford. Hume, D 1978, A Treatise of Human Nature, 2nd edn, Clarendon Press, Oxford. Hurka, T 2001, Virtue, Vice and Value, Oxofrd University Press, Oxford. Hurka, T 2007, 'From Thick to Thin: Two Easy Moral Reduction Plans', Unpublished Manuscript. Hurley, S 1989, Natural Reasons, Oxford University Press, Oxford. Hursthouse, R 1995, 'The Virtuous Agent's Reasons: A Reply to Bernard Williams', in R Heinaman (ed.), Aristotle and Moral Realism, UCL Press Limited, London. Jackson, F 1998, From Metaphysics to Ethics, Oxford University Press, Oxford. Jardine, N 1995, 'Science, Ethics, and Objectivity', in J Altham, R Harrison (eds.), Mind, World, and Ethics, Cambridge University Press, Cambridge. Joyce, R 2001, The Myth of Morality, Cambridge University Press, Cambridge. Kekes, J 1984, 'Moral Sensitivity', Philosophy, vol 59, no. 227, pp. 3-19. Kekes, J 1993, The Morality of Pluralism, Princeton University Press, Princeton. Kim, J 1993, Supervinience and Mind, Cambridge University Press, Cambridge. Koons, J 2003, 'Why Response-Dependence Theories of Morality are False', Ethical Theory and Moral Practice, vol 6, pp. 274-94. Kovesi, J 2004 [1967], Moral Notions, Cybereditions Corporation, Christchurch, New Zealand. Kraybill, D 2001, The Riddle of Amish Culture, Johns Hopkins University Press, Baltimore. Krichin, S 2003, 'Particularism, Generalism, and the Counting Argument', European Journal of Philosophy, vol 11, no. 1. Krichin, S 2007, 'Moral Particularism - An Introduction', Journal of Moral Philosophy, vol 4. 316 Kuhn, T 1996 [1962], The Structure of Scientific Revolutions, 3rd ed., University of Chicago Press, Chicago. Lang, G 2001, 'Rule-Following Considerations and Metaethics', European Journal of Philosophy, vol 9, no. 2. Lawrence, G 1993, 'Reflection, Practice and Ethical Scepticism', Pacific Philosophical Quaterly, vol 74, pp. 289-361. Lear, J 1983, 'Ethics, Mathematics, and Relativism', Mind, vol XCII. Li, C 1997, 'Shifting Perspectives: Filial Morality Revisited', Philosophy: East and West, vol 47, no. 2, pp. 211-32. Little, M 2000, 'Moral Generalities Revisited', in B Hooker, M Little (eds.), Moral Particularism, Oxford University Press, Oxford. Lovibund, S 1983, Realism and Imagination in Ethics, University of Minneapolis Press, Minneapolis. Lovibund, S 2002, Ethical Formation, Harvard University Press, Cambridge, MA. MacIntyre, A 1984, After Virtue, University of Notre Dame Press, Notre Dame, ID. MacIntyre, A 1988, Whose Justice? Which Rationality?, University of Notre Dame Press, Notre Dame. Mackie, J 1977, Ethics: Inventing Right and Wrong, Penguin Books, London. Major, B 2005, 'Moral Discourse and Descriptive Moral Properties', The Philosophical Quarterly, vol 55, no. 220. McDowell, J 1986, 'Review of "Ethics and the Limits of Philosophy"', Mind, New Series, vol 95, no. 379, pp. 377-86. McDowell, J 1996, Mind and World, Harvard University Press, Cambridge, Mass. McDowell, J 1998a, 'Aesthetic Value, Objectivity and the Fabric of the World', in Mind, Value and Reality, Harvard University Press, Cambridge, Mass., MA. McDowell, J 1998b, 'Are Moral Requirements Hypothetical Imperatives?', in Mind, Value and Reality, Harvard University Press, Cambridge, Mass., MA. 317 McDowell, J 1998c, 'Non-Cognitivism and Rule-Following', in Mind, Value and Reality, Harvard University Press, Cambridge, Mass., MA. McDowell, J 1998d, 'Truth and Projection in Ethics', in Mind, Value and Reality, Harvard