University of Southern California Dissertations and Theses

Thomas Reid on singular thought

(USC Thesis Other)

Thomas Reid on singular thought

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THOMAS REID ON SINGULAR THOUGHT
by
Marina Radiana Folescu
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PHILOSOPHY)
August 2013
Copyright 2013 Marina Radiana Folescu

Dedication
To the memory of my beloved grandparents:
Patea and Mitic˘ a Umbr˘ arescu, and Ionel Folescu
ii
Acknowledgments
If it takes a village to raise a child, how many villages does it take to raise a philosopher? It
depends how one counts, but, in my case, it took at least two academic villages: I consider
myself an offspring of both UCLA and USC.
While my affiliation with the UCLA Department of Philosophy has been informal, I
would like to thank the faculty there, especially Tyler Burge, Andrew Hsu, David Kaplan,
and Terence Parsons, who encouraged me to pursue my love of philosophy, at a time when
it seemed impossible to do so. My undergraduate training was in art history, and, to pre-
pare for a doctorate in philosophy, I started attending the UCLA Philosophy of Language
Workshop. The first day I went, it felt like being thrown into the ocean and told to learn
quickly how to swim and avoid the big mean sharks. Most of what was discussed in that
first quarter went over my head, but somehow I felt at home. Every time I went, I felt
excited and motivated to go again. Every time I gave a presentation, I felt that this is what I
want to do for the rest of my life: search for the truth, in the company of people engaged in
a similar quest. I cannot begin to express the gratitude I have for this wonderful institution
and for its great effect it had in my initial formative years.
The USC School of Philosophy has proved to be the perfect environment for me to grow
as a philosopher. I am extremely grateful that USC took a chance with me, when almost no
one else would: my background in art history was seen as a liability, but not at USC! They
were prepared to receive someone interested in art, history, and philosophy. In this regard,
iii
I would like to thank George Wilson, who showed me first hand that an interest in analytic
aesthetics, in general, and in film theory, in particular, can be successfully cultivated in a
Philosophy Department. I am deeply indebted to Scott Soames, who was one of my close
mentors for the first two years in the program. He got to know me better than I knew
myself. Scott recommended that I should dedicate myself more closely to the study of the
history of philosophy and philosophy of mind, because that is where he thought my heart
was. I didn’t see it at that time, because I thought I felt at home studying vagueness, direct
reference, and the problem of predication, but, of course, he was right. I want to thank him
and James Higginbotham, for providing me with such a rigorous training in the philosophy
of language. Together with Robin Jeshion, who joined the USC Faculty in my fourth year,
they helped me shape my interest in the history of philosophy, focused on the philosophy
of language and mind.
Mark Schroeder, who was the Director of Graduate Studies during most of my years
at USC, deserves special thanks. He has gone well above and beyond the call of duty in
helping me successfully navigate all the obstacles the PhD program and life itself put in
my path. I am grateful for his constant guidance, help, and support, especially throughout
the most stressful period in my life. I think that Mark is sometimes more invested in our
success than we ourselves are, and he does everything in his power to help us achieve it.
I am happy to have met so many wonderful people, fellow travelers on the road to get-
ting a PhD in Philosophy. I thank each and every graduate student at USC and UCLA, for
taking part in lively philosophical conversations, and for being good intellectual friends.
The following deserve special thanks: Hrafn Asgeirsson, Rima Basu, Brian Blackwell,
Lee-Ann Chae, Tiffany Chang, Justin Dallman, Ashley Feinsinger, Brian Hutler, Rachel
Johnson, Ben Lennertz, Jen Liderth, Matt Lutz, Eliot Michaelson, Eileen Nutting, Lewis
Powell, Johannes Schmitt, Justin Snedegar, Cecilia Stepp, Tamar Weber, and Aness Web-
ster. They laughed at my jokes, offered sympathy when I was miserable, and, most impor-
iv
tantly, they made me laugh when I needed it the most!
Throughout the years, I have benefited greatly from going to conferences and presenting
my half-baked ideas in front of patient and supportive audiences. In this regard, I would
like to thank: Todd Buras, Rebecca (Becko) Copenhaver, Stephen Daniel, Raffaella De
Rosa, Todd Ganson, Mike LeBuffe, Yitzhak Melamed, Elliot Paul, Patrick Rysiew, Ericka
Tucker, and Tom Vinci.
As I reach the end of my PhD training, I cannot believe what a journey this was! I would
not be here, writing this dissertation, without my committee members: Ed McCann, Jim
Van Cleve, and Gideon Yaffe. I thank Ed for taking the time to talk to me about philosophy,
life, and the movies, especially since my attachment to Reid’s brand of common-sense
philosophy has almost always clashed with his greater enthusiasm for Locke and Hume’s
“skepticism”.
I am extremely grateful to Gideon, who was the first to encourage me to specialize in the
history of Early Modern philosophy. He read one of my early papers, and saw the potential
for a dissertation, right there and then. And he said so right away. At the end of my second
year, I already knew that I don’t have to spend any more time fishing out for a dissertation
topic; it had already presented itself, and all I had to do was stick with it. This would
be enough to make Gideon anyone’s advising hero. But there is more: whenever doubts
crowded in, all I had to do was talk to him, and I would find out the best way to tackle them.
Instead of feeling discouraged by the latest devastating objection that he’d found against
my arguments, I would be excited to work through it. And this is mainly because of his
attitude towards mistakes: we all make them, and without making them, chances are that
we wouldn’t uncover the truth. But most of all, I thank Gideon for showing me that doing
philosophy is fun, and for insisting that I be extremely careful in choosing research topics:
whatever I do, I need to be happy to spend a long time in their company.
I would also like to thank Jim, for his dedication to Reid and to my project, and for
v
knowing to be there whenever I needed him, but also to give me room to breathe and grow
whenever I needed to mull things over on my own. Besides being a great philosopher
and historian of philosophy, Jim is a great teacher: he always takes whatever you tell him
seriously and thinks it over, before reacting to what you said. You learn quickly to recognize
when he thinks you have a good point and when he takes your mere suggestion of an insight
and, by adding a sentence or two, makes it into the core of a chapter. Sometimes he just
asks a “mere” clarificatory question, which first confuses you greatly and in a year or so,
when you finally have an answer to it, clarifies and re-categorizes everything you’ve been
thinking on an issue. Above all, I thank Jim for always treating me like a colleague: junior
to him, to be sure, but engaged in the common project of finding the best path to the truth.
Finally, I would like to thank my family, who supported me unconditionally throughout
all my decisions, and who helped me relocate to the US, from far (my parents) and near
(my sister and brother-in-law). I thank my husband Alexandru Radulescu for everything: I
couldn’t have done any of this without you!
vi
Contents
Dedication ii
Acknowledgments iii
Abbreviations x
Abstract xi
Introduction 1
1 PerceptualandImaginativeConception: TheDistinctionReidMissed 10
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.2 Perceptual Conception . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.2.1 The Non-conceptual Character of Perceptual Conception . . . . . . 14
1.2.2 Perceptual Conception Selects Properties from Complete Objects . . 29
1.3 Imaginative Conception . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1.3.1 The Conceptual Character of Imaginative Conception . . . . . . . . 32
1.3.2 Imaginative Conception “Bestows” Qualities on Incomplete Objects 34
1.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2 PerceptionandItsObjects 43
vii
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.2 Immediate Objects of Perception – Textual Evidence . . . . . . . . . . . . 48
2.2.1 Evidence That Qualities Are Immediate Objects of Perception . . . 51
2.2.2 Evidence That Bodies Are Immediate Objects of Perception . . . . 58
2.3 Which View Should Reid Have Adopted? . . . . . . . . . . . . . . . . . . 63
2.3.1 The Textual Evidence . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.3.2 An Argument Against the Qualities-only View . . . . . . . . . . . 64
2.3.3 Dispelling An Objection Against the Qualities-and-bodies View . . 69
2.3.4 Is There Any Reason Favoring The Bodies-only View? . . . . . . . 71
2.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3 TheMechanismofPerception 75
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.2 Original versus Acquired Perception . . . . . . . . . . . . . . . . . . . . . 77
3.2.1 Acquired Perception Is the Fruit of Experience and Learning . . . . 79
3.2.2 Only Two of Our Senses Give Us Original Perceptions . . . . . . . 81
3.2.3 Some Secondary Qualities Are Originally Perceived . . . . . . . . 83
3.2.4 What About Bodies? . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.3 The Two Stages of Original Perception . . . . . . . . . . . . . . . . . . . . 91
3.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4 MemoryandthePerceptionofEvents 117
4.1 Memorial Conception Is Sometimes (Pre-)conceptual . . . . . . . . . . . . 119
4.2 The Objects of Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
4.2.1 An Inconsistent Triad . . . . . . . . . . . . . . . . . . . . . . . . . 138
4.2.2 The Need for A Metaphysics of Events . . . . . . . . . . . . . . . 143
4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
viii
5 ImaginationandItsNonexistentObjects 149
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
5.2 Ficta as Complex Universals . . . . . . . . . . . . . . . . . . . . . . . . . 153
5.2.1 The View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
5.2.2 Did Reid Hold This View? . . . . . . . . . . . . . . . . . . . . . . 156
5.2.3 Problems With the View . . . . . . . . . . . . . . . . . . . . . . . 159
5.3 Ficta as Abstract Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5.3.1 The View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5.3.2 Did Reid Hold This View? . . . . . . . . . . . . . . . . . . . . . . 172
5.3.3 Philosophical Problems With This View . . . . . . . . . . . . . . . 175
5.4 Ficta as Complex Tropes . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
5.4.1 The View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
5.4.2 Did Reid Hold This View? . . . . . . . . . . . . . . . . . . . . . . 185
5.4.3 Advantages of This View . . . . . . . . . . . . . . . . . . . . . . . 190
5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
Bibliography 195
ix
Abbreviations
EHU Hume, D. (2000). An Enquiry concerning Human Understanding. Oxford:
Clarendon Press. (Original work published in 1758-76).
T Hume, D. (2007). A Treatise of Human Nature. Oxford: Clarendon Press. (Orig-
inal work published in 1739-40).
Essay Locke, J. (1979). An Essay Concerning Human Understanding. Oxford: Claren-
don Press. (Original work published in 1700).
EAP Reid, T. (2010). Essays on the Active Powes of Man. Pennsylvania State Univer-
sity Press. (Original work published in 1788).
EIP Reid, T. (2002). Essays on the Intellectual Powers of Man. Pennsylvania State
University Press. (Original work published in 1785).
IHM Reid, T. (2003). An Inquiry into the Human Mind on the Principles of Common
Sense. Pennsylvania State University Press. (Original work published in 1764).
x
Abstract
In this dissertation, I offer a new reading of Thomas Reid’s philosophy of mind, in which I
bring attention to his so far overlooked view that we can entertain a singular thought about
an individual substance. What emerges is a picture of Reid’s theory as psychologically
sophisticated, anticipating some very recent and influential views in empirical psychology.
Two of Reid’s works, An Inquiry into the Human Mind on the Principles of Common
Sense (1764) and Essays on the Intellectual Powers of Man (1785), address central issues in
philosophy of mind, psychology and vision science. Our ability to have singular thoughts
is one of the recurring themes throughout his investigation. I discuss the way in which sin-
gular thoughts occur in conceiving, perceiving, remembering, and imagining, and I explain
how this singularity is brought about in some seemingly difficult cases.
Reid does not use the terminology of “singular thought”, which became widespread in
the twentieth-century, once the lessons from Russell and Quine had been fully appreciated.
In Chapter 1, however, I show that his interest in our ability to think about individual
substances in the world places him squarely within this tradition.
According to Reid, perception gives us immediate knowledge of the external world,
namely that type of knowledge that the skeptic denies we can have. In Chapter 2, I argue
that bodies, together with their qualities, are, in Reid’s terminology, immediate objects of
original perception. This is a controversial view in Reid scholarships, but I show that other
scholars overlook some key passages when they offer their reading.
xi
Perception of bodies is singular: according to Reid, one does not use a description of
the qualities of bodies, in order to perceive those bodies. In Chapter 3 I argue that this is
compatible with his view that perception of bodies is a two-stage process: at the first stage,
we perceive qualities; at the second, the bodies having those qualities.
In Chapter 4 I discuss the way in which individual conceptions enable memory to sup-
ply us with singular thoughts about the objects of the past. Given Reid’s overall theory of
memory and perception, he should have restricted the domain of memory to objects and
tropes. However, surprisingly, he claims that we also have episodic memories of events.
This is impossible according to his overall theory, because, on his understanding of percep-
tion of motion and change, we cannot perceive events, and episodic memory is grounded
in the previous perception of its objects.
However, one of the most striking consequences of Reid’s theory of singular thought
comes in his theory of imagination. According to Evans’ influential view, by definition
a singular thought is such that it is guaranteed to be about an existent thing. In contrast,
Reid’s theory, as I reconstruct it in Chapter 5, entails that we can imagine nonexistent
particulars, in a singular way.
xii
Introduction
The main goal of this dissertation is to determine Reid’s account of the contents of thought
(broadly construed). Scholars usually agree that, in opposition to his predecessors (Locke,
Berkeley, and Hume), Reid thinks that our mind is able to get in contact with the objects
in our environment, in an immediate manner. Our access to the world is not mediated by
mental entities, the so-called “ideas” of the British Empiricism; the mind apprehends the
objects in the world immediately.
Since Reid argues that the objects of the world are individuals, he thinks that we can
entertain thoughts about such objects, qua particulars. He even calls such thoughts “in-
dividual conceptions,” (i.e. thoughts about particulars) to differentiate them from what he
calls ”general conceptions” (i.e. thoughts about general entities, universals.) Guided by this
distinction, I investigate what is the nature of these individual conceptions, or thoughts: are
they, in some important respects, like what is usually called “singular thoughts”? What
is their specific feature(s) that makes it possible to think immediately about the particu-
lars surrounding us? I offer a new reading of Reid’s philosophy of mind, in which I bring
attention to his so far overlooked view that we can entertain a singular thought about an
individual substance – we can think immediately about a particular object. What emerges
is a picture of Reid’s theory as psychologically sophisticated, anticipating some very recent
and influential views in empirical psychology.
Two of Reid’s works, An Inquiry into the Human Mind on the Principles of Common
1
Sense (1764) and Essays on the Intellectual Powers of Man (1785), address central issues in
philosophy of mind, psychology and vision science.
1
Our ability to have singular thoughts
is one of the recurring themes throughout his investigation. By contrast to Hume, who be-
lieves that we do not have notions of individual substances, but only of bundles of qualities,
Reid argues that the human mind is able to think in an immediate manner about individ-
ual objects in the external world. This is a very broad claim, allowing for the fact that
“thinking” is a catch-all term for Reid, designed to refer to every operation of the mind:
conceiving, perceiving, remembering, and imagining are all ways of thinking about exter-
nal objects. I discuss the way in which singular thoughts occur in each of these faculties
and explain how this singularity is brought about in some seemingly difficult cases.
It is a substantive issue to know what types of objects our mind is able to get in contact
with, and what faculties are responsible for this contact. There are two types of particulars
Reid is concerned with: physical qualities of bodies, which, on his view, are to be under-
stood as tropes, and the bodies themselves. Both types of particulars are found in the world
and we can think about both of them. This is an important issue, which distinguishes Reid’s
philosophy from contemporary (to us) approaches to singular thought. In the literature on
singular thought, it is usually assumed that the object of such thoughts are just the bodies,
without their qualities. After all, the qualities of such objects are (usually) thought to be
universals, not particulars. The view that I’m attributing to Reid focuses on qualities as
they are found in objects, so that the qualities themselves are individuals and this enables
1
All references will be made to the Edinburgh Edition of Thomas Reid, Knud Haakonssen, general editor,
published in the USA by The Pennsylvania State University Press, University Park, Pennsylvania:
An Inquiry into the Human Mind on the Principles of Common Sense: A Critical Edition (1764),
Derek R. Brookes, editor, 1997 (IHM henceforth);
Essays on the Intellectual Powers of Man: A Critical Edition (1785), Derek R. Brookes, ed., Knud
Haakonssen, intro., Brookes and Haakonssen, annot., 2002 (EIP henceforth).
Citations will be given by using chapter or essay, section and page numbers. In some cases, the relevant lines
will also be mentioned.
2
us to think about them singularly. Reidean singular thoughts are about both types of partic-
ulars: qualities and objects. Moreover, the primitive type of singular thought, on this view,
turns out to be about tropes, and not about the physical objects themselves. One cannot
have singular thoughts about physical objects without having singular thoughts about their
qualities.
Reid does not use the terminology of “singular thought”, which became widespread
in the twentieth-century, once the lessons from Russell and Quine had been fully appre-
ciated. There are, however, important points of convergence between his ideas and those
of Russsell’s, whose work is considered to have laid the foundations for many of the de-
velopments on singular thought. His distinction between singular and general propositions
is the starting point of much of the debates in contemporary philosophy of language and
mind. According to him, thinking about the objects that we commonly interact with can be
done in either of two distinctive ways: one can think about them via descriptions, namely
as those things that satisfy a determined set of conditions. For example: I can think about
my blue beach ball as whatever object of mine that is round, blue, made of rubber and
used to play volley-ball on the beach in sunny weather. My thought is successful in being
about the blue beach ball, because that is the only object that simultaneously satisfies all
the conditions.
Alternatively, I can think about my blue beach ball directly: if I pack my bag for the
beach and I look around for things that could come in handy, I can think to myself “That
will provide us with something to do besides lying on the beach”, while visually singling
out the blue beach ball. In this case, my thought is again about the blue beach ball, but it
does not refer to it as the satisfier of a complex definite description, but it picks it out by
using something like a demonstrative, in thought. This demonstrative thought is said to be
singular.
This is a common starting point of any theory about singular thought. Of course, there
3
is more to the discussion. One important aspect concerns the so-called perspective that a
certain object is considered from: it is true that I visually attend to the blue beach ball from
a certain distance and I see it at a certain angle. When I think about it and single it out as an
object I would need to take with me to the beach, I do not “conceptualize that perspective
and use that conceptualization as my means of thinking of” it (Jeshion (2010a, p. 1).) This
perspective then does not interfere with the singularity of the thought I have about the beach
ball. My thought is singular, even though the perspective in question helps me determine
what object I’m singling out. It is important to remember this point when talking about
Reid’s theory of singular thought. The perspective can help us connect with something in
our environment, but, if it does not enter the content of the thought, the thought in question
will still be singular.
The literature on perception is divided between two main theories concerning the con-
tent of perceptual experiences.
2
The importance of perception can hardly be exaggerated
here: our thoughts about the world are at the foundation of any other kind of thought (be
it singular, or general). It may be even said that singular thought starts with perception:
or, at least, this is one of the chief tenets of one of these two theories. On this theory (es-
poused, among others by Burge (2010) and Campbell (2002)), perceptual experiences are
singular, in the sense that they involve the individual objects of the world. For instance, on
Burge’s view, each perceptual experience contains an individual and a general component
and takes the form “this red” – where “this” is supposed to pick out an individual body, and
”red” is the attribution of a general property to that individual. This is done automatically,
non-inferentially, and non-propositionally, by our perceptual apparatus.
The alternative theory claims that the contents of our perceptual experiences are always
general, namely existentially quantified propositions-like entities, which pick out objects
2
For more details, see Jeshion (2010a, p. 10).
4
in the world in virtue of those objects satisfying the given set of properties. No singular
thoughts are ever entertained in perception. On this theory, the content of a perceptual ex-
perience is something like “there is a body of such-and-such size and shape occupying my
visual field, to the left”. The object in question, that happens to satisfy all these conditions,
is not part of my thought: only its properties are, and they are thought of as being instan-
tiated in a body. This construal of the contents of our perceptual experiences is, in a way,
object-independent. It does not matter which body is there, as long as there is a body with
all the properties I attribute to it, by way of a description.
3
The former theory is preferable, not least because it allows for singular perceptual
thoughts, but I argue that the logical space is not exhausted by these two: there is an inter-
mediary to consider. Moreover, not even the first theory is free from problems: it entails
that our perceptual apparatus is so designed as to be able to process and to attribute general
properties to individual objects. But how does one’s mind reach these general properties?
It is usually argued that, by a process of abstraction, after getting in contact with many
individuals of the same kind (in our example, red bodies), one will be able to form and
deploy the general concept “red”. This is Russell’s explanation of how one comes to be ac-
quainted with universals.
4
However, it seems that one needs to identify that those many red
bodies surveyed in order to form the concept “red” are indeed red, even before this concept
is acquired. I don’t think that there are many people today arguing that one cannot perceive
a red thing without possessing the concept “red” (e.g. Burge (2010) certainly doesn’t), but
why then argue that the structure of a perceptual experience is given by both an individual
and a general element? Why not say instead that this content contains two individual el-
3
There are more or less sophisticated versions of this view, espoused, among others by Searle, but all
suffer from a common defect: they do not account well for the phenomenology of our experiences, which
seem to be about bodies and their properties, and not about existentially quantified propositions. For more
details see Jeshion (2010a, p. 10).
4
See, for instance, Russell (1999).
5
ements: namely the body in question and a certain shade of red, that may resemble other
shades more than others? Once we recognize this fact, we will categorize the body as being
red.
Given this theory about how general concepts are acquired, I think this to be a bet-
ter starting point for a theory of perception, and thus of singular thought: the structure of
our perceptual thoughts is given by two individual elements – the body and the quality.
Throughout this dissertation, I argue that such a theory was espoused by Reid. On this the-
ory, the story of singular thought starts with perception and, more precisely, with one of the
main ingredients of this faculty, namely conception.
5
Telling this story will include Reid’s
arguments concerning the way in which any perceptual experience starts from a quality;
our mind gets in contact with the body that quality belongs to, but only after apprehending
that quality. The content of such an experience could be given by a phrase like “this red”,
but in this case, “this” refers to the particular red trope that one is experiencing at a given
moment. Our mind apprehends this quality by sensing and conceiving it. This type of
conception of a physical quality, based on the sensation one has when in contact with that
quality is perhaps best described as Russellian acquaintance.
6
One’s mind is perceptually
aware of a quality and can go on from there to perceive the body. I want to emphasize that
the sensation of the quality has a crucial role to play in the perception of the body having
that quality. I argue however that Reid’s story of singular thought does not end with physi-
5
Van Cleve (2013) argues that it is not obvious that Reid thinks that conception is an ingredient of percep-
tion, memory and imagination, and that there are passages indicating that conception is more of a concomitant
of these faculties, which are simples, not compounds. For the present purposes, it does not matter greatly
whether conception is an ingredient or a concomitant of these other faculties, as long as it is understood that
even as concomitant, it is necessary for perception, memory and imagination. I have a slight preference for
thinking that conception is an ingredient, and so I will continue to speak this way throughout the dissertation.
Some of the representative passages for thinking that conception is an ingredient of memory, for instance,
can be found in: EIP II. 20, p. 232 and EIP III. 2, p. 255.
6
See the arguments offered in support of this idea by Alston (1989) and Van Cleve (2004). There are
some issues with understanding perceptual conception as acquaintance, since acquaintance is supposed to be
always direct, whereas conception can sometimes be indirect, or, as Reid would say, relative. I will discuss
these issues in Chapter 1 and in Chapter 2.
6
cal qualities. There is a way of describing as singular the perceptual experience of the body
the quality belongs to, even though that experience is closely connected with the quality
itself.
To sketch this explanation here, it has to be acknowledged that our thoughts about the
bodies whose qualities are initially perceived can be singular, but they cannot be had in-
dependently of the qualities themselves. The crucial bit here is that we do not think about
the object as something (whatever it is) that satisfies a complex description, made up of
particular properties. But, we think about the object, with that quality, as being a particular
in the world. It’s as if I’m always going to think (singularly) about my blue beach ball
by having a thought with the content: “that blue ball” (where both the color and the body
are individuals, picked out in my thought.) To put it a bit more technically, it is as if one
were given the qualities of a body in perception and then one’s perceptual apparatus used
something like a “dthat” operator, in order to get to the content (namely the body itself)
of those qualities. Once this is done, the body in question is going to be the subject of
singular thoughts. I argue that this is done by perception, and not by a different faculty,
such as judgment. This accommodates Reid’s views, according to which bodies are actu-
ally perceived, and not simply inferred to exist. A singular thought à la Reid is thus doubly
singular: this is different from both the object-dependent and the object-independent con-
temporary theories concerning the contents of perception. As it will become apparent in
Chapter 5, this result helps explain how we can have singular imaginative thoughts.
The structure of the dissertation is as follows. In Chapter 1, I explain Reid’s views
on the faculty of conception, bringing to light a distinction between perceptual concep-
tion and imaginative conception (or conception simpliciter). This distinction should have
been made explicit by Reid, since he endorses some of the distinguishing characteristics.
It is important to make this distinction apparent, since it turns out that for Reid, percep-
tual conception is non-conceptual, whereas imaginative conception is conceptual. While
7
in perception we just grasp the world around us, in imagination we “construct” the nonex-
istent using concepts, as building materials. This distinction was missed by the secondary
literature on Reid and it is important for understanding how singular thought functions in
perception and in imagination.
According to Reid, perception gives us immediate knowledge of the external world,
namely that type of knowledge that the skeptic denies we can have. Scholars are divided
concerning the issue of what can be originally and immediately perceived: Lehrer & Smith
(1985), Nichols (2007), and Buras (2009) think that the only candidates here are primary
qualities of bodies. On their view we cannot have perceptual singular thoughts about ma-
terial objects. Others, e. g. Van Cleve (2004) and Copenhaver (2010), argue that some
secondary qualities can be perceived, but they are silent regarding the perception of bodies.
In Chapter 2, I argue that bodies, together with their qualities, are, in Reid’s terminology,
immediate objects of original perception, but that the perception of bodies cannot occur in
the absence of our having sensations of their qualities first.
This, however, does not entail that perception of bodies cannot be singular: according to
Reid, when perceiving bodies, one isn’t employing a complex description of their qualities.
Still, I argue, this is compatible with the view I’m attributing to him, namely that perception
of bodies is a two-stage process. The first stage involves the perception of qualities, while
the second stage involves the perception of the body having those qualities. These stages
are not inferential in nature, and the perception of bodies is thus immediate (in the relevant
sense).
After I present Reid’s views on the mechanism of perception in Chapter 3, I discuss
some contemporary psychological research, to show how Reid’s theory could be devel-
oped. In particular, I present the feature-integration theory of attention, which discusses
the different stages of human visual perception and the formation of object files (e. g.
Treisman et al. (1977), Wolfe (2003), and Chan & Hayward (2009)). Of course, Reid did
8
not have access to the findings of these psychologists, and I don’t claim to find their theory
in his work. However, I argue that their theory embodies a model which is anticipated in
his writings, and that he would have taken their experiments as supporting his theory.
In Chapter 4, I discuss the way in which individual conceptions enable memory to
supply us with singular thoughts about the objects of the past. Following Van Woudenberg
(1999) and Copenhaver (2006), I begin by presenting the extent to which Reid anticipated
some of the distinctions made by contemporary psychologists and philosophers of mind
working on memory (e.g. Tulving (1983) and Burge (2003)). But there is a difficulty that
has not been noticed by these and other Reid scholars: given his overall theory of memory
and perception, Reid should have restricted the domain of memory to objects and tropes.
However, surprisingly, he claims that we also have episodic memories of events. This is
impossible according to his overall theory, because, on his understanding of perception of
motion and change, we cannot perceive events, and episodic memory is grounded in the
previous perception of its objects. Without his knowledge, Reid engenders an inconsistent
triad. I show how this puzzle can be solved, by adopting a broadly Reidean solution, while
emphasizing that this is not Reid’s solution. He is unaware that his account of memory has
this problem; but, I argue, if he were aware, he could have solved the puzzle in the manner
I suggest. With this caveat, it turns out that we can have singular (episodic) memories of
individual properties, objects, and events.
One of the most striking consequences of Reid’s theory of singular thought comes in
his theory of imagination. According to Evans’ influential view, by definition a singular
thought is such that it is guaranteed to be about an existent thing. By contrast, Reid’s theory,
as I reconstruct it in the last chapter, Chapter 5, entails that we can imagine nonexistent
particulars, in a singular way. The reason is that imagination differs from both perception
and memory: it does not involve any belief of the existence of its subject matter, and it can
be successful even when its subject matter does not exist.
9
Chapter1
PerceptualandImaginativeConception:
TheDistinctionReidMissed
1.1 Introduction
Conception has a prominent role to play in Reid’s philosophy of mind, as is apparent from
his Essays on the Intellectual Powers of Man. The present investigation concerns Reid’s
explanation of how objects (be they real or nonexistent) are conceived. According to him,
conception functions in two different ways: it is either an ingredient in another act of
thinking, such as perception or memory, or it is exercised on its own, sometimes about
objects that do not (and will never) exist. Fictional objects can be remembered, but to do
so the mind needs an initial, independent grip on them, which can only be achieved by an
exercise of “bare conception”.
1
This chapter shows that there is a deep-rooted tension in Reid’s understanding of con-
ception: although the type of conception employed in perception is closely related to the
1
These two ways of employing conception are discussed by Reid in EIP, I. 1, p. 24 (lines 19–22; 29–33),
and especially in EIP, IV . 1, p. 295–96.
10
one employed in imagination, three fundamental features distinguish perceptual conception
(as the former will be referred to throughout this dissertation) from imaginative conception
(as the latter will be called henceforth).
2
These features would have been ascribed by Reid
himself to conception as involved in perception, but not to conception as involved in imag-
ination. He should have recognized them as marking the former as a different kind from
the latter, and he should not have hastily lumped perceptual and imaginative conceptions
together.
The first and most important distinguishing trait concerns the fact that perceptual con-
ception of bodies does not essentially proceed by way of concepts. A child who does not
possess the concept “red” will be able to perceive a certain red object as it is, namely red. If
perceptual conception were propositional and if it were, as Reid thinks, an essential ingre-
dient in any perceptual experience, the child would not be able to perceive the red object
accurately, unless he had the concept “red” in his conceptual repertoire. Later in the chap-
ter, a more detailed explanation will show why considerations like this weigh in favor of
thinking that perceptual conception is non-conceptual or non-propositional. What has been
said so far, however, should be enough to make apparent the contrast with imaginative con-
ception. According to Reid, to imagine something one must combine different attributes
with which one was previously acquainted. To be able to do so, one must think of a specific
color, a specific size, shape, etc., put together in a certain way. This would not be possible if
the imaginer did not possess the respective concepts. In this sense, imaginative conception
is conceptual or propositional.
2
Reid argues that imagination, “when it is distinguished from conception,” is a species of conception,
namely the one that can be employed about visible objects (EIP, IV . 1, p. 306). It may be a bit misleading to
call the second type of conception ‘imaginative conception’. The use of this label should, however, highlight
the related, but different structures of perception and imagination. For a perceptual experience to take place,
two necessary conditions must be satisfied: the perceiver must conceive the object and have an irresistible
belief in the existence of the object perceived. By contrast, only one of these conditions is necessary for
imagination: the imaginer must conceive a certain object, without regard to its existence.
11
Second, the substances of which we have a relative notion in perception are complete
objects. By contrast, most of the objects of imagination are incomplete. This is a conse-
quence of how these objects are “given” to us: the power of imagination combines several
qualities together, but there is no requirement that in doing so imagination will completely
specify all the qualities had by such an object. For instance, very rarely an imaginer will
wonder what is the color of the stomach of a particular imagined centaur. If one does so
wonder, one can imagine it being a vivid shade of purple, if one so decides. However,
whether to undergo such an act of imagination or not is entirely open to that imaginer. This
issue is not already settled, as it is in the case of a particular existing horse. The point
here is that perception is always about complete objects, while imagination is about, at
most, completable, but incomplete objects. The nature of the objects conceived on a given
occasion makes a difference for what type of conception one employs.
Third, perceptual conception makes salient properties from the substance to which
those qualities belong, whereas imaginative conception bestows properties on the imag-
ined incomplete object.
The last two features are very closely connected: they imply one another, and thus they
stand or fall together. Consequently, they will be discussed together, in this chapter.
The starting point for establishing the fundamental distinction between perceptual and
imaginative conception is Reid’s characterization of perception in EIP. The next section
will explain how conception functions, when it is an ingredient in perception. A good way
of understanding perceptual conception is to liken it to the relation of ‘having in mind’,
discussed in Donnellan (1997). This explanation provides a richer context for developing
the other two distinguishing features of perceptual conception.
The last section discusses Essay IV of EIP, where Reid characterizes the power of
conception and imagination. The explanation of how imaginative conception operates is
based on the following model: the mind starts by considering a certain set of properties
12
and then orients itself towards an object that is supposed to uniquely correspond to that set.
In the course of defending this thesis, the contrast between the two kinds of conception
will fully emerge: while perceptual conception makes salient some of the properties had
by the complete perceived object, imaginative conception works by ‘creating’ an imagined
incomplete object.
1.2 PerceptualConception
According to Reid, when one conceives a material substance, as part of a perceptual experi-
ence, one does not classify or subsume that body under concepts. He argues that conception
proper is not propositional, and that it implies no judgment about what is conceived:
Thus we see that the words conceive, imagine, apprehend have two meanings,
and are used to express two operations of the mind, which ought never to be
confounded. Sometimes they express simple apprehension, which implies no
judgment at all; sometimes they express judgment or opinion. [...] “I conceive
an Egyptian pyramid.” This implies no judgment. [...] the thing conceived
[the Egyptian pyramid] may be no proposition, but a simple term only, as a
pyramid, an obelisk (EIP, I. 1, p. 25).
The interpretation offered here takes this idea one step further: not only is conception
proper not propositional, in the sense that it does not affirm or deny anything of the object
conceived, but, when employed in perception, it is not even conceptual. A mental counter-
part of the term “Egyptian pyramid” is not necessary in order for someone to perceive such
a pyramid, when in its presence. One might possess such a concept, and one might use
it when one sees a pyramid, but perception can still occur in its absence. In other words,
propositional concept-attribution may be deployed alongside perception, but this is neither
necessary for perception, nor is it the result of the way the material substance is conceived.
Rather, the concepts deployed in perception may belong to other operations of our minds
13
(e.g. judging), occurring at the same time as perceptual conception. Reid’s writings sup-
port this idea by indicating that perception (much like other operations of the mind) never
occurs in isolation and that we must pay special attention to distinguish what is specific to
perception from other faculties. The following two passages back up this interpretation:
Most of the operations of the mind, that have names in common language, are
complex in their nature, and made up of various ingredients, or more simple
acts; which, though conjoined in our constitution, must be disjoined by ab-
straction, in order to our having a distinct and scientific notion of the complex
operation (EIP, I. 1, p. 37).
This observation is made here only, that we may not confound the operations
of different powers of the mind, which, by being always conjoined after we
grow up to understanding, are apt to pass for one and the same (EIP, II. 5, p.
97).
Section §1.2.1 starts with a discussion of some of the secondary literature on Reidian
perception, where the consensus seems to be that perception of bodies does take place,
but that the conception associated with perception of bodies is necessarily conceptual. An
objection to this view will be raised and an alternative interpretation will be offered, arguing
that, given some cases that Reid himself discusses, Reidian perceptual conception of bodies
must be non-conceptual. This discussion leaves out perception of qualities, which has its
own special conditions, different from those necessary for the perception of bodies.
In §1.2.2, the other two features distinguishing perceptual from imaginative conception
are discussed in order to emphasize why Reid should be understood as saying that sub-
stances are complete objects and that perceptual conception works in making qualities of
substances salient to the mind.
1.2.1 TheNon-conceptualCharacterofPerceptualConception
In perception, we learn that the world is populated with objects and also what kind of
objects they are. There are two sides to any perceptual experience: one ‘physical’; the
14
other one ‘mental.’ The former is determined by the impression that the object perceived
makes on the perceiver’s organs of sense, the nerves and the brain. The latter relates to the
operation and the content of perception, understood as a faculty of our mind.
Concerning the physical side, no perceptual experience can occur in the absence of
an impression made by the object on a certain organ of sense (EIP, II. 2, p. 74). These
impressions correspond “exactly to the nature and conditions of the objects by which they
are made,” whereas our sensations and perceptions correspond only in a varying degree to
these impressions (EIP, II. 2, p. 76). This gives us a first characterization of perception: it is
not only direct, but also, for lack of a better term, ‘objectively’ perspectival. Our perception
depends on the “medium which passes between the object and the organ” (EIP, II. 2, p.
74). This medium is also responsible for making our perception more or less distinct. A
perceptual experience always gives us information about an object perceived under certain
conditions, which, in the case of visual perception, include, but are not restricted to: light,
distance, position. These conditions constitute the aforementioned perspective: it matters
for how somebody perceives a chair whether he sees it in broad daylight or at dusk, or
whether he sees it from three feet or from ten:
The objects in this room will be seen by a person in the room less and less dis-
tinctly as the light of the day fails; they pass through all the various degrees of
distinctness according to the degrees of the light, and at last, in total darkness,
they are not seen at all (EIP, II. 5, p. 97).
The way this objective perspective contributes to how something is perceived is signif-
icant for the distinction between perceptual and imaginative conception. Perceptual con-
ception is always of the object as it is presented there and then to the perceiver. By contrast,
imaginative conception has more to do with how the conceiver puts together certain quali-
ties the object conceived is thought to have. The imaginer, as opposed to the perceiver, is
entirely responsible for supplying a given ‘subjective’ perspective, from which the object
15
is imagined.
Reid characterizes the mental side of perception as having the following key ingredi-
ents: a conception of the object perceived and an irresistible and immediate belief of its
existence (EIP, II. 5, p. 96).
3
Perceptual conception presents the object to the perceiver,
while the belief allows the perceiver to assert that the object exists.
To see how everything works together, let us take as an example an instance of visual
perception (this is just for heuristic purposes; an instance of any type of perception, in any
of the perceptual sense-modalities, would work just as well). Suppose that someone sees a
chair. This means, first, that the chair in question must exist. Second, a certain impression
on the perceiver’s organ(s) of sight must be present. Third, this impression must give rise
to a visual sensation, which prompts the perceiver to perceptually conceive of the object
and believe in its (present) existence, in the place and in the position it is perceived to be.
To paint a complete picture of how perception works, we must understand what it is
that we perceive when we see, for instance, a chair. The most straightforward reading
of the passage in which Reid gives his official characterization of perception (EIP II. 5,
p. 96) indicates that when we see a chair, we perceive the chair, and not just some of its
qualities. A perceiver sees the chair with its qualities, but he does not see qualities and
then infers that they belong to a certain solid object, used for sitting. However, this reading
is in tension with several other passages from EIP. When Reid talks about the objects of
perception proper, he argues that they “are the various qualities of bodies” (EIP II. 17, p.
3
According to Copenhaver (2010, p. 291) the belief component of a perceptual experience has the role
of supplying the perceiver with information regarding what type of object he is currently perceiving. She
argues that this belief predicates certain properties to the object that the perceptual conception presents to
the perceiver. However, perceptual belief does not have this role: Reid argues that, in perception, one is
seized by a belief that the object perceptually conceived exists when it is thus conceived. So the only property
that perceptual belief seems to be responsible for attributing to the object is existence (on the assumption
that existence is such a property). This issue will not be further developed in this chapter, even though the
problem of concept-application is all the more interesting, on the assumption that perceptual conception is
not conceptual.
16
200). Previously, at the opening of the EIP, he argued that the following is one of the
principles taken for granted:
All the things which we immediately perceive by our senses, and all the things
we are conscious of, are things which must be in something else as their sub-
ject. Thus by my senses, I perceive figure, colour, hardness, softness, motion,
resistance, and such like things. But these are qualities and must necessarily
be in something that is figured, coloured, hard or soft, that moves, or resists. It
is not to these qualities, but to that which is the subject of them, that we give
the name of body (EIP I. 2, p. 43).
The straightforward interpretation of these two passages is that material substances are
not immediately (in the sense of ‘directly’) perceived; only primary qualities and color
have this privilege, according to Reid (EIP II. 17, p. 202).
4
The notion one forms of both
secondary qualities and bodies is relative (e. g. EIP II. 17, p. 203, and EIP II. 19, p. 219).
However, one must not think that Reid believes that we know that there exists a substance
because we infer that the qualities perceived are bundled together in a certain body. The
belief in the existence of the material substance must be non-inferential if we are to count
as perceiving that substance. The perception of material substances must be, in a certain
sense immediate, namely non-inferential.
5
One word of caution to the reader: Reid uses “direct” primarily as an antonym to “rel-
ative”, and not as a synonym to “immediate”. This contrast is made apparent by what he’s
saying in regards to the notions (or conceptions) one forms of primary qualities versus the
notions one forms of secondary qualities. Reid argues that we have direct notions only
of primary qualities. The notions of secondary qualities that we have are relative to our
sensations of secondary qualities. As a consequence of this, we can have the following
4
For more on Reid’s inclusion of color, a secondary quality, on the list of things straightforwardly per-
ceived, see the discussion on the perception of primary qualities versus the perception of secondary qualities,
in Chapter 3.
5
For more on the distinction between different notions of “immediacy” which Reid used, see Buras (2008).
17
situation: we can form an immediate perceptual notion, which is, nonetheless indirect,
namely relative. This is the only consistent interpretation of Reid’s being explicit that color
can be perceived, despite its being a secondary quality, of which we can have only relative
(i. e. indirect) notions. Since Reid thinks that we cannot have direct notions of colors via
our senses, if “direct” were a synonym for “immediate”, we would not be able to perceive
colors, on the assumption that perception is always immediate (in the relevant sense).
6
There are other options here, however: first, Reid may be interpreted as saying that
both substances and qualities are perceived, but that perception of qualities is not necessary
for the perception of bodies. In this case, the type of conception associated with the per-
ception of bodies is not necessarily conceptual. It is not conceptual, in the sense that the
body under consideration is not conceived just as that which uniquely instantiates all the
qualities perceived. If this were the case, the content of that perception would be given by
a complex definite description denoting the body in question; this cannot be done without
deploying concepts. Moreover, as it will be shown later on, if the perceiver is mistaken
regarding one quality, then the body in question is not actually perceived. On the current
interpretation, however, the body can be perceptually conceived and hence perceived, even
without correctly perceiving and hence conceiving all of its qualities.
Second, one can argue that both qualities and the objects they belong to are perceived,
but that perception of the latter cannot happen without the perception of the former. On this
view, perceptual conception of bodies involves concept-attribution and is propositional.
The former interpretation is the one developed throughout this chapter: whenever an
object makes an impression on our organs of sense, we have a sensation that suggests the
conception of the qualities had by that object and also a conception of the body having those
qualities. The conception of the body in question is non-propositional and the belief that the
6
These issues will be explained in more detail in Chapters 2 and 3.
18
body exists is non-inferential. This interpretation leaves open the possibility that perceptual
conception of qualities is conceptual, in the following sense. Unless one is in a position
to deploy concepts of specific colors, sizes, etc., one is not in a position to perceptually
conceive and hence perceive the respective qualities. This is a topic for a different project,
one that would be primarily concerned with concept acquisition; consequently, it is not
further discussed in this dissertation.
Chappell (1989), Yaffe & Nichols (2009), and Gallie (1997) favor different versions
of the latter interpretation, that could be thought of as “the perception of bodies by the
perception of qualities” view.
Chappell argues that for hardness to be perceived, an impression of hardness must be
made on the organs of sense and that, in turn, gives rise to a certain sensation. Such a
sensation would “suggest the quality of hardness [to the perceiver], that is, prompt him
both to conceive of this quality and believe that it exists in the hard body that first started
the perceptual process, which conception and belief [would] constitute his perception of
that quality” (Chappell (1989, p. 59)). According to him, we perceive qualities that we
believe to exist in the world and we also believe that they belong to certain bodies. On
this interpretation, it is unclear by what process someone starts with the perception of the
quality of hardness and ends up with the belief that the quality exists in the hard body. It is
possible that Chappell thinks that this missing step is supplied by an inference. Someone
perceives a quality; qualities cannot exist by themselves; hence there is an object which
is such-and-such (in this case, a hard body) that is perceived. According to Chappell,
the conception of body one forms in perception is necessarily connected with the quality
perceived and, as such, necessarily attributive. One is thus unable to conceive the body
in the absence of a mode of conception: in the example that he gives, one necessarily
conceives of the body as a hard substance, by first conceiving the quality of hardness. If
19
this process is inferential, and it seems to be, it cannot count as perception for Reid.
7
Yaffe & Nichols (2009, n. pg.) argue that “to conceive of an object is to be aware of that
object as the bearer of some particular property. [...] To perceive an object is to be aware
of it in a particular way, as the possessor of a particular quality, and, at the same time, to
be convinced that the object exists and is as you conceive it to be.” This passage indicates
that, in order to perceive a body, one must perceive the property the object is understood
to possess. Otherwise, one could not be perceptually aware of an object as the bearer of
that property. If the property in question is not perceived, but just thought to be perceived
(or plainly misperceived), the perception of the material substance is in question, since the
only way of being perceptually aware of the object is by correctly attributing a property
to that object. One consequence of Yaffe & Nichols’s interpretation is that perception of
objects, namely bodies, is conditionally dependent on perception of their qualities. The
tight connection they allege to exist between the conception of an object and the perception
of a property had by that object rules out the possibility of illusions, in which the property in
question is misperceived, but the object itself is actually perceived. The following passage
shows that Reid’s theory allows for such cases, thus Yaffe & Nichols’s interpretation is
problematic:
[I]n perception, the notion which our senses give of the object may be more or
less clear, more or less distinct, in all possible degrees.
Thus we see more distinctly an object at a small than at a great distance. An
object at a great distance is seen more distinctly in a clear than in a foggy day
(EIP II. 5, p. 96).
The idea here is that distinctness, or, more precisely, lack thereof, could be responsible
for visual illusions: for instance, in thick fog, a bright kite may look lie a menacing dragon.
7
Passages indicating that Reid thought that material substances can be perceived can be found throughout
EIP, and there is nothing to prevent this type of perceptual experience from occurring in the absence of
(accurate) perception of qualities. See, for instance, EIP II. 19, p. 219 (lines 29-35), and EIP IV . 3, p. 327.
20
To see why this is important, consider this: sometimes someone perceives an object, but
he is mistaken with regards to what properties that object has. In fact, he is mistaken even
with regards to what kind of object it is; all he knows is that there is an object there, and
that object causes him to see a certain shape, color, size, etc. To make the problem more
vivid, let us take an example, adapted from Donnellan (1997, p. 372-73). Suppose that
on a sunny afternoon, a group of friends goes for a walk in the park. In the distance, one
of them sees a man carrying an umbrella and says to the others: “How funny: that man
carries an umbrella! He is probably a tourist; it never rains in LA, in the summer.” To this,
one of the others replies: “What man with an umbrella? That is just a rock you’re seeing,
though it looks like a man carrying an umbrella.” The object of perception was the rock,
although the first perceiver took it to be a man with an umbrella. The question is: how did
the perceiver conceive of the rock? On Yaffe & Nichols’s interpretation, it follows that the
perceiver conceived of the rock as the bearer of the shape, color and size of a man with
an umbrella. But if this is true, it is unclear in what sense the perceiver can truthfully be
said to have been perceptually conceiving of the rock itself. Rocks have rock-properties,
not man-with-an-umbrella-properties. If the only way the perceiver was aware of that rock
was as the bearer of some properties (e.g. man-with-an-umbrella-properties), as Yaffe &
Nichols suggest, he cannot be said to have had any de re thoughts about the rock itself,
since rocks do not have man-with-an-umbrella-properties.
One could argue that, since a man with an umbrella and the rock in question have
common shape properties, the perceiver conceived of the rock as a presently existing thing
with a certain shape. So the argument above would show at most that rock-properties can
be incorrectly ascribed, but not shape-properties. However, a similar argument could be
generated for shape. To see how this works, think about a shape-illusion, for instance
the Hering illusion, in which two straight parallel lines appear to be curved. What this
illusion (and others like it) shows is that the object in question is perceived, even when the
21
perceiver is wrong about its shape. In this particular case, the straight shape of the lines is
misperceived, but the lines themselves are perceived. In this sense, it is accurate to say that
one had a de re perception of the two lines, although he had been wrong about their shape
all along.
Reid would describe situations like these by arguing that the rock (or the straight line)
was the object of perception, even though some of its qualities were misperceived. How-
ever, Yaffe & Nichols describe this situation differently: the perceiver misperceived the
object itself, because he misconceived it as the bearer of nonexistent properties, and there
was no other way of conceiving that object, in that situation. But illusions, understood as
perceptions of objects whose properties are misperceived, do happen and Reid has a way
of explaining why such anomalies occur.
8
What Reid argues that happens with animal and young children when they perceive
also counts against Yaffe & Nichols’s interpretation. Contrary to their claim that perceptual
awareness of an object is dependent on our correct attribution of a property to that object,
Reid argues that animals and infants do perceive, but they do not have the intellectual
abilities to distinguish between qualities and objects (EIP II. 19, p. 219, lines 18–25).
Rather, they perceive these things jumbled together. One way to interpret this is to think
that one can perceive an object, without relying on its properties, because sometimes no
distinction between objects and properties is actually made.
The interpretation in Gallie (1997) is similar to that of Yaffe & Nichols’s, but he takes it
one step further: he argues that conception in general, thus including perceptual conception,
requires one to have an ability for linguistic representation. He thinks that perceptual con-
ception cannot function in the absence of one’s prior grasp of concepts, and of words used
to express those concepts. Therefore, perception itself cannot properly function without the
8
See, for instance, his discussion of how the senses can be fallacious and how they could be improved, in
EIP II. 22 and II. 21, respectively.
22
perceiver having such a prior grasp and use of concepts. In reply to this interpretation, Reid
would point out that animals and young children are quite capable of perceiving the outside
world. It is not our superior perceptual abilities that distinguish us from animals, but our
powers to process the information we are fed by our senses: “brute animals, who have the
same senses that we have, cannot separate the different qualities belonging to the same sub-
ject, and have only a complex and confused notion of the whole” [emphasis added] (EIP
IV . 3, p. 327). What animals lack is our power of abstraction and analysis. Moreover, while
it is true that one can have a more or less accurate notion of the objects around oneself,
Reid nonetheless argues that such a notion is not formed in perception: “The child has all
the notion of it [a jack for roasting meat] which sight gives; whatever there is more in the
notion which the man forms of it, must be derived from other powers of the mind” (EIP
II. 5, p. 97). Gallie (1997, p. 321) unnecessarily over-intellectualizes a process thought by
Reid to be basic, and shared by us with infants and animals with no linguistic abilities.
Furthermore, one may perceptually conceive of things, as one often does in perception,
that are not always communicable through language. For instance, in perceiving a red
body, one thereby conceives that particular shade of red. But, if asked what shade of red
that is, the perceiver may be at a complete loss. Since our perceiver cannot verbalize what
particular shade of red he perceived and hence perceptually conceived, Gallie would have
to say that no perception actually took place. In support of Gallie, one might argue that
our perceiver perceived a red thing, since he is able to say that he is seeing a red body,
even though he does not know what particular shade of red that was. To make the problem
more apparent, then, think about a color that does not yet have a name (or whose name
is not known by a certain perceiver). A perceiver would not be able to deploy a linguistic
representation of that color, and, on Gallie’s interpretation, someone like that would not
actually be in a position to perceive that color. This is an unwanted consequence. Moreover,
Gallie did not provide enough textual evidence to show that this is Reid’s view. If we attend
23
to such evidence, passages like the ones quoted above support a different interpretation:
animals and human beings perceive even before they are able to exercise any linguistic
abilities.
Therefore, according to Reid, and contrary to the interpretations discussed so far, we
have a way of perceiving an object, and hence conceiving it, without necessarily perceiving
its qualities, and without necessarily thinking about that object under a certain concept –
e.g. “hard body”, or “red ball”, etc. Perhaps perceiving a body’s quality of being a thing
or a presently existing thing might be required, but such qualities are not relevant here,
since they are common to all perceived substances and would not be enough to distinguish
between one body and another. This result should make the issue concerning the conceptual
or non-conceptual character of perceptual conception easier to address and develop.
The previous discussion brings us to the crux of the problem: just what is this perceptual
conception of material substances? Reid’s first attempt to explicate conception is in Essay
I, where he defines some of the main terms of the book. “[T]o conceive, to imagine, to
apprehend [...] signify an act of the mind which implies no belief or judgment at all” (EIP
I. 1, p. 24). Thus to conceive means to apprehend something, to entertain a thought, a belief,
a sensation. To perceive a material substance, however, something else is needed: one must
also believe that what one presently conceives also exists. Such a belief cannot be formed
unless that object is conceived by the perceiver, on a given occasion. The conception had
by a perceiver may be clear or obscure; either way, a conception of the object is necessary
for the formation of the existential belief. Conception is more primitive than belief – there
can be no belief without conception, whereas there can be conception of an object without
belief in its existence (EIP, II. 20, p. 228). According to Reid, the faculty of conception is a
basic operation of the mind; it always has an object, although sometimes its object may not
exist. As it will become apparent from section §3, the type of conception employed about
nonexisting objects is importantly different from the type used to think about existing ones.
24
In perception, we conceive of real and existing objects, along with some of their quali-
ties. Our conception of an object, formed in perception, ranges between clear and obscure,
many times depending on external circumstances. The same can be said about our knowl-
edge of an object: it too comes in degrees. The conception of an object had in perception is
distinct from the notion of that object, had by understanding what qualities the object has
and how they fit together. A carpenter knows a lot more about chairs than a layperson does
just by looking at the chair in front of him. The impression made by the body in front of
him on his visual organs leads to his conceiving of the chair, but not necessarily as a chair,
i.e. as an object used for sitting. In order for perceptual conception to be propositional,
it should be necessary for a perceiver to conceive of the qualities of the chair as qualities
of such an object and conceive of the chair itself as an object used to sit on. Moreover,
such a perceiver would be required to have concepts denoting all these qualities and the
object itself (as a chair), for this type of conception to be conceptual. Whereas, Reid thinks
that one can perceptually conceive a chair, in a different way, i.e. non-descriptively and
non-propositionally:
Thus the notion which a child has of a jack for roasting meat, will be acknowl-
edged to be very different from that of a man who understands its construction,
and perceives the relation of the parts to one another, and to the whole. The
child sees the jack and every part of it as well as the man: The child, therefore,
has all the notion of it which sight gives; whatever there is more in the notion
which the man forms of it, must be derived from other powers of the mind,
which may afterwards be explained [emphasis added] (EIP II. 5, p. 97).
This passage supports the interpretation found in Alston (1989, p. 43): perceptual con-
ception of material substances is not about the use of ‘general concepts,’ and it can actu-
ally operate in their absence. The powers of classification can be used, but they are not
necessary for perception. And if they are used, the information they provide must be dis-
tinguished from the information provided by perception proper. This lesson is drawn from
25
Reid’s claim that the notion the child has by sight of that object is identical to the one the
man has by sight of the same object. Reid contrasts the conception of an object one has in
perception with the more sophisticated notion of the same object, at which one may arrive
after careful contemplation and after employing several other faculties (e. g. abstraction and
generalization). Alston is right in arguing that conception is not about subsuming an object
under a concept and thinking about it as being of a certain kind. The passage just quoted
adduces important evidence in favor of the idea that perceptual conception of substances
is not propositional. Just to perceive, one does not need to have and use general concepts.
If such concepts are however available, one will probably use them and have a richer type
of experience. Reid’s suggestion is that this type of experience is a compounded one, with
perception being just one of its components, together with abstraction and generalization
bringing in general concepts.
9
Van Cleve (2004) presents another argument against the idea that perceptual conception
of bodies is conceptual: if the conception involved in perception were conceptual (i.e.
formed with concepts and impossible in their absence), it would already be present in the
belief of the existence of the object perceived. In such a case, “[f]orming a conception
of an object would be entertaining some proposition about it and the belief component of
perception would consist in affirming that proposition” (Van Cleve (2004, p. 107).) But
then, one would need to have and be able to use a whole array of concepts when one
perceives that a chair is in front of oneself. This would be inconsistent with Reid’s idea that
infants and all kinds of animals can perceive objects, together with their qualities.
Let us make things more explicit. Van Cleve (2004, p. 108) proposes to understand
9
This issue brings to mind the controversy surrounding Reid’s notion of acquired perception. Some Reid
scholars believe that acquired perception is not proper perception, because it is inferential; while others
think that it is no more inferential than original perception. For more details, see Van Cleve (2004) and
Copenhaver (2010). Another way of drawing the line between original and acquired perception is to think
that the former involves the non-conceptual type of conception, while the conception employed by the latter
is fully conceptual. This issue will not be further developed in this chapter.
26
perceptual conception as a kind of Russellian acquaintance, since “it is not constituted
by conceptualization or judgment.” Van Cleve does not construe conception as the full-
blown Russellian version of acquaintance, since that would make it impossible for Reid to
argue that we perceive external objects.
10
Accordingly, perceptual conception should be
understood as the awareness of an object one has while perceiving that object. Contrary
to Russellian acquaintance, Reidian perceptual conception is not supposed to give one full
knowledge of the thing perceived. Examples like the one where someone perceives the rock
and thinks it is something else would not be possible: on Russell’s view, if the perceiver is
actually acquainted with the rock, he cannot think instead that he is perceptually acquainted
with a man with an umbrella. Whereas, on Reid’s view, this situation is entirely plausible.
To gain a better grip on this notion of perceptual conception, Donnellan’s relation of ‘hav-
ing in mind’ might be better suited here. Donnellan, unlike Russell, does not presuppose
that the ‘having in mind’ relation gives one knowledge about the respective object. Thus,
someone will be said to have the rock in mind, in the situation under consideration, even
though he incorrectly ascribes man-with-an-umbrella properties to it. Reid’s terminology
suggests that this is a better approximation: he uses ‘apprehension’ to indicate that to con-
ceive something often just means to entertain a thought about that thing, without judging
that thought to be true or false.
11
For this to happen, perceptual conception must be understood as having a non-conceptual
10
Russell (1910) argued that we are acquainted only with sense-data (which are mind-dependent), univer-
sals and ourselves; material substances cannot be objects of acquaintance.
11
In discussing Reid’s notion of conception, Wolterstorff (2001, p. 6) uses the same idea of having in
mind, or securing a mental grip onto something, to characterize it. Wolterstorff’s usage is suggestive and
appropriate. However, his notion of ‘having in mind’ is different from Donnellan’s: Wolterstorff argues that
the mental grip in question is secured by deploying a singular concept, understood to be something like a
definite description (Wolterstorff (2001, p. 15)). Whereas, one of Donnellan’s points is to show that one can
have something in mind, even though a definite description used to describe it is not satisfied by that thing.
Donnellan’s notion allows one to form singular thoughts about external objects, in the absence of correct
application of concepts to that thing, whereas Wolterstorff’s does not. For the reasons discussed in the main
text, Reid’s notion of perceptual conception is better captured by Donnellan’s.
27
character, in the sense that its content is not given by a (propositional) description-like en-
tity. Perceptual conception of bodies is best understood as direct awareness of the objects
perceived. One may raise an objection here: to be able to perceptually conceive substances,
one must have at least very general concepts, like “body”, “solidity”, etc., otherwise one
would not be able to individuate a body and distinguish it from others. So, perceptual con-
ception cannot be non-conceptual, in the way discussed so far. There are several things
one could say in reply to such an objection. First, it should not be too difficult to perceive
where a body ends and another one begins: no two bodies occupy the same space and this
can be apparent, in the absence of any concepts of body. Different sense-modalities which
are employed for acknowledging the presence of material substances in one’s environment
may help distinguish among different bodies. For instance, if someone touches something
with his left hand while also seeing something to his right (while he has his head turned to
the right), that person should be able to notice that he is dealing with two different things.
This should be so, even when that person is not able to conceptualize what kind of things
the two objects are.
Second, even if the same sense-modality is employed, one should still be able to ac-
knowledge the presence of two bodies, even though no concept of body or solidity, is
present to the mind of the perceiver. For instance, if one touches something with a hand
while also eating something (and thus touching it with his tongue) one should be able to
register the difference between the two things, even though one would not be able to de-
scriptively characterize the two objects.
For all these reasons, perceptual conception of bodies should be understood to be non-
conceptual.
28
1.2.2 PerceptualConceptionSelectsPropertiesfromCompleteObjects
Existing substances are complete objects, in the sense that for some given quality Q, that
quality is (or is not) a constituent of a particular substance that one perceptually conceives
on a given occasion. The fact that that quality is (or is not) a constituent of the object is
independent of somebody’s perceptually conceiving it. Reid does not speak of substances
as being complete objects, in this sense. However, he argues that all contingent objects
cannot be known by normal, limited minds, but only by the mind of their creator (EIP,
VII. 1, p. 545). There is more in a substance than meets the eye. The fact that these are
created objects suggests that once created, no other constitutive qualities are going to be
added to them; moreover, each created material substance is a unique individual. Reid
takes this claim of individuality very seriously, arguing that even the qualities belonging to
these substances are individuals, since only individuals are things that exist; universals are
nonexistent, in his view. The following passage supports this interpretation:
To this I answer, that the whiteness of this sheet is one thing, whiteness is
another; the conceptions signified by these two forms of speech are as different
as the expressions: The first signifies an individual quality really existing, and
it is not a general conception, though it be an abstract one; the second signifies
a general conception, which implies no existence, but may be predicated of
every thing that is white, and in the same sense. On this account, if one should
say that the whiteness of this sheet is the whiteness of another sheet, every man
perceives this to be absurd; but when he says both sheets are white, this is true
and perfectly understood [emphasis added] (EIP V . 3, p. 367).
The individuality of the qualities does not have a marked role to play in how we under-
stand the notion of completeness attributed here to Reid; its importance resides in the fact
that it helps distinguish between one object and the next.
To better understand the idea that substances are complete objects, let us think about
them as supporting complete constellations of qualities. We do not have perceptual (or
29
other type of) access to all the qualities that make up a complete set; however, we can con-
ceive of qualities that are not given to us in perception. But it is important to understand that
perceptual conception of substances only enables us access to a limited subset of qualities.
This is reminiscent of Locke’s distinction between nominal and real essences (Essay, III.
iii-v), with the caveat that Reid, in contrast to Locke, thinks that individuals have essences
from which all their qualities flow.
12
The real essence is immutable, complete, and not
discoverable by mere mortals; so, instead, we make do with attributing nominal essences
to individual substances. The latter type of essences are neither immutable nor complete:
they can be changed by either introducing or subtracting qualities from the designated set.
Part of the set that makes up the nominal essence of a substance is constituted by quali-
ties that are made available to us by the perceptual conception of the substance in question.
When we perceive a particular horse, for instance, what we perceive is a complete object,
since a horse is a created individual material substance. However, the perceptual con-
ception that is an ingredient of somebody’s perception will not present to his mind every
quality had by the respective horse. It will not even present every perceivable quality had
by that horse. For instance, the horse under consideration might have a tail that is 40 cm
long, and a perceiver will certainly perceptually conceive the horse as having a tail; but
he will not conceive that tail as being 40 cm long. His perceptual conception is not fitted
to discriminate such details. But, arguably, the length of the horse’s tail is a perceivable
quality of that horse. It may thus be said that perceptual conception selects qualities, in the
sense that it makes salient some properties of the substance perceived and only those are
12
The idea that there is more to a substance than meets the eye, or, more generally, the mind, is also
reminiscent of Locke, in a different way. Locke (Essay, II. xxxi. 1; xxxi. 3 and xxxi. 6) thought that our
ideas of substances are incomplete, and thus inadequate, because there is more in the thing than in the idea;
whereas our ideas of mixed modes, or relations are adequate, and complete, because everything we think of
a mode is in the idea we have of it. On the present interpretation of Reid, nonexistent things (some of which,
at least, would be counted among mixed modes by Locke) are incomplete things, even though, in a Lockean
jargon, one can say that our ideas of nonexistent things are still adequate. But substances are still seen by
Reid as being complete things, and our conceptions of them quite inadequate.
30
presented to the perceiver. Although any constellation of qualities inherent in a substance
is complete, whenever someone perceives the substance, he only perceptually conceives a
subset of those qualities. One would need to have perceptual access to the real essence of
things, in order to be in principle able to perceptually conceive all of their qualities. Even
then, it is doubtful that perceptual conception can present one with all the qualities of a
certain substance: one must not forget that perception is perspectival. It would be thus
counterintuitive to argue that one can perceive and hence perceptually conceive the tail of
a horse, if one just has a frontal view of the respective horse. Perceptual conception of
a substance is quite selective concerning what properties (if any) it makes salient to the
perceiver’s mind. Both this trait and the fact that the objects of perception are complete are
not shared by imaginative conception.
1.3 ImaginativeConception
The structure of imagination is analogous to that of perception, with some qualifications: a
perceptual experience is evoked by a sensation of the external object and involves a type of
conception and a belief that the object conceived exists; an imaginative experience is not
evoked by sensation and it only involves a type of conception, and no belief concerning
the existence of the object conceived. Although structurally the two faculties are quite
similar, this section shows that imaginative conception is different from the perceptual kind.
Whereas the latter is non-conceptual, the former is fully conceptual, i.e. propositional. In
a way, more thinking power is needed for someone to imagine a centaur than just to see a
horse. Furthermore, whereas perceptual conception makes salient some of the properties of
complete objects, imaginative conception “bestows” on the imagined objects the qualities
they are imagined to have. Imagining a nonexistent object logically entails the ascription
of certain known (and conceptualized) qualities to that object. Section §1.3.1 develops
31
the claim that imaginative conception is conceptual, while section §1.3.2 explains how an
imaginer constructs incomplete imagined objects, out of a pre-determined set of qualities.
1.3.1 TheConceptualCharacterofImaginativeConception
Reid dedicates a whole essay to what he calls “conception”, but the issues he addresses
there are different from the ones raised by perceptual conception. Some of the things he
says in Essay IV do not apply equally well to perceptual conception and this indicates that
what he calls “bare conception” is interestingly different. The distinction between percep-
tual and imaginative conception to which this chapter draws attention is best supported by
Essay IV .
Regardless of whether one finds it compelling and helpful to assimilate perceptual con-
ception to Russellian acquaintance, according to Reid (EIP, IV . 1, p. 308-309), the type of
conception employed by imagination is analogous to the type of knowledge one can have
by description, as Russell (1910) suggested. In order to be able to imagine a winged horse,
one must either be acquainted with such an animal, or have the concepts denoting the bits
and pieces that would make up such an animal and a way of putting them together, such that
the result would be a winged horse. Minimally speaking, an imaginer of a winged horse
would need to know what wings are, what horses are and be able to form a conception
of how wings could be attached to a horse. Since winged horses do not exist, our imag-
iner cannot be acquainted with such a beast. Imaginative conception is active in a sense in
which perception is not, namely it must “construct” its objects; hence one must have a prior
grasp of the components from which those objects are constructed. In order to be able to
imagine a certain new (not previously heard) sound, an imaginer would need to be able to
imagine both the pitch and the tone of that sound. Without having any concepts denoting
the two different characteristics of the sound, such an imaginer would be in no position to
32
carry out his act of imagination. One must be able to identify each component in such a
way that it can be attributed to an imagined object; an imaginer must have something like
a concept denoting the components used to construct the imagined object.
A congenitally blind person cannot conceive colors (either perceptually or imagina-
tively), Reid argues, because such a person cannot be acquainted with color. But something
more than just perceptual acquaintance is required for imagination. One must not forget
that Reid thinks that imagination is a certain type of bare, or simple conception (EIP, IV .
1, p. 306). And conceiving an object, be that object an actual object of sense, and not just
a mythological creature, in the absence of any other operation of the mind of which con-
ception might be an ingredient, does not just happen out of the blue, because “conception
of objects is not the first act of the mind about them. External objects are perceived by our
senses before they are simply conceived.”
13
Moreover, “we must have judged or reasoned
before we have the conception or simple apprehension of judgment, or of reasoning” (EIP,
IV . 3, p. 327).
According to Reid then, one must not only be acquainted with the bits and pieces of
what one (imaginatively) conceives, but also be able to separate the conception-component
of perception (or consciousness, or belief / judgment), and reflect on it. This analysis alters
the nature of conception itself: the simple conception is different from the conception-
component employed in perception. What in perception was non-conceptual becomes fully
conceptualized when what Reid calls “conception” (and hence imagination) is exercised
by itself. We should not be mislead by his use of a single term, namely “conception”,
to refer to both an ingredient of all the other operations of our mind, and to a faculty
that can simply present an object to the mind, without expressing any judgment about it.
The process of acquiring a perceptual conception is different from that of having a simple
13
This excerpt is to be found in Aberdeen MS 2131/8/ii/02 and is reproduced here from Nichols (2007, p.
46).
33
conception, including an imaginative one. Whereas perceptual conception just happens in
perception because we are constituted in a certain way, so that our sensations suggest such
conceptions, “simple conceptions are got by analyzing more complex operations” (EIP IV .
3, p. 327). So, simple conceptions, including imaginative conceptions, are different from
perceptual conceptions: to be able to use this type of conception, someone must have a
fully conceptualized understanding of the object conceived.
Although imaginative conception is derived from perceptual conception, the former
can still be different from the latter; this derivation simply means that someone cannot
have a purely imaginative conception of a yellow winged-horse, unless one was previously
acquainted, in perception, with yellow, wings and horses. This does not contradict the thesis
that perceptual conception is non-conceptual: once we analyze the information we have in
perception, we can form concepts about all sorts of things and only when we have those
concepts can we employ imaginative conception, in the way envisaged by Reid. The fact
that we are supposed to know every little detail which we use when we imagine a mythical
creature shows that imaginative conception is conceptual. It is quite counterintuitive to
claim that we are going to imagine a certain centaur without construing that object under
the concept “man-horse“, or something similar. The first difference in character between
the two types of conception should be by now apparent.
1.3.2 Imaginative Conception “Bestows” Qualities on Incomplete Ob-
jects
Although imaginative conception of nonexistent objects must start from known ingredients,
the way of combining those ingredients is entirely up to the imaginer. We can imaginatively
conceive things that do not exist, as clearly and distinctly as we can perceptually conceive
things that do exist. However, it is we who arrange the parts and combine the attributes,
34
in the case of imagined nonexistents. By contrast, nature is responsible for how things
are in the real world, and, according to Reid, we cannot know entirely what powers of
combination it used in putting together the objects that populate the world. It is different
with the objects that are imagined by us, and known to have no existence. According to
Reid, we may “form an endless variety of combinations and compositions, which we call
creatures of the imagination. These may be clearly conceived, though they never existed”
(EIP, IV . 1, p. 310). The interesting question is: how can such nonexistent individuals be
conceived?
By way of reply, let us look at how Parsons (1980) argues that nonexistent objects
are to be conceived, according to Meinong. One may worry that this is anachronistic:
but the claim here is not that the whole Meinongian ontological system is supported by
Reid’s philosophy. But some of the things Reid does say indicate that he might have been
of a Meinongian inclination.
14
This does not mean that Reid thought that there are three
separate levels of existence: existence, subsistence, and nonexistence. Thinking of Reid as
a Meinongian kindred spirit makes sense of his idea that when one imagines a centaur one
does not form an image of a centaur in one’s mind, but one conceives an animal, with a
certain body, with internal life and motion, despite its obvious non-existence (EIP IV . 2, p.
321-22). By imagining a centaur, one understands what it would be like to be in perceptual
contact with such an animal. This does not confer existence or subsistence to a nonexistent
object; it just suggests how one can imagine a centaur, without necessarily bringing an
image to one’s mind.
According to Parsons, in a Meinongian framework, to any non-empty set of proper-
ties there corresponds an object. By employing such a method one can include in one’s
ontology nonexisting, and even impossible objects. For example, the set {goldenness,
14
For more on this issue, see Nichols (2002b).
35
mountainhood} is correlated with a golden mountain, which is an object. There are no
real, existing golden mountains, so this object is nonexistent.
15
Imaginatively conceiving
a centaur proceeds in the following way: an imaginer conceives a certain set of properties
(including, but not limited to, being an animal that is half man and half horse) and thinks
that a certain imaginary object corresponds to it. To emphasize, an imaginer will not be
said to only have a certain set of properties in mind, but also a certain object, “constructed”
out of those properties by his act of imagination. This set of properties is nothing more
than a set constituted by attributes that Reid says are necessary for imagining things that
have no existence.
If we look at what Reid claims to happen when one imagines a centaur, this sugges-
tion takes a clearer shape: in imaginatively conceiving a centaur, an imaginer puts certain
properties together in such a way that the result is a certain nonexistent animal with human
head and torso, and with the body of a horse:
This one object which I conceive, is not the image of an animal, it is an animal.
I know what it is to conceive an image of an animal, and what it is to conceive
an animal; and I can distinguish the one of these from the other [...]. The thing
I conceive is a body of a certain figure and color, having life and spontaneous
motion (EIP, IV .2, p. 321-22).
The mythical animal imagined in this way is a certain individual; one does not just
imagine the property of centaurhood, in general, as Gallie (1997, p. 320) thought. Reid is
explicit on this issue: one imaginatively conceives a body with certain specific attributes.
Moreover, imagining a certain centaur is not problematic: one way this can be done is
by combining in one’s imagination certain individual attributes, with which the imaginer
is acquainted via perception. For instance, someone might imagine a centaur by thinking
that it has the body of his neighbor’s horse and the torso of his neighbor. These bits of
15
Parsons (1980, p. 18-19). The set {roundness, squareness} is correlated with an impossible object. This
chapter will not discuss the difference between possible and impossible nonexistent objects.
36
bodies do exist in reality and belong to individuals (and hence are themselves individuals).
By combining two individuals in this manner, the resultant thing should be an individual
itself – albeit a nonexistent one. This is not to say that the universal centaurhood cannot
be conceived; it can, but someone conceiving it will not be said to imagine anything. Reid
argues that universals cannot be imagined, since imagination is related to objects of sense,
and universals are not objects of sense, but he thinks that they can be conceived, without
being thought to exist or not.
16
According to Reid, imaginatively conceiving a centaur is
not like conceiving a triangle, namely a universal, as Gallie (1997, p. 321) argues.
One may ask certain questions here about what this imaginative conception of an in-
dividual centaur might look like: for instance, will the imagined centaur have a particular
color, be of a particular height, etc.? There is nothing problematic in answering these
questions in the manner suggested above: initially, the imaginative conception of a cen-
taur is less distinct and it can be made clearer by adding more individual properties to the
set of properties with which one begins. At this point, someone might worry that adding
more specifications to a certain conception will not make that conception clearer, but it will
change it, so that the imaginer has a different conception all together. This worry might be
addressed in either of the following ways: (i) one might argue that the initial conception
is the same as the richer one, as it was already suggested; or (ii) one might argue that the
initial conception is different from the richer one. Identity conditions for conceptions are
not that clear, and making them so would take us too far from the concerns of the present
chapter. However, there are several things to say here: we use imagination when we inter-
pret stories about mythological creatures, for instance centaurs. Some of these stories even
name their native centaurs and once such a name is introduced, it is used throughout the
story to refer to the same centaur. Think about the following situation: at the beginning
16
Reid talks about the distinction between imagining a particular and conceiving a universal in EIP, V . 6,
p. 394.
37
of the story we are told that Flane is a brave centaur; by the end of the story, we will have
learned that Flane has brown hair and green eyes, that he is the friend of Miradora, and
uses a red bow to shoot his human enemies. If we set score by (ii) above, we would have to
say that the conception of Flane we had at the beginning of the story is different from the
one we had at its end. This is not how we talk about fictional objects, though. According
to Reid, this should count as reason enough to prefer (i) to (ii).
17
The issue is, of course, more complicated than this. More importantly, however, it is
not peculiar to this theory of imagination, or to this kind of theory of fictional characters,
as Parsons (1980, p. 190) correctly notes. According to him, a similar issue can be raised
about set theory, more generally and even about the meaning of scientific terms. Indeed, it
is not altogether clear when a certain scientific term entirely changes its meaning and when
it simply gains a richer one, while its content remains the same.
18
Even if we accept (i) and agree that a certain conception becomes clearer as the imaginer
specifies more attributes the centaur imagined is supposed to have, one may raise another
problem. Earlier, it was claimed that nonexistent objects are incomplete, in the sense that,
for some property P, it is undecided whether a certain object has it or not. The worry
then is this: just how incomplete are these nonexistents? If an agent engaged in a certain
imaginative act makes his initial conception clearer and clearer, by specifying more and
more attributes that object might have, will not that imaginer make the object in question
eventually complete? The simple answer is to say that such a situation might indeed occur.
17
For more on the issue concerning how a name can be introduced to refer to something that does not yet
exist and continues to refer to that same thing even after it starts to exist, see Jeshion (2010b, p. 116-117).
Reid comes close to this issue when he talks about how someone “may conceive a machine that never existed”
(EIP V . 4, p. 375). If somebody conceiving such a machine were to build it, and name it, the name would
refer to the conception and to the actual object, once it is finished. The issue concerning centaurs is different,
since they will never exist, but sufficiently related to see that if we use names to talk about them, and Reid
argues that we do, there is not much to prevent us from preferring (i) over (ii).
18
For more on this issue, see the discussion in Parsons (1980) and the works he cites, which are classics
concerning the issue of meaning change: Field (1973) and Parsons (1975).
38
But this is not troublesome, since it was not claimed that nonexistent objects are necessarily
incomplete. Moreover, this actually helps the case this chapter makes, by emphasizing that
it is the imaginer’s action that completes the object; it is not given to that imaginer as
complete, from the beginning.
A more sophisticated answer takes into account the great difficulty raised by such an
aim. Nonexisting objects are, in principle, completable, but this does not mean that it is
easy to do so. Nor does it mean that this task only requires a little bit of concentration on
the part of the imaginer. On the contrary, it requires quite a lot and it is unclear that such a
task can be achieved in an ordinary lifetime. To see why this is so, let us think about what
ordinarily happens when someone actively imagines something. According to the theory
developed here, an imaginer will be engaged in an act of imagination for a while, then
he will probably return to more pressing issues. While the imaginer is thus engaged, he
will not be concerned to specify all the attributes of the centaur he is imagining. He will
probably think about it as having a particular shape, size, maybe even overall color, but he
will not think about how many hairs the centaur has on his head, or what color his bow
and arrows are or even whether his stomach is more human-like than horse-like. If we are
only interested in offering a theory which explains how imagination actually works, then
the worry above is not that worrisome; nonexistent objects are mainly incomplete, maybe
completable, but this is of no consequence. The moral of the story is the same: there is
an important distinction between the objects of perception and those of imagination. The
latter are incomplete or, at least, not as complete as the former.
However, let us think, for the sake of the argument, what might happen if one took it
upon himself to specify everything that can be specified about a particular centaur, thus
completing the conception and the nonexistent object which corresponds to it. What would
such an exercise entail? Our imaginer will have to specify not only what color the centaur’s
body has, the size of his body, the size, shape and color of his stomach, etc., but also, for
39
every second of a day in the “life” of that centaur, what type of food he ate, how much, what
internal processes his cells underwent, etc. If we think that to change but one such quality
of the centaur means changing the world of which he is a part, because every object in a
world is the product of that world, then the imaginer will have to sit down and describe the
possible world to which this centaur belongs, in such great detail, that it will take him ages
to do so. Moreover, it is not only a question of not having enough time to do this, but also a
question of what power of computing our imaginer’s mind must have. This task is akin to
that performed by God for every created thing in the world. So, yes, a nonexistent object
is not necessarily incomplete; but to successfully complete it, one must be very much like
God, and no mortal human-beings are like that. So, from a human-being’s point of view,
most of the objects of imagination are incomplete.
19
This difference in the nature of existent and nonexistent objects, namely that the former
are complete, whereas the latter are (for the most part) incomplete, is the basis for the sec-
ond difference between perceptual and imaginative conception. In perceptually conceiving
existent objects, one must roughly know only that this body is different from that one. Per-
ceptual conception makes certain properties of objects salient to the mind. These properties
may be very general ones, such as recognizing that something is a body, or that something
is solid. But in imaginatively conceiving a nonexistent object, one must know something
more specific about that object, since otherwise it would make no sense to say that this
nonexistent object is different from that one – there is no way of individuating them in
space and time, since they do not exist. The only way of distinguishing between Magorian
and Bane of the Harry Potter novels is by refining each imaginative conception such that
different properties are included in each set, corresponding to each of the two nonexistent
19
This qualification is needed because there are very simple objects that can be imagined completely,
without requiring so much from the imaginer. For instance, imagining an electron coming into existence
and being destroyed after just one second will require much less from an imaginer than imagining a centaur
killing a human being in a one on one fight.
40
objects.
20
Moreover, each of these two objects is nothing it is not conceived to be, and it
is something inasmuch as we think of attributing certain qualities to it. Nature does not
interfere with the process of combining and recombining these qualities. Imaginative con-
ception bestows properties on the objects of imagination, and it is up to the imaginer to
make such an object more or less complete. This is the third way of distinguishing between
perceptual and imaginative conception.
1.4 Conclusion
This chapter presented a distinction which may be implicit in EIP, but is not discussed as
such by Reid. There is a sharp difference between the conception employed in perception
and the one employed in imagination. The arguments discussed here showed that there are
three ways in which perceptual conception differs from imaginative conception. First and
foremost, perceptual conception does not essentially proceed by way of concepts, i.e. it
is non-conceptual and not propositional, whereas imaginative conception is entirely con-
ceptual. This distinction is true to the spirit of Reid’s philosophy: for instance, he thinks
that the presupposition of existence is not required for conception and hence conception
is a different faculty from perception. In the same vein, the presupposition of attribution
of a quality to an object, which is necessary for imaginative conception, is sufficient for a
finer-grained distinction between different types of conceptions.
Second, the objects of perceptual conception are complete objects; the objects of imag-
inative conception are incomplete – one can explicitly endow an imagined centaur with
more properties, while one imaginatively conceives it, thus turning it into a more definite
object, in one’s imagination. But this is not true in the case of the objects of perception –
20
Such a differentiation works on the assumption that something akin to a Lockean principle of individua-
tion is applicable even to imagined objects: no things of the same kind could be co-occurrent.
41
they are what they are, nothing more or less, and perception does not alter this. Third, per-
ceptual conception selects or makes salient properties from the object perceived, whereas
imaginative conception bestows properties on the object imagined.
If the arguments and analyses presented here are correct, this distinction is real, even
though overlooked by Reid. Further research is needed in order to establish what conse-
quences this distinction has on the rest of Reid’s philosophy.
42
Chapter2
PerceptionandItsObjects
2.1 Introduction
It is an open question in the contemporary philosophy of perception whether the human
perceptual system is providing us with representations as of bodies. Some psychologists
and philosophers argue that bodies are not immediately perceived, and that we only per-
ceive their features and then infer that bodies having those features exist. Others argue that
bodies are indeed objects of perception and that no inference is required for perceiving their
shape or the fact that they are bodies.
1
A lot depends on the answer to this question: if bod-
ies are not objects of perception, then is there any reason to believe that singular thought
is formed in perception? De re thought must start somewhere, and if perception does not
represent the world as being segmented into things, one may reasonably argue that we are
not acquainted with things at all. Maybe we do perceive free-floating properties, but bun-
dles of properties are not enough to give us a robust understanding of the world around us.
If all we have are properties and bundles thereof, it becomes very difficult to think of the
1
For more details on this debate, Burge (2010, p. 438-42). Burge argues that body is a perceptual attribu-
tive and that our perceptual apparatus gives us representations as of bodies. On the other side of the debate
is Elisabeth Spelke, whose work proved to be very influential. Burge gives some insightful arguments to
prove that she does not provide enough reasons to think that the representation of body does not originate in
perception. Her original arguments are presented in Spelke (1988).
43
bundling relation as anything other than arbitrary. In other words, without explaining how
bodies are perceived, we are burdened with Hume’s skepticism about the external world.
Thomas Reid is credited with shredding the veil of skepticism to pieces.
2
But the exact
details of this overhaul are still not well settled: there is a tension in the secondary literature,
engendered by a corresponding tension in Reid’s text itself, regarding the perception of
bodies. The problem is even more acute, since this tension has gone largely unnoticed.
One purpose of the present chapter is to show that perception of bodies is indeed possible
on Reid’s view, and to situate his view in the larger context of the debate mentioned above
between psychologists and philosophers of perception.
To do so, one must first note that some scholars interpret Reid as saying that only pri-
mary qualities of bodies can be perceived. Secondary qualities and bodies are not objects of
perception proper; at most they can be objects of acquired perception.
3
Other scholars just
take it for granted that we can (originally) perceive material substances, without noticing
that they are attributing inconsistent views to Reid, and without discussing either the way
this would fit into his theory, or what problems this position might raise for Reid’s theory of
perception.
4
To address this tension, this chapter discusses several arguments showing that
bodies can indeed be objects of perception proper, on Reid’s view. This result is beneficial:
it arms Reid with an answer to certain types of skeptical challenges, as shown in §2.3.4.
Throughout this chapter, it will be assumed that the objects of original perception are
immediate objects of perception. Any discussion concerning the immediate objects of per-
ception must start with an explanation of the relevant notion of immediacy. According to
2
For a good exposition of Reid’s reaction to skepticism see Lehrer (1989c).
3
Such an interpretation can be found in Lehrer & J. C. Smith (1985).
4
This category is represented, among others, by Buras (2002) and Copenhaver (2000). Interestingly,
Copenhaver (2000, p. 20), cites one problematic passage for the view that bodies are objects of perception,
but the focus of her article is different and thus she takes it for granted that, for Reid, the phrase “objects of
perception” denotes bodies.
44
Buras (2008), there are three senses of immediacy that Reid employs in his work. First,
something is said to be immediately perceived or known if it is not known in virtue of
a propositional inference. Thus, one can know that a mathematical axiom is true immedi-
ately, by just contemplating it; whereas one only knows that a mathematical theorem is true
by having proved that it follows from one or more axioms. The knowledge of the theorem
is mediated by the inference needed to reach the conclusion.
Second, something is said to be immediately perceived if there are no mental entities,
e. g. ideas, precluding a perceiver to connect with a body.
Third, something is said to be immediately perceived if its perception is direct. For
instance, Reid claims that the notion we have of primary qualities, in perception, is direct,
while the notion we have of secondary qualities is relative or indirect.
For the present purposes, “immediacy” will be taken in the first and second sense,
but not in the third: this chapter argues that something is immediately perceived if that
perception is not based on a propositional inference, and if no mental entities are mediating
between the mind and the external object. On the present interpretation, something can be
indirectly, but immediately perceived: secondary qualities and bodies. As we will see in
§2.2.2, immediacy and directness are seen to come apart by other Reid scholars, as well.
For instance, Copenhaver (2000, p. 18) endorses such a position. Moreover, this lack of
directness should not be seen as more worrisome than the fact that even the perception of
primary qualities is not direct, in a certain sense. A certain quality cannot be perceived,
unless a physical impression is made on the organs of sense, and a sensation is then being
felt. Only then, will the sensation suggest a conception and belief in the existence of the
quality in question. In this sense, the perception of the quality in question is indirect.
5
There is a reason why interpreters have paid less attention to this issue and have offered
5
According to Van Cleve (2004, 114-19), this chain must be in place for the perception of both primary
and secondary qualities.
45
interpretations that are not always consistent with each other: Reid is not explicit with
regards to how bodies or their qualities can be immediately and originally perceived. Still,
the logical space is restricted by the text to three possibilities.
First, physical qualities could be said to be the only immediate objects of perception,
while the bodies to which they belong are only (at most) inferred to exist.
6
Henceforth, this
will be called “the qualities-only view”.
7
Its main disadvantage is that, since bodies are not
immediate objects of perception, we do not have a simple reply to the skeptic doubting our
knowledge of bodies.
Second, logic allows for the possibility that only bodies are immediate objects of per-
ception. On this view, a body is first perceived and then, by an inferential process, certain
qualities are attributed to it. This view is only apparently supported by some passages in
Reid. But it is not a good view to have. Usually, bodies are distinguished by the qualities
they have. On this view, this is not possible: they are first perceived as distinguished ob-
jects, and then one attributes some qualities to them; but how is the initial process taking
place, in the absence of perception of qualities? This view is mentioned here only for the
sake of completeness, and it will be very briefly discussed in what follows, under the name
of “the bodies-only view”.
Third, the immediate objects of perception could be both qualities and the bodies to
which they belong. This view can come in many flavors; I will defend one of them ac-
cording to which qualities and the bodies to which they belong are not simultaneously
perceived: qualities are the first to be taken in by the perceiver, and then the body itself.
However, the perception of the body is still immediate, in the relevant sense: the existence
6
On this view, matter is not immediately perceived, although it is known to exist, in virtue of the reliability
of perception and the truth of the relevant first principles. Since qualities can be seen as its modifications,
perceiving them would be enough to somehow connect our mind with matter itself. I will clarify this point
when the time comes.
7
This view is quite reminiscent of Locke, once we acknowledge that qualities are sometimes called “ideas”
and that they are considered to be the immediate objects of perception, in Locke’s philosophy.
46
of the substance is not propositionally inferred; it is automatically known, by perception,
and there are no mental intermediaries (i.e. ideas) that preclude the perceiver from connect-
ing with the body in question.This will be called “the qualities-and-bodies view.”
There is not enough textual evidence to adjudicate between the first and the third possi-
bilities: the majority of passages support the qualities-only view. In other places, however,
Reid argues that bodies are also perceived; and at least one passage may be interpreted as
supporting the bodies-only view. Although this last interpretation does not have a lot of
back up, the fact that some passages support the first view, while others support the third
view remains; Reid is genuinely inconsistent regarding this issue. To address this incon-
sistency, this chapter will look at the broader context, and conclude that there are strong
reasons, internal to Reid’s philosophy, for thinking that both qualities of bodies and the
bodies to which they belong are objects of perception. Although the text is vexed, and pas-
sage by passage, it cannot be established what was Reid’s view on this issue, the qualities-
and-bodies view is the one he should have adopted, given his larger epistemological and
metaphysical concerns.
Matters are not so simple, however. Bodies and qualities are very different categories of
things, so can we say that we perceive both in exactly the same way? Aren’t we somehow
losing the ‘unity’ of the faculty of perception? I acknowledge these as hard question,
and answering them is complicated, but I think I can provide a good explanation of how
everything fits together. Furthermore, we should not be troubled by the fact that Reid does
not explicitly discuss this view; since it allows us to rule out a possible inconsistency,
attributing the third view to Reid is the most charitable reading of his works.
This chapter has the following structure. The next section discusses the representative
passages talking about the immediate objects of perception, from Reid’s 1764 An Inquiry
into the Human Mind on the Principles of Common sense, and his later 1785 book, Essays
on the Intellectual Powers of Man. We will see that the problem with Reid’s characteriza-
47
tion of the objects of perception is not that he changed his mind from one book to the next;
the problem is that he is not consistent even within each book. The last section presents
the reasons supporting the qualities-and-bodies view over both the qualities-only and the
bodies-only views. They are based on Reid’s philosophical system as a whole, and indi-
cate that he should have consistently maintained that qualities of bodies, and the bodies
themselves, are immediate objects of perception. This chapter argues that Reidian percep-
tion is best understood as a two-stage process: perception of an external quality leads the
perceiver to perception of the material substance to which the quality belongs.
2.2 ImmediateObjectsofPerception–TextualEvidence
In both of his books, Reid offers an analysis that uncovers some of the necessary ingredients
of perception. There are two official characterizations, which are almost identical, but for
an important detail.
In IHM, Reid argues that:
[. . . ] the perception of an object implies both a conception of its form, and a
belief of its present existence. I know moreover that this belief is not the effect
of argumentation and reasoning; it is the immediate effect of my constitution
[emphasis added]. (IHM 6. 20, p. 168, 32-6)
Whereas in EIP, he states:
If, therefore, we attend to that act of our mind which we call the perception
of an external object of sense, we shall find in it these three things. First,
Some conception or notion of the object perceived. Secondly, A strong and
irresistible conviction and belief of its present existence. And, thirdly, That this
conviction and belief are immediate, and not the effect of reasoning [original
emphases]. (EIP II. 5, p. 96)
Two other faculties are essentially involved in every perceptual experience: it is nec-
essary (but not sufficient) that a conception of the object and a belief in its present exis-
48
tence are simultaneously occurring. Both conception and belief play an important role in
establishing whether a certain experience should be classified as either illusion or hallu-
cination, rather than perception. Something counts as a perceptual experience only if its
belief component is actually true; if it is false, that experience will be either an illusion or
a hallucination, depending on further issues.
8
The discrepancy between the two characterizations is this: in the passage from IHM,
Reid writes that the perception of an object involves “a conception of its form”, whereas in
the passage from EIP, he says that it comprises “a conception of the object perceived.”
The characterization of perception from IHM suggests that Reid supports the qualities-
and-bodies view, although the passage does not conclusively rule out the bodies-only view.
The second passage expresses the following thesis: in a perceptual experience, the object
of perception must coincide with the object of conception. The first passage does not say
anything about this, but it makes sense to think that one cannot believe that a certain object
can exist without having some conception of that object. In EIP, Reid argues that no belief
can exist in the absence of a conception of the thing believed, so that conception is always
necessary for belief (EIP II. 20, p. 228). There is no specific place in IHM where he
endorses something similar; however it is intuitive to think that a conception of a thing is
required for a belief about it. Thus, if the object of conception differs from the object of the
existential belief, there is no sense in which this conception is necessary for the occurrence
of a perceptual experience. The conception of the object of the existential belief would be,
in turn, the one required for perception. So, a similar thesis has to be at work in IHM. On
this assumption, there are three ways to unpack the passage from IHM.
First, suppose that ‘object of perception’ denoted ‘quality of substance’ only. We would
8
Nichols (2007), at p. 34 comes close to understanding Reidian perception in this way, but he does not
specify that the belief in existence involved in perception is key to classifying a certain experience as percep-
tion or not.
49
then need to explain away the following unnatural paraphrase of the respective passage:
the perception of a quality implies both a conception of the quality’s form and a belief in
its present existence. This is not good: the phrase ‘form of an object’ is probably best
understood to denote its figure. Qualities do not have figures; at most, some are figures.
Bodies have figures or shapes, and any other kinds of qualities that make up the form of
the body, if ‘form’ is used in loose Aristotelian fashion. So this passage does not favor the
qualities-only view.
Second, one might think that this passage indicates that only bodies are immediate ob-
jects of perception, and their qualities are inferred to exist, since it talks about the object
perceived, the conception of its form (i.e. qualities), and a belief in its existence. This
reading is problematic: it collides with the thesis regarding the necessary identity between
the object of perceptual conception and that of perceptual belief. Any kind of belief pre-
supposes a conception of the object that belief is about. The conception of the qualities of
an object can be presupposed by an existential belief about those qualities, but not by an
existential belief about the object itself. We end up with three components here, if we want
to read this passage as supporting the bodies-only view: (i) a conception of the form of the
object, (ii) a conception of the object, and (iii) a belief in the existence of the object. The
conception of the form does not play any role in the make-up of the existential belief about
the object, if ‘object’ is understood to mean body. Moreover, it is quite a mystery where
this conception of the object qua body comes from.
One can, instead, offer the following reading, that takes into account the aforementioned
thesis and makes sense when ‘form’ is understood to mean figure or collection of qualities:
the perception of a body and (some of) its qualities implies both a conception of the body
and its form, and a belief in the present existence of the body and (some of) its qualities.
This reading is not subject to the same objection as the first or second readings: one does
not simply perceive qualities, all by themselves, or just bodies; one perceives the body
50
together with its qualities.
Further indication that this passage is not to be understood as supporting the qualities-
only view is provided by its location, in the text of IHM. Just a few lines above this passage,
Reid is trying to establish that in perception, the object perceived is always distinct from
the operation it is perceived by. And to exemplify, he speaks of ‘a tree that grows before
my window; there is here an object which is perceived, and an act of the mind by which it
is perceived’ (IHM, p. 168, lines 19–21.) There is no reason to think that he changes the
subject mid-sentence, and suddenly starts talking about qualities. This passage then is best
understood as saying that the immediate objects of perception are material bodies, together
with their qualities.
By contrast to IHM, the formulation in EIP does not indicate whether Reid favored
the qualities-only view, the bodies-only view, or the qualities-and-bodies view. Here ‘ob-
ject’ may mean any of the three and in what follows I argue that ‘object’ should stand for
‘qualities-and-body.’ To fully address this issue, we need to go beyond the official charac-
terization of either IHM or EIP.
2.2.1 EvidenceThatQualitiesAreImmediateObjectsofPerception
The passages discussed in this section favor the qualities-only view. that qualities of ma-
terial substances are among the immediate objects of perception. Some passages quite
unequivocally support the qualities-only view, while others lend themselves, with a little
bit of imagination, to the reading that the qualities-and-bodies view is the correct one.
In IHM, Reid does not say a lot about what are the objects of perception. But, in the
course of analyzing the sense of sight, while talking about the phenomenon of acquired
perception, he writes as if the things perceived, both originally and acquiredly, are just
qualities of bodies.
The contrast Reid draws between original and acquired perception is not the focus of
51
the present investigation, but it is worth a brief discussion, because it will help us better
understand what can be said to be immediate in a perceptual experience. According to Reid,
some things can be perceived originally – everyone is born with a capacity to touch objects
and perceive their shape, for instance. Other things need experience and learning, in order
to be perceived. For instance, a shepherd will learn to perceive by sight the weight of his
sheep, and an art historian to distinguish by visual perception a Titian from a Rembrandt.
Reid calls the latter cases acquired perceptions, because, by habit, one acquires a certain
ability with which one was not born: to associate certain perceptions with others. The art
historian learns to associate a certain disposition of colors and shapes with a style, and to
recognize that style as characterizing Titian.
Scholars disagree whether to think that acquired perception is a perceptual phenomenon,
rather than an inferential one. Copenhaver (2010) offers a compelling analysis of acquired
perception, arguing that it does not lack any of the features that perception proper has.
Van Cleve (2004), on the other hand, argues that acquired perception is not perception
proper. The present interpretation sides with Van Cleve (2004): at least in one place, in his
EIP, Reid himself argues that acquired perception is not perception proper:
Acquired perception is not properly the testimony of those senses which God
hath given us, but a conclusion drawn from what the senses testify. [. . . ] I
have therefore given the name of acquired perception to such conclusions, to
distinguish them from what is naturally, originally, and immediately testified
by our senses. (EIP II. 22, p. 247)
Another remarkable fact is that at least one secondary external quality, namely color, is
said by Reid to be originally perceived. This goes against the interpretation offered by some
scholars, according to which secondary qualities are all objects of acquired perception.
Here is what Reid says: “By this sense [sight] we perceive originally the visible figure and
color of bodies only, and their visible place” (IHM 6. 20, p. 171). This passage and others
52
like it are not discussed by the scholars arguing that secondary qualities are only objects of
acquired perception.
9
Whether one thinks that cases of acquired perception are to be counted as real percep-
tion is a substantive issue. This chapter focuses on establishing what counts as objects of
perception, in the sense of being immediately and originally perceived. If something turns
out to be an object of acquired perception, it will not count as an immediate object of orig-
inal perception, and there are good reasons to believe that it will not count as an immediate
object of perception at all, since reasoning is involved in this process. Reid does not always
use the full phrase ‘immediate objects of original perception’; but he should be understood
as talking about such objects whenever he discusses the objects of perception. This is be-
cause perception must be original, in order to count as perception at all, according to the
passage quoted above, and original perception is immediate, in Reid’s view, as seen from
the passages where he offers the “official” characterization of perception (IHM 6. 20, p.
168, and EIP II. 5, p. 96). For this to be so, its objects must be immediately perceived –
this is the sense in which its objects are immediate.
Going back to the main issue, this is the central place in IHM, where Reid discusses the
perception of the qualities of bodies:
Our perceptions are of two kinds: some are natural and original, others ac-
quired, and the fruit of experience. When I perceive that this is the taste of
cyder, that of brandy; that this is the smell of an apple, that of an orange; [...]
these perceptions, and others of the same kind, are not original, they are ac-
quired. But the perception which I have by touch, of the hardness and softness
of bodies, of their extension, figure and motion, is not acquired, it is original
(IHM 6. 20, p. 171).
This passage implies that the things immediately perceived are either primary or sec-
9
This is true of Lehrer & J. C. Smith (1985, p. 27) and of Buras (2009). Nichols (2007, p. 224) acknowl-
edges that color might be problematic for this view, but he does not offer any additional explanation of why
this secondary quality is special.
53
ondary qualities of bodies, and not bodies. Some matters are left unclear: what exactly is
original, and what is acquired, in the perception of the taste of cider? There is a quality
belonging to a certain substance, that we call its taste; we take notice of it by tasting; and
then we learn to associate that taste with that particular substance. We learn, in other words,
that a particular taste is the taste of cider. What is crucial for the present purposes is that
perceptual experiences start from tastes, smells, qualities of hardness and softness, under-
stood as external qualities of bodies. We are not told whether bodies enter the picture at the
perceptual level, or at a later, inferential level. Going just by the evidence presented by this
passage, the qualities-only view seems to fare better than the qualities-and-bodies view.
Nonetheless, I want to stress that the qualities of bodies, and not the bodies themselves are
the focus of this passage. I take this to suggest that even if the qualities-and-bodies view
turns out to be plausible, the qualities of bodies are still quite special, from a perceptual
point of view: the body they belong to cannot be perceived unless they first give rise to
certain sensations to the perceiver. This is not to say that bodies are not immediate objects
of perception; after all, no mental entity is required to exist for an agent to perceptually
experience a body. However, the perception of qualities comes at an earlier stage (concep-
tually): the body they belong to is taken in by the perceptual faculty after the perceiver gets
a hold of its qualities. This is my conjecture, at least; so far, Reid’s text has only provided
us with some reasons to believe that qualities are, in some sense, more special than bodies.
In the end, IHM is not explicit regarding which of the qualities view Reid favored: both
qualities-only and qualities-and-bodies seem plausible. The bodies-only view, on the other
hand, seems to be excluded, at this point.
The text of EIP does not provide an unequivocal answer to the question regarding
the immediate objects of perception either. The qualities-only view is well represented
throughout the book; and Reid seems to be as his clearest when he argues in favor of this
view.
54
In the opening essay of the book, Reid argues that:
All the things which we immediately perceive by our senses, and all the things
we are conscious of, are things which must be in something else as their sub-
ject. Thus by my senses, I perceive figure, colour, hardness softness, motion,
resistance, and such like things. But these are qualities, and must necessarily
be in something that is figured, coloured, hard or soft, that moves, or resist. It
is not to these qualities, but to that which is the subject of them, that we give
the name of body (EIP I. 2, p. 43.)
Qualities of bodies are the only things one perceives immediately, according to this
passage. The senses provide us with information about qualities of bodies, as external
from our minds. By a first principle of common sense, we then know that bodies having
those qualities must exist.
10
This much is clear: whenever we perceive a quality we know
that a body having that quality exists. It is not clear that we know that that body exists
by perception; if anything, one might think that reasoning is involved and helps the first
principle deliver the result that bodies exist too. In §2.2.2, it will be argued that no inference
is drawn here; the first principle has a different role, which does not rule out the possibility
of bodies being perceived, as well as their qualities. But this passage tilts the scales towards
the qualities-only view. Everything else, including bodies, must be things that we either do
not perceive, or we perceive non-immediately.
The belief of the existence of bodies that we naturally form by perceiving their qualities
raises some interesting issues. As previously discussed (in connection with the passage
from IHM, p. 168), any belief presupposes a conception of the thing believed; in this case,
a belief of the existence of a body, that is the substance all its perceived qualities inhere
in, presupposes a conception of the body itself. So, two of the necessary conditions for
perception are met, and thus Reid is faced with a challenge: he has to explain why we
should not include bodies among the (immediate) objects of perception. What is it about
10
Reid argues that attributes cannot exist by themselves, and that they need a subject in which to inhere;
see EIP I. 2, p. 43 and VI. 6, p. 495.
55
the relation between a person and a material substance, which makes that relation non-
perceptual? Why should only the relation between a mind and a physical quality be thought
as perceptual, as Reid seems to be suggesting here? I argue that by Reid’s own lights there
is no reason to exclude bodies from the things that can actually be perceived. This passage
is a bit misleading, in this respect. This is even more so, since it would actually be quite
useful for Reid to argue that bodies qua bodies (and not as constellations of qualities) can
be perceived. Such an argument would help him deal with the skeptic (real or imaginary),
who argues that it is not enough to be certain of the existence of qualities, in order to be
certain that what we take to be the material world actually is the way we think it is; we
need to also be certain of the existence of the bodies the qualities perceived belong to. For
these purposes, it is better to have immediate and direct access to material substances; and
in §2.3 I explain how this immediate access to bodies is possible.
11
The next relevant passage is the opening sentence of the section titled “The Objects
of Perception; and first, Of primary and secondary Qualities”, in the second essay of EIP:
“The objects of perception are the various qualities of bodies” (EIP II. 17, p. 200). These
must be understood as being immediate objects of perception, as was shown above. If Reid
had wanted to indicate that other kinds of things besides qualities of substances are to be
counted as objects of perception, he would have said something like “qualities of bodies
are among the things perceived”. And then it would not be unreasonable to expect a more
or less detailed discussion of what is interesting about the other types of things that can be
objects of this faculty. However, such a discussion is nowhere to be found in the rest of that
section.
11
This discussion is related to (but importantly different from) what Copenhaver (2004) argues concerning
how, on Reid’s theory, sensations are supposed to mediate between the objects of perception and the perceiver.
My interpretation does not conflict with hers: the issue I’m raising concerns the immediacy of perception
of bodies. If Copenhaver’s interpretation is correct, it turns out that for Reid, perception is mediated by
sensation, irrespective of whether its objects are qualities or bodies.
56
Moreover, the next section (EIP II. 18) extends the list of things that can be immediate
objects of perception, but that list contains only qualities, too:
Besides primary qualities and secondary qualities of bodies, there are many
other immediate objects of perception [emphasis added.] Without pretending
to a complete enumeration, I think they mostly fall under one or other of the
following classes. 1st, Certain states or conditions of our own bodies. 2d,
Mechanical powers or forces. 3d, Chemical powers. 4th, Medical powers or
virtues. 5th, Vegetable and animal powers (EIP II. 18, p. 211.)
12
This passage reinforces the claim that Reid had the immediate objects of perception as
his target, even in the preceding section, dealing with the primary and secondary qualities
of bodies. This list extends the initial one, but bodies are still not on it. We are told that this
list is not exhaustive, but if bodies were to be regarded as immediate objects of perception,
they would not have been left out, since they are more important than some of the things
Reid does include on this list.
13
If we are setting score by these three passages, which are unequivocally addressing
the nature of the immediate objects of perception, bodies are more like an after-thought
12
This talk of powers should not be confusing: these powers are nothing more than occult qualities – a
class to which even secondary qualities are seen to belong. Reid is here continuing the Lockean tradition.
13
This passage suggests that not only primary qualities, but also secondary ones and other occult qualities
are immediate objects of perception, regardless of the fact that one cannot have direct, but only relative
notions of secondary and other occult qualities. So, I would like to note that the notion of immediacy and that
of directness do not necessarily overlap, for Reid. One can have an immediate perception of something that
one can get only a relative notion of. This is interesting for deciding whether to classify Reid’s philosophy
as direct realism. Reid seems to think of “direct” as the antonym of “relative”. And he thinks that one can
have immediate perceptions of things that we do not have direct notions of (e.g. secondary qualities). So, I
would argue that Reid is not a direct realist, but, at most, an “immediate realist.” This may seem to suggest
that the difference between him and Locke, for example, is not very important, in this respect: since ideas
have a real existence for Locke, and they are the immediate objects of perception (and thought), one could
argue that Locke is an “immediate realist,” as well. But this would be wrong: the sense in which Reid is said
to be a realist is different from that in which Locke would be a realist. It is the thesis that external objects are
immediately perceived, without a mental mediator, that makes Reid a realist, unlike Locke. So, the relevant
notion of immediacy has to do with ‘mental’ immediacy, i. e. there is no other mental entity that is to be
found between the perceiver and the object (Buras (2008).) To emphasize though, the notion of immediacy
and not that of directness is the most important one here.
57
of perception rather than its immediate objects: in order to think about them, one must
perceive something else first, i.e. their qualities.
2.2.2 EvidenceThatBodiesAreImmediateObjectsofPerception
Contrary to the conclusion of the previous section, other textual evidence indicates that,
according to Reid, our senses give us knowledge of bodies. This is why a section of the
Second Essay, dedicated to discussing our external senses, is concerned with what infor-
mation our senses provide about matter and space.
14
According to Reid, our senses give us information about bodies, but we are not told
that we know that bodies exist, by perceiving them, directly. On the contrary, the notion we
have of bodies, in perception, is relative, in the sense that we have a conception of matter
only inasmuch as we have direct conceptions of its qualities. However, just because we
can only have a relative notion of something, it does not mean that we cannot perceive that
thing originally and immediately. A case in point is color, which is a secondary quality, of
which we can only have relative notions, but we perceive it originally (IHM 6. 20, p. 171;
EIP II. 21, p. 236). So, the fact that the notion we have of bodies is relative does not count
against our ability to perceive them originally and immediately.
15
There is an important problem here, however: since one can only have relative notions
of bodies, which are relative to their qualities, it seems clear that our senses inform us
of their qualities first. One way to understand what Reid says about our knowledge of
bodies due to the external senses is to think that perception is a two-stage process: one
14
Reid uses “matter”, “material substance” and “body” interchangeably, as indicated here: “[w]e give the
names of matter, material substance, and body, to the subject of sensible qualities; and it may be asked what
this matter is?” (original emphasis; EIP II. 19, p. 217).
15
It is true that we can only have relative notions of secondary qualities, too. However, Reid is claiming
that secondary qualities are objects of sense, and that they are immediate objects of perception. One can
immediately perceive something that one can only relatively conceive. So, having only relative notions of
bodies should not preclude us from immediately perceiving them.
58
perceives qualities and, through them, one also perceives the body. However, this process
is still immediate, in the relevant sense: the qualities of bodies, and not mental entities
representing them, are what is necessary for someone to perceive a body.
This point is the important one: since the qualities a body has are external physical
entities, not mental ones, Reid’s system does not collapse into an unsophisticated idealism.
As Copenhaver (2000, p. 18) observes, Reid’s opposition to mediation refers only to the
implausibility of the existence of “a mediating mental entity that represents external objects
in virtue of some intrinsic quality possessed by that entity”. Other mediating physical
factors, such as the medium through which the relevant information travels, for instance,
do not interfere with the immediacy of the perceptual experience. With respect to the
immediacy of the perception of bodies, qualities of objects should be regarded in the same
way as the rays of light.
Moreover, reasoning is not involved in this process: it is not in virtue of a perceiver’s
thinking that he knows a body having a certain color exists, whenever he sees that color.
It is in virtue of our constitution that we perceive bodies whenever we perceive qualities.
The first principle Reid talks about in connection with the perception of bodies, like other
first principles of contingent truths, has the role of an explanatory rule: it just is a natural
law natural by which the human mind functions.
16
This first principle is not something to
which a perceiver appeals, as one would to a step or a rule, in an inference, whenever he
perceives a quality, in order to believe that a body having that quality exists. In this sense,
the perception of bodies is immediate, because it is not based on an inference.
Bearing this in mind, let us look at the relevant passages:
The objects of sense we have hitherto considered are qualities. But qualities
must have a subject. We give the names of matter, material substance and
16
This interpretation is developed in Lehrer (1989c).
59
body, to the subject of sensible qualities, and it may be asked, what this matter
is (EIP II. 19, p. 217.)
Although the passage starts out as if Reid wanted to say that there are other objects
of sense, besides qualities, it soon moves away from this question to that of investigating
the nature of material substance. This quote does not settle the issue, but other passages
indicate that bodies are not objects of sense, in the same way as their qualities: the qualities
of bodies need to facilitate the perception of the bodies themselves.
17
I perceive in a billiard ball, figure, colour, and motion; but the ball is not figure,
nor is it colour, nor motion, nor all these taken together; it is something that
has figure, and colour, and motion. This is a dictate of Nature, and the belief of
all mankind. As to the nature of this something, I am afraid we can give little
account of it, but that it has the qualities which our senses discover (EIP II. 19,
p. 217–18.)
We are said to perceive color, figure, etc, in a billiard ball. It is difficult to understand
how this is possible without perceiving the billiard ball itself. The qualities of a body need
a body to belong to; they cannot exist by themselves. But without perceiving the body they
belong to, how can we know to which body they belong? In this particular case, it is a
dictate of nature that the perceived qualities belong to a material substance, but it cannot be
a dictate of nature that they belonged to a billiard ball, instead of something else. Billiard
balls are not natural substances, so the attribution of such-and-such qualities to this type
of bodies cannot be explained by invoking natural laws. This passage indicates that the
qualities-and-bodies view is better suited to explain how different pieces of the puzzle are
held together in perception.
The following passage reinforces this interpretation.
17
When Reid talks about space, in the same section 19, of the second essay, he makes it clear that space
is not an object of sense. We might be inclined to think that if he wanted to say something similar about
matter, he would have been explicit about it. But this issue is not crucial to my thesis: I do not claim that
Reid thought that matter is not an object of sense, just that it cannot be so in the absence of any perception of
its qualities.
60
It seems therefore to be a judgment of nature, that the things immediately per-
ceived are qualities, which must belong to a subject; and all the information
that our senses give us about this subject, is, that it is that to which such quali-
ties belong. (EIP II. 19, p. 218)
We are told that we immediately perceive qualities, and that they need a subject in which
to inhere, which cannot be anything but a body. Since no quality can exist by itself and since
perception tells us that qualities exist, their subjects must exist too. Reid thinks that we do
not reach this conclusion by reasoning: it is a dictate of nature, and immediately available
to everyone undergoing a perceptual experience. By our constitution, we understand that
this is how things are. So, the belief in the existence of a material substance, supporting
the qualities thus perceived, is as immediate as the belief in the existence of the material
qualities themselves.
Thus, by perception of qualities, we have a relative notion of the subject to which they
belong, and an immediate belief in the existence of this subject. Two necessary conditions
for perception are therefore met for body, in a similar manner as they are met for the sec-
ondary quality of color.
18
There is no reason to discount the idea that bodies are perceived,
any more than there is for the idea that color is perceived.
One last piece of indirect evidence that bodies, as well as their qualities, are immediate
objects of sense is the following passage:
I think it requires some ripeness of understanding to distinguish the qualities
of a body from the body. Perhaps this distinction is not made by brutes, nor
by infants; and if one thinks that this distinction is not made by our senses, but
by some other power of the mind, I will not dispute this point, provided it be
granted, that men, when their faculties are ripe, have a natural conviction, that
sensible qualities cannot exist by themselves without some subject to which
they belong (EIP II. 19, p. 219.)
18
These two are necessary conditions for perception, according to both IHM 6. 20, p. 168, and EIP II. 5,
p. 96.
61
Reid argues that when our faculties are still not fully developed, our minds do not
draw any distinction between qualities and the bodies that have them. But if they were
not objects of the same faculty, how could we jumble them together? He must think that
there are two types of things that can be immediate objects of perception, and they are
objects of perception in the same way, in our infancy. On the one hand, there are qualities
of bodies, and on the other, there are the bodies themselves. Not until we acquire our
fully developed minds, do we understand that there are two types of objects that can be
immediately perceived: qualities and bodies. An adult will be able to understand that there
is an asymmetrical relation obtaining between qualities and bodies: the former belong to
the latter (not the other way around). Until a certain ripeness of understanding is reached
this asymmetry is not acknowledged.
To conclude this section, let us note that in the section on memory, Reid writes: “[w]e
perceive material objects and their sensible qualities by our senses” (EIP III. 2, p. 257).
Since whatever is perceived, properly speaking, is originally perceived and whatever is
originally perceived is also immediately perceived, this passage supports the idea that bod-
ies and their qualities are immediately perceived.
Based on the evidence discussed so far, Reid seems to be endorsing both the qualities-
only view and the qualities-and-bodies view, but not the bodies-only view. At the outset, I
said that one passage might be seen as providing support for the last view. Let us look at
it and see whether it does indeed support the bodies-only view and not the qualities-and-
bodies view:
[...] if I may trust the faculties that God has given me, I do perceive mat-
ter objectively, that is, something which is extended and solid, which may be
measured and weighed, is the immediate object of my touch and sight. And
this object I take to be matter, and not an idea. (EIP II. 11, p. 154)
Reid writes this in the course of discussing Berkeley’s idealist system. His main con-
62
cern is to argue that what we perceive is physical and external to our mind, not just an idea,
which is mental and thus internal to our mind. Reid argues that matter is the immediate
object of one’s perception – at least of the senses of sight and touch. If we think that the
use of the definite article indicates that matter, namely body (since Reid uses these terms
interchangeably), is exclusively the object of perception, this passage will be seen to con-
tradict all the passages surveyed so far. But this is not the only interpretation possible: the
use of the definite article indicates exclusivity, but of a different kind. The contrast here
is between things of the physical world and things of the mind, and by using the phrase
“the immediate object of my touch and sight” to describe matter, Reid flags his opposition
to Berkeley and the tradition to which he belongs. No mental entities, be they ideas or
something else, are immediate objects of perception; the only things that have that role are
material objects.
Read in this light, this passage supports the idea that bodies, too, are immediate objects
of perception, just like their qualities are. It does not support the bodies-only view, but
the qualities-and-bodies one. One problem remains however: Reid is inconsistent in his
claims. He sometimes argues that only qualities can be immediately perceived; othertimes,
he says that bodies are immediate objects of perception, too. To address this issue, the next