University Press, Cambridge, Mass., MA. McDowell, J 1998e, 'Values and Secondary Qualities', in Mind, Value and Reality, Harvard University Press, Cambridge, Mass., MA. McDowell, J 1998f, 'Virtue and Reason', in Mind, Value and Reality, Harvard University Press, Cambridge, Mass., MA. McDowell, J 2000, 'Replies', in M Willaschek (ed.), McDowell: Reason and Nature, Transaction Publishers, London. McGinn, C 1984, The Subjective View, Clarendon Press, Oxford. McLaren, A 1997, The Trials of Masculinity, Chicago University Press, Chicago. McNaughton, D 1988, Moral Vision, Blackwell Publishing, Oxford. McNaughton, D & Rawling, P 2000, 'Unprincipled Ethics', in B Hooker, M Little (eds.), Moral Particularism, Oxford University Press, Oxford. Milgram, E 1995, 'Inhaltsreiche ethische Begriffe und die Unterscheidung zwischen Tatsachen und Werten (‘Thick Ethical Concepts and the Fact/Value Distinction’)', in C Fehige, G Meggle (eds.), Zum Moralischen Denken, Suhrkamp, Frankfurt. Milgram, E 2009, 'D'ou venons-nous.Que sommes nous.Ou allons-nous?', in D Callcut (ed.), Reading Bernard Williams, Routledge, New York. Miller, A 2003, Introduction to Contemporary Metaethics, Blackwell Publishing, Oxford. Moody-Adams, M 1994, 'Culture, Responsibility and Affected Ignorance', Ethics, vol 104. Moore, A 1991, 'Can Reflection Destroy Knowledge?', Ratio (New Series), vol 4. Moore, A 1997, Points of View, Clarendon Press, Oxford. Moore, A 2002, 'Quasi-Realism and Relativism', Philosophy and Phenomenological Research, vol LXV, no. 1, pp. 150-56. 318 Moore, A 2003, 'Williams on Ethics, Knowledge and Reflection', Philosophy, vol 78, pp. 337-54. Moore, A 2006, 'Maxims and Thick Ethical Concepts', Ratio (new series), vol 19, no. 2, pp. 129-47. Moore, A 2007, 'Realism and the Absolute Conception', in A Thomas (ed.), Bernard Williams, Cambridge University Press, Cambridge. Moran, R 2001, Authority and Estrangement, Princeton University Press, Princeton. Morrison, B 2003, 'Mind, World, and Language: McDowell and Kovesi', Ration (new series), vol XV, no. 3, pp. 293-308. Murdoch, I 1970, The Sovereignty of Good, Routledge & Kegan Paul, London. Nagel, T 1971, 'The Absurd', Journal of Philosophy, vol LXIII, no. 20, pp. 716-27. Nagel, T 1986, The View from Nowhere, Oxford University Press, New York. Norman, R 2007, 'Particularism and Reasons - A Reply to Kirchin', Journal of Moral Philosophy, vol 4, pp. 33-39. Nozick, R 1981, Philosophical Explanations, Harvard University Press, Cambridge, Mass. Payne, A 2005, 'A New Account of Thick Concepts', The Journal of Value Inquiry, vol 39, pp. 89-103. Peacocke, C 1992, A Study of Concepts, MIT Press, Cambridge, Mass. Peirce, C 1871, 'Fraser's "The Works of Bishop Berkeley"', North American Review, vol 113, pp. 449-72. Peirce, C 1931, 'How to Make Our Ideas Clear', in C Hartshorne, P Weiss (eds.), The Collected Papers of Charles Sanders Peirce, Vol. 5, Cambridge, Mass., Harvard University Press. Perrett, RW 1999, 'Musical Unity and Sentential Unity', British Journal of Aesthetics, vol 39, no. 2. Platts, M 1979, Ways of Meaning, Routledge and Kegan Paul, London. 