section discusses reasons why Reid should have consistently endorsed the qualities-and-
bodies view.
2.3 WhichViewShouldReidHaveAdopted?
The qualities-and-bodies view agrees more with the rest of Reid’s philosophy than the
qualities-only view, which is the only other viable option. To show that this is so, let
us look at some arguments against the qualities-only view and some advantages of the
qualities-and-bodies view.
63
2.3.1 TheTextualEvidence
As already indicated, the textual evidence does not straightforwardly favor any of the views:
there are some passages that entail that qualities are objects of perception, while others
entail that (only) bodies play that role. There are two options here: either Reid meant that
both qualities and bodies are objects of perception, or he was speaking loosely whenever
he seemed to be contradicting himself. He most probably speaks loosely about perceiving
bodies only. Whenever he gives examples, he speaks as if we perceive bodies; on the other
hand, whenever he tries to make things precise, he speaks as if only qualities were objects
of perception. The passages talking about perceiving the tree outside one’s window can be
reinterpreted as saying that one immediately perceives some tree-type qualities and, since
those qualities can’t exist by themselves, one does in fact perceive a tree (but going through
its qualities first.)
19
In what follows, I show that there are independent reasons favoring the
combined view.
2.3.2 AnArgumentAgainsttheQualities-onlyView
There are at least two reasons why Reid should have supported the qualities-and-bodies
view. In what follows, “immediacy” will be understood (just as before) to mean that per-
ception does not go through any mental intermediaries, and there is no reasoning involved
for an experience to count as immediate. The immediate objects of perception, in this
sense, are things in the world, and not some other entities belonging to the mind of the
perceiver, e.g. ideas somehow derived from physical objects.
The first reason is meant to clear away some obstacles to the qualities-and-bodies view,
rather than to directly argue for it. This works equally well against the qualities-only view
19
This imprecision would not be without precedent: Reid’s view about causation is that, strictly speaking,
only agents are causes (either God or, to a lesser extent, people). But often he talks as if physical objects
caused other things to change. This is Reid speaking loosely.
64
and the bodies-only view. But since the latter is a bad view to have anyway, and lacks
clear textual support, the focus will be on showing that the former should not be endorsed
by someone with Reid’s philosophical commitments. This argument will be called “no
knowledge of bodies without knowledge of qualities”, and before presenting it, let us first
look at a more detailed analysis of the process of perception.
Reid argues that two other faculties are necessary ingredients of every perceptual expe-
rience. The official characterization of perception, in EIP, requires the perceiver to have a
conception of the object perceived and “a strong and irresistible conviction and belief in its
[i.e. the object’s] present existence” (EIP II. 5, p. 96.) Before going any further, I would
like to explain how “perceptual conception” and “perceptual belief” work.
20
With regards to perceptual conception, I subscribe to the view that it is akin to Rus-
sellian acquaintance, as Alston (1989) and Van Cleve (2004) have argued. Alston (1989,
p. 43) argues that perceptual conception is not about the use of “general concepts”, and
hence its role is not to subsume an object under a concept. Another way to put this point
is that perceptual conception does not have any kind of conceptual content; its role is to
present the bare object to the mind of the perceiver, without focusing the attention on what
kind of object that is.
21
Perceptual conception does not function as a faculty that explicitly
predicates properties of objects; its content is not given by descriptive propositions.
However, we do form some notion of what kind of properties and / or bodies we per-
ceive, so, if conception is not responsible for indicating what attributes the substances have,
one could think that that role is played by “perceptual belief”. Copenhaver (2010) defends
such a view, but there is not enough textual evidence to support this interpretation. Reid
20
The marker “perceptual” indicates that both conception and belief, when involved in perception, have
certain characteristics that set them apart from other types of conception and belief, respectively.
21
This is a controversial point among Reid scholars. However, since a full presentation and evaluation of
this debate would take us too far from the concerns of the present chapter, I simply invoke the Alston – Van
Cleve arguments, since I believe they conclusively show that perceptual conception is not descriptive. For
more on the other side of the controversy, see Gallie (1997) and Buras (2008).
65
consistently says only that, in perception, one is seized by a belief that the object perceptu-
ally conceived exists when it is thus conceived.
22
This can be seen as attributing a certain
type of property, namely existence, to the object of perceptual conception. However, this is
not full-blown predicate-attribution, not even on the assumption that existence is indeed a
predicate. One may believe that something exists without thinking that it is a certain kind
of something. For instance, someone may see something at a great distance and have no
idea whether it is a rock or a man with an umbrella, and still believe that what he sees
(whatever that is) does indeed exist. Furthermore, perceptual belief attributes the same
property to every object that is perceptually conceived. In order for perceptual belief to
work as Copenhaver (2010) says it does, it would have to attribute properties like colors,
shapes, sizes to the objects perceptually conceived, and Reid says nothing about perceptual
belief having this role. So, perceptual conception presents an object to the mind, without
describing it as some specific type of object, and perceptual belief just affirms the existence
of the thing thus conceived.
On this understanding of perceptual conception and belief, material substances cannot
be the only immediate objects of perception. If bodies were the sole objects of perception,
we would not get enough information to specify where a body ends and another begins.
Nothing would help us identify the figure, size, color, etc. of the respective body.
Since neither conception nor belief has this role, and since the object of each is the
same as the object perceived, we need something else to explain how it is that we actually
partition the world the way we do. In the absence of such an explanation, we do not get
knowledge of the characteristics of things, in perception. Thus material substances cannot
be the sole objects of perception.
Nor can qualities of material substances be the only immediate objects of perception.
22
For instance, this is clear from passages like IHM 6. 20, p. 168, and EIP II. 5, p. 96, where Reid is
offering his “official” characterization of perception.
66
If all that we perceived were just colors, shapes, sizes, etc, we would have no way of
entertaining singular thoughts about the substances themselves, in perception. I could never
think about my friend that I’m just seeing, without that thought being entirely descriptive.
If taken to the extreme, this view would have bodies be logical reconstructions out of the
physical qualities one perceived, as in Russell’s logical atomism, where bodies are nothing
more than logical constructs out of our sense data.
23
A body, then, would not be perceived, but understood to exist as that thing, whatever it
is, that satisfies a certain description. The table I’m currently touching would be thought
of as being whatever object simultaneously has this particular texture, together with this
particular rectangular shape, and this particular size. This is a counter-intuitive view. My
intuition here aligns with that of those philosophers, Reid among them, who think that
singular thoughts about material substances are required for knowledge of the material
world.
24
Without being able to supply singular thoughts about real, existing objects, per-
ception would fail to satisfy our continued interest in material objects, even when certain of
their qualities have changed. We want to be able to talk about the tree outside our window
both in summer and in winter, even though many of the visible qualities that it has during
the summer are altogether absent in the winter. In other words, our interest in an object sur-
vives the loss of some of its qualities, and Reid is one of the first philosophers to recognize
and attempt to explain the role of perception in the production of singular thoughts.
Immediate perception of bodies has another advantage for Reid. In some of the pas-
sages discussed so far, Reid argues that our perceptual faculty cannot stop at qualities; it
has to have a way of getting information about the body having those qualities. This is
23
See, for instance, Russell (1999).
24
I will not talk about the complex issues surrounding the notion of singular thought in this chapter, but
I want to draw attention to the fact that one mark of singular thought is the ability to use proper names to
directly pick out particular objects. Reid argues that proper names do have this function, and this indicates
that his philosophy can accommodate the notion of singular thought (EIP IV . 1, p. 303.)
67
meant to help Reid rebut a particular brand of skepticism, which claims that having percep-
tual knowledge of qualities of objects is not enough to gain knowledge of the world. This
skeptic would say that, in order to claim knowledge about the material world, one’s mind
must have a way of getting directly to the bodies populating that world.
25
Reid, of course,
would want to show that this type of skepticism is as wrong and contrary to common-sense
as any other kind. These reasons should rule out the qualities-only view.
The combined view is the only one able to do all this work. If the immediate objects of
perception are both qualities and the bodies they belong to, we learn quite a lot about the
external world by perception. We conceive of and believe in the existence of color, figure,
size, etc., as instantiated in the body currently perceived. Even if this type of conception is
non-conceptual, in the sense that it does not descriptively inform me that red
35
is presently
existing in my immediate environment, we will end up knowing more about the body to
which those qualities belong, if we take both its qualities and itself to be immediate objects
of perception. Namely, in the first instance we learn that there are qualities, and since
qualities cannot exist by themselves, we also learn (in the second stage of perception) that
they belong to a body that is currently existent. We move past the qualities, to their subject,
and we are thus capable of entertaining singular thoughts about that body itself. On the
bodies-only view, we would only know that there is a material substance existing before
me – but it would be hard to see how we learn what kind of substance that is. On the
qualities-only model, we would only know that certain qualities presently exist – but this
would not give us the substance itself; it would just be some sort of an after-thought of
perception.
25
Stillingfleet accused Locke of endorsing this type of skepticism. Even if he was wrong, scholarly speak-
ing, and Locke did not support such a view, this well-known debate shows that alleged examples of this type
of skepticism did indeed exist. Reid was probably aware of this debate and he might have been influenced by
it in his interpretation of Locke; so it is not unreasonable to think that he might have responded to this type
of skepticism. For more on this issue, see the correspondence between Stillingfleet and Locke, published in
Locke (1823).
68
2.3.3 DispellingAnObjectionAgainsttheQualities-and-bodiesView
The first reason did not directly favor the combined view, but showed that both of the other
views are wrong. The present argument is meant to dispel an apparent objection to the
qualities-and-bodies view. For all that has been said so far, one might think that we can
perceive bodies only if we know their essences, since otherwise our notions of them are
quite obscure. But we do not have such knowledge, according to Reid: only the creator of
these substances has access to their essences. So, the objection goes, we cannot perceive
bodies. To show that this objection does not work, I argue that knowledge of a substance’s
essence is not required for perception of that substance.
Let us look at the relevant passage, and see what conclusions we can draw:
[...] our conception of [individuals] is always inadequate and lame. They are
the creatures of God, and there are many things belonging to them which we
know not, and which cannot be deduced by reasoning from what we know:
They have a real essence, or constitution of nature, from which all their quali-
ties flow; but this essence our faculties do not comprehend: They are therefore
incapable of definition; for a definition ought to comprehend the whole nature
or essence of the thing defined. (EIP IV . 1, p. 303.)
Even if we thought that we can only perceive things we know the nature of, this would
not show that bodies cannot be objects of perception. At most, this would show that bodies
cannot be perceived without other things being perceived first, namely their qualities. By
perceiving qualities and then bodies, we would not find out what is the nature of material
substances, but we would definitely learn something about them: they have a certain color,
a certain shape, etc. Physical qualities do not have natures that are hidden to us, even
though they are created things, too. Although Reid is silent concerning the issue of whether
qualities have essences, one thing is clear: even if they do have something like an essence,
that nature is going to be importantly different from the nature of a body. Reid is endorsing
the Aristotelian idea that all the qualities of an object flow from its essence (from its form);
69
if qualities themselves would have forms, they would not be determined by the essence
of the body they belong to, but by their own forms. Although the logical space seems to
allow for this possibility, there is no way of making sense of it, in Reid’s metaphysics. This
would amount to saying that qualities themselves have qualities, which flow from their
forms. But, this seems to be the wrong way to go.
To grasp the difference between qualities and bodies, with regards to the issue of
essence, it is helpful to think about the essences of substances as consisting of sets of qual-
ities – the essence of a substance is going to be made up of a very restricted set of qualities,
containing only special properties. Since qualities of substances are not composites of sim-
pler entities, there is no analogous way of thinking about some hidden nature of qualities:
they are given to us in perception in their totality. Thus, qualities can be seen as playing
an intermediary role in perception: through them, we perceive the bodies they belong to,
even though we are never capable of really knowing the essences of these substances, i.e.
knowing which qualities belong to the special set.
This is not to say that perception does not give us knowledge of the material world, on
the contrary. This knowledge is not perfect, however: if the skeptic, in his theory, requires
the non-skeptic to show that one can know the essence of things, then Reid is not proving
that type of skeptic wrong. But he is proving other skeptics wrong: those who claim
that we cannot know that there is an external world containing different types of material
substances. In this, Reid seems to be an externalist: although objects have essences which
are hidden to us, just by interacting with these objects, we learn a lot about them. This, in
turn, is enough to give us knowledge of the external world.
The objection would have some force, on the assumption that we perceive only bodies.
If that were the case, Reid would need to explain why our perception of bodies does not
give us knowledge of their essences. If perception were just a relation between a perceiver
and a body, such a limitation would be mysterious, and Reid would have to agree that it’s
70
possible to have full access to the special set of qualities qua the essential set. But we
cannot do that since, in perception, we do not have access to all the qualities of a body.
This problem disappears once we acknowledge that we perceive some qualities of a body,
and not just the body.
The difference between the natures of qualities (if there are such things) and those of
bodies explains why our knowledge of the material world is restricted in the way it is. Let
us suppose that we can in fact interact, in perception, with each and every essential quality,
on the assumption that our faculties were more perceptive than they actually are. What
we are said to lack, on this picture, is knowledge that a certain set of those qualities is
the essential one. Such knowledge seems to be required by the bodies-only view, and is
actually impossible to obtain, according to Reid. On his understanding of perception, we
can acquire an immediate knowledge that certain qualities exist in the real world, and also
immediate knowledge that certain qualities belong to certain bodies. Perception gives us
enough information to know that certain qualities belong to certain bodies; but not enough
information to know which qualities constitute the essence of a certain body.
Let us take stock. Reid should not adopt the bodies-only view, since on this view,
perception would not provide crucial information regarding what type of object one is
perceiving. He should also not adopt the qualities-only view, since on this view perception
is incapable of giving us singular thoughts about the objects perceived. He should have
adopted the combined view, especially since such a view does not entail that perception
gives one knowledge of the substances’ essences.
2.3.4 IsThereAnyReasonFavoringTheBodies-onlyView?
There seemingly is, namely Reid would have an easy reply to a certain kind of skeptical
challenge, as discussed below. But, I argue that Reid can provide an answer to one chal-
lenge he is faced with even without helping himself to the bodies-first view. Given the
71
problems this view cannot deal with, his whole theory fares better on the qualities-and-
bodies view.
Let us see what advantage the bodies-first view has. If material substances, without
their qualities, were immediate objects of perception, the skeptic Reid has in mind would
be closer to being refuted. As already mentioned, this particular brand of skepticism does
not really claim that there is no material world out there, but that we cannot know that
material substances actually exist. Our senses give us information only about physical
qualities, so why suppose that there is anything more that holds these qualities together?
Reid would not have a good answer, if his considered view were that only qualities are
objects of perception. But he does, on the qualities-and-bodies view.
Arguing that we perceive matter immediately, directly, and without the help of any kind
of physical intermediaries, like qualities, would constitute a strong answer to this type of
challenge. And in the passage from EIP II. 11, at p. 154, in the section criticizing Berkeley’s
theory of ideas, Reid seems to make this argument. If matter can be perceived immediately,
it has to exist. On the other hand, if only qualities of bodies were perceived immediately,
rather than the material bodies they belonged to, one could claim not only that there are no
material substances, over and above bundles of qualities, but also that matter does not exist
altogether, and that the qualities we allegedly perceive are only ideal fabrications.
Reid wants to answer both of these challenges: while the one about there not being
a material world at all seems too extravagant (and so, easily dismissed by anyone with
enough common-sense), the one about the sole existence of qualities, and not also of the
substances they belong to seems to pose greater problems to his theory. This is why he goes
to great lengths to establish that it is easy to know that qualities cannot exist by themselves,
and argues that this is, in fact, a first principle of common-sense. We are naturally consti-
tuted so that we automatically believe that the qualities perceived belong to certain material
substances. It might have been easier to deal with the skeptic’s challenge if Reid did not
72
need to make appeal to first principles in order to argue that material substances exist. If
he could instead argue that material substances are immediately perceived, he would have
a good answer to this type of skeptical challenge.
However, given that he spends so much time arguing that qualities of substances are
immediate objects of perception, I think there is another way of understanding the passage
from EIP II. 11, p. 154. That passage is found in the middle of the discussion of Berkeley’s
ideal system and Reid is criticizing Berkeley, because arguing that there is no material
world, Reid thinks, blatantly contradicts common-sense. It is Reid’s keenness to prove
Berkeley wrong that makes him choose the easiest path, without regard to what he’s saying
in other places. Or without regard to what he should say, given his other theoretical tenets,
namely that it is possible to disprove Berkeley and others like him by arguing that both
qualities and substances are immediate objects of perception.
Moreover, there is another way of answering the skeptical challenge regarding the ex-
istence of material substances. Both qualities and material substances can be taken to be
immediate objects of perception, in the relevant sense of mental immediacy: one will per-
ceive qualities first, in order to also perceive the bodies, but qualities of bodies belong
entirely to the physical realm, and thus are not mental intermediaries. To fully unpack
Reid’s notion of perception, one needs to realize that it works as a two-stage process: first,
the perceiver’s mind latches on to qualities and then the mind passes on to bodies, in virtue
of the perceiver being constituted in a certain way. Although such an explanation is re-
quired in order to fully appreciate Reid’s theory of perception, I will not undertake this task
here, since the immediate purpose of this chapter was to show that both bodies and their
qualities can be objects of perception. In the following section, I explain how perception
is supposed to work, on this assumption. This should prove reason enough for showing
that Reid does not need to say that matter (without its qualities) is immediately perceived,
in order to disprove this peculiar brand of skepticism (that I do not explicitly attribute to
73
Berkeley, but see as a logical possibility).
2.4 Conclusion
Although the evidence that Reid consistently thought that qualities and material substances
are immediate objects of perception is not conclusive, this chapter put forward several
arguments to show that his writings not only allow for this view, but also overall support
it better than any of the competing ones. Material substances are immediate objects of
perception; they are just not bare objects of perception: we need to first perceive some of
their qualities, in order to perceive them. Bodies do indeed populate the world around us;
we just interact with them with the help of their qualities.
Thus, this chapter argued, one should not disregard the passages favoring the qualities-
and bodies view, even though they are at odds with those supporting the qualities-only
view. While one should acknowledge this tension, one should also look at the broader
context, provided by Reid’s general philosophical commitments, before deciding which
of these views is better, given those commitments. The qualities-and-bodies view is the
view which Reid should have endorsed, and would probably have endorsed, if he had been
made aware of this tension in his writings. An explanation of how perception of bodies
actually works, and what role the qualities of bodies actually play, is needed, but such an
explanation would have taken us too far from the main considerations of the present chapter
and it will be discussed in Chpater 3. It turns out that in this regard, too, Reid’s philosophy
is aligned with contemporary research on vision theory and he should be recognized as a
worthy predecessor.
74
Chapter3
TheMechanismofPerception
3.1 Introduction
In Chapter 2, I argued that Reid is one of the first philosophers to think that primary qual-
ities, secondary qualities and the material substances to which these belong are immediate
objects of original perception. In this chapter, I explain how these different things can be
immediately and originally perceived, by providing a detailed picture of the mechanism of
original perception.
To do so, I first discuss Reid’s distinction between original and acquired perception.
Even though the arguments advanced in chapter three showed that secondary qualities,
contrary to what some Reid scholars think, can be immediate objects of original percep-
tion, there may be some lingering doubt that bodies can be so perceived.
1
To eliminate
this doubt, I discuss the reasons that might make one think that one perceives bodies in
an acquired fashion, and show that they do not support this conclusion. This excursus
in the realm of acquired perception is needed because I agree with the diagnosis of ac-
1
The camp of scholars who think that secondary qualities are objects of acquired perception is represented
mainly by Buras (2009), Nichols (2007), and Lehrer & J. C. Smith (1985).
75
quired perception in Van Cleve (2012), where the author argues that acquired perception
is not perception proper, in Reid’s view. On this interpretation of acquired perception, if
it is thought that bodies are objects of acquired perception, it turns out that bodies are not
actually perceived. If Reid had adopted this view, his attempt to refute skepticism about
the existence of material substances would have been ruined. Given that it turns out that
acquired perception is not perception proper, contrary to Rebecca Copenhaver (2010), a
different explanation of how bodies are perceived is needed.
In the second part of this chapter, I explain how original perception works, according
to Reid. I interpret him as arguing that perception is a two-stage process: at the first stage,
one perceives qualities of objects (primary and some secondary ones); at the second stage,
one perceives bodies, that the perceptual mechanism supplies as that to which the qual-
ities of the first stage are bound. These two stages are sequential: one cannot perceive
bodies without perceiving some of its qualities first. However, the binding of features to
objects is not a propositionally inferential step, and it is not mediated by mental intermedi-
aries. To develop this explanation of the mechanism of original perception, I draw on some
contemporary psychological literature discussing the different stages of human visual per-
ception and the formation of object files. Although Reid did not have access to the findings
of these psychologists, I argue that this model is not only anticipated in his writings, but
also that he would have accepted some of their experiments as supporting his conclusions.
Moreover, this model explains how qualities of bodies and bodies are originally perceived,
without being perceived in exactly the same way. This is a welcomed consequence, given
the metaphysical distinctions between bodies and their qualities.
76
3.2 OriginalversusAcquiredPerception
To draw the distinction between original and acquired perception, let us take a look at some
relevant passages. In IHM, Reid argues for the distinction thus:
Our perceptions are of two kinds: some are natural and original, others ac-
quired and the fruit of experience. When I perceive that this is the taste of
cyder, that of brandy; that this is the smell of an apple, that of an orange; that
this is the noise of thunder, that the ringing of bells; this the sound of a coach
passing, that the voice of such a friend; these perceptions, and others of the
same kind, are not original, they are acquired. But the perception I have by
touch, of the hardness and softness of bodies, of their extension, figure, and
motion, is not acquired, it is original.
In all our senses, the acquired perceptions are many more than the original,
especially in sight. By this sense we perceive originally the visible figure and
colour of bodies only, and their visible place: but we learn to perceive by
the eye, almost every thing which we can perceive by touch. The original
perceptions of this sense, serve only as signs to introduce the acquired. (IHM
6. 20, p. 171)
This quote indicates that there are two types of perceptions, and that everyone has more
acquired perceptions than original ones. We are also given a list of the things that are
originally perceived: they are primary qualities and color, which belongs to the class of
secondary qualities. We are not told whether this list is exhaustive, but I would like to note
that the category body is not included here. This last category is not included on the list
presented in EIP and, moreover, color seems to have been scratched off that list too. Here
is how Reid draws the distinction in his later work:
Some of our perceptions by the senses may be called original, because they
require no previous experience or learning; but the far greatest part is acquired,
and the fruit of experience.
Three of our senses, to wit, smell, taste, and hearing, originally give us only
certain sensations, and a conviction that these sensations are occasioned by
some external object. [emphasis added]
77
Thus we learn, that a certain sensation of smell is produced by a rose; [. . . ].
Here it is evident that the sensation is original. The perception, that the rose
has that quality, which we call its smell, is acquired. In like manner, we learn
all those qualities in bodies, which we call their smell, their taste, their sound.
These are all secondary qualities [. . . ].
By the other two senses, we have much more ample information. By sight,
we learn to distinguish bodies by their colour, in the same manner as by their
sound, taste and smell. By this sense, we perceive visible objects to have ex-
tension in two dimensions, to have visible figure and magnitude, and a certain
angular distance from one another. These I conceive are the original percep-
tions of sight. [emphasis added]
By touch, we not only perceive the temperature of bodies as to heat and cold,
which are secondary qualities, but we perceive originally their three dimen-
sions, their tangible fire and magnitude, their linear distance from one another,
their hardness, softness or fluidity. These qualities we originally perceive by
touch only; but, by experience, we learn to perceive all or most of them by
sight. (EIP II. 21, p. 235-236.)
Several things Reid says here are worthy of attention and further discussion. First, ac-
quired perception is the result of experience and learning. So, if no learning and habituation
is involved in a certain mental state (or experience), that state may be called original. Sec-
ond, only two of our senses give us original perceptions – sight and touch. The other three
– smell, taste, and hearing – give us acquired perceptions only. However, by sight we can
also have acquired perceptions, of things that we originally perceive by touch. Reid thinks
that this is something of a rule for acquired perception: we learn to perceive something by
one sense, that was originally perceived by another. Third, the distinction between primary
and secondary qualities of bodies might be thought to coincide with the distinction between
original and acquired perception.
2
I discuss these points in turn.
2
This last point has been recently made by several Reid scholars and I will come back to it later on.
78
3.2.1 AcquiredPerceptionIstheFruitofExperienceandLearning
While original perception is a given of the way we are constituted, in order to have ac-
quired perception something else must happen. One must first acknowledge that some of
one’s original perceptions are almost always accompanied by some other perceptions; then,
this association of perceptions must become habitual. After experiencing this type of asso-
ciation over and over again, it becomes so strong, that whenever one has perceptions of the
former kind, one will automatically have perceptions of the other kinds too. At least, this
is what happens, in theory.
3
The problem here is that Reid does not talk only about experience and habituation,
but about learning. And learning is something that makes one think that more than just
habituation is at play, in order for acquired perception to be developed. More often than
not learning involves inference and judgment. And if reasoning is involved in a certain
mental experience, that type of experience is not to count as perception, in Reid’s eyes.
4
This reading of the text echoes the one found in Van Cleve (2012). There, he offers
five arguments to show that acquired perception is not perception proper, contrary to what
Copenhaver (2010) has argued. The first and last argument of Van Cleve’s are the ones that
interest me most. First, he argues that, since acquired perception is the product of learning,
it is not immediate and it probably involves some type of inference. Although once the
learning is complete, from a psychological point of view, no inference is taking place and
one automatically is led to conceive and believe in the existence of the object of acquired
perception, from an epistemic point of view, things are not so simple. My belief, had by
acquired perception, that the bell I hear on a certain occasion is large “is justified only
because on many past occasions when I heard similar sounds, I knew by sight or touch that
3
For more on this issue, see Van Cleve (2012).
4
See, for instance, EIP, II. 5, p. 96.
79
they came from a large bell. Had I not been justified on those sundry occasions in what
I believed about the source of the sound, I would not be justified now. Thus my current
belief is not epistemically immediate” (Van Cleve (2012, p. 8)). The role of learning here
is crucial: without it providing justification for the beliefs had by acquired perception, we
would have no reason to trust the testimony of our senses, on those particular occasions.
This is enough to make acquired perception a different type of experience altogether.
Moreover, as Van Cleve (2012, p. 19) points out in his last argument marshaled to show
that acquired perception is not a type of perception, Reid himself thinks that this is so. Here
is what he says, in conclusion to the discussion of acquired perception in EIP:
Acquired perception is not properly the testimony of those senses which God
hath given us, but a conclusion drawn from what the senses testify.
That such conclusions are formed even in infancy, no man can doubt; nor is
it less certain that they are confounded with the natural and immediate per-
ceptions of sense, and in all languages are called by the same name. We are
therefore authorized by language to call them perception, and must often do so,
or speak unintelligibly. But philosophy teaches us in this, as in many other in-
stances, to distinguish things which the vulgar confound. I have therefore given
the name of acquired perception to such conclusions, to distinguish them from
what is naturally, originally, and immediately testified by our senses. (EIP II.
22, p. 247)
For the sake of preserving the normal use of the words, Reid agrees to call acquired
perception “perception”; however, common language and common sense diverge on this
occasion, and this is good, for clarity’s sake. To avoid confusion, one must acknowledge
that acquired perception is not perception proper, because it is the product of learning,
namely of “drawing conclusions from what the senses testify.”
In light of this conclusion, one should be very careful not to restrict too much the class
of things that can be originally perceived. If only primary qualities are objects of this type
of perception, as some Reid scholars think, the picture of the world we have by perception
is too incomplete to provide the type of knowledge that Reid argues we can have just by
80
employing this original faculty of our minds. As I argued in Part I of this chapter, however,
other things should be included on the list of things that can be originally perceived, namely
secondary qualities and bodies. I turn now to discussing some refinements of the view
advanced there.
3.2.2 OnlyTwoofOurSensesGiveUsOriginalPerceptions
The converse of this is that the other three senses, namely smell, taste and hearing, do
not give us original perceptions; in Reid’s words, they only give us “original sensations.”
This is a peculiar phrase, since sensation is always supposed to be original; Reid never
talks about acquired sensation. He probably uses it to emphasize that there is an original-
acquired dichotomy concerning these three senses, and that perception can at most be talked
about at the acquired level. The use of this phrase also indicates that Reid thought that
the sensations with which one is provided by these senses are at the same level with the
perceptions provided by the other two senses: both are original, and part of our constitution.
Acquired perception is not part of our constitution and may or may not develop for these
three senses that are constitutively limited to be sensation-providers.
This line of thought is surprisingly modern, on at least two counts: first, the distinction
between original and acquired perception is made mainly by the senses involved, and not
by whether one type of perception is about primary qualities and the other about secondary
qualities. The distinction between primary and secondary qualities, inherited by Reid from
Locke, is interesting to him, but of greater interest is the mechanism of perception and how
it is capable of putting our minds in contact with the external world. There is a powerful
move from metaphysics to the science of psychology here, that one should not ignore.
Second, and more important for the present project, psychologists today and philoso-
phers of mind interested in human perception and the role it plays in justifying knowledge
81
of the external world, draw a similar, if not entirely overlapping distinction: three of our
senses, namely touch, vision and hearing, are seen as giving us perceptions, whereas the
other two are thought to supply us only with sensations. The numbers are reversed, but the
way of drawing the distinction is similar to Reid’s. To show how poignant this parallel is,
let us look at what Tyler Burge writes:
The chemical senses (smell and taste) seem largely to be non-perceptual sen-
sory systems, unless they are supplemented by input from other sources. Con-
stancies are not prominent in the workings of these senses. Although we hu-
mans can locate, map, and remember areas occupied by certain smells, these
abilities seem to rely partly on homing capacities that are not perceptual, or
on other senses or conceptual capacities. There is, I think, objectification in
humans’ determining quality and type of the taste of food or wine. The food is
taken to have a taste, in addition to producing a taste on the tongue. This objec-
tification seems to depend on conceptual association and conceptual memory.
I see no apriori reason why smell and sensory taste could not be largely per-
ceptual systems. The reasons why they are not seem to be empirical. (Burge
(2010, p. 415))
This passage echoes Reid in at least two places: first, Burge argues that smell and taste
give rise to sensations only; second, something like Reid’s notion of acquired perception
seems to be at play, when Burge talks about how we determine “quality and type of the
taste of food or wine.” Whereas, Reid would say that we learn, by observing the constant
association of certain perceived qualities, that “this is the taste of cyder, that of brandy;
that this is the smell of an apple, that of an orange” (IHM 6. 20, p. 171), Burge argues
that we manage to use something like conceptual association to identify these qualities of
food. This conceptual association is something we acquire by becoming habituated with
the objects in question and by learning what to associate with what, and when.
Second, Reid himself thinks that this division between the three and the two senses is
entirely empirical: his idea that we are constituted this way is based on scientific obser-
vation, and not on apriori reasoning. Moreover, he sees no reason against thinking that
82
we could have been constituted in a different way, namely with the actual purely-sensory
senses giving us perceptions, as well as sensations: “Perhaps we might have been so made,
as to taste with our fingers, to smell with our ears, and to hear by the nose. [. . . ] We
might perhaps have been made of such a constitution, as to have our present perceptions
connected with other sensations” (IHM 6. 21, p. 176).
Similarities and discrepancies aside, it is interesting to note that Reid’s intuition has
been consolidated by modern science and that Burge’s interest in the origins of objectivity
has integrated the scientific discoveries in a philosophical program that sees knowledge and
justification as core issues. This shade of modernity in Reid’s philosophy has been mostly
missed by the scholars focusing on the importance of the traditional distinction between
primary and secondary qualities. Moreover, this focus is mostly misplaced, as it can be
seen from what follows.
3.2.3 SomeSecondaryQualitiesAreOriginallyPerceived
Although I have already argued that secondary qualities are objects of original perception,
I would like to develop and qualify this claim, in light of the passages quoted above.
Given that the senses that give us only original sensations are those connecting us
mainly with secondary qualities, it seems fair to draw the conclusion that secondary qual-
ities are objects of acquired, and not of original perception. If we add to this the idea
that acquired perception is perception proper, there seems to be no problem regarding this
issue. But, as already mentioned, not everyone thinks that acquired perception is percep-
tion proper and, more importantly, Reid himself is not of this opinion. Moreover, there
are several places throughout EIP where Reid specifically says that secondary qualities are
“immediate objects of perception” (e. g. EIP II. 17, p. 200; EIP II. 18, p. 211). If we
couple this with his view that acquired perception is not perception proper, it follows that
83
secondary qualities must be objects of original perception, if they are objects of perception
at all.
Let us first look at what made Lehrer & J. C. Smith (1985), Nichols (2007), and Buras
(2009) think that secondary qualities are not objects of original perception. Lehrer & J.
C. Smith (1985, p. 27) and Buras (2009, p. 344) argue that there is a distinction in the
responses that we have to primary and secondary qualities, due to the different types of
principles associated with the way we take notice of the presence of the two different types
of qualities. There are particular principles, “which account for the particular conceptions
of primary qualities”, and there is also a general, inductive principle, “which accounts for
our conception of secondary qualities, smells, tastes, sounds, colors, and so forth, as what-
ever is the cause of the associated sensation” (Lehrer & J. C. Smith (1985, p. 27)). Both
types of principles are principles of our natural constitution, so there is nothing more origi-
nal in the former type than in the latter. Buras (2009, p. 344) takes this idea one step further
and argues that the distinction between original and acquired perception is supported by
this distinction between the principles that are used to give rise to the conception of the
quality perceived (either originally or acquiredly).
This is not the whole story: one can find another distinguishing trait talked about in
both Lehrer & J. C. Smith (1985) and Buras (2009). Namely, the perceptual conception
of primary qualities is direct, whereas the perceptual conception of secondary qualities is
relative, which, for Reid means “indirect”. The conception is relative, or indirect, in that it
is always connected with the sensation in which that perception originated. According to
these philosophers, anything of which one can have a direct notion is something that can be
originally perceived, and everything else, of which one has only a relative notion, is going
to be acquiredly (if at all) perceived.
However, and this is a very important point, there is no separate discussion of color in
84
either Lehrer & J. C. Smith (1985) or Buras (2009).
5
Why is this secondary quality special?
Because Reid acknowledges that it is a secondary quality that is originally perceived (IHM
6. 20, p. 171). So, it has all the traits had by any other secondary qualities – they are
perceived in virtue of the general inductive principle and the notion one has of them is
relative – and still it is explicitly said to be an object of original, and not of acquired
perception. This indicates that one should not draw the distinction between original and
acquired perception based on the distinction between primary and secondary qualities.
But Lehrer & J. C. Smith (1985) and Buras (2009) seem to think that color is not special.
In Lehrer & J. C. Smith (1985), color is grouped with several other secondary qualities, and
said to be an object of acquired perception: “These conceptions of secondary qualities, in
that they involve the operation of the inductive principle, are acquired and not original”
(p. 28). In Buras (2009), color is not included on the list of objects of either original or
acquired perception.
If one wonders why these authors were not sensitive to the special case of color, one
finds a possible answer in Nichols (2007, p. 224). Based on what Reid says in one of the
passages quoted above, it may seem that he changes his mind with regard to color in his
later work (EIP II. 21, p. 236). However, according to Van Cleve (2012), this is not so. Reid
did not change his mind in the years intervening between the publication of IHM and that of
EIP. Van Cleve (2012) gives two reasons to support this claim: one of them has to do with
the scope of the pronoun “these” in Reid’s sentence – “These, I conceive, are the original
perceptions of sight” (EIP II. 21, p. 236, line 4). As Van Cleve observes, this pronoun may
be “meant to refer to refer back not just to the qualities mentioned in the previous sentence,
but to color, which was mentioned in the sentence just before that” (Van Cleve (2012, p.
5
A similar discussion of the fact that color is special for Reid and that other Reid scholars seem to have
missed this point, can be found in Van Cleve (2012). I develop this point here because I want to draw a
somewhat different lesson from it than the one found in Van Cleve (2012).
85
20)). The other, more robust reason why Van Cleve thinks that Reid did not change his
mind regarding color is that later on the same page, Reid “implies that color is originally
perceived” (Van Cleve (2012, p. 21)). This is what Reid says: “The eye originally could
only perceive two dimensions, and a gradual variation of colour on the different sides of
the object” (EIP II. 21, p. 236, lines 26-28). We see that Reid still thinks that color is an
object of original perception.
Color is indeed special: it is a secondary quality that is originally perceived. This is
contrary to what the scholars discussed above interpret Reid as saying, namely that all sec-
ondary qualities are objects only of acquired perception. The problem with their interpreta-
tion is that it disregards what Reid says, without explaining why. One finds a suggestion that
the distinction between primary and secondary qualities does not perfectly coincide with
the one between original and acquired perception in Nichols (2007, p. 224). The problem
here is that Nichols does not defend the idea that color can be originally perceived, but
that there are certain primary qualities that can be acquiredly perceived, for instance visible
figure.
6
This is why the two distinctions are not perfectly isomorphic, as he says. Although
this observation is correct, it does nothing to explain why Reid is thought to be inconsistent
with regards to color.
Color is not even the only secondary quality that is originally perceived. Reid does
not include any other secondary qualities on the list of things originally perceived, but Van
Cleve’s conjecture that there are others is quite apt: “if our senses enable us to localize
a secondary quality (as they do in the case of colors, textures, and temperatures), we may
have original perceptions regarding it” (Van Cleve (2012, p. 21)). I think that this conjecture
is correct, since it is the only explanation of why Reid includes color among the things that
6
In footnote 91, in Buras (2009, p. 349), the author makes a similar observation, but he does not think
that this changes anything in his interpretation. I think it makes his thesis weaker, in the sense that he
should recognize that not all secondary qualities are acquiredly perceived, just as not all primary qualities are
originally perceived. His interpretation should accommodate this issue.
86
can be originally perceived. It also explains why pain, which is originally just a sensation,
can be seen as contributing something to original perception, as Reid argues in EIP, II.
18, p. 211. Pain helps one localize a disorder in one’s body. It is unclear whether pain is
a secondary quality, but it is quite clear it is not a primary quality. Reid thinks that pain
belongs to a larger class to which secondary qualities also belong, namely occult qualities
(EIP, II. 18, p. 217). So, pain is more like secondary qualities than anything else.
Additionally, Van Cleve is on to something, given how Burge, starting from psychol-
ogy, explains why some of our senses are just sensory, and not also perceptual. One feature
that the sensory senses lack and the perceptual ones possess is a capacity to accurately
localize and represent the source engendering the sensation. “This capacity exhibits dis-
tance and location constancies” (Burge (2010, p. 414)). While exhibiting constancies is not
what makes a sense be perceptual, rather than sensory, it is nonetheless a good indication
that perception does indeed take place. Again, modern science is not exactly aligned with
Reid’s theory of perception, since what prompted Burge to write that sentence was the ob-
servation that hearing is a perceptual and not purely a sensory sense; whereas, it was Reid’s
thought that hearing is one of the senses that give us original sensations only.
7
However,
this discrepancy should not prevent us from noticing the bigger picture: localization has
some important role to play in how the distinction between the two categories of senses is
drawn. Since Reid draws the distinction between original and acquired perception, based
on how our senses actually function, and on the observation that some are only sensory,
and not also perceptual, Van Cleve’s conjecture gains a lot of support.
7
Several psychological studies explaining why we should think that hearing is a perceptual sense are
referenced by Burge (2010) in footnote 53, on p. 415.
87
3.2.4 WhatAboutBodies?
To my knowledge, no one has argued that bodies are perceived in an acquired fashion.
There are several reasons why this is so: first, Reid scholars do not appear to think that
there is anything special or problematic with the perception of bodies; they just seem to
assume that we can perceive them tout court. Second, Reid does not discuss bodies when
he talks about the distinction between original and acquired perception. As I have already
argued in Chapter 2, he almost never talks about our perception of bodies – the few passages
where he does say something about this issue seem to be just an afterthought. Third, since
the objects of acquired perception are thought to be mostly secondary qualities, there does
not seem to be enough to prompt someone to think that we perceive bodies in the same
manner as secondary qualities.
I want to dispute this last reason: there is something that makes the perception of bodies
more similar to the perception of secondary qualities, than to the perception of primary
qualities. This may be surprising or not, but if secondary qualities are thought to be objects
of acquired perception, the same should be true about bodies. My position here is that
bodies, just as primary and secondary qualities, are immediately and originally perceived.
There is an explanation as to how this is possible and I will offer it shortly after taking a
closer look at exactly what bodies and secondary qualities have in common, from the point
of view of perception.
The arguments discussed in §3.2.3 should provide enough evidence that, although there
is a distinction that Reid draws between primary and secondary qualities, that distinction is
not sufficient to understand what differentiates original from acquired perception. However,
several Reid scholars believe that the former distinction very nearly coincides with the
latter. So, what was it that mislead Lehrer, Smith, Nichols, and Buras? The answer is
easy enough to find: Reid argues that the notions of primary qualities that one has are not
88
only distinct, but also direct, while the notions had about secondary qualities are usually
confused and also relative. Reid uses “relative” as an antonym to “direct”, but we are told
precious little more about what this relative notions are supposed to be.
8
For instance,
Buras (2009) argues that, “although Reid never says so explicitly, the distinction between
perception involving absolute [i.e. direct] and relative conceptions precisely overlaps the
distinction between original and acquired perception” (p. 348).
If Buras were correct, Reid should have argued not only that secondary qualities are
objects of acquired perception, but also that bodies are such objects. This is because Reid
says that our notions of secondary qualities are relative to the sensations we have of those
qualities (EIP II. 17, p. 202), and that our notions of bodies (or material substances) are
relative to the primary and secondary qualities of those bodies (EIP II. 19, p. 219). Buras
does not address this point. Neither do others, who think that secondary qualities are ob-
jects of acquired perception because perceivers can only have relative notions of them.
Here is what Lehrer & J. C. Smith (1985, p. 29) say, in conclusion of the discussion of
primary versus secondary qualities, and original versus acquired perception: “Finally, to
complete our sketch of Reid’s theory of perception, we note that, in addition to conception
of qualities and objects, there is the ‘immediate and irresistible’ belief in their existence
[. . . ]”. I emphasized the word “objects” in the quote, because its use there is surprising:
objects (probably understood to mean material substances), were not mentioned before, in
8
Wolterstorff (2001) and Buras (2009) argue that “relative” means “descriptive”, which, in the philosophy
of language and mind, has come to be regarded as the opposite of “acquaintance”. I think this is incorrect,
for the following reason: it over-intellectualizes the process of perception. On this view, small children and
animals would be required to have access to and to use the concept “sensation” in order to be able to perceive.
According to these commentators, the content of a color perception, for instance red, would be given by a
definite description like: “the color which produces the present sensation in me”. Wolterstorff (2001, p. 117)
actually notices that this is so and then he says that this is why children do not perceive. This may be true or
not (it’s probably false, if we are to believe recent contemporary psychological findings), but it certainly is
incorrect as an interpretation of Reid. He thinks that animals and infants do perceive, but not like adults do
– because for adults, other faculties aide perception and make it clearer. However, in many places, Reid says
that the conception of a thing that an infant and an adult have in perception is exactly the same. This should
indicate that understanding “relative” to mean “descriptive” is not the way to go, as far as Reid is concerned.
89
conjunction with either original or acquired perception. However, the same feature that
makes secondary qualities objects of acquired perception, in the eyes of these scholars,
should make bodies objects of acquired perception as well. It is surprising that this is not
so.
It would indeed be undesirable to claim that bodies are objects of acquired perception,
for two reasons. First, Reid does not say this: bodies are not included on the same list as
some secondary qualities. Neither are they on the list of the things originally perceived. I
argued that Reid’s text allows for the possibility that bodies are immediately and originally
perceived. Second, our perceptual access to the external world would be very limited, if
only primary and some secondary qualities were to be perceived.
What I want to emphasize here is that everyone should notice that if relative notions
make problems for the perception of secondary qualities, they would make troubles for
the perception of bodies, too. So here is my proposal: one should not claim that just be-
cause we have relative notions of something, that thing is necessarily an object of acquired
perception. Not all secondary qualities are objects of acquired perception, and neither are
bodies, even though our notions of both categories of things are only relative. Therefore,
the distinction between original and acquired perception should not be drawn along the
same lines as the distinction between primary and secondary qualities. This is a distinction
that coincides with a distinction between our senses: by some of them, we only have ac-
quired perceptions; by the others we have original and acquired perceptions. This may not
be as clear a way of drawing the distinction, as the one proposed by the scholars discussed
above, but it is Reid’s way.
Armed with a distinction between original and acquired perception that is not grounded
in the distinction between primary and secondary qualities, let us focus on explaining the
mechanism of original perception of qualities and bodies.
90
3.3 TheTwoStagesofOriginalPerception
I argue that original perception, for Reid is a two stage process (at least): first, one perceives
primary and some secondary qualities, and then one perceives the bodies to which these
qualities belong. The second stage of perception involves some psychological processing;
however, this kind of processing is automatic and non-inferential, in the sense that it does
not involve the mechanism of propositional inference-drawing, employed by reasoning.
When Reid argues that perception must be non-inferential, this is the sense he has in mind:
There is no reasoning in perception, as hath been observed. The belief which
is implied in it, is the effect of instinct. But there are many things, with regard
to sensible objects, which we can infer from what we perceive; and such con-
clusions of reason ought to be distinguished from what is merely perceived.
When I look at the moon, I perceive her to be sometimes circular, sometimes
horned, and sometimes gibbous. This is simple perception, and is the same in
the philosopher, and in the clown: but from these various appearances of her
enlightened part, I infer that she is really a spherical figure. (IHM 6. 20, p.
172).
This quote should make it apparent that Reid’s notion of inference is more robust than
Hume’s: not every transition of the mind, from one (impression or) idea to another is sup-
posed to be an instance of inferential processing. While Hume makes room for “inferences
from experience”, which are the “effects of custom, not of reasoning” (EHU 5.5; SBN 43),
Reid does not.
9
9
Compare this notion of inference to the one in the Treatise: “[. . . ] having found, that after the discovery
of the constant conjunction of any objects, we always draw an inference from one object to another, we shall
now examine the nature of that inference and of the transition from the impression to the idea.” (T 1.3.6.3;
SBN 88) So, there is a close-knit relationship between drawing an inference and making a transition from an
impression to an idea, at least in the case of the cause and effect relation. Hume does not say that that type
of inference coincides with such a transition, but he implies it. Moreover, we can wonder what is the nature
of this transition, or inference: “Since it appears, that the transition from an impression [. . . ] to the idea of
an object [. . . ] is founded on past experience, and our remembrance of their constant conjunction, the next
question is, Whether experience produces the idea by means of understanding or of the imagination; whether
we are dertermin’d by reason to make the transition, or by a certain association and relation of perceptions.”
(T 1.3.6.4; SBN 88-9) The conclusion to the investigation started by this question is that it is not reason that
91
There is no direct textual evidence to support the claim that original perception is a
two-stage mental process; but some of the things Reid says in connection with material
substances uphold such an interpretation. The starting point for this interpretation is Reid’s
argument that:
It seems to be a judgment of nature, that the things immediately perceived
are qualities, which must belong to a subject; and all the information that our
senses give us about this subject, is, that it is that to which such qualities be-
long. From this it is evident, that our notion of body or matter, as distinguished
from its qualities is a relative notion [. . . ]. (EIP II. 19, p. 218-9)
Although the beginning of this quote makes it sound that bodies are not immediately
perceived, I have argued in §2.2 of Chapter 2 that one cannot draw a conclusion like this
based solely on one passage; the claim that bodies are objects of perception cannot be
proved or refuted by making an inventory of all the passages in which Reid talks about what
are the immediate objects of original perception. Reid seems to be genuinely inconsistent
regarding this issue; so my strategy was to consider this topic from a broader philosophical
perspective and see whether Reid would have been, in principle, against the idea that bodies
can be originally perceived. The answer to that investigation was negative.
The rest of the quote is of utmost importance for the current investigation. Reid argues
that by our senses we gain certain information about material substances, namely that they
are the things to which the qualities we perceive belong. Additionally, we are told that the
notion we have of bodies – derived from perception, I gather – is relative to the qualities
they are supposed to have. This last idea is in need of clarification: the notion we have of
bodies can be relative (i) directly to the qualities of bodies, (ii) relative to the qualities a
body is perceived to have or (iii) relative to the qualities, through the sensations we have of
determines us to make this inference. Be that as it may, even from the passage quoted, it is clear that Hume
thinks that drawing an inference is not always a process associated with reasoning. This is something that
Reid would never accept.
92
those qualities (be they primary or secondary). This passage does indicate, however, that
we do not have sensations of bodies qua bodies, otherwise the notion we have of material
substances would be relative to the sensations we have of them, and not to their qualities.
This is contrary to what is suggested in Buras (2009, p. 334): “[. . . ] when describing
the function of sensations, Reid often speaks of the mind passing “immediately” or “in-
stantaneously” from sensations to the perception of bodies and their qualities.” (Emphasis
added).
To see which of these alternatives is better suited to accommodate Reid’s philosophical
commitments, let us look at each possibility in greater detail.
TheNotionofBodyIsRelativeDirectlytoitsQualities
In some ways, it would be good if one could argue that option (i), stating that our perceptual
notions of bodies are relative directly to qualities, is what Reid had in mind. If this were
the case, the conception we have of bodies would be relative to something physical, and
not mental, and no skeptical doubts could be raised against Reid’s perceptual realism. On
this view, cases of perceptual illusion, where we perceive a body having a certain quality
that it does not actually have, would not be able to arise and perception of bodies would be
almost infallible. This should already tell us that (i) is not Reid’s preferred option, since he
believes that perception is a fallible faculty of the human mind. Furthermore, this option
is not viable, for a different, more substantial reason: it is unclear how our notion of a
body can be relative to its physical qualities, directly. We need to have a mental grip on
the qualities of that body; nothing physical can reside in our minds, and we would need
nothing short of that for this option to be a real possibility.
93
TheNotionofBodyIsRelativeDirectlytothePerceptionsofitsQualities
Option (ii), stating that the notion we have of bodies is relative to the perception of its
qualities, is not good, either, for the following reason. It may so happen that we mistakenly
perceive a body to be red, when it is, in fact, orange. Our perception of the body is not
entirely veridical; we are said to have a perceptual illusion of that body: there is a body
out there, with which we are in causal contact, but that body is not red, but orange.
10
But our perception of its color is not veridical at all: there is nothing red out there; we
are just hallucinating this quality and mistakenly attribute it to the body in question. If
our notion of the body were relative to the perception of the quality, this situation could
not obtain. Since the quality in question is not actually perceived (but hallucinated), and
since the perception f the body depends on a non-veridical perception of the quality, the
perception of the body in question will turn out to be just as non-veridical as the one of
the quality. Hallucination of a quality leads to a hallucination of a body, and not to a
perceptual illusion of a body. But illusions should be allowed by a perceptual theory like
Reid’s. On a perceptual spectrum, a perceptual illusion is closer to a veridical perception,
than is a hallucination, which is plainly non-veridical perception. Having a perceptual
illusion brings us closer to knowledge than a hallucination, so perceptual illusions should
be positively regarded.
One may object to this discussion of option (ii) by pointing out that the notion of per-
ceptual illusion cannot be found anywhere in the Reidean corpus. It is true that he does not
talk about perceptual illusions as the one I have described, but he comes close to accounting
for this type of phenomenon, in his discussion of the “third class of errors, ascribed to the
fallacy of the senses, [which] proceeds from ignorance of the laws of Nature” (EIP II. 22,
p. 249). His explanation for how to account for the mistake in such a case is not entirely
10
Note that although color is a secondary quality, I am allowed to use it in this example, since, as I have
already discussed, Reid thinks that color is originally perceived by vision.
94
satisfactory: he argues that it is ignorance, sometimes understood as inattention, of these
laws that makes somebody think that there is something red out there, when there actually
is something orange. However, what is interesting here is that something like perceptual
illusion is on Reid’s list of mistakes that the senses could be thought to give rise to.
To see whether his explanation is feasible, let us attribute the mistake to the fact that the
lighting is uncommon and that makes our perceiver “see” red. If our perceiver is not paying
enough attention to how the uncommon lighting interferes with his perception, it may be
the case that this is why our perceiver does not see orange, even though the body in question
is orange. The problem with Reid’s idea here is that no matter how hard our perceiver may
be trying and how much attention he may be paying to the quality of light and the tricks
it plays on his eyes, he might still be subject to the same perceptual illusion; the body in
question might still look red to him.
11
A further problem with this proposed solution is
Reid’s insistence that errors like this are not to be imputed to the senses, but to the the
judgments one makes regarding the objects of sense; the judgment, and not the perception
is fallacious. However, the case I described seems to be a case where perception is not
entirely veridical; and other cases like this could be discussed, where the error cannot be as
easily dismissed as an error of judgment. Although Reid’s explanation of this phenomenon
may not be entirely accurate, it is remarkable that he even considers something like this to
be worthy of discussion.
Another place where Reid comes close to talking about perceptual illusions is at the
beginning of the chapter on perception, where he says that “even in perception, the notion
which our senses give of the object may be more or less clear, more or less distinct, in all
possible degrees. [...] An object at a great distance is seen more distinctly in a clear than in
11
One problem with some perceptual illusions is that they do not go away, even when we know that we
should not be subject to them. A case in point is the Müller-Lyer optical illusion, where two parallel lines
appear not to be equal, even though we can establish, by measuring them, that they are equal.
95
a foggy day.” (EIP, II. 5, p. 96). There are certain illusions that can be occasioned by fog
and, although fog does not explain away the illusion I described (since I set up the example
without mentioning fog), the quality of light is acknowledged by Reid to interfere with
the distinctness of perception. If the degree of distinctness is sufficiently low, the alleged
perceptual experience may not be, in fact, veridical, in the way described above, and thus
it should count as a perceptual illusion, instead.
Moreover, option (ii) cannot adequately characterize Reid’s theory, because it would
come too close to indirect realism, for his taste. If the notion we have of bodies is relative
to the perception of its qualities, in what way are we departing from “the way of ideas”? We
would not be mistaken in saying that a perception of a quality is a mental intermediary with
intentional content, in a very similar way in which an idea is such an intermediary between
the perceiver’s mind and the world. Reid would not be content with this interpretation of
his theory of perception, since he regards himself as opposing this type of indirect realism,
while being a representative of direct realism.
12
TheNotionofBodyIsRelativetoitsQualities,viaTheSensationsWeHaveofThem
This leaves us with option (iii), which states that the notion of body we have in perception
is relative to its qualities, through the sensations we have of those qualities (be they primary
or secondary). This option has at least two important advantages over (i) and (ii), and none
of their disadvantages.
First, the sensations we have can never be mistaken (not even in the way that our per-
ceptions are sometimes said to be), according to Reid:
It is impossible that there can be any fallacy in sensation: For we are conscious
of all our sensations, and they can neither be any other in their nature, nor
12
For a discussion of Reid’s direct realism, and an explanation in what way it is actually direct, see Copen-
haver (2004).
96
greater or less in their degree than we feel them. It is impossible that a man
should be in pain, when he does not feel pain; and when he feels pain, it is
impossible that his pain should not be real, and in its degree what it is felt to
be; and the same thing may be said of every sensation whatsoever. (EIP II. 22,
p. 243)
Cases of sensorial hallucinations or illusions, akin to perceptual hallucinations and il-
lusions, can never arise, on Reid’s assumption that we have immediate introspective access
to our sensations. This is an assumption that I share. The sensations had whenever some-
one hallucinates something are real: it is because someone feels something, that he thinks
that he perceives an external object. Under normal circumstances, such sensations would
suggest that certain external qualities exist; and they do just that, even under abnormal
conditions. In cases of hallucination, it just turns out that it is not true that the respective
qualities exist, but the sensations suggesting them are there. In this sense, the perceptual
experience is not veridical, but the sensorial experience is so by default. This ensures that,
in cases where a certain quality is not actually perceived, as when someone sees an orange
body as red, the perceiver still has a perceptual connection with the body presumed to have
that quality. One perceives a body, while misperceiving its color, by having a veridical sen-
sation of one of its qualities. Even though the perceptual apparatus mistakenly interprets
an orange-sensation as a red one, it also correctly interprets that same sensation as being of
an external quality, which must belong to a body. Thus one has a perception of a colored
body, even though only an illusion of a red one. Neither option (i) nor (ii) could account
for this situation; since option (iii) can, it is preferable.
This way of accounting for the phenomenon referred to as “perceptual illusion” is
closely related to Reid’s explanation of what happens in cases of phantom limb pain. In
contemporary literature, it is usually agreed that phantom limb pain represents a case of
hallucination, whereas cases of referred pain are instances of illusions. Reid does not talk
about referred pain (it may be the case that this phenomenon was not known in his time),
97
but the case of phantom limb pain that Reid discusses is actually closer to the phenomenon
I was talking about than we might initially think.
We say, that the man had a deceitful feeling, when he felt a pain in his toe after
the leg was cut off; and we have a true meaning in saying so. But, if we will
speak accurately, our sensations cannot be deceitful; they must be what we feel
them to be, and can be nothing else. Where then lies the deceit? I answer, it
lies not in the sensation, which is real, but in the seeming perception he had
of a disorder in his toe. This perception, which Nature had conjoined with the
sensation, was in this instance fallacious. (EIP II. 18, p. 214)
Thus, in cases of phantom limb pain, a person has a painful feeling of an amputated
limb, when, for instance, he thinks that his big toe of the right foot, which he lost to di-
abetes, is hurting. Reid argues that the painful sensation is real, but the perception of the
unknown disorder of the body, which is thought to have caused the sensation, is not veridi-
cal. This is a case of hallucination, in that the toe to which the non-existent disorder is
misattributed does not exist anymore. But, with just a little bit of imagination, one can
construe the case under the guise of perceptual illusion, if one thinks that the non-existent
disorder is actually attributed to that human body, on its whole. Under the former guise,
there is no disorder, and there is no “body” (i.e. big toe); under the latter guise, there is no
disorder, but there is a body (namely that person’s body, minus the big toe in question). I
do not ask the reader to undergo this exercise of imagination because I want to identify a
plausible case of perceptual illusion discussed by Reid; I have already conceded that the
text does not overtly address such a case. I conjecture, however, that the line separating
these two phenomena is not that clearly drawn by Reid, and, more importantly, only if we
think that option (iii) is correct, can we make sense of Reid’s explanation of these confused
phenomena. Namely, we have something like a perceptual notion of a body, relative to a
quality of that body, even though that quality is mistakenly thought to exist. This is the crux
of the problem, and, although the issue is made clearer by discussing a case of illusion, than
98
one of hallucination, from a conceptual point of view, the finer distinctions between these
two phenomena do not matter. What can be said to explain cases of phantom limb pain can
be adapted to explain cases of perceptual illusions, even though Reid does not account for
the latter. Such explanations are made possible by option (iii) alone.
Second, although sensations are mental acts, and always needed for perception, they
are not to be regarded as mediating perception of bodies, in the same way as perception of
qualities would be mediating the perception of bodies, as stated by option (ii). Sensations
do not have the same type of intentional content as perceptions of qualities do; they do
not represent external qualities of bodies. Whether sensations have intentional content at
all is controversial. In analyzing sensation, Reid argues that “[s]ensation is a name given
by Philosophers to an act of the mind, which may be distinguished from all others by
this, that it hath no object distinct from the act itself” (EIP I. 1, p. 36). Passages like
this have been interpreted by Buras (2005) to mean that sensations have themselves as
objects, and by Ganson (2008) to mean that sensations do not have objects at all. I will
not settle this dispute here; what is relevant for the present argument is that sensations do
not have as objects anything external to the mind. Qualities of objects are not represented
by sensations, in the way in which they are by perceptions. This is an important point,
regardless of whether sensations represent themselves or nothing at all. The perception of
bodies is as direct and immediate as the perception of its qualities. If sensations do not
interfere with the immediacy of perception, in the case of qualities, they will not interfere
with it, in the case of bodies, either.
The special role sensations play in the chain of perception has been recognized by al-
most everyone interested in Reid’s theory of perception. Moreover, several articles dis-
cussing Reid’s views on sensation and perception, such as Immerwahr (1978), Pappas
(1989), and Wolterstorff (2000), flagged the worry that Reid’s appeal to sensations makes
him a representative of indirect realism, something that Reid would have opposed.
99
Copenhaver (2004) offers compelling reasons to dispel this worry. The apparent prob-
lem is Reid’s insistence that sensations are signs of external qualities, an idea he borrows
from Berkeley. This has been understood to mean that sensations have a representational
character – they are seen to represent “objects and qualities in virtue of information they
carry and we interpret” (Copenhaver (2004, p. 69)). It is true that Reid regards sensations to
be signs of external entities, namely qualities and bodies, but it is important to understand
“in virtue of what Reid regards sensations as signs. Reid thinks that the representational
relation that holds between sensations and objects is an external relation – it is fixed by
extrinsic properties of sensations and objects, properties they have in virtue of being con-
nected to one another by God.” (Copenhaver (2004, p. 71)). It is not because sensations
do represent the qualities in question that they manage to be signs of them. It is because
God, or Nature designed us in this way: it is by our constitution that sensations signify
something external to our minds. Recall what Reid was saying in IHM: there is no intrinsic
connection between our sensations and what they signify; we could have been so made
that our tactile sensations would have signified tastes, instead of textures, and our hearing
sensations would have signified smells, and not sounds. (IHM 6. 21, p. 176).
But, one can object, even though sensations do not have the same type of intentional
content as perceptions do, if they have any type of intentional content, they can still be seen
as mediating between perceivers and the external world. The evidence presented above
indicates that sensations do not represent anything (other than themselves), but it does not
indicate that they are not mental intermediaries. However, it is important to understand of
what kind of physical or mental intermediaries Reid is afraid here. As Copenhaver (2004,
p. 72) aptly puts it, “Reid opposes the idea that perception requires an internal relation
between mediating mental entities and objects.” Namely, Reid is against the “way of ideas”
because ideas are representational mental entities that mediate the access of the perceiver
to the external world. But sensations are not like that. The following argument shows that,
100
in this respect, sensations are no more problematic than, for instance, the rays of light are
for visual perception.
As it is noted in Copenhaver (2000, p. 18), unless a perceptual experience and the per-
ceived object are numerically identical, there are going to be intervening steps and maybe
even entities, that might be said to mediate between the perceiver and the world. But this
is a trivial notion of mediation, to which Reid is not opposed:
[. . . ] there are certain means and instruments, which, by the appointment of
Nature, must intervene between the object and our perception of it. [. . . ] If the
object is not in contact with the organ of sense, there must be some medium
which passes between them. Thus, in vision, the rays of light; in hearing, the
vibrations of elastic air; in smelling, the effluvia of the body smelled, must pass
from the object to the organ; otherwise we have no perception. (IHM 6. 21, p.
174)
The things talked about in the above quote are all of a physical nature. But, Reid uses
the same word, “medium” to talk about signs, in general, which are said to signify things
to the mind: “[. . . ] any kind of sign may be said to be the medium by which I perceive
or understand the thing signified. The sign, by custom, or compact, or perhaps by nature,
introduces the thought of the thing signified” (EIP II. 9, p. 134). This quote is used in
Wolterstorff (2000, p. 11) to draw the conclusion that, since sensations are said to be signs
of external qualities, it is clear that sensations are “a sort of medium between the external
object and our perception thereof.” This is true, and sensations can be seen to be this sort
of medium but, in light of the previous quote, being a medium is not necessarily a bad thing
and does not have the consequence of turning Reid’s realism from direct to indirect. If we
put these together, we are only entitled to conclude that sensations are, on the mental side
of things, what the rays of light are on the physical side.
If on option (ii), it turned out that perception of bodies cannot be immediate, in this non-
trivial sense, on option (iii), the notion we have of bodies in perception is had immediately,
101
by appeal to the sensations we have of their qualities. Perception of bodies is immediate, in
the relevant sense: it is not mediated by mental entities, which can be either states or acts,
with (the relevant kind of) representational content.
Sensations play an important role in the perception of both primary and secondary
qualities. If we are only interested in the perception of external qualities – both primary
and secondary, inasmuch as we can talk of original perception of secondary qualities –
there can be no talk of sensoriless perception. The only case of perception of an external
quality where doubts have been raised is the perception of visible figure. However, as
shown in Yaffe (2003a) and Yaffe (2003b), this is not so. Yaffe argued that “Reid takes
visual sensations to suggest both color and figure, and so holds that visible figure is indeed
suggested by a sensation, just like every other perceived quality” (Yaffe (2003a, p. 135))
13
This is a good precedent to the view advanced here: if it is possible for the sensation of
color to suggest both color and visible figure, nothing seems to preclude the sensation of
color to suggesting also a body. There are no body-sensations suggesting bodies; but there
are quality-sensations suggesting bodies. Thus, the sensation one has of an external quality
performs double (and sometimes triple) duty, being instrumental in the perception of the
quality in question and of the body having that quality, as well. The alternative, where
one is able to perceive a body without having any sensation suggesting it, would contradict
the “Standard Schema” (Wolterstorff (2001, 96-110)) and hence an explanation would be
needed to show why Reid gives up his schema, in this particular case. The burden of proof,
therefore, is on the scholars arguing that the perception of bodies does not make use of any
13
This is the so-called “unpopular view”, since the majority of Reid scholars have argued that visible figure
is perceived without the help of any kind of sensation. Yaffe (2003b, pp. 105-11) discusses the passages that
support the view that the perception of visible figure goes through a sensation of color and compellingly
addresses the difficulties raised by some passages that seemingly contradict this view. The popular view,
arguing that no sensation suggests visible figure, and that the perception of such figure is sensoriless, is
supported by: Chappell (1989), Lehrer (1989b), and Wolterstorff (2001). In the rest of this chapter, it will
be assumed that Yaffe proved this interpretation wrong, since it would make Reid’s position on perception
incoherent.
102
kind of sensation. This is the only alternative to the view defended here, since nowhere
does Reid talk about our having specific sensations of bodies.
This option allows a perceiver to have singular thoughts (in the form of perceptions) of
bodies, even if the perceiver is mistaken about a certain quality that body is thought to have.
Once a perceiver has a sensation of a quality, that sensation suggests the quality, which is
then conceived and believed to exist. Once this happens, the perceiver, armed with both the
sensation and the perception of the quality, is in a position to perceive the body to which
that quality belongs. The object then becomes the focus; the properties do not disappear,
but are effaced, and the perceiver can keep his attention on the object. The next section
presents a model explaining how this is possible.
LessonsfromThePsychologicalLiteratureontheBindingProblem
Having argued that the perception of bodies is dependent on our having a mental grip on
their qualities, let us now look at a model that makes explicit the two stages of perception:
at the first stage, one perceive qualities, having had sensations of them, and, at the second
stage, one perceives a body having those qualities. The literature discussed in what follow
argues that human visual perception is like this; the results can be applied to tactile per-
ception, without too many modifications (the features registered by touch will probably be
different from the ones registered by vision, but that should not be surprising). It is unclear
whether the sense of hearing lends itself to the same type of analysis, since the objects of
hearing tend to be quite different from material substances, which are objects of both vision
and touch. This is not problematic for the view advanced here, since psychologists still dis-
pute whether hearing is a perceptual, as opposed to just sensory, sense, as it was mentioned
before. Moreover, Reid thinks that hearing does not give us original perceptions, so we
should not expect the results discussed to apply to more than the two senses that give us
original perceptions.
103
I do not claim that Reid endorsed such a model; the only claim I make is that the
psychological findings discussed here are a natural way of understanding how Reid’s theory
could have been developed. If Reid’s perception is to be understood as a two stage process,
it is normal to look for answers concerning exactly how these stages come about and what is
the connection between the two. While Reid himself did not carry out such an investigation,
the model presented here is appropriate to his own findings.
The model in question was proposed as a way of understanding the role of attention in
perception, and it was later developed and adopted as a model for how perception of objects
works. This is known as the “feature-integration theory of attention” (FIT), and it was
initially proposed by Treisman et al. (1977), and subsequently developed by Treisman, in a
series of articles, some of which are talked about here. A series of experiments conducted
on normal subjects (with no known physical impairment) revealed the fact that “features [i.
e. qualities of objects] are registered early, automatically, and in parallel across the visual
field, while objects are identified separately and only at a later stage, which requires focused
attention” (Treisman & G. Gelade (1980, p. 98)). Not all features of objects are registered
at the first stage of perception, also called the “preattentive stage”. This stage of early
processing selects only a small set of basic features; researchers do not always agree which
features belong to the first stage of visual processing, however. While Treisman (1998,
p. 405) argues that color, size, contrast, tilt, curvature, and line ends are detected in the
first stage of visual processing, other researchers think that “movement and differences in
stereoscopic depth are also extracted automatically in early vision” (Treisman (1998, p.
405); see also Wolfe (2003)). These features are identified as existing in the visual field
and perceived automatically. Another way of describing the first stage of visual processing
is “as the extraction of features from patterns of light” (Treisman (1998, p. 399)). Light is
essential for human vision, of course, a fact understood by Reid and explained by him as
constituting part of the physical medium needed for visual perception to occur. It is also
104
interesting to note that the features identified by Treisman (1998) as belonging to the first
stage of visual perception are all qualities that could be originally perceived in vision, by
Reid. He is silent about movement
14
, but he would place stereoscopic depth in the class of
qualities visually perceived in an acquired manner.
According to FIT, objects in our environment, together with their settings, are identified
at the second stage of visual perception. The mechanism of the perception of objects is not
yet well understood and explained. However, it was observed that after the existence of the
basic type of features is registered via the first stage of perception, those features are then
“bound” or attributed to an object. Focused attention is needed in order to bind features
to objects, and the search for bound features (or relevant conjunctions of features) is done
“serially, or one at a time, as if a mental spotlight were being moved from one location to
another” (Treisman (1988, p. 399)). This is one difference that has been found between
the first stage of visual perception and its later stages: the search for unbound features is
done in parallel, whereas the search for bound features is done serially. No explanation
has been offered for this disparity, but it has been observed that the second stage cannot
operate in the absence of the first one, whereas the first stage can operate correctly even
when the second stage does not, as it is evidenced mostly by patients with certain types of
brain damage.
15
All this talk of attention, however, does not imply that perception of objects involves,
in any capacity, propositional inference-drawing. In perceiving objects, we do not first
perceive their features and then infer that an object must exist and intentionally direct our
14
His theory of movement perception is quite complicated, and it is unclear whether he thinks that move-
ment can be actually perceived. A more detailed discussion of movement will be taken up in the chapter on
Memory.
15
In particular, the case of a certain patient, referred to as R. M., is illustrative here: while he can easily
identify multiple unbound features, he cannot do the same either with conjunctions of two or more features,
or with objects; he can only see one object at a time. A more detailed discussion of his case and its relevance
for normal subjects follows shortly.
105
attention to binding the features perceived to that object. This is important, in light of the
previous discussion of the difference between Reid and Hume on inference and reasoning.
Even though visual perception has two stages, the second stage does not depend on the
perceiver’s capacity for drawing inferences and it is not based on reasoning. Here is what
a psychologist argues about perception of objects: “The perception of meaningful wholes
in the visual world apparently depends on complex operations to which a person has no
conscious access” (Treisman (1998, p. 399)).
16
On Reid’s understanding of inference, one
cannot use reason to draw an inference if one is unconscious of doing so. This is the way in
which perception of objects is not inferential, if it is understood to work according to this
model, and this is why objects should not be said to be perceived in an acquired manner
(acquired perception probably involves reasoning, as previously discussed).
The way of integrating the features perceived at the pre-attentive stage into a whole, at
a later stage, is not unique to perception. Psychologists have observed that we have a gen-
eral “capacity to integrate information across time, space, attributes and ideas” (Treisman
(1999b, p. 105)). The human mind has a great associative power, and, many times, cre-
ativity is understood as finding the correct patterns in what seems to be a jumbled mixture
of pieces of puzzles. However, this way of processing in perception, as in other faculties,
is quite puzzling, and researchers have been trying to “understand how we can respond to
relations to relevant subsets of the world, but not to relations between arbitrarily selected
parts or properties” (Treisman (1999b, p. 105)). The main interest here is in the so-called
“binding problem”, or set of problems, that identifies what types of seemingly disparate
features are combined by the human mind into meaningful wholes.
16
This is possible due to a very technical notion of attention, which can and does occur without awareness,
according to Treisman (2003, p. 109). Reid does not use a similar notion of attention, his being more or less
synonymous with awareness. However, this is of no consequence; even if he had thought that this process
is not identical with attention, he would have still endorsed the idea that there is some mental processing
involved in perception, distinct from inference. This is all that matters here.
106
In its most intuitive form, the binding problem for visual perception could be stated
thus: a visual scene usually contains more than one object, and it may be difficult, in
some cases, to know which features belong to which objects. This is even more problem-
atic in cases where a certain object occupies multiple locations, or when other objects are
occluding it, thus breaking its continuity (Treisman (1999b, p. 105)). The researchers in-
vestigating the binding problem argue that determining which features belong to an object
is a process that takes place every time a perceiver looks upon a scene.
There is no discussion of a “binding problem” as such, in Reid. This is surprising,
given his theory of perception: we perceive qualities of objects and then, due to the fact
that a first principle of our constitution is activated, we form the conception and belief that
there is an object having those qualities. Things can go wrong here: just because one is
guaranteed that a certain perceived quality cannot exist by itself, and hence must belong to
an object, the principle of our constitution to which Reid appeals does not ensure that the
quality in question will be correctly attributed to the object having it. Moreover, nothing
ensures that several distinct qualities are attributed to the same object. The first principle in
question is only responsible for correlating a quality with an object; no more. So, one can
have a binding problem, on Reid’s theory, even though he does not discuss the issue.
There is some indirect evidence that he was aware of the difficulty of segregating par-
ticular objects in a given scene, a problem increased by the possibility of their occluding
one another. For instance, Reid observes that:
I can directly conceive ten thousand men or ten thousand pounds, because both
are objects of sense, and may be seen. But whether I see such an object, or
directly conceive it, my notion of it is indistinct; it is only that of a great mul-
titude of men, or of a great heap of money; and a small addition or diminution
makes no perceptible change in the notion I form in this way (EAP I. 1, p. 11).
The direct perceptual conception of ten thousand men we might have in perception does
107
not tell us a lot about that multitude: there are too many objects in that visual scene for us
to be able to segregate their individual features and recombine them in a veridical way.
The treatment of the binding problem in the contemporary psychological literature is
important because it presents evidence supporting the claim that features are bound to ob-
jects in perception at a later stage, after they themselves are first perceived. According to
Treisman (1996), the binding problem comes in different flavors. She identified at least
seven different binding problems for visual perception: (i) property binding, concerning
the way in which different properties are bound to the objects to which they belong; (ii)
part binding, concerned with the way in which the parts of an object are segregated from
the background and bound into a coherent whole, which is especially difficult when partial
occlusions disrupt the continuity of the objects in the scene; (iii) range binding, “in which
particular values on a dimension (e. g. purple on the color dimension) are signaled by ratios
of activity in a few distinct populations of neurons (e. g. those sensitive to blue and to red
to signal purple)” (Treisman (1996, p. 171)); (iv) hierarchical binding, concerned with the
binding of the features of shape-defining features to the surface-defining properties carry-
ing them; (v) conditional binding, where the interpretation of a certain property (e. g. the
direction of motion) may depend on another (e. g. depth); (vi) temporal binding, concern-
ing the way in which successive states of a certain object are integrated across temporal
intervals; (vii) location binding, concerning the way in which objects are bound to their
locations (Treisman (1996, p. 171)).
Some of these binding problems are more relevant to Reid’s theory of perception, which
is less technical than the one Treisman and her colleagues proposed. Property binding
represents the core of the theory, but part binding, temporal binding and location binding
are important, as they give rise to puzzles showing that perceiving objects and motion, for
instance, is not as automated a process as one might initially think. In what follows, I focus
on property binding.
108
The case of the patient R. M., who suffered damage to both sides of his occipito-parietal
cortex and consequently displayed the usual symptoms of the Balint’s syndrome, presented
evidence that one can have real issues with binding features to objects (Treisman (1999a,
p. 94)). His condition is characterized by three primary problems: the inability to perceive
more than one object at a time (simultagnosia), the inability to reach towards objects on
which one is focused (optic apraxia), and problems in changing the object on which the eyes
are focused (optic ataxia). His case was extensively studied by psychologists, especially
by Friedman-Hill et al. (1995) and Robertson et al. (1997). In experimental lab settings,
R. M. did not show any difficulties in binding different features to a single object, if those
features were presented sequentially. Since he suffered of simultagnosia, he was unable
to “bind simultaneously presented features to the correct objects” (Treisman (1999a, p.
100)). This observation is in line with my interpretation of the role of the first principle
of our constitution, which says that qualities cannot exist on their own. In Reid’s view,
this principle is supposed to ensure the move from qualities of objects to an object to
which those qualities belong. It provides a one-to-one correlation between a quality and an
object. Even if experience kicks in and expands this type of correlation, helping to attribute
many qualities to a single object, the principle in question, as previously discussed, is not
responsible for binding several qualities to different objects in the scene. The Balint’s
syndrome patient offers support for the hypothesis that something like this first principle of
our constitution can still function well, and help bind different qualities to a single object,
when the qualities are presented one after another, even when more specialized functions
are damaged.
The most frequent binding error that R. M. was observed to make was seeing illusory
conjunctions, when he was presented with more than one object. When presented with two
or more objects, he would, for instance, incorrectly attribute the color one of the objects
had to one of the other objects presented in the display. He was shown “some very simple
109
displays containing just two coloured letters selected from T, X, and O in red, blue or
yellow, and asked to tell the first letter he saw. In some sessions, even with exposures
as long as 10s, he made binding errors, reporting one letter in the colour of the other”
(Treisman (1999a, p. 94)).
Examples of illusory conjunctions are not limited to pathological cases. Certain ex-
periments showed that normal subjects can make the same type of binding errors R. M.
was seen to make. The paradigmatic experiment for testing whether normal subjects are
susceptible to illusory conjunctions of features was proposed by Treisman & H. Schmidt
(1982).
17
Subjects were presented with a display showing a line of colored shapes or let-
ters, flanked by two black digits. They were asked to report what the two digits were, so
that they would not pay close attention to the colored shapes. Subjects reported correctly
what the digits were, but the reports of the stimuli between the two digits were inaccurate
and showed that the subjects were susceptible to a large number of illusory conjunctions.
Interestingly, it was irrelevant whether the stimuli between the two digits were letters or
shapes; the subjects reported illusory conjunctions regarding all the features tested – i. e.,
color, shape, size and solidity. The conclusions drawn from this series of experiments were
not based solely on the verbal reports of the subjects, but also on simultaneous and succes-
sive matching tasks, thus indicating that these are neither failures of memory nor of verbal
labeling (Treisman & H. Schmidt (1982, p. 138)).
The existence of illusory conjunctions of features, both in pathological, but especially
in normal cases, indicates that features are separately coded by the perceptual mechanism,
since otherwise there would be no reason to explain why it is possible to recombine them
incorrectly. This tendency to form illusory conjunctions by recombining features which
17
The psychological literature also discusses some experiments meant to show that binding errors arise for
the binding of parts as well as for the binding of features. These will not be described here, since they do not
raise any special issues, but for more on them, see Treisman (2003).
110
characterize separate objects supports the idea that our visual perception starts by per-
ceiving features and then, with the help of attention, moves to perceiving conjunctions of
features and hence objects (Treisman (1999a, p. 96-97)).
The experiments described by Treisman and her colleagues have been contested, on
the basis of being quite difficult to find evidence of perception of unbound features. If
no unbound, or free-floating features can be perceived, the objection goes, then the case
for illusory conjunctions, which is based on the possibility of free-floating features to be
haphazardly combined, is pretty weak. The data supposed to help establish that unbound
features can be perceived is inconclusive, according to some researchers. While Treisman
& G. Gelade (1980) argued that the identification of features takes precedence over the
localization of features (and thus that a certain feature can be unbound from its location),
Sagi & B. Julesz (1985) argued for the opposite claim, and Green (1992) thought that
there is an equality between the two types of operation. I will not delve deeper into this
controversy; for scientific research, this just is business as usual. However, I would like
to point out that refinements to the original FIT were proposed, and, with some further
amendments, psychologists do accept this model of visual perception, according to Wolfe
(2003) and Chan & W. G. Hayward (2009).
18
As already mentioned, visually perceiving objects, on this paradigm, means that several
features perceived at the preattentive stage are bound to an object representation, at the later
stages of vision. According to several researchers, visual perception is object-driven: the
preattentive stage just supplies the perceiver with the necessary materials for the perception
of objects to be ensured. It is the second, attentive stage of perception that is responsible
18
For instance, it was initially thought that the preattentive stage coincides with the operations of early
vision, while the later stages with those of later vision. This was shown to be untrue, since it was observed
that attentive processing itself uses some of the information stored in the early visual cortex, according to
Wolfe (2003, p. 71). Chan & W. G. Hayward (2009, p. 126) discuss alternative theories and whether they can
explain the data better than FIT. Their conclusion is that FIT fares better than a search-based account in all
the experiments described.
111
for our having singular thoughts in perception. This is important for Reid: if these psy-
chologists are right, one can perceive the object, and thus have a singular thought about
it, via perception, even in the absence of veridical perception of its qualities. So, although
the possibility of illusory conjunctions is seen as proving that visual perception starts with
qualities of objects, it does not show that visual perception stops with features of objects.
Several researchers argue that the best evidence for the idea that perception is object-
driven comes from motion tracking experiments.
19
Kahneman et al. (1992) made the ob-
servation that as long as the attention is guided by those features that control the unity and
continuity of an object over a certain interval of time, a certain object will be perceived
as the same object, even though some of its qualities are seen to change over that interval
of time. For instance, the object can be seen to change its shape and color, and still be
perceived as the same object, as long as its continuity was not broken down. According
to Kahneman et al. (1992, p. 217), “[i]f the appropriate constraints of spatiotemporal con-
tinuity are observed, objects retain their perceptual integrity and unity”, even when their
spatial location, sensory properties and any descriptive labels used to pick them out change.
They think that “some form of object-specific representation, addressed by its present lo-
cation and by its continuous history of travel through space over time” is responsible for
the object-specific perceptual phenomena. The example used to illustrate this is that of the
onlookers in a Superman movie, who would say things like: “It’s a bird; it’s a plane; it’s
Superman!”, without registering any change of referent for the pronoun “it”. They see and
talk about a single object, initially thought to be a bird, then a plane, then Superman.
In IHM, Reid presents a similar case, and he talks about it in the same manner:
Walking by the sea-side in a thick fog, I see an object which seems to me to be
a man on horseback, and at the distance of about half a mile. My companion,
19
For a good review of the literature on motion tracking, see Scholl (2001). For a detailed discussion of
some of the experiments conducted see, in particular, Pylyshyn (1989).
112
who has better eyes, or is more accustomed to see such objects in such circum-
stances, assures me, that it is a sea-gull, and not a man on horseback. Upon a
second view, I immediately assent to his opinion; and now it appears to me to
be a sea-gull, and at the distance only of seventy or eighty yards (IHM 6. 22,
p. 183).
The object we are told, appeared to him to be first a man on horseback, and then a sea-
gull. However, the pronoun “it” refers to the same thing throughout, even though there is
a considerable conceptual distance between a man on horseback and a sea-gull. Moreover,
Reid thinks that perception is at play here, and not any other faculty, that might involve
reasoning, as it is apparent from the following:
The mistake made on this occasion, and the correction of it, are both so sudden,
that we are at a loss whether to call them by the name of judgment, or by that
of simple perception.
It is not worth while to dispute about names; but it is evident, that my be-
lief, both first and last, was produced rather by signs than by arguments and
that the mind proceeded to the conclusion in both cases by habit, and not by
ratiocination (IHM 6. 22, p. 183-184).
So, according to both Reid and contemporary psychologists, visual perception can work
in such a way that, no matter what sensory qualities we identify as belonging to an object,
if certain spatiotemporal conditions are met, an object will be perceived as a single object,
even when it undergoes certain radical apparent changes.
To explain this phenomenon, the psychologists supporting FIT argued that “a scalable
‘window of attention’ scans a ‘master map’ of locations, selecting the features currently ac-
tive in corresponding locations of various specialized feature maps, and suppressing those
in other locations to prevent erroneous bindings. The selected features are combined to
form an ‘object token’.” (Treisman (1996, p. 172)). This ‘object token’ is more commonly
known under the label ‘object file’, and is a representation of the object, where its features,
observed at the preattentive stage are coded, stored and updated. To employ an analogy,
113
perception works just like the police, opening a file during the investigation of a crime,
where the information on the crime is collected and updated. The same is true whenever
an object comes under the incidence of visual perception: a file is opened, where all the
perceptual information about that object is entered (Treisman (1998, p. 410)).
20
In the
examples above, what ensures that the perceivers think about the same object through all
those changes is the fact that only one object file is opened and the successive manifes-
tations of that object are allocated to the same file. The object’s perceptual continuity is
thus maintained through the changes, by there being just one object file where everything
is updated. This is possible on the assumption that the object remains still or “it changes
location within constraints that allow the perceptual system to keep track of which file it
should belong to” (Treisman (1998, p. 410)).
The very idea that psychologists talk about how our perceptual system creates, updates
and maintains an object file, while the object undergoes changes of location, shows that
they think that perception is object-driven, and hence that our perceptions are mainly de re.
This, despite our perception starting from individual properties of the objects in questions.
Throughout changes in positions or color, or even form, we still think of that as being
one object and not just whatever happens to satisfy a certain complex description at each
given moment. If the description we start with is different from the one with which we
end, we cannot say that there is something that remains the same. If our way of perceiving
and thinking about objects were always determined via a complex definite description, the
alleged referent of the description would change, should one predicate in the description
change. There is a way of putting this in language more appropriate to Reid’s time: if
objects were just bundles of qualities, and we would only have access to such bundles
20
The literature on object files is quite diverse and it addresses several loosely connected issues. Here, I
only focus on the literature that discusses issues relevant to how singular thoughts about objects are possible,
even though we start from their qualities, and even in the absence of correct attribution of qualities to objects.
114
via these qualities, then changing a quality would change the bundle. In the Superman
example, we would not be able to keep “it” referring to the same thing: the first bird-
bundle is different from the last Superman-bundle. It is in virtue of an object file that we
can keep referring to the same object, even though we misattribute qualities to it.
The model then is this: we start with qualities, and open an object file; we can make
changes to that file, as long as certain spatiotemporal constraints are observed. Once we
have the object file, we are able to refer, in perception, or in thought derived from per-
ception, to the object itself, while its qualities remain in the background. This is a model
different from both bundle theory and its more modern variant, descriptive theory. One
does not need to perceive the object whole, at once, without having any perception of any
of its qualities, in order for that perception to count as de re. On this two-stage model, de re
perceptions are possible, even though, to speak Reid’s language, the perceptual conception
of the object, and hence the perception itself, is relative to the object’s qualities.
Let us take stock: the interpretation of Reid presented here proposed that his theory
sees perception as a two-stage process. At the first stage, qualities of objects are perceived,
and at the second stage, the objects to which those qualities belong come under perceptual
scrutiny. Contemporary literature on visual perception describes it as a two-stage phe-
nomenon. This contemporary model is a natural extension of Reid’s theory, one he might
have endorsed, given the opportunity.
3.4 Conclusion
This chapter discussed how to best understand the mechanism of perception, on Reid’s
theory of perception. The first sections distinguished between original and acquired per-
ception, arguing that for Reid, perception proper must be original, and not acquired. The
important lesson of that discussion was that qualities of bodies and bodies themselves are
115
objects of original, and not acquired perception, for Reid.
The last sections argued that, for Reid, original perception is a two-stage process: the
first stage is dedicated to the perception of qualities, while the second, to that of the objects
having those qualities. Drawing on psychological literature concerning visual perception, I
offered a possible model that would extend Reid’s theory, while preserving its most impor-
tant tenets: original perception must be non-inferential and de re original perceptions must
be possible.
If the arguments and analyses presented here are correct, Reid’s theory of perception is
more modern than initially thought and aligns itself with contemporary research on vision
theory and perception.
116
Chapter4
MemoryandthePerceptionofEvents
One of the main projects that Reid undertook was to explain the different mechanisms and
offices of the faculties of the human mind, in order to understand how it is that we have
knowledge of the external world, with all its objects and persons. Perception, we have
seen, is credited with giving us knowledge about presently existing objects. Analogously,
and maybe unsurprisingly, Reid argues that memory gives us knowledge about things that
existed in the past. It is a first principle of common sense “[t]hat those things did really
happen which I distinctly remember” (EIP VI. 5, p. 474). Moreover, this type of knowl-
edge, again, as in the case of knowledge acquired via perception, is “immediate knowledge
of things past” (EIP II. 1, p. 253). These remarks might lead us to believe that there is not
much to be said about memory that has not already been said about perception. A simple
formula seems to be at work here: take whatever is true about knowing presently existing
things via perception and project that to the past. The result is a product of remembering, of
memory. This seems to be Reid’s train of thought, when he argues that two key necessary
ingredients (or concomitants) of memory are a conception of a past object and the belief in
its past existence: “[m]emory implies a conception and belief of past duration” (EIP III. 1,
p. 254). This principle mirrors the one he gave for perception: any perceptual experience
117
implies a conception of a presently existing object and the belief that it exists (EIP II. 5,
p. 96). Although both perception and memory are simple original operations of our minds,
and hence cannot be defined, knowing that conception and belief are necessary for both
perception and memory is a good starting point for further explanation and clarification.
Reid’s account of memory has attracted less interest from Reid scholars than his ac-
count of perception. This is probably because in his view memory is so closely related to
perception, that when all is said and done, not much is going to be importantly different for
memory. There are some very good articles explaining how it is possible to have knowledge
of the past and what consequences this has for Reid’s direct realism.
1
There are yet other
articles concerned with explaining Reid’s famous criticism of Locke’s theory of personal
identity, which is taken to involve memory essentially.
2
This chapter builds on the work of
the scholars explaining how knowledge of the past is possible, in Reid’s view, and it will
not contribute to the discussion of his criticism of Locke’s theory of personal identity.
The focus here is on a puzzle that is not discussed in the secondary literature, which
arises from Reid’s conviction that the present is just a moment of time, understood as an in-
divisible point, and nothing more. This makes him argue that motion cannot be perceived,
despite what the vulgar thinks. There is no direct solution to this puzzle; the only way of
defanging it is by having Reid give up one of his commitments. The problem is acute, as
will emerge from the discussion below, since, on the understanding of Reid’s account of
memory in the current secondary literature, having knowledge about past events is a wel-
come result of one’s use of one’s faculty of memory. However, Reid’s account of memory
doesn’t allow us to actually remember events, despite his insisting that it does. Moreover, it
1
This is (more or less) an exhaustive list of articles dedicated to explaining Reid’s theory of memory:
Copenhaver (2006c), Hamilton (2003), Van Woudenberg (1999), and Van Woudenberg (2004).
2
See, for instance, Yaffe (2009a), for an explanation of Reid’s contribution to the debate concerning the
criterion of identity for persons. As a side note, Locke’s theory of personal identity might not, in fact, involve
memory as closely as Reid, and many other philosophers, thought it does. For a good explanation of various
ways of reading Locke and understanding his principle of individuation for persons, see Weinberg (2011).
118
might turn out that his account of memory is too restrictive. If Reid is right, our knowledge
of the past, and even of the present, is far more limited than we might have initially thought.
The structure of the chapter is as follows. In §4.1, I discuss what type of conception
is involved in memory, and I present arguments for thinking that “memorial conception”,
as I will refer to it, is, at least sometimes, conceptual. I propose that we think about the
conception ingredient or concomitant employed by the faculties of perception, memory,
and imagination, as arranged on a continuum, with perceptual conception and imaginative
conception placed at the opposites ends, and with memorial conception placed somewhere
in between. In other words, memorial conception is a hybrid, having both non-propositional
and propositional elements in its content, and it might be called, “pre-conceptual”.
3
This
might be surprising, but it is a useful result.
In §4.2, I discuss whether it is possible to remember events. Strictly speaking, on Reid’s
view, this is not possible. However, if we are willing to be broad Reideans in our commit-
ments, rather than follow the letter of the text, remembering events turns out to be possible.
But, the theory requires a metaphysics of events, such that what is remembered includes
internal and external objects. The internal objects are events or states, like perceiving,
sensing, being conscious of, while the external objects are the ones populating our regular
environment, like flowers, trees, and people. This will help explain how episodic memory
functions.
4.1 MemorialConceptionIsSometimes(Pre-)conceptual
Reid argues that whenever we remember something, we have a conception of the thing re-
membered and a belief that that thing existed (EIP III. 1, p. 254). There are two questions
3
As I argued in Chapter 1, imaginative conception is conceptual, whereas perceptual conception is non-
conceptual.
119
that one can raise at this point: what counts as an object of memory? What type of concep-
tion is Reid talking about here? I discuss the question about conception in this section and
the question about objects in the next.
We have seen in Chapter 1 that the conception we have of things perceived is primarily
non-conceptual, whereas the conception we have of things imagined is primarily concep-
tual. We may wonder whether the conception involved in memory is like the one involved
in perception, or rather like the one involved in imagination. Reid is not very helpful here;
he does not discuss memorial conception as a special kind.
However, the issue arises as soon as we are wondering how exactly it is possible to have
“immediate knowledge of things past” in the present. Several commentators have noted
how improbable it is to think that we can presently have an awareness of the past. Hamilton
(2003, p. 231) argues that having any type of immediate awareness of the past would be
like having a “telescope into the past”, which we obviously don’t have. Copenhaver (2006c,
p. 181) is very careful to emphasize what exactly we are aware of in memory, according
to Reid: “the events we remember are past rather than present and so cannot be objects
of a current apprehension.” Copenhaver further adds that if the things remembered were
objects of a current apprehension, there would be no distinction between perception and
memory, since in perception, the object perceived is present to the mind, namely currently
apprehended. But Reid thinks that such a distinction between memory and perception is
real and that we can uncover it if we get sufficiently clear about how these faculties work.
Van Cleve (2013) argues that Copenhaver (2006c) and Hamilton (2003) deny the pos-
sibility of direct awareness of things past based on an argument that could be used to show
that any awareness of the past is problematic, and not just direct awareness of the past. I
think that Van Cleve is right in his discussion of that argument, but Copenhaver (2006c)
and Hamilton (2003) were denying that we can have direct awareness of the past in order
to show that Reid’s distinction between perception and memory is correct. Their worry
120
seems to be that once we allow direct awareness of the past, there would not be anything
sufficiently different between a perception and a memory.
It might seem natural to draw the distinction between the two faculties by referring to
the objects of perception and memory: an object of perception must be present, whereas
an object of memory must be past. At some point, Van Cleve thought that this is, indeed,
Reid’s view and now he argues that such a difference in the presentness of the objects of
perception and the pastness of the objects of memory is at least necessary to make the
distinction between the two faculties (Van Cleve (2013)).
However, on closer scrutiny, it becomes apparent that this difference cannot also be
sufficient for the distinction. Such a procedure would only work for distinguishing between
perceiving and remembering events. A current event is different from a past event; the past
event doesn’t exist anymore, even though it might involve the same objects that the current
one does. A current object, on the other hand, like the tree I’m currently seeing might not
differ at all from a past one; the same tree, three days ago, would have looked the same, in
a favorable climate, like the one in Florida, for instance. How are we to know when we are
remembering the younger tree, as opposed to seeing the older tree, if the only difference
lies in whether the object is present or past? For all I know, I might be remembering the
tree while it seems to me that I’m currently seeing it.
The fact that we might remember events differently from how we remember objects
that we’ve once perceived is not discussed in the literature. However, Reid might have an
answer, explaining how we know that we’re perceiving something, instead of remembering
the same thing, on any given occasion. And this answer applies to both events and objects.
Reid claims that: “I am conscious of a difference in kind between sensation and memory,
and between both and imagination” (IHM 2. 3, p. 29). Van Cleve (2013) also notices Reid’s
way of drawing the distinction and, to explain what this difference in kind might consist
in, he employs Russell’s help. According to Russell, if we experience the same object
121
in two intrinsically different ways, as we do in perception and memory, the two types of
experience must involve different relations to that object (Russell (1992, p. 54)). Although
Russell doesn’t cite Reid, we can find a similar idea in his writings, as Van Cleve (2013)
notes. A certain sensation, “a smell, for instance, may be presented to the mind three
different ways: it may be smelled, it may be remembered, it may be imaguned” (IHM 2. 3,
p. 27). We thus see that when the object of memory and perception is the same, Reid thinks
that the distinction between memory and perception must be drawn internally.
My interpretation therefore agrees with the one in Copenhaver (2006c): the distinction
between perception and memory is at least partly determined by the internal workings of
each of these faculties. Both perception and memory involve other two faculties: concep-
tion and belief. For each of these faculties, the conception is of the object perceived or
remembered, and the belief is an immediate existential belief, which I have argued that it
is best understood as the assertive force of mental operations with veridicality conditions.
4
So, if we are to look at the inner workings of perception and memory for the distinction be-
tween them, we need to look no further: the conception employed in perception is different
from the conception employed in memory, and this difference is sufficient to distinguish
between the two faculties.
The difference between the two types of conception is easy to state, but a bit more dif-
ficult to explain: perceptual conception is non-conceptual, whereas memorial conception
is not. It has been argued that perceptual conception is like Russellian acquaintance (see
Van Cleve (2004, p. 108)), or like Donnellan’s notion of ‘having in mind’, to get rid of
some of the problems that acquaintance carries with it (see Chapter 1). Memorial concep-
tion, however is not acquaintance-like. According to Copenhaver, memory preserves the
4
For more on this issue, see Chapter 1. The basic idea is that, if this belief were fully fledged existential
judgment, brutes and infants would not be able to perceive (or remember) anything, until they acquire the
concept of existential quantification. This over-intellectualize a mental process which we are said to share
with animals and infants.
122
previous apprehension, namely that which was conceived in perception. Thus “Reid is able
to account for memory in terms of previous, rather than present awareness or apprehension
[...] Memory preserves past apprehension by conceiving of an event previously appre-
hended and believing, of this event, that it happened”(Copenhaver (2006c, p. 181-182)). I
take Copenhaver to be arguing that memorial conception is preserving the conception in-
volved in perception; it is not itself acquaintance, but, at most, the preservation of such an
acquaintance.
This interpretation is indeed supported by something that Reid says at the opening of
the Essay on memory:
It is by memory that we have immediate knowledge of things past: The senses
give us information of things only as they exist in the present moment; and this
information, if it were not preserved by memory, would vanish instantly, and
leave us as ignorant as if it had never been. (emphasis added; EIP III. 1, p.
253)
I suggest that the preservation Reid is talking about here is done in part by the con-
ception ingredient of memory; this conception has the office of interacting with perceptual
conception in such a way as to preserve information and help us remember that information.
Does this mean that Reid’s account of memory is related to a “trace” theory of mem-
ory, or to an “idea” theory of memory, both of which, according to Van Cleve (2013), he
criticizes? Briefly, according to a trace theory of memory, the physical impressions caused
by external objects would persist after the object is not present to the senses, and thus,
these traces would enable our memories. According to an idea theory of memory, on the
other hand, the immaterial impressions (in Hume’s parlance, this time) would be somehow
stored by memory and revived, with a lesser degree of vivacity, whenever somebody would
remember something.
From the passage quoted, we can see that Reid talks about information-preservation,
and this information does not seem to be of a physical variety, like a Reidean impression.
123
So, if a trace theory of memory talks about the preservation of the impressions themselves,
Reid’s account is not a trace theory.
However, it is more difficult to answer the question regarding the connection between
Reid’s account and an idea theory of memory. It is certain that Reid wants to distance
himself from such a view, given that he considers all things related to ideas to have dubious
epistemic consequences. It is unclear whether he manages to do so. Copenhaver (2006c,
p. 181) argues that, according to Reid, the object of a memory is not a past apprehension;
in other words, the object of memory is not something like a Humean impression, that
one simply receives via perception. But, according to Hume, the object of memory is not
an impression either: it is an idea derived from an impression, which, although has less
vivacity than the original, “is somewhat intermediate betwixt an impression and an idea”
(T 1.1.3.1; SBN 9). In the passage quoted above, Reid talks about memory preserving
the information from the senses, and not about what exactly is the object of memory; so,
one way of distinguishing his theory from Hume’s is to say that however the information
the senses give us is preserved by memory, the objects of memory are external, and not
internal, as Hume’s ideas.
However, if we look at another passage in Reid, things become less clear. Reid objects
to both Locke and Hume, by arguing that past impressions (or ideas, or perceptions) cannot
actually be revived, and thus an impression cannot reappear as an idea, as Hume thinks.
Against Locke, Reid argues that memory is already presupposed in the revival of a past
perception, and thus it cannot be explained by Locke’s principle, that “The mind, as it
were, paints them anew upon itself”(Locke, Essay, II. x. 2). Here is what Reid says: “the
mind, which paints the things that have ceased to exist, must have the memory of what
they were, since every painter must have a copy either before his eye, or in his imagination
and memory” (EIP III. 7, p. 285). So, there can be no revival of past perceptions without
memory. This entails that memory must be something else, since memory is needed to
124
bring about this revival. Additionally, if memory were just the revival of a past perception,
the distinction between perception and memory would collapse, and this would not be a
good result (EIP III. 7, p. 286).
Against Hume, Reid argues that it is incomprehensible to say that a past impression can
reappear as an idea, and that this is done by memory. This issue is related to the one Reid
found objectionable in Locke, and it should be dismissed as well, on the same grounds:
if the existence of the perception or impression is interrupted, a numerically identical per-
ception or impression cannot be revived later on, unless the original was already preserved
somehow all along. This, again, presupposes memory and does not offer an explanation of
what memory is.
Impressions and ideas are fleeting perishable things, which have no existence,
but when we are conscious of them. If an impression could make a second
and a third appearance to the mind, it must have a continued existence during
the interval of these appearances, which Mr. Hume acknowledges to be a gross
absurdity. It seems then, that we find, by experience, a thing that is impossible.
(EIP III. 7, p. 288)
These objections lose their force, when compared to what Reid says in another place:
I must therefore have perceived it [the transit of Venus over the sun] when it
happened, otherwise I could not now remember it. Our first acquaintance with
any object of thought cannot be by remembrance. Memory can only produce a
continuance or renewal of a former acquaintance with the thing remembered.
(emphasis added; EIP III. 2, p. 255)
If we understand “acquaintance” to mean something close to a Humean “impression”
or a Lockean “perception”, then memory, for Reid, plays a similar role, at least some of
the time it is employed: it renews or revives the acquaintance we had with an object via
perception. Somebody may object to my interpretation here by saying that we should not
think about acquaintance as being so closely connected to Humean impressions or Lock-
ean ideas. To this I reply: as discussed above, the conception involved in perception is
125
best understood as enabling the perceiver to have in mind, or to be acquainted with, the
object of perception. Perceptual conception is thus non-conceptual. It is automatically per-
forming its function, whenever the perceiver has a sensation caused by an external object,
which suggests the respective conception. It has been argued that Humean impressions and
ideas are like images and as such they can be understood to have non-conceptual (or non-
propositional) content. So far, the similarities are greater than the differences. The main
distinction between Hume and Reid is that Reid thinks that we do perceive external objects,
whereas, at least according to him, Humean perception stops at these images derived from
objects. Be that as it may, I find it problematic that Reid thinks that the acquaintance itself
can be renewed by memory. On this view, it is difficult to see how the object of mem-
ory turns out to be whatever external object was once perceived, and not the perceptual
conception or acquaintance one established in perception with that external object.
If we focus on the idea that Reid believes that “memory produces a continuance of a
former acquaintance”, then what Copenhaver (2006c) is saying seems correct: memory
preserves a past apprehension of or an acquaintance with the object of a past perception. If
it is preservation of acquaintance, the object of memory is not the acquaintance itself. But,
the model of memory as repository of acquaintances, and not of ideas, seems to present
itself here as a natural candidate.
5
However, if we focus on what Reid says immediately after claiming that memory pro-
duces a continuance of a former acquaintance, namely that memory produces “the renewal
5
Reid criticizes Locke for arguing that memory should be understood as such a repository of ideas (EIP
III. 7, p. 284); but if I am correct, he falls into the same trap. Maybe Reid criticizes Locke for arguing that
ideas which do not exist any longer can be laid into a storehouse and then revived. By contrast, Reid does not
say that the acquaintance we originally had with an object of perception goes out of existence: more often
than not, that acquaintance is continued by memory. However, it is still difficult to explain how and why there
is a need for memory to renew that acquaintance in some cases: has the person remembering something lost
his initial grasp of the object once perceived? If so, how does memory revive that original grasp? I would
like to say that I don’t exactly know how to answer these questions, but I’d like to also point out that we can’t
just ignore them.
126
of a former acquaintance”, Reid seems to be very much a product of his age. On this
interpretation, one function of memory (although probably not the sole one) is to help us
remember, by renewing it, the initial contact, call that “perceptual conception” or “acquain-
tance” that we had with an object, in perception. In this sense, one object of memory is the
initial acquaintance itself: this operation of the mind presents again to the mind that initial
apprehension.
Why is it important to know how close Reid’s account is to that of Locke’s or Hume’s?
Reid is responding to the theories advanced by his predecessors and contemporaries and
sometimes, unbeknown to him, he is influenced by what they say. Other times, it is difficult
to understand how to address important issues that he didn’t overtly address but should
have. Without knowing what type of conception is the one employed in memory, it might
be impossible to establish what kinds of objects one can remember and the manner of
remembering them.
A surprising omission in this regard is Reid’s ignoring the fact that memory rarely
perfectly preserves the information the senses give us. By applying his favorite method for
analysing the faculties of our mind, a philosopher might think that the information memory
preserves is not the whole of the information received via one’s senses, in perception. If
I think about what I remember, even episodes I was a part of in my recent past, there are
some details that disappear, or are effaced. The details that remain are the “essential” ones,
we could say. It is as if our mind uses an alembic: everything that is remembered is what is
left after a process of evaporation. As in distillation, so in memory: the things that remain
are concentrated.
6
The information that is preserved by memory undergoes a process of
conceptualization: we remember the color red of a rose that we saw yesterday, as red. We
6
Interestingly, in talking about abstraction, Reid employs an analogy between intellectually analysing an
object and chemically decanting a compound into its ingredients. As with any analogy, he warns us that it is
not perfect and that it can lead us into error, but he employs it nonetheless (EIP V . 3, p. 370). My inclination
to common sense leads me to believe that this process starts with (or maybe presupposes) memory.
127
might not remember exactly the shade that it had, but we would know that it was red, if
prompted to say what color the rose had. By contrast, in perception we are apprehending
the object whole, without necessarily deploying our conceptual apparatus. In memory, in
order to preserve what we perceived, we must employ concepts. Without any conceptual
grip on something we perceived, we would not be able to remember something we see for
the first time, such as a fruit that we hadn’t seen before. This might explain why infants
do not remember many things; they do not have the concepts needed to fix the information
that they get in perception. Once they acquire more and more concepts, they are able to
retain more and more from what they perceive.
One indication that this reconstruction is on the right track is Reid’s choice to talk
about memory after talking about perception, and before bare conception and abstraction.
He argues that infants and brutes perceive, in the same way as adults do; but in talking about
memory, Reid makes no such conjectures.
7
On the contrary, we are given to understand
that memory develops later in life:
In the gradual progress of man, from infancy to maturity, there is a certain
order in which his faculties are unfolded, and this seems to be the best order
we can follow in treating of them. The external senses appear first; memory
soon follows, which we are now to consider. (EIP III. 1, p. 253)
This passage doesn’t show that memorial conception is conceptual; it just indicates
that Reid thinks that memory comes later in the development of a human being. One
explanation of why this happens is to think that the ability to form concepts must also be
underway, and maybe even interact with memory in such a way that memory (or one type
thereof) is actually instrumental in our ability to form general concepts. But this is not the
only possible reading of this passage.
7
Reid seems to think that brutes have something like memory, but that it is different from the type of mem-
ory humans have. This is the reading I favor of the following passage: “If we could not record transactions
according to their dates, human memory would be something very different from what it is , and perhaps
resemble more the memory of brutes” (EIP III. 7, p. 294).
128
There is just one place where Reid (at most) alludes to the fact that memory’s preserving
of what was once perceived is not perfect:
There are cases in which a man’s memory is less distinct and determinate, and
where he is ready to allow that it may have failed him; but this does not in the
least weaken its credit, when it is perfectly distinct. (EIP III. 1, p. 254)
There are two ways of understanding this quote: we can think that memorial conception
is non-conceptual, or that memorial conception is more (or pre-)conceptual. A good case
can be made for each of these readings. Let us first assume that memorial conception is
non-conceptual, just like perceptual conception. Let us further assume that here Reid is
addressing a case where the perception of the object was distinct and determinate, and thus
our acquaintance with the object, via perceptual conception, was distinct and determinate.
However, some of that distinctness and determinacy is lost in memory, and the acquaintance
provided by memorial conception, although still non-conceptual, is somewhat blurred. This
way of reading the passage again brings Reid very close to Hume: it is as if memorial
conception, in this case, is less vivid than our perceptual conception was.
We could instead think that the loss of distinctness and determinacy Reid is talking
about here has to do with the aforementioned process of distillation: we could think that
there are details that disappear completely, or are made less vivid, but that what remains,
and what is actually remembered, is more vivid, than what was perceived, in some respects.
If we understand memorial conception to be pre-conceptual, and hence different from per-
ceptual conception, its objects are part of our conceptual apparatus that gives them a higher
degree of focus (or vivacity).
This latter reading of the quoted passage assumes that there can be degrees of con-
ceptuality. On this assumption, “non-conceptual” just means that the object is given to us
entirely, as a thing that does not have this or other property. “Conceptual” means that we
attribute properties to the object in question, explicitly, in virtue of our having acquired the
129
respective concepts. If perception gives us non-conceptual information, one would be said
to be able to perceive something he has never encountered before. If perception didn’t give
us any non-conceptual information, we would not be in a position to perceive something
without knowing what it is. To say that a faculty is conveying “pre-conceptual” information
is to think that the information in question is made up of non-conceptual and conceptual
elements. To illustrate this, let us think about simplified versions of the accounts given by
Hume, Russell, and Frege. If we think that Humean impressions are image-like, they are
encoding non-conceptual information. If we think that Fregean thoughts are purely pred-
icative (made up of general concepts only), they are encoding only conceptual information.
And if we think that Russell’s singular propositions are hybrids, made up of physical indi-
vidual objects and general concepts, the information encoded by these propositions is, on
the present view, pre-conceptual.
The idea that there might be degrees of conceptuality makes sense, if we think about
the fact that acquiring concepts is a process, that takes time, and involves comparing and
contrasting of information derived from the senses. Concepts don’t just happen on us: we
acquire or develop them, after encountering many instances of a general concept in our
daily lives. We can think that the initial information we have via the senses is at one end of
the conceptual continuum, having non-conceptual elements, and the end result, the general
concept is at the other end of that continuum. And everything in between, in this process
of acquiring concepts, is to be called “pre-conceptual” and understood as being a hybrid of
non-conceptual and conceptual elements.
It is difficult to say exactly which reading of the above passage should be preferred:
we are not given enough to draw a firm conclusion. What I want to emphasize, though,
is that we may not need to draw a firm conclusion: it is enough if we think that some
type of memory, in virtue of the memorial conception it employs, uses something close
to concepts, whereas other types of memory do not. Someone might wonder why I am
130
forcing this view on Reid, as it may appear I am. The answer is easy to give: as it will be
apparent from the next section, Reid’s account of memory must make recourse to the idea
that memorial conception is, at least sometimes, more conceptual than not (or, as I said
before, pre-conceptual). Otherwise, we would not be able to remember some of the things
he says we can remember (namely, events).
Moreover, on this understanding of memorial conception, it is easier to understand
how imagination functions. We construct nonexistent things, in our imagination, not from
scratch, but by using materials that were furnished to us by perception. But to be able to
retain those materials we need memory: only the stuff that is preserved by memory is going
to be used in imagination to think (even singularly) of Don Quixote, for instance. In this
respect, the conception component of memory must be working in a similar way as the one
of imagination, at least some of the time. Let us suppose that the arguments I offered in
Chapter1 to support the idea that imaginary conception is conceptual are correct. On this
assumption, those arguments should be understood to show that memorial conception is
closer to the imaginary kind, than to the perceptual kind.
Is there any direct evidence in Reid that he thinks this way? No, as far as I can see; but
there is nothing directly opposing this interpretation either.
Moreover, there is some circumstantial evidence that I am interpreting Reid correctly
here: he argues that for infants, and for people suffering from a disorder of the mind, it
may not be obvious that what they remember is actually remembered, and not imagined, or
vice-versa:
Perhaps in infancy, or in a disorder of the mind, things remembered may be
confounded with those that are merely imagined; but in mature years, and in
a sound state of mind, every man feels that he must believe what he distinctly
remembers [. . . ]. (EIP III. 1, p. 254)
On my interpretation, this is possible because memory and imagination function sim-
131
ilarly: the conception that is an ingredient or concomitant of these two faculties is con-
ceptual. There is a need for an additional ingredient or concomitant, namely of belief, to
help us realize, in difficult cases (not just when we are not sound of mind, as Reid argues)
whether we really remember something happened, or whether we just imagine it did.
This interpretation is bolstered by what Reid says about the notions we have of exter-
nal things (by our senses), and of the operations of our minds (by consciousness). These
notions, which are “first notions”, are neither simple, nor distinct, says Reid in criticizing
Locke’s attempted reduction of all our ideas to ideas of sensation and ideas of reflection. It
is the use of our memory that aids us in making these conceptions distinct, once they are
accessed via our external or internal senses. In talking about reflection, and the role it plays
in understanding and categorizing the world around us, Reid writes:
For, although our first notions of material things are got by the external senses,
and our first notions of the operations of our own minds by consciousness,
these first notions are neither simple nor clear. Our senses and our conscious-
ness are continually shifting from one object to another; their operations are
transient and momentary, and leave no distinct notion of their objects, until
they are recalled by memory, examined with attention and compared with other
things. (emphasis added; EIP III. 5, p. 269).
Memory is cited as being the next step on the road to obtaining simple and clear no-
tions of external and internal things. Our acquiring the notion of duration, for instance,
presupposes memory. The interaction between memorial conception and the more ratio-
nally developed faculties of abstraction, generalization and reflection is not made entirely
explicit, but memory seems to be doing double-duty. On the one hand, it retains the infor-
mation we have from perception and consciousness – and because of that we are entitled
to think that memorial conception is not conceptual. On the other hand, it helps us form
certain concepts, by sublimating some of the information we have from perception – and
because of this, I think we should be entitled to think that memorial conception is more
132
conceptual than perceptual conception. This, I take it, indicates that memory is the first
step on conceptualizing these things, that our senses record. For this reason, I argue that
the conception employed by memory is pre-conceptual, and not purely of the acquaintance
type, like the one employed by perception.
8
4.2 TheObjectsofMemory
Let us now turn to answering the other question I announced I will address: what are the
objects of memory? Reid is more helpful here than he was in indicating what kind of
conception is an ingredient of memory. After taking a look at the most relevant passage, I
will briefly discuss what kinds of memory there are, in order to know exactly what kind of
memory Reid talks about and what types of objects it can have.
Here is what Reid says at the beginning of the Essay on memory:
Things remembered must be things formerly perceived or known. I remem-
ber the transit of Venus over the sun in the year 1769. I must therefore have
perceived it at the time it happened, otherwise I could not now remember it.
Our first acquaintance with any object of thought cannot be by remembrance.
Memory can only produce continuance or renewal of a former acquaintance