319 Putnam, H 1978, Meaning and the Moral Sciences, Routledge and Kegan Paul, New York. Putnam, H 1981, Reason, Truth and History, Cambridge University Press, Cambridge. Putnam, H 1987, The Many Faces of Realism, Open Court Publishing, LaSalle, IL. Putnam, H 1990, 'Objectivity and the Ethics/Science Distinction', in Realism with a Human Face, Harvard University Press, Cambridge, Mass. Putnam, H 1992, Renewing Philosophy, Harvard University Press, Cambridge, Mass. Putnam, H 1994, Words and Life, Harvard University Press, Cambridge, Mass. Putnam, H 2002, Collapse of the Fact/Value Dichotomy, Harvard University Press, Cambridge, Mass. Quine, W 1990, The Pursuit of Truth, Harvard University Press, Cambridge, Mass. Quinn, W 1987, 'Reflection and the Loss of Moral Knowledge: Williams on Objectivity', Philosophy and Public Affairs, vol 16, no. 2, pp. 195-209. Rabinowicz, W & Ronnov-Rasmussen, T 2004, 'the Strike of the Demon: Pro-Attitudes and Value', Ethics, vol 114, pp. 391-423. Raz, J 1979, 'Kelsen's Theory of the Basic Norm', in The Authority of Law, Clarendon Press, Oxford. Raz, J 1999, 'Notes on Value and Objectivity', in Engaging Reason, Oxford University Press, Oxford. Rorty, R 1982, Consequences of Pragmatism, University of Minnesota Press, Minneapolis. Rorty, R 1994, 'Taylor on Truth', in J Tully (ed.), Philosophy in the Age of Pluralism, Cambridge University Press, Cambridge. Ryle, G 1971, 'The Thinking of Thoughts: What is 'Le Penseur' Doing?', in Collected Papers, Vol. 2, Routledge, London. Scanlon, T 1995, 'Moral Theory: Understanding and Disagreement', Philosophy and Phenomenologial Research, vol LV, no. 2. 320 Scheffler, S 1987, 'Morality Through Thick and Thin: A Critical Notice of Bernard Williams 'Ethics and the Limits of Philosophy'', Philosophical Review, vol 96, pp. 411-38. Schiffer, S 1990, 'Meaning and Value', Journal of Philosophy, vol 87, pp. 602-14. Skorupski, J 2007a, 'Internal Reasons and the Scope of Blame', in A Thomas (ed.), Bernard Williams, Cambridge University Press, Cambridge. Skorupski, J 2007b, 'What is Normativity?', Disputatio, vol 2, no. 23, pp. 247-69. Sreenivasan, G 2001, 'Understanding Alien Morals', Philosophy and Phenomenological Research, vol LXIII, no. 1. Stevenson, C 1937, 'The Emotive Meaning of Ethical Terms', Mind, vol 46. Stocker, M 1970, 'Morally Good Intentions', The Monist, vol 54. Stocker, M 1973, 'Act and Agent Evaluations', Review of Metaphysics, vol 27, no. 1. Stocker, M 1976, 'The Schozophrenia of Modern Ethical Theories', The Journal of Philosophy, vol 73, no. 14. Stout, J 2001, Ethics After Babel: The Langauges of Morals and Their Discontents, Princeton University Press, Princeton. Strawson, P 1961, 'Social Morality and Individual Ideals', Philosophy, vol 36, no. 136, pp. 1-17. Tappolet, C 2004, 'Through Thick and Thin: Good and Its Determinates', Dialectica, vol 58, no. 2, pp. 207-21. Tasioulas, J 1998, 'Relativism, Realism and Reflection', Inquiry, vol 41, pp. 377-410. Taylor, C 1989, Sources of the Self, Harvard University Press, Cambridge, Mass. Thomas, A 2006, Value and Context, Clarendon Press, Oxford. Thomas, A 2007, 'The Nonobjectivist Critique of Moral Knowledge', in A Thomas (ed.), Bernard Williams, Cambridge University Press, Cambridge. Twain, M 1994, Adventures of Huckleberry Finn, The Easton Press, Norwalk, CN. 321 United Nations 2004, Compendium of United Nations Standards and Norms in Crime Prevention and Criminal Justice, viewed 15 March 2008, http://www.uncjin.org/ Standards/Compendium/pt2h.pdf. University of Southern California 2008, University Governance: Scampus 2007-08, viewed 15 March 2008, http://web-app.usc.edu/scampus/wpcontent/uploads/ 2007/08/university_governance.pdf. Urmson, J 1950, 'On Grading', Mind, New Series, vol 59, no. 234, pp. 