with the thing remembered. (emphasis added; EIP III. 1, p. 253-55)
The first sentence of the quote above seems like a very common-sensical thing to say:
to remember something, I must have had an initial grasp of that thing by a different faculty,
at a previous time. In contemporary psychological and philosophical literature on memory,
this is regarded as an essential condition on at least one type of memory. The condition is
called “The Previous Awareness Condition” and it is usually held that it is needed for what
8
If someone objected to this interpretation by arguing that memorial conception just helps us retain non-
conceptual information, gotten by perception, and then, by comparing, we reach conceptual information of
the kind discussed in the main text, I would reply by arguing that Reid does not talk about the faculty of
“comparing”, and so memory should be the faculty responsible for aiding us in forming general concepts and
using them, in the way presented in the text.
133
is known as episodic memory, and not for what is known as semantic memory. According
to Tulving (1983) there are three types of memory: procedural – whereby one remembers
how to perform certain actions (e. g. one remembers how to ride the bike, or how to bake
a cake); episodic – whereby one remembers an experience that one underwent or an event
one witnessed (e. g. I remember running my first 5k race); and semantic – whereby one
remembers that so-and-so is the case, where the fact remembered may be something that
happened before one’s time (e. g. one remembers that Napoleon was defeated at Waterloo.)
The procedural type of memory is not of interest here: Reid has nothing to say about it.
But the other two types are quite interesting, especially from the point of view of any
discussion of singular thought. Some philosophers (e.g. Burge (2003)) have argued that
no de re memory is possible in the absence of episodic memory: if one loses the ability
to remember actions that one was the agent or witness of, and one can only have semantic
memories of those actions, it is thought that that person will only be able to offer a de dicto
description of what happened. In this respect then, episodic memory seems to be more
fundamental and ground semantic memory.
The Previous Awareness Condition distinguishes episodic memory from semantic mem-
ory, although some philosophers argue that there is something like a Previous Awareness
Condition for semantic memory, too: one can only remember something if one previously
learned (or knew) that.
9
The distinguishing trait of episodic memory, however, still re-
mains: only for this type, and not also for the semantic type, the Previous Awareness Con-
dition concerns previous experiences or perceptions, and not facts learned at one time or
another. Reid might be thought to have a broad Previous Awareness Condition in mind, as
Van Cleve (2013) points out, when he writes that one can only remember “things formerly
perceived or known.”
9
Copenhaver (2006c, n. 22) points out that Sydney Shoemaker argues that semantic memory must be
subject to a different Previous Awareness Condition.
134
This distinction, although not explicitly drawn by Reid, is useful for understanding that
the type of memory he is mostly concerned with is episodic memory. At least, one forms
this impression by looking at the examples Reid gives, almost all of which use language
usually associated with episodic memory.
10
In fact, Van Woudenberg (2004, p. 207) argues
that Reid is only concerned with episodic memory, and that there are no cases that he
discusses of semantic memory: we have “reason to believe that Reid thinks of memory as
being objectual in nature”. However, Copenhaver (2006c) disagrees; she thinks that Reid
is concerned with both types of memory, but that episodic memory is seen by him as being
fundamental. Van Cleve (2013) sides with Copenhaver (2006c), and I do, too. In support
of this, Van Cleve (2013) cites the following passage, where Reid uses a that-clause to
report a memory that he has: “I believe that I washed my hands and face this morning [. . . ]
[because] I remember it distinctly” (EIP III. 1, p. 255-56). Van Cleve goes on to argue that
“it” in the quote is bound by the that-clause in the previous sentence, i. e. “that I washed my
hands and face this morning”, and that, since Reid says that he remembers that he washed
in the morning, he means that he remembers the actual event of washing his hands and
face. Otherwise, Van Cleve argues, Reid would have said that he knows he did it, just as he
always says, when he doesn’t actually remember something, but he is sure that that thing
happened in his past. For instance, Reid says that “I know who bare me, and suckled me,
but I do not remember these events” (EIP III. 4, p. 264).
On behalf of Van Woudenberg (2004), I would like to say that, although I think that
Copenhaver (2006c) and Van Cleve (2013) are right, the evidence for their claim is not that
strong. In the passage quoted above, Reid says first that he believes that he washed his
hands and face in the morning, and not that he remembers doing so. And then he adds that
his belief is supported by the fact that he has a memory of something related to that belief.
10
Semantic memory can also be distinguished from episodic memory at the language level, as Martin
(2001) discusses: semantic memory is reported using “that-clauses”.
135
This indicates that the distinction between the belief component of a memory and a belief
formed on the basis of a memory, which Van Woudenberg (2004, p. 208) draws, may be
correct. Read in this light, this passage would show that Reid is remembering an event –
the washing of the hands and face – and has a belief that he washed his hands and face
formed on the basis of remembering that event. If this is right, Van Woudenberg (2004)
is right that all cases of memory are cases of episodic memory for Reid, and that we can
form explicit beliefs, different from the memories we have, on the basis of the memories
themselves.
For the present purposes, however, it does not matter who is right. The problems I am
about to discuss are engendered by Reid’s arguing that there is episodic memory; whether
he thinks that episodic memory is the only type of memory has little bearing on this issue.
Having established that the main (or possibly only) type of memory Reid is concerned
with is episodic memory, we are now armed to tackle the question about the objects of
memory. It should come as no surprise that events are considered to be objects of memory,
and not facts. After all, this is the domain of episodic memory; facts are usually retained by
semantic memory. This, however, raises the following worry: given that episodic memory
is characterized in part by the Previous Awareness Condition, and given that we remember
events, this entails that we can be the witnesses of (or the agents of) events. This means
that we are capable of perceiving events, if these are events going on in the external world,
or be conscious of mental events, since consciousness operates like perception, except that
its objects are the currently occurring operations of our minds.
The idea that events are objects of some operations of our minds, namely of perception
and of consciousness, is intriguing. Reid dedicates a whole (long) Essay to the study of
perception, and nowhere in that discussion does he mention events as things that can be
perceived. In previous chapters (Chapter 2 and Chapter 3), I have pointed out how difficult
it is to establish that we can perceive physical objects, in addition to the qualities they have,
136
if we are to keep close to the letter of Reid’s text. And now we are asked to add something
else to the list of things that we are able to perceive: “I remember the transit of Venus
over the sun in the year 1769. I must therefore have perceived it at the time it happened”
(emphasis added; EIP III. 1, p. 254). At the beginning of the Essay on memory, Reid just
throws this in: we can perceive events, because we can clearly remember them, and we
would not be able to remember them, unless we perceived them at a previous time.
And what about consciousness? Reid talks about this operation of our mind throughout
the whole book, and the observations he makes are brief and inconclusive. We are told that
consciousness operates just like perception; the only difference being that its objects are
internal, and not external. This is an important difference. The fact that both internal and
external objects can be thought about with the help of operations confined to the present
moment seems to be one of the few things they have in common.
The following question arises: can we remember things that we took notice of via
consciousness? Reid answers this question affirmatively, in a passage already quoted (EIP
III. 5, p. 269). But then, the memory employed to remember a certain type of sensation is
surely going to be episodic, which means that our mental operations, which we remember,
must be construed as being events, and not states, if they are to be objects of episodic
memory. So, just like in the case of perception, consciousness must be able to register
events, if we are to use episodic memory to remember sensations had at a previous moment
of time.
If this reconstruction of Reid’s views on memory is correct, and I am not alone in
thinking that it is, as discussed above, there are several issues it raises. First, there is
a conflict between Reid’s thinking that both perception and consciousness are confined
to the present moment, and that no succession can be either perceived or registered via
consciousness. Second, we do not know anything about Reid’s account of the metaphysics
of events. And this, in turn, makes it difficult to understand how we have access to external
137
objects in memory. Recall that, for Reid, memory must be immediate, in the sense that
there must be no mental intermediaries between the mind and the object of a particular
memory. However, depending on how we understand events, this might not work out. In
what follows, I discuss each of these issues in turn.
4.2.1 AnInconsistentTriad
Reid argues that we cannot be said to actually perceive or be conscious of any kind of
succession, because both perception and consciousness are confined to the present moment.
As we’ve already seen (EIP III. 5, p. 269), he thinks that both perception and consciousness
are always shifting from one object to another, and that these objects is all we are aware of,
and not the continuous moving from one to the next. In order to be able to perceive or be
conscious of succession, we would need to be aware of the shifting itself, but Reid thinks
this is impossible.
It may here be observed, that if we speak strictly and philosophically, no kind
of succession can be an object either of the senses, or of consciousness; be-
cause the operations of both are confined to the present point of time, and there
can be no succession in a point of time; and on that account the motion of
a body, which is a successive change of place, could not be observed by the
senses alone without the aid of memory. (EIP III. 5, p. 270)
I would like to note several issues this passage gives rise to. First, it is a claim made
by the philosopher, who wants to make things precise and thus will think that neither ex-
ternal nor internal succession can be noticed. In fact, Reid adds that this is a departure
from common sense, due to the meaning the philosopher assigns to “present”, which is
different from the one the vulgar assigns to the same word. In the introduction to his book,
Reid argued that departures from common sense and from common usage of words are
sometimes allowed, provided that in doing so the meaning of the words do not become un-
recognizable. Thus, using “idea” to mean anything from perceptions to thoughts, as Hume
138
allegedly does, should not be allowed, because it distorts the common language to a high
degree. Naturally, we should expect that whenever Reid thinks that such departures from
the way the vulgar talk are required, we would be provided with some justification. In this
case, however, Reid is not giving any arguments to support the view that the present time
is confined to a single “indivisible point of time, which divides the present from the past”
(EIP III. 5, p. 270). Moreover, Reid seems to be confused regarding what he is doing here:
he says that the philosopher and the vulgar might seem to be contradicting each other, but
that this contradiction is only apparent. He says that “[i]t arises from this, that Philosophers
and the vulgar differ in the meaning they put upon what is called the present time, and are
thereby led to make a different limit between sense and memory” (EIP III. 5, p. 270). The
contradiction he’s referring to seems to me to be quite real, and not apparent, as he’s sug-
gesting. After all, based on the meaning the philosopher ascribes to the word “present’, it
follows that one cannot perceive succession. Whereas, based on the meaning the vulgar
give to the word “present”, it follows that one can perceive succession.
Another issue I would like to note in connection with this passage is the consequence
of the view that succession cannot be perceived (which Reid embraces): we are unable to
perceive the motion of bodies. He believes that we actually need memory in order to be
able to perceive a body move. According to Reid, we infer that a body has moved, because
we see the position it is currently at, and we remember that it occupied other positions,
for the duration of time we were observing it.
11
The following passage gives the details of
Reid’s view:
[W]hen as philosophers we distinguish accurately the province of sense from
11
As pointed out in Van Cleve (2013), Reid’s view is incompatible with the doctrine of the specious present,
advocated by William James and Bertrand Russell, according to which we have access to a little bit of the
past, which is perceived as though it were present. Thus, the positions a body was at several moments earlier
would temporally belong to this specious present. On this view, there would be no problem with saying that
we actually perceive motion.
139
that of memory, we can no more see what is past, though but a moment ago,
than we can remember what is present; so that speaking philosophically, it is
only by the aid of memory that we discern motion, or any succession what-
soever: We see the present place of the body; we remember the successive
advance it made to that place; The first can then only give us a conception of
motion, when joined to the last. (EIP III. 5, p. 271)
So, motion is not perceived; motion is, at most, inferred. Or, as Reid puts it, we need
memory to know that motion happens. Here we are, at last, equipped to reconstruct the
inconsistent triad Reid’s view gives rise to:
1. People remember bodies moving – “I remember the transit of Venus over the sun in
the year 1769.” (EIP III. 1, p. 254)
2. People only remember what they previously perceived – “Things remembered must
be things formerly perceived or known.” (EIP III. 1, p. 254)
3. It is impossible for people to perceive motion. (EIP III. 5, p. 270)
This is not a devastating objection to Reid’s views on memory; but he should have
noticed that he is reaching this impasse, if he wants to hold on to all of his commitments
regarding perception, consciousness, and memory. Interestingly, I do not think that we
can generate a parallel inconsistent triad for consciousness: the first item on what should
be a corresponding list seems to be missing. As far as I can tell, Reid does not say that
we can be conscious of any internal motion; on the contrary, he stresses that we are con-
scious of particular sensations, and that our mind is fleeting from one thought to the next.
12
These sensations, perceptions, etc. have duration and Reid argues, contra Locke that we
do not derive our notion of duration from registering, via consciousness, any succession
of thoughts in our mind. The notion of duration is antecedent to the notion of succession,
12
Here “thought” is, as always, a very general notion, encompassing all operations of our minds.
140
according to Reid, since even without being aware of any succession, if we focus all our
attention on a specific sensation, at any given moment, we will think that that sensation
took some time, i. e. had a certain duration (EIP III. 5, p. 272).
There are two ways of resolving this inconsistency, as Van Cleve (2013) notes: either
give up (3), and say that motion can be perceived, or liberalize the Previous Awareness
Condition. If Reid were to do the former, he would also need to give up the idea that
perception is restricted to the present, if he wanted to hold on to the idea that the present
is an indivisible point of time. Or, he could give up his view that the philosopher’s notion
of present is different from the vulgar’s one, in which case he could adopt something like
the specious present view. If the present extended a little bit into the past, there would be
nothing to prevent motion from being perceived, even if perception would be restricted to
the present moment.
If Reid wanted to hold on to (3), he could liberalize the Previous Awareness Condition
in the following way, according to Van Cleve (2013): people only remember things that
they have previously experienced. Experiencing would not be restricted to perceiving, but
may be, instead, a compound of perceiving, remembering, comparing and inferring. In
other words, one would be said to experience the motion of a body, if one registered its
motion, in the way Reid suggests: see the different positions of a body on a trajectory,
remember the ones it already traveled through, and infer that it must have moved from one
position to the other.
I think that best option would be to give up (3) by way of reouncing the idea that the
present moment is just one point in time; in other words, Reid should not create an unneces-
sary tension between his philosophy and common sense, on this issue. The other option is
not as attractive. In contemporary philosophy of mind, the Previous Awareness Condition
on episodic memory is more liberal than the one Reid explicitly uses. As Van Cleve (2013)
points out, experiencing is the key notion. However, including memory in the compound
141
of things that experience is constituted by is problematic. Let us take the case of motion,
again, to illustrate the problem. In order to remember the transit of Venus over the sun in
the year 1769 somebody must have experienced the respective transit, at a previous moment
of time. That is, our subject saw several positions that Venus was at, in the course of the
year 1769, remembered that Venus was at several previous positions, during that year, and
inferred that Venus must have traveled from the original position to the one it was at, at the
end of the year. So, memory does double-duty: to register motion, one already uses mem-
ory; and to remember motion, one must use memory twice over. Why is this a problem?
Suppose that one hallucinates some of the positions the body was supposed to travel to over
the course of our subject experiencing its movement. Based on that hallucination, our sub-
ject infers that the body was at p
1
, and then at p
2
, etc. According to the notion of experience
we are working with, this should provide enough material for our subject to remember the
motion of the body in question, at a later date. After all, the body was perceived to be at p
1
,
hallucinated at p
2
, and then inferred to have moved from p
1
to p
2
. This is problematic for
Reid’s account of memory, which is supposed to give us immediate knowledge of the past:
there is no knowledge of the body having moved, in this case, although the motion seems
to have been experienced by our subject, at a previous moment.
Moreover, if inference is used in this way, as part of the experience that falls under the
incidence of the Previous Awareness Condition, it is unclear in what way the knowledge
of the past we have via memory is immediate. One notion of immediacy attributed to Reid
by Buras (2008), and endorsed throughout the present work, has to do precisely with the
non-inferential character of the belief involved in perception and memory. However, if
inference is lending a helping hand here, this notion of immediacy is lost, and its loss is
of consequence for Reid’s anti-skeptical project. Immediacy is key to arguing against the
skeptic; hence, immediacy should not be given up.
142
A consequence of this discussion is that motion of bodies should be added to the list of
things that we can perceive, in addition to bodies and their qualities. To do so, Reid must
only give up his commitment to the philosopher’s notion of the present. Since motion of
bodies can be perceived, if the present moment is more extended than an indivisible point
in time, motion of bodies can also be remembered: the Previous Awareness Condition on
episodic memory is satisfied. Moreover, since motion can be perceived, and remembered,
we should not be surprised by the idea that events can be perceived and remembered.
However, to everyone hastily adding events to the list of things that can be perceived, I
say: “Not so fast!” The reasons for this are discussed in the section below.
4.2.2 TheNeedforAMetaphysicsofEvents
As already pointed out, it is difficult enough to think that we can perceive bodies, on Reid’s
theory of perception. The most direct textual evidence regarding the objects of perception
indicates that qualities of bodies are certainly on that list. The textual evidence supporting
the idea that bodies having those qualities are also objects of perception is inconclusive, but
I argued that Reid needs to have that category on the list. We have just seen, that motion
should be added to the list, since there are no good arguments offered in support of the idea
that the present moment should be restricted in the way Reid says it should. But there are
additional problems to be considered before penciling in events, of all kinds.
These problems have to do mainly with the idea that we can remember everything that
we have once perceived or were conscious of, given that Reid is abiding by the Previous
Awareness Condition on episodic memory. In the case of motion, we will thus remember
the object undergoing movement, and the trajectory of that object itself. We might also
remember several individual positions that object was at, during the course of moving from
the initial to the last position, but this does not seem necessary, once we reject Reid’s
143
restrictions on how to construe the present.
So, here is one problematic case regarding our ability to remember events: perceiving,
and sensing can be construed as events. One is conscious of perceiving something while
perceiving it. While it is true that the object of perception is external, it is also true, on
Reid’s view, that the object of consciousness, in that case, is internal. For instance, if I
am said to perceive the tree outside my window, I will be also said to be conscious of
perceiving the tree outside my window. There is a somewhat clear division here between
the objects of perception and the objects of consciousness. The division is maintained even
when one is said to perceive an event, like this: I witness a car crashing into a wall. The
perception is of the event of crashing, while the consciousness is of the event of perceiving
(the event of crashing). These issues are not discussed by Reid, but this is common sense,
if we want to say that we are able to perceive events unfolding.
What happens with this division between perception and consciousness when memory
comes on the stage? It disappears: one could remember perceiving the car crashing – i. e.
an internal event – or one could remember the car crashing – i. e. an external event. The
only condition that needs to be satisfied is the Previous Awareness Condition, and this is
satisfied in both cases: in the former, one was conscious of the internal event; in the latter,
one perceived the external event. The problem here is that Reid treats these two cases as
one. He is unaware of the distinction and of its relevance. Let us first look at what he says
in this regard:
Suppose that once, and only once, I smelled a tuberose in a certain room where
it grew in a pot, and gave a very grateful perfume. Next day I relate what I
saw and smelled. When I attend as carefully as I can to what passes in my
mind in this case, it appears evident, that the very thing I saw yesterday, and
the fragrance I smelled, are now the immediate objects of my mind when I
remember it. [. . . ] I beg leave to think with the vulgar, that when I remember
the smell of the tuberose, that very sensation which I had yesterday, and which
has no more any existence, is the immediate object of my memory. (IHM 2. 3,
144
p. 28)
In this passage, Reid, intentionally or not, uses the ambiguity he found in the way we
refer to some sensations. For the case of smell, we call “smell” the unknown cause that is
the external cause of our sensing in a certain way, e. g. the physical distribution of particles
in the tuberose; and we also call “smell” the very sensation that we have on the occasion of
being presented with a tuberose. It is unclear whether Reid wants to say that in the case he
discusses we have a memory of the external object, namely the quality of the tuberose, and
a memory of the internal object, namely of the sensation we had on the occasion of being
presented with a tuberose. If I were to take an informed guess, I would say that he is not
careful enough to acknowledge that he changes the subject from the beginning of the quote
to the end of the quote. One lesson we can draw from this passage is that even Reid can
equivocate!
The more substantive lesson I want to draw from this passage is that in cases like this,
it is unclear what the object of memory is. Do we remember the external object and,
simultaneously, the internal object? Do we remember the external object in virtue of our
remembering the internal one? Can we remember the internal object without remembering
the external object, or just the external object, without remembering the internal one?
Reid offers too little in the way of an explanation for us to be able to answer these
questions. However, we can analyze the consequences of answering them in different ways.
For Reid’s theory to work as an alternative to a skeptical system, the object of mem-
ory must be apprehended by the mind in an immediate fashion: no mental intermediaries
can exist between the mind and the object of memory. So, it will not do to argue that, in
order to remember the (once perceived or sensed) external object, one must remember the
perception itself. In other words, for Reid’s theory to not be subject to some of the skep-
tical challenges, by his own lights, we cannot say that I remember smelling the fragrance
145
of the tuberose, and thus I remember the fragrance of the tuberose. I must be able, for
the sake of Reid’s anti-skeptical commitments, to remember the fragrance, and, possibly
simultaneously, to remember smelling the fragrance of the tuberose.
The question, of course, is whether this is possible. Recall that Reid is primarily con-
cerned with episodic memory, which involves events. The fragrance of the tuberose can-
not be understood to be an event, in its own right; whereas smelling the fragrance of the
tuberose can be so understood. It thus appears that we have a problem: if Reid is talking
about episodic memory, then we need mental intermediaries, to give us the correct object
of this type of memory, namely events. In the case of the tuberose, the remembering mind
cannot reach the fragrance, in the absence of remembering the act of smelling. This is bad
for Reid, because we lose the relevant immediacy of the memorial experience.
I think that this problem can be solved as follows: there should be no problem with
construing the objects of memory as events, as long as we have a metaphysics of events. It
is usually held, in the literature on events, that the objects involved in a particular event are
essential for how we understand where the event in question begins and where it ends.
13
For instance, every soldier involved in a battle is essential to that battle, as are the weapons,
the hills, and the blades of grass on the field.
I argue that we can adopt this position, on Reid’s behalf: what one remembers are events
constituted by the original internal object (e. g. having the sensation of smelling) and the
original external object (e. g. the fragrance of the tuberose). The sensation or perception in
question is not to be understood as playing the role of a mental intermediary, but rather as
being a constitutive part of an event. In this way, whenever one says things like “I remember
the tuberose you showed me yesterday”, one should be understood as meaning something
like “I remember seeing / smelling the tuberose you showed me yesterday”. This works in
13
For more on the relationship between events and the objects involved in them, see Lombard (1986) and
Parsons (1991).
146
more complicated cases, too. For instance, assume that someone witnessed a car crashing
into a wall, and then his testimony is heard in a court of law. What the witness is going to
relate are his memories of seeing the car moving from a certain position to another, at high
speed, etc.
I think this is the only way in which Reid can solve the issue. By having a meta-
physics of events, that involve the original act of perceiving or sensing the external object
or event, he can argue that the remembrance itself is immediate; it is not mediated by men-
tal intermediaries. Of course, this assumes that events can be perceived or introspected
via consciousness, otherwise the Previous Awareness Condition cannot be satisfied. As we
have seen in the previous section, this is something that Reid’s view doesn’t immediately
allow for. However, given that otherwise he reaches an inconsistent triad, I argued that he
should give up some of his commitments, and allow for events to be both perceived and
introspected.
4.3 Conclusion
In conclusion, I would like to recapitulate the main findings of this chapter. Reid dedicates
an entire Essay to the faculty of memory and he argues that it functions in much the same
way as perception does, with an important exception. Whereas perception gives us im-
mediate access to presently existing objects, memory gives us immediate access to objects
existing in the past. He talks mainly about episodic memory, just as other Reid scholars
have argued. However, by contrast to other Reid scholars, I identified several points of
tension in Reid’s theory of memory.
First, I discussed what type of conception is an ingredient or concomitant of mem-
ory. If my arguments are correct, memorial conception is closer to imaginative conception
than to perceptual conception. I have thus argued that memorial conception is conceptual,
147
even though it is based and probably derived from perceptual conception, which is non-
conceptual. This result was instrumental in establishing what are the objects of memory.
Second, I investigated how coherent is Reid’s discussion of what are the objects of
memory. As other Reid scholars have argued, there is a necessary condition that must
be satisfied, in order for memory to function in the way Reid says it does: the Previous
Awareness Condition on episodic memory. Although strictly speaking, Reid only talks
about perception as giving the scope of the Previous Awareness Condition, I argued that he
needs more. He needs to say that one can remember not only things previously perceived,
but also things previously sensed (like fragrances, etc.) and things previously introspected,
via consciousness (like acts of smelling, witnessing, etc.) If I am correct in arguing that his
original position is inconsistent, the only modification that would resolve that inconsistency
is to have a particular kind of metaphysics of events, according to which the original mental
faculty is a constitutive part of the event remembered.
It remains to be seen if such a metaphysics can be constructed, but this is the topic
of a different project. All we need to know for now is that Reid does not have such a
metaphysics and he needs one in order to dispel the inconsistent triad.
148
Chapter5
ImaginationandItsNonexistentObjects
5.1 Introduction
Thinking about the nonexistent is difficult. This difficulty becomes twofold when one
aims to explain how it is possible to think about different nonexistents, and it seems to
increase even more when somebody tries to think about different nonexistents of the same
kind. And yet, this level of difficulty does not seem to stump avid readers, movie-goers,
or opera buffs; it seems easy to think of Sherlock Holmes and Watson, even though they
are nonexistent people. However, several questions could easily be asked: what is it that
the name “Watson” refers to? How is it possible to distinguish between Ronan and Bane,
two of the centaurs of the Forbidden Forest at Hogwarts? These, and related questions,
have been asked and answered throughout the history of philosophy, and various theories
explaining how to understand, interpret, and appreciate fiction have been offered.
1
As shown in Chapter 1, the content of imaginative experiences, according to Reid, is al-
ways conceptual. This is in sharp contrast to perceptual experiences, whose content can be,
and almost always is, non-conceptual. After establishing what exactly are the contents of
1
Some of the most recent, but also classic, texts that come to mind are Kripke (2013) and Parsons (1980).
149
perceptual and memorial experiences, in the previous chapters, I now turn to investigating
the content of imaginative experiences. There are two questions this chapter will answer:
What objects can we imagine? and How do we imagine the things we imagine?
As a preview, the answer to the first question should be straightforward: we can imagine
existing and nonexisting objects. For Reid, imagination proper is a species of conception. It
“signifies a conception of the appearance an object would make to the eye, if actually seen”
(EIP V . 6, p. 394). As it will become clear later in the chapter, this notion of imagination
is too restrictive. I argue that Reid would have no problem with extending it to include
non-visual imaginative experiences, like aural or tactile. The crucial aspect of imagination
that must be preserved is that it goes along with perception, in a very literal sense. Just
as visual imagination is about capturing the appearance an object would make to the eye,
if seen, aural imagination would capture the appearance an object would make to the ear,
if actually heard. This will be discussed in more detail, but it is good to know from the
beginning what sorts of things can be imagined. Another aspect of imagination, according
to Reid, is that it constrains the imaginer to entertain a thought about something, without
any belief regarding the existence of that thing, since it is a type of bare conception, or
simple apprehension. In this, imagination differs from both perception and memory.
The answer to the second question is more complicated, due in part to Reid’s not being
very explicit on this issue. The greater part of this chapter is dedicated to showing that
on a consistent reconstruction of Reid’s theory of imagination, one can entertain singular
imaginings of both existent and nonexistent objects. Indeed, once we understand how
closely related imagination and perception are, according to Reid, we have all the tools to
understand how it is possible to imagine three distinct individual centaurs. The very idea
that one can imagine categories of things that one cannot perceive, such as the property
of knighthood or of elfhood, or simply a proposition, will be shown to be nonsensical,
on this theory. Our minds have the power of entertaining propositions or thoughts about
150
non-perceptual properties (and objects), but that power, which Reid calls “conception”, is
different from imagination.
In what follows, I present and evaluate three different theories that explain what we
imagine when we imagine Pegasus, Don Quixote or the land of Oceana. These theories
purport to throw light on the nature of entia rationis or fictional objects. Once we know
what kinds of things one is supposed to imagine, according to Reid, we can understand
whether the act of imagination itself is singular or general. To have a singular thought,
according to Reid, means no more and no less than to be able to think about an individual
object, while to have a general thought means that the object thought about is general,
namely a universal (EIP V . 2, p. 360 & 364). Reid does not employ the terminology of
singular thought versus general thought, but the distinction he draws between individual
conception and general conception maps nicely onto it, since conception is a necessary
ingredient of every other operation of the mind (EIP IV . 1, p. 295). One cannot have a
belief about something without conceiving that thing as well; and the belief will be said
to be individual or general, depending on whether the conception employed is singular or
general.
This connection to singular thought is quite important, and will closely guide the recon-
struction of Reid’s theory. He stresses how important it is to understand that imagination
works almost like perception, with the exception that it can be employed about nonexistent
objects. So, allowing for singular imaginings is key to understanding his theory. Thus,
out of the three theories about to be discussed, the one that allows for such imaginings is
to be preferred. Another criterion of judging which theory should be attributed to Reid
is the principle of charity, so dear to historians of philosophy. If we find that attributing
a certain view to Reid has the consequence of him holding inconsistent views, we have
a strong indication that we have to look for an alternative interpretation, provided that he
is not explicitly contradicting himself. A welcome result of charitable interpretation, of
151
course, is that by engaging in it we also come to appreciate the strengths and weaknesses
of alternative theories. So, the purpose of this chapter is not only to argue that a certain
theory of imagination is more suited to be attributed to Reid, given the textual evidence
we have, but also to understand what it would take for a theory of imagination to allow for
singular imaginings to exist.
Here are the three theories, briefly. First, one could argue that ficta are bundles of uni-
versals, especially if one thought that actually existing things are such bundles. I call this
“the complex universal view”, since a fictional object is thought to be composed of many
different simple universals, put together by the imaginer’s mind, in an act of imagination.
Throughout the history of philosophy, there were several representatives of this view. Ac-
cording to the standard interpretation of Reid, advocated, among others, by Gallie (1997)
and Wolterstorff (2001), this is the view that he held. Other representatives include Bertrand
Russell, with regards to existent objects, and proponents of Meinong-inspired theories of
fiction, such as Terence Parsons (Parsons (1980).) Although the standard interpretation is
identifying this as Reid’s view, I will show that, on such a reconstruction, his views are
inconsistent.
Second, one can hold that there are such things as fictional objects and fictional charac-
ters, which are not concrete objects, like real flesh and blood people, but abstract, just like
numbers, or nations. This theory will be referred to as “the abstract object view”, and it is
advocated primarily by Saul Kripke (Kripke (2013)) and Nathan Salmon (Salmon (1998)).
No Reid scholar has argued that he held this view with regards to what ficta are, but, based
on the text, something close to this view might be attributed to him. No inconsistencies
emerge, on this view, but also no explanation of how imagination works. Since Reid is
mostly interested in talking about the active role of the mind in exercising its faculties,
and not primarily in fictional discourse, a different view is better suited to reconcile all his
commitments.
152
Third, the direct opposite of the first view presents itself as an alternative: “the complex
trope view”, according to which ficta, but not real existing objects, are bundles of tropes,
or abstract particulars. The asymmetry between existing and nonexisting objects should
not be regarded as problematic: it is not surprising to argue that there is something that
holds the tropes together in the case of existing objects, which is missing in the case of the
nonexistent. This something is the substance that has those tropes as qualities; once we
abstract away and think about the tropes distinctly from their worldly home and combine
them in our imagination, the substance disappears. The only thing left is the combination
of abstract particulars, reached at by a process of abstraction from the tropes inhering in
the substances of the world. On the reconstruction offered here, Reid is the proponent of
such a view, with regards to ficta. The advantages of this view over the other two will be
fully discussed in what follows, but to whet the reader’s appetite, I will say that on such a
view singular imaginings of nonexistents are possible, and maybe even encouraged.
The structure of the chapter should be obvious. Each of these theories will be discussed
in detail and evaluated for their consistency and accuracy vis-a-vis Reid’s overall views.
5.2 FictaasComplexUniversals
5.2.1 TheView
How does one construe nonexistent objects? The most important aspect of a fictum is its
blatant nonexistence; since one cannot look around oneself to see what a unicorn looks like,
one must rely purely on one’s own mind to envisage such a fictional animal. Granted, there
are stories about unicorns and there are images (usually paintings or drawings) of unicorns
that can help someone think about such a magical creature. But, given that one is told to
think about, or imagine, a unicorn, without knowing any of the painting or stories, what
153
is one supposed to do? Usually, there are some instructions given: imagine a white horse
with a horn attached to its forehead. And then, it’s easy: all one has to do is think about
a white horse (any old one) and a horn and combine them in their mind to think about a
unicorn.
Is that unicorn an object, in the way a certain horse is? That depends on the particular
view about objects that one has. If one thinks that there is no more to a real existing thing
than its properties, then yes, a unicorn is an object, almost in the same way a horse is. This
is an old dispute, that Hume and Reid (among others) were divided by. Hume is the well-
known champion of the bundle theory, with regards to both material objects and immaterial
minds, whereas Reid is an advocate of the old Aristotelian view that there are substances
which have properties. However, for both Hume and Reid, the properties in question are
thought of as tropes, and not universals.
2
The view under consideration here proposes that the properties in question are univer-
sals, not tropes, and that real existing tables, chairs, and humans are nothing more than
collections of such properties. Russell, who was a supporter of this view at one time,
thought that things are bundles of universals, existing in a certain place, at a given moment
of time:
[W]herever there is, for common sense, a “thing” having the quality C, we
should say, instead, that C itself exists in that place, and that the “thing” is to
be replaced by the collection of the qualities existing in the place in question.
(Russell (1967, p. 98))
Russell’s view is essentially eliminativist. We are acquainted with universals, which
do exist, but not with objects. In other words, he restates the Humean thesis that once an
object is stripped bare of its properties (which for Russell are universals, not tropes), there
2
Hume argues against general abstract ideas, saying that even if one wants, for instance, to think about
a triangle, in general, what one does is to bring to one’s mind a particular triangle, as a representative of its
kind (T. 1. 1. 7. 6; SBN 20).
154
is nothing left; there is nothing more to the object that the bundle of properties our senses
inform us of existing at a certain place, in a certain time.
This view can be easily adapted to talk about fictional objects. It is surprising that
Russell himself didn’t do this. One reason for his hesitation may be his thinking that it is
essential to talk about universals being tied to a place and time for construing what he calls
“things”.
3
And to adapt the bundle-theory view of things to ficta one must be allowed to
scratch out all talk of properties being tied to a spatiotemporal location. To construe ficta
in this way, one must abstract from the real existing universals that one encounters in one’s
day-to-day life and use these abstract notions to make the corresponding fictional bundle.
An interesting thing to notice here is that on such a view, ficta are not actually things. To
be a thing, you need to be made up of universals which are in fact tied to a place and a
time; without this, you are just a bundle of abstract notions. On this view, then, a unicorn is
not an object, whereas a real horse is, even though, on the surface, they are made up of the
same stuff: universals. This is important here, and it will be as important to keep in mind
when we consider the other two views.
To finalize the adaptation of the bundle theory of objects to ficta, one should notice that
in talking about a centaur, one would think about that centaur as being a set of universals.
4
On this picture, there is nothing more to that centaur, which is a set, than its members,
namely universals. For instance, one could construe such as set as having the following
members: the property of having a horse’s body, together with the property of having a
3
Without tying universals to a time and place, Russell’s bundle-theory would be open to an objection
raised by James Van Cleve: “If a thing were nothing more than a set of properties, any set of properties would
fulfill the conditions of thing-hood, and there would be a thing for any set. But in fact there are many sets
without corresponding things – e. g. the set being an alligator, being purple” (Van Cleve (1985, p. 95)). This
might be a reason why Russell didn’t extend his theory to fictional objects.
4
I use the notions of sets here, because sets are fairly well understood. There are other options here:
classes, fusions, or perhaps bundles, understood as a special kind of collection, might do as well. Choosing
between these options is not trivial, but it remains true that thinking about a centaur consists in thinking about
a construct made up of universals.
155
human torso, together with the property of being an animal, etc. In other words, whenever
one thinks about a centaur, on this view, it is part of one’s thinking to think about the
complex property of centaurhood. This extension of the bundle theory of objects to fictional
objects and characters is quite natural, especially if one thought that real existing things
were nothing more than bundles of universals. Meinongian theories of fictional discourse
construe fictional objects in this manner, although these theories are not explicit bundle
theories. On a Meinongian account, what is important is that to any set of properties there
corresponds an object, and not whether there is more to an object, be it real or fictional,
than the properties making up that set.
5
5.2.2 DidReidHoldThisView?
According to the Standard Interpretation of Reid, he did hold the view that entia rationis
are nothing more than complex universals, made up by combining several simple universals
into a whole that represents to the mind the fictional object or character under considera-
tion. This interpretation is defended, among others, by Gallie (1997) and by Wolterstorff
(2001). Wolterstorff (2001, p. 73-74) argues that, although Reid’s language might remind
the reader of Meinong and his idea that the substances that exist are a subset of those that
have being, one should not be too hasty in attributing this view to Reid. What Wolterstorff
argues against here is the idea that fictional characters and objects should be construed as
substances (which is what actual objects are). And, if these “things” are not objects, what
can they be? Here’s Wolterstorff’s answer: “[f]ictional characters, fictitious beasts, plans
for unbuilt buildings – all are, on [Reid’s] view, not nonexistent particulars but complex
universals – person-types, animal-types, building-types, etc” (Wolterstorff (2001, p. 74)).
On this interpretation, whenever one wants to think about a centaur, what one needs
5
This was discussed in some detail in Chapter 1. Here, I will only explain what are some of the problems
of holding such a view, both as an interpretation of Reid, and in general.
156
to think about is the property of centaurhood, in general, which is nothing more than a
bundle of simple properties, held together by the one conceiving the centaur in question.
To conceive a centaur, thus, requires no more than to conceive the meaning of the word that
expresses that attribute in language. This is true about more important concepts, like felony
or justice, and it should be true about the conception of centaur too. After all, felonies and
centaurs have the same degree of "concrete" existence: zero.
6
This interpretation is prima facie supported by the text. Reid argues that our concep-
tions are of three kinds:
They are either the conceptions of individual things, the creatures of God; or
they are conceptions of the meaning of general words; or they are the creatures
of our own imagination; and these different kinds have different properties
which we have endeavored to describe. (EIP IV . 1, p. 305)
On the Standard Interpretation, the last two kinds of conceptions are grouped together,
in the idea that their objects share something that is essential to them: they do not exist.
Even though universals and creatures of our fancy do not exist in the way in which concrete
particulars that can be perceived exist, they all can be thought about, and the mind can
reason about them. One can have a conception of triangles, in general, and of the color
white, in general, without tying that conception to any instance of whiteness in particular
(e. g. the whiteness of the sheet of paper I’m writing on). This is one way in which “bare"
conception differs from both perception and memory: one can conceive things, without
even considering the question of their existence. As Wolterstorff (2001, p. 69-70) puts
it, to exercise one’s bare conception means nothing more than to mentally apprehend (or
entertain) an object of thought. One can do this, even though that object does not and will
not ever exist.
6
Reid talks about this issue in EIP IV . 1, p. 303, and his arguments there are reminiscent of what Locke
thought about mixed modes (Essay, II. xxii. 1-2).
157
Moreover, Reid argues that it is not just simple universals that one can conceive of, such
as those derived, by abstraction, from the particular colors one perceives. One can conceive
products of universals, that is, things that do not and may never be instantiated: “one may
conceive a machine that never existed. [. . . ] Such compositions are things conceived in the
mind of the author, not individuals that really exist” (EIP V . 4, p. 375).
In addition, this is an attractive interpretation, given that one might think that nonexis-
tence goes hand-in-hand with generality. After all, Reid argues, just like Locke before him,
that everything that exists is an individual. He does not explicitly say that everything that
does not exist is a universal, but he does argue that universals, because they are general, do
not exist. Here is one place where Reid makes that point:
[E]very mathematical figure is accurately defined, by enumerating the simple
elements of which it is formed and the manner of their combination. [...] It
is not a thing that exists, for then it would be an individual; but it is a thing
conceived without regard to existence. (my emphasis; EIP V . 4, p. 373)
Wolterstorff (2001) and Gallie (1997, p. 321) explicitly endorse the idea that if some-
thing does not exist, then it is a universal, since if it were to exist, it were an individual.
They do more, however, than just draw this conclusion: they assume that there can be no
nonexistent individuals, and that all “conceptions of things that don’t exist – imaginings
of such things – are just special cases of general conceptions” (original emphasis; Wolter-
storff (2001, p. 74)). Wolterstorff (2001) presents as evidence for this claim the following
passage:
Some general conceptions there are, which may more properly called compo-
sitions or works than mere combinations. Thus one may conceive a machine
which never existed. He may conceive an air in music, a poem, a plan of ar-
chitecture, a plan of government, a plan of conduct in public or in private life,
a sentence, a discourse, a treatise. Such compositions are things conceived in
the mind of the author, not individuals that really exist; and the same general
conception which the author had may be communicated to others by language.
158
Thus, the Oceana of HARRINGTON was conceived in the mind of its author.
The materials of which it is composed are things conceived, not things that
existed. [. . . ] And the same may be said of every work of the human under-
standing. (EIP V . 4, p. 376)
I agree with the Standard Interpretation that, according to Reid, bare conception works
in this way: one may conceive complex universals, which do not exist, and which are
composed, in the mind of the conceiver, from simple universals, which are derived from
the individuals one is acquainted with, in perception.
7
However, one should pay attention
to the things Reid lists in the passage quoted above: they are all abstract things that can be
communicated through language, like sentences, discourses, and treatises. I do not think
that the list in question is exhaustive, but one should note that the passage in question does
not talk about what it is like to conceive Don Quixote or Pegasus. Neither does it say
anything about what it is like to imagine such fictional characters.
The Standard Interpretation of Reid makes a good case for thinking that centaurs, uni-
corns and the like are just like machines that never existed, or plans of government never
put into practice, namely complex universals, that will be thought of by exercising our bare
conception. Compelling as this interpretation may be, however, I show that it is incorrect:
it fails to see that it attributes a gross inconsistency to Reid and it also fails to appreciate
the strength of his system, dedicated to explaining human faculty psychology.
5.2.3 ProblemsWiththeView
In what follows, I talk about two types of problems this view has: first, I argue that this
is not Reid’s considered view, contra the Standard Interpretation. Second, I argue that this
view comes with serious philosophical problems that should make anyone interested in
7
See, for instance, what Reid says about this in EIP V . 4, p. 376: “Nature has given us the power of
combining such simple attributes, and such a number of them as we find proper.”
159
fictional discourse and imagination dig deeper and find a better explanation of how we are
capable of thinking about nonexistent objects.
Reid’sViewisDifferent
The main flaw of the Standard Interpretation of Reid is that its supporters do not acknowl-
edge that Reid distinguished between (bare) conception and imagination proper, which he
sees as a subspecies of conception. On behalf of the Standard Interpretation, I should point
out that Reid does not spend too much time on explaining what imagination is and how it
differs from conception. However, Reid is interested in understanding the specific office
of each faculty of the human mind. I will not argue that imagination is a different faculty
from conception – it is just as well if we think about it as being a sub-faculty. However, for
someone interested in singular thought and in how our minds are capable of thinking about
individual objects, it is important to discuss this distinction and account for it.
To begin, Reid argues that universals cannot be imagined; they can, however, be con-
ceived. Here is what he says about imagination: “[It] signifies a conception of the appear-
ance an object would make to the eye, if actually seen. An universal is not an object of any
external sense, and therefore it cannot be imagined; but it may be distinctly conceived” (my
emphasis; EIP V . 6, p. 394). This runs contrary to Wolterstorff’s move, who thought that
“bare conceptions of things that don’t exist” are equivalent to “imaginings of such things”
(Wolterstorff (2001, p. 74)). This is strike one against the Standard Interpretation.
In addition, Reid argues that creatures of fancy, like centaurs, can be imagined. It would
be indeed surprising to argue that fictional entities cannot be imagined, since that would run
contrary to common sense. Reid calls fictional characters “creatures of imagination” (EIP
IV . 1, p. 301) and he endorses a passage from Berkeley: “I can imagine a man with two
heads, or the upper parts of a man joined to the body of a horse” (Berkeley, A Treatise
Concerning the Principles of Human Knowledge, Introduction, 10, p. 68; cited by Reid,
160
EIP V . 6, p. 393-94). So, centaurs can be imagined and, since universals are not the kinds
of things that can be imagined, centaurs must be different from universals. The only other
option is to think that centaurs and the like are individuals, even though nonexistent ones.
I will come back to this later in the paper; for now, I would like to emphasize that imagina-
tion, although a subspecies of bare conception, when taken “in its strict and proper sense”
(EIP V . 6, p. 393), cannot have universals as its objects.
Another important point that the Standard Interpretation misses is that it is a common
practice to use proper names to talk about fictional characters and object. “Pegasus” and
“Don Quixote” are examples of such names, known by many of Reid’s eighteenth century
readers and interlocutors. On the assumption that fictional entities are to be construed as
complex universals only, this is problematic, because Reid thinks that proper names can
only be used for individuals, and not for universals, which by their nature are general.
Following the Standard Interpretation of Reid results in an inconsistent triad:
1. Some entia rationis are bearers of names (e. g. Pegasus; don Quixote; Oceana):
We can give names to such creatures of imagination, conceive them distinctly,
and reason consequentially concerning them, though they never had an exis-
tence. They were conceived by their creators, and may be conceived by others,
but they never existed (EIP IV . 1, p. 302).
2. Proper names are linguistic devices used to “express individuals in language”, and
anything that is so expressed is an individual:
An individual is expressed in language either by a proper name or by a general
word joined to such circumstances as distinguish that individual from all others;
if it is unknown it may, when an object of sense and within reach, be pointed
out to the senses; when beyond the reach of the senses, it may be ascertained
by a description, which, though very imperfect, may be true and sufficient to
distinguish it from every other individual (EIP IV . 1, p. 303).
3. All conceptions of entia rationis are general (according to the Standard Interpreta-
161
tion).
There are two ways out of this bind, so far as I can see: accept that Reid is inconsistent
in his treatment of conception and imagination of fictional creatures, or reject the Standard
Interpretation.
One might worry that Reid’s view is indeed inconsistent. This is not what the philoso-
phers advocating the Standard Interpretation say, but, if we look at the beginning of one of
the passages quoted, it may seem that Reid equates conception with imagination:
There are conceptions which may be called fancy pictures. They are commonly
called creatures of fancy, or of imagination. They are not the copies of any orig-
inal that exists, but are originals themselves. Such was the conception which
SWIFT formed of the island of Laputa and of the country of the Lilliputians;
CERV ANTES of DON QUIXOTE and his Squire; and Sir THOMAS MOORE
of that of Utopia. (EIP IV . 1, p. 301)
If we compare this passage to the passage where Reid calls the conception of a machine
that never existed a general conception (EIP V . 4, p. 375), we might think that he is being
inconsistent. On the one hand, he thinks that these conceptions are conceptions of univer-
sals, and hence they cannot be imagined. On the other hand, he thinks that Laputa and Don
Quixote are creatures of imagination, and thus they can be imagined. I argue that he is not
inconsistent, because some of these things can be conceived, as general types, in the man-
ner suggested by Wolterstorff (2001), but some of them can also be imagined, as individual
nonexistent things. There is nothing preventing Reid from endorsing this picture: Cervantes
might have conceived of a medieval knight, doing certain deeds, but to put “flesh” on his
conceptions, he imagined a particular knight, doing particular deeds, and even named him
“Don Quixote”, as any person would with a sufficiently well-contoured individual. The
main point of contention here is that, according to the Standard Interpretation, one is not
able, on Reid’s understanding of conception, to imagine a particular nonexistent fictional
162
character. To see that this is indeed possible on Reid’s view, we must take a closer look at
his theory of imagination, which I will do, in the last section of this chapter.
Once the distinction between conceiving and imagining a nonexistent thing is properly
understood, it becomes clear that Reid is not inconsistent; so the Standard Interpretation
should be rejected, since it is attributing an inconsistent view to Reid, where no such view
is warranted. The Standard Interpretation should be replaced, and I will explore two alter-
natives in what follows. But first, I would like to discuss some philosophical problems that
the view under discussion has.
PhilosophicalProblemsWithThisView
There are additional reasons that do not recommend this view. We should then be cautious
in attributing it to Reid, in the absence of clear-cut textual evidence that he endorsed such
a view. So far, the textual evidence does not warrant such an interpretation, but one might
think that independent philosophical reasons do. This section shows that philosophically
speaking, this is a problematic view, due to several considerations.
One immediate problem with this view is to know what exactly is the complex universal
denoted by “centaur”. If we think about the complex universal as a cluster of simple uni-
versals, this can be represented with the tools of set theory. Let us suppose that we have a
certain set, which contains different simple properties, which stands for (or is) the complex
property of centaurhood. This is how Meinongian-inspired theories of fiction represent the
universe.
8
The problem here is that, depending on which simple properties one adds to
the respective set, one can have different complex properties of centaurhood. For instance,
there could be a property of centaurhood that includes color among its specifications, which
would be different from the property of centaurhood that does not include color among the
8
See, for instance, Parsons (1980) and Deutsch (1991).
163
universals specifying it, etc.
One might try to solve this issue by arguing that there are certain properties that are
essential to how we think about centaurs; those and only those should specify the property
of centaurhood. Allegedly, this is how we decide whether to count a three-legged dog
as belonging to the class of all dogs, because it has all the essential properties of dogs.
The property of having four legs does not belong to that set of essential properties, so one
should not care about its missing one leg, if one is interested only in classifying it correctly.
If one thinks that existing objects are just clusters of universals, the problem of identifying
the subset of essential properties is very acute: without knowing what are the essential
properties that specify the complex universal dog, how is one supposed to know which
individuals to classify as dogs? The other option would be to follow Locke (Essay III.
iii-iv) in thinking that we do not know what is the real essence of things, and we classify
them according to their nominal essences, which are entirely socially constructed, and in
permanent change.
For the property of centaurhood, the problem seems to be less pressing: there are no
centaurs, and no real essence of the fictional species to speak of. Or, one could think that the
real essence and the nominal essence coincide here, just like in the case of abstract notions,
like justice, felony and triangles, to name a few. However, centaurs are modeled after horses
and other real existing animals, and not after triangles and mathematical notions. This is
important, since it might be easy to define mathematical notions, and even other notions
that seem to be less precise (like justice, or felony), but it is quite complicated to define the
property of horse-hood, namely to specify all and only the properties belonging to the set
that would precisely separate all horses from the non-horses of these world.
9
9
Some may think that something should be said to be a horse only if that thing has a specific DNA
sequence – that is the real essence of an individual; and all individuals with that type of DNA should be
grouped together. This is a view defended initially by Kripke (1980, p. 156-58). Without getting into too
many details, this would not work, for the same reason that the original position doesn’t. The equine genome
164
In the case of the property of centaurhood, one is told that it is crucial to have at least
two (simpler) properties in there: that of being a torso of a man, together with that of being
the body of a horse. But if we do not have a good understanding of how to define horse-
hood, we will not have a precise way of specifying a complex universal that is based, at
least in part, on such a definition. Our difficulty in knowing the real essences of things
will transfer to any attempt of specifying the real essence of fictional creatures. It may
seem that anything goes, then, in the case of centaurs, but if we end up with different sets,
with several different members, it will not be possible to say which one of those sets is
the specification of the complex universal centaurhood. To do so, we would need to know
the real essence of man-torsos, the real essence of horse-bodies, and the way of combining
these two elements to “create” a centaur. The discussion above, however, should indicate
that this is extremely difficult, and on some views, even impossible.
This problem is not original with this particular way of construing nonexistent objects.
Since fictional creatures are, to a large extent, indeterminate (and clearly incomplete), it
is to be expected that one will not have an easy time specifying the property of centaur-
hood. One way to eliminate this problem is to embrace it and argue that just like there are
many different types of triangles, there are many different types of centaurs: colored ones,
non-colored ones, three-legged, non-legged, etc. So, we have different sets, with different
members, and we think of each of these sets as specifying some property of centaurhood.
One obvious problem with this solution is that it is quite arbitrary. This is not how we
normally speak, and if we want to track natural language and its uses in fictional discourse,
we should do better. We usually say that we conceive centaurs, without having a precise
way of specifying the complex universal centaurhood, in the same way in which we do
contains a multitude of genes, and some of them might have great mutations that would normally disqualify
an individual having them from being classified as a horse. So, where does one draw the line: 3 mutated
genes do not constitute a problem, but 300 do? Or are some genes “more essential” than others?
165
for the complex universal triangle. It is true that we can think about an isosceles triangle
as opposed to a regular one, but we know what the two have in common: triangularity.
Whereas, if we try to say that I am thinking about a colored centaur, and not about a three-
legged one, it is not so easy to say what the two fictional objects have in common. This is
what we are actually after: the property of centaurhood and an explicit way of specifying
it.
Moreover, this philosophical problem puts additional pressure on the Standard Inter-
pretation of Reid. That interpretation relies on our being able to think about the property
of centaurhood, in general, and not about a family of properties, that may have some at-
tributes in common. After all, the supporters of this interpretation argue that thinking about
a centaur is just like thinking about felony or justice, which can be specified. One might
think that it is difficult to specify what exactly is justice and felony; and I would agree.
However, justice is nothing more than we, as people living in a community, decide to be,
and the same goes for felony. Justice only has a nominal essence (or, we might think that its
nominal essence coincides with the real one). This is not so for centaurhood: this property
inherits its imprecision from the imprecise properties that are bundled together to form it.
The nominal essence and the real essence come apart for horses, and, of course, they do
for bodies of horses, too. There is a greater difficulty that one encounters here: it is unclear
what the family of properties have in common and how we should specify their shared
traits. This is, at least in part, due to the connection that the property of centaurhood (or
the family of properties of centaurhood) has with the property of horsehood (or the family
of properties of horsehood).
Another problem with this view is that one cannot entertain any singular thoughts about
fictional characters and objects. On this view, in thinking about fictional entities, one thinks
about complex universals. There is no particular centaur that one thinks about when one
reads the Harry Potter novels; there is just a bunch of properties, including the bunch which
166
is the property of centaurhood. One should recall what Russell said in the passage quoted
at the beginning of this section: instead of saying that a thing is red, one should actually
just say that “there is redness here, now". To give the logical form of such a statement, one
would need to employ the technical apparatus of definite descriptions. By the time Russell
adopts the view that things are bundles of universals, he is not the champion of singular
thought, although he is credited with making that notion precise.
10
In the case of fictional entities, the fact that one can employ definite descriptions to talk
about them might seem like a god-send: (almost) everyone has access to the universals
denoted by the predicates making up the descriptions, and no one has access to anything
more, because there is nothing more. Centaurs, unicorns, and the like do not exist, so the
best we can do is to use non-denoting definite descriptions to talk about them. But once we
use this mechanism, we are in the realm of general or de dicto thought, instead of that of
singular or de re thought.
However, as Reid was keen on noting, this is not how we talk about fictions and the
characters and objects they contain. We do use names to name fictional entities, and even
if we are not strict Millians about proper names, we do seem to be interested in the well-
being of the heroes and heroines, and not in how several universals are combined to give the
“meaning” of a fictional name. At least on the surface, our thoughts about Desdemona have
the same structure as our thoughts about Thomas Reid: both types are singular thoughts
about some particular entities. It is probable that this singularity is partly the effect of
our using proper names to talk about such entities. This is what Reid thought, and unless
we have good reasons and a principled way of distinguishing between real proper names
and fictional ones, we should agree with him. In the last section of this chapter, when I
discuss what I think is Reid’s considered view, I will say more about how one can entertain
10
One can make the case that Russell still accepted singular thoughts about sense-data, but these are not
the kinds of singular thoughts that interest us here.
167
singular thoughts about the nonexistent. For now, I just want to emphasize that it is not
likely that we can think about three different nonexistents of the same kind (e. g. three
different centaurs talked about in the same story) without employing any notion of singular
thought and proper names. Moreover, if we think that natural language behaves uniformly,
there is hardly any reason recommending a rift between regular and fictional discourse, just
because we don’t know how to understand the idea that we can entertain singular thoughts
about fictional entities.
A third problem is the so-called “problem of creation”, which is problematic for Meinon-
gian theories of fiction, in particular. The complex universal view is essentially elimina-
tivist. On this view there are no fictional objects (just like there are no real things, either).
There are only bundles of properties, which are understood to be general, and not tropes.
Fictional objects and characters are not referred to by the use of proper names; as explained
above, whenever there is apparent reference to such a thing, we actually have a “hidden”
complex definite description that is employed to talk as if we had real things behind our
words. Such descriptions are, of course, non-denoting. Given these assumptions, it is
difficult to see how one can argue that authors of fiction create fictional characters and ob-
jects, when they seemingly write about them; there are no such things, so the creative pro-
cess must result in something else. And according to philosophers like Nelson Goodman,
Kendall Walton, and Harry Deutsch, authors are thought “to create things like sentences,
descriptions, stories, Pickwick-pictures, props in games of make-believe”, depending on
the specific details of the theory under consideration.
11
One might think that certain Meinongian-inspired theories of fiction do have a reply to
this problem: we do have objects that names name. To any set of universals there corre-
sponds an object; some such sets concern existing objects, while others nonexistent ones.
11
Lamarque (2010, p. 188).
168
In the case of the latter, however, the problem of creation is still acute, because one subclass
of nonexistent objects is represented by fictional characters and objects. And these objects,
qua nonexistent objects, are timeless and eternal. They do not have a beginning, and they
do not have an end: these categories only apply to existing objects. All that is required
for the construal of a nonexistent object, on this view, is that we have some properties held
together by co-membership in a set. And sets, just like properties, are eternal. Any combi-
nation of these properties (a set) exists from the moment its members exist. So, an author,
on such a view, would not even be credited with combining the universals together: they
belong to the set for all time.
To sum up this problem, if one has an eliminativist view, one does not think that any
artist has ever created a fictional character or object; if one is a Meinongian, one will believe
that there are fictional character and objects, but not as a result of any creative process of
an artist. This problem runs against common sense and how we think about fiction. This is
another reason against the idea that ficta are nothing more than complex universals.
Given all these problems that the view under scrutiny has – it is neither a good inter-