145-69. Vayrynen, P 2008, 'Thick Evaluative Terms and Gradability (unpublished manuscript)'. Wiggins, D 1991, 'Moral Cognitivism, Moral Relativism and Motivating Moral Beliefs', Proceedings of the Aristotelian Society, pp. 61-85. Wiggins, D 1998a, 'A Sensible Subjectivism?', in Needs, Values, Truth 3rd ed., Clarendon Press, Oxford. Wiggins, D 1998b, 'Truth and Truth as Predicated of Moral Judgments', in Needs, Values, Truth 3rd ed., Clarendon Press, Oxford. Wiggins, D 2005, 'Objectivity in Ethics: Two Difficulties, Two Responses', Ratio, vol 18, no. 1, pp. 1-26. Wiggins, D 2006, Ethics: Twelve Lectures on the Philosophy of Morality, Harvard University Press, Cambridge, Mass. Williams, B 1972, 'Comments on 'Contraception and Chastity'', The Human World, vol 9, pp. 41-51. Williams, B 1973, 'Morality and the Emotions', in The Problems of the Self, Cambirdge University Press, Cambridge. Williams, B 1981a, 'Internal and External Reasons', in Moral Luck, Cambridge University Press, Cambridge. Williams, B 1981b, 'Moral Luck', in Moral Luck, Cambridge University Press, Cambridge. Williams, B 1981c, 'The Truth in Relativism', in Moral Luck, Cambridge University Press, Cambridge. Williams, B 1981d, 'Wittgenstein and Idealism', in Moral Luck, Cambridge University Press, Cambridge. 322 Williams, B 1985, Ethics and the Limits of Philosophy, Harvard University Press, Cambridge, Mass. Williams, B 1986, 'Reply to Blackburn', Philosophical Books, vol 27, no. 4. Williams, B 1995a, 'Ethics and the Fabric of the World', in Making Sense of Humanity, Cambridge University Press, Cambridge. Williams, B 1995b, 'Moral Incapacity', in Making Sense of Humanity, Cambridge University Press, Cambridge. Williams, B 1995c, 'Replies', in J Altham, R Harrison (eds.), Mind, World and Ethics, Cambridge University Press, Cambridge. Williams, B 1995d, 'What does intuitionism imply?', in Making Sense of Humanity, Cambridge University Press, Cambridge. Williams, B 1995e, 'Who Needs Ethical Knowledge?', in Making Sense of Humanity, Cambridge University Press, Cambridge. Williams, B 1996a, 'History, Morality and the Test of Reflection', in Sources of Normativity, Cambridge University Press, Cambridge. Williams, B 1996b, 'Truth in Ethics', in B Hooker (ed.), Truth in Ethics, Blackwell Publishing, Oxford. Williams, B 2002, Truth and Truthfulness, Princeton University Press, Princeton. Williams, B 2005 [1978]a, Descartes: The Project of Pure Enquiry, Routledge Press, New York. Williams, B 2005b, 'Pluralism, Community and Left Wittgensteinianism', in In The Beginning Was The Deed, Princeton University Press, Princeton. Williams, B 2006a, 'Philosophy as a Humanistic Discipline', in Philosophy as a Humanistic Discipline, Princeton University Press, Princeton. Williams, B 2006b, 'Political Philosophy and the Analytic Tradition', in Philosophy as Humanistic Discipline, Princeton University Press, Princeton. Williamson, T 2009, 'Reference, Inference and the Semantics of Pejoratives', in J Almog, P Leonardi (eds.), The Philosophy of David Kaplan, Oxford University Press, Oxford. 323 Winch, P 1958, The Idea of a Social Science, Routledge & Kegan Paul, London. Wittgenstein, L 1969, On Certainty, Harper Torchbooks, New York. Wittgenstein, L 1974 [1922], Tractatus Logico-Philosophicus, Routledge, London. Wittgenstein, L 2001 [1951], Philosophical Investigations, Backwell Publishing, Oxford. Wittgenstein, L 2007 [1967], Zettel, University of California Press, Berkeley, CA. Wolf, S 1987, 'The Deflation of Moral Philosophy', Ethics, vol 97. Wolf, S 2007, 'Moral Psychology adn the Unity of the Virtues', Ratio (New Series), vol XX, no. 2. Wong, D 2006, Natural Moralities: A Defense of Pluralistic Relativism, Oxford University Press, Oxford. Wright, C 1992, Truth and Objectivity, Harvard University Press, Cambridge, Mass. Zangwill, N 1995, 'The Beautiful, the Dainty, and the Dumpy', British Journal of Aesthetics, vol 55, no. 4, pp. 317-29.