pretation of Reid and it engenders several philosophical problems that it doesn’t solve – we
should find an alternative. The next section explores such an alternative.
5.3 FictaasAbstractObjects
5.3.1 TheView
A clear candidate for an alternative to the view that ficta are complex universals is the view
that they are things, substances, or, simply put, objects. This works for existing objects:
instead of thinking about them as being nothing more than bundles of properties, we should
be thinking instead that they are the things that have those properties and thus hold them
169
together.
Let us see how one can apply this idea to fictional objects; after all, it is not as easy
to think that there are nonexistent substances. Indeed, this is not how one should think
of fictional object and characters; only existing things are substances, and, in some very
precise sense, fictional characters and objects do not exist. The main idea then is to think
that proper names used in fictional discourse directly refer to abstract objects. As proposed
by Kripke (2013) and later developed by Salmon (1998) this view is primarily a semantic
one. The main question it aims to answer is: What do fictional proper names refer to? The
answer, however, should be informative for philosophers interested in the correlated ques-
tions: What are abstract objects and how do we think about them? Without answering these
questions, the view cannot hope to make much semantical progress. And in developing this
view, Kripke does make some metaphysical claims regarding the nature of abstract objects,
which become apparent by considering an example.
Let us think about the sentence:
(1) Hamlet exists.
Since this is primarily a semantic view about fictional discourse, let us play the seman-
tics game for a while and ask the question: Is sentence (1) true or false, on Kripke’s view?
If the sentence is false, in a robust sense, then we learn something very important: fictional
characters and objects do not exist, and understanding how a fictional name works poses
significant problems. It would not be very helpful to say that “Hamlet” directly refers to
something, since that something is said to not exist. Maybe a move like Russell’s, in which
definite descriptions are thought to give the logical form of a sentence like (1), would be
warranted here, after all.
12
However, in this case, fictional proper names could not be di-
rectly referential.
12
However, the previous section showed why that is a bad move to make.
170
The alternative is to think about this sentence as true. In this way, “Hamlet” can be
directly referential, since the entity it names is said to exist. How can this be, given that no
real flesh-and-blood Danish melancholic prince called “Hamlet” ever existed? According
to the advocates of the view, we are spoiled for riches: this sentence can be evaluated as
true in two distinct ways. It can be said to be true (a) from “inside the fiction”, as a sentence
about the story under discussion; and it can be said to be true (b) from “outside the fiction”,
as a sentence about facts pertaining to the real world.
For the first option, Kripke argues that fictional sentences should be understood not
as expressing propositions and as referring to anything straight, but as expressing pretend
propositions and attaining pretense reference. This is mainly because no such prince as
Hamlet ever existed.
13
To signal this idea of pretense reference, Kripke proposes to use an
intensional operator “according to the story”, in order to evaluate a sentence like (1). In a
true or false test administered to students in an English Literature class, marking a sentence
like “Hamlet soliloquizes” as true would earn full points. This is because such a sentence is
true, according to the story. The same goes for our sentence (1) above: it is true, according
to the story, that Hamlet exists. Somebody might have qualms in thinking that Hamlet does
exist in the real world, but once we use the intensional operator, there is no reason to think
that the sentence is not true. A direct consequence of this way of evaluating sentence (1)
is that we have a name that is thought to stand for something, even though that something
does not exist in the concrete way that I do, but only according to a certain story.
The other way of evaluating sentences containing names of fictional entities is by treat-
ing them as reports about the real world. As such, they are evaluated “straight”, or from
outside the fiction. The idea that sentences like (1) above can be evaluated as reports about
13
A word of caution: even if a fiction is based on a real-life person, say Napoleon, one could still talk about
pretense reference and pretend propositions being expressed by the story, since the fictional Napoleon may
have quite distinct characteristics from the real-life one.
171
the real world may seem prima facie implausible. How can we assert truly that Hamlet ex-
ists? Kripke thinks that it’s easily done: Hamlet does exist in the real world, although not
as a flesh-and-blood Danish prince. The idea is not that such a prince existed but doesn’t
any longer. Hamlet, once created by Shakespeare, started existing in the real world as an
abstract object. This is what all fictional characters and objects are supposed to be, accord-
ing to Kripke: abstract objects whose existence depends on their authors having written
stories about them, having painted them, etc. In the case under consideration, Hamlet’s
existence is no more mysterious than the existence of a nation, like France or England
(Kripke (2013, p. 55-78)). On this view, an abstract object exists in virtue of concrete
activities among people, such as writing fiction, or reading or performing plays. In this,
they are more closely related to nations, which are abstract objects also existing in virtue of
concrete activities among people, albeit of a different nature. And, fictional characters and
objects are unlike another category of abstract objects, namely numbers, whose existence
is wholly independent of agents.
5.3.2 DidReidHoldThisView?
The philosophical project Reid is mainly interested in is not a semantical one, so, if we read
Kripke’s picture as purely semantical, we will not find many points of convergence in Reid
and Kripke. However, if we look at Kripke’s theory as having metaphysical import, as it
obviously does, since it makes certain assumptions about the nature of the things referred to
by fictional names, we will find that there are certain things with which Reid would agree.
As we saw, Reid argues that fictional characters and objects can be bearers of names
(EIP IV . 1, p. 303). By this he commits himself to the idea that they are individuals,
since only individuals are named; general terms are used to group things together based on
their sharing a certain feature, and proper names are used to distinguish among individuals
172
of the same kind. Kripke offers a simple explanation of how ficta can be individuals; an
explanation that Reid himself did not include in his discussion of the topic.
In addition, Reid thinks that fictional characters and objects are created by their authors.
On this, his view is closer to Kripke, rather than to Meinongians. Here is one place where
Reid makes this clear: “They were conceived by their creators, and may be conceived by
others, but they never existed” (EIP IV . 1, p. 302). Thus we see that, according to Reid,
ficta are objects of thought and discourse once their authors conceived them for the first
time and shared those conceptions with others, through a specific medium, be that visual
or written. There must be more (or maybe less) to fictional characters and objects than
just eternal general properties that can be comprehended together in eternal sets. What this
“more” can be is yet to be explored; but the idea here is that Reid, just like Kripke, seems to
think that Cervantes was writing about a particular medieval knight, even thought entirely
fictional, when describing the fantastic deeds of Don Quixote.
This is, however, where the agreement between Reid and Kripke stops. Reid argues
that whenever somebody is conceiving a centaur, one is thinking about an animal with
spontaneous motion and life, and not about an idea, or abstract notion. Neither an idea nor
an abstract notion could be understood to have spontaneous motion and life.
This one object which I conceive, is not an image of an animal, it is an animal.
I know what it is to conceive an image of an animal, and what it is to conceive
an animal; and I can distinguish the one of these from the other without any
danger of mistake. The thing I conceive is a body or a certain figure and colour,
having life and spontaneous motion (EIP IV . 2, p. 321-22).
The abstract object view does not easily accommodate this idea. An abstract object
is not an animal, and it is unclear whether it can thus be thought to have internal life and
motion. One may try to use the intensional operator “according to the story”. But it is not
easy to do this either. We have two options: we can either give wide scope to the intensional
173
operator, or we can give it narrow scope.
14
Let us suppose that we want to evaluate a sentence like:
(2) Whenever I imagine a centaur, I imagine a real animal, with spontaneous
life and motion.
On the first option, if we give wide scope to the operator, we get the following reading
of (2): “According to the story, whenever I imagine a centaur, I imagine a real animal, with
spontaneous life and motion.” On this reading, however, it’s unclear how to understand the
word “real", or even “animal”, etc. Since these words are inside the scope of the operator,
to say that the animal is real or that the object in question is an animal does not take us
very far. The animal imagined is real, according to the story; and according to a fantasy,
anything might be real. Also, on this reading, it would be quite difficult to find a story that
talks about centaurs just as much as it does about the personal acts of imaginings of the
readers. This reading thus entails that the story says something about me imagining, which
is clearly false, about most stories. However, this is required, if we give wide scope to the
operator in question.
15
On the other option, if we give narrow scope to the operator in question, we have this
result: “Whenever I imagine a centaur, I imagine that, according to the story, there is a
thing [an abstract object] that is an animal, with spontaneous life and motion.” This looks
like the right reading of the above sentence; but it isn’t a good understanding of what Reid
is trying to accomplish in the passage quoted above. His point is that what I imagine is a
14
On Kripke’s view, we must give the operator wide scope; however, in the case under scrutiny, we are not
evaluating a sentence according to any particular story, and we can get ambiguous readings of the sentences
we have to evaluate. The problem here is that there is not one single story of the Centaur, according to which
we can evaluate any and every sentence talking about a centaur. Other issues also arise, as discussed in the
main text.
15
Kripke’s view does not require this: any act of imagination must be kept apart from what is true in
the fiction, according to the fiction. So, it should not be too surprising that this move does not work here.
However, for Reid, it is important to know how we imagine what we imagine; so, we should be able to give
a satisfactory reading to a sentence like (2).
174
certain animal with life and spontaneous motion, and that I can do that even when there is
no clear-cut prescription, in the form of a (detailed) story, to guide my act of imagination.
A certain sentence might be true or false according to a particular story, but my act of
imagining is true, just in virtue of being an act of imagining, independently of there being
a certain story that would make it so. In other words, Kripke’s view only takes us so far; it
is not suited to explain how we, as readers, or even authors, relate to these abstract objects.
Moreover, on this reading of sentence (2) we would be faced with this puzzling situation:
how does the author of the story, the one who is said to imagine the fictional character or
object for the first time, imagine it, if there is no story to help him do it? A sentence like (2)
would come out false for Cervantes sitting down, at his desk, giving shape to Don Quixote,
and not due to any fault of his imagining abilities. This variation in truth value of such
sentences is not encouraging.
In the end, Reid’s and Kripke’s projects are different, even though it could be argued
that Reid had the same semantical intuitions as Kripke and Salmon. As a semantical view,
the abstract object picture might have been adopted by Reid. However, Reid is determined
to find answers to epistemological questions, and the view under consideration does not
help in this respect; its scope is too limited.
5.3.3 PhilosophicalProblemsWithThisView
As I mentioned at the beginning of the chapter, my interpretation of Reid is driven both
by textual interpretation and by charity. If a view has certain irremediable philosophical
weaknesses, that should constitute a good reason to not attribute that view to Reid. This is
more so, if there is no unequivocal textual evidence indicating that Reid actually held such
a view. This principle constituted the basis for rejecting the Standard Interpretation of Reid
and it will be used in similar fashion to argue that, even though attractive in many respects,
175
the abstract object view should not be attributed to him, either.
Let us see what are some of the problems of Kripke’s view. Kripke offers his picture of
how to interpret fictional discourse and how to understand fictional reference as an alterna-
tive to a Meinongian view. According to him, such a view mistakenly posits two types of
existence: real existence and being.
16
However, Kripke’s view that fictional names refer to
abstract objects is not a good alternative. He argues that it is one thing to say that Hamlet
exists according to the story, and quite another to say that Hamlet exists simpliciter, as a
real object. Obviously, if we want to attribute any kind of existence to a fictional person,
we must acknowledge that such a person exists in a different way from a real person, since
a fictional person is not a flesh and blood one; it can’t be touched, or sensed in any other
way. Kripke’s proposed solution is to say that there is a systematic ambiguity in the se-
mantic role of predicates. On the one hand, if one talks about real persons, one applies the
predicates used to describe them, straight. On the other hand, if one talks about fictional
characters and objects, one applies the predicates in question according to a certain rule,
such that the resulting sentence is true if the people or objects talked about are so described
in the story (Kripke (2013, p. 59-60)).
On this picture, the predicate “exists” (just like any other predicate that can be used in a
story) is ambiguous, so unbeknown to us, we operate with two distinct notions of existence:
existence simpliciter and existence according to the story X. Let us call the latter “abstract-
fictional existence”, to indicate the kinds of objects this predicate applies to. What is this
notion of existence? It just seems to be an artificial construct, used to allow us to talk about
fictional characters and objects as if they were real. However, this notion is not as intuitive
as Kripke would like us to think. He places fictional characters and objects in the real
world, in some sense. A natural question is whether we can ask about where exactly these
16
Incidentally, this is not Meinong’s view, although many philosophers think so. See Van Cleve (1996) for
an explanation why Meinong’s view was, in fact, different.
176
objects exist; for example, nations (though in some sense abstract) exist roughly wherever
their citizens do. Maybe one could say that fictional objects exist wherever there are people
talking about them, but this does not make any progress on the question whether the type
of abstract-fictional existence that they are supposed to have is some sort of second-rate
existence, and how it differs from ordinary existence.
Moreover, Kripke does not explain how the abstract-fictional existence of fictional ob-
jects is supposed to be correlated with the attributes their home stories indicate them to
have. It is clear what it would take for a real person to soliloquize. It is less so when we try
to apply the predicate “soliloquize” to a fictional object. If we think that it is true that Ham-
let (qua abstract object) soliloquizes, we should think again: according to Kripke, such a
sentence would be false, about the real world, where we have real reference to an abstract
object. We need an intensional operator to make sense of a sentence like that, talking about
how Hamlet soliloquizes, according to the story. However, once we do that we lose the
advantage the view in question allegedly had: we need an explanation of what it would
take for an abstract object to soliloquize, or play the violin, etc.
On Kripke’s view, we have two possible readings of a sentence saying that Hamlet
soliloquizes: (i) According to the story, Hamlet soliloquizes; (ii) Hamlet, according to the
story, soliloquizes. The first reading, (i), is de dicto, while (ii) is de re. However, in (i) we
do not have reference to a real (albeit abstract) object; and in (ii) the problem is that we
don’t know what it means that an abstract object soliloquize. We cannot just assume we
know what it means to say that an abstract object soliloquizes in a different manner from
a real object. Kripke does not give us enough to understand what the “discovery” of the
systematic ambiguity at the level of predicates is supposed to do.
The problem, as I see it, is that Kripke argues that any predicate can be meaningfully
applied to fictional characters and objects. It is not nonsensical to say that Hamlet solilo-
quizes or that Hamlet exists; on the contrary, the sentence saying that Hamlet exists is even
177
true, when evaluated at the real world. It may be false to say that an abstract object plays
the violin or is a slayer of dragons, but according to some stories, such false, but mean-
ingful sentences, would be quite true. There are no restrictions, since, in a story, anything
can happen, from time-travel to change of essence. It is not so with other types of abstract
objects, however. It is nonsense, and not just a falsehood, to say that France is singing an
opera, or that Two is painting a beautiful rose. To say that France is singing an opera is to
make a category mistake, and not just to assert a falsehood. Whereas, in talking about fic-
tional characters and objects, on Kripke’s view, one simply cannot be subject to a category
mistake.
This matters: on this view, there is a stark difference between fictional characters and
objects and other types of abstract objects. However, one of the main points of this view
was to assuage our worries regarding the fact that even thought people like Don Quixote
or Hamlet do not actually exist, we can still talk about them, just like we can talk about
things like nations or natural numbers, which, also do not concretely exist. But we found
that fictional characters and objects are disanalogous from other types of abstract objects,
in an important respect. This should indicate that the worries we started with are not easily
assuaged.
Moreover, the whole view depends on the “discovery” or postulation that there is a
deep-rooted systematic ambiguity in the use of predicates. There seems to be no other ex-
planation for postulating this ambiguity at the level of ordinary language, except its role in
making the philosophical theory work. This move goes against common-sense philosophy,
and as such, would not be acceptable to Reid. It is true that a Kripke-style explanation
of how we think about fictional characters and objects has obvious advantages over the
complex universal view: it allows for singular imaginings, if we accept the idea that the
object imagined is an abstract object, and it avoids the problem of creation. On this view,
Cervantes created Don Quixote when he started writing about him, just like the citizens of
178
France created their nation when they rose against their monarchs. Both Don Quixote and
France have a beginning, and in this they differ from eternal abstract objects, like numbers,
or sets.
But the fact that Reid would not be happy with us reading him as a supporter of this
view, and the fact that it is quite unclear how to make the view work, signal the need for
another alternative to the view we started with, that Don Quixote and Pegasus are nothing
more than complex universals.
5.4 FictaasComplexTropes
5.4.1 TheView
Reid does not give us a lot when it comes to explaining exactly what it is we are supposed
to be imagining when we are imagining Pegasus or Don Quixote, or even how it is that we
imagine nonexistent things. A careful analysis of some of the things he did say indicated
that the Standard Interpretation results in a contradiction, when seen in the larger context
of Reid’s philosophical commitments. This prompted the attempt to reconcile his views
with a Kripkean-style analysis of fictional discourse, given that there is reason to believe
that Reid would have been congenial to some of Kripke’s motivations for advocating such
a picture, chiefly among them his interest in singular thought. But, as we have seen, the
costs of adapting such a view to the Reidean vision are too great. Let’s see if we can do
better.
Here’s the proposal: what if we think that fictional objects are actually nothing more
than complexes of tropes? On this picture, the ‘things’ thought about when people imagine
Pegasus, Don Quixote and the like are nothing more than bundles of tropes (or abstract
particulars, as they are sometimes referred to in the literature). This view is inspired both
179
by the complex universal view and the abstract object view. In the logical space, it finds
itself somewhere in between the other two views, having the advantages of each, and none
of their disadvantages. Or so I argue.
To begin, let us see what are tropes and why they are sometimes called “abstract partic-
ulars” (which should not make us think that they are identical to Kripke’s abstract objects,
although those are also particulars of sorts). Tropes are instance-properties, individuals in
their own rights, which either belong to the individual objects of the world, or completely
make-up such objects, depending on one’s metaphysics. As will become clear later on,
Reid thought that the real world contains objects that have properties, which are tropes;
other philosophers, Hume among them, thought that there is nothing more to an individual
than a collection of tropes bundled together. For the present purposes, it is important to
keep in mind that tropes are particulars, and not universals. They have a beginning of exis-
tence in time and are tied to a place of origin; they are not shareable or repeatable. In other
words, tropes are pieces of the world, just as much (or maybe more, if one is a Humean) as
trees or mountains.
As pieces of the world, tropes come packaged together in a specific way, but we can
single them out, think about them in isolation from the other tropes inhering in the same
object, and even combine them with other tropes that we singled out in this way, to form
new collections, different from the ones actually found in nature. This procedure should
seem no more mysterious than a procedure by which my mind selects several trees from
a certain grove of trees and “places” the selected ones in a different, imaginary grove,
together with a favorite lake, which is actually far from the original grove.
To make things more precise, let us look at what Campbell (1983, p. 130), one of the
champions of modern trope theory, writes about thinking about individual tropes:
To single out that temperature [of my right hand, at a given moment of time
and place] for thought and discussion on its own, we must perform an act
180
of abstraction, selectively ignoring what is found along with the temperature.
Tropes are brought before the mind by an act of abstraction. That is the sense,
and the only sense, in which they are abstract entities.
In other words, to single out a trope and think or talk about it, in isolation from the other
tropes with which it is combined in the real world, we must use abstraction, to “tease” it
apart from the place in which it is naturally found. However, the trope in question will
always be thought of as belonging to an object: it is inextricably connected with the object
it is a quality of. The idea here is that if I want to think about a whiteness trope, I will do
it under a very specific concept, for instance “the whiteness of this sheet of paper”. I will
not think about that trope under the concept of whiteness, in general. The general concept
is also available, probably by performing another operation of abstraction, on all whiteness
tropes I know of. Only then, can I be said to think about the universal whiteness. On
this picture, universals are entirely abstract; they are not part of the furniture of the world.
Tropes are entirely concrete; they are the physical qualities found in the world. Abstract
particulars are, for lack of a better term, mental representation of the physical qualities in
the world.
17
“Trope”, therefore, is a word that has an ambiguous meaning: it may designate
either the physical quality or the mental representation of that quality. Only in the latter
sense can it be said to be abstract. However, a trope is a particular, in either of these senses.
Let us see how this picture can help us understand how it is possible to imagine nonex-
istent things. Here is the proposal: whenever I am told to imagine a centaur, I engage
in an act of imagination in which I combine several abstract particulars, usually found in
real existing objects, with which I am acquainted. For instance, in imagining the centaur
Bane (of the Harry Potter novels), I combine in my imagination the torso of my next-door
17
These mental representation are not mental images, or ideas, that we have of the things in the world.
They are, rather, ways (or modes) of apprehending bits of the world, and, as such, they are not intermediaries
between our minds and the world. They should be no more mysterious or worrisome than the fact that we can
change our perspective on an object of the world, if we change our position. Perspectives are not considered
to be such intermediaries, and ways of apprehending the world should not be regarded as intermediaries,
either.
181
neighbor Ben and the body of my next-door neighbor’s horse, Shane. The torso, and the
body in question are tropes (or maybe made up of tropes, depending on the trope theory to
which one is subscribing), really existing in the world. My mind has the power to imagine
a combination of tropes that does not actually exist in the world, just as it has the power
of perceiving a combination of abstract particulars, that does exist. The important thing
to note is that, in the case of imagination, the components of the respective combination,
namely the tropes that my mind grasps, exist and are objects of my senses.
For this proposal to work, one must understand imagination as being very closely re-
lated to perception. The construction blocks used in imagination are the things we had
perceived and now remember.
18
The tropes we encounter in our day-to-day lives and per-
ceive as being qualities of the things around us can be “detached” in our minds and used to
construct imaginary objects. Those are the tropes that we can imagine and we should think
about imagination as having sense modalities, just like perception.
Let us suppose that there are three perceptual senses, and hence that we have visual per-
ception, tactile perception, and aural perception (as argued in Chapter 3). The same sense
modalities would be used to imagine things. The imaginer would use the abstract particu-
lars derived from perceptible tropes, so that, on the present proposal, it might actually be
better to talk about visualizing, “tactilizing” and “heararizing”, than imagining.
To better understand why this is so, think about the following case. Suppose that we
think about two different knights: Sir Gawain and Sir Peter (Strawson). If trope theory is
correct, a particular instantiation of knighthood is among the tropes Sir Gawain had, while
a different one is among the tropes that Sir Peter had. Let us think what would happen
if someone is prompted to imagine the knighthood trope associated with Sir Gawain, but
18
Or, rather, they are a small subset of the things perceived. There are things that are perceived without us
being aware of perceiving them. (The literature on the phenonmenon of change-blindness makes a good case
that such phenomena exist. For more on this issue, see Block (2007)). Such things are probably not available
to be used in constructing an imaginary object.
182
not the one associated with Sir Peter. I argue that the imaginer would not know what to
do: what is it that is characteristic to the trope of Sir Gawain but not to that of Sir Peter?
It isn’t easy to say. Or, in any event, it is not as easy to say as it is when somebody is
prompted to imagine the whiteness of this sheet of paper, as opposed to the whiteness of
that other sheet of paper. One reason for this difficulty is that there is nothing particularly
visual, or tactile, or aural about knighthood. Knighthood is not the kind of thing that can be
sensed. Although it is true that different knights instantiate the universal knighthood and
thus have the appropriate tropes among their qualities, such tropes cannot be perceived and,
I argue, neither can they be imagined. So, to help an imaginer “construct” a fictional knight,
one would need to provide some materials, namely abstract particulars, that are usually
associated with knights: shining armor, a horse, a knighting ceremony, etc. This would
work for imagining a Medieval knight, like Don Quixote, for instance. To imagine a knight
like Sir Peter, however, one would not have enough to go on, in addition to imagining that
that object is a person. To get the desired effect, one might need to do a trick, for instance,
to explicitly say something like, “oh, and by the way, this imagined person is a knight”.
Or, one could imagine Sir Peter in the middle of his knighthood ceremony, or one could
imagine people calling him “Sir”.
Knighthood is not perceived in the same way in which colors, shapes or sizes are. Better
still, in some sense, one does not perceive knighthood at all – this is just a social role some
people fill, just like the role of being an accountant, or a medical doctor. There are certain
insignia that can be associated with social roles like these, and, due to this association, we
could claim that if one is imagining a lady with a crown on her head, for instance, one
is therefore imagining a queen. But what one is actually imagining in this case is a lady
wearing a crown; queen-hood can be imagined no more than knighthood. Some tropes are
not perceivable: instances of queen-hood and of knighthood are like that. In the same vein,
tropes like these cannot be imagined either. If we were strictly abiding by etymology here,
183
there is no particular image that one can produce whenever one is told to produce an image
of a queen. After all, Queen Elisabeth II is still the Queen of England, even when she’s not
wearing her crown. In stark opposition, there are specific images that one can produce (or
indicate) whenever one is told to produce an image of redness, or squareness, etc.
The main idea here is that, on the current proposal of how imagination works, one can
only imagine the right kind of tropes (and combinations thereof): only the perceivable ones.
This is a restricted notion of imagination, but the restriction in question is not arbitrary. We
might think about this type of imagination as the kind artists employ when they create their
works. If we just think about visual artists, for a moment, they would be hard pressed to
represent such notions as knighthood, or courage, etc. If someone were to paint a painting
of Don Quixote, they might use the props described above: show Don Quixote in an armor,
on a horse, etc. But if they wanted to paint a modern-day knight, of the ilk of Sir Peter, they
would be hard pressed to come up with some symbols making it explicit that that person
is a knight, too. They might, for instance, represent the fictional modern-day knight in the
act of being knighted. This is because the tropes in question are not part of our perceptual
vocabulary, and we need to appeal to “tricks” like the ones described to make them explicit.
I don’t mean to deny that there are stories with knights, fictional or otherwise. The
point is just that, according to the view under discussion, imagination has no role to play,
if we put aside knighthood paraphernalia, and try to think of knighthood on its own. The
faculty involved here is (bare) conception, not imagination. This distinction is not ad-hoc:
knighthood, just like virtue, is a human construct; its essence is fully known to us, and in
these cases, imagination is replaced by conception, whenever we attempt to think about
these notions. The most we can do, in such cases, is to imagine persons who instantiate
these concepts; but the concepts are always general.
184
5.4.2 DidReidHoldThisView?
I argue that this is indeed Reid’s considered view: one imagines nonexistent combinations
of abstract particulars, starting from the sense-experiences they had of the tropes making
up those combinations. To do this, I first show that Reid was a trope theorist, and then I
argue that only on such a view can one think, as he did, that Don Quixote is an individual.
Let us look at the following passage which indicates that Reid was a trope theorist, at
least with regards to the properties of existing individuals:
[T]he whiteness of this sheet is one thing, whiteness is another; the conception
signified by these two forms of speech are as different as the expressions: The
first signifies an individual quality really existing, and is not a general concep-
tion, though it be an abstract one: The second signifies a general conception,
which implies no existence, but may be predicated of every thing that is white,
and in the same sense. (EIP V . 3, p. 367)
He argues that the whiteness of this sheet of paper belongs only to this sheet of paper,
and that it is not shareable by other substances. This is clear indication that the color of a
particular object is a trope; and the conception we have of such a trope is both individual and
abstract, to use his terminology. In contemporary parlance, we would say that the whiteness
of a particular sheet of paper, as apprehended by the mind, is an abstract particular.
More evidence that Reid is a trope theorist is provided by his view that qualities of
objects are objects of sense (e. g. EIP II. 17, p. 200). To be such objects, they must be
individuals, because Reid argues that only individuals can be objects of sense. Universals
cannot be perceived, in part because they are not individuals (and also because they do not
exist, according to Reid); they are general notions, reached at by a sophisticated process
of abstraction. It is worth noting that one uses abstraction in two different ways: to single
out tropes and think and talk about them in isolation, as Campbell (1983) was explaining,
with the result of having an abstract particular to think about and combine with others, as
185
one’s fancy sees fit. This is one level of abstraction that Reid talks about. Another way in
which abstraction can be used is to form general notions, by observing what several abstract
particulars have in common. Thus, a universal is found at a higher level of abstraction than
an abstract particular.
Reid’s belief that only individuals can be objects of sense helps him find a solution to
a serious problem. The problem is generated by his idea that only objects of sense can be
imagined. Other types of objects, for instance universals or even non-perceivable tropes,
can be conceived, but not imagined. Reid argues that, although imagination is a sub-species
of conception, as the Standard Interpretation also noted, it can be used to think about certain
types of things, in a way in which conception cannot. Here is some evidence supporting
my interpretation over the Standard one:
[Imagination] signifies a conception of the appearance an object would make
to the eye, if actually seen. An universal is not an object of any external sense,
and therefore it cannot be imagined; but it may be distinctly conceived. (EIP
V . 6, p. 394)
It is important to note that Reid thinks that a universal is not an object of sense, and that
this makes him say that a universal cannot be imagined. Imagination goes hand-in-hand
with sense-perception, according to him.
So, here is the aforementioned problem: if one argues, as Reid does, that centaurs
and the like can not only be conceived, but also imagined, does that mean that centaurs,
unicorns and our favorite medieval fictional knight are objects of sense? Reid seems to
be forced to say “yes”, if he is committed to the idea that only objects of sense can be
imagined. However, it is quite nonsensical to claim that centaurs are objects of sense: to
my knowledge, no one has ever encountered such a fantastical creature and, according to
some accounts of fiction, it would be impossible to have such an encounter.
19
Thus, it
19
See, for instance, the Addenda in Kripke (1980, p. 156-64) for an explanation why it is impossible to
186
seems impossible to claim that centaurs can be objects of imagination, since they so clearly
are not objects of sense.
Moreover, a very simple valid argument can be given to show that centaurs are, in fact,
universals. This argument just puts together several of Reid’s claims regarding this topic,
discussed above:
1. All and only objects of sense can be imagined.
2. Centaurs aren’t objects of sense (because they don’t exist.)
3. Therefore, centaurs can’t be imagined (Conclusion 1.)
4. All and only universals can’t be imagined.
5. Therefore, centaurs are universals. (Conclusion 2.)
In evaluating this argument, one should note that premises (1) and (2) entail (3) and
that Reid is prima facie committed to the truth of both (1) and (2). Premise (4) seems to
be true as well, since Reid argues that universals cannot be imagined, and since the world
of thought is divided by him into individuals and universals.
20
Given these, the conclusion
(5) does logically follow from (3) and (4). In addition, the argument from (3) to (5) seems,
at least on the face of it, sound. This might have been the reasoning behind the claim of
the philosophers advocating the Standard Interpretation that individual centaurs cannot be
thought about, since they do not exist and since it is more plausible to think about them as
universals.
ever find a unicorn in the real world.
20
One might think that there are other things, beside universals, that cannot be imagined. One such instance
is provided by impossibilities, according to Reid. However, there is textual evidence that impossibilities are
considered to be a type of universals. For instance, mathematical impossibilities, which can be conceived,
are said to be universals. Such impossibilities can be conceived, as any universals can, but they cannot be
imagined: “Mathematicians often require us to conceive things that are impossible, in order to prove them to
be so. This is the case in all their demonstrations, ad absurdum. Conceive, says Euclid, a right line drawn
from one point of the circumference of a circle to another, to fall without the circle; I conceive this, I reason
from it, until I come to a consequence that is manifestly absurd; and from thence conclude, that the thing
which I conceived is impossible.” (EIP IV . 3, p. 332)
187
I argue that Reid would have not accepted this argument, and that his alleged endorse-
ment of the claims above is not real. To argue, however, that individual centaurs can be
imagined, as I take Reid to be arguing, the inference from (1) to (5) must be blocked. Here
is the strategy for blocking the derivation of claim (5) above: once we show that (3) is false,
(5) does not follow any longer. To show that (3) is not true, we must give up either (1) or
(2). Premise (2) is too intuitive to reject. So, the only possibility is to show that premise (1)
is not something that Reid would accept in its current form. To block the inference from
(1) to (5) I propose to modify premise (1) so that it fits better Reid’s view. The modified
premise (1), once coupled with (2), will no longer entail (3).
Let us think about what happens if we thought that imagining a centaur were possible.
I could do it in the following way: I could construct a centaur in my imagination, starting
from several individual qualities of bodies, or tropes, with which I am acquainted from
perception. Whenever I imagine a centaur, I am said to imagine a certain human head
and torso joined to a certain body of a horse. Both the human head and torso and the horse
body are individual parts of individuals, having certain individual qualities. On this picture,
I would be said to imagine an individual centaur, since it is “constructed” out of individuals,
and not out of universals.
Is the resulting centaur an object of sense? No; for that to be possible, it would have
to exist, and my constructive act of imagination, no matter how creative, cannot result
in a flesh-and-blood animal. However, the imagined centaur should be understood to be
a compound of objects of sense, namely “the upper parts of a (certain) man”, which are
joined by my imagination in a different manner than they actually are, to another object of
sense, namely “the body of a (certain) horse.”
Reid would have to endorse this picture of how imagination works to connect us with
nonexistent objects, for his claim that one can imagine centaurs to have any traction. We
can now modify the first premise of the argument above like this: “All and only things
188
that are either objects of sense or combinations of objects of sense can be imagined.” This
modification not only invalidates the previous argument, so that the conclusion that centaurs
are universals cannot be derived any longer, but it also validates the following welcome
argument:
1’. All and only things that are either objects of sense or combinations of objects of sense
can be imagined.
2’. Centaurs are such combinations.
3’. Therefore, centaurs can be imagined.
This argument describes a way in which a centaur, which is the final product of an act
of imagination, is not an object of sense, in the same way as a real table is: that is, it is not
an existing substance. An existing horse is a combination of tropes, with the proviso that
the tropes are actually inhering in an existing substance. The horse is what holds together
its shape, its size, its color and all the other qualities that it has. In the case of an imagined
centaur, the same is not true. The imaginer’s mind (or, if we want to be very specific, the
respective act of imagination) can be said to play the role of the substance for real, existing
objects. It plays such a role, not in the sense that the abstract particulars inhere in the mind
of the imaginer, but in the sense that it is in virtue of the imaginer’s acting in a certain way
that those abstract particular are combined together.
On this picture, a centaur is imagined as a cluster of individual qualities. This is im-
portant, since, in a sense, the centaur in question is an individual itself. Just like a cluster
of universals was thought to be a complex universal, so a cluster of individuals should be
understood to be a complex individual, or, in keeping with the terminology discussed here,
a complex abstract particular. This is the view I am attributing to Reid: fictional characters
and objects are collections of tropes, combined in ways in which nature did not intend to
combine them. The tropes themselves inhere in a real substance and thus exist in virtue
of the substance’s holding them together. The centaur that is the imagined collection of
189
abstract particulars does not exist; however, the power of imagination can bring such a
combination before one’s mind.
I want to emphasize that Reid does not explain in any detail how all this comes about. I
am attributing this view to him based on two features of his theory: he thinks that there are
tropes and that one can think about them, singling them out, by a process of abstraction.
Also, he argues that only objects of sense can be imagined. Unless we want to violate
common sense and say that centaurs and the like cannot be imagined, we must either say
that they are objects of sense (which is absurd!) or explain how closely related they are
to objects of sense. The picture I presented here does the latter: it explains how centaurs
are “made up” of objects of sense, without being objects of sense themselves. This should
satisfy Reid’s requirement that we only imagine objects of sense, since a centaur thus con-
strued is as close as it can be to such an object.
5.4.3 AdvantagesofThisView
If we think that fictional characters and objects are related to imagination in this way and
thus they must be construed as complexes of abstract particulars (or tropes), we can re-
cuperate a notion of singular thought that would apply to nonexistent objects. On Reid’s
view, to have a singular thought just is to have an individual conception of something. The
contrast here is with general conceptions, which are general, not as acts of the mind (they
still are singular acts of the mind), but in virtue of the object conceived. In other words,
for Reid, to have an individual conception means no more than to conceive an individual
object; and to have a general conception means to conceive a general object, or a universal.
Since I argued that, on the view I’m attributing to Reid (based on both textual evi-
dence and philosophical reconstruction) one can imagine individuals, like Pegasus, and
Don Quixote, I will draw the obvious conclusion and say that one can thus have singular
190
imaginings of Don Quixote or Pegasus.
These are not to the exclusion of having general conceptions of winged-horses or
comical-medieval knights. One can still conceive a centaurhood universal. However, on
the view I’m attributing to Reid, one can do more: one can conceive an individual nonexis-
tent. Someone might wonder why this is important. There probably are many reasons, but
I will just mention the best one, which has to do with logic. Reid was arguing that we can
“reason consequentially” about creatures of imagination (EIP IV . 1, p. 302). To be able to
do so, one must have just one thing that one thinks about; the identity of the subject must
be maintained throughout the piece of reasoning in question.
However, there are worries that arise even if one could help oneself to the axioms
and theorems of such a logic: for instance, it is unclear how one could do existential or
universal instantiation if there are no objects that could be thought to instantiate properties
like centaurhood, or winged-horsehood, etc. One way to get out of this problem is to accept
that one can have individual conceptions about individual centaurs, in the form of singular
imaginings, as previously described.
Since combinations of individuals (namely the properties that are understood to be ab-
stract particulars) are individuals themselves, only not substances, to imagine a particular
centaur is possible and just means to have a singular thought about such a centaur.
Combinations of universals, on the other hand, are not to be understood as individuals;
they just are universals, in their own right, only of a higher level of complexity than the
universals making them up.
Only if we adopt what I argue to be Reid’s view, do we have real reference to nonex-
istent particulars. In this way, we can imagine Ronan, Bane, and Magorian of the Harry
Potter novels as being three different nonexistents belonging to the same kind, namely three
different centaurs. This is not possible on the complex universal view.
Moreover, if we adopt this picture of how imagination works to connect us with nonex-
191
istents, there is no violation of ordinary language whenever we speak of fictional characters
and objects. We do not need to postulate a systematic ambiguity at the level of predicates in
order to understand how fictional names work, as we must do on the abstract objects view,
endorsed by Kripke (2013). All we need to do is understand that fictional names do refer to
individuals, and that these individuals are different from real flesh-and-blood individuals;
they are not substances, but just combinations of really existing tropes.
The last advantage of this view is that we do not have a creation problem, as we have
on the complex universal view. On the current view, the combinations of tropes we are
imagining are not made up of eternal properties. Nonexistents turn out to be combinations
of abstract particulars, which do exist in the real world, inhering in various substances.
It takes some mental effort to tease tropes apart, single them out as abstract particulars
and then re-combine them according to unnatural laws, entirely at the discretion of the
imaginer. Without a Cervantes, there would not have been a Don Quixote, even though
the tropes Don Quixote was imagined to be made up of did really exist at the time when
Cervantes was writing his story.
The converse of this idea is that, just as fictional characters come into existence, by
being imagined by their authors, they also can go out of existence. If all copies of Don
Quixote are destroyed and everyone familiar with the story dies, Don Quixote himself can
be said to die as well. There will be no one left behind to imagine this character and his
deeds. This idea should be neither surprising, nor disturbing. Or, at least, it should not be
more surprising than the fact that languages die out with their last native speakers.
5.5 Conclusion
Let us take stock: in this chapter, I argued that there are three possible views that explain
what it is that we imagine when we imagine nonexistents: the complex universals view, the
192
abstract objects view, and the complex tropes view. The third view, the complex tropes one,
is the view I attributed to Reid, based on direct textual evidence and his overall philosophi-
cal commitments. This is in contrast to the Standard Interpretation, which I showed to lead
to attributing an eventual inconsistency to Reid. I also discussed the advantages this view
has over the other two, making it preferable philosophically, not only historically speaking.
There are some issues to which this view does not have a definitive answer. In this, it
does not fare any worse than other views about fictional discourse and the metaphysics of
fictional characters and objects; but some of the issues are interestingly difficult. It should
be pretty obvious that we do have an individuation problem, given that the fictional objects
we are talking about are quite nonexistent. Differentiating between two nonexistents, on
the current proposal, is a voluntary act, and it depends solely on the imaginer: he must
choose different tropes with which he is acquainted to “create” the combination that is
Ronan, which ends up being different from the one that is Bane – again referring to two
of the centaurs talked about in the Harry Potter novels. This proposal gets things right
in ways in which neither the complex universals view nor the abstract objects view do.
On the complex universal view, one cannot do the same as one does when using tropes:
for instance, a certain color universal is the same, no matter how one chooses to use it,
so it would be really difficult to distinguish two brown centaurs on that view. If nothing
else, one can be a lot more specific when describing a centaur on the view I attributed to
Reid. It is even less clear how one is supposed to distinguish among different centaurs on
the abstract objects view. Maybe the names are supposed to help doing that, but certainly
two different centaurs, in the same story, can both be named “John”. This would pose a
problem, especially for how we think about these objects, in the real world. It is easy to
think about distinct numbers, and the same goes for nations. However, fictional abstract
objects do not share this feature with the other types of abstract objects.
As expected, then, some version of the individuation problem is to be expected when
193
we deal with nonexistent objects, and the complex trope view is not any different in this
respect. However, on the present view the reverse situation can arise: suppose that you and
I live close to each other and our daily experience will make us be acquainted with pretty
much the same objects. Given that we interact mainly with the same tropes, we might end
up having the same (or sufficiently closely connected) abstract particulars in our mental
arsenal. So, we might think that you and I will use the exact same tropes to constitute our
own Sherlock Holmes that we end up imagining while reading some of the Conan Doyle
stories. The two combinations of tropes are qualitatively the same (at least up to a point),
so we might think that they are not two, but just one – on the hypothesis that qualitative
identity is the only kind of identity we can have for nonexistent objects. However, they are
distinct combinations, because one of them is imagined by me, while the other is imagined
by you.
I do not see this result as especially problematic, but more work needs to be done to
clarify the details here. In a way, this is to be expected: we do not interact with real
substances here, so we should not expect to have numerical identity hold whenever we
have qualitative identity holding. That is true for real existing objects, but there is no deep
metaphysical reason requiring that to be true for nonenxistents. It might seem that the
abstract objects view fares better in this respect. After all, if we have objects, even though
abstract, we can have numerical identity, or so we would expect. But is this really so? What
is the abstract object that “John Watson” names? The one with the war wound in his leg,
or the one with the war wound in his shoulder? There is no answer here.
On the current picture, we might think that the stories give us a core to work with, and
we, as audience of the stories, use our imagination to mold that core into something that is
unique to each and every imaginer. This is a proposal that should be developed further, and
the reasons favoring the complex trope view recommend that we do develop it.
194
Bibliography
Alston, W. P. (1985). Thomas Reid on Epistemic Principles. History of Philosophy Quar-
terly, 2(4), 435–452.
Alston, W. P. (1989). Reid on Perception and Conception. In M. Dalgarno, & E. Matthews
(Eds.) The Philosophy of Thomas Reid, (pp. 35–47). Dordrecht: Kluwer.
Atherton, M. (1984). “Suppose I am Pricked with a Pin": Locke, Reid and the Implications
of Representationalism. Pacific Philosophical Quarterly, 65(2), 149–165.
Bach, K. (2010). Getting a Thing Into a Thought. In R. Jeshion (Ed.) New Essays on
Singular Thought, (pp. 39–63). Oxford: Oxford University Press.
Beanblossom, R. E. (1975). In Defence of Thomas Reid’s Use of Suggestion. Grazer
Philosophische Studien, 1, 53–62.
Beanblossom, R. E. (1998). Reid and Hume on the Nature of Belief. Reid Studies, 1(2),
17–32.
Ben-Zeev, A. (1986). Reid’s Direct Approach to Perception. Studies in History and Phi-
losophy of Science, 17, 99–114.
Benbaji, H. (2007). Is Thomas Reid a Direct Realist About Perception? European Journal
of Philosophy, 17(1), 1–29.
195
Berkeley, G. (1975). A Treatise Concerning the Principles of Human Knowledge. In
Philosophical Works. London: J. M. Dent and Sons. (Original work published in 1710).
Block, N. (2007). Consciousness, Accessibility and the Mesh between Psychology and
Neuroscience. Behavioral and Brain Sciences, 30, 481–99.
Borge, S. (2007). Some Remarks on Reid on Primary and Secondary Qualities. Acta
Analytica, 22(1), 74–84.
Bourdillon, P. (1975). Thomas Reid’s Account of Sensation as a Natural Principle of Belief.
Philosophical Studies, 27, 19–36.
Brody, B. A. (1971). Reid and Hamilton on Perception. The Monist, 55, 423–441.
Brooks, G. P. (1976). The Faculty Psychology of Thomas Reid. Journal of the History of
the Behavioral Sciences, 12, 65–77.
Buras, T. (2002). The Problem with Reid’s Direct Realism. Philosophical Quarterly,
52(209), 457–477.
Buras, T. (2005). The Nature of Sensations in Reid. History of Philosophy Quarterly,
22(3), 221–238.
Buras, T. (2008). Three Grades of Immediate Perception: Thomas Reid’s Distinctions.
Philosophy and Phenomenological Research, 76(3), 603–632.
Buras, T. (2009). The Function of Sensations in Reid. Journal of the History of Philosophy,
47(3), pp. 329–353.
Burge, T. (2003). Memory and Persons. The Philosophical Review, 112(3), 289–337.
Burge, T. (2010). Origins of Objectivity. Oxford: Oxford University Press.
196
Campbell, J. (2002). Reference and Consciousness. Oxford: Oxford University Press.
Campbell, K. (1983). Abstract Particulars and the Philosophy of Mind. Australasian Jour-
nal of Philosophy, 61(2), 129–41.
Carey, S., & Xu, F. (2001). Infants’ Knowledge of Objects: Beyond object-files and object
tracking. Cognition, 80, 179–213.
Castagnetto, S. V . (1992). Reid’s Answer to Abstract Ideas. Journal for the Philosophical
Research, 17, 39–60.
Chan, L. K. H., & W. G. Hayward (2009). Feature Integration Theory Revisited: Dissoci-
ating Feature Detection and Attentional Guidance in Visual Search. Journal of Experi-
mental Psychology: Human Perception and Performance, 35(1), 119–132.
Chappell, V . (1989). The Theory of Sensations. In M. Dalgarno, & E. Matthews (Eds.) The
Philosophy of Thomas Reid, (pp. 49–63). Dordrecht: Kluwer.
Copenhaver, R. (2000). Thomas Reid’s Direct Realism. Reid Studies, 4(1), 17–34.
Copenhaver, R. (2004). A Realism for Reid: Mediated but Direct. British Journal for the
History of Philosophy, 12(1), 61 – 74.
Copenhaver, R. (2006a). Is Thomas Reid a Mysterian? Journal of the History of Philoso-
phy, 44(3), 449–466.
Copenhaver, R. (2006b). Thomas Reid’s Philosophy of Mind: Consciousness and Inten-
tionality. Philosophy Compass, 1(3), 279–289.
Copenhaver, R. (2006c). Thomas Reid’s Theory of Memory. History of Philosophy Quar-
terly, 23(2), 171–187.
197
Copenhaver, R. (2007). Reid on Consciousness: Hop, Hot or For? Philosophical Quar-
terly, 57(229), 613–634.
Copenhaver, R. (2010). Thomas Reid on Acquired Perception. Pacific Philosophical Quar-
terly, 91(3), 285–312.
Cummins, P. (1974). Reid’s Realism. Journal of the History of Philosophy, 12, 317–340.
Cummins, P. (1976). Reid on Abstract and General Ideas. In S. F. Barker, & T. L.
Beauchamp (Eds.) ‘Thomas Reid: Critical Interpretations’, Temple University Philos-
ophy Monographs.
Cummins, P. (1990). Pappas on the Role of Sensations in Reid’s Theory of Perception.
Philosophy and Phenomenological Research, 50(4), 755–762.
Daniels, N. (1976). On Having Concepts “By Our Constitution”. In S. F. Barker, & T. L.
Beauchamp (Eds.) ‘Thomas Reid: Critical Interpretations’, Temple University Philoso-
phy Monographs.
David, M. (1985). Nonexistence and Reid’s Conception of Conceiving. Grazer
Philosophische Studien, 25/26, 585–599.
DeRose, K. (1989). Reid’s Anti-Sensationalism and his Realism. Philosophical Review,
98, 313–348.
DeRose, K. (2005). Direct Warrant Realism. In A. Dole, & A. Chignell (Eds.) God and
the Ethics of Belief: New essays in Philosophy of Religion. Cambridge: Cambridge
University Press.
Deutsch, H. (1991). The Creation Problem. Topoi, 10(2), 209–225.
198
Donnellan, K. (1997). Reference and Definite Descriptions. In P. Ludlow (Ed.) Readings
in the Philosophy of Language, (pp. 361–82). Cambridge, Mass.: MIT Press.
Duggan, T. (1960). Thomas Reid’s Theory of Sensation. Philosophical Review, LXIX,
90–100.
Duggan, T. (1978). Ayer and Reid: Responses to the Skeptic. The Monist, 61(2), 205–219.
Ellos, W. J. (1981). Thomas Reid’s Newtonian Realism. Washington: University Press of
America.
Ellos, W. J. (1983). Thomas Reid’s Analysis of Sensation. New Scholasticism, LVII(1),
107–114.
Faurot, J. H. (1952). The Development of Reid’s Theory of Knowledge. University of
Toronto Quarterly, 21(3).
Faurot, J. H. (1978). Thomas Reid on Intelligible Objects. The Monist, 61(2), 229–244.
Field, H. (1973). Theory Change and the Indeterminacy of Reference. Journal of Philoso-
phy, 70, 462–481.
Friedman-Hill, S. R., L. C. Robertson, & A. Treisman (1995). Parietal Contributions
to Visual Feature Binding: Evidence from A Patient with Bilateral Lesions. Science,
269(5225), 853–855.
Gallie, R. (1989a). Hume, Reid and Innate Ideas: A Response to John P. Wright. Method-
ology and Science, (pp. 218–229).
Gallie, R. (1989b). Thomas Reid and “The Way of Ideas”. Dordrecht: Kluwer Academic
Publishers.
199
Gallie, R. (1993). Lehrer on Reid on General Conceptions. British Journal for the History
of Philosophy, 1(2), 125–138.
Gallie, R. (1997). Reid: Conception, Representation and Innate Ideas. Hume Studies,
23(2), 315–35.
Ganson, T. (2008). Reid’s Rejection of Intentionalism. Oxford Studies in Early Modern
Philosophy, 4, 245–263.
Ganson, T. S. (2002). Reid on Colour. British Journal for the History of Philosophy, 10(2),
231 – 242.
García-Carpintero, M. (2010). Fictional Singular Imaginings. In R. Jeshion (Ed.) New
Essays on Singular Thought, (pp. 273–299). Oxford: Oxford University Press.
Green, M. (1992). Visual Search: Detection, Identification and Localization. Perception,
21, 765–777.
Greenberg, A. R. (1978). Reid, Berkeley and Notional Knowledge. The Monist, 61(2),
271–281.
Griffin-Collart, E. (1977). Perception et sens commun: l’immatérialisme de Berkeley et le
réalisme de Reid. Annales de l’Institut de Philosophie.
Haldane, J. (1993). Whose Theory? Which Representation? Pacific Philosophical Quar-
terly, 74(3), 247–257.
Hamilton, A. (2003). ‘Scottish Commonsense’ about Memory: A Defence of Thomas
Reid’s Direct Knowledge Account. Australasian Journal of Philosophy, 81(2), 229–245.
Heath, P. (1978). Reid on Conceiving and Imagining. The Monist, 61(2), 220–228.
200
Holt, D. C. (1989). The Defence of Common Sense in Reid and Moore. In M. Dalgarno, &
E. Matthews (Eds.) The Philosophy of Thomas Reid, (pp. 145–158). Dordrecht: Kluwer.
Hooker, M. (1976). A Mistake Concerning Conception. In S. F. Barker, & T. L. Beauchamp
(Eds.) ‘Thomas Reid: Critical Interpretations’, Temple University Philosophy Mono-
graphs.
Hopkins, R. (2005). Thomas Reid on Molyneux’s Question. Pacific Philosophical Quar-
terly, 86(3), 340–364.
Hume, D. (2000). An Enquiry concerning Human Understanding. Oxford: Clarendon
Press. (Original work published in 1758-76).
Hume, D. (2007). A Treatise of Human Nature. Oxford: Clarendon Press. (Original work
published in 1739-40).
Hummel, J. E. (2001). Complementary Solutions to the Binding Problem in Vision: Impli-
cations for Shape Perception and Object Recognition. Visual Cognition, 8, 489–517.
Immerwahr, J. (1978). The Development of Reid’s Realism. The Monist, 61(2), 245–256.
Jeshion, R. (2010a). Introduction to New Essays on Singular Thought. In R. Jeshion (Ed.)
New Essays on Singular Thought’, (pp. 1–36). Oxford: Oxford University Press.
Jeshion, R. (2010b). Singular Thought: Acquaintance, Semantic Instrumentalism and Cog-
nitivism. In R. Jeshion (Ed.) New Essays on Singular Thought, (pp. 105–140). Oxford:
Oxford University Press.
Kahneman, D., A. Treisman, & & B. J. Gibbs (1992). The Reviewing of Object Files:
Object-specific Integration of Information. Cognitive Psychology, 24, 175–219.
201
Kaplan, D. (1989). Demonstratives. In J. Almog, J. Perry, & H. Wettstein (Eds.) Themes
from Kaplan, (pp. 565–614). Oxford: Oxford University Press.
Kripke, S. (1980). Naming and Necessity. Cambridge, Mass.: Harvard University Press.
Kripke, S. (2013). Reference and Existence. Oxford: Oxford University Press.
Lamarque, P. (2010). Work and Object: Explorations into the Metaphysics of Art. Oxford:
Oxford University Press.
Lehrer, K. (1978). Reid on Primary and Secondary Qualities. The Monist, 61(2), 184–191.
Lehrer, K. (1985). Reid on Conception and Nonbeing. Grazer Philosophische Studien,
25/26, 573–583.
Lehrer, K. (1987). Beyond Impressions and Ideas: Hume versus Reid. The Monist, 70(4),
383–397.
Lehrer, K. (1989a). Conception Without Representation – Justification Without Inference:
Reid’s Theory. Noûs, 23(2), 145–54.
Lehrer, K. (1989b). Reid on Evidence and Conception. In M. Dalgarno, & E. Matthews
(Eds.) The Philosophy of Thomas Reid, (pp. 121–144). Dordrecht: Kluwer.
Lehrer, K. (1989c). Thomas Reid. London and New York: Routledge.
Lehrer, K. (1998). Reid, Hume and Common Sense. Reid Studies, 2(1), 15–26.
Lehrer, K., & J. C. Smith (1985). Reid on Testimony and Perception. In D. Copp, & J. J.
MacIntosh (Eds.) ‘New Essays in Philosophy of Mind. Series II’, Canadian Journal of
Philosophy, vol. suppl. 11, (pp. 21–38).
202
Locke, J. (1823). The Works of John Locke, vol. 4. London: Printed for Thomas Tegg, W.
Sharpe and son, G. Offer, [et al.].
Locke, J. (1979). An Essay Concerning Human Understanding. Oxford: Clarendon Press
(Original work published in 1700).
Lombard, L. B. (1986). Events: A Metaphysical Study. London: Routledge and Kegan
Paul.
Madden, E. H. (1986). Was Reid a Natural Realist? Philosophy and Phenomenological
Research, 47(2), 255–276.
Magnus, P. D. (2008). Reid’s Defense of Common Sense. Philosophers’ Imprint, 8(3),
1–14.
Maione, M. (1999). Hume, Reid and the Theory of Meaning. In L. Formigari, & D.
Gambarara (Eds.) New Italian Studies in Linguistic Historiography. Münster: Nodus
Publikationen.
Malherbe, M. (1989). Thomas Reid on the Five Senses. In M. Dalgarno, & E. Matthews
(Eds.) The Philosophy of Thomas Reid, (pp. 103–120). Dordrecht: Kluwer.
Malherbe, M. (1991). Reid et la possibilité d’une philosophie du sens commun. Revue
Metaphisique et Morale, 96(4), 551–571.
Martin, M. (2001). Out of the Past: Episodic Recall as Retained Acquaintance”. In C.
Hoerl, & T. McCormack (Eds.) Time and Memory. Oxford: Oxford University Press.
McKitrick, J. (2002). Reid’s Foundation for the Primary/Secondary Quality Distinction.
Philosophical Quarterly, 52(209), 478–494.
203
Nadler, S. M. (1986). Reid, Arnauld, and the Objects of Perception. History of Philosophy
Quarterly, 3, 165–174.
Nelkin, N. (1989). Reid’s View of Sensations Vindicated. In M. Dalgarno, & E. Matthews
(Eds.) The Philosophy of Thomas Reid, (pp. 65–78). Dordrecht: Kluwer.
Nichols, R. (2002a). Learning and Conceptual Content in Reid’s Theory of Perception.
British Journal for the History of Philosophy, 10(4), 561–90.
Nichols, R. (2002b). Reid on Fictional Objects and the Way of Ideas. The Philosophical
Quarterly, 52(209), 582–601.
Nichols, R. (2007). Thomas Reid’s Theory of Perception. Oxford: Oxford University Press.
Pappas, G. S. (1986). Common Sense in Berkeley and Reid. Revue Internationale de
Philosophie, 40, 292–303.
Pappas, G. S. (1989). Sensation and Perception in Reid. Noûs, 23(2), 155–167.
Pappas, G. S. (1990). Causation and Perception in Reid. Philosophy and Phenomenological
Research, 50(4), 763–766.
Parsons, K. P. (1975). A Criterion for Meaning Change. Philosophical Studies, 28, 367–
396.
Parsons, T. (1980). Nonexistent Objects. New Haven: Yale University Press.
Parsons, T. (1991). Tropes and Supervenience. Philosophy and Phenomenological Re-
search, 51, 629–632.
Pelli, D. G., & Katharine A. Tillman (2008). The Uncrowded Window of Object Recogni-
tion. Nature Neuroscience, 11, 1129–1135.
204
Pitson, A. E. (1989). Sensation, Perception and Reid’s Realism. In M. Dalgarno, & E.
Matthews (Eds.) The Philosophy of Thomas Reid, (pp. 79–90). Dordrecht: Kluwer.
Pylyshyn, Z. W. (1989). The Role of Location Indexes in Spatial Perception: A Sketch of
the FINST Spatial Index Model. Cognition, 32, 65–97.
Recanati, F. (2010). Singular Thought: In Defense of Acquaintance. In R. Jeshion (Ed.)
New Essays on Singular Thought, (pp. 141–190). Oxford: Oxford University Press.
Reid, T. (2002). Essays on the Intellectual Powers of Man. Pennsylvania State University
Press. (Original work published in 1785).
Reid, T. (2003). An Inquiry into the Human Mind on the Principles of Common Sense.
Pennsylvania State University Press. (Original work published in 1764).
Reid, T. (2010). Essays on the Active Powers of Man. Pennsylvania State University Press.
(Original work published in 1788).
Robertson, L., A. Treisman, S. Friedman-Hill, & M. Grabowecky (1997). The Interaction
of Spatial and Object Pathways: Evidence from Balint’s Syndrome. Journal of Cognitive
Neuroscience, 9(3), 295–317.
Rollin, B. E. (1978). Thomas Reid and the Semiotics of Perception. The Monist, 61(2),
257–270.
Russell, B. (1910). Knowledge by Acquaintance, and Knowledge by Description. Pro-
ceedings of the Aristotelian Society, 11, 108–128.
Russell, B. (1967). An Inquiry into Meaing and Truth. Baltimore: Penguin Books. (Origi-
nal work published in 1940).
205
Russell, B. (1985). The Philosophy of Logical Atomism, chap. Logical Atomism, (pp.
157–181). La Salle: Open Court. (Original work published in 1924).
Russell, B. (1992). Theory of Knowledge. New York: Routledge. (Original work published
in 1913).
Russell, B. (1999). The Problems of Philosophy. Dover Publications. (Original work
published in 1912).
Rysiew, P. (2002). Reid and Epistemic Naturalism. Philosophical Quarterly, 52(209),
437–456.
Sagi, D., & B. Julesz (1985). "Where" and "What" In Vision. Science 7, 228(4704),
1217–1219.
Salmon, N. (1998). Nonexistence. Nous, 32(3), 277–319.
Scholl, B. J. (2001). Objects and attention: The State of the Art. Cognition, 80, 1–46.
Schulthess, D. (1986). Antoine Arnauld et Thomas Reid, défenseurs des certitudes percep-
tives communes et critiques des entités représentatives. Revue Internationale de Philoso-
phie, 40, 276–291.
Shoemaker, S. (1970). Persons and their Pasts. American Philosophical Quarterly, 7(4),
269–285.
Silver, B. (1974). A Note on Berkeley’s ‘New Theory of Vision’ and Thomas Reid’s Dis-
tinction Between Primary and Secondary Qualities. The Southern Journal of Philosophy,
12, 253–263.
206
Sleigh, R. C. (1976). Reid and the Ideal Theory of Conception and Perception. In S.
F. Barker, & T. L. Beauchamp (Eds.) ‘Thomas Reid: Critical Interpretations’, Temple
University Philosophy Monographs.
Smith, J. C. (1986). Reid’s Functional Explanation of Sensation. History of Philosophy
Quarterly, 3, 175–194.
Spelke, E. (1988). Where Perceiving Ends and Thinking Begins: The Apprehension of
Objects in Infancy. In A. Yonas (Ed.) The Minnesota Symposia on Child Psychology,
vol. 20, (pp. 197–234). Hillsdale, NJ: Lawrence Erlbaum.
Stecker, R. (1992). Does Reid Reject/Refute the Representational Theory of Mind? Pacific
Philosophical Quarterly, 73(2), 174–184.
Tebaldi, D. A. (1976). Thomas Reid’s Refutation of the Argument from Illusion. In S.
F. Barker, & T. L. Beauchamp (Eds.) ‘Thomas Reid: Critical Interpretations’, Temple
University Philosophy Monographs. Philadelphia.
Treisman, A. (1988). Features and Objects: The Fourteenth Bartlett Memorial Lecture.
The Quarterly Journal of Experimental Psychology, 40(2), 201–237.
Treisman, A. (1996). The Binding Problem. Current Opinion in Neurobiology, 6, 171–178.
Treisman, A. (1998). Features and Objects in Visual Processing. In G. Mather, F. Ver-
straten, & S. Anstis (Eds.) The Motion Aftereffect, (pp. 399–413). Cambridge, Mass.:
MIT Press.
Treisman, A. (1999a). Feature Binding, Attention and Object Perception. In
J. G. W. Humphreys (Ed.) Attention, Space and Action, (pp. 91–111). Oxford: University
Press.
207
Treisman, A. (1999b). Solutions to the Binding Problem: Progress Through Controversy
and Convergence. Neuron, 24, 105–110.
Treisman, A. (2003). Consciousness and Perceptual Binding. In A. Cleeremans (Ed.)
The Unity of Consciousness: Binding, Integration, Dissociation, (pp. 95–113). Oxford:
Oxford University Press.
Treisman, A., & G. Gelade (1980). A Feature-Integration Theory of Attention. Cognitive
Psychology, 12, 97–136.
Treisman, A., & H. Schmidt (1982). Illusory Conjunctions in the Perception of Objects.
Cognitive Psychology, 14, 107–141.
Treisman, A., M. Skyes, & G. Gelade (1977). Selective Attention and Stimulus Integra-
tion. In S. Dornic (Ed.) Attention and Performance VI, (pp. 333–361). Hillsdale, NJ:
Lawrence Erlbaum.
Tulving, E. (1983). Elements of Episodic Memory. Oxford: Oxford University Press.
Tulving, E. (1985). How Many Memory Systems Are There? American Psychologist,
40(4), 385–398.
Van Cleve, J. (1985). Three Versions of Bundle Theory. Philosophical Studies, 47, 95–107.
Van Cleve, J. (1996). If Meinong Is Wrong, Is McTaggart Right? Philosophical Topics,
24(1), 231–254.
Van Cleve, J. (1999). Reid on the First Principles of Contingent Truths. Reid Studies, 3,
3–30.
Van Cleve, J. (2003). Reid Versus Berkeley on the Inverted Retinal Image. Philosophical
Topics, 31, 425–455.
208
Van Cleve, J. (2004). Reid’s Theory of Perception. In T. Cuneo, & R. Van Woudenberg
(Eds.) The Cambridge Companion to Reid, (pp. 101–133). Cambridge: Cambridge Uni-
versity Press.
Van Cleve, J. (2006). Touch, Sound, and Things Without the Mind. Metaphilosophy, 37(2),
162–182.
Van Cleve, J. (2007). Reid’s Answer to Molyneux’s Question. The Monist, 90(2), 251–270.
Van Cleve, J. (2008a). Reid on Single and Double Vision: Mechanics and Morals. Journal
of Scottish Philosophy, 6, 1–20.
Van Cleve, J. (2008b). Reid’s Response to the Skeptic. In J. Greco (Ed.) The Oxford
Handbook of Skepticism, (pp. 286–309). Oxford: Oxford University Press.
Van Cleve, J. (2012). Four Questions About Acquired Perception. In Mind, Knowledge
and Action: Essays in Honor of Reid’s Tercentenary. Unpublished manuscript.
Van Cleve, J. (2013). Problems from Reid. Unpublished book manuscript.
Van Woudenberg, R. (1999). Thomas Reid on Memory. Journal of the History of Philoso-
phy, 37(1).
Van Woudenberg, R. (2000). Perceptual Relativism, Scepticism, and Thomas Reid. Reid
Studies, 3(2), 65–85.
Van Woudenberg, R. (2004). Reid on Memory and the Identity of Persons. In T. Cuneo, &
R. Van Woudenberg (Eds.) The Cambridge Companion to Thomas Reid, (pp. 204–221).
Cambridge: Cambridge University Press.
vanMarle, K., & Scholl, B. J. (2003). Attentive Tracking of Objects Versus Substances.
Psychological Science, 14, 498–504.
209
Vernier, P. (1976). Thomas Reid and the Foundations of Knowledge and his Answer to
Skepticism. In S. F. Barker, & T. L. Beauchamp (Eds.) ‘Thomas Reid: Critical Interpre-
tations’, Temple University Philosophical Monographs. Philadelphia.
Weinberg, S. (2011). Locke on Personal Identity. Philosophy Compass, 6(6), 398–407.
Weldon, S. (1982). Direct Realism and Visual Distortion: A Development of Arguments
from Thomas Reid. Journal of the History of Philosophy, 20(4).
Wolfe, J. M. (2003). Moving Towards Solutions to Some Enduring Controversies in Visual
Search. Trends in Cognitive Sciences, 7(2), 70–76.
Wolfe, J. M., & Kyle R. Cave (1999). The Psychological Evidence for A Binding Problem
in Human Vision. Neuron, 24, 11–17.
Wolterstorff, N. (2000). Thomas Reid’s Account of the Objectivated Character of Percep-
tion. Reid Studies, 4(1), 3–16.
Wolterstorff, N. (2001). Thomas Reid and the Story of Epistemology. Cambridge: Cam-
bridge University Press.
Wood, P. (1994). Hume, Reid and the Science of the Mind. In M. A. Stewart, & J. P. Wright
(Eds.) Hume and Hume’s Connexions, (pp. 119–139). Edinburgh.
Yaffe, G. (2003a). The Office of an Introspectible Sensation: A Reply to Falkenstein and
Grandi. Journal of Scottish Philosophy, 1(2), 135–140.
Yaffe, G. (2003b). Reid on the Perception of Visible Figure. Journal of Scottish Philosophy,
1(2), 103–115.
210
Yaffe, G. (2009a). Beyond the Brave Officer: Reid on the Unity of the Mind, the Moral
Sense, and Locke’s Theory of Personal Identity. In S. Roeser (Ed.) Reid on Ethics.
Palgrave Macmillan.
Yaffe, G. (2009b). Thomas Reid on Consciousness and Attention. Canadian Journal of
Philosophy, 39(2), pp. 165–194.
Yaffe, G., & Nichols, R. (2009). Thomas Reid. Stanford Encyclopedia of Philosophy.
Accessed on 19 May 2011.
URLhttp://plato.stanford.edu/entries/reid/
Yolton, J. W. (1984). Perceptual Acquaintance: From Descartes to Reid. University of
Minnesota Press.
211

Abstract (if available)

Abstract In this dissertation, I offer a new reading of Thomas Reid's philosophy of mind, in which I bring attention to his so far overlooked view that we can entertain a singular thought about an individual substance. ❧ Two of Reid's works, An Inquiry into the Human Mind on the Principles of Common Sense (1764) and Essays on the Intellectual Powers of Man (1785), address central issues in philosophy of mind, psychology and vision science. Our ability to have singular thoughts is one of the recurring themes throughout his investigation. ❧ Reid does not use the terminology of ""singular thought"", which became widespread in the twentieth-century, once the lessons from Russell and Quine had been fully appreciated. In Chapter 1, however, I show that his interest in our ability to think about individual substances in the world places him squarely within this tradition. ❧ According to Reid, perception gives us immediate knowledge of the external world, namely that type of knowledge that the skeptic denies we can have. In Chapter 2, I argue that bodies, together with their qualities, are, in Reid's terminology, immediate objects of original perception. This is a controversial view in Reid scholarships, but I show that other scholars overlook some key passages when they offer their reading. ❧ Perception of bodies is singular: according to Reid, one does not use a description of the qualities of bodies, in order to perceive those bodies. In Chapter 3 I argue that this is compatible with his view that perception of bodies is a two-stage process: at the first stage, we perceive qualities