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
On the variety and inter-relation of ethical judgments
PDF
Cause and consequence: confronting ethical challenges in models of morally informed legal practice
PDF
Descartes's creation doctrine and modality
PDF
Numbers, symbols, and induction: the genetic and finite concepts of number
PDF
The other side of authenticity: authenticity, liberalism, and resistance
PDF
Literature as narrative ethics: ethics, religion, and scripture in Barbara Kingsolver's The poisonwood bible
PDF
Interorganizational knowledge networks: the case of the biotechnology industry
PDF
Chemical applications of density functional theory as an analytical tool
PDF
Just imagining things: Hume's conception-based account of cognition
PDF
Development of high-frequency (~100mhz) PZT thick-film ultrasound transducers and arrays
PDF
Negotiating pluralism and tribalism in liberal democratic societies
PDF
Consumers' subjective knowledge influences evaluative extremity and product differentiation
PDF
Crowdsourcing for integrative and innovative knowledge: knowledge diversity, network position, and semantic patterns of collective reflection
PDF
Hobbes, education and the long run stability of nations: toward a Hobbesian model for contemporary religious theocratic socities
PDF
Language and the structure of Berkeley's world
PDF
An evaluation study of the Executive Master of Business Administration (EMBA) degree: examining the transfer of knowledge and leadership practices
PDF
Hellenistic historiography and the sciences: practices and concepts in Polybius' Histories
PDF
Developing a neural prosthesis for hippocampus: proof-of-concept using the in vitro slice
PDF
Believing someone: the omission of the interpersonal component of testimonial belief and the concomitant limitations of merely believing a proposition
PDF
Visual reasoning: representations in global politics, economics, and society
Asset Metadata
Creator
Pryor, Francis J.
(v)
Core Title
Thick concepts, reflection, and the loss of ethical knowledge
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Philosophy
Publication Date
09/08/2009
Defense Date
07/08/2009
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
Bernard Williams,ethical knowledge,ethics,OAI-PMH Harvest,thick concepts
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Dreher, John (
committee chair
), Willard, Dallas (
committee member
), McCann, Edwin (
committee member
), Olson, Thomas H. (
committee member
)
Creator Email
jeff_pryor@hotmail.com,pryorv@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m2590
Unique identifier
UC1225816
Identifier
etd-Pryor-3191 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-264193 (legacy record id),usctheses-m2590 (legacy record id)
Legacy Identifier
etd-Pryor-3191.pdf
Dmrecord
264193
Document Type
Dissertation
Rights
Pryor, Francis J., V
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
Bernard Williams
ethical knowledge
thick concepts