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Thomas Reid: motives and the anatomy of the mind

PDF

Language and the structure of Berkeley's world

PDF

Just imagining things: Hume's conception-based account of cognition

PDF

The scope and significance of George Berkeley's language model

PDF

Minimal sensory modalities and spatial perception

PDF

Merely verbal disputes in philosophy

PDF

The case for moral skepticism

PDF

The mind of the slave: the limits of knowledge and power in Roman law and society

PDF

The built environment, tour complexity, and active travel

PDF

Medics of the soul and the body: sickness and death in Alta California, 1769-1850

PDF

Legal spectres, narrative ghosts: mothers and the law in the Victorian novel

PDF

Point singularities on 2D surfaces

PDF

Contrastive reasons

PDF

The constitution of action

PDF

Theism and secular modality

PDF

Effect of mindfulness training on attention, emotion-regulation and risk-taking in adolescence

PDF

Responsibility and rights: a search for a principled distinction between criminal law and tort law

PDF

Can election reforms improve democracy?

PDF

Crime and punishment: examining proportionality issues within criminal law

PDF

Scripted voices: persona and speech in Senecan philosophy

Asset Metadata

Creator Folescu, Marina Radiana (author)

Core Title Thomas Reid on singular thought

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Philosophy

Publication Date 07/03/2013

Defense Date 06/07/2013

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag conceptual content,history of philosophy of mind,Imagination,memory,non-conceptual content,OAI-PMH Harvest,philosophy of perception,philosophy of psychology,Thomas Reid

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Van Cleve, James (committee chair), McCann, Edwin (committee member), Yaffe, Gideon (committee member)

Creator Email folescu@usc.edu,radifol@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-283735

Unique identifier UC11292633

Identifier etd-FolescuMar-1734.pdf (filename),usctheses-c3-283735 (legacy record id)

Legacy Identifier etd-FolescuMar-1734.pdf

Dmrecord 283735

Document Type Dissertation

Format application/pdf (imt)

Rights Folescu, Marina Radiana

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA