University of Southern California Dissertations and Theses

Contract, from promise to commodity

(USC Thesis Other)

Contract, from promise to commodity

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 1

Contract, From Promise to Commodity
Erik Encarnacion
University of Southern California
Department of Philosophy
Dissertation
Submitted to Satisfy the Requirements of the Doctor of Philosophy Degree
Degree Conferred by the Graduate School, August 2016

2

Table of Contents
Acknowledgements ................................................................................................................................................ 3
Chapter Zero: Introduction .................................................................................................................................. 9
Chapter One: Contracts and the Practice Theory of Promising .................................................................... 28
Chapter Two: Contracts and Agency Theories of Promising ......................................................................... 63
Chapter Three: Contracts and Interest Theories of Promising ...................................................................... 90
Chapter Four: Contracts and the Assurance Conception of Promising ....................................................... 116
Chapter Five: Contract as Commodified Promise ............................................................................................ 154
Works Cited ............................................................................................................................................................ 201
Cases Cited .............................................................................................................................................................. 203
Statutes Cited .......................................................................................................................................................... 203

3

Acknowledgements
I consider myself extraordinarily lucky in life, to have been given so much in the way of opportunity
and support, and I know without doubt that I won’t be able to recall every person that has helped me along
the way. And I’m certain that the people I do recall cannot be thanked nearly enough.
I want to start with Gideon Rosen, Alexander Nehamas, and Mark Greenberg. Gideon taught the
course, “Freedom and Responsibility,” and I was hooked on philosophical inquiry from then on. My very
first paper was a disaster. But the next one “hit all the right notes,” he wrote. (It’s funny how little things can
make such a big difference, a lesson that I have not always remembered in commenting on my own students’
work.) He, of course, might very well think that he deserves neither praise nor blame, especially not blame,
for whatever causal role he played in my decision to pursue an academic life. But no one did more to light the
fire than Gideon.
I admired Alexander’s fearlessness in tackling big-picture questions about the meaning of life, love,
beauty, and friendship—and doing so without sacrificing clarity or rigor. More importantly Alexander is a
kind and genuine person. Mark Greenberg taught an intro legal philosophy course that I think about to this
day. The course introduced the usual suspects—Hart, Dworkin, and so on. I knew I was not finished with
legal philosophy, and somehow I suspected I never would be—though I confess I think much less about
general jurisprudence than I used to. When I asked Mark whether I should go to law school or graduate
school, he in effect asked why I wouldn’t want to do both. And so I did. My time at Princeton was extremely
academically rewarding.
I continued learning about law and its theoretical foundations at Columbia Law School. I took a
course with Joseph Raz, who was remarkably patient with a “mere” law student during his office hours.
Jeremy Waldron had the greatest influence on me. I thank him for his patience despite my persistent
pestering. He very kindly agreed to write letters of recommendation for me years after I had graduated from
law school. And I will never forget his course, Perspectives on Legal Thought. Its syllabus was crafted
dialectically. Each theoretical perspective on law was presented as a reaction or response to a prior dominant
school of thought. All perspectives—including anti-philosophical views ranging from American Legal
4

Realism, Critical Race Theory, to Critical Legal Studies, to law and economics—were treated respectfully,
even when he did not endorse the view in question. Jeremy’s course was a tremendous, foundational
education in legal theory, and his teaching style and method of syllabus construction is a model I hope to
carry on if and when I ever teach a similar course.
I want to thank the University of Southern California’s philosophy department for admitting me.
When I finally decided that the siren song of academia was too strong to resist, I cobbled together some
philosophy of law papers that I thought were pretty good, and submitted the package to graduate schools in
philosophy. Looking back, I am far less certain that those papers qualify as even “pretty good,” so I count
myself very lucky to have been admitted to USC in the first place, let alone with the financial support
provided by the Provost Fellowship. I don’t know who was responsible for letting me through the gates. But
I want to thank whoever you are.
Let me thank Scott Soames and Mark Schroeder. They have set the tone at the top, demanding
excellence, seriousness, and professionalism—and walking the walk with their tireless work ethics. Mark’s
advice on nearly any topic, ranging from career planning to the fine art of structuring a paper, is invariably
valuable. His intense dedication to each graduate student boggles the mind. Meetings with Mark aimed to
ensure that we were making the most of our time here. Every time he weighs in on an issue he changes my
thinking. It’s almost unfair how routinely insightful he is.
But above all other faculty members before and during my time at USC I want to thank Andrei
Marmor. One’s experience as a graduate student depends largely on whether you feel like you have sound
guidance from at least one faculty member who you believe takes your work seriously—someone who will
always ultimately “have your back.” I was lucky enough to have several faculty members who I could count
on. But Andrei was an ideal mentor. He was amazingly prompt about giving feedback and never pulled
punches in doing so. He has a keen sense of which lines of inquiry in legal philosophy are promising or
fruitful and which are moribund or otherwise to be avoided. He invited me to various events in legal
philosophy that I did not even know about. No less important, Andrei provided much-needed perspective
when my self-doubt began to take hold, sometimes by telling a story about his own academic experiences. I
5

could not have asked for a better adviser. I hope only that all graduate students can experience the same level
of support and guidance from a faculty member that I experienced.
Others made my time at USC a happy one. Gary Watson and Greg Keating deserve special mention.
Gary and Greg co-taught a course on the philosophy of tort law my first year. It was by far my favorite
course at USC, and in retrospect, this course set me on my path towards becoming a scholar of private law
theory. (Yet another example of luck. What if they had opted not to teach the course? I shudder to think.)
Gary somehow found time to work with me on my projects, and served as a characteristically thoughtful and
deep commentator on my dissertation committee, despite numerous other commitments and an accelerated
timeline that took even me by surprise. Greg also found time to discuss my work despite his primary
appointment in the law school, and despite his considerable administrative and academic burdens. What’s
more, Greg spent a lot of time shepherding me through the daunting law teaching job market and reaching
out to others on my behalf. He did this, time and again, and tirelessly, without once expressing irritation at my
steady stream of job-related e-mails. Beyond this, Greg’s own work in tort law has influenced my thinking,
especially about strict liability. I learned much from both Gary and Greg, both in their scholarly work and as
models for academic service.
Thanks also to the other members of my committee, Jon Quong and Becca Stone. In some respects I
regret not having been in Los Angeles long enough to have worked more closely with them given their own
fascinating scholarly work, which overlaps to a large degree with my own interests. Despite the long-distance
nature of our exchanges, their comments were invaluable and saved me from errors, forcing me to abandon
some arguments that were insufficiently well developed. I am especially grateful to Becca for keeping me
honest on the law of contract, pointing out messy nuances in the law that I had either forgotten or did not see
while wearing my philosophical blinders. The dissertation is, I hope, somewhat less embarrassing as a result.
And somehow they managed to do all of this despite yet another young philosopher, Kira, vying for their
time. I wish I were as organized in my life as they apparently are in theirs.
I want to thank other current and former faculty members that I have worked with in various
capacities, either as a student in a seminar, a TA, or simply as a reader of my work, including: Kenny
6

Easwaran, Steve Finlay, John Hawthorne, Shieva Kleinschmidt, Janet Levin, Frank Lewis, Sharon Lloyd, Ed
McCann, Jake Ross, Gabriel Uzquiano-Cruz, Kadri Vihvelin, Ralph Wedgwood, and Gideon Yaffe. Susan
Wolf allowed me to sit in and participate in a wonderful seminar on responsibility while I was living in North
Carolina during a fall semester, and for that I’m really grateful. I owe Robin Jeshion special thanks for
agreeing to be my teaching mentor and for being a wonderful faculty sounding board, one who is genuinely
concerned about the well-being of graduate students.
The graduate students at USC were as willing to grab a beer and talk about pop culture as discuss
philosophy—or invite you and your fiancée into their Long Beach home for Thanksgiving dinner even when
they don’t know you very well (ahem, Justin Snedegar). And some of the best feedback on my work has come
from fellow graduate students. So grad life was so thrilling in no small part because of the likes of Ara
Astourian, Matt Babb, Josh Crabill, Justin Dallman, Alex Dietz, Marina Folescu, Amanda Gorman, Keith
Hall, Michael Hatcher, Ben Lennertz, Alida Liberman, Jen Liderth, Joe Horton, Nicola Kemp, Nick
Laskowski, Woo Ram Lee, Shyam Nair, Abelard Podgorski, Indrek Reiland (despite the jiu jitsu beatings I
received form him), Kenneth Silver, Julia Staffel, Aaron Veek, Myles Webster, Jessica Wilson, and Jon
Wright. I hope our paths continue to cross early and often.
In my cohort, specifically, I want to thank Rima Basu and Caleb Perl. In addition to being great,
sharp philosophers they are also hilarious and fun loving and down to earth. The same goes for Mike
Ashfield, Renee Bolinger, Maegan Fairchild, and Tanya Kostochka, who sometimes felt like they were in my
cohort. I count myself lucky to know them. Matt Lutz, my roommate, never held back his views on anything.
I miss our Game of Thrones viewing nights. If I regret anything, it is that time is so short, and that I’m such a
homebody; I wish I could have gotten to know all of you even better.
I was lucky enough to become part of a philosophy of law subculture while at USC. I could talk shop
with Steve Bero, Kory DeClark, Mike Pressman, Alex Sarch, and Aness Webster. It was incredibly gratifying
having a few lawyers who, like me, were crazy enough to pass up potentially lucrative careers for the scholarly
life (though Kory looks like he is well on his way towards reversing this order). They also kept me honest in
my theorizing about the law. I owe a special thanks to Steve, Alex, and Aness for reading so much of this
7

dissertation, some of it at the very last minute, and providing very thoughtful feedback throughout. I hope
they will continue to exchange papers with me, and that they will continue to point out the flaws in my work
with gusto.
Now let me get more personal—second-personal, that is. Thank you Steve Bero and Beth Snyder for
your friendship. You were like family when my own was often so far away. You made my time here
immeasurably better than it would have been otherwise. Apart from having excellent taste in music and
introducing me to Neutral Milk Hotel, Beth is a master storyteller. Her tales are legendary and guaranteed to
make you laugh until you cry or otherwise wet yourself. As for Steve, what can I say? Like me, Steve
graduated from Columbia Law School in 2006 (but I didn’t know him then!); like me, Steve had clerked for a
federal judge (two, to be precise!); like me, Steve had spent time as an associate in a fancy law firm; and Steve
also has deep interests in legal philosophy. This could have been a recipe for disaster, actually, or at least a
recipe for a surefire frenemy. But Steve is, and here I gush again, simply one of the best and most virtuous
people I know, full stop. Steve, I want to be like you when I grow up.
Family support remains essential. Claudia Encarnacion, Rosa Ramirez, and especially Edgar Ramirez
all provided crucial help by taking care of our little Evan during critical times. It is no exaggeration to say that
we could not have finished our respective degrees in a timely fashion without your help. And thank you,
again, Claudia and Jose Luis Encarnacion, mom and dad. At times you struggled to understand our
unorthodox life plans. Hell, I did, too. But you never ceased trying to understand or providing moral support.
I could never have asked for anything more. Thank you and I love you.
Of course I need to thank you, Jahanett. I was practically on my way out the door to Memphis to
take up a clerkship when we met in New York. And, in full disclosure, I told you that I had the long-term
goal of becoming a legal scholar. I was pleasantly surprised when you said that you’d stick with me despite
this precarious moonshot of a pursuit. We thought that the distance between Dallas and Memphis was long.
And it was. But after a brief reunion in Dallas, and after our engagement, the admissions gods frowned upon
us and sent us even further apart to opposite coasts. Medical school forced you to open up open up cadavers
in Winston-Salem while I was forced to take meta-logic in Los Angeles. We both had to hold our noses it
8

seems. And at times not having each other right there was very, very difficult. Graduate student life is hard, but
so is life as a medical student, and I dare say you had a far less forgiving schedule and far less supportive
learning environment than I did. But every night things would be okay because we would talk and put each
other at ease. Those phone calls were without fail the highlight of my day, every day. So I never once doubted
the mutual respect, love and support that we have for each other, despite the long distance and long time
apart. Thank you for not wavering.
You are generous and kind and smart and driven and funny. You are one of the loves of my life, the
other of course being our baby boy. I could not have done it without your love, support, and patience from
afar. But I am so, so glad that the long distance stuff is over with. I hope only to be able to reciprocate as you
begin your career as a physician.
I dedicate this dissertation to you, my love.
E.E.
9

Chapter Zero: Introduction
1 A Tale of Two Promise Theories: Fried and Shiffrin
Many scholars accept that contractual obligations are a form of promissory obligation. This seems
plausible given their respective normative structures. When you make a promise, you voluntarily commit to
do something, and you are not entitled to forego that commitment absent highly circumscribed conditions
including release by the promisee or the existence of a justification or excuse. The mere fact that a better
opportunity has come along, for example, normally does not count as an excuse or justification to back out of
a promise.
Contractual obligations have similar normative structures. They are also voluntarily undertaken, and
the conditions under which you can avoid contractual obligations are also limited. But there are also palpable
differences between contracts and ordinary, run-of-the-mill promises. Some are obvious. Roughly, contractual
obligations are creatures of law, while promissory obligations are not. This difference might suggest yet
further normative and evaluative differences. These differences matter to the extent that we are interested in
evaluating claims that contractual obligation is rooted in the norms and values that give rise to promissory
obligations more generally.
Despite these apparent differences, Promise Theories of contract law attempt to justify contract law at
least in part in terms of promissory morality and to show the extent to which the common law of contract is
itself guided by moral norms governing making and keeping promises. This dissertation is an exploration of
Promise Theory. Before turning to the specific contributions that I hope to make, and what motivates the
dissertation as a whole, a little background about Promise Theory will be helpful. In particular, I should say a
few words about what I take to be the two most important contributions to Promise Theory in the legal and
philosophical literature in the last fifty-or-so years.
The most systematic Promise Theory of contract law to date appears in Charles Fried’s book, Contract
as Promise. Fried argued that legal norms governing contractual obligations do and should track moral norms
governing ordinary promissory obligations. For example, Fried observed that promises, to generate binding
moral obligations on promisors, must be accepted by a promisee. Lo and behold, contractual obligations, like
10

promissory ones, must be accepted in order for them to become legally binding. Fried argued more generally
that many of the stable structures of the common law of contract can be usefully understood in terms of the
structural features of promissory morality, including the moral norms governing: the formation of promises
(including whom may form morally binding promises), the duty that promises be kept, the conditions under
which they need not be kept, the conditions under which broken promises may be justified and excused, the
moral demands that arise when promises are broken, and the like.
Fried recognized that the mapping of the legal on the moral is not always a perfect. There is not
always a one-to-one correlation between what promissory morality requires and what the common law of
contract requires. But this just meant, according to Fried, that either we have a strong reason to reform
contract law, or that contract law’s divergence from the demands of promissory morality must come from
some other extremely important moral consideration, such as the importance of ensuring that disputes arising
from broken promises are capable of being fairly administered by courts of law.
One reform that Fried proposed was to jettison the doctrine of consideration. This common law
doctrine holds that no promise is binding in contract law unless it is bargained for or exchanged for either
another promise or valuable “consideration”—cash or something else with market value. The rationale for
this requirement remains disputed. But it is a firmly entrenched formal requirement in the common law of
contract, and although it has certain exceptions, most theorists of the common law try to explain and justify
rather than eliminate the doctrine. Fried, however, saw no point to it. Because there is no moral analogue in
promissory morality, and because the doctrine (he thought) was internally incoherent, we should get rid of it.
This was a call for a fairly radical reform given, again, how firmly entrenched the doctrine was and remains.
Fried nevertheless sought to preserve other doctrines not readily explicable in terms of promissory
morality. In the case of legal remedies, Fried saw no problem with the duty to mitigate damages. This duty
imposes on plaintiffs in contract cases a duty to try to minimize the damages that they would be able to
recoup from those who breached their contracts. So if someone A breaches his contract to buy some widgets
from you, then you should (according to the duty) try to minimize the amount that A owes by trying in good
faith to resell the widgets originally set aside for A. This requirement does not seem to be demanded by
11

promissory morality. Fried, acknowledging this, tried to justify the requirement as a manifestation of
something like a duty of beneficence.
Fried’s book has been influential, but mainly as a cautionary tale about how not to argue for a Promise
Theory. The criticisms have been varied. The most obvious objection is the conceptual one: that contractual
obligations are not promissory obligations. The two forms of commitment are simply too different, it is
argued, to justifying treating the former as a form of the latter. Others criticized the specific way that Fried
executed his task, claiming that Fried misunderstood the implications of his own commitments. For example,
Fried argued that the natural and appropriate remedy for breach of contract is expectation damages. This relief
aims to provide to the defendant the “benefit of the bargain,” i.e., the monetary equivalent that the plaintiff
would have realized if the defendant had actually performed her contractual obligation. If selling something
to A per the contract would have realized $100.00 in net profit to the plaintiff-seller, then the plaintiff is
entitled to $100.00. But several critics pointed out that the remedy for breach of contract should not be
expectation damages, but rather specific performance, since promising to do something primarily commits one to
actually perform, not to choose whether to perform or pay the monetary equivalent of performance. Specific
performance, where possible, better reflects the character of promissory obligations on Fried’s own terms.
Apart from these specific criticisms, Fried’s project was subject to more fundamental objections that
target the very project of articulating a Promise Theory of contract law. Two objections stand out. The first is
what we might call the accusation that Fried’s project was moralistic in a pejorative sense. As many have
pointed out, the fact that some act is morally wrong or morally required does not normally suffice to justify
making that act illegal or legally obligatory. My being rude to a family member during Thanksgiving is wrong;
but it should not, as a matter of principle, be subject to legal sanction. The moralism objection’s motivation
usually comes from liberal political thought—whether grounded in the Harm Principle or theories that limit
states action to matters of justice. The law should not be enlisted to enforce moral virtue in general, so it is
not obvious why the law should be enlisted to enforce the virtue of keeping one’s promissory commitments.
For his part, Fried replied by treating promissory morality as within the realm of rightness rather than
goodness or virtue. Contract law simply protects promissory rights in the way that other areas of law, like tort
12

law and criminal law, protect property rights and rights to bodily integrity. But this is not satisfying. My
promise to meet a good friend for lunch should not, it seems, place me on the hook legally should I forget,
even though my good friends obtain a moral right to my showing up on time when I make the promise. True
enough, my friend probably won’t sue me. But the point is that contract law should not in principle empower
my friend to sue me if he so desired. So the stock objection of excessive moralism remains problematic for
Fried as with virtually any attempt to articulate a Promise Theory.
This raises the second stock objection. Even assuming that some promises ought to be subject to legal
enforceability, promissory morality provides no resources to determine which ones. For structural reasons, this objection
should not be surprising. Any (non-debunking) theory of promissory morality provides nothing more or less
than a theory of what promising is, which acts constitute a promise, and a theory explaining when and why
those acts generate morally binding, special obligations. With the exception of utilitarian accounts of
promissory obligations, which provide the same answer that it provides to all normative questions, it is simply
beyond the purview of these theories to suggest the conditions under which contract law ought to enforce
promises. So we should not be surprised that this second stock objection is tailor made to embarrass attempts
to justify and guide contract law by reference to the demands of promissory morality.
Despite these objections, recent years have seen a Promise Theory revival. I will go out on a limb and
attribute this revival to a paper by Seana Shiffrin, published in the Harvard Law Review, called The Divergence
of Contract and Promise. The paper was published in 2007 but has already amassed 255 citations to date
according to a quick Google Scholar search. Her chief innovation, in my opinion, is showing a new approach
to Promise Theory. She rejects Fried’s attempt to directly argue for one-to-one mappings between the moral
domain of promising and the domain of contract law doctrine. Instead, she seeks an indirect,
“accommodationist” approach, which begins with an assumption rooted in political morality. Very roughly,
even if the state should not seek to directly enforce moral virtue, Shiffrin argues, this does not mean that the
state should promote moral vice or express approval of morally vicious attitudes and conduct, or otherwise
make it excessively difficult to live a minimally morally virtuous life. This commitment allows Shiffrin to point
13

out the ways that contract law doctrine, in her view, either promotes or approves of morally mistaken
conduct and attitudes towards promising.
The main problem appears to be, according to Shiffrin, the problem of opportunistic breach. An
opportunistic breach is a breach, roughly, where one party breaks his contractual commitment in order to
secure a different contractual commitment to another party, where the two commitments are incompatible.
For example, suppose that I promise to sell my widgets to B for $500.00 and B agrees. Now suppose that C
comes along later on and offers $2,000 for those same widgets. If I turn around and sell to C rather than B, I
have breached my contract with B opportunistically in order to secure the better deal with C.
The issue isn’t simply that I broke my promise to B. To see why, and to understand Shiffrin’s critique
of contract law, requires knowing the remedies available to B under the common law of contract. B is entitled
to expectations damages but nothing more. Assume that B’s plan was to turn around and re-sell the widgets
on the retail market for $600, which would realize $100.00 in net profit. Expectation damages means that B is
entitled to only $100 when he sues me. B is not entitled to specific performance; that is, he is not entitled to make
me hand over the promised widgets, if he happened to catch wind of my tempting offer from C. Nor is B
entitled to punitive damages, which operate as a kind of fine or punishment for my rank opportunism.
Indeed, B lacks any legal vehicle for expressing his dismay over my disrespect for or my infringement of his
promissory right. My opportunism is treated no differently by contract law than a different case in which a
hiccup on my production line prevented my timely completion of the widget delivery. What’s more, B might
even have a duty to try to mitigate the damages that I owe to B, according to the aforementioned duty of
mitigation. This just adds insult to injury.
Now we are in a position to see why contract law looks problematic from Shiffrin’s point of view. If
C offers $2,000 for the same widgets that I have promised to B for $500.00, it would be instrumentally
rational of me to break my contract and simply pay B the $100.00 to which he is owed. And I can do this
under current contract law since punitive damages and specific performance do not hover in the background
as additional incentive to avoid breaching the contract. Contract law provides little in the way of additional
14

incentive or opprobrium to discourage my breaching; indeed, the rationales sometimes proffered for the rule
against specific performance and punitive damages appear to endorse such breaches.
This will not do, according to Shiffrin. The problem is not necessarily the kind that Fried might
identify, which is that legal norms governing contractual obligation lack an analogue found in ordinary
promissory morality. The problem lies in the way the law treats those trying to live minimally decent moral
lives. Contract law sets up an environment that creates incredible pressure to breach opportunistically, both
intrinsically by what it expresses and by creating improper incentives to breach, and cultivates in citizens
improper attitudes towards their promissory obligations. We begin to think that breaking our leases is not a
problem so long as we pay a fee, or if liquidated damages clauses are not provided, so long as we pay
expectation damages to the property owner. Property owners are often subjected to duties to mitigate the
damages caused by our lease-breaking mischief by securing new tenants to fill their now-empty lodgings,
again, adding insult to injury. And this is all an affront to promissory morality. If this sounds moralistic it is
because it is; Shiffrin’s view is certainly moralistic but indirectly so, in a way more likely compatible with the
demands of liberal political morality, at least arguably.
Shiffrin’s accommodationist view arguably provides resources to answer the stock objections. Unlike
Fried, Shiffrin does not appear to try to justify contract law simply on the grounds that doing so enforces the
moral duties that constrain our behavior anyway. Accommodationism requires making it possible, without
undue burden, for minimally virtuous moral agents to exercise their agency in ways that accord with the
demands of morality. This is no less compatible with liberal political theory, the thought might be, than
requiring that the state no place undue burdens on the beliefs of sincere religious believers. But it is less clear
that the accommodationist view explains which promises should be subject to legal enforcement, or why some
but not all of our promises should make us liable to lawsuits. Perhaps the state should not stand ready to
enforce promissory obligations when doing so would be self-defeating from the perspective of
accommodating the morally virtuous, though it is not clear whether this can explain why my friend’s promise
to meet me for lunch should not be in-principle enforceable.
15

Like Fried’s book, Shiffrin’s article spawned a minor cottage industry.
1
A few sympathetic voices
emerged, notably Fried himself, voicing tentative support for her criticisms of contract law. But by and large
the reaction has been critical. Some rehearse the conceptual objection that contractual obligations are simply
not a form of promissory obligation. Others accepted the premise that contracts involve promises, resisting
instead Shiffrin’s interpretation of ordinary promissory morality, adding that contract law already
accommodates minimally decent moral agents just fine. And within this vein of commentary another split
emerged. The first view argues that there really is no “divergence” between contract and promise, defending
the view that the common law of contracts largely harmonized with the demands of promissory morality. The
second view holds that although, yes, contracts are promises, and although, yes, contract law in some ways
fails to adequately reflect or respect the norms governing run-of-the-mill promises as they appear in ordinary
non-legal life, this is nevertheless a desirable situation worth preserving. Contractual promises and ordinary
promises are simply, according to this group, fundamentally different kinds of promises such that it does not
make sense for contract law to try to hew closely the informal practices of promising as they occur in the
context of, say, close relationships. Or so it has been argued.
2 From Promise Theory to Robust Promise Theory
I have singled out Fried and Shiffrin not just because I believe that they are the most influential
Promise Theorists, but also because I think that they offer the most interesting Promise Theories. I will call
them examples of Robust Promise Theory. This subset of Promise Theory is distinguished by a particular view
that they have towards the common law of contract and the way that they argue for their position. Their
accounts are the most interesting Promise Theories because they are implicitly committed to the view that
most common law jurisdictions—including Australia, England, and the United States—contain contract law
doctrines that inadequately respects the demands of promissory morality, and for that reason, stands in need
of reform.

1
Authors who respond to Shiffrin include Aditi Bagchi, Barbara Fried, Charles Fried, Jody Kraus, Daniel
Markovits, Liam Murphy, Michael Pratt, and many others.
16

We have already seen an example of this from Fried. Fried held, at least in his 1981 book, that the
doctrine of consideration should be rejected, at least in part on the grounds that promising contains no moral
analogue. The implication of this revision, if accepted, would be that many unilateral promises—promises
that are not made in exchange for anything valuable—would be in principle legally enforceable; this is a
significant expansion of the protection for promissory rights, since the doctrine of consideration functions to
screen out unilateral promises and leave them largely unprotected in common law jurisdictions.
We have also already seen some suggestions from Shiffrin, who points out that the lack of punitive
damages, the rare enforcement of specific performance, and the plaintiff’s duty to mitigate damages,
collectively and individually suggest that promissory rights need not be taken as seriously as they should be.
This provides some grounds, albeit indirectly for Shiffrin, to considering reforming our contractual practices.
What makes these versions of Promise Theory Robust is that they offer fairly concrete criticisms and
recommendations for reform, and moreover, these criticisms and recommendations flow from each author’s
strong views about what promissory morality demands from promisors.
In a moment I will explain what motivates this dissertation’s overall project: to assess whether a
particular theory of promissory obligation serves the agenda of Promise Theory, or to identify any problems
in trying to ground Promise Theory in specific theories of promissory obligation. But I will also discuss,
where convenient, whether a given theory serves the agenda of a Robust Promise Theory that aims to
strengthen contractual rights. Sometimes I do this to illustrate ways that Promise Theory might be brought to
bear on practical issues in contract law. But mostly, on the few occasions when I invoke Robust Promise
Theory, I do so in the service of making points about Promise Theory more generally.
3 This Dissertation and Its Motivations
So what? Originally, this dissertation’s aim was simple to state but difficult to execute: to provide a
new Promise Theory of contract law that avoided some of the difficulties of the predecessors. A Promise
Theory, to my mind, should have some ability to justify contract law and to do so in a particular way. Contract
law involves, quite minimally, a set of legal rules pursuant to which people can promise to do things for each
other and expect that these promises will be legally enforceable. Contract law also confers a right on
17

promisees to sue their promisors for not keeping their promises. So ideally, a Promise Theory would be able
to say something about why the state should be in the business of enforcing promises at all. A Promise
Theory should also have some ability to guide the shape that contract law doctrine takes. This might mean that
Promise Theory should have the ability to tell us the conditions under which contractual promises should be
binding, or what the proper remedies should be for breach of contract. And we should not be surprised if the
justificatory and guidance roles of Promise Theory inter-relate.
I said that the original aim of this dissertation was to come up with a new Promise Theory. The
motivation for this aim was nothing more than a vaguely optimistic sense that we might do better than Fried
or Shiffrin. But I quickly realized that part of the problem in defending Promise Theory involves identifying a
plausible conception of promissory morality well suited to the task. Now, in principle this task involves
simply presupposing the correct conception of promissory obligation, and then drawing out the implications of
this theory (if any) for a Promise Theory of contract law. Charles Fried did this when he adopted a
conventionalist account of promising, Dori Kimel did this later on when he adopted an interest-based theory
of promising en route to defending a kind of quasi-Promise Theory, and others take on the same approach
time and again, singling out or presupposing that their respect views (the correct one, of course) represents
the right way to explain promissory obligation, and proceeding to draw lessons about what contract law
should or should not do. None devote much time to considering plausible competing theories of promissory
obligation or what they offer Promise Theory. Unsurprisingly, they spend even less time considering how
their substantive Promise Theories would change, if at all, if a different theory of promising were correct.
Critics of Promise Theory sometimes take a similarly myopic focus. Their critiques gain traction only
to the extent that they make correct assumptions about what promising involves. Some writers assert, for
example, that promising necessarily involves a promisor’s subjectively intending to be morally bound, and
then claim that no such subjective intention is required to form a binding contract, offering this mismatch as
a refutation of Promise Theory. But not every theory of promising agrees that validly promising requires
subjectively intending to be bound, morally or otherwise. So these critics might miss their target entirely. As
long as Promise Theorists rely on their own preferred understandings of promising, and as long as the critics
18

do the same, we will make little progress on the question of whether Promise Theory is a viable normative
approach to contract law.
Of course the question of which theory of promising is correct is important. A Promise Theory
should not be built upon crumbling foundations. For that reason, the first four chapters considers objections
to the theories of promising on their own terms, independent of concerns about contract theory. As we will
see, some theories (in my view) better stand up to scrutiny. But instead of proceeding to presuppose or argue
for a single “best” theory of promissory obligation and trying to build a Promise Theory on top of it, I take
up an additional task in the first four chapters of this dissertation, one that is more difficult to state. After
considering four plausible conceptions of promissory morality, I then ask what more-or-less unique
challenges each might face in grounding a Promise Theory of contract law. This task looks critical—and most
certainly is critical—but the motivation was ultimately a sympathetic one. My hope was that one of these
theories would clearly emerge as best able to serve the agenda of Promise Theory, given its agenda of
justifying and guiding contract law, bracketing (for the most part) the question of whether that theory is
correct or true. I had hoped to make some progress for Promise Theory that would go beyond simply
assuming that one single theory of promising is the correct one at the outset.
After all is said and done, I am skeptical that any of the four theories that I identify “best” serves
Promise Theory’s agenda in any obvious way, which should not be surprising given my almost exclusive focus
on the challenges that they face rather than the opportunities they provide. And because I spend most of my
time discussing how each theory presents raises more-or-less unique challenges for Promise Theory, I say
nothing at all about the stock objections to Promise Theory until Chapter Five. So any progress I have made
in trying to advance Promise Theory by investigating prominent views on promissory obligation is quite
modest. I mostly end up pointing out pitfalls to avoid for future Promise Theories that wish to take up a
particular conception of promissory obligation and run with it.
But I am getting ahead of myself; I will summarize the chapters below. Let me first say a few words
(perhaps not enough) about methodology. The first four Chapters discuss a prominent and plausible
conception of promissory obligation and its implications, if any, for Promise Theory. It is reasonable to
19

wonder why I evaluate these theories and not others in the vast an expanding literature on promising. To
make the task manageable, I selected the theories using two criteria. The first is fairly obvious. I sought
influential, well-regarded, and highly cited philosophical contributions to recent philosophical debates about
promissory obligation. But this criterion under-determines the class. The second criterion was that I wanted
to select theories that were independently motivated; I wanted them to answer to some important question in
the theory of promising that was at some remove from contract law. My thought was that if the theory of
promising begins to look too much like a theory of existing contract law, the theorist is trying (intentionally or
not) to make the theory of promising fit contract law rather than to respond to some independent
philosophical worries about promising.
To ensure this independent motivation, each of the theories that I discuss represents a way of
responding to one of the central questions in the theory of promising, Hume’s Challenge. The challenge has
been characterized in various ways, but perhaps the simplest is as a basic question about promising: how is it
possible that we can by bind ourselves morally by simply declaring an intention to undertake an obligation to
perform some action for someone else? Interpreting this question and why it is supposed to be puzzling is no
easy task. But as a first pass, notice that, try as I might, it looks near impossible to impose on myself a moral
obligation to exercise simply by willing such a moral obligation into existence. This is so even if my New
Year’s Resolution to exercise involves a bona fide commitment. Still, this commitment lacks a moral flavor.
So how do promissory commitments acquire their moral flavor? The philosophical theories of promising that I
focus on all propose answers to this question and to that extent are independently motivated.
Unfortunately, despite my hope that one theory of promissory obligation would emerge as the one
that best serves Promise Theory’s agenda, Promise Theorists have reason to worry about relying on each of
the four theories. What’s more, the first four Chapters do not addresses the two stock objections to Promise
Theory. So even if one of the theories did emerge as the best one, more work would need to be done. This
concern about the stock objections motivates the fifth and final Chapter, where I propose an answer to the
stock objections according to which Promise Theory plays a supporting rather than central role in justifying
20

and guiding contract law. The role is still an essential one, as we will see, but implicitly concedes that some of
the critics were correct: Promise Theory alone cannot serve to justify contract law.
4 Chapter One: Contracts and the Practice Theory of Promising
Chapter One discusses the so-called Practice Theory of promising. The Practice Theory responds to
Hume’s Challenge by asserting that our ability to self-impose moral obligations depends on a widespread and
valuable social practice of promising. Our moral obligations to keep the promises we make derive from
general background obligations to comport with the dictates of valuable social practices—including
promissory practices. It is an implication of this view that without a widely accepted habit or practice of
making and keeping promises, no genuine moral obligations to keep one’s promises emerges. I argue in
Chapter One that Practice Theory makes it difficult for Promise Theories to establish what I call the
“normative priority” of informal promissory practices over the formal ones established by contract law. This
normative priority is required for any criticism of contract law or call for its reform grounded in claims about
what promissory morality requires. Indeed, this normative priority of promise over contract is presupposed by
Promise Theorists, which try to try to justify and guide contract law by reference to promissory obligations
and the practices associated with them.
I will argue that this normative priority is difficult to justify if Promise Theory presupposes a Practice
Theory of promising. Here is one way to put the difficulty. The problem, at least initially, is that under
Practice Theory, it is not obvious why one set of promissory practices (the one that constitute informal
promissory norms) should provide some justification and guidance for another set of promissory practices
(the one that constitute contract law), rather than vice versa. After all, why reform or criticize contract law for
failing to adequately respect or reflect or somehow fit informal promissory practices, rather than seeking the
reform of those informal promissory practices in light of contract law? This would also mitigate the alleged
incongruity between promise and contract.
Here is another way to put the point. In the first instance, contract law and ordinary promises are
simply two versions of the practice of promising that are morally on par, at least qua being social rules and
conventions that share basically the same normative structure (insofar as they are both identifiable as forms
21

of promising). So if ordinary promissory expectations must govern contract law the social rules themselves
look unlikely to explain why; the normative priority must come from elsewhere. The natural place to look is
in the value underlying the ordinary practice of promising. But even if we assume that contract law and
promising serve the same value (e.g., autonomy), that value may very well count against accepting certain
putatively promise-respecting reforms—at least those suggested by Robust Promise Theorists. More
generally, all the normative work in favor of reforming or designing contract law would be performed by the
values supported by the practice of promising, not any claims about what the informal promissory practices
demand.
Of course much more will be said in Chapter One. But if I am correct, the fact that Practice Theories
of promising provide a shaky foundation for Promise Theory is a surprising result. After all, Charles Fried
explicitly rested his Promise Theory on a Practice Theory of promising. And Seana Shiffrin asserts, in her
more recent defense of Promise Theory, that her criticisms of the common law of contract hold even if
Practice Theory is true. Chapter One calls both positions into question.
5 Chapter Two: Contracts and Agency Theories of Promising
Chapter One motivates a search for non-practice-based theory of promissory obligation, one that
faces fewer difficulties in accounting for the normative priority of promissory norms over contract law. To
that end, I turn to other theories of promissory obligation that answer Hume’s Challenge without relying on
social practices of making and keeping promises. Chapters Two and Three turn to Normative Power
Theories (NPTs) of promising. NPTs in effect respond to Hume’s Challenge directly, arguing that we do have
the power to impose on ourselves moral obligations to perform actions by making promises; that is, we have
a moral power to bind ourselves simply by declaring the intention to do so under certain circumstances. This
moral power, in turn, is grounded either in moral agency itself or in certain interests or values—not
necessarily presupposing that any convention or entrenched social practice exists to confer that power on us.
Because the Normative Power Theories that I discuss do not rely on social practices to explain the power to
promise, I will refer to them Moral Power Theories.
22

Chapter Two has several aims. First, I argue that one of the most well-developed and recent attempts
to ground a moral power to promise in moral agency, offered by Seana Shiffrin, is incomplete. The Chapter
begins with Shiffrin’s elegant suggestion that promising to do something involves nothing more than
transferring to someone a right to decide to do otherwise. The suggestion is elegant because the theory’s
inputs involve little more than normative items that seem relatively uncontroversial; it is relatively
uncontroversial (though like anything else in philosophy not wholly uncontroversial) that moral agents have
moral rights, and the power to give up, waive, or transfer these rights to others. If we can capture the moral
power to promise in terms of these comparatively less controversial powers, Shiffrin will have provided one
compelling answer to Hume’s Challenge. Shiffrin also offers an independent, transcendental argument in
favor of the claim that we have a moral power to promise. The argument holds, very roughly, that being a
minimally decent moral agent requires having the moral power to promise, and because minimally decent
moral agents exist, we can infer that we have the moral power to promise.
Chapter Two criticizes both the rights-transfer view and the transcendental argument. The main
worry with the rights-transfer view is that it does not rule out the possibility that rights transfers are revocable. I
call this the Irrevocability Problem. Imposing promissory obligations are normally exercises that are
irrevocable; one morally cannot unilaterally un-promise to perform a validly promised act. I argue that, with
respect to Agency Theories, explaining why rights transfers are irrevocable is difficult because they lack the
resources to do so; Agency Theory must go beyond mere moral agency and rely on moral norms that govern
those agents to explain the irrevocability of promising. This effectively goes beyond the Agency Theory’s
agenda. With this problem in mind, I construe Shiffrin’s transcendental argument in favor of a moral power
to promise as an attempt to establish that we have a sui generis power to irrevocably impose promissory
obligations on ourselves. I argue, however, that her argument does not succeed because she does not show
that promising is indispensable to live a minimally decent moral life.
Returning to contract law, I argue that a Promise Theory that relies on an Agency Theory of
promising will be similarly incomplete. One of the core features of contractual obligations is that they are
supposed to be irrevocable. Indeed, one motivation for Promise Theories to begin with was the fact that both
23

promissory and contractual commitments share important normative features, including this one. But relying
on Agency Theory makes justifying this very feature difficult without relying on some other normative
resources that go beyond the building blocks of moral agency. So lacking an account of irrevocability in the
context of a theory of promising creates a gap in the contractual one.
6 Chapter Three: Contracts and Interest Theories of Promising
Chapter Three examines another set of Moral Power Theory, Interest Theories. Interest Theories
claim that normative powers—and the moral power to promise in particular—are grounded primarily in
certain interests or values rather than in the basic capacities that make us moral agents. Essentially, it is
claimed that we have certain moral powers because it would be good to have them. Joseph Raz and David
Owens, the most prominent Interest Theorists, emphasize different interests. Early on, Raz emphasized the
interest in forging special bonds with others; more recent work emphasizes the interest in control over our
normative lives. David Owens argues that making sense of promising requires understanding promising in
terms of so-called normative interests—and in particular, the interest we have in having authority over others.
Much of the Chapter puts pressure on Raz’s early views and Owens’s theory. Against Raz’s earlier
views, developed in the 1970s and 1980s, that the value of the power of promising lies in its connection with
forging special obligations, I argue that this view faces a dilemma. If we understand this view as supposing
that that promissory obligations forge intrinsically valuable relationships, I argue that this is not so; not all
genuine promisor-promisee relationships are intrinsically valuable. If we understand Raz’s view as an
instrumental one, focusing on the usefulness of the power to promise in forging and managing intrinsically
valuable relationships, then it is not obvious why intrinsically valuable special relationships should take pride of
place, as opposed to other values that plausibly are also instrumentally served by having the power to
promise. I take Raz’s unacknowledged shift to a different view, which grounds the power to promise in our
interest in exercising control over our lives, as evidence indicating that he recognizes some difficulty in his
early views. I also claim that this later view is more plausible in any event.
I advance several objections to Owens’s authority-interest-based theory of promissory obligation. For
one thing, we should doubt whether his theory is meaningfully different from an Agency Theory of
24

promising of the kind encountered in the previous Chapter. Independent of this concern is a more basic one:
merely having the authority to place others under an obligation does not seem intrinsically valuable,
independent of other non-normative interests that having that authority might serve. The idea verges on
power fetishism. But this is precisely what Owens claims: that we have a sui generis “normative” interest in
having authority over others, which grounds the power to promise.
Returning to contract law, I argue that Interest Theories of promising invite the question of whether
the interests that ground the legal power to contract must be primarily or exclusively the same as those that
ground the moral power to promise, whatever the latter interest might be. Joseph Raz, who comes close to
suggesting that the same interests ought to ground both normative powers, does not supply an argument in
support of this claim. Obviously this does not show conclusively that Interest Theory provides a poor
foundation for contract law, though it does complicate matters, since the interests or values that ground
promising—whatever they might be—also compete with other interests or values that ground the legal power
to contract. This is not a wholly unique concern for Interest Theory; some version of this worry remains
operative for all attempts to articulate a Promise Theory. But Interest Theory highlights the problem in
perhaps its starkest form.
7 Chapter Four: Contracts and the Assurance Conception of Promising
This chapter turns away from attempts to ground the moral power to promise directly in some
agential capacity or interest or value. I argue that the right way to account for promissory obligations,
following T.M. Scanlon, is to focus first on what makes breaking a promise to someone wrong when it is
wrong. In this chapter, I draw on Scanlon’s work to argue in favor of a few moral principles that purport to
explain the wrongness of promise breaking, including primarily a modified moral principle of fidelity. The
principle holds, roughly, that if one person x tries to assure another y that x will perform some action for y,
and x and y both know that x is trying to do this, this generates an obligation for x to perform that action for
y, absent special justification. This is a principle of fidelity that grounds promissory obligation.
The moral principles that determine promissory obligations avoid the particular problems faced by
the previous views. The Scanlonian approach does not depend on the existence of a widespread practice of
25

promising; we may form promissory obligations without such a practice because the principle holds
regardless of current practices. So the view avoids the difficulties of Practice Theory. The assurance
conception also captures the most salient features of promising just as well as the two Moral Power Theories
encountered in the prior chapter. In particular, the notion that promising involves an assignment of rights
(per Agency Theory) falls naturally out of the view. x’s triggering an obligation under conditions set forth in
the principle is tantamount to assigning the correlative performance right to y. And importantly, according to
the view I favor, the irrevocability of promissory obligations is simply a built-in feature of the moral
principles that ground the moral power to promise, since x must perform the promissory act absent special
justification. This avoids some of problems faced by the two Moral Power Theories. So if we are interested in
grounding contract law in a theory of promissory morality, then the Assurance Conception looks like the best
candidate from the views canvassed so far.
This is not to say that the fidelity principles face no specific problems with respect to contract law. I
consider a problem that is built in to Scanlon’s theory from the beginning. I consider what I call the right-
kind-of-reasons objection, which adds the following: a fair implication of Scanlon’s account is that the right
kinds of reasons for complying with a promise necessarily include moral reasons or other-regarding reasons.
But the right kinds of reasons for complying with contracts do not necessarily include moral or other-
regarding reasons. And if the set of right-kind reasons for complying with two forms of commitment differ,
then the commitments are fundamentally different in nature. So promising cannot provide the foundation for
contract because they are fundamentally different. I attempt to respond to this objection but do not purport
to do so decisively. My main goal in articulating the objection is to point out a potential pitfall in trying to
ground a Promise Theory in a Scanlonian account.
8 Chapter Five: Contracts as Commodified Promises
Throughout the first four Chapters, I avoid the so-called stock objections to Promise Theory,
objections that purport to hold no matter which theory of promissory obligation is selected to ground
Promise Theory. Instead, I point out how each of the four theories of promissory obligation raise or highlight
more or less unique problems for attempts to ground Promise Theory in those theories.
26

Chapter Five changes course. I will assume for the sake of argument that some correct theory of
promissory obligation best serves or can be reconciled with Promise Theory. Even so, we would need to
answer the stock objections to Promise Theory. Recall what those objections were: first, that Promise Theory
does not explain why states should be in the business of enforcing promissory obligations in the first place,
and second, Promise Theory faces the related problem of explaining why some but not all promises should
be in principle enforceable by legal authorities.
This Chapter begins by giving an Instrumental Justification for contract law. The justification is a
familiar one: I argue that contract law is justified by reference to its ability to facilitate market economies. The
state has a legitimate interest in fostering and regulating commercial markets given the many benefits they
bring. Facilitating markets, however, practically requires something like a contract law, because having a
robust market economy requires enforcing some promises. Promissory obligations, apart from being
characteristically irrevocable, are also characteristically inter-temporal and future-oriented. The law’s readiness
to enforce certain promissory obligations provides the needed assurances to mitigate trust deficits and
provide the requisite assurance of performance.
So in response to the first stock objection, why enforce promises, the Instrumental Justification
answers: to facilitate markets. The Instrumental Justification also answers the second stock objection, which
insists that Promise Theory does not provide an answer to the question of which promises will be enforced.
The Instrumental Justification answers: commercial promises. The Instrumental Justification thereby provides an
essential role for the enforcement of promises, vindicating (in a sense) Promise Theory by locating it in a
larger instrumental project.
This does not read Promise Theory out of the picture entirely. Contract law’s pursuit of markets is
not unlimited, any more than the criminal law’s pursuit of criminal deterrence permits criminal law to punish
the innocent. The state enlists individuals, and their status as holders of promissory rights and obligations, in
the service of a market-facilitating public policy. Individuals’ moral statuses as holders of these rights and
obligations must be adequately respected. Chapter Five does not, however, explain what this means or what
implications this claim would have for contract doctrine. Instead, I turn to another, more overlooked issue:
27

what does it mean for a promise to be a commercial one, and what implications does this have for contract law
and theory?
I sketch an answer to these questions in this Chapter, advancing a view I call Contract as Commodified
Promise (CCP). This view holds, first, that the fact that a promise is commodified provides a strong pro tanto
reason favoring its legal enforcement, and second, that the fact that a promise is not commodified, or is
commodified improperly, provides a strong pro tanto reason against its enforcement. Explaining these claims
will take up most of the Chapter. But one of the main motivations for the view is that not only do contracts
typically involve the enforcement of commercial promises (as predicted by the Instrumental Justification), it is
that we can usefully explicate the nature of commercial promises in terms of promissory rights and
obligations that are treated as commodities available for purchase or sale. I hope to show how these claims illuminate
certain existing doctrines in contract law, including the doctrine that illegal contracts and contracts against
public policy are void, and the doctrine limiting the enforcement of so-called donative promises.
CCP also provides new normative resources to recurring debates about contract remedies, including
claims made by Robust Promise Theorists that contract law fails to adequately respect promissory morality.
In particular, I will argue that we can make sense of these claims or re-characterize them in terms of debates
about whether contract law over-commodifies or under-commodifies promises. I do not attempt to resolve
these disputes, but rather to show how they can be cast in a new and potentially useful light.
In many ways, the dissertation that I had in mind to begin with evolved, as evidenced by my fairly
drastic shift in course from Chapter 4 to Chapter 5. I suppose some evolution is par for the course when
writing a dissertation. It’s called “learning,” I’m told. But in many ways the views expressed in Chapter 5 are
sketchy and will have to remain so for present purposes. Making them less sketchy will require nothing less (I
fear) than a new dissertation, a project that I am eager to defer indefinitely.
28

Chapter One: Contracts and the Practice Theory of Promising
1 Introduction
Most common law jurisdictions bar plaintiffs from collecting punitive damages for their breach-of-
contract claims.
2
Courts in these jurisdictions also decline to enforce penalty clauses, even when agreed to at
arm’s length by parties of roughly comparable bargaining power.
3
But things could have been different. After
all, civil law jurisdictions enforce penalty clauses.
4
And France recently debated making punitive relief
available even absent prior agreement.
5
Commentators have reportedly called for similar reforms elsewhere—
even in common law jurisdictions.
6

Scholars have begun to lay the groundwork for this kind of punitive damages reform in contract
law—and, indeed, more wide-ranging reforms that would make the protection of contractual rights more
robust in general. Professor Seana Shiffrin, for example, has argued in recent years that the rule barring
punitive damages represents only one of a number of ways—including the duty to mitigate damages—that
the common law of contract inadequately respects the demands of promissory morality.
7
She draws

2
Unless, that is, accompanied by an action constituting an independent tort for which punitive damages are available.
One scholar traces this exception, where it exists in the United States at least, to South Carolina. Laurence P. Simpson,
Punitive Damages for Breach of Contract, 20 OHIO ST. L. J. 284 (1959) (tracing the origin of the independent-tort exception in
the United States to a case decided in South Carolina in 1904). The Supreme Court of Canada has upheld punitive
damages awards in exceptional cases. See, e.g., Whiten v. Pilot Ins. Co., [2002] 1 S.C.R. 595, 596 (involving “exceptionally
reprehensible” conduct of defendant insurance company).
3
See, e.g., E. ALLAN FARNSWORTH, CONTRACTS § 12.18 (4th ed. 2004).
4
Jack Graves, Penalty Clauses as Remedies: Exploring Comparative Approaches to Enforceability, 29 TOURO L. REV. 681, 682
(2013) (observing that “penalty clauses are generally enforceable in civil law systems”).
5
Solène Rowan, Reflections on the Introduction of Punitive Damages for Breach of Contract, 30 OXFORD J. OF L. STUDIES 495, 513
(2010) (“[T]here has recently been debate in France regarding whether the unavailability of punitive damages should be
revisited. A proposal to reform the Civil Code 133 has, amongst other things, propounded the idea of punitive damages
being added to the arsenal of French remedies.”).
6
See id. at 496 n.4 (2010) (collecting citations); Seana Shiffrin, The Divergence of Contract and Promise, 120 HARV. L. REV.
708, 752-53 (2007) [hereinafter Shiffrin, Divergence of Contract] (concluding that “there may be reason to reexamine these
[contract] doctrines and their justifications, and to strive for greater convergence between promise and contract,”
including the doctrine providing for “the general unavailability of punitive damages in contract”); Charles Fried has also
expressed sympathy for relaxing the rule against punitive damages. See Charles Fried, Convergence of Contract and Promise,
120 HARV. L. REV. F. 1, 7 (2007) (“As for the risk of a mistaken outcome that the disappointed promisee is forced to
encounter when the defaulting promisor does not tender expectation damages at the outset, that is where Professor
Shiffrin’s plea for punitive damages has some plausibility.”).
7
See Shiffrin, Divergence of Contract, supra note 6, at 752-53 (concluding that “there may be reason to reexamine these
[contract] doctrines and their justifications, and to strive for greater convergence between promise and contract,”
including the doctrine providing for “the general unavailability of punitive damages in contract”).
29

normative force for these claims by relying on further assertions about what promissory morality requires.
Others sympathize with Shiffrin.
8
Together they represent particularly robust and reform-minded version of a
general position that I will call Promise Theories of contract law. Promise Theories aim to justify contract law
and guide its doctrine by reference to the moral demands and expectations governing the formation and
performance of promises. Shiffrin’s version counts as robust given that her Promise Theory gives rise to fairly
specific criticisms against the common law of contract on the grounds that it inadequately respects
promissory morality. For this reason, I will call Shiffrin’s view a Robust Promise Theory.
This paper concerns the viability of Promise Theory generally and Robust Promise Theory in
particular. One of my claims is that their plausibility depends on the underlying theory of promissory morality
that they presuppose.
9
This claim matters because scholars periodically evaluate promise-based theories of
contract law with scant attention to the promissory theories underlying them, as if these theories were
fungible. Even Shiffrin argues in favor of more robust protection of contractual rights while assuming that
the underlying theory of promissory obligation matters little in pursuing her argument.
10

This first claim follows from my second and main claim: I will argue that one of the most influential
families of views in the philosophical literature on promising—a view held by David Hume and John Rawls,
among others—undermines, if true, Promise Theories of contract. Members of this family are called Practice
Theories of promising.
11
Practice Theories hold that (1) promissory obligations depend on the existence of a

8
See Charles Fried, Convergence of Contract and Promise, 120 HARV. L. REV. F. 1, 7 (2007); see also Liam B. Murphy, Contract
and Promise, 120 HARV. L. REV. F. 10, 15 (2007) (“Treating the morality of promise as itself parasitic on a social practice
that is worth having because of its good effects, I see contract law as valuable primarily because it supports, and in fact
partly constitutes, that social practice. … Contract theorists that agree with this are unlikely to recommend the kind of
divergence between contract and promise that Professor Shiffrin finds troubling.”).
9
To my knowledge, only Jody Kraus has acknowledged something like this point. See generally Jody Kraus, The
Correspondence of Contract and Promise, 109 COLUM. L. REV. 1603 (2009).
10
See Seana Shiffrin, The Divergence of Contract and Promise, 120 HARV. L. REV. 708, 720 n.18 (2007) (“Most of this Article’s
points, however, do not depend on the rejection of conventionalism but only on the moral rules of promising being
understood in the broad terms I describe.”). Suffice it to say that, if I am correct, then this claim is false.
11
DAVID HUME, A TREATISE ON HUMAN NATURE 520-21 (1978) (eds. L.A. Selby-Bigge, P.H. Nidditch), and JOHN
RAWLS, A THEORY OF JUSTICE 303 (1999). See also Charles Fried, CONTRACT AS PROMISE: A THEORY OF CONTRACTUAL
OBLIGATION 17 [hereinafter, FRIED, CONTRACTS AS PROMISE] (“There exists a convention that defines the practice of
promising and its entailments.”). A more recent defense can be found in Niko Kolodny and R. Jay Wallace, Promises and
Practices Revisited, 31 PHIL. & PUB. AFFAIRS 119 (2003).
30

widely followed social practice of making and keeping promises that (2) is justified because it promotes or
allows us to realize certain values.
12
Meanwhile, Robust Promise Theories hold: (3) that the fact that
contractual obligations are promissory ones justifies a great deal of contract law, and moreover, to the extent
that the law inadequately reflects the promissory nature of contractual obligations, we should reform contract
law.
13

These claims look plausible and easily reconciled. But I will argue that if (1) and (2) are true, this
undermines support for (3). Likewise, if (3) is true, then this should put pressure on one’s commitment to (1)
and (2). This is significant because one of the most influential contract law theorists—Charles Fried—
endorses all three claims.
14
More generally, once we recognize what makes this tripartite package problematic,
this should motivate Robust Promise Theorists of contract law to reject Practice Theories of promising and
vice versa.
15

To preview the argument, I claim that Robust Promise Theories, in accepting (3), presuppose that
the informal social practice of promising, and the rules constitutive of that practice, are privileged or take
normative priority over the rules constitutive of contract law. There is no other way to make sense of (3): if we
should reform contract law to reflect the informal practices of promising (rather than vice versa), then our

12
See, e.g., HUME, supra note 1 (arguing that promising is a practice established for “mutual advantage” or “common
benefit”).
13
See FRIED, CONTRACT AS PROMISE, supra note 11; see also Shiffrin, Divergence of Contract, supra note 6 at 752-53.
Although it is clear that Seana Shiffrin accepts something like (3), she does not hold a practice-based view of promissory
obligation—so she rejects (1) and (2). See Seana Shiffrin, Promising, Intimate Relationships, and Conventionalism, 117 PHIL.
REV. 481 (2008). Assumption also (3) seems to be at work in Frank Menetrez, Consequentialism, Promissory Obligation, and
the Theory of Efficient Breach, 47 UCLA L. REV. 859, 879–80 (2000) and Peter Linzer, On the Amorality of Contract Remedies —
Efficacy, Equity, and the Second Restatement, 81 COLUM. L. REV. 111 (1981).
14
See generally FRIED, CONTRACTS AS PROMISE, supra note 11.
15
Liam Murphy accepts (1) and (2), and appears also to accept (3), as well. He argues that we probably should not permit
efficient breach in contract law given that doing so will likely undermine the practice of making and keeping promises
and agreements, and regards this as a very strong if not decisive reason to avoid adopting this notion in contract law. It is
hard to see how this form of argument does not presuppose a stance akin to (3). Murphy’s apparent presupposition of
(3) is further bolstered by his support for some of Shiffrin’s prescriptive claims, which clearly do presuppose or require
something like (3). See Liam B. Murphy, Contract and Promise, 120 HARV. L. REV. F. 10, 15 (2006) (“Treating the morality
of promise as itself parasitic on a social practice that is worth having because of its good effects, I see contract law as
valuable primarily because it supports, and in fact partly constitutes, that social practice. … Contract theorists that agree
with this are unlikely to recommend the kind of divergence between contract and promise that Professor Shiffrin finds
troubling.”).
31

ordinary, informal practices of promising somehow take normative priority over the formal practices that
constitute contract law. Indeed, even ordinary Promise Theories presuppose this normative priority to some
extent, insofar as they try to ground contractual obligation in promissory obligation.
The problem is this: If the informal social practice of promising is valuable primarily because it
promotes or realizes some value, then it is an open question as to whether more “faithfully” enforcing that
practice promotes or realizes that value better than enforcing a version of the practice that would count as a
“defective” version of the informal one. Indeed, if contract law promotes or realizes that value better by
imperfectly respecting the informal practice, then this undermines the grounds for privileging the informal
practice. I’ll further suggest that modern common law of contract plausibly does promote or realize certain
values better as compared to a contract law that would more faithfully respect the informal practice, as
Robust Promise Theorists propose. Indeed, adopting a practice-based theory of promissory obligation makes
arguing for doctrinal reform on the grounds of Promise Theory—of both Robust and non-Robust varieties—
much more difficult. And if this is true, then, as my first goal suggests, the plausibility of Promise Theory is a
function of the theory of promissory morality that Promise Theory presupposes.
2 Practice Theories of Promising
Some background on Practice Theories—including their motivation and content—is necessary.
Practice Theories of promising start with a puzzle. Somehow we are able to create for ourselves moral
obligations to perform certain actions for others by expressing an intention to do so. David Hume claims,
most prominently, that this seems puzzling because we are able to, by word or deed, magically transform an
undertaking that might have been morally insignificant into something morally obligatory.
16

Practice Theorists aim to explain, in two steps, how this is possible. The first step points out that we
have a valuable social practice of promising. The practice is constituted by certain social rules, which is to say
that there are patterns of conduct that are widely accepted by many people in a community.
17
Not all rules

16
DAVID HUME, A TREATISE ON HUMAN NATURE (1978) (eds. L.A. Selby-Bigge, P.H. Nidditch)
17
ANDREI MARMOR, SOCIAL CONVENTIONS 3 (2009) (indicating that one of the hallmarks of a social rule is that it is a
“rule that, by and large, followed by a population” and further suggesting that social conventions are a special kind of
social rule).
32

that constitute the practice are equally important, of course. Practice Theorists are interested primarily in the
central constitutive rules of promising—the rules that make a practice promising and not some other activity.
18
Call
any non-central features of the practice ancillary. Ancillary features vary from community to community, and
sometimes amount to mere linguistic differences (“I promise you that…” instead of “Yo te prometo que…”).
Other ancillary rules might include certain presumptions about the content of promises once made. Perhaps, in
some promising cultures, there is a presumption that—absent express statement to the contrary—the very
person who made them, as opposed to third-party substitutes must discharge promissory obligations. Other
cultures might lack such a presumption. In such cultures or sub-cultures, making a promise implicitly suggests
that either the promisor or someone else will perform the action in question. But it seems that, regardless of
whether a promising culture contains this presumption, we would still recognize the culture as having a
practice of promising as opposed to some other practice.
With this central/ancillary distinction in mind, we should point out two central rules that Practice
Theorists (and others) have emphasized as particularly important. First, if you behave in such a way that
constitutes making a promise (as determined by the particular rules and conditions governing the formation of
promises), then you must keep the promise unless the promisee excuses your performance or unless non-
performance is otherwise justified.
19
Keeping your promise requires actually doing what you have said you will
do for another person. Second, the grounds for excuse and justification are tightly circumscribed. For example,
if you promise to do X at time t, then your mere preference to do something other than X at t, even if it is a

18
Actually, Rawls claims that the rule requiring keeping one’s promise is a constitutive convention. See Rawls, supra note 11, at
303. It is not obvious that this is correct, though undoubtedly some rules associated with promising are conventional. See
DORI KIMEL, FROM PROMISE TO CONTRACT: TOWARDS A LIBERAL THEORY OF CONTRACT 8-10 (2003); ANDREI
MARMOR, SOCIAL CONVENTIONS 136-140 (2009). In any event, Rawls and other practice theorists need only claim that
the rules constitutive of promising are social rules whose existence depends on a critical mass of the population regularly
following them and taking them to be binding, but which are not themselves moral principles. See Rawls, supra note 11,
at 306 (“Even the rule of promising does not give rise to a moral obligation by itself.”) (emphasis added).
19
See, e.g., RAWLS, supra note 1, at 303 (asserting that “[i]n the case of promising, the basic rule … reads roughly as
follows: if one says the words ‘I promise to do X’ in the appropriate circumstances, one is to do X, unless certain
excusing conditions obtain” and that the previously stated rule “may be taken as representing the practice as a whole”).
33

rather strong preference, does not by itself excuse you from doing X.
20
These features are deeply ingrained in
the practice of promising, so much so that failing to understand them amounts to failing to understand the
practice. I’ll leave open the question whether other norms associated with promising qualify as central. The
important point is that the two central rules just highlighted are, to Practice Theories, actually existing social
rules. This means that, without these rules, the practice of promising does not exist in a society.
21

En route to explaining, in the second step, how promising generates moral obligations, Practice
Theorists point out that the practice of promising is valuable and try to explain why. These explanations do
not list the innumerable ways we might enlist promises in the service of some value, obviously. By promising
to go to the movies with you, I’ve thereby enlisted a promise to reinforce our friendship and to patronize the
arts. But we do not therefore generalize and conclude that the value of the practice of promising is to bolster
friendships and encourage aesthetic pursuits. Theorists try to identify the value or values that the practice of
promising in general serves.
We can specify the relation between the practice and its value in two ways. The practice might be
valuable because it promotes some further practice-independent value. On this view, the relation between the
practice and the value is instrumental: the practice is a useful means to promoting some valuable end.
Different Practice Theorists emphasize different values. Utilitarians argue that having a well-established social
practice of promise keeping generates more welfare than the alternative of not having one or having a weak
one. After all, promising facilitates cooperation and trust, which in turn increases welfare overall by
facilitating mutually beneficial transactions.
22
Charles Fried emphasizes autonomy rather than welfare. He

20
Rawls mentions other features of the practice of promising, such as the excusing condition that promises obtained in
reliance on deceit are not binding. But he acknowledges that these features of the practice would need specification.
RAWLS, supra note 11, at 303.
21
P.S. Atiyah has criticized Practice Theorists for failing to give detailed descriptions of what the practice consists of. See
P.S. ATIYAH, PROMISES, MORALS, AND THE LAW 138 (1981) (finding it “odd” that Practice Theorists “make no attempt
to inquire into the rules of this practice [of promising]”). This charge seems unfair, once we understand Practice
Theorists as primarily interested in the central constitutive rules, rather than the ancillary ones that may vary quite widely.
22
For an extended discussion of utilitarian attempts to account for promising on similar grounds, see P.S. ATIYAH,
PROMISES, MORALS, AND THE LAW (1981) (chapter 3).
34

observes that the same coordination-enhancing and trust-facilitating function of the practice is valuable
primarily because it enhances the conditions of autonomy by increasing our “options in the long run.”
23

But the relation between promising’s value and the practice might be non-instrumental. The ordinary
practice might be valuable because engaging in the practice instantiates, realizes, or manifests values.
24
The
very act of committing ourselves to others’ projects, and enlisting others in the service of our own, is a way of
exercising autonomy. To be sure, we can exercise our autonomy without a practice of promising. But the
practice’s value might be important primarily as a powerful vehicle for enabling individuals to realize that
important value in their own lives, regardless of whether it tends to increase the amount of that value either in
their lives or overall in a population.
25
The social practice makes this vehicle possible.
This completes the first stage. We’ve seen how, according to Practice Theory, (1) promising and
hence promissory obligation depends on the existence of social rules that (2) promote or instantiate certain
values. But a challenge remains. To see why, focus on the claim that the central constitutive rules of
promising are social rules. This opens up a challenge for Practice Theories. They still must explain how
promises generate moral obligations, and accordingly, why breaking a promise is morally wrong. After all, if
the rules governing promises are like rules of a game, or other social rules, we might worry that Practice
Theory cannot capture certain important moral differences between these two activities. Someone who
decides to play tennis ordinarily does not, for example, do something morally wrong in deciding to abandon
the match midway through, at least not without adding some details to the story. She just forfeits. The general
point is that declining to follow a social rule is not ipso facto morally wrong—it depends on the reasons that we
have for following it.
26

23
FRIED, CONTRACT AS PROMISE, supra note 13, at 13-14.
24
I will use the terms “instantiate,” “realize,” and “manifest” interchangeably.
25
Dori Kimel claims that this manifestation relation reflects the ways that the practice of promising is intrinsically
valuable. DORI KIMEL, FROM PROMISE TO CONTRACT: TOWARDS A LIBERAL THEORY OF CONTRACT 28 (2003). As I
use the term, an activity is ‘intrinsically valuable’ just in case a complete explanation of the practice’s value need not
mention anything beyond the activity itself. I take this to capture the common idea that intrinsically valuable activities are
activities that are valuable in themselves. I address the idea of how promising might be an intrinsically valuable practice
when I address objections later on.
26
I thank Andrei Marmor for this formulation of the point.
35

The second stage tries to mitigate this worry. One thought associated with Hume posits a general
duty not to undermine valuable social practices. And since breaking a promise involves undermining the
valuable social practice of promising, breaking promises violates this duty and is therefore morally wrong.
Rawls gives another response. He argues that complying with the practice of promising, once invoked, is
required by fairness.
27
Breaking promises is wrong because it involves unfairly free-riding on a just and
beneficial social practice. Finally, Charles Fried, argues that by making a promise we invite another’s trust in
us—trust enabled by having a widely accepted practice of promising. Respect for others demands that we
must keep our promises to avoid abusing that trust.
28
Breaking promises is therefore wrong because it usually
involves abusing others’ trust.
All of these Practice Theories share a common strategy to explain why we generate moral obligations
to keep our promises once we make them, and hence why breaking a promise involves moral wrongdoing.
First, posit a general duty to behave in certain ways vis-à-vis a social institution or practice, then explain how
breaking promises involves failing to behave in this way. These various possible relations between the social
practice of promising, its value, and promissory obligations will be important for my main argument later on.
3 Robust Promise Theories of Contract: General Commitments
Having examined Practice Theory and its commitments to (1) and (2), let’s examine Robust Promise
Theory. Its distinctive claim is (3), which holds that understanding contractual obligations as promissory ones
justifies a great deal of contract law, and moreover, to the extent that contract law inadequately reflects the
promissory nature of contractual obligations, we should reform contract law. Notice that (3) is not merely the
claim that contracts are promises, or that contract law aims to enforce some promissory obligations. Many
contract theorists accept these “orthodox” views.
29
But not all of them would accept (3). After all, one might
accept that contracts are promises while denying that this tells us anything important about what contract law
should look like. Richard Craswell, for example, argues that philosophical theories of promising tell us

27
See RAWLS, supra note 1, 306.
28
See FRIED, CONTRACT AS PROMISE, supra note 13, at 16-17.
29
See STEPHEN A. SMITH, CONTRACT THEORY 56 (2004) (calling the view that contractual obligations are promissory
ones the “traditional and still orthodox” view).
36

nothing useful about what the remedies should be for breaches of contract.
30
Other theorists might accept
(3)’s first conjunct but reject the idea that we have reasons to reform contract law on the grounds that it fails
to reflect informal promissory morality. Contract law might, after all, serve goals or values other than merely
enforcing certain private commitments.
31
If so, then it is not obvious why we should shape contract law to
better respect the informal practice of promising as (3) suggests.
But an influential subset of contract theorists—Robust Promise Theorists—maintain that not only
does contract law enforce promissory obligations, and not only does this insight help justify contract law, they
also hold that to the extent that the law inadequately reflects the fact that contractual obligations are
promissory ones, we should reform contract law.
32
This is commitment (3). To illustrate some reforms floated
by Robust Promise Theorists, Charles Fried has argued that we should eliminate the long-standing doctrine of
consideration.
33
As already noted, Seana Shiffrin suggests that punitive damages should be more readily
available to plaintiffs in breach-of-contract actions, even though common law jurisdictions reject this remedy
in claims for breach of contract.
34
All in all, Robust Promise Theorists argue that these doctrines are
objectionable from the perspective of ordinary promissory morality. Indeed, we probably should reform

30
See Richard Craswell, Contract Law, Default Rules, and the Philosophy of Promising, 88 MICH. L. REV. 489, 489-490 (1989); see
also LOUIS KAPLOW & STEVEN SHAVELL, FAIRNESS VERSUS WELFARE 162-63 (2006) (arguing that the promise-based
theories of contract “does not necessarily have the aforementioned—or, indeed, any—implications for the rules of
contract law because, in an important sense, the theory is incomplete”).
31
Dori Kimel argues, for example, that promising and contracting play similar instrumental values associated with
promising by “facilitating co-operation and mutual reliance,” while at the same time serving a very different value of
detachment, which is valuable because it gives us freedom from personal or intimate relationships that are often in the
background of, or fostered by, the informal practice of promising. Regardless of whether Kimel is correct, a general
point remains, which is that contract law, qua public institution, might legitimately serve public goals or values in
choosing whether and how to enforce certain promises. As Kimel puts the point, if the question is whether contract law
enforces promissory obligations, a very plausible answer might be “yes, but…” DORI KIMEL, FROM PROMISE TO
CONTRACT: TOWARDS A LIBERAL THEORY OF CONTRACT 65, 80 (2003).
32
FRIED, CONTRACTS AS PROMISE, supra note 11, at 1 (“The promise principle, which in this book I argue is the moral
basis of contract law, is that principle by which persons may impose on themselves obligations where none existed
before.”); id at 6 (stating that the goal is to show the principles relating to promises “generate[] the structure and
accounts for the complexities of contract doctrine”); id. at 17 (“The moralist of duty thus posits a general obligation to
keep promises, of which the obligation of contract will be only a special case—that special case in which certain
promises have attained legal as well as moral force.”).
33
FRIED, CONTRACTS AS PROMISE, supra note 11, at 28.
34
Shiffrin, Divergence of Contract, supra note 6 at 752-53.
37

contract law to make it less objectionable in the aforementioned ways. This reflects Promise Theorist’s
commitment to (3).
4 Promise Theory: Contractual Remedies Debates
For the purpose of later discussion, let’s take a closer look at the question of contractual remedies for
breach-of-contract claims. The question arises in cases like the following. Suppose that A and B have a
contract such that A promises to pay B money after B paints A’s house. Now suppose that B never paints the
house. A can sue B for breach of contract. The question of remedies is the question of what remedy A should
receive if she prevails, if not already specified by A and B in their prior agreement.
I will focus on the following menu of remedial options, which are most frequently discussed:
• If a court awards A specific performance, this means that the court will mandate that B
perform pursuant to the terms of the contract to the extent possible, which would entail
mandating that B actually paint A’s house. The goal is to give B the benefit of B’s bargain by
forcing A to paint the house.

• Expectation damages are monetary. Like specific performance, the goal is to give the non-
breaching party the benefit of her bargain with the breaching party, but through financial
compensation rather than actual performance. Here, the benefit of the bargain would be a
painted house for the agreed-to price. So B might not be entitled to any recovery if B manages to
hire a replacement painter for less than or equal to the amount that A was charging. After all, B
had not yet paid A any money, since they agreed that payment would follow performance by A.
But if B’s replacement painter C charges more, giving B the benefit of B’s bargain would allow B
to sue A for the extra amount that C charged over and above B’s agreed-to price. This would
guarantee that A would obtain the benefit a painted house, at price charged by B.

• Reliance damages are measured by reference to the losses that A suffered in relying on B’s
promise. The goal is to undo the damage done to A by B’s failure to keep his promise by placing
A in the position A would have occupied but for contracting with B. Suppose that A foreseeably
made vacation plans in reliance on B’s promise, but had to forgo those plans on account of the
breach. If reliance damages are awarded, B will have to pay A an amount to offset the losses
associated with the foregone vacation, assuming that A would not have made those plans but for
entering into a contract with B.

• Punitive damages aim, like fines, to sanction the wrongdoer. So if B’s breach of contract
warrants condemnation, or exhibits behavior that ought to be specially deterred, punitive
damages seem appropriate.

Now consider two legal rules that currently govern contract law’s remedies. First, contract law’s
default remedy is expectation damages. The term “default” as used in contract theory means that courts will
award expectation damages by default if the parties had not agreed in advance what the remedy would be in
38

cases of breach. The term also implies that this is not the exclusive remedy available to plaintiffs. Sometimes
reliance damages or specific performance are awarded instead of expectation damages. But these are the exceptions.
The second rule governs punitive damages. Punitive damages, as mentioned in the Introduction, are not
available in common law jurisdictions.
35

Both of these remedial rules are fixtures in common law jurisdictions. But Robust Promise Theorists
have criticized both as inadequately respecting the practice of promising. Consider the expectation damages
default. Many commentators, both sympathetic to Promise Theory and not, have observed that if the right
that a promisee obtains from a promisor is that the promisor actually perform a given promise, the most fitting
remedy for a violation of this right would be specific performance, not expectation damages.
36
Expectation
damages arguably suggest that every contractual obligation is an obligation to either perform or pay damages,
even though promises are not ordinarily understood to be disjunctive in that way.
37

The strict rule against punitive damages also arguably fails to reflect promissory morality. In general,
according to our ordinary, informal practices of promising, a broken promise is a moral wrongdoing eligible
for blaming attitudes like condemnation. This is not to say that it is always appropriate to morally blame a
promise-breaker—sometimes there are good, all-things-considered reasons to break a promise. But

35
For a discussion of possible recent exceptions in Canada and the UK, see Solène Rowan, Reflections on the Introduction of
Punitive Damages for Breach of Contract, 30 OXFORD J. OF L. STUDIES 495, 498-99 (2010) (citing English and Canadian cases
suggesting limited recognition of punitive relief for certain extreme breach-of-contract cases).
36
Charles Fried originally argued that expectation damages were the “natural” remedy for breached contracts. FRIED,
CONTRACT AS PROMISE, supra note 13, at 18. He was roundly criticized for this suggestion, and ultimately conceded the
point. Compare Shiffrin, Divergence of Contract, supra note 6, at 722-23 (claiming that the law’s failure to routinely demand
specific performance represents a “divergence” between contract law and promissory norms); DORI KIMEL, FROM
PROMISE TO CONTRACT: TOWARDS A LIBERAL THEORY OF CONTRACT 95 (2005) (criticizing Fried’s “neglect of specific
performance”); KAPLOW & SHAVELL, supra note 30, at 160 & 161 n.18 (pointing out that although Promise Theorists of
contract should “have a preference for a remedy of specific performance—requiring the contract to be performed—for
such a remedy amounts to a requirement that the promisor keep his word,” Fried “does not even mention specific
performance as an alternative to expectation damages”); Craswell, supra note 30, at 518 (arguing that Fried “cannot be
relying on existing nonlegal practices [to justify or explain the expectation measure of damages], for studies of those
practices show that people often do not demand (or offer) expectation damages in cases of unexcused
nonperformance”), with Charles Fried, Contract as Promise Thirty Years On, 45 SUFFOLK. U. LAW REV. 961, 968 (2012)
(“Both the moral criticism and the economic defense of expectation damages persuade me that I had overstated the case
for the connection between the promise principle and expectation damages[.]”).
37
Some have argued that, if we understand contract law as enforcing disjunctive promises (promises to either perform or
pay damages), then the alleged difficulty raised by Promise Theorists for expectation damages goes away. See, e.g., Daniel
Markovits & Alan Schwartz, The Myth of Efficient Breach: New Defenses of the Expectation Interest, 97 VA. L. REV. 1939 (2011).
But for the sake of argument I will assume that Promise Theorists are correct.
39

sometimes, and maybe even frequently, moral blame seems appropriate for failing to live up to one’s
promissory obligations. And if contractual obligations are just a type of promissory obligation, then moral
blame seems appropriate for breaching contracts as well. Ordinarily, the remedial vehicle for expressing
condemnation in civil litigation is punitive damages. But this vehicle is ruled out when the case involves
breach of contract. From the perspective of promissory morality, this is a deficiency. As Shiffrin puts the
point, contract law’s rule against punitive damages “fails to use its distinctive powers and modes of
expression to mark the judgment that breach is impermissible as opposed to merely subject to a price.”
38

Robust Promise Theorists thus view the strict rule against punitive damages as suspect and ripe for reform.
There are at least six ways to interpret the kind of criticism put forth by Robust Promise Theorists.
The first I mention only to set aside: Robust Promise Theorists might be arguing that the state, in purporting
to enforce promissory obligations, entirely fails to do so given the unavailability or curtailment of certain
remedies. This critique would be forceful if the only remedy for a rights violation of contractual rights were
extremely weak—say, limited to nominal damages. Because contract remedies and expectation damages in
particular are not weak, however, we should not interpret the complaints of Robust Promise Theorists this
way. Expectations damages are not weak.
Five other interpretations seem more plausible:
• Defectiveness Claim. Even though the law’s remedies vindicate promissory rights to some
extent, they render contractual right a defective form of promissory right. A non-defective
contractual obligation would have legally recognized analogues for all of the appropriate remedial
options available in the informal practice of promising. For example, if punitive reactions are
warranted and available in the informal practice, then they should likewise be warranted and
available in the form of its legal analogue, punitive damages. The unavailability of punitive
damages makes contractual rights and obligations defective qua promissory rights and
obligations.

• Disrespect Claim. Contract law disrespects promisees as holders of promissory rights. For
example, promisees participating in the informal practice of promising are entitled, by default, to
performance rather than performance or payoffs. But current contract law affords us, by default, to the
latter disjunctive entitlement—even in many cases in which the promisee clearly preferred the
former. This disjunctive right and obligation suggests that one might abandon, for whatever
reason, one’s prior promise to perform provided that one has enough money to buy her way out
of performance. This kind of reasoning collides with the second constitutive rule of promising,

38
Shiffrin, The Divergence of Contract, supra note 13, at 724.
40

which sharply limits what counts as an excuse or justification for abandoning our promissory
commitments. So the law’s default remedy, arguably, inadequately respects of the promisee’s
status qua promisee.
39

• Undermining Claim. Contract law undermines the informal practice of promising.
40
The norms
of contract law might cause citizens to take the promissory obligations less seriously. Perhaps the
ban on punitive damages, for example, conveys that there’s nothing inherently morally wrong in
not keeping one’s voluntary commitments. In turn, this suggests promissory morality warrants
less respect and accordingly undermines the valuable moral culture of promising.

• Conflict Claim. Making a contract necessarily involves making a promise. We are thus, through
a single act, liable to the expectations of promissory morality and legal liability. But the legal
regime is in some sense less demanding than the corresponding moral practice of promising
would imply. For example, contract law allows us, in effect, to purchase our freedom from our
promissory commitments if we have enough money to satisfy expectation damages. This makes
it harder for the morally decent person to do the right thing in accordance with moral norms,
given the incentives not to.
41

• Purposive Claim. Promising is an informal practice capable of being understood and followed
pursuant to norms that exist independently of law. Accordingly, since contract law’s point or
purpose is to enforce certain promises, or at least reinforce the informal practice,
42
the law must
respect our ordinary understanding of this informal phenomenon. Enforcing certain promises is
the point or aim of contract law.

Some contract theorists offer in effect global responses to these claims, arguing that contractual
obligations are not promissory obligations after all.
43
Another view argues that contracting does involve
promising but implicates importantly different values from those in play when engaged in the informal
practice of promising.
44
Accordingly, we should avoid taking promise-based criticism of extant contract law
too seriously.

39
Seana Shiffrin takes this worry a step further, adding that contract law’s doctrines fail to support or accommodate
moral agency, by undermining the conditions for morally virtuous agency. Shiffrin argues that individuals have a basic
interest in being morally virtuous. But contract law doctrines either directly or indirectly threaten to “make it harder for a
morally decent person to behave decently.” So, absent strong reasons rooted in the administration of the law, if a law
invokes a concept like promise, it must respect moral agency by accommodating the reasonable (normative) expectations
of virtuous moral agents regarding the ordinary practice of promising. Shiffrin, The Divergence of Contract, supra note 13, at
710.
40
Insofar as having a flourishing informal practice of promising is conducive to virtuous moral agency, a version of this
claim can be found in Shiffrin, The Divergence of Contract, supra note 13, at 752-53.
41
Shiffrin appears to make this point. See citation supra note 39.
42
See, e.g., Joseph Raz, Promises in Morality and Law, 95 HARV. L. REV. 916, 933 (1982).
43
See, e.g., Randy E. Barnett, Contract is Not Promise; Contract Is Consent, 45 SUFFOLK L. REV. 647 (2012); Michael G. Pratt,
Contract: Not Promise, 35 FLA. ST. U. L. REV. 801 (2007-2008).
44
See generally Aditi Bagchi, Separating Contract and Promise, 38 FLA. ST. L. REV. 709 (2011) (arguing that contractual
obligations differ from promissory ones by replacing the reasons ordinarily motivating promise keeping with the threat
41

My response differs. I offer no global denial that contracts involve promising. Nor do I emphasize
allegedly different values realized or promoted by the two practices.
45
Finally, I do not assume that a
particular conception of promissory morality is correct, full stop. I argue instead that if Practice Theory is
presupposed this renders Robust Promise Theory’s criticisms of contract law vulnerable. In particular, the
above-listed critiques associated with Promise Theory assume (3), which in turn presupposes that our
informal, ordinary practice of promising, and the expectations shaped by that practice, are somehow
privileged or should take normative priority over contract law. In the next section, I will argue that
presupposing a Practice Theory of promising makes justifying this priority problematic, even assuming that
contracting involves promising.
If I am correct, this problem generalizes and extends beyond Robust Promise Theory. This is
because non-Robust Promise Theory is also committed to the normative priority claim. That is, if one holds a view
that promissory morality provides the moral foundations of contract law, you will also be committed to the
normative priority of promise to contract. Focusing on the Robust Promise Theorists just brings out the issue
starkly given their penchant for actually recommending certain concrete and fundamental reforms that focus
the investigation. So from here on out I will speak strictly in terms of “Promise Theory” and “Promise
Theorists” rather than specifying that I am speaking of the “Robust” variety in particular.
5 A Problem for Combining Promise Theory and Practice Theory
With this background in place, let us turn to the main argument. To begin, recall the claim
presupposed by (3): that the informal, ordinary practice of promising takes normative priority over the
doctrines of contract law, such that the latter ought to strive to accommodate the former. This claim
underwrites various criticisms of current contract law put forth by Promise Theorists, regardless of whether

of legal sanction); KIMEL, supra note 36 (arguing that contract’s intrinsic value lies in the value of personal detachment,
while promising instantiates the value of an attached relationship). Nothing I say contradicts these claims, though I
emphasize that my argument goes through even if we assume that the informal and formal practices pursue or manifest
the same values.
45
See especially Aditi Bagchi, Separating Contract and Promise, 38 FLA. ST. L. REV. 709 (2011); DORI KIMEL, FROM PROMISE
TO CONTRACT: TOWARDS A LIBERAL THEORY OF CONTRACT (2003).
42

the criticisms are understood as Defectiveness Claims, Disrespect Claims, Undermining Claims, Conflict
Claims, Purposive Claims—or some combination thereof.
But holding a Practice Theory of promises makes explaining or justifying this priority difficult. To see
why, notice that one option is ruled out. Suppose that the rule that we ought to keep our promises, when
properly understood, simply reflects some basic moral principles—such as the principle that we ought to do
for others what they expect us to do when we have intentionally caused them to expect our performance.
46

These principles bind us regardless of whether actually existing social rules embody them. And to the extent
that contract law purported to enforce these moral principles but failed to do so, or did inadequately, or
provided incentives to avoid complying with them, then contract law would be obviously problematic.
47

Indeed, we should criticize contract law on these grounds, insofar as the law Undermines our commitments to
basic moral principles, or Disrespects us as holders of special moral rights that we hold simply in virtue of the
truth of certain moral principles.
But, again, this option is ruled out for Practice Theorists, since the central rule that one ought to keep
one’s promises, and any other constitutive or ancillary rule, are social rules—not moral principles. This hampers
the priority claim. After all, contract law is a social practice constituted by social rules, too. Why should
contract law aspire to reflect the informal practice, rather than vice versa? The priority must flow from
somewhere besides the social rules themselves.
One natural place to look is in Practice Theorists’ explanations for how the social practice enables us
to generate binding moral commitments. We saw one view associated with Hume, which holds that we have a
general obligation not to undermine valuable social institutions and practices, and thus we must comply with
the social rule that requires keeping promises to avoid undermining the valuable social practice of promising.
Rawls grounds promissory duties in a general obligation to avoid free riding on just and useful social
institutions.

46
For a similar view, see Chapter 7 of T.M. SCANLON, WHAT WE OWE TO EACH OTHER 304 (1998).
47
Scanlon actually holds that contracts are importantly different from promises, but I use his theory of promising as an
illustration here. See T.M. Scanlon, Promises and Contracts, THE THEORY OF CONTRACT LAW: NEW ESSAYS (Peter Benson,
ed., 2001).
43

But neither Hume nor Rawls—nor any other Practice Theory that uses the same general strategy to
explain why promises morally bind us—can easily explain the normative priority. Each Practice Theory that
tries to explain the duty to comply with the requirements of the practice presupposes some already-existing and
valuable social practice of promising. But contract law, in its current form, is also a valuable social practice
constituted by social rules. Contract law is similar to but also differs from the informal practice of promising
in many ways. So assuming that modern contract law, as it stands, is generally valuable, the same general
moral considerations favoring abiding by the rules constituting the informal practice of promising—such as
they are—likewise favor following the rules constituting the formal practice of contracting—such as they are.
That is, these general principles explain just as well why we should comply with the terms of current,
unreformed contract law. This hardly suffices to explain priority. So the moral principles that Practice Theories
rely on at the second stage of their theories do not suffice to explain the normative priority of promising.
48

Perhaps we are looking in the wrong place. Practice Theorists also invoke certain values at the first
stage of their theories. More specifically, the practice’s value derives from its constituting a vehicle that
promotes certain values and/or empowers individuals to realize them. So perhaps these values, and the relation
that the practice of promising bears to them, explain why the informal practice of promising takes priority
over the formal practice of contract law in cases where the latter inadequately reflects the former.
This response seems weak. To see why, consider in turn each of the two primary relations between
values and the practice of promising. First, the practice is instrumentally valuable. But notice what
instrumentalism implies. If the practice of promising P1 is essentially a means for promoting some value like
autonomy, it remains an open question whether legally enforcing a “defective” version of P1, say, P2 serves
those values better than legally enforcing P1. And if this question is open, then we cannot assume that we
should have reason to reform contract law on the grounds that it “inadequately” reflects the informal practice

48
At this point the Practice Theorist might try to argue that, unlike the informal practice of promising, contract law is
unjust or not valuable by comparison. To substantiate the alleged injustice, Practice Theorists would need to look
beyond the respective practices of promising and contracting to locate standards of justice suitable for criticizing the
latter but not the former. Once located, these standards that would ground the calls for reforming contract law, not the
mere fact that contract law fails to reflect promising adequately. I address the idea that promising is simply more valuable
than contract, in some sense, when I address possible objections later on.
44

of promising. We cannot, in turn, assume that (3) is true. After all, it might be the case that contract law better
promotes values like autonomy precisely because it strategically departs from, and thus fails to adequately reflect,
the informal practice of promising. This makes it more difficult to sustain criticism of contract law on the
basis of its alleged instrumental inferiority to a “more perfect” enforcement of the informal practice.
But maybe the practice of promising enables individuals to realize certain values in ways that current
contract law cannot. The relation between the practice and values is not just that the former promotes the
latter; rather, it is something like an instantiation relation. The practice makes possible certain ways of
realizing values that would be more difficult without it. Few people doubt that promising greases the wheels
for cooperation, enabling us to enlist in others’ projects and enlist others in our projects, in ways that tend to
be mutually beneficial. Charles Fried, as we saw, sees these commonplaces about promising as reflecting the
underlying autonomy value that promising enables.
49
Promising is an important way of exercising autonomy.
But once again it seems unlikely that, simply by locating the value of promising in another relation
with respect to the practice, we will thereby establish the normative priority of promising. First, a variant on
the prior argument still goes through: contract law, by systematically under-enforcing or otherwise
“inadequately” enforcing the informal practice of promising might better instantiate the value of, say,
autonomy, than the alternative of enforcing more faithfully that informal practice. This is an open question.
Second, it is not obvious why informal practices of promising instantiate a given value, say, autonomy, better
than the formal practice of contract. Perhaps the value is instantiated in both practices in a way that is morally
on par. Finally, the possibility remains that contract law is designed so as to instantiate some values that are
ordinarily irrelevant in informal promissory context. Contract law, in common law jurisdictions, is a mode of
voluntary commitment that is constituted in part by the judgment that breaches of commercial arrangements
typically ought not warrant punitive reactions.
50
Punitive attitudes might be perfectly appropriate, by contrast,
in the informal promissory context.

49
See FRIED, CONTRACT AS PROMISE, supra note 13, at 13-14.
50
See, e.g., Applied Equipment Corp. v. Litton Saudi Arabia Ltd., 869 P.2d 454, 515 (Cal. 1994) (“This limitation on available
[consequential] damages [like punitive damages] serves to encourage contractual relations and commercial activity by
enabling parties to estimate in advance the financial risks of their enterprise.”).
45

But perhaps we are once again looking in the wrong place. The Purposive Claim might show how to
justify the priority on purposive or functionalist grounds. Perhaps, as the law sometimes claims, the essential
purpose or function of contract law is to enforce certain promises. This explains the priority without needing
to rely on either the general principles Practice Theorists invoke (at stage two of their accounts), or the
general values that the practice instantiates or promotes (at stage one). But here again Promise Theorists are
at a particular disadvantage if they assume a Practice Theory of promising. If general principles that explain
promissory obligation likewise explain why we should follow contract law such as it is, and if the values that
the informal practice instantiates or promotes are realized equally well or better by contract law such as it is,
then we lose any sense of why we should accept the Purposive Claim—that contract law essentially serves the
function of enforcing some promises. Maybe so, maybe not. Or, even if contract law does enforce some
promises, we lose any sense of why it must do so in the ways that more fully reflects the understandings of
those engaged in the informal practice—at least the values of the informal practice, and the moral norms
making the binding, do not explain the priority. Once again, Practice Theory renders the promise-based
arguments for reform an uphill battle for Promise Theorists.
6 The Status Quo (Plausibly) Promotes Autonomy Better: Reliance and Specific Performance
In this section and the next, I argue that current Anglo-American contract law remedies might better
serve values that Practice Theorists attribute to the informal morality of promising, even if by doing so in a
way that somehow fails to adequately respect that very same informal practice. The point will illustrate how,
once a Promise Theorist presupposes a Practice Theory of promising, the normative grounds on which the
Promise Theorist argues for stricter enforcement of promissory rights may well support an argument for laxer
enforcement instead. To illustrate, I will assume the Practice Theory endorsed by Charles Fried, which holds
that the main value promoted and instantiated by the practice of promising is autonomy.
51
Fried,
unfortunately, does not make explicit the conception of autonomy that he relies on. So we are left to piece
together his thoughts from his comments throughout Contract as Promise.

51
See Fried, supra note 13, at 7, 13, 57.
46

The conception of autonomy that Fried appears to have in mind relates to liberal political morality:
We realize the value of autonomy in our own lives when we can reasonably dictate the direction that our lives
take.
52
One of the conditions for realizing this value is having meaningful options from which to freely
choose.
53
A well-established practice of promising is valuable, on this view, both because it multiplies our
options over time and because the practice empowers us to enlist others in our projects and thereby extend
their reach.
54
Respecting others requires respecting their free choices without unreasonably interfering with
them—including by not frustrating the projects of others by failing to live up to one’s promises.
55
This is the
picture of autonomy that Fried appears to have in mind.
Before exploring how contract law’s remedies might actually promote or instantiate autonomy better,
we should address a threshold objection by Richard Craswell, which simply denies that autonomy has
anything useful to say about remedies in contract law.
56
Craswell argues that
[a]ny damage measure is consistent with the ideal of individual autonomy, as long as it is adopted
solely as a default rule, since any default rule expands the promisor’s options by making it easier for
her to make a certain kind of promise. Fried must therefore invoke some other value in order to
decide which of the many damage rules to select as a starting point.
57

52
See Fried, supra note 13, at 7 (“[W]hatever we accomplish and however that accomplishment is judged, morality
requires that we respect the person and property of others, leaving them to free to make their lives as we are free to
make ours. This is the liberal ideal.”).
53
Free or autonomous choice requires that one’s choice among these options be, among other things, free from
coercion, duress, manipulation, and exploitation. For a discussion of duress and unconscionability doctrine, see, e.g.,
Fried, supra note 13, at 93-102.
54
See Fried, supra note 13, at 13 (claiming that promising “increase[s] one’s options in the long run”).
55
Fried, supra note 13, at 57 (“The moral force behind contract as promise is autonomy: the parties are bound to their
contract because they have chosen to be.”).
56
See Craswell, supra note 36, at 515 (“Just as any default rule would be consistent with the obligation to tell the truth,
any default rule would also be consistent with individual freedom, as long as the parties are allowed to change the rule by
appropriate language.”).
57
See Craswell, supra note 36, at 519.
47

Craswell argues, in other words, that so long as a remedy or other rule of contract law is simply a
default rule that parties have the freedom to alter via the terms of their contract, any default rule that is chosen
is equally compatible with autonomy.
There is a sense in which Craswell’s argument is correct, and a sense in which it is misleading.
Craswell runs together what we’ve called the constitutive value of the practice of promising and its
instrumental value. If we view Craswell’s claim as one concerning the instantiation relation, it seems plausible.
That is, if we treat the practice of promising as valuable primarily because it creates a vehicle that “expands
the promisor’s options,” and more generally, because it empowers individuals to realize, manifest, or exercise
their autonomy by respecting their choices, then Craswell’s point is plausible. After all, if we focus on
contracts formed between equals with roughly equal bargaining power (an important qualification), the parties
may freely choose under the terms of their agreement to abandon expectation damages and choose another
form of relief (within certain other legal limits). This is what it means for expectation damages to be a default
remedy. But if this is correct, then the default rule of expectation damages embodied within contract law,
even if a “defective” rule from the perspective of informal promissory morality, is equally compatible with the
constitutive value of promising—i.e., that it respects and enables autonomous choice in ways made much
more difficult without that practice. Contract law also respects and enables autonomous choice, and plausibly,
performs no worse than the ordinary practice in this respect. So there is no priority of the informal practice
over the legal one in instantiating the value of autonomy.
58

But Craswell’s argument is also misleading if he intends to suggest broadly that autonomy is wholly
silent with respect to the choice of default remedy—such that we need to “invoke some other value” in
choosing a default rule. That is, even if any default remedy is compatible with autonomous choice, it hardly
follows that every default rule’s “starting point” promotes autonomy equally well. Here Craswell overshoots the

58
Dori Kimel objects that some contract remedies might protect the practice of promising or individuals’ contractual
interests better than others.

KIMEL, supra note 36, at 93-94. And presumably some remedies do so in a way that preserves
the “natural continuity” between right and remedy. Id. at 94. While I accept the basic point that some remedies are better
than others in protecting interests and practices, this seems to miss Craswell’s point, which is that this alleged “natural
continuity” is not determined by the value of autonomy alone. Craswell could point out that this alleged continuity
between rights and remedies is the province of, say, corrective justice. This simply bolsters Craswell’s point, since
autonomy alone does not identify the remedy without the aid of corrective justice norms.
48

mark. A default rule R1 according to which you are presumed an organ donor unless you actively opt out
(and can easily do so), is equally compatible with autonomy as R2, according to which you are presumed not
to be an organ donor unless you actively opt in (and can easily do so). But countries adopting R1 rather than
R2 tend to have significantly higher organ donation rates, and moreover, we have good evidence suggesting
that the choice of default rule explains the disparity.
59
Consequently, many more people needing organs get
them in countries adopting R1. In turn, a higher rate of people in need of organs goes on to pursue their
important life projects unencumbered by severe medical problems. So even though two default rules of organ
donation may be equally consistent with autonomous choice (assuming roughly equal bargaining power), default
rules may not all promote overall autonomy equally well. Accordingly, the threshold objection fails to establish
that any default rule promotes a given value equally well, even assuming it establishes that any default remedy is
equally compatible with autonomous choice. A value like autonomy still might help in selecting between
alternative remedial rules.
In what follows, I continue to assume for the sake of argument that having a remedy of expectation
damages as of right, and disallowing punitive damages, reflects a failure of contract law to adequately respect, in
some sense, the informal practice of promising. I will nevertheless argue that, plausibly, Anglo-American
contract law’s remedial rules tend to promote autonomy better than the alternative of having some a default
rule requiring specific performance to the extent possible, and having a rule permitting punitive damages
routinely—both of which are rules floated by contemporary Promise Theorists.
60
The rest of this section will
discuss specific performance; the next takes up punitive damages. And if I am correct, then this further
undermines (3) given commitment (1) and (2), since we would no longer have a reason flowing from the
values that the informal practice secures to privilege the informal practice in designing contract law.

59
See RICHARD H. THALER & CASS R. SUNSTEIN, NUDGE (2008).
60
Shiffrin, Divergence of Contract, supra note 6, at 752-53 (concluding that “there may be reason to reexamine these
[contract] doctrines and their justifications, and to strive for greater convergence between promise and contract,”
including the doctrine providing for “the general unavailability of punitive damages in contract”); Again, Charles Fried
has also expressed sympathy for relaxing the rule against punitive damages. See Charles Fried, Convergence of Contract and
Promise, 120 HARV. L. REV. F. 1, 7 (2007).
49

Consider how expectation damages, as compared to the alternatives of reliance and specific
performance, improve our ability to make financial plans and predict our debts and expected income.
Predictability, of course, is not everything: a rule banning all contract-based lawsuits would be highly
predictable but not very helpful. But predictability, especially financial predictability, matters in planning our
lives. It matters for autonomy. Expectation damages in particular allow us to estimate, fairly precisely, what
the upper-bounds price of our contractual liabilities will be. After all, expectation damages involve giving a
person the benefit of her bargain, and this will typically not exceed the value embodied in the underlying
contract.
61
Now, expectation damages are not perfectly clear to all parties all the time. Sometimes the award
turns on market prices, which fluctuate. B’s failure to paint A’s house forced B to compensate A in the
amount representing the difference between market price and B’s original price quote (assuming the last-
minute price is higher). But it is nevertheless widely understood that expectation damages provide some
relatively clear sense of the monetary value of one’s contractual obligations. B, in determining whether he can
afford to not perform, is in principle well situated to determine the market price of hiring last-minute
painters, and figuring this into his decision not perform.
Contrast expectation damages with reliance damages and specific performance. Awards of reliance
damages are notoriously unpredictable in advance of making a contract. Not only is the fact of loss-in-reliance
relatively unpredictable, the amount often is, too. The litany ways in which a party might potentially rely on a
contractual obligation is often unforeseen and hence difficult to take into account financially. Limiting
reliance damages to foreseeable damages helps. But foreseeability is itself a difficult determination as compared
to knowing in advance the extent of your financial liabilities as determined by the terms of the agreement. If

61
Dori Kimel has argued that although the practice of promising would naturally suggest a default rule of specific
performance, contracts typically govern detached commercial relationships where expectation damages are roughly
equally good from the point of view of the victim of breach. Id. at 101. As between two equally satisfactory default rules,
one which involves more government intrusion in our lives (specific performance) than the other (expectation damages),
Kimel argues that the harm principle militates in favor of expectation damages rather than specific performance. Id. at
103-09. I find much appealing with Kimel’s overall point, which I take to be consistent with my own. But the harm
principle plays no obvious role in the analysis; a direct appeal to the autonomy-promoting advantage of default rule of
expectation damages seems all that is necessary (as I will argue in a moment). Nor does my argument assume that
expectation damages and specific performance are equally satisfactory. I worry that this begs the question. And even if
this is true, denying the claim would still allow for the possibility, as I argue, that (plausibly) a default rule of expectation
damages rather than specific performance yields net positive autonomy gains.
50

greater uncertainty makes planning our important commitments more difficult, and if this in turn tends to
impede autonomous choice, reliance damages seem comparatively undesirable from the perspective of
autonomy. It helps all parties involved to know that, if you fail to paint my house in time pursuant to the
terms of our agreement, I can immediately hire someone else and make you pay for any difference in value
between the market price for last-minute painters and the agreed-to price. I do not need to worry whether my
reliance damages, whatever they might be, are enough to offset the costs of hiring a new painter at the last
minute. And you do not need to find the time and resources to perform a promise that you’ve already failed
perform (for whatever reason). As between reliance and expectation damages, it seems that expectation
damages performs better in terms of promoting autonomy by fostering financial predictability.
Specific performance is another matter. Predictability is not the main threat to autonomy. To see
why, let us begin with the idea, likely to be endorsed by Promise Theory, that courts ought to order specific
performance rather than expectations damages where possible. Recall the argument motivating this suggested
reform. Absent express agreement to the contrary, when we promise to do x, we ought to do x. When the
promisor fails to do x, then the best remedy will and should aim to give the promisee the promised benefit—
i.e., the defendant’s doing x to the extent still possible. And forcing the promisee to perform will do just this.
True, specific performance is not a perfect substitute. If you promise to paint my house on Monday and do
not show up, the court cannot rewind the clock to Monday. But the court can, under a specific performance
remedy, still make you paint the house. Those endorsing Promise Theory thus regard or should regard
specific performance as the next best thing from the perspective of giving a person the benefit of the bargain:
a promised performance, not merely damages. So specific performance, not expectation damages, ought to be
the default.
I grant for the sake of argument that informal promissory morality most naturally recommends
awarding specific performance when available. But one might doubt that a specific performance default
performs worse than expectation damages in securing autonomy. First, in cases involving specific
performance of contractual terms mandating the payment of money, specific performance differs only in
form as compared to court-ordered payments for damages. If the latter are comparatively unproblematic
51

from the perspective of autonomy presumably the former should be as well. Second, specific performance
does not look troublesome in cases where both parties would prefer payment of damages.
62
Suppose that the
court orders you to paint my house as promised. Assume further that neither of us wants you to paint my
house any more—i.e., both would prefer payment to performance. Presumably, we can negotiate a new
agreement pursuant to which you waive your right to enforcement in exchange for, say, the amount necessary
to hire someone else to paint my house. But this is tantamount to expectation damages. So really the default
remedy makes no difference.
But this analysis does not fully establish that specific performance is on par with expectation damages
in promoting autonomy. One problem mirrors Craswell’s oversight. Just because parties can in principle
negotiate around a specific-performance default does not necessarily show that the choice of default will not
have disparate negative effects on autonomy. Indeed, autonomy problems loom when parties do not share a
preference for specific performance. Suppose that, post-judgment, the plaintiff would prefer a payment to
specific performance, but that (for whatever reason) the breaching party now prefers to perform pursuant to
the court’s judgment. There is a sense in which the plaintiff gets the benefit of her bargain—the house gets
painted. So her plan still gets executed.
But unsurprisingly the plaintiff might wish to have no further dealings with the defendant. They had
to litigate, after all, and the plaintiff might simply prefer to deal with someone new without the history of
rancor and litigation potentially interfering with the business relationship going forward. Being forced into
continued dealings looks deficient from the perspective of autonomy, since the plaintiff loses control over her
choice of agent. The court, in effect, is mandating that the plaintiff work with someone she no longer wants
to. The promisee/plaintiff loses the option to forswear the commercial relationship that had been
problematic in the past.

62
Why would this occur in the first place? After all, presumably it is the plaintiff who pursues specific performance. So
why suppose that the plaintiff would abandon this pursuit? I am assuming, of course, that specific performance is a
“sticky” default, such that the court will award it (if available) even if the plaintiff would prefer money damages. This
thought is also motivated by the logic of Promise Theory: plaintiff should get neither more nor less than what they are
promised, which is a performance, not a money substitute.
52

True, the plaintiff may waive the right to performance, but then she is left with nothing—at least
assuming that specific performance is a “sticky” default as mentioned. (We will relax this assumption in a
moment.) So the choice for this set of cases is not an attractive one: to either have one’s plan executed by
someone one no longer wishes to be involved in that plan or being forced to incur additional expenses to find
someone new. Assuming both results are at odds with one’s ability to pursue one’s plans as one wishes, this
hardly seems a pro-autonomy result.
Now suppose the reverse: that the plaintiff prefers specific performance but the defendant does not.
Here too autonomy problems loom. The problem is familiar: specific performance, at least in the case of
service contracts, smacks of “involuntary personal servitude”.
63
As one commentator remarks, “we have a
strong prejudice against any kind of involuntary personal servitude,” to the extent that “[w]e insist upon
liberty even at the expense of broken promises.”
64
This represents the “conventional wisdom” on why
specific performance is typically disfavored in common law courts, especially when specific performance
would call for performance of services or labor.
65
The idea that the court may coercively order you against
your wishes to perform work that you do not wish to perform, especially when you are willing to buy out
your contract, threatens autonomy.
Contrast this situation with the default rule of expectation damages. If you promise me to do X, and
you do not perform, often my project will be such that I need someone else to do X if you are no longer willing
or able to do so. This will require my going out on the market to find someone else. A default rule of
expectation damages gives me the ability to pursue alternative plans of action without wondering whether I
will be able to recoup the cost. This would allow me to pursue my projects close to as originally planned. To
illustrate, recall the painting case. If my remedy, as of right, is to require you to paint my house, then I must in
effect wait for a court to demand that you execute pursuant to the terms of our agreement, even though you

63
ARTHUR LINTON CORBIN, CORBIN ON CONTRACTS § 1204, at 401 (2006).
64
Id.
65
Nathan Oman, Specific Performance and the Thirteenth Amendment, 93 MINN. L. REV. 2020, 2021 (2009). Oman ultimately
argues against the claim that specific performance of personal service contracts would violate the Thirteenth
Amendment’s constitutional ban on involuntary servitude. But even so the underlying concern for autonomy would
remain, and the rule disfavoring specific performance of service contracts as a prophylactic would have traction.
53

have already failed to do so. The far better rule from the perspective of autonomy would empower me to
execute an alternative plan by hiring another painter if you fail to live up to your end of the bargain. Such a
rule would allow me to immediately hire a replacement without wondering whether I’ll be able to recoup the
difference in cost (if any). Assuming that these sorts of circumstances occur routinely, and that expectation
damages better promote autonomy than specific performance in such cases, the expectation damages seem
preferable to specific performance from the perspective of autonomy.
One might object that the argument against specific performance targets a straw man: that specific
performance would be foisted upon a plaintiff who did not want to perform or take nothing at all. The far
better reform from the perspective of autonomy would be to allow the plaintiff to choose a remedy as
between specific performance and expectation damages once the court determines that the contract has been
breached. But there are two problems with this suggestion. One is internal to the logic of Promise Theory.
This disjunctive remedy looks to deliver more than the promisor bargained for. The promisor obtained a
promised performance, so if this is what is feasible, this is what should be delivered. The suggested reform,
however, gives the promisor the option of choosing between performance and damages—i.e., more than the
benefit of the bargain. So I am unsure of whether this remedy would be available on the grounds of Promise
Theory.
But even if this reform were consistent with Promise Theory, notice that allowing the plaintiff to
choose remedies does not eliminate the conventional-wisdom problem. That is, the reform would still allow
plaintiffs to enlist the state to force people to perform services they no longer wished to perform, raising the
same threat to autonomy. Now the Promise Theorist might further object that the defendant’s preference not
to perform does not warrant protection once he has broken his promise to perform. The defendant has
broken his promise; why does his autonomy warrant equal consideration with the plaintiff seeking specific
performance of that very same promise? But surely the promise breaker’s autonomy, post-judgment, is
relevant qua human being, even if not equally so. Nor are we assuming that the defendant to wholly escapes
his obligations at the expense of the promisor’s own. The defendant must pay damages, and by design,
expectation damages aim to give the plaintiffs the benefit of the bargain. And even if the “benefit” is
54

uncertain, we might reasonably worry about the risks posed by allowing individuals to enlist the state to force
others to do their bidding against their will, even if they have previously promised to perform. This
reasonable worry about autonomy warrants a prophylactic rule strongly disfavoring specific performance.
And recall the dialectic. I aim to show that the normative priority of promise over contract cannot be
explained once we assume that the practice of promising is merely a social practice like contract law. They are
practices and as such are prima facie morally on par. The alleged normative priority becomes even harder if we
can mount a plausible argument that the values justifying the informal practice actually recommend rejecting the
categorical preference for protecting expectations shaped by our informal understanding of promises in
shaping contract law’s remedies. Plausibility is what I need, not certainty. And that is what we have just
illustrated, at least with respect to Promise Theory’s preference for specific performance.
7 How the Status Quo Might Better Promote Autonomy: Punitive Damages
The other recommended reform that Promise Theory endorses involves relaxing the rule barring
punitive damages. Let me state the rule under discussion more carefully. The relevant legal norm holds that
merely establishing an intentional breach of contract—e.g., knowingly or willfully breaching a contract—will
not suffice to justify a demand for punitive damages. This is not the rule holding that egregious, fraudulent, or
abusive conduct in and around the contract’s formation or performance will not justify punitive damages.
Canada has adopted such a rule or has come close to it.
66
The idea floated by Promise Theorists is that any
opportunistic breach would potentially be subject to punitive damages, not just the most extreme cases. This
proposal makes sense if we assume that intentionally breaking a promise is a moral wrongdoing worthy, like
any other intentional wrongdoing, of moral condemnation.
It is tempting to argue that punitive damages are disfavored from the perspective of autonomy
because they are unpredictable. Indeed, some have cited concerns about “commercial uncertainty” to justify

66
See Solène Rowan, Reflections on the Introduction of Punitive Damages for Breach of Contract, 30 Oxford J. of L. Studies 498-
99(2010) (citing Canadian cases suggesting limited recognition of punitive relief for certain extreme breach-of-contract
cases).
55

excluding punitive damages.
67
As already noted, too much unpredictability undermines an important
condition of autonomy. Autonomy-based considerations would appear to bolster the common law rule
against punitive damages in breach-of-contract actions.
But this argument falls short. Nothing about punitive damages per se generates the problem of
predictability, given that we can reform the administration of punitive damages to make them more
predictable, both in terms whether and how much punitive damages would be available. One suggestion is to
impose a strong but defeasible presumption that breaches are not deliberate and hence not subject to punitive
damages.
68
And we might limit punitive damages to, say, 20% of the expectation damages awards, thereby
tethering punitive damages to the fairly determinate baseline fixed by expectation damages. So if punitive
damages threaten autonomy it is not due to unpredictability per se.
69

But autonomy worries loom elsewhere. The goals of punitive damages are to deter conduct, exact a
form of retributive justice, or express condemnation.
70
But perhaps contract law ought to avoid these
pursuits, lest it over-deter breaches of contract. Consider the following. Suppose that you agree to paint my
house for $X on Monday, but later on someone else (Jim) asks you to paint her house for $5X on Monday.
You cannot do both, though by accepting the later, conflicting contract, you will be able to compensate me
enough to hire some adequate substitute on the market—even if the substitute is pricier. Stated in terms of

67
See BRIAN H. BIX, CONTRACT LAW: RULES, THEORY, AND CONTEXT 99 (2012) (noting that California had
experimented with a permissive punitive damages rule, but later abandoning it because it was “too hard to apply and
creat[ed] too much commercial uncertainty”); see also Applied Equipment Corp. v. Litton Saudi Arabia Ltd., 869 P.2d 454, 515
(Cal. 1994) (“This limitation on available [consequential] damages [like punitive damages] serves to encourage
contractual relations and commercial activity by enabling parties to estimate in advance the financial risks of their
enterprise.”); Laurence P. Simpson, Punitive Damages for Breach of Contract, 20 OHIO ST. L. J. 284, 284 (1959) (“Basically
the reason is an unwillingness to extend the doctrine of punitive damages to the commercial field thus introducing
uncertainty and confusion into business trans- actions, and also the feeling that substituted performance in terms of
compensatory damages is an adequate remedy for the aggrieved party to a contract.”).
68
See STEPHEN A. SMITH, CONTRACT THEORY 418 (2004).
69
We might also doubt whether predictability matters: those who do not breach contracts willfully or maliciously need
not worry about predictability of punitive damages. See Rowan, supra note 5, at 500 (“Predictability would be unaffected
for upstanding businesses which aim to discharge their contractual obligations. Those in wilful default, in contrast, are
not deserving of certainty and should not be heard to complain.”).
70
See, e.g., Pac. Mut. Life Ins. Co. v. Haslip, 499 US 1, 19 (1991) (asserting that “under the law of most States, punitive
damages are imposed for purposes of retribution and deterrence.”); Whiten v. Pilot Ins. Co., [2002] 1 S.C.R. 595, 635
(“[T]here is a substantial consensus that coincides with Lord Pratt C.J.’s view in 1763 that the general objectives of
punitive damages are punishment (in the sense of retribution), deterrence of the wrongdoer and others, and
denunciation”).
56

efficiency, this is a Pareto improvement, since after the compensation happens no one is worse off, and at
least two people are better off—Jim and I.
71
But if I were forced to paint your house, either by specific
performance or because I would face public sanction expressed through punitive relief, at least one person
will be worse off in this situation: me. So the idea, roughly speaking, is that certain cases of non-performance
plus compensation are welfare enhancing. To the extent that specific performance and punitive damages
disrupt such efficient non-performances, this hampers welfare-enhancing transactions.
So far this sounds like the dubious notion of efficient breach. But we might approach the same idea
through the lens of autonomy, and can articulate the idea in terms that do not presuppose that the law
actively endorses breach or otherwise signs onto other efficient-breach heresies.
72
Assume that committing
ourselves to others and ensuring that others will commit themselves to our projects are important ways that
we exercise our autonomy. This by itself does not tell us what the limits of these commitments should be.
Plausibly, our commitments should be strong but limited to some extent. Recognizing a legal right of the
desperately poor to voluntarily contract themselves into slavery does not respect autonomy in the right way.
And an exercise of that “right” seems, at best, a highly defective exercise of autonomous choice under
conditions that call into question whether it was an autonomous choice at all.
Less dramatically, no one doubts that we should, in certain circumstances, be permitted to back out
of our voluntary commitments—promissory or otherwise. Contract law’s innovation is, on the
characterization on offer, to specify that doing so requires satisfying the financial expectations of the non-
breaching party. In principle, at least, the non-breaching party can use the funds to pursue her own projects
as originally intended (painting a house, etc.), while the breaching party will have more resources at her
disposal to advance her own projects. A rule that made punitive damages more readily available might end up
making contractual agreements sub-optimal from the perspective of our ability to change our plans in light of
changing circumstances. Contracts will be too confining. This, in turn, might be sub-optimal from the

71
Alternatively, we might view this as a Kaldor-Hicks efficient transaction in which I actually compensate the party
initially made worse off: you. I am, of course, assuming no transaction costs and that you located a perfect substitute.
72
Efficient breach doctrine has been roundly criticized as “empty” by economists and morally and legally perverse by
others. See, e.g., Daniel Markovits & Alan Schwartz, The Myth of Efficient Breach: New Defenses of the Expectation Interest, 97
VA. L. REV. 1939 (2011).
57

perspective of autonomy. Contract law’s compromise position is to hold that it is a wrong to breach one’s
contract without voluntarily paying enough to give the promisee the benefit of the bargain. In short, perhaps contract law
expresses the judgment that there is nothing intrinsically wrong with breaking a contract, even
opportunistically, provided that one compensates for the breach up front and is not otherwise abusive
towards the promisee.
This might initially sound incoherent. Consider a stock objection to efficient breach theory.
Breaching a contract is an actionable wrong—this is why damages are available. And if it is wrong then the
notion that we may “permissibly” back out of our agreements, provided that we pay a price for doing so, is
normatively incoherent. To say that something is wrong is to say that it is impermissible to do it. This is so
regardless of whether we try to justify breaking contracts in terms of efficiency or autonomy.
But this claim of incoherence misfires. To see why, consider recent work interpreting certain forms
of strict liability in tort law. In Vincent v. Lake Erie Transportation Company, a ship captain chose to fasten his
ship to a dock in a violent storm, without the dock owner’s permission (there was no time to secure it). But in
all likelihood this trespass saved himself, his ship, and his crew in the process—even though he damaged the
dock as a result.
73
The Vincent puzzle is to figure out why the shipping company is liable to pay if the captain
behaved as any reasonably prudent person would. That is, if the captain was justified in appropriating the dock
to save lives then on what grounds is he liable in tort—i.e., the law of wrongs. Arguably, part of the answer
involves the property tort involved: one may be strictly liable for torts like trespass even if perfectly
reasonable. Due respect for others’ property rights requires compensation for even justified appropriation of
that property.
74
Sometimes—such as in Vincent’s case involving private necessity—we may very well be all-
things-considered justified in appropriating someone else’s property for our own purposes. But this does not
obviate the need to pay damages to show respect for those whose property we have used.

73
Vincent v. Lake Erie Trans. Co., 124 N.W. 221 (Minn. 1910).
74
Greg Keating has recently argued that some strict liability claims do involve a wrong—namely, failing to voluntarily pay
for one’s reasonable infliction of harm. See Gregory C. Keating, Strict Liability Wrongs, in PHILOSOPHICAL FOUNDATIONS
OF THE LAW OF TORTS 292, 296 (2014) (“These liabilities [such as the one imposed in Vincent do not condemn the
infliction of harm, per se. They condemn harming without repairing.”).
58

The possibility of strict liability “wrongs” in tort law makes intelligible the idea that breaching a
contract may be justified (if doing so can pay off other contracting parties to their satisfaction) yet still involve
an actionable wrong (if in breaching the promisor party fails to come forward voluntarily with the full
payment). Contract law does not actively encourage breach any more than Vincent actively encourages
trespasses. By the same token, if one party is well positioned to breach and as a result can pay off the
promisee such that the promisee is virtually no worse off, contract law does not necessarily want to express
the judgment that these transactions are morally reprehensible or fully unjustified. Indeed, these transactions
tend to promote autonomy rather than undermine it, since the breaching party is better able to pursue her
ends, and the non-breaching party is (given expectation damages) virtually no worse of towards pursuing his
own. But allowing punitive damages for all such “opportunistic” breaches forces the promisor to forego
opportunities, even if the promisee would be virtually no worse off after compensation.
75
The bar on punitive
damages, I submit, plays a role in striking the right balance between deterring breaches, protecting contractual
rights, yet not locking people into deals to the detriment of individual autonomy.
Someone might object that my argument proves too much. Perhaps expectation damages or the
absence of punitive damages are sub-optimal, as well, either awarding too much or too little from the
perspective of autonomy. But I do not claim to establish that expectation damages are wholly optimal.
Indeed, there’s good reason to believe it is not, especially given that the non-breaching party often must pay a
lawyer to secure the expectation damages to which she is originally entitled when the other party fails to
perform. The recovery will invariably be less than cases in which the breaching party voluntarily pays the
expectation damages. This suggests that expectation damages are too little. And maybe there’s an empirically

75
One might argue that the argument implausibly assumes that the promisee will receive “virtually” the benefit of the
bargain. First, if it is a problem it is a problem fundamental to the calculation of expectation damages. The problem is
not unique to my argument. It might also be argued that, if the promisee were truly indifferent between a payment of
expectation damages and performance, then the promisor and promisee should simply renegotiate the terms of the
contract. The promisor should be able to negotiate a release of the contractual obligation for the price that the promisee
sets. But the leverage this creates may defeat the purpose: the promisee might be able to obtain even more than the
benefit of the original bargain. Contractual rights on this view become as strong as property rights, subject to the
promisee’s discretion as to pricing. It is far from obvious that the promisee’s extreme leverage over the promisor leaves
the parties with more autonomy than the status quo alternative.
59

grounded argument suggesting that expectation damages or the lack of punitive damages actually award too
much by way of compensation from the perspective of autonomy.
I do not purport to settle the matter. The preceding arguments are speculative, and depend on
empirical facts. Nor have I tried to establish or prove that the default remedy of expectation damages and the
rule against punitive damages together promote autonomy better than the alternatives recommended by
Promise Theorists. The point is just to make plausible the idea that these status quo rules might do a better job in
promoting autonomy, by first emphasizing the predictability gains we obtain with the current common law
regime, as well as the possible autonomy-promoting virtues of the breach-and-pay model articulated above—
an approach which would be ruled out by alternative remedial regimes advocated by Promise Theorists, such
as allowing punitive relief for opportunistic breaches. To the extent that contract law’s current remedial
regime promotes some value (like autonomy) better than a reformed version of the law that somehow better
“respects” our ordinary practice of promising, we have no reason to adopt the suggested reforms if we accept
Practice Theory. After all, adopting the suggested reforms would be pointless without showing some value-
based gains; it is not enough to point to the putative fact that the informal practice would be taken less
seriously, undermined, or inadequately respected. At least when we adopt Practice Theory.
Let’s take stock. I have argued that those who accept claim (3)—that we should reform contract law to
better reflect the putative fact that contractual obligations are promissory ones—face problems if they also
accept two further claims associated with Practice Theories of promising—namely, that (1) promissory
obligations depend on the existence of a widely followed social practice
76
that (2) constitute or promote certain
values, which in turn justify that practice. This was because (3) presupposed that the informal practice of
promising is somehow privileged over extant contract law. This privilege is difficult to justify when one
adopts a Practice Theory. The main problem, as Section 5 argued, is that it is contingent whether contract
law’s alleged inadequate reflection of informal promissory morality better realizes or promotes values than the
alternative of enforcing more perfectly ordinary promissory morality. Sections 6 and 7 press further, arguing

76
See citations collected supra note 11.
60

that, plausibly, Anglo-American contract law’s remedial regime does promote autonomy better than would a
comparable law that adopted Promise Theorists’ recommended reform of contract remedies.
8 Objection 1: A Two-Tiered Approach
Promise Theorists might object as follows. What justifies a practice might differ from what justifies
individual acts in accordance with that practice.
77
A game’s value might be entertainment. But that does not
mean that the only value at stake for actions covered by the game’s rules is the value of entertainment. If I
cheat in a game that my colleagues take seriously, this disrespects or manipulates them. And this might be so
regardless of whether they discover my cheating. My colleagues might be entertained despite my secretly
winning money at their expense—thus realizing the practice’s value while still wronging them.
This objection misses the point. I’m not denying that the values that justify a social practice might differ
from the reasons against contravening the rules of that practice. We should not cheat at poker even though
the values associated with poker (entertainment, perhaps) are independent of the reasons against cheating (we
should not manipulate others). My point is simply that once we acknowledge that the rules governing the
informal practice of promising are merely social rules justified by the values the practice is there to realize or
promote, we lack any reasons for privileging the informal practice over contract’s formal practice. The fact
that contracts are tantamount to a “defective” form of the informal phenomenon is, by itself, of no moral
import apart from the values justifying the informal practice, especially if the formal practice of contract law
better promotes those values. Even if contract law inadequately respects the informal practice (e.g., by not
doing enough to guard against certain opportunistic breaches), that “disrespect” ultimately might not be
problematic when viewed from the perspective of the values served by the informal practice in the first place.
9 Objection 2: Can Promising Constitute a Sui Generis Value?
Another response might target the assumption, implicit throughout my discussion, that the values at
stake in the informal practice of promising—autonomy, in particular—can be understood independently of
that practice. Maybe the relevant value is not independent. Perhaps the value promoted or realized by the
practice of promising is not simply, for example, autonomy; rather, perhaps it is autonomy-enabled-by-promising.

77
See, e.g., John Rawls, Two Concepts of Rules, 64 PHIL. REV. 3 (1955).
61

The informal practice is thus somehow intrinsically valuable—just like games can be regarded as intrinsically
valuable by those who play them.
It is tempting to reject this response as circular. But the circularity is not necessarily vicious. The
value of playing basketball might be described in terms of, say, exercising, or increasingly specific terms, such
as having fun while exercising. These explanations do not mention basketball itself. But a complete statement
of the value might be self-referential. Someone might value exercising and having fun while playing basketball.
To say that the value of playing basketball lies in part in playing basketball is just to say that basketball is
intrinsically valuable—valuable for its own sake. To point out other features of basketball that are valuable
does not diminish the importance of these other things. It just means that the most complete statement of the
reasons why one plays basketball mentions basketball itself.
Similarly, we might think of promising as an intrinsically valuable social practice, participation in
which should be valued for its own sake for those who participate in it. Maybe a complete statement of the
reasons why promising is valuable will mention—in addition to autonomy or whatever other value we like—the
practice of promising itself. Suppose that the informal practice of promising is somehow intrinsically valuable,
as understood in this way. And suppose that the Practice Theorist who accepts Promise Theory as well (like
Fried) also accepts that promise keeping is intrinsically valuable.
If so, the very same thing might be said about contract law. We can say that contract law, in addition
to realizing the values of autonomy and trust, also realizes the commercial virtue of contract keeping, as defined
by the rules of contract law, including the remedies available to victims of breaches. Contract law is as much a
practice as promising allegedly is. So why not suggest that there is a contract-law-dependent value realizable
only in conformity with that practice, which happens to enforce the rules in a particular way, including by
disallowing punitive remedies? What’s more, suppose further that contract law is better at facilitating autonomy
and trust than the ordinary practice of promising. Would there still be strong reasons to force contractual
norms to better respect promissory morality, insofar as contractual ones do a better job of promoting
autonomy and trust? I do not see how.
62

10 Conclusions
I have tried to argue that two prominent views of promising and contract, respectively, sit uneasily
with each other. The first is Practice Theory: that promissory obligations exist only if there is an actually
existing practice of promising constituted by social rules. The second view is that properly respecting
promissory morality requires that we reform current contract law by strengthening its remedial regime. This is
Promise Theory. I have argued that holding one view makes holding the other more difficult. If the rules
constituting the informal practice of promising are simply social rules, we have no prima facie reason to
prioritize them over the social rules constituting the formal practice of contract law. We must look beyond
the rule of promise keeping for the relevant priority. And if the main justification for the informal practice of
promising is that it promotes or enables us to realize certain values, then it is possible that contract law, to the
extent that it fails to fully “respect” informal promissory morality (as claimed by Promise Theory), might
actually do a better job (from the perspective of these same values) than more “adequately” enforcing the
informal practice. Hopefully, I’ve given some examples of how contract law might succeed in precisely this
way.
There are several important upshots. First, Charles Fried—the most influential Promise Theorist—
adopts a Practice Theory. This undermines the force of his promise-based critiques of the common law of
law, or at least imposes a previously unappreciated burden on his calls for reform. Second, I hope that it is
clear that this problem is not faced uniquely by Fried’s account. Any future Promise Theory that assumes a
Practice Theory will face similar obstacles. But perhaps the most important lesson is the most general one. If
we want to criticize contract law for somehow failing to respect our rights as promisees, the particular
conception of promissory obligation that we presuppose matters. We cannot simply assume that all
conceptions share a common denominator that serves equally well in grounding our promise-based critiques
or reconstructions of contract law.
63

Chapter Two: Contracts and Agency Theories of Promising
1 Introduction
David Hume famously observed that
since every new promise imposes a new obligation of morality on the person who promises, and
since this new obligation arises from his will; ‘tis one of the most mysterious and incomprehensible
operations that can possibly be imagin’d, and may even be compar’d to transubstantiation, or holy orders,
where a certain form of words, along with a certain intention, changes entirely the nature of an
external object, and even of a human creature.
78

So promising involves conjuring new moral obligations by willing them into existence. Or as Joseph
Raz puts the point, promising involves creating new obligations by “communicat[ing] an intention to
undertake by that very act of communication an obligation to perform a certain action.”
79
But how can we
willingly impose a new moral obligation on ourselves by communicating an intention to do so? This is
Hume’s Challenge.
80

78
DAVID HUME, A TREATISE OF HUMAN NATURE 524 (1960) (L.A. Selby-Bigge, ed.) (emphasis in original).
79
Joseph Raz, Promises and Obligations, in LAW, MORALITY, AND SOCIETY: ESSAYS IN HONOR OF H.L.A.
HART 218 (P.M.S. Hacker and Joseph Raz, eds. 1977); see also JOHN FINNIS, NATURAL LAW AND NATURAL
RIGHTS 298 (1980) (“Centrally, then, a promise is constituted if and only if (i) A communicates to B his
intention to undertake, by that very act of communication (in conjunction with B’s acceptance of it), an
obligation to perform a certain action (or to see to it that certain actions are performed), and (ii) B accepts
this undertaking in the interests of himself, or of A or some third party, C.”).
80
What makes Hume’s Challenge challenging is, in part, a certain common-sense deontological picture of
morality. Morality is supposed to guide our conduct, at least in part, and is supposed to determine what we
may and may not do, regardless of how we’d like to act. To see this point stated in terms of reasons more
generally, see Gary Watson, Promises, Reasons, and Normative Powers, in REASONS FOR ACTION 158 (David Sobel
and Steven Wall, eds., 2009) (“To be real, reasons must constrain and guide the will (insofar as we are
rational); otherwise there is no normative reality to which the will must answer. If it were sufficient for having
a reason to x merely to decide that one has a reason to x, then having or not having that reason would be no
constraint at all.”); see also David Enoch, Giving Practical Reasons, 11 PHIL. IMPRINT 1, 1 (2011) (“If, as seems
likely, ‘reason must constrain and guide the will’, how is it that we can create reasons at will, for instance by
making a request?”). Furthermore, the possibility of imposing moral obligations on ourselves by promising
seems somehow at odds with this picture. What is to prevent us from making immoral promises? Wouldn’t
doing this provide a loophole to morality? These are important questions any theory of promissory obligation
must face. Like Watson, I am tempted by the view that the moral power to promise is constrained by other
moral requirements, such that I doubt that immoral promises are morally binding. See Gary Watson, Promises,
Reasons, and Normative Powers, in REASONS FOR ACTION 176 (David Sobel and Steven Wall, eds., 2009)
64

Moral Power Theories of promising give an answer. They hold that promising indeed involves
undertaking an obligation to do something for someone else by means of communicating an intention to do
so, and moreover, that there is nothing particularly mysterious about this ability. After all, we have the power
to consent to conduct that would otherwise violate our rights. Legitimate authorities have the power to impose
obligations on others by fiat. And if we already implicitly accept that there are some non-mysterious moral
powers, we have reason to suspect that promising might qualify as yet another one.
Seana Shiffrin provides one of the most fully developed and recent Moral Power Theories, a version
of what I call an Agency Theory of the moral power to promise. Agency Theories argue that the moral power is
grounded in basic capacities that we have as moral agents, such as the capacity to commit to things in general.
The power to promise is nothing more than a specific instance of this general capacity; we would not fully
understand moral agency without understanding that agents are the kind of beings able to form
commitments. As we will see, Shiffrin’s particular version of Agency Theory explains promissory
commitments in terms of our basic ability to transfer rights to others. She bolsters this account with an
independent, transcendental argument for the existence of sui generis moral powers to promise.
Although I am sympathetic to Agency Theory, I ultimately think that it faces what I call the
Irrevocability Problem. The problem is how to explain why promissory obligations are not ordinarily
revocable given that they are voluntarily undertaken. Shiffrin appears to recognize the problem, though not in
these terms. She argues that full moral agency requires the ability to give up one’s rights irrevocably, and since
some of us have full moral agency, we have reason to believe that we have the moral power to transfer rights
by promising. I argue, however, that her argument fails. Indeed, there is a general reason why we cannot
ground irrevocability in the capacities of moral agents as rights holders: The Irrevocability Problem, stated
differently, is the problem of explaining why it is morally wrong to revoke a promise. This calls for an
explanation in terms of moral norms that govern our capacities as agents, rather than in terms of these capacities.

(“When acting as promised is neither possible nor permissible, no reason to carry it out follows.”). But I do
not take up the question of moral constrains on valid promising in this dissertation.
65

I therefore conclude that although Agency Theory provides some good insights about the inter-definability of
promising and other things we do with rights like transferring rights, Agency Theory is incomplete.
Finally, I return to contract law, and the basic question of whether Promise Theorists of contract
would do well by trying to justify contract law in terms of Agency Theory. I will argue that a Promise Theory
that relies on an Agency Theory of promising will be incomplete, given the Irrevocability Problem. One of
the core features of contractual obligations is that they are supposed to be irrevocable. Indeed, one of the
initial motivations for Promise Theories of contract law was the observation that both promissory and
contractual commitments share this normative feature. But relying on Agency Theory makes justifying this
very feature difficult without relying on some other normative resources that go beyond the building blocks
of moral agency. So lacking an account of irrevocability in the context of a theory of promising creates a
justificatory gap in the contractual one.
2 The Concept of Normative Power: From Capacious to a Narrow Understanding
Before turning to these arguments, let me set forth some working definitions that will carry through
the rest of the paper. Let’s start with a capacious understanding of ‘normative powers’ held by Wesley
Hohfeld and adopted by Judith Jarvis Thomson. In this capacious sense, ‘normative powers’ are abilities to
change someone’s rights or obligations by saying or doing certain things.
81
Some authors have specified the
relevant normative changes in terms of reasons for action.
82
In discussing promises and contracts, however,
we are ultimately concerned with how our words or deeds manage to affect rights and obligations.

81
JUDITH JARVIS THOMPSON, THE REALM OF RIGHTS 57 (1990) (“Following Hohfeld, I will say that a
power is an ability to cause, by an act of one’s own, an alteration in a person’s rights, either one’s own rights
or those of another person or persons, or both.”); W. N. HOHFELD, FUNDAMENTAL LEGAL CONCEPTIONS
(1919) (W. W. Cook, ed.). Thompson expresses the point in terms of rights, but the capacious sense of
normative powers will assume the same thought can be expressed in terms of rights or obligations.
82
See Ruth Chang, Commitments, Reasons, and the Will, in 8 OXFORD STUDIES IN METAETHICS 74, 75 (2013)
(characterizing normative powers as “the power to confer reason-giving force on something through an act
of will” in which “you have created reasons for yourself by willing something to be a reason”). JOSEPH RAZ,
THE AUTHORITY OF LAW 17-18 (2d ed. 2009) (asserting that a normative power is “an ability to change
protected reasons,” which are reasons that count both as reasons in favor of performing a certain actions and
as exclusionary reasons for “disregarding reasons against it.”). Here, Raz abandons an earlier formulation,
which formulated normative powers solely in terms of the ability to affect exclusionary reasons. See id. 18 n.18
(claiming to have erred IN JOSEPH RAZ, PRACTICAL REASON AND NORMS 101 (1999) (“A normative power
is an ability affect exclusionary reasons which apply to one’s own or other people’s actions.”)).
66

Accordingly, I will simplify things by focusing on our normative powers to change rights and obligations,
assuming without argument that these normative concepts can somehow be translated into reasons.
83
So the
capacious sense of normative powers is, for our purposes, abilities to change someone’s rights and obligations.
84

One way that we exercise a normative power, in this capacious sense, is by promising. By promising,
we impose obligations on ourselves to perform an action for others
85
(and conferring corresponding rights on
others to demand performance) by saying or doing certain things under circumstances.
86
Regardless of how
those circumstances are specified (a task for any full theory promising), there’s a sense in which all normative
theories of promissory obligations are attempts to explain how we have the normative power to promise. That
is, normative theories of promising try to answer the question, how we are able to impose obligations on
ourselves to others by promising? So from this perspective all writers attempting to give normative
explanations of promising are “normative power” theorists.
But there is also a narrower understanding of normative powers in common currency, especially
among Normative Power Theories of promising. For example, David Owens offers the following definition:
Narrow Normative Power. Agent A has a normative power in circumstances C if and only if A has
the ability in C to (i) create or change someone’s obligations by (ii) intentionally communicating the
intention of so doing.
87

83
Here I follow David Owens in focusing on obligations rather than reasons more generally. See DAVID
OWENS, SHAPING THE NORMATIVE LANDSCAPE 4 (2012) (stipulating that normative powers are the ability
to “change what someone is obliged to do by intentionally communicating the intention of hereby so
doing.”). For a similar focus on requirements, see Gary Watson, Promises, Reasons, and Normative Powers, in
REASONS FOR ACTION 155 (David Sobel and Steven Wall, eds., 2009) (stating that a normative power is the
power to “create or rescind a practical requirement at will”).
84
Sometimes I will simply speak in terms of obligations to simplify.
85
Connie Rosati has argued that we can make genuinely binding promises to ourselves. This is controversial.
But if she is correct, then understand “others” to include our future selves as well. See Connie Rosati, The
Importance of Self-Promises, in PROMISES AND AGREEMENTS: PHILOSOPHICAL ESSAYS (Hanoch Sheinman, ed.
2011) (Chapter 5).
86
I use ‘circumstances’ here as placeholder term. Some writers hold that promising communicates not only an
intention to do something, but also an intention to be obligated to do that thing. But I take it that all agree
that, minimally, what is communicated by promising is an intention to do something.
87
This comes, with modifications that will be explained, from DAVID OWENS, SHAPING THE NORMATIVE
LANDSCAPE 128 (2012) (quotation marks omitted, roman numerals added).
67

To see what motivates this narrower formulation, contrast Narrow Normative Power with Hohfeld’s
capacious one. Hohfeld’s formulation is, in effect, (i) only: virtually any way that we behaved that entailed a
change of someone’s obligations qualifies as an exercise of a normative power. Those who construe
normative powers narrowly argue, however, that this formulation is over-inclusive because it allows that, in
David Enoch’s words, “merely [] manipulat[ing] the non-normative circumstances in such a way as to trigger
a dormant reason that was there all along”.
88
To illustrate, suppose that Antony insults Cleopatra with the aim
of incurring an obligation to make amends. (Why? Suppose that apologizing face-to-face is the only way that
Anthony can get close to her.) Even so, this doesn’t mean that Antony has thereby exercised a normative
power. Element (ii) functions to rule out cases like this, and it appears to succeed: Nothing in the act itself—
nothing in Antony’s insult—communicates an intention to undertake an obligation to change any obligations,
regardless of Anthony’s peculiar aims, given that insults ordinarily do not communicate an intention to
thereby change obligations.
Defenders of the Hohfeldian framework might point out that Narrow Normative Power simply
picks out a subset of normative powers captured by Hohfeld’s formulation, and is therefore not strictly
inconsistent with it. Defenders might also insist that Hohfeld’s definition can accommodate the
aforementioned worries about over-inclusion. But because nothing in particular hangs on which conception
we employ, pursuing this conceptual dispute probably will not bear much fruit. More important, I think, is to
be aware of the two ways of understanding the idea of a normative power, and that Moral Power Theories of
promising, as they are called in the literature, typically operate with the understanding akin to the one
expressed in Narrow Normative Power. To fix ideas, I will rely on Narrow Normative Power in explaining
Agency Theory.
More important still is to have some examples of normative powers in mind. The following
normative powers count as such, regardless of whether one accepts the Hohfeldian formulation or Narrow
Normative Power:

88
David Enoch, Giving Practical Reasons, 11 PHIL. IMPRINT 1, 4 (2011).
68

Orders. When a legitimate authority orders a subordinate to do something, the authority thereby—
by virtue of the communicated order—places the subordinate under an obligation to do that thing.
Consents. When a person consents to certain conduct, he says something or does something that—
by virtue of the communication—permits someone S to do something to him that S might not
otherwise be entitled to do. Granting permission, in turn, involves waiving a right or suspending
others’ obligations.
Promises. When we promise we intentionally communicate our taking on, by virtue of that very
communication, an obligation to do something for someone else.
Notice how these powers operate. Ordering creates new obligations in part by communicating an
intention to impose an obligation.
89
By consenting to surgery, you thereby permit the surgeon to operate on
you, rescinding the surgeon’s obligation not to cut you. As for promising, one’s ability to form binding
obligations just is, according to Moral Power Theorists, the ability to undertake obligations to perform actions
for others by communicating the intention to do so. Whereas consenting involves exercising a normative
power that normally grants permissions, thereby suspending others’ obligations towards us, promising
imposes new obligations on the promisor and confers certain rights on the promisee. We simply do have the
power to impose obligations on ourselves by our communicating the intention to do so (provided that other
circumstances obtain). Or so hold Moral Power Theories.
3 The Nature of Moral Powers: Interests or Agency
We have only set the conceptual stage so far. It is possible that, even though we have the concept of
a normative power, the concept does not actually pick out anything in the world. But this seems implausible.
Normative powers are ubiquitous. Umpires have normative powers to eject players from baseball games; by
ejecting them, umpires impose on the players an obligation to leave the baseball field. More generally, social
practices and institutions, including legal institutions, confer powers on agents all the time. And although

89
Joseph Raz claims that having the normative power to order is tantamount to having legitimate authority.
JOSEPH RAZ, PRACTICAL REASONS AND NORMS 100 (1999). But this is controversial. See Andrei Marmor,
An Institutional Conception of Authority, PHIL. & PUB. AFFAIRS 238, 238 (2011) (asserting that “practical
authorities are not necessarily legitimate”).
69

Hume does not use the term ‘normative power,’ even he accepted that social practices and conventions could
arise that in effect delegate normative powers to individuals, including the power to promise. This was
Hume’s solution to the Challenge he posed, as well as the solution of Practice Theorists more generally.
Much more controversial—and what Hume in effect denied—is the claim that agents can have
normative powers absent any social practices, conventions, or institutions to confer them. Call a normative
power a moral power if and only if moral agents can possess it without its being conferred on them by an
actually existing social practice or convention.
There are two schools of thought on how to establish that moral agents have moral powers. The split
resembles another well-known division in rights theory. Interest Theories of rights hold that rights are best
explained and individuated in terms of interests that they serve. Roughly, we have moral rights when we have
interests sufficient to justify holding others under duties.
90
Will or Choice Theories hold that rights fundamentally
empower moral agents to control the duties of others.
91
A version associated with Kant tries to ground rights
in a single innate, fundamental right to be one’s own “master”—to control one’s means to ends that one
chooses, consistent with each other moral agent’s equal right to do the same. Other rights derive from this
innate right, which partly constitutes moral agency itself rather being in the service of any interest of that agent.
92

But on Will Theory, no matter the version, one’s moral agency and one’s status as a holder of rights are
explained in terms of each other.
A similar divergence appears in conceptions of moral powers. Some writers—notably, Joseph Raz,
Dori Kimel, David Owens, and Ulrike Heuer—explain normative powers in terms of interests or values.
These are Interest Theories of moral powers. Others—notably, Ruth Chang, Seana Shiffrin, and Gary Watson—

90
JOSEPH RAZ, MORALITY OF FREEDOM (1986) (Chapter 5).
91
See Leif Wenar, Rights, STANFORD ENCYCLOPEDIA OF PHILOSOPHY (2015) (“In Hohfeldian terms, will
theorists assert that every right includes a Hohfeldian power over a claim. In colloquial terms, will theorists
believe that all rights confer control over others' duties to act in particular ways.”),
http://plato.stanford.edu/entries/rights/.
92
ARTHUR RIPSTEIN, FORCE AND FREEDOM: KANT’S LEGAL AND POLITICAL PHILOSOPHY 30 (2009).
70

explain moral powers as constitutive of moral agency.
93
Call members of this group Agency Theories. Both
families share the belief that moral powers explain our ability to make morally binding promises, and do so
without necessarily relying on the existence of power-conferring social rules.
94

The rest of this Chapter focuses on Agency Theories. Agency Theorists acknowledge that some
normative powers, the umpire’s ejection power, for instance—can be conferred by social practices. But they
deny that we must explain all normative powers that way, and in particular, that we must understand the power
to promise by reference to power-conferring social practices. Instead, we have certain basic moral powers in
virtue of being moral agents.
95
Moral powers are constitutive of being moral agents. One of those powers is
to make morally binding promises. And insofar as the power is a genuinely moral one, Agency Theorists hold
that we would have this power to promise even if no social practices were in place to guide our behavior or
empower us to make promises.
To explain this power, Agency Theorists begin by observing that we have moral rights independent
of any institutions or widely accepted social practices to create or maintain those rights. Once this much is
granted, Agency Theorists argue that it is a short step away to answer Hume’s Challenge. Recall the
Challenge: How is it possible to undertake an obligation by willingly communicating an intention to do so? By
what normative mechanism? Instead of arguing that entrenched social practices of promising confer this

93
Ruth Chang has not, to my knowledge, taken a stand on promising; her project—Hybrid Voluntarism—is a
broader meta-normative theory that insists that the will is a non-exclusive “source” of normativity. That is,
under constrained circumstances, certain facts can become reasons for actions simply by our willing that them
to be reasons of this kind. This seems to have implications for promising. But to my knowledge, Chang has
not extended this view explicitly. See generally Ruth Chang, Commitments, Reasons, and the Will, in 8 OXFORD
STUDIES IN METAETHICS (2013); see also Seana Shiffrin, Promising, Intimate Relationships, Conventionalism, 117
PHIL. REV. 481 (2008); see also Seana Shiffrin, Immoral, Conflicting, and Redundant Promises, in REASONS AND
RECOGNITION: ESSAYS ON THE PHILOSOPHY OF T.M. SCANLON 155-177 (R. Jay Wallace, et al., eds. 2011);
Gary Watson, Promises, Reasons, and Normative Powers, in REASONS FOR ACTION 165 (David Sobel and Steven
Wall, eds., 2009).
94
See e.g., Gary Watson, Promises, Reasons, and Normative Powers, in REASONS FOR ACTION 165 (David Sobel
and Steven Wall, eds., 2009) (claiming that some normative powers, including the power to promise, are not
be grounded instrumentally, but rather as “integral to how we think of our moral standing”); see also id.
(“[P]romissory power is inherent in the kind of moral standing that we have in mind when we speak of
ourselves as autonomous beings, a standing to determine what determine what others may permissibly do.”)
95
I discuss Shiffrin’s views at length below. See also Gary Watson, Promises, Reasons, and Normative Powers, in
REASONS FOR ACTION 165 (David Sobel and Steven Wall, eds., 2009).
71

power on us (as the Practice Theorists maintain), Agency Theorists explain promising in terms of our
inherent capacity as moral agents to manipulate certain moral rights in certain ways, such as transferring or
waiving them.
Seana Shiffrin defends one of the most completely developed, contemporary versions of this view, so
I focus primarily on her claims in what follows.
96
She explains the moral power to promise in terms of our
ability to transfer certain rights to promisees. The particular right at issue is, roughly, a right to decide to
perform some action other than that which has been promised: By promising to φ, the promisor transfers to
the promisee, and thereby gives up, her right to decide to act otherwise. The rest of this section elaborates
and defends Shiffrin’s particular rights-transfer against the claim that we cannot make sense of promissory
obligations by invoking the notion of a rights transfer of a right to decide to do otherwise than some action.
Let us evaluate this proposal in more detail below.
4 Agency Theory: Promising as Transferring Rights?
Shiffrin claims that promising involves nothing more than one’s transferring to someone else the
right to decide to do otherwise (than that which has been promised). The transferee is the promisee in this
transaction, and the transferor is the promisor. Shiffrin thus seeks to respond to Hume’s Challenge by
reducing a supposedly mysterious power—mysterious on par with “transubstantiation”—to the familiar
power that we have to transfer things to one another. We are able to bind ourselves, according to Shiffrin, by
giving certain rights to others. So contra Hume there is thus nothing mysterious about promising or at least
nothing more mysterious than transferring a right.
To evaluate this simple claim, let us start with the question: why rely on the notion of a rights transfer?
According to Shiffrin, transfers conveniently explain important features of the relationship between promisor
and promisee:

96
Seana Shiffrin, Promising, Intimate Relationships, Conventionalism, 117 PHIL. REV. 481 (2008); see also Seana
Shiffrin, Immoral, Conflicting, and Redundant Promises, in REASONS AND RECOGNITION: ESSAYS ON THE
PHILOSOPHY OF T.M. SCANLON 155-177 (R. Jay Wallace, et al., eds. 2011) (defending the view spelled out in
2008). Gary Watson defends a similar view. See
72

The transfer explains why the promisor is bound to perform: he no longer has a (moral) right to
decide whether to do otherwise. It further explains why the promisee has the power to waive
performance: the moral permission of the promisor to decide to act otherwise now lies in the hands
of the promisee and the promisee may exercise this permission. The theory thereby explains the basic
structure of the promissory relationship with satisfying simplicity.
97

Shiffrin appears to suggest that transfers account for the core structure of the promisor-promisee
relationship. First, when you transfer the right to decide to do otherwise than x, you no longer have the right
to decide to do anything other than x. But if you lack the right to decide to do anything other than x, then
you must do x. And if you must do x, then you are obligated to do x. Therefore, when valid, a promise to do
x obligates you to do x. Second, once the right to decide is transferred, the transferee (now promisee) obtains
the right to decide whether the transferor (now promisor) may do otherwise. This putatively explains how the
promisor acquires the right to waive performance.
One quibble with this passage is that transfer appears to play no essential role in capturing the features
of promissory obligation. To see why, consider Shiffrin’s assertion that lacking the right to decide to do
otherwise than x for person P is tantamount to being obligated to refrain from doing other than x. Assume that
this is correct.
98
But if so, then any method by which a person can give up that right will succeed in generating
the obligation. This might include, say, waiving that right or abandoning it. The important thing is the giving up
of a right to decide rather than transferring it per se. And if the giving up the right suffices to generate the special
or directed obligation to P, then this also suffices to guarantee that the correlative promissory right winds up
in the hands of P. So it is not obvious, at least from the passage above, that the notion of a transfer plays any
essential role Shiffrin’s rights-transfer version of Agency Theory.
More serious problems loom. To begin, notice that the passage above contains an invalid inference.
It does not follow from the fact that one has given up or transferred a right to decide to do otherwise than x

97
Seana Shiffrin, Immoral, Conflicting, and Redundant Promises, in REASONS AND RECOGNITION: ESSAYS ON THE
PHILOSOPHY OF T.M. SCANLON 157 (R. Jay Wallace, et al., eds. 2011) (defending the view spelled out in
2008).
98
Shortly I will argue to the contrary.
73

that one must do x. In joining the military and entering basic training, you give up the right to decide whether
you wake up at time t. This does not yet entail that you thereby have an obligation to wake up at time t. You
neither have the obligation to wake up at t nor lack that obligation. It is undecided one way or the other unless
and until the officer has exercised that right and orders you to wake up at t (or some other time).
99
So merely
transferring or otherwise giving up that right does not guarantee that one has imposed an obligation on
oneself. And that is precisely what an account of promising must explain: how it is possible to impose an
obligation on oneself.
One might reply that the relevant right is really a power to decide whether to do other than x. The
suggestion is that the promisee obtains from the promisor the power to impose an obligation on the
promisor. But this means that, pre-transfer, the promisor possessed that same power. (This is because
transfers involve symmetries of gain and loss; the transferee gains, after the transfer is complete, exactly the
object that the transferor loses.) But if the promisee gains a moral power to impose a moral obligation on the
promisor, the symmetry of gains and losses implies that the promisor had the power to self-impose moral
obligations. If so, we have a viciously circular explanation of the moral power to promise, since the attempted
explanation depends on one’s possessing a moral power to impose moral obligations on oneself. This is
precisely what we are trying to establish, after all. To avoid circularity, Shiffrin’s proposed right must be a
more quotidian claim right to decide whether to do something other than x, one that places others under the
duty not to interfere with that decision. This is why I interpret Shiffrin’s right as a claim right.
Here is another reply to the invalid-inference objection. Shiffrin might point out that transferring the
right to decide is usually and in all relevant respects like promising. Assuming that the recipient of the right to
decide wants the promisor to perform the proposed act, she will simply exercise that right by, say, ordering
performance. This fails, however, to fit the phenomenology of promissory morality. To see why, assume that
A possesses a right to decide whether to do something other than to give B a watch. Now assume that A can

99
Stephen Smith appears to make a similar point in criticizing rights-transfer theories of promising, but it is
not entirely clear given his penchant for describing the relevant rights in terms of liberty rights, when it is far
from obvious that rights-transfer theorists have in mind liberty rights to begin with. See STEPHEN A. SMITH,
CONTRACT THEORY 101 (2004).
74

transfer this right to B and does so. So now B possesses the right to decide whether A does something other
than give B a watch. And A has no right to decide whether to do something other than to give B a watch. But
this is perfectly consistent with A’s still having no obligation to give B a watch. A might have no obligation either
way; A’s promissory obligation, if any, is held in abeyance until B decides the matter. So, again, giving up that
right does not yet entail that A is obligated to give B a watch unless and until y exercises that right to decide what
A should do.
That is not how promising works.
100
When one makes a promise, one thereby undertakes an
obligation; one does not, first, delegate authority to impose a promissory obligation and then, second, wait for
the promisee to exercise that authority. Shiffrin’s account fails to capture the right timing of the relevant
promissory rights. The right to a performance—the promisor’s self-binding—coincides directly with the
promisee’s right to waive performance (the right which corresponds to the putative right to decide whether
the promisor performs). The right to decide does not come first dragging the performance right in its wake, if
at all.
This complaint about timing is not just about whether Shiffrin’s account captures all the right data
points as a descriptive matter. After all, Shiffrin might reasonably point out that nearly any philosophical
account has some revisionist aspects built in. The worry is that the inaccurate timing predicted by her account
indicates an order-of-explanation mistake. That is, Agency Theorists generally and rights-transfer theorists in
particular must explain first and foremost how it is that promisees end up with promissory rights—rights to
performances—not the derivative waiver rights that come along with them. Once we explain how rights to a
performance end up in the hands of promisees, we get waiver for free. After all, if I have a right to your
performance, then I may waive that right. This explains the limited sense in which I have the right to decide
whether you perform. But Shiffrin tries to use this derivative and very limited right to explain obligations to
perform. This looks exactly backwards.

100
For a similar worry in the context of so-called rights-transfer theories of contract, see STEPHEN A. SMITH,
CONTRACT THEORY 101 (2004) (“Contracting parties do not own rights ‘to the performance of a future act
(by them)’, which they can transfer to another party.”). Similar points are made in David Owens, Does a
Promise Transfer a Right? in PHILOSOPHICAL FOUNDATIONS CONTRACT LAW 79 (Gregory Klass, et al., eds.
2014).
75

Something is amiss. If the culprit is the fact that Shiffrin emphasizes a right to decide, it is worth
pausing to consider why this right is emphasized. This characterization is initially surprising. After all, if
promising involves transferring a right to another person, the most natural suggestion is the right to a
performance.
101
So what motivates characterizing the right in terms of a right to decide to do otherwise?
Although Shiffrin does not address this question directly, the answer probably relates to the initial theoretical
choice of explaining the moral power to promise in terms of the power to transfer. Recall the assumption that
I called the symmetry of gains and losses, which we can call Symmetry for short. Symmetry is a basic
constraint on whether some transaction counts as a genuine transfer. To satisfy Symmetry, recall, A must
possess some object o before the transfer but not afterwards; and B must possess o after the transfer but not
before. Now we can see why the relevant right cannot be the “natural” suggestion—i.e., the right to a
performance. Even granting that B has a performance right after the transfer, the promisor lacks the same
right before the alleged transfer: he has no performance right that correlates to anyone’s obligation to
perform it. True, A has a “right” to perform an action, whether characterized as a liberty to perform or a claim
right that corresponds to a duty of others not to interfere with that performance. But this is not the same as
having a right to a performance that correlates to another’s obligation to perform. That bipolar obligation
does not arise until after exercising the moral power to promise.
Symmetry does not rule out Shiffrin’s preferred formulation of the right. When we promise to φ,
according to Shiffrin, we transfer a right to decide whether to do something other than φ.
102
Suppose that A

101
Or perhaps a right to demand a performance. For an example of this claim, see Joseph Raz, Promises and
Obligations, in LAW, MORALITY, AND SOCIETY: ESSAYS IN HONOR OF H.L.A. HART 214 (P.M.S. Hacker and
Joseph Raz, eds. 1977) (observing that “promises are commitments undertaken through being communicated
to an addressee who acquires a right to demand the performance of the promise and the right to release the
promisor from his obligation.”) (emphasis added). This suggestion, however, reverses the order of
explanation: a right to demand a performance, if legitimate, is grounded in a prior right to a performance.
And that is what we are trying to determine: how that right ends up in the hands of the promisee. Shiffrin
resists Judith Jarvis Thomson’s characterization—that promising involves giving a claim to the promisee—on
similar grounds. See id. at 172 n.7.
102
The final formulation of the right is more complex: a right to decide, as between the promisor and promisee. This
avoids certain objections based on the alleged possibility of conflicting yet morally binding promises. Immoral,
Conflicting, and Redundant Promises, in REASONS AND RECOGNITION: ESSAYS ON THE PHILOSOPHY OF T.M.
SCANLON 164 (R. Jay Wallace, et al., eds. 2011) (defending the view spelled out in 2008).
76

promises B that A will buy groceries. Notice that, before promising, A has the right to decide to do
something other than buy groceries for B. But A lacks that right after promising—he gives it up by
promising. Before the promise, B lacks the right to decide that A will do something other than buy groceries
for B. But B has this right after promising. So at first glance the Shiffrin’s transfer-of-rights approach satisfies
Symmetry.
103
This appears to explain the motivation behind Shiffrin characterization of the relevant right as a
right to decide rather than directly in terms of a right to a performance.
Regardless of the motivation for characterizing the relevant right as a right to decide, there is one
final—and most important—worry about Shiffrin’s rights-transfer view: the Irrevocability Problem. Indeed,
even if we could somehow replace the notion of transfer with some other normative power like assignment that
does not appear constrained by Symmetry, and even if we would thereby be freed to sketch an account
directly in terms of a right to performance (in lieu of the derivative right to decide to do otherwise), the
Irrevocability Problem would remain.
The worry is this. Assuming that we have the power to transfer, assign, and otherwise allocate
performance rights to someone else, what makes this allocation irrevocable? Consider this argument. To be
morally valid, exercising the power to transfer something requires doing so with the consent of the transferor.
But many theorists—Shiffrin included—take consent to be prima facie revocable. If I consent to your touching
me, I can revoke that consent at any time. And transfer arguably works the same way. I can lend you
something—one way of transferring that thing to you—but I can normally revoke that transfer at my
discretion. So we need to rely on some normative resource to explain why certain transfers or consents are
irrevocable, a resource that probably goes beyond the constituents of moral agency.
Nothing in what we have seen so far answers this objection. Of course Practice Theorists stand ready
to fill the void. We very well might have moral rights, they might point out, and perhaps we can transfer or

103
David Owens rejects without argument Shffrin’s formulation of the relevant right, insisting that her
formulation “is hard to elucidate further.” David Owens, Does a Promise Transfer a Right?, in PHILOSOPHICAL
FOUNDATIONS OF CONTRACT LAW 82 (Gregory Klass, et. al., ed. 2014). He apparently assumes that the
same asymmetry transfer-related problems arise for the right to decide to do otherwise, proposing that we
should “drop[] the transfer metaphor” and abandon the “attempt to identify some act or acts that the
promisor used to have the right to perform (or omit), which right the promisee now possesses.” Id. at 82.
77

assign them to others by our mere say so. But to do so irrevocably—as promises typically do—requires social
practices or conventions that stipulate that revocations of rights transfers, under certain practice-defined
conditions, are irrevocable. Against this suggestion, Agency Theorists might simply insist, more or less
dogmatically, that being a moral agent with moral rights and liberties presupposes being able to give some of
them up irrevocability. Perhaps. But Agency Theorists who seek to more than dogmatism need an argument.
Shiffrin tries to do just this.
5 Shiffrin’s Transcendental Argument
So far we have seen an attempt to try to explain promissory obligations in terms of a supposedly less
controversial power that we have to transfer certain rights to others. Although the attempt raises difficulties,
this is not the only way to establish the existence of a moral power to promise. Shiffrin’s main argument is a
transcendental one.
The form of argument is this. She argues that, to exercise our moral agency in the world in minimally
respectful ways, we must have some candidate normative power M. Because we are able to exercise our
agency respectfully towards one another even in the absence of a shared social practice establishing M, it follows that
we have M and it is a moral power—i.e., a normative power that we have in the absence of shared social
practices establishing that power.
Using this strategy, Shiffrin argues that we have a moral power to consent, and then turns to
promising. I reconstruct Shiffrin’s argument regarding consent as follows:
P1. Full moral agency requires the capacity to interact with other moral agents in minimally respectful
ways.
P2. Interacting with other moral agents in minimally respectful ways requires the ability to permit
others to interact with us and gain permission to do the same—i.e., a normative power of consent.
P3. Exercising this capacity doesn’t necessarily require the existence of any particular social practice
that confers this power on agents.
P4. A normative power that does not necessarily require the existence of any particular social practice
that confers that power on agents is a moral power.
78

C. Therefore, full moral agency requires the moral power to consent.
On the further assumption that there are full moral agents among us, we can conclude that at least
some of us we have the moral power to consent in this way.
Let me explain the premises. P4 is simply definitional; the remaining premises do the heavy lifting.
Some might resist P1 insofar as they harbor doubts about moral agency and respect. These are fraught
concepts, which Shiffrin does not fully discuss, and which cannot be fully addressed here. Still, we should say
something. Full moral agency requires the ability to recognize and conform to moral demands and
expectations, including by respecting the rights of others. This includes the ability to avoid violating others’
rights; doing the bare minimum, morally speaking, involves not violating others’ rights. Whether a hermit
who studiously avoids violating others’ rights counts as a full moral agent depends on whether the hermit has
the capacity to interact with others in minimally respectful ways given the chance. But if the hermit fully lacks
this capacity even if given the chance, the hermit lacks full moral agency.
P2 connects minimally decent interaction with the power of consent. Having the capacity to engage
with others in minimally respectful requires the ability to give and obtain consent. Consider how difficult it
would be to interact with one another in minimally respectful ways without the ability to grant and obtain
permissions to interact with others in mutually beneficial arrangements, let alone basic interactions that
presuppose the giving and gaining of permissions. Imagine that some highly intelligent being B is thrown into
the world with other highly intelligent agents. Now imagine that B—but no other highly intelligent being—
lacks the ability to give and obtain permissions. All of B’s interactions with others come at extremely high risk
of violating others’ rights. Would B, despite its intelligence, be able to engage with others in minimally
respectful ways?
No. Part of being a full moral agent involves having the capacity “[t]o forge meaningful
relationships,” which in turn requires, among other things, “the ability to interact within the same physical
space, to share the use of property, and to touch one another.”
104
And doing all of this in minimally respectful
ways requires being able to do so with permission. So the ability to give and obtain permission, and hence

104
Seana Shiffrin, Promising, Intimate Relationships, Conventionalism, 117 PHIL. REV. 481, 502 (2008).
79

consent, is a sine qua non of full moral agency.
105
The upshot is that at least one moral power—the power to
give and obtain consent—is an indispensible ingredient of moral agency itself.
P3 goes further. Not only is the moral power to consent indispensable, no entrenched social practice
confers this power on us. Roughly, “social practices” are patterns of human activity constituted and regulated
by social rules that are generally followed by a populace.
106
Something like P3 is necessary to show, as is
Shiffrin’s aim, that our ability to consent does not necessarily hinge on such practices to confer that power on
us the way that the power to eject players is conferred on umpires. To bolster the point, consider a case in
which two moral agents (Adam and Eve) encounter each other for the first time in the state of nature. They
share no prior practices—even linguistic practices—in common (cf. Scanlon 1998). Plausibly, these agents still
have the capacity to consent. After all, Eve has the capacity to withdraw from or decline to form any
relationship whatsoever with Adam. Likewise, Adam must obtain consent from Eve before forming a
relationship with her. But, as strangers encountering each other in the state of nature, there is no sense in
which this capacity to engage with one another consensually rests on established social practices generally
adhered to in a population. Cases like these are not conclusive. But they do provide some prima facie support
for P3.
The argument about consent seems plausible. Indeed, I’ll grant it for the sake of argument. Why
shouldn’t the Agency Theorist, then, try to ground promising in consent, given the plausibility of this
argument? This is tempting. After all, if a promise is made without the promisor’s consent, defeats the validity
of the promise. So consent seems a necessary condition on promissory obligation. So perhaps consent also
grounds promissory obligation. The main reason to avoid this suggestion, however, is the Irrevocability
Problem. As we’ve seen, promises are importantly normatively different from consent. Promises involve

105
It is worth noting that, in law, the terms the legal “age of consent” and having legal “capacity” are
synonymous. Moral power theorists take the moral analogue seriously.
106
Conventional rules are a species of social rules, at least in a narrow sense of “conventional.” See ANDREI
MARMOR, SOCIAL CONVENTIONS 8-9 (2009) (explicating conventions as arbitrary in the sense that, for every
conventional rule, there is at least one other rule that could have been implemented for roughly the same
purposes).
80

undertaking an obligation irrevocably. But consent normally operates differently, since it often can be
withdrawn. This was the Revocability Problem.
This chief difference between promising and consenting motivates Shiffrin to offer an independent
transcendental argument in order to establish a power to promise. The argument is similar in form to the
argument about consent, but aims—as I construe the argument—to capture qualities of promissory
obligations, including their irrevocability. Just as full moral agency seems impossible without the power to
consent
a conception of a moral world without the power to promise seems unsustainable. … Meaningful,
moral relations depend on agents having the ability to make binding promises.
107

And because we can have meaningful, moral relations with others, we have the ability to make
binding promises. Shiffrin adds that it would be “ludicrous” to think that this capacity depends on the
existence of a widespread social practice of promising, causing her to reject Practice Theories of promissory
morality.
108

Let’s try to examine this argument more carefully. I reconstruct it below:
p1 Full moral agents have the capacity to form and maintain minimally decent close relationships
with others.
p2 Forming and maintaining these relationships does not necessarily presuppose widespread social
practices that empowers us to form and maintain these relationships.
p3 Maintaining these relationships does require the ability to give up (irrevocably) certain liberties not
to perform actions for others—i.e., a normative power to promise.
p4 A normative power that does not necessarily require the existence of any particular social practice
that confers that power on agents is a moral power.
C. Therefore, full moral agents have a moral power to promise.

107
Seana Shiffrin, Promising, Intimate Relationships, Conventionalism, 117 PHIL. REV. 481, 502 (2008).
108
Seana Shiffrin, Promising, Intimate Relationships, Conventionalism, 117 PHIL. REV. 481, 502 (2008).
81

If we further assume that most of us qualify as full moral agents, most us have a moral power to
promise.
Again, p4 just defines what moral powers are, as opposed to other normative powers conferred by
social practices. The other premises are more important. As for p1, Shiffrin disclaims any attempt to
“establish definitively that the power to conduct meaningful, equitable relationships is inherent in the
capacities of an autonomous moral agent.”
109
But she nevertheless asserts that it is “implausible” that a
person who lacked the ability to form minimally decent close relationships—i.e., relationships manifesting
high levels of trust, on par with family or friends—could fully realize her agency in the world.
110

p2 asserts that minimally decent interpersonal relationships do not necessarily depend on the
existence of social practices. This initially looks implausible. Our relationships are so bound up with social
practices that it might seem difficult imagine what kind of anemic relationships remain when we strip away
those practices. Many relationships, like marriage, are constituted by social practices. But Agency Theory does
not deny that widespread social practices play an important role in forming or maintaining close relationships.
They just deny what Practice Theorists assert—namely, that having widely accepted, deeply entrenched social
norms are necessary for promises to be morally binding. So even if there is a sense in which all close
relationships involve “social practices,” this is not the sense relevant for proposing an Agency Theory that
resists Practice Theory. Recognizing this distinction renders p2 plausible: Our ability to form and maintain
close relationships with, say, our family members does not necessarily depend on any social practices of the
relevant kind.
111
The initial process of relationship formation and maintenance does not presuppose any such

109
Seana Shiffrin, Promising, Intimate Relationships, Conventionalism, 117 PHIL. REV. 481, 520 (2008).
110
Seana Shiffrin, Promising, Intimate Relationships, Conventionalism, 117 PHIL. REV. 481, 520 (2008).
111
Shiffrin appears ready to accept that we can form close relationships and commitments without social
practices of any kind. Citing T.M. Scanlon, Shiffrin asserts these relationships involve “brute” commitments
to others, which may even be formed and established “extralinguistically, albeit without the nuances
accessible through a common language (Scanlon 1999, 296–97).” Regardless of whether this stronger claim is
true, I think, for the reasons above, that she does not need it to advance her position.
82

practices. It is unclear how social practices could have been established without the stability inherent in social
relationships that existed prior to those practices.
112

Much more might be said about p2’s plausibility. Still, I am willing to grant p2 for the sake of
argument. As we will see in the next section, p3 represents the most problematic premise. The remainder of
this section tries to explain p3 and Shiffrin’s arguments in favor of it.
p3 holds that close relationships presuppose the ability to self-impose obligations (irrevocably) by
giving up (by waiver or transfer) our liberty to decide to do other than a certain action for someone—i.e., by
promising. We might wonder why. Shiffrin argues that, without the ability to give up certain rights
irrevocably, forming and maintaining close relationships is very difficult indeed:
Parties could not bind themselves through the communication of their intention to do so. At most,
they could say to one another “I intend to φ,” but this would always ride alongside an implicit
disclaimer: “but I am not morally bound to you to φ just because you know of my intention through
my declaration; nor would I be bound if you acted in reliance upon my declaration.”
113

But why should the ability to maintain minimally decent relationships depend on the ability to self-
impose obligations by irrevocably giving up certain rights? The main argument is simple. Unless we actually
obligate ourselves in this way we would always lack confidence and security in our relationships. After all, in
the context of a relationship with another person, we would never be morally justified in relying on that
person’s representations to us about their future plans. No matter how clear my representations about my
intended future conduct are, and no matter how clear and sincere my initial intentions to engage in certain
joint activities with others, I would still have the right not to perform.

112
In a similar vein, Judith Jarvis Thomson argues that if no promissory transaction could be valid absent a
social practice of promising in the background, then it is unclear how the social practice could get off the
ground in the first place. See JUDITH JARVIS THOMSON, THE REALM OF RIGHTS 303-04 (1990). The Practice
Theorist’s answer would have to be, roughly, that the initial rounds of mutually beneficial cooperative
transactions did not generate moral obligations. They did not arise until members of a community routinely
and regularly yielded patterns of cooperation came to expect its members to conform to that pattern. The
expectations of conformity generated, roughly, moral expectations. This is roughly the Humean picture. But
even this picture, which presupposes a community as an input, seems difficult to imagine without some stable
relationships, even if simply familial ones, at the foundation.
113
Seana Shiffrin, Promising, Intimate Relationships, Conventionalism, 117 PHIL. REV. 481, 502 (2008).
83

Indeed, retaining this right (a right to do something other than to perform), or giving it up while
reserving the option to take it back, is at odds with maintaining minimally morally decent relationships. If we
had no power to promise—i.e., no power obligate ourselves irrevocably by our expressing our intention to do
so under appropriate circumstances—others would have no grounds for reasonably demanding our
performance based on our representations. We could always correctly claim, absent such a power, that we still
have the right not to perform. Others have legitimate claims to rely on us only if we lack that right and can’t get
it back. So, without the power to promise, including the all-important power to bind ourselves irrevocably, we
lack the ability to justifiably rely on one another morally, which in turn makes having a close relationship with
that person practically impossible.
In addition to this main argument, Shiffrin points out two relationship-destroying dynamics that
might arise if we were to lack the power to commit by promising. Suppose A wants to do something only if B
will participate. Also suppose that B does not want to participate as much as A does. Two unhealthy
dynamics might arise without the power to promise. First, A may try to coerce, cajole, or create other
incentives to try to B’s defection. These methods may become outright manipulative. The second dynamic
involves B. In recognizing that A wants to pursue some activity together only if B does, B might exploit his
bargaining position, so to speak, and try to extract extreme concessions from A. These dynamics of
manipulation and exploitation are made more likely once A lacks the ability to simply obtain a right to
demand performance from B.
So much for p3. If the premises are true, then it follows that full moral agents have the power to
promise, and that this power does not necessarily presuppose social practices of promising. And assuming we
count as full moral agents, we likewise have this power. Now let’s evaluate the argument. I will focus my
critical attention primarily on p3.
6 Evaluating Shiffrin’s Argument
Recall the dialectic. Agency Theorists typically try to explain how we can impose promissory
obligations on ourselves by reference to commonplaces about how we can dispose of our rights. Shiffrin
argues that promising is nothing more than transferring a liberty to do other than some action. But one
84

feature of promises remained unexplained: the fact that they are normally irrevocable. Agency Theorists
would do well to provide some argument that the ability to irrevocably alienate certain rights depends only on
our having capacities as full moral agents. Otherwise we might suspect that we must invoke Practice Theory
to solve the Revocability Problem—i.e., that the practice of irrevocably alienating rights to performance.
Unfortunately, I do not think that Shiffrin’s argument succeeds. I will grant, for the sake of
argument, Shiffrin’s claim that we have a moral power to consent. I will also grant p1 and p2. The main
problems surround p3. That premise—that promising is required to maintain close relationships with
others—looks questionable at best.
Let’s proceed in reverse order. One of the claims bolstering p3 was that, without promising, we lack
a vital tool for mitigating certain unhealthy dynamics—exploitation and manipulation. But it’s not clear how
having such a tool avoids those dynamics when people don’t antecedently care for, trust, or love each other.
Suppose we live in a world without the power to promise. This doesn’t entail that we lack the ability to care
or love others. And if we have those abilities, it’s unclear why we would be motivated to exploit or manipulate
those we love or care about, even knowing full well that lacked the ability to give rights in the manner that
Shiffrin emphasizes. And even if we had the ability to give up certain rights (however the content of the right
is specified), the possibility of unhealthy dynamics still exists. A would still try to up the ante, coerce, and
otherwise manipulate B in cases where B is reluctant to give up his right to do other than that which he has
promised to do. This still threatens the relationship between A and B. Likewise, B might still try to exploit his
position by extracting relationship-damaging concessions from A. The main problem is not the lack of moral
powers in such cases; the problem is the antecedent morally dubious attitudes of the participants in the
relationship.
But what about the main argument for p3? The argument depends on a further, questionable claim:
that we could not reasonably rely on others without having the power to waive certain rights by making
promises. This seems false. Close-knit families and good friends needn’t rely on promises. Consider families
first. Family members may need little more than the ability to be motivated by love and reciprocity. Parents,
motivated by love, do a lot for their children. As the children grow, they recognize that their parents have
85

sacrificed much for their benefit. After recognizing this, they are motivated by love and reciprocity to helping
their parents when reasonably required. No promises need be part of this story so far. Nor are promises
required to set the wheels of reciprocity in motion. Apart from love, parents have moral duties of care over
their children, which require them to attend to their needs. Duties of kindness and gratitude also motivate
members of this family to behave well towards one another. Perhaps there are moral duties of loyalty, further
strengthening the ties that bind them. Family members can maintain minimally decent relationships with each
other without bringing promises into the picture. And we can tell similar stories about friendships and other
close relationships. So Shiffrin’s claim that promising is practically necessary to maintain minimally decent
close relationships looks doubtful.
Notice another difficulty. Assume that we have a moral power to consent. If so, transcendental
arguments for a distinctive moral power to promise, as opposed to, say, consent become harder to establish.
Recall our imagined being, B. Assume, as we did earlier, that B lacks the capacity to make binding promises.
Now, however, assume that he has the ability to give or obtain consent. We can easily imagine that this
person can have respectful, minimally decent, close relationships with others. After all, in lieu promising to
perform action X for agent B, the agent can simply consent to being held accountable in some way to B for not
performing a certain action. Having the ability to interact grant others permission to interact with him will
make it much easier for this being to form and maintain relationships with others, even if he lacks the power
to promise. Indeed, in many cases consenting might serve functions on par with promising, even if they’re
not strictly identical.
But Shiffrin observes that nothing provides a perfect substitute for a full-throated promise. In
response to a similar objection—i.e., that perhaps we don’t need the power to promise given the ability to
communicate one’s intentions—Shiffrin responds that “[c]ommunicated intentions of this kind can do some
work, but they are not always sufficient.”
114
Shiffrin might similarly insist in response to my consent-based
objection that the power to consent likewise can’t fully substitute for promising. After all, she might add, as I

114
Seana Shiffrin, Promising, Intimate Relationships, Conventionalism, 117 PHIL. REV. 481, 507 (2008).
86

have already emphasized, that consent is paradigmatically the kind of change in the normative landscape that
can be revoked. Normally, promising cannot be undone.
But this response moves the goal posts. The original objection was not that, without the power to
promise, there would be no absolutely perfect substitute for the moral power to promise. Rather, to make the
transcendental-style argument work, Shiffrin must argue that it would be practically impossible to carry on as
moral agents capable of having minimally decent moral relationships without that power. And once our
agents have the power to consent, coupled with other values and norms, this hampers Shiffrin’s ability to
substantiate this claim with respect to promising.
115

7 Revisiting Revocability
To summarize, it is not obvious that we can describe promising in terms of rights transfer along the
lines that Shiffrin has in mind. Even if we grant that moral agents have something like a moral power to
consensually transfer certain rights, this runs into the Irrevocability Problem: consent is normally revocable, while
promissory obligations normally are not. But as we have seen Shiffrin also offers a transcendental argument
for the claim that we have the moral power to promise. If successful, this would answer the Irrevocability
Problem, since the argument tries to show how minimally respectful moral agency requires (at least in
intimate relationships) having the ability to irrevocably grant others promissory rights. But the argument is
not convincing. So the Irrevocability Problem goes unanswered in at least one prominent Agency Theory.
This observation holds not just for Shiffrin’s views, however. Answering the Irrevocability Problem
requires understanding what irrevocability means. What makes promissory obligations normally irrevocable

115
There are other transcendental arguments available. Perhaps there is, for example, a more fundamental
moral power capable of grounding both consenting and promising. On this view, promising are
manifestations of something like the will, understood as a capacity even more basic than the power to
consent. The powers to promise derive, on this view, from the will. Similarly, Immanuel Kant had an
explicitly will-based theory of contracts, according to which contracts amounted a joinder of wills. See
ARTHUR RIPSTEIN, FORCE AND FREEDOM: KANT’S LEGAL AND POLITICAL PHILOSOPHY 109 (2009)
(claiming that both contract and consent require that choices be joined, and that consent “is to be understood
as two persons uniting their wills to create new rights and duties between them”); id at 111 (“[P]romises an
create enforceable obligations to perform future acts only because people have the more general rightful
capacity to make arrangements with each other that change those relations.”). I ignore these views, however,
because I find the will to be an elusive concept, and because Kant’s views on contract strike me as formal to
point of emptiness. But this is obviously too quick a dismissal and warrants further analysis that I cannot
provide here.
87

when they are irrevocable has little to do with formal features of rights or the deep structures of moral agency
itself. The irrevocability of promissory obligations means that it is morally wrong to revoke a promise, not that
revocability is literally impossible or presupposed by moral agency itself. Substantive claims about moral
requirements are needed to explain this wrongness, and hence are needed to explain irrevocability.
It is telling, then, that Shiffrin herself looks to supplement her Agency Theory. After laying out her
rights-transfer account of promising, according to which promising involves “forswear[ing] certain
opportunities that may or may not be to the promisor’s advantage,” she asks, why is it that “this forswearance
be morally binding?”
116
This question, which squarely raises what I have called the Irrevocability Problem,
comes some thirty pages deep into her article. This is surprising, given the centrality of the answering the
question to giving a full account of promissory morality.
Her answer owes too much to the problematic transcendental argument discussed above, so I won’t
discuss it in detail. But I will point out the kernel of truth in it. Shiffrin observes, as others before her, that in
making a promise, one solicits another’s trust, and in many cases revoking a promise represents an abuse of
trust. Similarly, promising involves attempting to give assurances. Backtracking defeats the purpose of these
attempts. In Chapter Four I will provide an independent account of promissory obligation that draws on
these themes. I think that the account works on its own, but in the final analysis, it is compatible with some
version Agency Theory: it can be viewed as an addendum that solves the Irrevocability Problem by explaining
why promises normally may not be revoked without relying on Practice Theory. Revoking a promise and
breaking it amount to the same thing.
9 Conclusion: Challenges for Promise Theories of Contract Law
It is time to return to contract theory. This dissertation’s first four chapters address the question,
which theory of promissory morality would best serve a Promise Theory of contract law? This question has
forced us to consider and evaluate, on its own terms, some of the most plausible recent work on promissory
morality. We will see that each theory has features that make it problematic as an engine capable of driving
Promise Theory’s agenda. In the case of Practice Theory, for example, we saw that it had difficulty explaining

116
Seana Shiffrin, Promising, Intimate Relationships, Conventionalism, 117 PHIL. REV. 481, 518 (2008).
88

the normative priority of promissory practices and expectations in relatively informal setting over promissory
practices as they occur in contract law, or to even convincingly explain why contract law could not be the law
governing a fundamentally different kind of commitment altogether.
We thus turned our attention to Agency Theory as a potential foundation for Promise Theory.
Agency Theories, despite their appeal, are incomplete. They function well in responding to Hume’s
Challenge—i.e., providing a plausible explanation of how it is possible to voluntarily change our rights and
obligations through communicating an intention to do so. But they do not explain how those changes
become irrevocable in the way that promises usually are understood to be.
This incompleteness makes Agency Theory inadequate as a theory fit for Promise Theory. Unless we
have a normative explanation of how and why promissory rights and obligation may not permissibly be
revoked, Agency Theory fails to give as a useful foundation for contract law. Contracts, like promises, are
generally understood to be irrevocable. This feature of contracts explains why those on the losing end of
broken contracts may sue those who break them. After all, if contractual obligations may be permissibly
revoked, a would-be contract breaker could simply assert that she no longer wished to abide by the terms of
the contract and thereby revoke her contractual obligation. No obligation would remain to enforce in the
courts. In short, Agency Theory, without more, provides very little in the way of a theory capable of justifying
contract law.
But maybe this is too quick. Perhaps the Irrevocability Problem points towards a very different kind
of Promise Theory. Start with a premise common in legal theory: that one of the main functions of law is to
correct for certain recurring problems in the application of moral requirements. This theme occurs in the
work of John Locke (who argued that the state and its positive law exists in part to adjudicate disputes that
arise when we are judges in our own moral cases), and has more recently been defended by Jeremy Waldron
and Scott Shapiro (who argue that the law functions to settle moral disagreements decisively, and for that
reason the determinants of law should not or do not include moral standards). And if we assume that the
Irrevocability Problem cannot be solved by Agency Theory, or any other viable theory of promissory
obligation, then perhaps one of contract law’s main functions is to establish a formal practice pursuant to
89

which certain promissory commitments will be irrevocable. Contract law on this view corrects a defect in
ordinary promissory morality, which is just that moral agents are incapable of generating moral obligations on
their own without recourse to certain practices of interpersonal accountability.
Contract law, on this account, would just correct a defect inherent in informal promissory morality
promissory. Agency Theory plays a role in explaining the possibility of promissory obligation, on this view,
but that role would not be complete. The story would go like this: Although we can transfer certain rights
without presupposing social practices of transferring these rights, a practice of holding that certain transfers
or waivers are irrevocable should be, according to the Practice Theories, a matter of social practice or
convention that serves some socially important interests. Contract law is one such set of practices or
conventions.
Of course Agency Theory would in effect become a Practice Theory of promises, insofar as explaining
the distinctive irrevocability of promissory obligations requires recourse to social practices. This resulting
hybrid theory might suffice qua theory of promissory morality. But the same time it would not necessarily be
a theory well suited for Promise Theory, for reasons that we have already seen in Chapter One. A genuine
Promise Theory will need an independent account of promissory morality capable of justifying and guiding
contract law. It is not obvious that an Agency Theory that depends so heavily on social practices will be up to
the task.
90

Chapter Three: Contracts and Interest Theories of Promising
1 Introduction
Recall Hume’s Challenge: Promising involves creating for ourselves moral obligations to perform
actions by declaring an intention to do so. But how is this possible? Moral Power Theory answers directly by
explaining why there is nothing unusual about having the moral power to promise, any more than it is
unusual that we have other moral powers like the power to consent. Promising is just one of a family of
moral powers, explained in the same way these other powers are explained.
In the previous Chapter, I discussed Agency Theory, which tries to ground the moral power to
promise in moral agency itself. I argued that Agency Theory has trouble explaining why promissory
obligations are normally irrevocable. This in turn makes Agency Theory a problematic foundation on which
to build a Promise Theory of contract law. After all, Promise Theory looks appealing in part because the
irrevocability of promissory obligations accounts for the irrevocability of contractual ones. But Agency Theory
threatens to deprive Promise Theory of this appealing feature.
Another approach to Moral Power Theory explains normative powers in terms of our interests or
values rather than in our moral agency. I will call these views Interest or Value Theories of normative powers,
or Interest Theories for short. These theories characteristically take a two-step approach. The first involves a
functional analysis of how the normative power to promise operates, which seeks to characterize the way that
promising works in our practical or moral reasoning. Here the project is descriptive and conceptual. This first
step amounts to fixing the data set, and so does not distinguish Interest Theories from any other theorist who
tries to make explicit the explananda. The distinctive feature lies in the second step: Interest Theorists offer
normative explanations of promising in terms of interests (aspects of our well-being) or values.
Joseph Raz and David Owens offer different candidate interests to play this second-step role. In
early writings, Raz focuses on the interest we have in forging special bonds with others. That is, Raz argues
that we have an interest in treating the interests of others—our nearest and dearest, roughly—with special
priority over the interests of strangers and even ourselves. This interest, Raz asserts, grounds the power to
91

promise. In later writings, Raz emphasizes that an interest in controlling aspects of our lives grounds our power
to promise, though Raz does not explain the motivation for this shift in emphasis.
Owens claims a different interest, an authority interest, grounds the power to promise. An authority
interest is an interest in being able to place others under obligations. Because we cannot explain this interest
and the power it grounds without invoking the irreducibly deontic concept of an obligation, the authority
interest counts as what Owens calls a “normative interest.” And we have an authority interest, Owens argues,
even if other more familiar non-normative interests, such as welfare or trust, would be served by our having
it.
Without pretending to give an exhaustive survey of all possible or even all plausible Interest Theories
of promising, I will argue that of three influential versions discussed in the philosophical literature (early Raz,
later Raz, and Owens), the later Raz looks the least problematic. Raz’s early view, developed during the 1970s
and 1980s, tries to ground the power to promise in the intrinsic value of special relationships. I argue that this
view faces a dilemma. If interpreted as the claim that all valid promises constitute intrinsically valuable
relationships between promisor and promisee, we should doubt whether this claim is true. On the other hand,
if we interpret Raz as making the instrumental claim that valid promises are useful in facilitating intrinsically
valuable relationships, then we should wonder why the intrinsic value of relationships takes pride of place in
the relationship, rather than some other values also served by promising.
Indeed, in more recent work, Raz argues that promising is useful for facilitating not only special
bonds, social coordination, and trust but more generally control over our lives. I will argue that this value is
more plausible than emphasizing on special bonds because the value of control is in some sense more general
than the value of special relationships.
After considering Raz’s views, I turn to Owens’s attempt to justify the power to promise in terms of
a so-called authority interest, which is one of a number of alleged normative interests. I identify several problems
with Owens’s view, not the least of which is the dubious claim that we have sui generis interests in having
authority over others. I seriously doubt that we have this interest, and argue that wanting authority for its own
sake represents morally dubious power fetishism.
92

Finally, the paper returns to Promise Theories of contract law. Interest Theory might very well
provide a solid foundation for a Promise Theory of contract law, though it raises an important difficulty. The
structure of justification for the moral power of promises holds that values or interests justify that power.
Interest Theorists are therefore seemingly committed to a similar structure of justification for the legal power of
contract law. But this means the interests or values that ground the contract power might fundamentally
differ—or perhaps should fundamentally differ—from the interests or values that ground the promise power.
Indeed, at the extreme, contract law might be an autonomous legal power grounded or justified in none of the
same interests that ground or justify promise. This presents a problem given that an Interest Theory based
Promise Theory of contract would be forced to explain why promising plays any justifying or guiding role if
the interests or values that ground the moral power differ significantly from the interests or values that
ground the legal power. This difficulty undermines the core mission of Promise Theory: to use promissory
morality to justify contract law and guide its doctrines.
2 The Value of Promising: Special Relationships
Let’s begin with the works of Joseph Raz. His Interest Theory begins by analyzing the concept of a
promise, which he holds to be a voluntary communication of an intention to undertake, by that very
communication, an obligation to perform some act for someone (the promisee).
117
In addition to this bit of
conceptual analysis, Raz describes how promising actually operates in our practical reasoning—how promises
function. Promises operate as what Raz calls protected reasons. Protected reasons are reasons in favor of
performing some action that are also protected against reconsideration or against certain countervailing
reasons that might in normal circumstances operate to undercut or defeat those reasons.
To illustrate, the fact that a movie is good might normally count as a reason for me to form an
intention to go to the movie. But my changing my mind come movie time might count against following
through on that intention. Now suppose that I promise you to go to the movie with you. According to Raz, I
now have a reason to go to the movie—the fact that I promised. What’s more, the normative logic of
promising prevents my reconsidering the decision to go to the movie; promising involves shielding this

117
Joseph Raz, Promises in Morality and Law, 95 HARV. L. REV. 916, 927 (1982).
93

reason to go to the movie against certain other reasons for not attending, including mere changes of mind.
My changed mind might operate as a reason not to go in normal circumstances. But merely changing my
mind does not enter into my deliberations as to whether follow through on my promise once I have promised
to go.
I will not contest Raz’s conceptual analysis. Nor do I think that Raz’s re-description of how
promising operates in terms of protected reasons raises any immediate difficulties. We could translate this
protected-reasons talk back into more familiar vocabulary without, I think, losing anything in translation.
Normally, we understand promises as giving rise to obligations. The obligations are such that certain things—
emergencies, for example—trump or nullify those obligations (though it is controversial which). But it is
nearly universally acknowledged that the normativity of promising does not permit ignoring one’s obligation
simply because one has changed one’s mind or because a somewhat better offer comes along later. The idea
of a protected reason elegantly translates this more complicated description into the vocabulary of reasons.
But analyzing the concept of a promise and describing how promises operate is one thing;
establishing that we have actually have a pre-conventional moral power to promise, and answering Hume’s
Challenge in the process, is quite another. Interest Theories like Raz’s hold, in effect, that having the moral
power to promise depends only on whether a moral principle that would confer that power on us is
justifiable; having a moral power to promise, that is, depends just on whether having that power would be
valuable, regardless of whether there are any actually existing social practices or conventions that allocate this
power on us.
118
Whether we have the moral power to promise is a normative matter all the way down. Like
Agency Theorists, Interest Theorists are inclined to see Hume’s Challenge as a non-problem.

118
Raz rejects Practice Theory. See Joseph Raz, Promises in Morality and Law, 95 HARV. L. REV. 916, 927 (1982)
(“While conventional ways of promising make it easier to promise, they are not, however, a necessary
condition of making promises.”); see also id. 927 n.21 (“One may promise by communicating an intention in
unconventional ways; and one could promise even in societies - if such ever existed - that do not recognize
the practice of promising.”); Joseph Raz, Promises and Obligations, in LAW, MORALITY, AND SOCIETY: ESSAYS
IN HONOUR OF H.L.A. HART 214 (1977) (eds. P.M.S. Hacker and Joseph Raz) (“Another common belief is
that promises can exist only if there is a social practice to that effect. It is admittedly difficult to conceive of a
human society in which some form of promising is not practised. But it is imaginable.”). David Owens’s
official view is a version of Practice Theory, insofar as explains promising partially in terms of a social
convention or practice. But for reasons that will become apparent I will treat his view as articulating an
94

So what makes having the power to promise valuable? Raz’s initially argues that the value of that
power depends on the intrinsic value of special relationships or special bonds that we can enter into
voluntarily.
119
He is relatively clear about this, asserting that the value of the ability to enter into special
relationship depends on “the intrinsic desirability of forms of life in which people create or acknowledge
special bonds between them and certain other individuals”.
120
The ability to enter into these relationships
derives from the intrinsic value of these relationships. This suggests that the value of the ability or power to enter
into these relationships is itself intrinsic but derivative.
As for the relationships themselves, Raz never gives an explicit account of special relationships. He
evidently has in mind relationships like friendships and relationships with what one might call voluntary
associations, like one’s workplace, club, or other organization. All of these relationships are voluntarily
undertaken and are constituted by members that exhibit partiality towards one another.
121
And surely these
relationships, at least many of them, are intrinsically valuable. The ability to enter into these relationships
derives its value from the underlying intrinsic value of those relationships.
Notice, too, that making a valid promise necessarily involves creating a special relationship between
promisor and promisee, which is constituted by a binding promise. The promissory relationship, according to
Raz, is only one form of special relationship that exists on a “spectrum” along with friendships or
relationships between colleagues.
122
So if it is in general valuable to have the ability to enter into special
relationships voluntarily, it follows a fortiori that it is valuable to have the ability to enter into promissory
relationships voluntarily in particular. And because it is quite plausible that having the ability to enter into

Interest Theory in what follows. Ulrike Heuer also reconstructs Owen’s views in a practice-free manner. See
Ulrike Heuer, Promising – Part 2, 7 PHIL. COMPASS 842, 848-50 (2012).
119
Joseph Raz, Promises in Morality and Law, 95 HARV. L. REV. 916, 928 (1982) (“The moral presuppositions of
this conception of promising are the desirability of special bonds between people and the desirability of
special relations that are voluntarily shaped and developed by the choice of participants.”); see also Joseph Raz,
Promises and Obligations, in LAW, MORALITY, AND SOCIETY: ESSAYS IN HONOUR OF H.L.A. HART 228 (1977)
(eds. P.M.S. Hacker and Joseph Raz).
120
Joseph Raz, Promises and Obligations, in LAW, MORALITY, AND SOCIETY: ESSAYS IN HONOUR OF H.L.A.
HART 228 (1977) (eds. P.M.S. Hacker and Joseph Raz).
121
See JOSEPH RAZ, MORALITY OF FREEDOM 175 (1986).
122
Joseph Raz, Promises in Morality and Law, 95 HARV. L. REV. 916, 928 (1982).
95

voluntary, partial relationships is valuable (given the intrinsic value of these relationships), having the power
to promise must be valuable, too.
This argument is appealing but also too quick. To see why, consider this parody. Insofar as having
the ability to voluntarily form special relationships is in general intrinsically valuable, a fortiori having the ability
to voluntarily form partial relationships by slapping others intentionally is intrinsically valuable. But having this
latter ability is not intrinsically valuable. So either there is a mistake in the reductio or in the argument sketched
in the preceding paragraph.
Raz would argue that the parody fails. The argument fails not because intentionally slapping someone
does not give rise to a special relationship. Intentionally slapping someone does give rise to a special
relationship, the one binding a wrongdoer (who owes compensation to the victim) to his victim (to whom
that compensation is owed). The victim’s claim for compensation rightly receives, or should receive, a special
priority in the wrongdoer’s practical deliberations about what to do next. So if the parody fails, it fails for
another reason.
Perhaps the parody fails in the following way. We might try to distinguish wrongdoer-victim
relationships as outside the scope of special relationships that are considered intrinsically valuable for the
purposes of justifying the power to enter into special relationships voluntarily. After all, wrongdoer-victim
relationships are not wholly voluntary, since victims suffer involuntarily even when wrongdoers inflict that
suffering voluntarily. Applying the voluntariness requirement to every member of the relationship properly
restricts the class to those special relationships that are entered into or sustained voluntarily by all participants.
Our ability to enter into these relationships voluntarily is what is intrinsically (but derivatively) valuable. In
turn, the power to promise represents an instance of this ability, and thus is itself intrinsically valuable. Raz’s
argument is rescued.
Although this response is plausible, it paves the way to a new worry, which is that it suggests that a
voluntarily entered into relationship is intrinsically valuable simply in virtue of being voluntarily entered into.
This is implausible. After all, some special relationships lack intrinsic value even though they are voluntarily
entered into, while others are intrinsically disvaluable. My relationship with T-Mobile, which I entered into
96

voluntarily, lacks any intrinsic value. There is no value in the relationship over and above instrumental
concerns. The moment T-Mobile ceases to provide reliable cell phone service, the relationship loses all value;
the moment I cease to pay my bills, T-Mobile terminates my service, and any further relationship involves
their trying to obtain payment from me. Any value in our relationship is instrumental only. As for disvaluable
voluntary special relationships, consider abusive, manipulative, or degrading special relationships. It is hard to
see why these relationship are any more intrinsically valuable or worth preserving than the relationship that
between, say, a negligent wrongdoer and victim. In turn, it is hard to see why having the power to enter into
these relationships would be intrinsically valuable. So restricting the set of intrinsically valuable relationships
to the set of those relationships voluntarily entered into fails to restrict the set enough.
Perhaps Raz can further restrict the set. Perhaps we lack the moral power to make binding promises,
and hence to form special promise-based relationships, to the extent that those relationships would be
intrinsically disvaluable. Maybe this is an attractive feature, not a defect, of Raz’s position. Any promises that
would constitute intrinsically disvaluable relationships, on this view, would fail to be binding, and hence
would fail to constitute genuine promissory relationships. These relationships therefore could not be said to
be intrinsically valuable because they do not exist.
Even so, the T-Mobile example is a harder case. This special relationship lacks any intrinsic value. So
why should the moral power to promise make possible this special relationship, which is exclusively of
instrumental value? I find it difficult to accept the claim that any promise to pay T-Mobile money in exchange
for cell phone coverage constitutes an instance of intrinsically valuable relationship. True, having the ability to
enter into the relationship is instrumentally valuable, but that is not the nature of the justification currently
under consideration. Raz posits a deeper, non-instrumental relation between the value of the power to
promise and the intrinsic value of special relationships, positing that this relation is just a specific instantiation
of a more general one that holds between abilities to forge special relationships and the intrinsic value of the
relationships themselves. Whether those relationships are intrinsically valuable therefore matters.
I think we can state the concern another, more direct way. On Raz’s view about the value of the
moral power to promise, this value derives from the fact that, when validly exercised, that power gives rise to
97

a special relationship (a promissory one) that is itself supposed to be intrinsically valuable. But promissory
relationships, even when validly exercised, are not necessarily intrinsically valuable (e.g., T-Mobile). So
whatever value the power to promise has, it does not have it in virtue of the fact that exercising that power
instantiates an intrinsically valuable relationship.
At this point a defender of Raz will question my attributing to him the claim that the power to
promise is intrinsically value only if and to the extent that it gives rise to intrinsically valuable promissory
relationships. This simply can’t be his view. The power to promise must, if we are to be charitable to Raz, be
valuable even if some exercises of that power might fail to give rise to a valuable relationship or a may even
give rise to a disvaluable one. Analogously, our ability to reason is valuable even if nothing guarantees that
this ability will be put to good use. So might our ability to forge special bonds by declaring an intention to be
bound might be valuable in and of itself.
This, I think, gives rise to a dilemma. If the ability to forge special obligations is suppose to be
intrinsically valuable—valuable in and of itself—this looks very close to a viciously circular explanation of
promising, tantamount to the claim that moral agents have the power to promise because that very same
power is intrinsically valuable. This interpretation, moreover, belies the fact that Raz repeatedly claims that
special relationships—the special bonds—are the bearers of intrinsic value, not the power to promise itself.
The other horn of the dilemma denies that Raz offers a circular explanation, and instead tries to
explain the power to promise instrumentally. This perhaps makes more sense as an interpretation of Raz’s
views since it does not attribute to him a viciously circular explanation. Indeed, if the validly exercised power
to promise does not guarantee certain intrinsically valuable relationships; if his claim is just that the power
merely makes more likely intrinsically valuable relationships; or if the power is merely useful for promoting
intrinsically valuable relationships, then we must interpret Raz’s claim as an instrumental one. But to see the
second horn of the dilemma, I will first have explain why interpreting Raz’s claim as an instrumental
justification is problematic in its own right. I attempt to do so below.
98

3 The Instrumental Value of Promising: From Special Relationships and to Control
We left off with a response from a Razian defender. One might argue for a different interpretation of
early Raz. The question is not whether any particular promissory relationship may be valuable. Of course it
might not be. The question instead is whether having the power to form these relationships is valuable. And here
the justificatory relationship between the power to promise and special relationships becomes clear. Having
the ability to form promissory commitments to others is incredibly useful for entering into or managing
relationships that are intrinsically desirable. Having the ability to make promises to loved ones, for example,
might enable us to reinforce those bonds that may already tie us together. And where the bonds are tenuous
or non-existing, having the ability to promise—having the ability to bind ourselves to perform actions for
others—might help us to develop new intrinsically valuable relationships. As a happy side effect, this useful
tool—the power to promise—allows us to enter into useful special relationships that are not necessarily
intrinsically valuable, like the relationship that I have with T-Mobile.
All of this might be true. Promising is useful for managing relationships given that they are useful in
managing others’ expectations about our future conduct. But if the justification for the power to promise is
instrumental, then why should the intrinsic value of some special relationships deserve pride of place in this
instrumental justification? That is, if we are going to justify the power to promise instrumentally, then it is no
longer obvious why we should primarily justify the power to promise in terms of promising’s ability to forge
special bond. Sometimes promising forges valuable bonds, sometimes not; sometimes promising can help
facilitate pre-existing relationships, but promising is neither necessary nor sufficient. The ability of promising
to facilitate valuable relationships looks like just one of a number of important values that might be served by
promising. Having the moral power to promise also facilitates our interest in: coordinating conduct with
others, inducing others to rely on our representations about our future conduct, gaining accurate information
about others’ future conduct, and building reservoirs of interpersonal trust, primarily by habitually following
on our promises. And so on. But none of these interests is entirely co-extensive with the interest that we have
in forging special bonds with others.
99

Indeed, perhaps in response to this worry, Raz’s justification for the moral power to promise has
shifted significantly over the years, though he does not explicitly acknowledge the shift. His more recent work
de-emphasizes this putative interest in forging special bonds in favor of a control interest:
In the case of promises the value of the power is that it expands people’s ability to fashion their lives,
or aspects of their lives, by their actions. Through their promises they commit themselves to others.
… Promises, being commitments to others, facilitate cooperation, the forging of relations that
presuppose dependence, trust and joint actions, and more. For the sake of brevity I will refer to the
value of having these powers as the value of enhanced control (of one’s life), though a somewhat
different explanation of their value is required when the powers are held by institutions.
123

In this passage Raz does mention the interest we have in the “forging of relations” with others, yet he
also mentions the important interests we have in “facilitating cooperation,” “trust,” as well as gesturing
towards “more” interests that might be served. He collects all of these interests that might be served by
promising under the catchall “control” interest—i.e., the interest we have in having control over our lives,
emphasizing especially the control we have in deciding which interpersonal commitments we form with
others.
Notice the shift. The intrinsic value of special bonds in themselves no longer takes center stage. That
interest is but one of several others, which share the common feature of serving the over-arching interest in
control. Raz’s motive for this shift is unclear. Again, he never acknowledges that a shift has taken place. But
Raz’s attempt to ground promising in control nevertheless looks plausible, mainly because it is stated at a
fairly high level of generality. And forging special bonds is simply one way of exercising control over our lives
in general.
Indeed, the interest we have in having control over our lives is stated at such a high level of generality
that it is hard to dispute. Perhaps all moral rights and powers are ultimately justified by reference to this
interest! Still, despite this high level of generality, we might try to say more about what kind of control Raz

123
Joseph Raz, Is There a Reason to Keep a Promise?, in PHILOSOPHICAL FOUNDATIONS OF CONTRACT LAW 61
(Gregory Klass, et. al., ed. 2014).
100

has in mind, other than providing a list of things that might be affected by promising (cooperation, forging
relationships, etc.). Once again, Raz is elusive, but I am tempted to interpret Raz as holding that promising is
valuable primarily because it is useful in facilitating normative control—i.e., control over our normative
situation—rather than causal control or control that effectuates in-the-world outcome.
Why normative rather than causal control? If we understand control causally—an interest that we
have in controlling outcomes, events, or states of affairs—then it cannot be an interest sufficient to ground
the power to promise. Having the power to causally control requires a tendency for that power, once
exercised, to be causally efficacious once employed, since causal efficacy is needed to control outcomes,
events, or states of affairs. To see why, consider a world in which nobody recognizes a power to promise.
Promissory acts thus routinely fail in this world. But it would seem strange to insist in this world that one has
the power to promise, one that is justified and grounded by the interest in exercising control over outcomes.
It would be tantamount to insisting that one had the ability to love another person in a world where no other
person existed. Abilities depend on external states of affairs as well as internal ones. This observation would
not worry Practice Theorists like Charles Fried, since they suppose that a robust practice of promising is a
necessary condition for promises to bind. But this causation issue should worry a Razian who denies Practice
Theory.
This is not to say that causal control is unimportant to Raz’s position. When I solicit a promise from
someone to do something for me, I want this promise to actually motivate that person’s future action. But
causal control is, as a theoretical matter, of secondary importance. The reason we can bring about changes in
the world, causal changes, depends on agents’ respecting the various changes in the normative landscape.
Promises can, for example, reliably facilitate cooperation among agents only if others respect the reason-giving
force of the promise to begin with. Without that presupposition, using promises to reliably secure
coordination will be more difficult. This is why I interpret Raz as grounding the power to promise in its
usefulness in securing normative control rather than causal control.
Raz’s more recent view on what justifies the power to promise, that the power is useful insofar as it
secures our ability to exercise normative control, is quite plausible. Indeed, when we return to discuss
101

problems faced by a Promise Theory grounded in Interest Theory, I will assume something like Raz’s more
recent view is the standard bearer for Interest Theory. Before we get to that point, however, let me address
David Owens’s recent Interest Theory, which has garnered some attention and support. As will be clear, I
think Owens’s position is problematic in ways that Raz’s later view is not.
4 Normative Interest Theories: The Authority Interest and Beyond
David Owens does not quarrel with Raz’s analysis of a promise as a voluntary undertaking of an
obligation via a communicated intention to do so. His main disagreement lies in the interest invoked to justify
and ground the power to promise. To understand the disagreement, we first need to introduce Owens’s
distinctive idea of a normative interest.
Owens claims that we have interests that have an irreducibly normative dimension.
124
He calls these
normative interests. One such interest is the so-called permissive interest. We have an interest in being able to grant
others permission to do things. This interest, argues Owens, grounds the power to consent, given that
consenting essentially involves granting permission. We have this interest, moreover, regardless of whether
the ability tends to serve any of our more familiar non-normative interests, such as health, friendship, trust,
welfare, and so on. This suggests that even moral agents who would systematically misuse this power to the
detriment of their non-normative interests would still have a permissive interest.
Owens makes similar claims about promising. The normative interest that grounds the power to
promise, however, is the so-called authority interest. Authority is understood as a normative power to change
others’ obligations to us simply by communicating the intention to do so. So we have an interest, according to

124
I should first note how his official view differs from the version that I will present. Officially, Owens
provides a different type of explanation of promising from Raz in an important respect. For Owens, we have
the power to promise only if people generally recognize and respect the practice of promising. In short,
Owens’s official view is a version of Practice Theory. The motivation for Owens’s reliance on Practice
Theory flows from his rejecting the view that interests and values alone can ground normative powers, a view
that Raz and others accept. Given the problems with Practice Theory, and my ultimate (implicit) argument
that it is false in Chapter Four, I will not address the motivations for Owens’s reliance on it. Suffice it to say
that Owens does not think we have authority unless others generally recognize that authority. Others have
implicitly denied that his Interest Theory requires accepting Practice Theory. See Ulrike Heuer, Promising – Part
2, 7 PHIL. COMPASS 842, 848-50 (2012).

102

Owens, in being able to change others’ obligations to us simply by communicating the intention to do so.
And the interest cannot be understood except by reference to the irreducibly deontic concept of obligation.
The authority interest is thus a normative interest. And just as the case with the power to consent, we have
this interest even if having this power would not serve any non-normative interest.
This last sentence requires justification. After all, if our putative normative interests obtain only in
virtue of certain more familiar non-normative interest (like health, etc.), then these more fundamental, non-
normative interests ultimately ground the power, not the alleged normative interests that supervene or
otherwise depend on them. The authority interest would therefore be inessential to explain that power. So we
need an argument for the claim that we would have any normative interests over-and-above the non-
normative ones. More specifically, to substantiate Owens’s theory of promising, we need an argument for the
claim that we have an interest in having authority for its own sake.
To bolster this claim, Owens places a lot of weight on a famous case involving Maklay and his
servant.
125
Maklay purportedly promised his servant that Maklay would not photograph him. The servant
apparently believed that photographs captured souls. We are asked to suppose that Maklay could have
photographed the servant in his sleep, but that doing so, intuitively, would have been wrong because it would
break the promise. Breaking this promise is wrong, on Owens’s view, because it arrogates authority that
belongs to the promisee, Maklay’s servant. This example supposedly generalizes. We might imagine other
cases in which the promisee’s having the power to demand compliance with the promise would make no
difference to the promisee’s non-normative interests—health, welfare, and so on. Nevertheless the promisor
does something wrong by breaking the promise. If Owens is correct, then we have an authority interest over
and above any non-normative interests.
This is Owens’s view in a nutshell. Let me now raise a few problems, focusing on difficulties unique
to his view. En route to the first problem let me try to clarify one more point about Owens’s theory. Notice
that, in comparing Owens’s account with Raz’s, Raz tries to ground the promisor’s power directly in the interest
in forging bonds, and later on, the control interest. After all, the promisor chooses whether to form a

125
DAVID OWENS, SHAPING THE NORMATIVE LANDSCAPE 126 (2012).
103

commitment with another person or institution, subject to the promisee’s waiver of the performance right.
But promisees rather than promisors have the relevant authority to validly demand performance of the
promised action. So how does this authority interest of the promisee ground the power held by the promisor?
Presumably, indirectly. I say “presumably” because, surprisingly, Owens does not explain in any
sustained way how our authority interests ground our power to assign that authority to others. The quick
discussion in the passage below is the best we get:
[S]ince it is the promisor who makes the promise, this raises the question as to what interest
the promisor has in the matter. The answer is reasonably obvious: we all have an interest in
being able to satisfy the needs of others should we so wish. Where the promisee has an
interest in acquiring authority over the promisor, the promisor often has an interest in being
able to grant this authority, an interest in having the power to make promises. Furthermore,
the promisor has an interest in being able to grant this authority by declaration, by simply
communicating the intention of so doing. Why so? Because the promisee has an interest in
being able to acquire authority simply by communicating the intention of so doing, simply by
accepting the promisor’s declarations.
126

This passage is startling. First, after having argued at length in favor of the existence and import of
normative interests generally and the authority interests specifically, the passage suddenly posits a different,
non-normative interest that grounds the promisor’s power to promise: the interest in satisfying others’ needs.
The interest is non-normative because its content does not refer essentially to any deontic concept. But this is
odd. Owens is claiming that the power to make promises is itself grounded in a non-normative interest.
Doesn’t this concede too much?
Now Owens might insist that the promisee’s authority interest still plays an indispensible role in
explaining this power, since it is this interest that needs satisfaction. But once we have the interest in satisfying
other’s needs on the table, we might begin to look elsewhere for other non-normative interests in the vicinity
capable of replacing the authority interest in the explanation. Consider this first attempt. Perhaps what

126
DAVID OWENS, SHAPING THE NORMATIVE LANDSCAPE 146-47 (2012).
104

grounds the power to promise would be a putative interest in satisfying others’ needs, while what grounds the
power to acquire rights to performance would be the non-normative interest in securing the desired future
actions of others. Why wouldn’t this or some other non-normative interest in the vicinity do just as well as
the putative authority interest in grounding the power to promise?
In fairness, there are innumerable non-normative interests that one might try to use to explain the
power to promise, and Owens cannot be reasonably expected to entertain all of them. He does in fact
consider a range of non-normative interests, including interests that we have in coordinating behavior or
securing accurate information. And the brisk theory sketched in the previous paragraph, which focuses on an
interest in securing future actions, might be problematic for reasons I have not considered. But raising this
alternative aims to illustrate a worry rather than to provide an alternative Interest Theory. That is, once Owens
himself leans on a non-normative interest to complete his picture of the normative power to promise, this
raises the concern that Owens has still moved too quickly in dismissing all non-normative interests as
incapable of grounding the power to promise. Owens’s puzzling reliance on the putative interest in satisfying
others’ needs re-opens the door to this kind of doubt.
But raising an inchoate worry is not the same as providing a refutation or even a strong objection.
And perhaps I am making too much of Owens’s invocation of an interest in satisfying needs. After all, after
having mentioned it, he quickly adds that the power to make promise lies simply in one’s interest in having a
power to promise. Presumably the interest we have in having the power to promise counts as a normative
interest, thereby obviating the worry that Owens has opened the door to a non-normative-interest-based
account.
But this interest is itself puzzling. Ignore the fact that he appears to conflate this interest with the
interest in satisfying others’ needs. Grounding the power to promise in the interest we have in having the power
to promise looks viciously circular. But suppose it is not. If not, it is no longer clear what motivates Owens to
rely on the promisee’s authority interest. After all, why not simply claim that the power to make a promise is
one’s normative interest in having the power to make a promise? This seems to be far more straightforward.
105

Given the confusion generated by the passage, let me try to reconstruct what Owens might have in
mind. Recall the issue: given that the authority interest represents the promisee’s interest, what grounds the
power to promise of the promisor, who, after all, is empowered to make promises in the first place? The best
answer for Owens should be, I think, that the authority interest should likewise ground the promisor’s power,
albeit derivatively. Here’s how: Let x and y be moral agents. Ordinarily y has an interest in deciding whether
to do some range of morally permissible actions. But if x has an interest in having exclusive authority over y
regarding whether to do some action within that range (i.e., if x has an authority interest), then there must be
some way for y to cede that authority to x. Otherwise x would lack any way to acquire that exclusive
authority; x’s interest would be wholly frustrated, since both x and y would both claim exclusive and hence
mutually incompatible authority to decide whether to perform that action.
At this point Owens posited that we have an interest in satisfying others’ needs, or at least not
frustrating them. But this response raised worrisome questions. Instead, we might observe that, from
experience we know that that x can succeed in acquiring that authority over y. This means that y must have the
power to cede that exclusive authority to x (to require y to perform an action for x), presumably by assigning
or transferring that power to x. Perhaps having presumptive authority over one’s own actions grounds one’s
ability to assign it to others. But that power to assign it just is the power to promise. So the authority interest
does ground the power to promise, albeit indirectly.
This argument tries to avoid introducing new non-normative interests or seemingly circular interests
into the grounding explanation. It attempts to rely on no other interest besides the authority interest. But by
the same token the argument brings difficulties of its own. Notice that it arguably abandons Owens’s Interest
Theory in favor of an Agency Theory. Specifically, notice that the argument asserts, first, that we have some
measure of authority over our own lives, second, that the ability to assign authority to another person is
somehow an ingredient of this authority, and finally, that assigning limited authority over an action just is the
power to promise. These premises seem essential to the argument.
But if we accept these premises it is difficult to see why we need to appeal to normative interests at
all, as opposed to certain abilities—such as the ability to consensually assign things to others—that appear
106

intrinsic to moral agency. We can see this by simply stripping away all the interests-talk from the argument
above. What remains is simply a power to cede authority to others over certain decisions and actions, a power
that we have in virtue of (presumably) being moral agents that can do this kind of thing. The fact that
promising serves putative normative interests looks incidental. Our moral capacities to do certain things is
fundamental, not the normative interests.
So far we have focused on one potential source of problems for Owens’s theory: how it explains the
power to make a promise relates to the promisee’s putative authority interest. But suppose that these issues raise
no difficulties for Owens’s account. A more direct problem concerns the centerpiece of Owens’s theory: the
authority interest. Recall that, to show that we have this normative interest, Owens must show that we would
have an interest in having authority over others regardless of whether having this authority would tend to
serve any non-normative interest. In other words, we must have an interest in having authority over others
for its own sake.
But it is not obvious that we have this interest. Owens appeals to cases to bolster his claim. The
Maklay case, which is supposed to illustrate a case in which making and keeping a promise serves only the
authority interest, takes center stage for Owens. But as others have observed, he does not eliminate other
plausible interpretations of the case convincingly. Respecting the promise of Maklay’s servant also serves the
latter’s interest in controlling outcomes, asserting that failing to respect Maklay’s promise would not
undermine Maklay’s control.
127
Here Owens is correct to the extent that Maklay does not exercise control
over what happens to his soul, but is mistaken if we understand the relevant control as control over whether
his photograph is taken. The control-interest-based Interest Theory supported by Raz looks like a live one.
Moreover, Owens himself, in the passage above, suggests another non-normative interest: an interest
in satisfying others’ needs. Owens might reply that the servant does not need to have his promise kept. But
once we allow that we have an interest in satisfying others’ needs, why not strong desires—such as the
servant’s strong desire not to be photographed? The point of raising these alternative, non-normative

127
Alida Liberman, Book Review, 124 ETHICS 201, 203 (2013) (reviewing DAVID OWENS, RESHAPING THE
NORMATIVE LANDSCAPE (2012)); see also DAVID OWENS, RESHAPING THE NORMATIVE LANDSCAPE 147
(2012) (discussing Maklay case in connection with control interests).
107

interests is not to provide the best or only interpretation of Maklay’s case, but rather to put pressure on
Owens’s claim that his own interpretation is the only plausible one. Other non-normative interests would be
plausibly served by keeping the servant’s promise.
Indeed, I suspect that Owens tacitly depends on the impurity of this case to bolster his favored gloss:
that we have a normative interest in authority. But if we truly want to establish that we have this interest we
must try to do a better job of eliminating other potential interests that might cloud our judgments. Taking our
cue from G.E. Moore’s test for intrinsic goodness,
128
let us consider two worlds, one in which we have some
modicum of authority over another and a virtually identical world in which we do not. We should, moreover,
try to design the cases such that this authority would not serve any of our actual non-normative interests or
the non-normative interests of the person subject to that authority. Once we describe these worlds, if we
would judge that it is better to live in the world in which we have the authority over the other, then perhaps
we will concur with Owens’s judgment that we have an interest in authority in and of itself.
Consider the following case, which strives to abide by these constraints. Suppose we have two
worlds, w@ and w. Let’s stipulate that w@ is the actual world (hence the “@”), the world you currently live
in. Now let’s stipulate that w is just like w@ with one exception. You have the authority to impose an
obligation on someone in Kazakhstan, who you don’t know and will never meet, to drink a cup of tea at time
t. We can stipulate that this person doesn’t much mind tea, and in fact regularly drinks it. Indeed, further
stipulate that, coincidentally, t coincides with the precise time at which the Kazakh happens to drink tea every
day anyway. So the obligation to drink tea at t, should you impose it, truly represents no inconvenience. And
the Kazakh knows he is under this obligation when it is communicated to him—but he doesn’t know where
this obligation comes from and it makes utterly no difference to his routine. (How is this obligation
communicated? Via some loudspeaker.) Finally, stipulate that the overall happiness of the person’s life is not
very much affected; suppose that this fairly arbitrary, from-the-loudspeaker obligation, should you choose to
impose it, will happen only once during his lifetime. But he knows it is an obligation nonetheless. Finally,

128
See G.E. MOORE, PRINCIPIA ETHICA (1903).
108

assume that no one is compensated for the performance of the obligation, once issued, or for your
imposition of the obligation.
The case is strange for a number of reasons, and it is not altogether clear that I have managed to
eliminate all the possible reasons for the Kazakh’s not wanting to be subject to this arbitrary authority, even if
it visits no inconvenience on his life. But imagine that, somehow, this one-off obligation, if and when it does
issue from you, really makes no material difference to his life. He is not bothered being subject to this one-off
authority. Again, were he not under this obligation, and lived instead in w@, he would drink his tea in exactly
the same way per his daily routine. We are truly trying to imagine that everything else is equal in this Kazakh’s
life in w versus his life in w@.
Finally, the question: Would having this authority to impose this obligation to drink tea at t on the
Kazakh make you any better off in any plausible sense? Remember you obtain no material benefit in terms of
well-being. You get no money. You just have this one-off normative power to impose this arbitrary obligation
on this person. Whether it is a moral obligation makes no difference. Indeed, precisely because it is arbitrary
and capricious, it looks like it is arguably an immoral normative power to have over someone. But if Owens’s
view is correct, this should not matter: merely by having this power is, all else equal, better than not having it.
There might be moral constraints on the exercise of that power; maybe the right thing to do on the balance of
moral reasons is to give it up immediately. But having this power is, on Owens’s view, better for you despite
overriding imperatives to abandon it.
But this seems morally dubious. My life is not better off for having this authority over someone else.
Indeed, I am inclined to think that desire to have authority, in and of itself and for no further generally
legitimate end, would look morally suspect. Those who would, other things being equal, prefer to have the
tea-obligating power, not because it improves Kazakh’s life, not because it even makes them even happier by
providing some dim form of entertainment, but for no other reason than to simply have that power, smacks
of a form of power fetishism that morally virtuous agents ought to eschew. Authority is after all a form of
domination or subjugation, even though in many cases authority is nevertheless all-things-considered
justifiable. But desiring to normatively dominate another for its own sake looks morally pernicious. So the
109

claim that we have an interest in this form of normative domination for its own sake looks similarly morally
implausible.
5 An Interest-Based Promise Theory of Contract Law?
Let’s take stock. We have seen three versions of Interest Theory. They share a methodological
commitment to the view that establishing the existence of the moral power to promise is a matter of
identifying the interests or values that would best justify that power. The descriptive question of whether we
have the moral power to promise turns on the normative question of whether having this power would be
good or valuable. I have argued that David Owens’s version of Interest Theory is problematic, as was the
early versions of Joseph Raz’s Interest Theory. Raz’s later approach, which holds that the power to promise
justified because it gives us greater control over aspects of our lives, looks plausible.
It is time to return to Promise Theories of contract law. In what follows, I will assume in the
background Raz’s more recent approach, though the issues I raise might not necessarily turn on presupposing
his particular view. Let’s focus instead on a particular problem that arises in the context of moving from
theories of Interest Theories of promising to Promise Theory.
The first issue involves the question of what grounds or justifies a legal power to make and enforce
promises through law. Recall that Interest Theories typically use interests or values to explain moral powers.
When we turn to contract law and its recognition of a legal power to bind oneself, Interest Theorists seem
committed to the same structure of explanation and justification. But in the case of making a legally enforceable
promise, the normative power essentially differs from the moral power to promise: it is a legal power to bind
oneself to others. So in justifying or explaining why we ought to have the legal power to contract, relying on
the control interests or the authority interest is simply the start. We need also to point to the desirability of
vesting in citizens the legal power to bind themselves to others via contracts. And this requires additional
argument.
Because Raz has discussed this question, it is worth spending a moment evaluating his remarks on
contract theory, cursory though they are. As we will see, at the heart of his comments lies an un-argued-for
assertion: that contract law primarily supports pre-institutional practices of promising rather than initiating a
110

different kind of voluntary commitment. Although this section focuses on Raz, I also move beyond his
remarks and argue that certain structural features of Interest Theory complicate the attempt to justify and
explain contract law in terms of promissory morality.
Raz begins by distinguishing between two (non-mutually-exclusive) goals that a legal framework
might have: supportive and initiatory. The law plays a supportive function when it aims to protect pre-
institutional social arrangements and practices. Initiatory law creates a new normative framework that is not
rooted in pre-legal or pre-institutional social arrangements. An example of an initiatory regime includes the
laws that make limited liability corporations possible. Limited liability corporations simply do not figure in
our pre-institutional, pre-legal normative practices.
With this distinction in mind, Raz asserts that contract law mainly plays a supportive role, and in
particular, aims to bolster practices that facilitate voluntary commitments, especially promissory practices.
129

This provides the foundation for Raz’s normative views about contract law. Like any normative theory of
contract, Raz’s aims to justify contract law and provide some non-trivial guidance in designing its doctrines. As
for justification, the law is justified in part on the grounds that it protects valuable social practices of making
and keeping commitments. As for guidance, the doctrines that contract law takes should be evaluated by
reference to their ability to vindicate the law’s goal of protecting these practices. Raz writes:
Because the predominant purpose of contract law is to support existing moral practices, one
might expect that the conditions of the validity of contracts will reflect common moral
conceptions concerning the validity of voluntary undertakings. Doctrines such as the
formation of contract, frustration, mistake, fraud, duress, unconscionability, and others
based on public policy considerations might be expected to mirror common moral views.
130

Another principle plays an essential role in justifying and guiding the law, according to Raz: the Harm
Principle. The principle holds that the law’s coercive authority is justifiably used only to the extent that it
protects individuals from harm. Raz combines this principle with the prior assertion that contract law aims to

129
Joseph Raz, Promises in Morality and Law, 95 HARV. L. REV. 916, 933 (1982).
130
Joseph Raz, Promises in Morality and Law, 95 HARV. L. REV. 916, 934-35 (1982).
111

protect the practice of voluntary commitments, yielding the over-arching view that contract law protects
promissory practices by protecting individuals against harm resulting from misuse of those practices.
One concrete recommendation flowing from this combination of principles is that the only prima facie
justified civil remedy flowing from broken promises is reliance damages—i.e., damages that result when a
plaintiff detrimentally relies on the formation of a contractual obligation that is not fulfilled. Expectation
damages—which aim to give plaintiffs the benefit of the contractual bargain (even when no harm flowed
from a broken contract)—are thus prima facie unjustified.
This particular claim, that the Harm Principle recommends reliance damages rather than expectation
damages, has been contested.
131
But I do not want to focus on whether Raz correctly draws out the
implications of the Harm Principle for contract remedies. The more pressing issue is Raz’s claim that contract
law plays a primarily supportive role. The nature of this assertion is not initially clear. Is it an empirical or
doctrinal observation or a normative assertion? If it is empirical or doctrinal (based, I suppose, on an alleged
congruence between informal social practices and contract law of a particular set of jurisdictions), Raz gives
no support establishing that this congruence exists. Little in Raz’s discussion, in other words, suggests that
Raz cares much about whether contract law actually does play a supportive role as a matter of official doctrine
or empirical fact, unless he takes this to be obvious. But it is not obvious. And in context Raz’s claim makes
more sense as at least in part a normative one: that contract law ought to primarily support promissory
practices.
But Raz gives no reasons to think that we ought to accept that contract law ought primarily to protect
existing promissory practices rather than play, in his own words, a primarily initiatory role. Indeed, Raz concedes
that contract law sometimes might serve an “initiatory function” sometimes despite its primarily supportive
role.
132
This point is obvious. Contract law, like any other rich, complex and highly developed field of law,
should probably be prepared to answer to or accommodate goals and values beyond the main one that
motivates and justifies that field as a whole. Despite this concession, Raz gives no explanation, let alone a

131
See DORI KIMEL, FROM PROMISE TO CONTRACT (2003).
132
Joseph Raz, Promises in Morality and Law, 95 HARV. L. REV. 916, 935 (1982).
112

satisfying one, of why contract law must primarily bolster pre-existing promissory practices rather than
primarily initiate a new one. Raz’s explanation therefore contains a gap.
Why does this matter? If contract law serves primarily to initiate a new form of voluntary
undertaking rather than reflect or bolster pre-existing promissory practices, then this calls into question the
deep connection between promise and contract. One possibility is that contracts represent an autonomous
form of commitment, one perhaps triggered by promises but ultimately transformed into something essentially
non-promissory.
Notice, moreover, that the putative value that underwrites the power to promise, most plausibly, the
control interest, does not foreclose this possibility. The interest that we have in controlling aspects of lives
would be served by legal rules that conferred the power on us to make certain commitments legally binding.
But to the extent that other interests might justify the legal power to contract but not the moral power to
promise, the institution of contract law will be a fundamentally different kind of power, not readily
answerable to the moral norms governing the exercise of the moral power or the making or keeping of
promissory obligations.
This is significant. Without ruling out the possibility that non-promissory interests, whatever they
might be, justify the legal but not moral power of promise, this undercuts the ability to rely on norms that
govern the making or keeping of promissory obligations to justify or guide contract law doctrines. In short,
once we recognize this point we lack any independent reason to afford special weight to promissory obligations
or relationships that are bolstered or reinforced by social practices. This creates a prima facie problem for
Promise Theories that aim to justify and guide contract law in promissory morality.
How might contracts differ fundamentally from non-contractual promises? Others have discussed
this point at length in ways that I will not be able to here. But one obvious way that legally valid contractual
obligations differ from morally valid promissory obligations: valid contractual obligations are necessarily
legally enforceable, whereas the latter are not necessarily legally enforceable. This difference is obvious but
important. Because the law is a publicly funded and sustained institution, there is at least one entity relevant to
any prospective enforcement action that is not necessarily relevant to all promissory actions—namely, the
113

public. Accordingly, a natural thought is that the legal enforcement of promises must be answerable to the
public’s interests, what is sometimes called matters “public policy,” in ways that run-of-the-mill interpersonal
promises need not. In other words, in addition to (or perhaps instead of!) whatever interests are served by
having or recognizing a moral power to promise, Interest Theorists arguably must articulate a legitimate public
policy interest that justifies our recognizing a legal power to contract.
This gives us a new perspective on Raz’s suggestion and where it falls short. Rather than thinking of
the Harm Principle as an external constraint on the exercise of enforcement actions, applicable just because
enforcement involves state action, we should think of the public’s interest (in being protected against harm
produced by the misuse of promissory practices) as an essential part of the justification for contract law. The
difference is subtle. Rather than thinking that there is some ideal contract law that we could design, which is
subsequently chipped away and modified by the harm principle, the ideal would be in the first instance
informed by the harm principle.
We can now revise our worry about Raz’s suggestion: it is far from obvious that our interest in being
protected from harm is the only public interest that grounds the legal power to contract. Indeed, in the final
chapter, I will suggest that concerns about the proper scope of protection of property rights and the proper
limits of commodification also count among the legitimate public interests at stake when deciding whether to
enforce promises. And although the passage above suggests that contract law must accommodate other
values and “public policy” concerns, I will argue that the commodification debates are central to justifying
and limiting the scope of contractual rights and obligations, not merely ancillary to them.
In fairness, this is not a concern unique to Raz’s later Interest Theory. All Promise Theories will
eventually need to explain how or why promissory morality takes pride of place, or at least plays an essential
role, in justifying contract law; all such theories will have to grapple with the question of the scope and role of
public policy concerns beyond the enforcement of promissory rights in justifying and guiding contract law.
But Interest Theory’s structure of justification makes these issues particularly salient, which is why raising
these concerns now makes sense.
114

6 Summary and Conclusion
Promise Theories of contract law that are based on Interest Theories of contract law cannot, by
themselves, foreclose the possibility that other interests and values might ground and justify the legal power to
contract. More specifically, the legal power might be justified or grounded in interests or values that might not
play any role justifying or grounding the moral power to promise or which might even replace the interests or
values grounding the moral power to promise. Either way, this raises the possibility that the power to
contract is grounded in very different interests and values from the non-legal power to promise—but most
plausibly, public policy concerns of some sort. It is difficult to guarantee a place for promissory commitments
in contract law if one’s Promise Theory is grounded in Interest Theories. This is problematic—though of
course not necessarily fatal for the prospect of an Interest Theory based Promise Theory of contract law.
Ultimately, I will argue in the final Chapter that the way to find an essential place in the law of
contract for promissory morality is, ironically, by beginning with an instrumental justification for contract law:
one that sees contract law as a service that facilitates markets. Interestingly, this raises issues surrounding the
commodification of promissory rights and obligations. That is, in contractual transactions, promises become
converted into or are treated as a form of commodity. Understanding contracts as commodified promises
also explains, among other things, the central role that the doctrine of consideration plays in common law
countries: consideration, which demands an exchange for a contract to be valid, ensures that the state gets
involved only when a transacting parties treat a promise as a commodity—i.e., something fit for exchange on
the market. This also explains why some promises seem so obviously outside the jurisdiction of contract law,
such as promises in the context of friendships, or gratuitous promises: these promises are not or should not
be treated as commodities fit for exchange.
133

Finally, and importantly for the purposes of this dissertation, the contract-as-commodified promises
view provides the resources needed to motivate the robust Promise Theorist—the promise theorist who aims

133
I will also argue that viewing contracts as involving commodified promises provices argumentative
resources to Promise Theorists, assuming that commodities are a form of property and property rights
typically receive robust legal protection, including punitive damages and strong forms of injunctive relief for
certain violations of property rights.
115

to strengthen contractual rights and obligations in common law countries. If we understand commodities as
necessarily genuine property interests, and contractual promises as commodified promises, then there is a
strong prima facie case that these promises should be protected as other property interests are, including the in-
principle availability of punitive damages, strong injunctive relief, and strict liability. This provides the
normative force for the reforms that robust Promise Theorists like Seana Shiffrin are looking for. Again, this
is only a preview, and I will discuss these issues at length in the last Chapter.
116

Chapter Four: Contracts and the Assurance Conception of Promising
1 Introduction
Hume’s Challenge asks how it is possible to voluntarily assume moral obligations. In the first
Chapter we investigated one answer supplied the Practice Theorist of promising. In the previous two
Chapters we saw Moral Power Theories, which attempted to answer Hume’s Challenge directly by arguing
that we simply do have the power to impose on ourselves moral obligations by promising. Agency Theorists
attempt to ground this power in moral agency itself, while Interest Theorists try to ground this power in
certain interest or values.
My own view is that we should begin from the perspective of the promisee: we should begin with the
question of why it is wrong to break a binding promise and work our way backwards. The most influential
recent view that takes a similar approach comes from T.M. Scanlon, who proposes that breaking promises
involves violating a moral duty of fidelity. The idea is, roughly, that you have a duty to do X if you
intentionally cause another person, who wants you to do X, to form expectations that you’ll do X. The act of
making a promise triggers this duty. And breaking a promise is wrong because it involves failing to fulfill this
obligation.
134

This Chapter defends a modified Scanlonian view of promising. I begin by addressing an influential
objection from Niko Kolodny and R. Jay Wallace. They argue that Scanlon’s account is circular.
135
My first
goal will be to defend Scanlon against their attack. Because their objection and its target are complex,
pursuing this goal will require a lot of stage setting. But given that some philosophers believe that their

134

See T.M. SCANLON, WHAT WE OWE TO EACH OTHER 295-397 (1998).
135

More accurately, Scanlon recognizes the possibility of a circularity objection, and provides a response.
Kolodny and Wallace argue that the response fails. See Niko Kolodny and R. Jay Wallace, Promises and Practices
Revisited, 31 PHIL. AND PUBLIC AFFAIRS 119 (2003).
117

critique succeeds, or remain agnostic on the issue, showing exactly where the criticism goes wrong will be
worth the effort.
136

Despite my defense of Scanlon, there is something wrong with his account. My second goal will be to
explain why. To preview, incurring a promissory obligation to someone requires, on Scanlon’s view,
successfully assuring that person that you’ll do as promised. But this feature of the view exposes the theory to
counter-examples. More importantly, the counter-examples point to some important features of promissory
obligations that we’d lose if we were to accept Scanlon’s account. We’d lose, for example, the right to hold
others accountable for their promises, at least in an important class of cases. I point out these difficulties in
order to pave the way to a modified Scanlonian view, one that gives attempts to provide assurance a central role in
generating promissory obligations obligation, while at the same time avoids the counter-examples.
Finally, I turn to a potential argument against using a Scanlonian view to ground a Promise Theory of
contract. To preview, the argument is that right kind of reasons for complying with a promise, as implied by
Scanlon’s account, differ fundamentally from the right kind of reasons for complying with a promissory
commitment, such that we can no longer be confident that they are fundamentally the same kind of
commitment. If so, then grounding Promise Theory in a Scanlonian account is a non-starter. Although I am
inclined to reject this argument, I offer no refutation of it; the argument remains a potential source for
concern in trying to build a Promise Theory on a Scanlonian view.
2 What Must a Theory of Promising Account For?
Before explaining Scanlon’s theory, let’s list desiderata that he thinks any account of promissory
obligation should satisfy, and then explain what motivates them.
137
Scanlon holds that any such account
should:

136
See, e.g., Nicholas Southwood and Daniel Friedrich, Promises Beyond Assurance, 144 PHIL. STUDIES 261, 270
n.15 (2009).
137

Scanlon does not collect these desiderata in one place; they are implicit in his discussion in Chapter 7 of
Scanlon, op. cit. pp. 295-327.
118

(i) Make intelligible the fact that the primary value of a promise is that it enables a promisor A to
assure the promisee B that A will perform as promised (i.e., the theory must preserve the value of
assurance).
(ii) Show how promises provide a distinctive kind of assurance—namely, assurance that the promisor
has moral reasons to do what she has promised to do (i.e., the theory must explain the distinctive kind
of assurance promises provide).
138

(iii) Explain how promissory obligations arise upon making a promise and don’t pre-exist the act of
making a promise (i.e., the theory must show how making a promise generates obligations).
(iv) Explain why breaking a promise isn’t just a wrongdoing, but rather why breaking a promise
wrongs the promisee in particular (i.e., explain why promissory obligation is bilateral).
(v) Explain what characteristically makes breaking a promise wrong without mentioning any social
practices in that explanation (i.e., the theory must be practice-free).
Let’s see what motivates these criteria. The first one holds that the primary value of a promise is that
it provides assurance to the promisee that the promisor will do as promised. It isn’t initially obvious why we
should accept this. After all, we might think that the primary value of a promise is that it allows the promisee
to justifiably hold a promisor accountable. Or maybe the primary value lies in giving us grounds for relying on
others. Or maybe there are many values, none of which takes priority over any other.
Scanlon rejects these suggestions. He thinks that they don’t capture why promises are valuable when
the promisee (a) will neither rely on the promise nor be in a position to hold or (care about holding) another
accountable but nevertheless (b) finds the promise valuable. To illustrate, Scanlon offers the case of a
colleague, Harold, who asks you to promise not to reveal some past indiscretion of his.
139
Although you think
the indiscretion isn’t anything to be embarrassed about, you nevertheless promise. Scanlon adds that (a)
Harold will be moving far away and won’t be in a position to hold you accountable (or care about doing so)

138
See Kolodny and Wallace, op. cit., p. 137 (“It is distinctive of promising [according to Scanlon] that A’s
public performance itself reflexively provides B with a reason to believe that A will do X that is connected to
A’s concern to avoid wrongdoing…”).
139
Scanlon, op. cit., pp. 302-304.
119

and won’t in any plausible sense be able to rely on your promise. Harold nevertheless (b) sees something
valuable in your promise—namely, your assurance that you won’t reveal his secret. Scanlon thinks that this
case suggests that the primary value of a promise is that it assures the promisee that you (as the promisor) will
do what you say you’ll do.
We might press Scanlon on this. For one thing, Scanlon arguably over-generalizes based on Harold’s
case. We could just as easily imagine a case in which the primary value of a promise seems to lie in something
other than the assurance provided. (Again, we will revisit precisely this suggestion later on.) Harold seems to
have some interest that he wants you to help him protect—namely, an interest in not having his indiscretion
revealed. And he seems to have this interest regardless of whether he obtains an assurance from you. Nor
does Harold’s case rule out the aforementioned possibility that there may be several valuable features of
promises, none of which is necessarily present in every case of a valid promise, but some of which may typically
be present in most ordinary cases.
So we should worry about Harold’s case, both about how we ought to interpret the case and whether
it establishes that the value of promises lies in assurance. But let’s assume for now that Scanlon is correct, and
that the primary value of promises lies in assurance, even if we aren’t persuaded by Harold’s case. We can still
ask what exactly makes assurance so valuable. Scanlon denies that experiential benefits (such as reduced
anxiety) explain why promissory assurance is valuable in all cases. But as for what precisely makes assurances
so valuable, Scanlon isn’t altogether clear. He seems to suggest that the value of assurances lies simply in
justifying the expectations of the promisee, full stop. But for now, the important point is that accounts of
promissory obligation should, according to Scanlon, explain how promises preserve the value of assurance.
Desideratum (ii) holds that a theory of promises should also account for the particular kind of
assurances that promises provide. The motivation for (ii) can be understood in light of another worry we
might have about (i). If the value of promising is to provide assurances (according to (i)), we might wonder
why we even need promises. After all, there might be many ways to assure people without promising. Scanlon
offers an example. Suppose that someone swears on his honor as a member of the Sacred Brotherhood of
Reindeer that he’ll do something for someone. Scanlon claims that this oath isn’t a promise, yet it might just
120

as well provide some people with assurance that the oath-taker will follow through.
140
Given such cases, we
might ask what distinguishes promissory assurances from other ways we might assure people.
Scanlon’s answer is reflected in (ii). Promises provide a special kind of assurance. In particular,
promisors tacitly assure the promisee that the promisor will be motivated by moral reasons to do what he promises
to do. Contrast this with two other ways you might assure someone. Again, you might swear a Reindeer Oath,
or you might tell someone that you will do X because it’s in your financial interest to do X. Neither type of
assurance, according to Scanlon, is tantamount to an assurance distinctive of promising, since neither depends
on the speaker’s being motivated by moral reasons to do X. For all we know, the reasons for doing X in
compliance with the Oath are non-moral reasons. And reasons of self-interest aren’t moral reasons, either.
Scanlon takes these cases to illustrate how promissory assurances are supposed to be distinctive.
Criterion (iii) says that an adequate theory of promissory obligation should tell us how promises
generate new moral obligations. The motivation for (iii) involves a puzzle about promising. When you make a
promise, you thereby undertake an obligation to do what you’ve promised. But this seems mysterious to some
philosophers. Making a promise involves (at least in part) communicating an intention to be bound by an
obligation (to do as promised). But it’s not immediately obvious how your communicating an intention to
undertake an obligation makes it the case that you actually have an obligation, and in particular, an obligation that
you didn’t have prior to the communication. This puzzle has been called the “generative” problem.
141
A
successful theory of promissory obligation should solve it.
The next criteria, (iv) and (v), relate more directly to the question of what makes breaking promises
wrong. Criterion (iv) captures the idea that promissory obligations are owed to particular persons, and that
when we break promises we wrong someone in particular—namely, the promisee.
142
Breaking a promise isn’t like

140
Ibid., pp. 298, 306.
141
Seana Shiffrin, “Promising, Intimate Relationships, and Conventionalism,” Philosophical Review, Vol. 117, p.
482 (2008).
142
Scanlon, op cit., p. 316 (“Unlike an obligation to comply with a just institution that provides some of the
public goods, the obligation to keep a promise is owed to a specific individual who may or may not have
contributed to the practice of promising.”).
121

driving drunk, which is wrongful even in cases in which nobody is harmed. Any theory of promissory
obligation, according to Scanlon, must account for this bilaterality.
Criterion (v) holds that we should strive to explain what makes breaking a promise wrong without
mentioning any social practices in that explanation. To understand what motivates this desideratum, it’s
helpful to have a rough idea of what a practice-based account of promissory obligation looks like, and what
motivates Scanlon to avoid these accounts. Practice-based accounts hold that breaking a promise amounts to
flouting a widely shared, just, and useful social practice of doing X when we say things like, “I promise to do
X.”
143
Our obligations to keep promises are thus ultimately obligations to comply with the requirements of
that practice. In turn, breaking a promise is wrong because it involves free riding on or undermining this
useful and just social practice. Philosophers who hold this view defend practice-based theories of promissory
obligation.
144
They are practice-based because the explanation of why it’s wrong to break a promise mentions
the fact that there exist certain social practices, and that explaining promissory obligation is impossible
without mentioning these practices.
Scanlon offers two reasons for resisting practice-based views—and thus for accepting (v). He argues,
first, these views have difficulty explaining the bilaterality of promissory obligations (as described in (iv)).
145

After all, practice-based accounts suggest that it’s wrong to break a promise because breaking promises
involves undermining or unfairly free riding on a social practice of promising that we all benefit from. But
this suggests that breaking a promise doesn’t necessarily involve a wrongdoing to anyone in particular.
The second reason for rejecting practice-based views is that the wrongdoing at issue when we break
promises might occur regardless of whether there were any particular practices commonly associated with
making and keeping promises. To bolster this claim, Scanlon offers a state-of-nature case, in which we
voluntarily and intentionally induce someone to believe that we’ll do something for her, but then fail to

143

Ibid., pp. 295-96; see also, e.g., John Rawls, A Theory of Justice (Cambridge, Mass.: Harvard University Press
1971), pp. 344-50.
144

See Rawls, op. cit. pp. 295-96.
145
Scanlon, op cit., p. 316.
122

follow through.
146
Scanlon thinks that this would be wrong even though, by design, the case involves no
established practices of making or breaking promises. And notice that the wrongdoing in the case seems to
involve the same kind of wrongdoing at issue when we break promises. These two thoughts—about
bilaterality and the state of nature—motivate Scanlon to search for practice-free accounts of promissory
obligations. Desideratum (v) reflects this motivation.
3 Scanlon’s Account
Now let’s turn to Scanlon’s attempt to satisfy these criteria. He articulates and defends moral principles that
aim to satisfy these criteria and thereby provide an adequate account of promissory obligation. Chief among
these principles is a principle of fidelity or Principle F:
Principle F. If (1) A voluntarily and intentionally leads B to expect that A will do X (unless B consents
to A’s not doing so); (2) A knows that B wants to be assured of this; (3) A acts with the aim of
providing this assurance, and has good reason to believe that he or she has done so [i.e., provide
assurance]; (4) B knows that A has the beliefs and intentions just described; (5) A intends for B to
know this, and knows that B does know it; and (6) B knows that A has this knowledge and intent;
then, in the absence of special justification, A must do X unless B consents to X’s not being done.
147

Scanlon claims that F is a true moral principle, and relies on it to explain promissory obligations.
148

In what follows I’ll assume, as do Niko Kolodny and Jay Wallace, that F is true.
Notice how Principle F helps Scanlon satisfy the desiderata. First, unlike practice-based theories, the
principle doesn’t refer to any social practices at all. To the extent Scanlon can use F to explain why breaking
promises is wrong, he’ll have succeeded in giving a practice-free account why breaking promises is wrong.
Principle F, in other words, helps Scanlon satisfy desideratum (v).

146

Ibid., pp. 296-297.
147

Ibid., p. 304.
148

Ibid.
123

Additionally, because the duty described in F is owed to someone in particular (represented by “B” in
Principle F above), this satisfies (iv)’s bilaterality criterion. Also notice how F can be used to answer the
question, raised in (iii), of how the act of making a promise succeeds in generating a moral obligation. Normally,
the act of making a promise triggers the new obligation described in F—the obligation to do something for
someone else unless that person consents to your not doing it. The obligation didn’t exist before that act
because the very act satisfies each of the six elements described in F needed to trigger the duty.
To illustrate, by saying “I promise to water your lawn while you’re on vacation” under ordinary
circumstances, I thereby (1) voluntarily and intentionally lead you to expect that I’ll water your lawn, under
circumstances in which (2) I know that you want to be assured of this (suppose). I also say these words (3)
with the aim of providing this assurance to you, and I have good reason to believe that I’ve provided that
assurance. Moreover, (4) you know that I have the beliefs and intentions just described (suppose), and that (5)
I intend for you to know this, and I also know that you know it. Finally, (6) you know that I have this
knowledge and intent. If something like this analysis holds in all ordinary cases of promising, then making a
valid promise involves triggering that duty, and hence breaking one will involve violating it.
The way in which Principle F purportedly satisfies the remaining two assurance-related criteria will
eventually raise some deep difficulties for Scanlon’s account. Indeed, much of this paper largely concerns
Scanlon’s troubles with assurances. The desiderata, recall, relate to (i) the value of assurance that promises
supposedly provide, and (ii) the particular kind of assurance provided. Scanlon’s way of ensuring that
promissory obligations preserve the value of assurance is by weaving assurance into the very fabric of the
principle, as seen in elements (1), (2), and (3) above. Less metaphorically, in order to trigger F’s obligation, A
must succeed in assuring B that A will do X. This guarantees a necessary connection between promissory
obligation and assurance, since the only way we can get promissory obligation as an output is by providing
assurances that one will do as promised as an input. This is how his account satisfies (i)—i.e., how the
account preserves the value of assurance.
But satisfying (ii) seems more difficult. According to (ii), any theory of promissory obligation must
explain how promissory assurances differ from other ways one might provide assurance. Scanlon claims that
124

promissory assurances are, roughly, assurances backed by moral reasons. But we can see an immediate risk of
circularity. In order to generate a moral obligation from Principle F, a promisor must first provide assurances
to the promisee. And to provide the assurance distinctive of promising, the promisor must first have a moral
reason to do as promised. But the duty contained in Principle F is supposed to provide precisely this moral
reason! We seem ensnared in a vicious circle. Kolodny and Wallace exploit this circularity problem to mount
their own objection, which I turn to next.
4 The Circularity Objection and Scanlon’s Way Out
So far we’ve seen Scanlon’s desiderata and his initial attempt to satisfy them all using Principle F. But
Niko Kolodny and R. Jay Wallace (hereinafter, “K&W”) have mounted a sophisticated attack against
Scanlon’s account. Their objection is complex, and I won’t retrace all of their steps. I’ll only go far enough to
allow me to pinpoint where I think it goes wrong. The upshot of their objection, however, is the claim that
Scanlon must rely on the social practice of promising after all; or more precisely, that we need a hybrid view
that relies not only the principles Scanlon advocates but also social practices. Stated in terms of the desiderata,
K&W claim that Scanlon must give up desideratum (v) because his account can’t remain practice-free.
149

The apparent difficulty arises because Scanlon tries to combine two desiderata. First, according to
Scanlon, A (the promisor) triggers the duty described in F only upon providing some assurance that A will do
X. This doctrine, as we’ve seen, is weaved into F itself in elements (1), (2), and (3). Second, as we’ve also
seen, promises provide a special kind of assurance—namely, assurances backed by moral reasons to do as
promised. But since triggering the duty in F first requires assurance, then the assurance can’t be backed by a
moral reason—a moral obligation in particular—generated by F itself. So Scanlon must explain how A can
provide B with the right kind of assurances—those grounded in moral reasons—needed to trigger the
obligation of F.

149
Kolodny and Wallace, op. cit., 131-136.
125

As K&W rightly note, two kinds of reasons are ruled out. A can’t try to assure B based on a “prior
reason.”
150
This would run afoul of criterion (iii), which requires showing how promising generates new (and
hence not pre-existing) obligations. Nor can Scanlon rely on non-moral, practice-based considerations. This
would run afoul of (v), insofar as Scanlon thinks we should avoid practice-based theories altogether.
Scanlon is aware of the circularity risk. He concedes that, to avoid circularity, the relevant moral
reasons must come from somewhere besides the duty described in F. To that end, he invokes the following
principle of due care, or Principle D:
One must exercise due care not to lead others to form reasonable but false expectations about what
one will do when one has good reason to believe that they would suffer significant loss as a result of
relying on these expectations.
151

(We might modify D so that it doesn’t prohibit games involving bluffing. But for our purposes, let’s bracket
these worries and assume that D is fully adequate as stated.)
Scanlon’s way out of the circularity seems to be this. Given D, it would be wrong for A to create the
impression that he has “a settled intention of doing X” unless A in fact had a settled intention to do X. Since
we assume that A takes moral requirements seriously, and hence that violating D isn’t an option for A, B may
rationally infer that A in fact has a settled intention to do X. This licenses B’s further assurance that A will do
X. Assuming that the other conditions needed to trigger F’s duty obtain, A now has the moral obligation to do
X without the circularity problem.
As K&W point out, Scanlon’s reliance on Principle D involves a partial concession. After all, the fact
that A has a moral reason to do X is supposed to explain the distinctive assurance provided by promising. But
Principle D doesn’t give A reason to do X; instead, it gives A a moral reason to avoid carelessly inducing B to
believe that A intends to do X. This is a big difference. But this partial concession isn’t as problematic as it
initially seems, as K&W explain:

150
Ibid., p. 132.
151
Scanlon, op. cit., p. 300.
126

[A]ny successful account of promising will have to concede at least this much to the circularity
objection. In order for principle F to become engaged in the first place, there has to be some reason
for B to believe that A will do X that is independent of A’s being obligated by F to do X, but that is
not a prior or nmp [non-moral, practiced-based] reason in the above senses (that is, a reason that A
had prior to making the promise, or that is grounded in nonmoral practices). ... Once assurance has been
generated in this way, Principle F can become engaged, and A’s concern to avoid violating F can provide B with a
further reason for assurance that A will do X, through a kind of “moral multiplier effect” (p. 308); but it is
important that assurance initially be secured without assuming that Principle F requires A to do X.
152

Scanlon’s idea, it seems, is that F does end up playing a role in providing assurance, based on moral
reasons, that A will do X. But the initial assurance required to trigger that additional assurance depends on a
moral reason other than the duty described in F. Principle D plays the supporting role of providing the
(perhaps weak) initial assurance. Because we assume (as we have throughout) that A and B respond to moral
reasons, and because A would be doing something wrong by risking assuring B that A will do X unless A in
fact intends to do X, Scanlon can rely on D to jump start the moral obligation to do X provided by F.
Scanlon’s way out of the circularity is initially appealing. Not only does D show how assurances can
be provided in connection with distinctively moral reasons (as all promises allegedly must, according to (ii)), it
seems to do so without relying on any social practices or prior reasons.
5 Kolodny and Wallace’s Reply
Despite the initial appeal, K&W argue that Scanlon’s response won’t work. Their response has several steps.
First, they argue that in order for A to provide the assurance to B needed to trigger F, A’s assurance must be
backed by compelling reasons. Then they argue that the reasons Scanlon offers—reasons that depend on
D—aren’t compelling enough. Finally, they claim that the only compelling reasons available to Scanlon are
either practice-based reasons or prior reasons. But Scanlon can’t rely on either kind of reason. As a result,
Scanlon’s account won’t work. There is an alternative—to rely on moral reasons that depend in part on social

152
Kolodny and Wallace, op. cit., p. 137 (emphasis added).
127

practices—but that would defeat one of Scanlon’s main goals: to provide a practice-free account of
promissory obligation.
Let’s turn to the first step: the argument that the reasons needed to back up A’s assurances must be
compelling. K&W begin by reconstructing the chain of reasoning that is supposed to assure B that A will do
X:
(a) If A does not believe that he will do X, then his attempting to get me to believe that he will do X
violates D.
(b) Since he would not violate D, he must believe that he will do X.
(c) The fact that A believes that he will do X is a good reason to think that A will do X.
(d) Therefore, A will do X.
153

Based on this reconstruction, K&W begin by asking why, under (c), A’s belief that he will do X gives
B a good reason to conclude that A will in fact do X. Here’s a thought: A’s doing X (or refraining from doing
it) is under A’s direct control. So “[i]f A believes that he will do X, it must be because A has formed the
intention of X-ing, and this in turn might seem to make it reasonable for B to conclude that A will actually do
X when the time comes.”
154

K&W argue, however, that, to be assured that A will do X, B must believe not just that A intends to
do X, but that A intends to do X based on “compelling reasons.”
155
To see why, notice that A can sincerely
form at t 1 an intention to do X at t 2 but can decide to abandon that intention on a “whim” before t 2—i.e.,
before it comes time to do X.
156
Nothing rules out the possibility that the kind of intention A has at t 1 might
be flimsy and easily overridden at t 2. Given this possibility, the assurance needed to believe that A intends at
t 1 to do X at t 2 isn’t enough, all by itself, to justify the belief that A will do X at t 2; to justify the belief that A

153
Ibid., p. 139 (quotation marks omitted).
154
Ibid., p. 139-40.
155
Ibid., p. 140.
156
Ibid.
128

will do X at t 2, B must believe, or be assured, that A’s intention (at t 1) to do X will persist in the face of
countervailing reasons that might arise later on.
So what kinds of reasons count as compelling in this way? K&W consider and reject a range of
options that Scanlon might try to invoke, but I won’t rehearse this discussion here.
157
For our purposes, it
suffices to say that the only reasons that they think are strong enough to ensure that A’s present intention to
do X will persist are prior reasons—reasons that A has before he makes a promise—and practice-dependent
reasons. Prior reasons can be quite strong, but Scanlon can’t rely on them because making a promise is
supposed to provide a new reason for the promisor to do as promised (according to criterion (iii)). Nor can
Scanlon rely on practice-dependent reasons, like Reindeer Oaths, since he seeks to provide (v) a practice-free
account of promissory obligation. So in giving a reason for B to believe that A will do X at t 2, A can’t (from
Scanlon’s point of view), rely on any reason that depends on the existence of a practice. So Scanlon is left
with no way out—at least without abandoning the hope of giving a practice-free account of promissory
obligation.
6 A Reply for Scanlon
Recall K&W’s response to Scanlon. Principle D, at best, provides moral reasons to be assured that at
t 1 we presently intend to do X at t 2. But that’s not compelling enough, they maintain, to provide assurances
that we’ll actually do X at t 2—which is the assurance needed to trigger the duty of fidelity. After all, our
intentions—even if sincerely formed at t 1—might nevertheless change on a whim. So to assure promisees in a
way to trigger a promissory obligation, we need to find some other moral reason strong enough to provide
the right kind of assurance. D isn’t enough.
But this is where K&W go wrong. They don’t correctly apply Principle D, and accordingly, they
overlook how it really does help to ensure that A’s present intention at t 1 to do X at t 2 isn’t easily changed on a
whim. To see how, suppose that at t 1 A induces B to expect that A will do X at t 2, and does so by expressing
(at t 1) the intention to do X at t 2. Suppose further that A knows that B would suffer a loss as result of relying

157
See ibid., pp. 139-144.
129

on those expectations and that B’s expectations are reasonable. Finally, and importantly, suppose that A does
all of this despite knowing that A will allow any fleeting whim to alter that intention to do X at t 2 at any time between t 1
and t 2. Has A exercised due care in the way required by D? K&W are committed to saying yes—but they’re
mistaken. Consider the following case:
Symphony. Gustavo says that he will go to the symphony only if Phil does. Phil sincerely tells
Gustavo that Phil intends to go to the symphony, knowing that Gustavo will rely on Phil’s stated
intention, and knowing that Gustavo will incur some expenses as a result. But Phil neglects to
mention that his intention to go to the symphony is such that any fleeting whim will sway Phil to
abandon that intention. Later on, Gustavo relies on Phil’s statement and goes to the symphony,
thereby incurring expenses that he wouldn’t have otherwise incurred. But Phil no shows because he
decided to watch cartoons instead.
In Symphony, I think that Phil violates Principle D precisely because he negligently or recklessly allowed Phil to
rely on his intention, to that person’s detriment. Phil led Gustavo to form reasonable but false expectations about
what Phil will do when Phil has good reason to believe that Gustavo would suffer losses as a result of relying
on these expectations. And notice why Phil failed to exercise due care: he had reason to believe that Gustavo
would rely on his stated intentions, and that Gustavo would incur costs he wouldn’t otherwise incur. And just
as importantly, Phil had reason to believe that his intention to go to the symphony was such that it could
easily be changed on a whim, but he failed to mention this fact to Gustavo.
These last two facts—the fact that Phil had reason to believe that his intention was whim-sensitive
and that Phil failed to mention this to Gustavo—are crucial in explaining why Phil failed to exercise due care
as required by Principle D. Suppose that Phil had warned Gustavo that his intention was easily abandoned—
and that, although he presently intended to go the Symphony, it was the kind of intention such that any new
reason to avoid going to the symphony might dislodge that intention. Now suppose that Gustavo relied on
Phil’s claims. In this altered case, Gustavo would have no grounds for complaint. Phil would rightfully tell
Gustavo, “Look, I told you my intention was such that any whim could alter it!” By disclosing the whim-
130

sensitivity of the intention, Phil would have exercised due care in preventing Gustavo from detrimentally
relying on his whim-sensitive intention.
Notice how this helps Scanlon. K&W claimed that a promisee can’t be assured, despite Principle D,
that a promisor’s present intention to do X will be sufficiently robust so as to avoid a sudden whim to
abandon that intention. But Symphony suggests that this is mistaken. A promisee can be assured that promisor’s
present intention to do X will be sufficiently robust because, if it were not, then the promisor would be
knowingly or recklessly or negligently inducing the promisee to rely on that present intention in violation of D.
Because both the promisor and promisee are motivated to avoid violating moral principles (it is assumed), the
promisee can rest assured that the present intention to do X is sufficiently resilient in the face of whims to do
otherwise.
Still, K&W might argue that the assurance that A has a resilient intention to do X still falls short of an
assurance that A will do X. But Scanlon doesn’t need the triggering assurance to be strong enough to assure
that the promisor will do as promised. The promisee receives this stronger assurance once the obligation to do
X (as described in F) is triggered. After all, once the obligation is triggered, the promisee has a new reason to
think that A will actually do X: namely, the new moral obligation described in F. But the assurance needed to
trigger the stronger assurance might be weaker, like the match that starts the blaze.
This completes my defense of Scanlon. If the defense succeeds, then K&W have failed to show that
Scanlon’s account is circular, or that he cannot show that promissory assurances are underwritten by practice-
independent moral reasons, as opposed to practice-dependent reasons. The strong reason to keep the
promise to someone in particular is supplied by Principle F; but the morality-based assurances needed to trigger
that duty is supplemented by a duty of care to the person to whom the promise is directed—a duty of care
spelled out in D.
7 Problems with Assurances: Putative Counter-Examples
Despite the foregoing defense, I nevertheless think that Scanlon’s account falls short. My objection is
simple. To preview, notice that, according to Scanlon, we can incur promissory obligations only if promisors
successfully assure would-be promisees. Thus, failing to assure entails failing to incur a promissory obligation.
131

But, I’ll argue, we can make binding promises to others even if we cannot succeed in assuring them—a
possibility that Scanlon can’t account for.
Let’s turn to one set of problem cases. Again, A can make a binding promise to do X for B only if A
succeeds in assuring B that A will do X for B. Notice that people who have a reputation for being unreliable
can’t undertake promissory obligations because they can’t succeed in assuring others. This might seem
worrisome.
Indeed, Scanlon raises this worry himself. He asks us to consider the case of the Profligate Pal, who is
wholly unreliable, routinely makes promises that he can’t keep, and is generally understood to be flaky.
158

Suppose that, in the past, your Pal has borrowed money from you, promising to repay you, but hasn’t done
so, and that he gives you no reason to believe that he will. Despite this poor track record, suppose that the
Pal begs you for yet another loan, promising to repay. Suppose that you give him money but don’t expect to
be repaid; indeed, suppose that nothing that the Pal says could assure you that you will be repaid. According
to Scanlon, your Pal is incapable of providing you with the assurances needed to trigger the duty described in
Principle F, and hence can’t incur promissory obligations. Even if the Pal thinks he has a promissory
obligation to repay a loan, he doesn’t. The Pal has no duty to repay (at least no duty grounded in a promise),
even though he has promised to do so, and you have no corresponding right to demand repayment if the Pal
fails to repay (at least no right grounded in a promise).
At first glance, Profligate Pal seems problematic. Scanlon’s account implies that the Pal incurs no
promissory obligation to repay, since he can’t provide you with the needed assurances. But the Pal seemingly
should repay the loan if the opportunity arises. We can put this idea in terms of moral rights and duties.
Suppose the Pal wins the lottery. The Pal seemingly has a duty to give some of his winnings to you to repay
his loans, including his most recent one. Likewise, arguably you have a right to demand some of those
winnings. So the case seems problematic.

158
Scanlon, op. cit., pp. 312-314.
132

But Scanlon has a response. He claims that Profligate Pal simply doesn’t describe a central case of
promising; it’s one of a number of “impure cases” that don’t undermine his overall account.
159
And even if
we have strong judgments that the Pal has incurred an obligation to repay, Scanlon points out that although
this might be true, the obligation needn’t be one triggered by the promise. Perhaps the Pal owes, for example,
“a debt of gratitude.”
160
Scanlon adds that “what he has received is a gift, tactfully described as a loan.”
161

Scanlon’s response is that the Pal might indeed have a duty to repay the loan, but that duty isn’t a promissory
obligation.
Scanlon’s suggestion seems plausible. There are many cases in which (a) it seems reasonable to
believe that someone has attempted but failed to incur a promissory obligation and (b) this failure coincides
with the fact that the promisor seems incapable of providing the promisee any assurance. A young man tries
to promise his dying mother that he’ll take care of the remaining family even though the mother knows he
lacks the financial wherewithal. A drug addict attempts to promise his incredulous sister that this will be his last
hit. And many might find it equally obvious that the Pal isn’t obligated by his promise to repay the loan. In all of
these cases, arguably, (a) and (b) obtain. So it’s tempting to infer that agents sometimes fail to incur
promissory obligations precisely because the would-be promisees can’t be assured.
But there are also cases where it seems that agents can incur promissory obligations despite failing to
provide assurance. Consider the Profligate Plumber:
Profligate Plumber. Suppose you live in a small town with only one licensed plumber. You know
that he’s totally unreliable, and that when he says he’ll arrive on time, he will surely be late. One day,
you call the plumber to schedule an appointment. You say, “You know you’re not really punctual. I
want you to promise to show up at noon on the dot.” The plumber says, “I promise. I know I’ve got

159
Ibid., p. 312.
160
Ibid., p. 314.
161
Ibid., p. 314.
133

to do better and I will be there at noon.” You are wholly unconvinced that he’ll show up at noon,
and sure enough, he’s very late.
162

It’s natural to think that the plumber made a binding promise and that he broke it. Scanlon might try
to explain away this natural thought, too. He might use the same strategy that he used in responding to
Profligate Pal. Maybe the Plumber’s obligation to show up on time isn’t a promissory obligation, but rather a duty
incurred by virtue of his role as a plumber. That is, he has certain standing duties to his clients by virtue of his
being a service provider, such as a duty of punctuality. The Plumber would have these duties regardless of
whether you tried to secure a promise from him. The case is arguably impure, too, and thus doesn’t raise any
unique concerns not already addressed in response to the Profligate Pal.

But Scanlon’s strategy seems less plausible as applied to Profligate Plumber. To see why, consider an
alternative version. Suppose that, instead of promising to show up at noon, the Plumber says, “I realize that
you want me to arrive at noon. And I’ll try my hardest to make it. After all, as a service provider, I have a duty
of punctuality. But I can’t promise. I’ve had trouble in the past keeping my promises, and I don’t want to
make any more promises that I can’t keep. So don’t be disappointed if I can’t make it.” Sure enough, the
Plumber shows up very late.
In this alternate version, I think that the Plumber might have done something wrong by showing up
very late, insofar he has violated his putative duty of punctuality. But the Plumber’s conduct seems less wrong
now. The only salient difference is that the Plumber expressly declines to promise. But you aren’t assured in
either version that the Plumber will be punctual. The best explanation for the additional wrong that occurs in
the first version is attributable to the Plumber’s breaking a promise. And obviously breaking a promise
presupposes having made a binding one. Yet Scanlon’s account wouldn’t allow for this possibility in this
case.
163

162
Andrei Marmor suggested this case to me.
163
Southwood and Friedrich, op. cit., p. 272, similarly offer Soccer Game. The story stars young Daisy and her
unreliable father, who never attends her soccer games even though Daisy desperately wants him to. One day,
her father sincerely promises to show up to Daisy’s next soccer game. But Daisy knows better than to get her
hopes up. When game day rolls around, Daisy’s father—as expected—fails to arrive. Instead, he goes out
134

Still, perhaps not everyone shares my view that the second version of the Plumber case is less wrong.
Some might think that Scanlon’s strategy to explain away Profligate Plumber works here, too. So let’s consider
another case, which I think is particularly difficult to explain away:
Incurable Pessimist. The Pessimist suffers from a peculiar psychological quirk that causes him to
have very low expectations about people in general—regardless of whether those low expectations
are warranted by the evidence. In other words, to his cynical mind, the world is populated with, in
effect, nothing but Profligate Pals and Plumbers. So he’s incapable of being assured that anyone will
do what she promises to do. But the Pessimist still makes and accepts promises. Each time a promise
is kept, however, he is very, very surprised—on par with your surprise in the Pal’s repaying his loans.
It seems, to me, possible to make binding promises to the Pessimist. But, according to Scanlon, we
can’t, since incurring a promissory obligation to someone B is possible only if B rests assured. But the
Pessimist cannot rest assured due to his mental quirk.
The normal way that Scanlon might respond is by pointing to some other obligation in the vicinity to
explain why we might think the person who attempts to promise might still be bound to perform. But this
route isn’t readily available here. Notice how Incurable Pessimist differs from Profligate Pal or Plumber: The Pal is
your friend and probably owes you a debt of gratitude. This makes it plausible for Scanlon to claim that his
duty to repay is grounded in obligations other than promissory ones (such as obligations grounded in
friendship or gratitude). The Plumber isn’t your friend but he is a professional service provider, and thus
arguably owes special duties to his clients. By contrast, the Pessimist is pessimistic to family and strangers
alike—regardless of whether they’re actually reliable. But it seems implausible that there will always be a special
duty in the vicinity ready to stand as a substitute for the promissory obligation. This makes it much harder to
explain away the case. So although I take Profligate Pal and Profligate Plumber to be deeply problematic for
Scanlon, Incurable Pessimist is problematic and more difficult for him to explain away.

drinking with friends. The case doesn’t appear to raise any special concerns apart from Profligate Pal. That is,
Scanlon’s responsive strategy—if it succeeds—would seem to succeed equally well against Profligate Pal and
Soccer Game. Scanlon could say, for example, that the dad is obligated to go because he has an ongoing duty to
support his child—but the obligation isn’t a promissory one. For similar doubts about Profligate Pal, see Shiffrin,
op. cit., pp. 486-89.
135

That said, not everyone may see Incurable Pessimist as problematic. The Pessimist’s distrusting attitude
might simply prevent him from maintaining certain special relationships. It is difficult to imagine, for
example, that he can have genuine friendships given his universally distrusting nature. And if this seems
plausible, why should we think that he is capable of entering into the relationship that holds between
promisor and promisee, one that might likewise presuppose trust?
This raises an interesting issue, a full treatment of which would require investigating the role of trust
in promissory obligation, as well as how promising contrasts with special relationships like friendship.
Although I cannot take up this project here, I agree that trust is very important, if not essential, for certain
relationships. But I’m not convinced that trust is essential for a promisor to bind herself to a promisee, for
reasons that will become clearer once I defend my positive proposal in Section 8. In any event, there are
independent reasons why we should think that the Pessimist can be a promisee (despite his distrusting nature),
and that the Plumber and Pal can make binding promises to us (despite our low expectations of them). I
discuss these reasons below.
8 Some Theoretical Costs of Scanlon’s Account
My case against Scanlon would be weak if it consisted solely of putative counter-examples. Scanlon’s
theory, like many philosophical views, faces challenging cases. And there’s always room to argue that one’s
philosophical account is revisionist. After all, asking any philosophical theory to conform to everyone’s
judgments may be unreasonably demanding.
But the cases just presented point to deeper problems. First, as we saw in the cases of the Pal and the
Plumber, Scanlon’s theory relieves flaky and unreliable people of abiding by promises that they attempt to
make. It’s difficult to think of another moral duty that works this way. My duty to avoid telling injurious lies
doesn’t abate after I tell sufficiently many of them. Presumably, whatever one’s theory of promising, that
theory should avoid implying that one’s having a proven reputation of failing to comply with promissory
136

obligations can effectively render those obligations less stringent or non-existent.
164
This is theoretical cost
number one—which is that Scanlon’s account Rewards Bad Behavior.
But the problem runs deeper. You don’t even actually need to be flaky and unreliable; having a
reputation for being flaky and unreliable suffices to deprive you of the ability to incur promissory obligations:
Son of the Profligate Plumber. One day the Profligate Plumber retires. His son—Reliable
Plumber—decides to move to the small town and take over the family business. However, due to
nasty and unsubstantiated rumors (which comes with the territory when one lives in a small town),
Reliable Plumber is plagued by an unwarranted reputation for being even less reliable than his father.
One day, you call the Reliable Plumber and ask him to promise to arrive at noon. Sure enough, he is
on time, and you are shocked.
According to Scanlon, although Reliable Plumber attempted to make a promise, he failed, since,
given his reputation, you can’t be assured that he’ll perform as promised. This seems problematic. Scanlon’s
account makes promissory obligations too sensitive to reputations—even when they’re false. So not only are
actual flakes prevented from incurring promissory obligations, those with unwarranted reputations for being
unreliable are likewise prevented from incurring promissory obligations. This seems like cost number two:
Scanlon’s account Dignifies Undeserved Reputations.
But there is a more problematic implication of Scanlon’s view. Recall the Pessimist. Due to an
unfortunate quirk, he’s incapable of being assured by anyone. We might wonder why he would even bother to
make and accept promises from anyone. My claim is that the Pessimist would still have good reasons to value promises
despite his chronic inability to be assured that anybody will follow through with their promises, and that Scanlon’s account
would unfairly deprive the Pessimist of these reasons. In particular, Scanlon’s account implies that the
Pessimist cannot rightly hold people accountable for their promissory obligations because they can’t incur any to him. That is,
even though the Pessimist isn’t capable of being assured, we should be troubled by the fact that Scanlon’s
account deprives him of something valuable: the right to hold others accountable for failing to follow

164
See also Shiffrin, op cit., p. 489 (“Why should a public history of prior breach make a contemporary
breach morally insignificant?”).
137

through on their promises. I maintain that it’s deeply problematic that the Pessimist is deprived of this right
despite his inability to be assured that they will follow through.
Let me say more about the right I have in mind. The right to hold someone accountable might be
spelled out in terms of Strawsonian reactive attitudes, such as being able to resent others justifiably for failing
to follow through on their promises. The Pessimist also loses, perhaps more importantly, the right to
justifiably demand either performance (of that which is promised) or compensation for losses that he incurs as a
result of promisors’ failing to follow through. However the right to hold another accountable is specified, the
Pessimist loses that right. This is a problematic implication of Scanlon’s view because it seems deeply unfair
to the Pessimist.
But it’s not just the Pessimist who loses out. Potential promisors—even genuinely flaky ones like the
Pal—do as well. This is because the ability of a promisor to bind oneself to another by making a promise is also
valuable. Suppose that I’m the Profligate Pal. Also suppose that it’s in my own interest to do X for you—that,
if I were rational, I would do X for you. Now suppose that if I merely intend to do X for you, the likelihood
that I’ll actually do X hovers around 40%. However, suppose that, if I can effectively confer a right to you to
hold me to a demand to do X for you, then the likelihood that I’ll do X increases to 51%. Often the
likelihood that I’ll do what I rationally ought to do increases by making a binding promise to you. So it’s not
hard to see how the ability to undertake obligations might be valuable to the promisor as well as the promisee.
Maybe the fact that one might lose the ability to bind oneself by one’s promises is simply another
reason to avoid being flaky. Indeed, this loss might even offset the worry that Scanlon’s account Rewards Bad
Behavior. That is, even if the Pal is rewarded in a sense, the reward comes with costs—namely, losing the
ability to bind oneself by one’s promises. But notice that, since Scanlon’s account Dignifies Undeserved
Reputations, a perfectly reliable agent likewise loses this ability. All of this is problematic for Scanlon in ways
that aren’t easily explained away.
9 A Scanlonian Way Forward
Given these difficulties, let me propose a new principle that will avoid them:
138

Principle F*. If (1) A acts with the aim of assuring B that A will do X, and (2) A has good reason to
believe that B knows that A acts with the aim of assuring that A will do X; (3) A intends for B to know that
he or she is acting with the aim of providing this assurance, and knows that B does know it; then, in the
absence of special justification, A must do X unless B consents to X’s not being done.
The most important revisions to Principle F are italicized. To understand what motivates the
revisions, recall the idea that originally motivated F: that inducing someone B to believe and expect that one
will perform an action can, under certain circumstances, trigger an obligation to perform. The problem with
this idea was that it seems possible to make binding promises even when we fail to change others’ beliefs or
expectations about whether we’ll perform. Thus the challenge: to find a plausible principle to cover these
cases, in a way that somehow remains faithful to the idea that changing others’ expectations can give rise to
obligations to perform as a result.
Principle F* overcomes that obstacle. Rather than requiring that B believe or expect that A will
perform, F* requires instead that A aim or try or attempt to induce B to form this belief or expectation. This is
reflected in element (1). But acting with the aim of assuring isn’t enough. Suppose that I tell you that I’m
going to attend your party in a language that you don’t understand. This isn’t enough, it seems, to give rise to
an obligation to attend. The obligation mentioned in F* is an obligation to someone B. So it seems natural
that B should play some role, even a relatively passive one, in generating the obligation. Our minds must meet
in some way.
The remaining conditions explain how our minds must meet. Suppose that, even though you and I
speak different languages, in context it is clear that I am trying to assure you that I’ll attend your party.
(Perhaps I am pointing to the time and date indicated on your invitation while smiling and giving you the
thumbs up.) So (3) is satisfied: I intend you to know that I am providing assurance, and you know that I’m
trying to assure you that I’ll attend. Now suppose that you start nodding, which I correctly infer means that
you understand me to be trying to assure you that I’ll attend. I now have (2) good reason to believe that you
know my intentions. I claim that in such circumstances I incur an obligation to attend the party. This party
139

scenario is thinly described. But Principle F* nevertheless seems to deliver the right results, suggesting that F*
is at least prima facie plausible.
We will continue to test the principle in the next section. But let me acknowledge an initial worry:
that F* makes incurring moral obligations too easy. Roughly, all that is required is my attempting to assure
you that I’ll do something, and that you and I both know that I am doing this. That’s it. Unlike F, F* omits
any reference to what B wants—or even what is in B’s interests. As a result, I can incur an obligation to you to
attend the party even if you don’t particularly care whether I attend, or even if you’d prefer that I not attend.
This concern—that F* makes incurring obligations too easy—can be fleshed out from the
perspective of A and B, respectively. From A’s perspective, we might worry that, by making incurring
obligations too easy, F* is too burdensome on A. But it is difficult to see why this would be so. True, some of
us are quick and careless in committing to social engagements, which suggests that some of us are quick and
careless in incurring moral obligations. But F* suggests that we just need to be careful in doing so. This hardly
seems unreasonably demanding.
We might also worry that Principle F* imposes too many burdens on B, insofar as A might incur
obligations to B even in cases where B does not want A to perform. Suppose, for example, that you really
don’t want me to attend your party because I’m boring. We might think that this should affect my obligations
to you; that in cases where B neither wants nor has an interest in A’s performance, A should not be obligated
to perform. But this thought would be mistaken, especially given our prior observations about the importance
of promissory obligations to promisors as well as promisees. Our ability to bind ourselves to others is an
important way to motivate us to do what is in our own interests. A principle that made A’s obligation to
perform wholly contingent on B’s interests or desires would significantly impair A’s ability to bind herself in
this way. The fact that F* avoids this implication is thus a good result.
But if we are still concerned about cases in which people incur unwanted obligations to us—cases in
which you want me to stay away from your party—then there are already solutions built into Principle F*. B
can simply consent to A’s non-performance. Additionally, the principle involves A’s obligation to B to
perform, which suggests that B has a correlative right that A perform. Thus, B can simply waive that right.
140

But if A still insists on performing despite B’s waiver, this unwanted burden isn’t a problem with Principle F*
so much as a problem with life. Sometimes life bequeaths us with gifts that we’d rather not have.
10 Why Principle F* Satisfies All the Criteria
Recall the criteria that an account of promissory obligation should satisfy, according to Scanlon:
(i) Make intelligible the fact that the primary value of a promise is that it enables a promisor to assure
the promisee that they will perform as promised (i.e., the theory must preserve the value of assurance).
(ii) Show how promises provide a distinctive kind of assurance—namely, assurances that the
promisor has moral reasons to do what he promises to do (i.e., the theory must explain the distinctive
kind of assurance promises provide).
(iii) Account for how promissory obligations arise upon making a promise and don’t pre-exist the act of
making a promise (i.e., the theory must solve the generative problem mentioned earlier).
(iv) Explain why breaking a promise isn’t just wrong, but rather why breaking a promise wrongs the
promisee in particular (i.e., explain why promissory obligation is bilateral).
(v) Explain what characteristically makes breaking a promise wrong without mentioning any social
practices in that explanation (i.e., the theory should be practice-free).
Given the objections I’ve raised against Scanlon’s view, we might add a new criterion:
(vi) Satisfy (i) through (v) without assuming that assurances must be prerequisites to incurring binding
promissory obligations.
The motivation for (vi) should be clear at this point. Scanlon’s account is prone to counter-example
because it fails to satisfy (vi), insofar as promissory obligations obtain only if the promisor succeeds in
assuring the promisee. But this is mistaken, since incurring promissory obligations should be possible in many
cases in which the promisee is not assured.
Principle F* satisfies the criteria. F* satisfies (iii) through (vi) straightforwardly. F* answers (iii), the
generative problem, since the duty doesn’t arise unless and until A acts with the aim of providing assurances in
such a way that enables B to know that A is trying to do so. The principle also captures (iv) the bilateral
nature of promissory obligation. The duty described in F* is a duty that A owes to B to do X on account of
141

trying to assure B that A will do X for B. F* also satisfies (v) because it’s practice-free: the principle neither
incorporates by reference nor presupposes any social practices.
165
And since F* strikes out language that
make the duty described in F contingent on each party having certain assurances or expectations of
performance, F* satisfies (vi). After all, the new principle eliminates the very feature of Principle F that gives
rise to concerns about circularity and counter-examples. Trying to provide assurances, and knowing that the
other person is trying to provide assurances, are required to trigger F*’s obligation, not succeeding in doing so.
It’s less obvious that Principle F* satisfies (i) and (ii). But I think it does. Recall the criteria: a theory
of promising should try to (i) explain why the primary value of a promise lies in its ability to provide
assurances, and in particular, (ii) show how those assurances are backed by moral reasons. In suggesting that
Principle F satisfies (i), Scanlon assumes that, in order to establish that the primary value of promising lies in
the assurances promises provide, a theory must show how every genuine promise necessarily or invariably
generates assurances. But this assumption isn’t well motivated. The primary value of knives, if knives can be
said to be valuable at all, is their ability to cut. But that doesn’t require that every knife must necessarily have the
ability cut things. Dull knives and ceremonial knives and decorative knives might lack this ability, and might
thereby be peripheral cases of knives, but that doesn’t mean that they’re not knives.
Likewise, we don’t need to assume that, in order to generate an obligation to a promisee, every
promise must necessarily succeed in assuring that promisee. We need only assume that each promise represents
an attempt to assure the promisee. What’s more, we can tell a sensible story about why promises ordinarily,
though not necessarily, succeed in providing assurances. Promises communicate that the promisor acts with the
aim of providing assurances, and in most ordinary contexts promises simply succeed in this aim by virtue of
succeeding as communications. (There’s nothing very special in this, any more than the fact that someone
asserts that p ordinarily gives us reason to think that that person also believes that p, even though there are non-
ordinary cases where this might not be true.) This renders intelligible the connection between assurances and

165
This doesn’t mean, however, that one can’t trigger a duty by invoking a social practice (like Reindeer
Oaths). Rather, the principle doesn’t rely on social practices essentially, any more than Scanlon’s original
Principle F does.
142

promises: promises aim to assure, and normally they succeed in assuring by virtue of succeeding in those aims.
And since the main aim of a promise is still to provide assurances—whether or not they always succeed in
that aim—it still gives us a way to make good on the claim that the primary value of a promise lies in
assurance. And this seems plausible: it seems that every binding promise involves such an attempt to assure
the promisee. Thus, the present account of promising satisfies (i) because promises still, on this account, have
the primary point, function, or aim of providing assurances.
So much for (i). Things are more complicated for the distinctive-assurance criterion, (ii), depending
on how we interpret it. On one interpretation, (ii) is implausible and should be rejected. If Scanlon thinks that
a theory of promising must explain why all binding promises necessarily provide assurances, this is a mistaken
desideratum that no theory can satisfy, since it rests on a false premise. That is, for reasons already discussed
at length, it’s not true that all binding promissory obligations involve assuring the promisee. So we should
look for a more relaxed interpretation of the criterion.
Here’s one suggestion: We should construe (ii) as holding that, if promisors succeed in providing
assurance to the promisee, then the promisee is assured that the promisor will be motivated by moral reasons
to comply with the terms of the promise. This construal of (ii) is wholly consistent with the possibility that
the promisee won’t actually be assured. And assuming that this construal is fair, the present account of
promissory obligation satisfies (ii) on grounds similar to those relied on by Scanlon. The moral reasons that
support the assurance that someone will perform as promised, if in fact the promise is assured, are supplied
by background principles like the principle of due care (Principle D). And attempts to assure by the promisor
implicitly rely on these background principles. But in the event that the promisee isn’t assured, this doesn’t
nullify the obligation generated by Principle F*. That obligation obtains once the requisite attempt to assure is
made under the other relevant conditions.
So far I’ve argued that F* satisfies the original Scanlonian desiderata, plus another one added to
ensure that the revised principle doesn’t run afoul of the counter-examples. Now I’d like to return to those
examples to see whether F* returns the correct verdicts in these cases—namely, that one should be able to
make binding promises in all of them.
143

Turning to Profligate Plumber, each of the prerequisites is easily satisfied. By promising to arrive at
noon, the Plumber (1) acts with the aim of assuring you that he’ll arrive at noon, and both you and he know
that he is acting with this aim, thereby satisfying (2) and (3). The same goes for Reliable Plumber. Thus, in the
absence of special justification (which there is none), the Plumbers make binding promises and hence are
morally obligated to arrive at noon unless you consent otherwise. Profligate Pal is also straightforward under
F*. The Pal promised to return the money, thereby trying to assure you that he will, and both you and he
know that he is trying to assure you in this way. He has therefore incurred a promissory obligation.
I admit, however, that people might have conflicting views in response to Profligate Pal. Some might
agree with Scanlon that, although the Pal tried to incur an obligation to return the money, he instead received
“a gift, tactfully described as a loan.”
166
But this characterization is mistaken. Privately characterizing the loan
as a gift doesn’t make it one. Scanlon fails to distinguish one’s attitudes towards an obligation from its
existence. One may try to construe taxes as gifts to the government in order to make payment easier to
swallow. But this does not by itself transform the obligation into a gift.
Likewise, treating the Pal’s loan as a gift might be aptly described as a plan to never enforce the
obligation or to forgive the debt if asked. But none of this nullifies the promissory obligation on the part of
the Pal unless these plans are executed. So I think F* renders the right verdict in Profligate Pal. The Pal has a
promissory obligation to repay his debt. Accordingly, my Scanlonian account doesn’t Reward Bad Behavior
or Dignify Undeserved Reputations. The ability to incur promissory obligations doesn’t depend on whether
the promisee can be assured, since assurances aren’t a prerequisite.
This leaves Incurable Pessimist. Scanlon made it impossible to incur promissory obligations to the
Pessimist, which robs him of remedial rights and also robs would-be promisors of the ability to bind
themselves to him. To avoid this result, Principle F* would have to render it possible to incur promissory
obligations to the Pessimist. And it does. When someone makes a promise to the Pessimist, he thereby tries
to assure him that he will perform as promised. Assuming that this person and the Pessimist both understand

166
Scanlon, op. cit., p. 314.
144

the person as trying to provide these assurances, the obligation contained in F* is triggered. So in cases where
the Pessimist doesn’t consent to non-performance, it will therefore be possible for the Pessimist to be a
promisee.
In sum, Principle F* easily handles the cases that were problematic for Principle F.
11 Some Objections
One might still have worries about Principle F*. One worry—a worry already broached—is that the
principle makes it Too Easy to incur binding obligations. Another is that it may be Too Hard. Here is one
putative example of the former. Suppose that I assure you that I’m going to punch you in the face, and you
and I both know this.
167
Surely I don’t incur an obligation to do so—quite the contrary. So if F* implies that
I’m bound to punch you in the face, then this would be very worrisome, and F* would indeed make it Too
Easy to make binding promises.
But the face-punching case, and others involving promises to perform immoral conduct, are ruled
out by the proviso, found both in F and F*, that duties of fidelity attach only in the absence of special justification.
Cases involving immoral promises are simply cases in which a special justification obtains: that a promise
proposes conduct that is itself morally wrong normally suffices to show that there can be no obligation to
perform it. To be sure, there is more to be said on the scope and content of this special-justification
condition, and many interesting borderline cases. But F* clearly avoids the implication—or has the resources
to avoid the implication—that immoral promises are binding.
Perhaps other cases show that F* makes incurring binding promises Too Easy. Suppose that a son
promises his dying mother that he will take care of the family even though he lacks the financial wherewithal.
I suggested that this is a case in which we might, quite reasonably, think that the son hasn’t made a binding
promise. Yet it’s also a case in which all of the elements of F* have been satisfied. After all, in promising, the
son tried to assure his mother that he’ll take care of the family, the mother (we can stipulate) wants the son to
do this, the son also knows that the mother wants this, and so on. Absent special justification, it seems

167
I thank an anonymous reviewer for raising this point.
145

(according to F*) that the son has incurred a binding obligation. If one has the strong intuition that the son
hasn’t incurred any binding obligation, then this case spells trouble for Principle F*.
There are several responses. First, I’m not uncomfortable at all with the thought that the son has
made a binding promise. So I don’t see this as a problem case. But suppose you disagree. Notice that, second,
deathbed promises pose puzzling cases for any account of promising. So the alleged difficulty F* faces
wouldn’t be unique. Finally, one might focus on the proviso that the obligation is binding absent special
justification; maybe one special justification for non-performance is the death of the promisee. In any event, I
really don’t see the deathbed promise as an insurmountable problem. Perhaps Principle F* makes it Too Easy
to incur an obligation. But the deathbed case doesn’t illustrate the point convincingly.
Another worry is that Principle F* makes it Too Hard to make a binding promise. Suppose that A
makes a lying promise to B to do X. A good result of any theory of promising, I take it, is that a lying
promisor is bound by her promises even if she has no intention of fulfilling them. Besides the wrong of
intentionally misleading someone, lying promisors commit another wrongdoing, which is breaking a promise.
Arguably F* doesn’t allow lying promisors to be so bound. After all, the lying promisor has no intention to
follow through on his promise.
But this objection misreads Principle F*. The principle does allow lying promisors to be bound.
Element (1) requires that the promisor act with the aim of assuring B. The lying promisor needn’t intend to do
X. Instead, he merely aims to provide B with assurance that A will do X. Lying promisors typically do aim to
provide this assurance. That’s how lying promisors manipulate others: by giving them false assurances and
getting others to rely on those false promises. So if Principle F* does make it Too Hard to make binding
promises, cases involving lying promisors don’t illustrate why.
12 Concluding Remarks on Promising
In the next section I will turn to some observations about the prospects for trying to use the
modified Scanlonian approach to provide a moral foundation for contract law. But let me conclude the
preceding discussion with some general remarks about what we want in a theory of promissory obligation.
One way to provide a theory of promissory obligation is by satisfying all the desiderata listed above—and
146

adding additional ones. Indeed, I’ve done this by adding a sixth criterion. But another approach—one
pursued by Kolodny and Wallace—is to lower the bar by subtracting from the laundry list. Kolodny and
Wallace reject, in effect, desideratum (v), claiming that we cannot explain promissory obligation without
mentioning social practices. But Kolodny and Wallace’s positive proposal—which hasn’t been discussed at
any length here—doesn’t address my sixth criterion. So even if my positive proposal is not ultimately
satisfying, the alternative proposed by Kolodny and Wallace is not waiting in the wings; they still need to
explain how we can incur promissory obligations to others when we don’t manage to convince them that
we’ll actually perform.
But we might nevertheless suspect, along with Kolodny and Wallace, that the list of desiderata needs
pruning. For my part, I think that (ii) might be problematic in ways not fully explored in this paper. The idea
that promises provide a distinctive form of assurance arguably lacks sufficient motivation. Scanlon’s example
of the Reindeer Oath, for example, was supposed to help motivate (ii). The Oath purportedly showed that
one could trigger obligations like those embedded in F or F* without thereby making a promise. But it’s not
entirely clear that the Oath isn’t a promise—and indeed, it’s difficult to think of cases in which the speech acts
that trigger those duties couldn’t plausibly be described as promises. So the quest for distinctiveness might be
ill advised—though I remain agnostic, even though the question of whether we ought to be motivated by
moral reasons to keep our promises arises again when investigating whether my account can provide a moral
foundation for contract law.
13 A Foundation for Contract Law? Reasons for Compliance
In this section, I want to consider an argument against using a Scanlonian view to ground a Promise
Theory of contract law. The argument tries to use one of the theoretical commitments of Scanlon’s view as a
lever to pry apart contract and promise and show that they are fundamentally different kinds of commitment.
Although I am inclined to reject this argument, I offer no decisive objection against it; the argument remains
a potential source for concern in trying to build a Promise Theory on a Scanlonian view.
The argument goes like this. On a Scanlonian view, to make a promise either suggests that one has
moral reasons apart from the promissory obligation itself to abide by the terms of one’s promissory commitment.
147

In the case of Scanlon, a family of principles including principle D provided these reasons. But regardless of
whether you accept the specifics, Scanlon does suggest that there are right reasons to comply with one’s
promissory obligations, and that these reasons are primarily other-regarding moral reasons. The fact that
promissory obligations create moral reasons for compliance is one of the things that are supposed to
distinguish promissory obligations from other forms of commitment.
But herein lies the potential problem. Arguably, there are no distinctively right reasons for complying
with one’s contractual commitment. One either does or does not abide one’s contractual terms. The fact that
one complies with those terms solely on the grounds of financial self-interest does not count as a moral strike
against one’s character. But the fact that the sole reason that one keeps a promise is because it is in one’s
financial self-interest, rather than for any other-regarding or moral reason, does (arguably) count as a strike.
One is keeping one’s promise for the wrong kind of reason.
To continue with the argument, let us add another assumption, which is a principle that allows us to
individuate kinds of commitment. Consider two forms of commitment, P and K. A principle of individuation
holds as follows: If the set of right-kind reasons for complying with P differs from the set of right-kind reasons for complying K,
then P and K are different kinds of commitment. And if P and K are different kinds of commitment, then we cannot
assume that the norms that rightly govern C prima facie also rightly govern K.
We are now in a better position to see a problem for trying to ground Promise Theory in a
Scanlonian theory. If moral reasons are necessarily included among the right-kind reasons for keeping a
promise, but are not necessarily included in the right-kind reasons for keeping a contractual commitment,
then these are fundamentally different kinds of commitment. Accordingly, we cannot assume that the norms
that rightly govern promises prima facie rightly govern contracts. And that would be bad news for trying to
ground a Promise Theory of contract law on a Scanlonian conception of promising.
I am not entirely persuaded, however, that the right reasons for complying with a mere promise
differ from those for complying with a contract. But at this point I should be more precise on what exactly
the objection is before responding to it. Here are the most plausible interpretations of what it means to say
that the “right reasons” for complying with contract and mere promise differ, respectively:
148

1. Self-regarding reasons permissibly motivate compliance with contracts but not compliance
with mere promises.
2. The best reasons for keeping one’s promises, but not necessarily one’s contracts, are moral
reasons.
3. Moral reasons (morally) must figure among the reasons motivating compliance with mere
promises but not compliance with contracts.
4. Moral reasons (morally) must be the sole or primary reasons motivating compliance with
promises but not contracts.
I doubt all of these claims. It seems too demanding and implausible to suggest, as (1) does, that self-
regarding reasons may not permissibly play any role motivating compliance with promises. I keep a dinner
date with a friend in part because I am motivated by moral reasons to keep my word, such as the reasons that
ground principles L and D—i.e., respectively, my duty not to have my friend incur losses in reliance on my
representations, such as wasted time, and my duty to exercise due care in my representations to others about
my future conduct. But surely I also have self-regarding reasons motivating my compliance with my promise:
I will enjoy eating tacos, having great conversation, and drinking a delicious margarita. Indeed, these might be
very strong reasons motivating my compliance. But it seems implausible, as (1) suggests, to hold that these
additional self-regarding reasons somehow improperly motivate my keeping the promise.
Perhaps self-regarding reasons are not wrong reasons to keep one’s promises; perhaps they are
permissible motivating reasons but are not, as (2) suggests, the best reasons. Let us grant this for the sake of
argument. Why couldn’t we say exactly the same thing about contracts? It seems correct—or no less
plausible—to say that the best reasons for keeping a contractual commitment are moral reasons, the same
reasons that ideally motivate compliance with mere promises. Someone who keeps her contractual
commitments because, say, “it is the right thing to do” seems to act more morally admirably than someone
who keeps the same commitments simply because it would be too financially costly to back out. This
suggests, in turn, that it is better, if not best, to keep one’s contractual commitments for moral reasons. For
this reason (2) seems implausible to me.
149

But perhaps (3) is more plausible, since it seems to capture our daily experiences with contractual
commitments. As indicated earlier, we might not feel compelled to keep our contracts by any sense of moral
duty; we might be entirely motivated by financial self-interest to comply with contractual obligations, such as
contracts with phone-service providers. And this seems, at first glance, morally unproblematic. By contrast,
keeping a mere promise without being motivated whatsoever by moral reasons appears (arguably) more
morally worrisome.
But we should have doubts about this alleged asymmetry. Even if we typically do not, as a descriptive
matter, expect people to comply with contractual commitments for moral reasons, this does not necessarily
mean that it would be morally inappropriate to hold people to these expectations. If the service provider
insisted that the only reason it keeps its contracts is profit, we might admire the candor while at the same time
not unreasonably being morally repulsed or at least worried by the amorality of the thought. At a minimum,
if the reader does not share this view—that contractual compliance ought be motivated in part by moral
reasons—then arguably we need to revise our own attitudes about the reasons that ought to motivate
contractual compliance. It hardly seems like too tough a bullet to bite.
Similar revisionism is called for with respect to (4), at least if we accept the premise that moral
reasons ought to be the primary reasons motivating compliance with mere promises. Revisionism is called for
if we habitually think that moral reasons may permissibly play no motivating role, and certainly not the
primary role, in securing compliance with contracts. Perhaps moral reasons should play this central role.
Someone seeking to ground contract in this assurance conception of promising might have a reason to accept
this kind of revisionism and thereby reject (4) as well.
Of course we might think that what makes (3) and (4) mistaken is that they are simply wrong about
the right reasons for complying with promises. I broached this idea at the end of Section 10, when I floated the
idea of rejecting desideratum (ii)—namely, the idea that what distinguishes promissory commitments from
other commitments is that the former convey that they the promisor has moral reasons to keep her promise.
I observed that this claim is not obviously true. If I make a promise to someone in circumstances that make
clear that I will suffer reputational harm if I fail to keep it, then we might all know that the primary
150

motivation or even exclusive reason for keeping my promise, in this circumstance, is to avoid reputational
harm. My making the promise fails to communicate that I have moral reasons to keep a promise even though in
fact I do have them. Nor is it obvious to me that keeping my promise for reputational reasons would be
impermissible reason, even if this were the main motivating reason, for keeping my promise. Indeed, the very
fact that I made the promise in these circumstances might succeed in generating greater assurance, and would
be more valuable, than making a promise that risks no reputational harm.
So far I have sketched a response suggesting that self-regarding reasons and moral reasons might
permissibly motivate compliance with both contracts and promises, and that they may therefore count as
right reasons for compliance with both kinds of commitments (at least on the most plausible interpretations
of the right-reasons objection). I also presented an alternative response, which supposed that moral reasons
for compliance do not necessarily figure among the right reasons for complying with mere promises. These
are obvious incompatible responsive strategies. But are they even necessary? There is another responsive
approach worth considering, which is to reject the premise that just because two kinds of commitments are
fundamentally different, this means that one kind cannot depend on another.
To see why, consider this example. At least in some Western cultures, a committed, sexually exclusive
relationship between betrothed is a prerequisite for marriage, a committed, sexually exclusive relationship
between two people is a prerequisite for engagement, and a committed non-sexual friendship between two
people is a prerequisite for having a committed sexually exclusive relationship those people. If we view these
interdependent relationships grounding each other, it looks as though marriage ultimately depends on non-
sexual friendship. But surely the reasons for being motivated to comply with the demands of a non-sexual
committed friendship will not fully coincide with the reasons for complying with the demands of a married
relationship. Still, many of the reasons for complying with the demands of a committed friendship will count
as reasons for complying with the demands of marriage. Marriage brings with it additional and often weightier
reasons for complying with its expectations and commitments, reasons that even include the prospect that
one’s spouse may have legal recourse for systematically failing to live up to those demands.
151

Despite the possibility of legal recourse, to deny that friendship (ideally) grounds marriages simply on
the basis of these differing sets of reasons looks like a dubious inference. Similarly, although the reasons to
comply with a contractual commitment include the prospect for legal enforceability, this is—plausibly, at
least—an additional reason that leaves the existing reasons for complying with promissory commitments
intact. If this is correct, then we can ground contractual promises in mere promises, even though it would not
be rational to be motivated to comply with promises on the grounds that they are legally enforceable (since
they are not).
As it turns out, this last claim—that legal enforceability simply provides an additional reason that
leaves the existing reasons to comply with mere promises in place—has been disputed. Dori Kimel argues
that the prospect of legal enforceability signals a certain level of distrust, which functions to keep
relationships detached.
168
This is very different from mere promises, which, due to their non-enforceability,
operate to foster trust when complied with. Aditi Bagchi argues that legal enforceability either displaces or
obfuscates the reasons for complying with mere promises.
169
This fundamentally alters the normative character
of these commitments, making a promise theory of contract, which aims to ground contractual obligation in
promissory obligation, a non-starter given their extreme differences.
I will not delve any deeper. The final chapter will try to justify the sense in which contract law
involves the enforcement of genuine promissory obligations, yet still manages to explain the sense in which
enforced promises are nonetheless normatively different in ways relating to the role that promises play in
markets and commodification. Suffice it to say that it is not obvious that the legal enforceability of a promise
either does or does not fundamentally alter the nature of the commitment, rendering the norms that govern
promissory obligations prima facie inapplicable to those governing contractual ones. Presumably some moral
norms govern interpersonal commitments qua interpersonal commitment, regardless of whether they are
purely promissory or contractual in nature. One must try, for example, to execute one’s interpersonal
commitments in good faith no matter the nature of the interpersonal commitment—whether promise, at-will

168
DORI KIMEL, FROM PROMISE TO CONTRACT (2003).
169
Aditi Bagchi, Separating Contract and Promise, 38 FLA. ST. L. REV. 709 (2011).
152

employment agreement, insurance contract, marriage, or something else. This duty does not vary across
commitments. But it is not entirely clear that this will be enough to give Promise Theory of contracts what
they try to do, which is to give price of place to promising in its justification of contract law.
14 Conclusion and Preview
So far the first four Chapters of this dissertation have focused on particular, plausible, and influential
theories of promissory obligation. Although my evaluations of these theories have been mostly critical, this
Chapter mostly defended my favored account of promissory obligation. But I have also argued that, each of
the four theories that I have investigated, including the one I favor, faces a relatively theory-specific difficulty
in providing a foundation for a Promise Theory of contract law. In the case of Practice Theory it was a
difficulty in accounting for the normative priority of the informal practices of promising over formal practices
of contract law; in the case of Agency Theory it was the Irrevocability Problem, which concerns how Agency
Theory can account for the irrevocability of promissory obligations, and in turn, contractual ones; and Interest
Theory lacked any theory-internal explanations for why the interests or values grounding the power to
promise should be identical to those that ground the legal power to contract.
In this Chapter, I raised an argument, which I called the right-kind reason argument, which focused
on the import that Scanlonian accounts place on the moral reasons for complying with a promissory
obligation. If moral reasons are not necessarily among the right kind of reasons for complying contractual
commitments, and if they are necessarily among the right kind of reasons for complying with promissory
obligations, then perhaps these commitments are so different that we should not expect promissory norms to
guide or justify contractual ones. Although I am inclined to think that this argument does not succeed, I do
not answer the worry decisively, either.
At this point I want to move in a more constructive direction. I want to bracket the critical question
of difficulties that accounts of promissory obligation face in grounding Promise Theory. Instead, I want to
address some more fundamental, stock objections to Promise Theory. After all, if these objections hit their
mark, and if there are independent reasons for thinking that a Promise Theory of contract law will not work if
grounded in the leading theories of promissory obligation, then maybe we should abandon Promise Theory.
153

I happen to think that the best way to save Promise Theory is to locate it within some other
approach to contract law, which shows why promissory obligations are indispensible aspects of this other
approach. I explore one such approach in the next and final chapter, where I investigate the implications of
the view that contract law serves the instrumental goal of facilitating markets, and that promises play an
essential role to this end. I call my view Contract as Commodified Promise.
154

Chapter Five: Contract as Commodified Promise
1 Introduction
The prior Chapters raised the question of which theory of promising provides the best foundation
for Promise Theories of contract law, focusing primarily on the more-or-less unique challenges faced by each
theory. For example, if contractual commitments are irrevocable, it would be nice if the theory of promising
presupposed by Promise Theory could account for this irrevocability. A theory of promising that fails in that
regard is prima facie defective given Promise Theory’s aim of justifying contract law and guiding its content.
This Chapter brackets the which-theory-works-best question. We will assume that some theory of
promising is correct, and that this theory, whichever it is, provides the foundation for a Promise Theory of
contract law. Stock objections to Promise Theory remain. The first is that Promise Theory gives no account
of why the moral obligations associated with promises should be enforced by law. Normally, the fact that
something is morally required does not suffice to justify its legal enforcement, so it is not obvious why things
would be different in the case of promissory rights and obligations. The second objection holds that, even if
we agree that some promises ought to be enforced, Promise Theory lacks any convincing account explaining
which ones ought to be enforced. So even if the moral requirement to keep one’s promises did suffice to justify
legal enforcement of promises, Promise Theory seems to place no limit on the kind of promises subject to
enforcement. This yields uncomfortable implications. My promise to meet you for lunch might in principle be
legally enforceable, and more precisely, might empower you to seek civil redress against me for failing to
abide by my promise.
Ultimately, I think that Promise Theory needs help from outside of promissory morality to answer
the stock objections. My proposal—Contract as Commodified Promise (CCP)—marries an instrumental
justification for contract law, in terms of contract law’s role in facilitating markets, with a promise-based
theory of contract. The instrumental justification is familiar in one form or another. According to the
justification, the law enforces promises because doing so facilitates robust markets, which in turn promotes
many well-known benefits including autonomy and efficiency.
155

Now, there are a variety of ways the state might promote markets. One is by adopting and regulating
an official currency. But the enforcement of promises is important given that promises have an inter-temporal
structure; one promises to perform a future action for someone in exchange for either presently conferred
value or another future action. This promotes a range of future-oriented transactions that parties might not
otherwise undertake given lack of trust or risk aversion. Legal enforceability mitigates somewhat the inherent
risk of non-performance that promisees face. The state’s willingness to enforce promises through contract
law is thus justified, at least in part, by reference to its instrumental goal of securing robust markets. The
instrumental justification thus explains, in response to the first stock objection, why the state should enforce
promises: to facilitate markets.
The justification also suggests a natural answer to the second objection. In response to the question
of which promises should be enforced, the most natural answer is that promises that facilitate market
exchanges should be the primary focus of contract law. In short, contract law should primarily enforce
commercial promises. This is not a novel answer by any means. Other contract theorists have sympathized with
the view that contract law ought to concern itself with commercial arrangements primarily, even exclusively.
The basic view is sometimes assumed without argument among economists engaged in contract theory. And
it is even argued for or presupposed by some who reject economists’ near exclusive concern for the virtue of
efficiency in the allocation of goods and services. But little work—none to my knowledge—has been done in
explaining what it means for a promissory transaction to be commercial in nature. This is not surprising. This
issue of whether a promise counts as commercial in nature rarely matters. And it is fairly easy to identify
paradigm cases of commercial transactions or relationships. Certain employment relations, merger
agreements, or consumer contracts seem unproblematically to involve commercial promises under any
plausible definition.
This Chapter nevertheless argues that contract theory has much to gain from explicating the idea of a
commercial promise in terms of commodification. To commodify something is to treat it as subject to market
156

norms, including treating it as an object available for purchase or sale.
170
I will argue that, in light of the
market-based, instrumental justification, contract law is and should be concerned primarily with enforcing
commodified promises—typically promises sold for money or otherwise exchanged for something of value.
More precisely, I will try to argue in favor of the position that I call Contract as Commodified Promise (CCP). The
position holds, first, that the fact that a promise has been commodified provides a strong pro tanto reason
favoring its enforcement through contract law. The second claim is that the fact that a promise has not been
commodified, or has been improperly commodified, provides a strong pro tanto reason against its enforcement.
Much will be said about commodification and CCP. But for now, I want to point out the dual
functions of CCP. The first that it represents an explication of the idea that contract law should enforce
primarily commercial promises. It makes this idea more precise. But the second, more important role of CCP
is as an interpretive lens through which to view contract law. An interpretive approach to legal theory, as I
understand it, contains both normative and descriptive aspects, with the aim of trying to provide the
normative resources to justify existing legal practices. At the same time, the normative aspects are relatively
independent of those practices, capable of justifying the reform of some of them when they too difficult to
square with the interpretive theory’s underlying claims. This is the spirit in which I offer CCP.
Let me preview some of the normative implications of CCP by focusing on its second claim, which
partly holds that the fact that promise has been improperly commodified provides a strong pro tanto reason
against enforcement. That claim, I will argue, has derivative and direct implications. Derivatively, the second
claim provides a natural explanation for why contract law does not and should not stand ready to enforce
certain contractual arrangements. For example, notice that criminal law bans certain market exchanges—such
as the sale of kidneys, drugs, babies, sexual labor, and so on. One attitude that contract law could take towards
these exchanges is a hands-off approach, in much the way tax law still stands ready to collect taxes from illicit
commercial exchanges, leaving it to other areas of law to regulate the underlying illicit trade. Perhaps contract
law should similarly stand ready enforce otherwise illegal contracts, or failing this, provide partial recognition

170
For a similar usage, see MARGARET JANE RADIN, CONTESTED COMMODITIES 15 (1996) (“One of the
earmarks of commodification, perhaps its central one, is that of sale; so commodification is undercut when
things are thought of as, or declared to be, not capable of sale.”).
157

of the promisee’s rights qua promisee by awarding only nominal damages. But contract law does not do this.
Illegal contractual terms are not enforceable. This is to say that promises to perform proscribed activity are not
enforceable in virtue of the fact that the underlying trade and their solicitations are banned. Contract law
passively and derivatively plays some role in regulating and restricting certain attempts to treat certain goods
as commodities. This reflects the norm that promises that are not properly commodified should not be legally
enforced, in line with CCP’s second claim.
But perhaps more interestingly there is a more direct sense in which promises themselves are
improperly commodified, in virtue of the nature of promise itself and independent of the content of that
promise. And contract law does and should attempt to take this into account, too. Certain promises
themselves, especially what I will call personal promises, made in the context of close personal relationships, are
not properly commodified. Consider this example. We do not treat our most intimate promissory obligations,
made among our nearest and dearest, as available for sale, and hence commodities. Indeed, mere promissory
obligations appear presumptively or more firmly “attached” to those who make them. When a friend
promises to do x for me I generally expect my friend to do x and not someone else hired to perform x in his
place. By contrast, contracts that create consumer debt not only involve the commodification of a promise in
their origination (“I promise to pay x amount to A under the following terms and conditions, in exchange for
y amount distributed under the following conditions…”), I may find myself paying someone other than A—i.e.,
someone or some institution other than the one I originally agreed to pay. And it is no secret why: contractual
obligations are themselves frequently treated on secondary markets as goods available for sale. This is so
given certain default rules in contract law but also permitted by the very terms of certain consumer debt
agreements. Personal promises are presumed not to be alienable in this way. In any event personal promises
are not commodities available for sale, whereas many contractual promises very well may be by design.
This example describes just one putative moral difference between presumptions that apply to
personal promises and those that apply to contractual ones. I will argue, in a normative vein, that if there are
valid moral norms governing public recognition and acceptance of commodification and limits thereto, we
should expect these norms to provide a rich resource from which to draw in characterizing and regulating the
158

commodification of promises themselves. In turn, because contract law is first and foremost the law governing
commodified promises, whatever moral norms govern commodification generally and commodification of
promises specifically should have potential bearing on contract law’s content. We should, moreover, expect
that these very same moral norms (i.e., governing commodification) will not apply in cases involve non-
commodified promises. Contractual obligations are promissory obligations, but they are also commodified
ones, raising additional normative issues.
My view has other implications for contract theory, as we will see in Section 6, in which I suggest
how CCP provides a way to re-frame debates about contract remedies. I do not ultimately endorse a position
in these debates. But I present the argument to illustrate one more way that CCP opens up new avenues for
discussion that to my knowledge have gone unrecognized.
2 An Instrumental Justification for Contract Law
Explaining CCP is easier given a background claim about what I call the Instrumental Justification for
contract law. The Justification begins with a series of claims about the well-known virtues of free markets. I
will not exhaustively discuss these virtues though a few words will be in order. The connection between free
markets and efficiency or preference-satisfaction is widely accepted by economists.
171
In a perfectly
competitive market with zero transaction costs, and perfectly transparent information about goods, services,
and prices, parties who are empowered to make voluntary exchanges with each other will eventually reach an
equilibrium such that nobody could make further transfers that would make them be better off. (One can be
“better off” according to economists if and only if a further preference-satisfying exchange is possible.) This
is because no further exchange could be made without at least one person in that transaction being worse off.
Every voluntary, mutually preference-satisfying transaction between two parties that occurs en route to the

171
Nate Oman, Markets as the Moral Foundation for Contract, 98 IOWA L. REV. 183, 187 (2012) (observing that
“[e]conomic theorists of contract certainly laud markets,” but criticizing their “single concern: the efficient
allocation of resources.”).
159

equilibrium is an ‘efficient transaction,’ while the state of equilibrium according to which no more efficient
transaction can be made is Pareto efficient.
172

We are far from the idealized conditions that bring about complete efficiency. Most obviously
transaction costs are not zero and information is far from fully transparent. But free market systems
nevertheless create strong incentives to produce goods and services that yield preference-satisfying
transactions between willing buyers and sellers, at least as compared with centrally planned economies. This is
largely due to the price system. Prices communicate how much producers are willing to sell a given good or
service for and are assumed to be a function of prevailing market value. Prices and quantities produced
fluctuate pursuant to familiar supply-and-demand functions. And goods or services that fail to attract any
buyers under any price point are quickly weeded out. To the extent that these market mechanisms operate
better than alternative good-allocating mechanisms in satisfying our preferences, and to the extent that
welfare is to a large extent a function of our ability to satisfy our preferences, markets do a comparatively
good job of enabling us to do so by allocating preference-satisfying goods or services to those willing to pay
for them at a given price.
Efficiency aside, markets promote freedom or autonomy (terms that I will use interchangeably here).
Debra Satz elaborates on the tight relationship between freedom and markets. She observes that markets
Present agents with the opportunity to choose between a set of alternatives (partly by
providing individuals with incentives to create the material wealth which is a precondition of
having an extensive array of choices). …Decentralize decision making, giving an agent alone
the power to buy and sell things without requiring him or her to ask anyone else’s
permission or take anyone else’s values into account; Place limits on the viability of coercive
social relationships by providing (at least formally) avenues for exit, Decentralize
information, thereby making abuses of power by authorities less likely; Allow people to

172
It has been observed that Pareto efficiency is an elusive public policy goal, given that for every policy
intervention or deregulation there is very likely at least one person who will be worse off as a result. This has
led economists to lean on the Kaldor-Hicks criterion of efficiency, which stipulates that state of affairs is
efficient if and only if anyone made worse off in a given equilibrium could in principle be compensated by the
persons made better off, sufficient that the post-compensation end result would be Pareto efficient.
160

experiment, to try new commodities, to develop new tastes, to opt out of traditional ways of
life; Contribute to the undermining of racial, ethnic, and religious discrimination by
appealing to the reciprocal self-interest of individuals in exchanging goods with one another
and by fostering anonymous exchange.
173

We might distinguish in this list the ways that markets directly and indirectly relate to freedom.
Markets themselves make possible certain exchanges, which are themselves manifestations or instantiations of
free choice. Markets that make available a wide array of goods and services indirectly make many options
available from which to choose. This goes far beyond the consumer’s freedom to choose among a dozen
forms of goods like toothpaste. A robust market system gives us much greater freedom in choosing where to
invest our labor, at least as compared with either a centrally planned economy or a market system that is very
unstable and (relatedly) not well regulated. The great diversity of goods and services in a robust market
economy gives us many options from which to choose when it comes to choosing a social role to play.
Roughly, it makes it much easier to choose one’s own boss or to become a boss oneself. And if one does not
like one’s boss, we have a much greater opportunity to exercise our “exit” option by quitting and finding a
new boss.
There are other benefits that have been discussed at length elsewhere.
174
But the advantages flowing
from our ability to satisfy our preferences and live freely stand out as particularly important. To obtain these
market benefits, certain preconditions must be in place. Securing these benefits requires that a given market
be relatively competitive and that information pertinent to buyers and sellers be relatively available. There
must be an initial allocation of property holdings, which can be more or less consensually transferred to
others, and which can be possessed more or less securely against coercive or otherwise involuntary
appropriation by others. More generally, for our purposes it is worth emphasizing that large-scale markets
need some kind of stable legal system to flourish.

173
DEBRA SATZ, WHY SOME THINGS SHOULD NOT BE FOR SALE 21 (2010).
174
See especially Nate Oman, Markets as the Moral Foundation for Contract, 98 IOWA L. REV. 183, 187 (2012)
(“Markets … are complex social institutions that can serve multiple functions.”).
161

Part of this legal system, to be effective, will require something like a contract law. Satz observes that
markets “require[] the existence of legal and regulatory frameworks to ensure that contracts are enforced and the
given property rights are respected,” adding that “all property rights are the products of laws and conventions
that back them up and enforce them.”
175
She is primarily concerned with the necessity of stable property
rights for a flourishing market. But contract law looks just as practically indispensable.
176
Most obviously,
contract law provides some measure of assurance that parties to a contract will actually follow through with
their agreements or face sanction in the form of potential legal liability.
Less familiar, contract law also serves a gap-filling function. Contracts cannot account for every
possible contingency that might make contractual performance more difficult, including but not limited to the
interpretation of the terms of a given contract. Contract law thus supplies a wide variety of “default rules”
governing contract interpretation or default contractual terms that apply absent express agreement to the
contrary. For example, under the Uniform Commercial Code, contractual rights of either the seller or buyer
of goods may be assigned to others “[u]nless otherwise agreed all rights of either seller or buyer.”
177
These
off-the-rack default rules allocate liabilities when contract disputes arise and where the contract itself is silent
or unclear. By mitigating potential sources of conflict and confusion arising from this silence, contract law’s
gap-filling function facilitates market transactions wholly apart from any enforcement function.
These functions plausibly explain certain ways that contract law facilitates market transactions. But at
their core they under-emphasize promises. Contract law facilitates markets by enforcing certain promises—

175
DEBRA SATZ, WHY SOME THINGS SHOULD NOT BE FOR SALE 26-27 (2010) (emphasis added).
176
It might be argued that, in principle, all we need is a system of property law; no contract law is necessary.
The property law would simply hold that the only valid exchange of property rights occurs simultaneously,
such that no outstanding or future-oriented performances or transfers would be permitted. I do not want to
claim that such a system is logically impossible. But the argument sketched here is not a transcendental
argument: it is not claimed that it is impossible to have a market economy without contract law, just that it is
practically impossible to have one that is successful or flourishing. This seems too plausible to be disputed to
me.
177
Uniform Commercial Code § 2-210(2) (there are other restrictions on assignment, which disallow
assignment “where the assignment would materially change the duty of the other party, or increase materially
the burden or risk imposed on him by his contract, or impair materially his chance of obtaining return
performance.”).
162

namely, commercial promises.
178
We will return to this idea in a moment when I discuss the commodification
of promises. For now, notice that the aforementioned assurance and gap-filling functions of contract law
imply a future-oriented perspective that we can explain by reference to contractual obligation’s promissory
foundation. Contract law provides some measure of additional assurance to parties to a contract regarding
future actions of promisors. Similarly, the gap-filling function is useful only when one contemplates that contract
law will govern transactions that are underway but not yet complete. Issues arise from contractual silences or
gaps only if a contractual obligation has yet to be satisfied or there is a question about whether that obligation
has been satisfied. This future-oriented perspective is easy enough to explain since they derive from the
promissory foundations of contract. Promises are necessarily future-oriented in the same sense. When one
makes a promise, one tries to assure another that one will perform an action in the future.
This future-oriented perspective—the promissory perspective—is vital for well-functioning markets.
Consider another way one might engage in a market transaction without contracts and their characteristically
future-oriented perspective. One might engage in a simultaneous or near-simultaneous exchange of goods or
services, in which something of value is exchanged almost immediately for something else. And once the
immediate exchange is complete, the transaction is complete. But it would be very difficult to sustain a large-
scale, robust market economy if all transactions were limited to this kind of near-immediate exchange—
exchanges that are essentially non-contractual. So instead we allow for the exchange of promises. Someone can
exchange actual, present value (cash, for instance) for a guarantee—a promise—of a future performance by
the other party. Or the exchange might be constituted as a mutual exchange of promises. One person

178
Limiting contract law to commercial transactions is commonplace in the law-and-economics approach to
contract theory, which often stipulate that the sole goal of contract law is to maximize the overall welfare of
individual transactions. See, e.g., Alan Schwartz and Robert Scott, Contract Theory and the Limits of Contract Law,
113 YALE L.J. 541, 544 (2003) (advancing the theory that “contract law should facilitate the efforts of
contracting parties to maximize the joint gains (the ‘contractual surplus’) from transactions,” as well as the
claim that “contract law should do nothing else.”). But this individual-transaction-specific perspective
threatens to ignore systemic benefits and harms justifying a system of law designed from the perspective of
highly idealized individual transactions among well-informed, arm’s length agents of roughly equal bargaining
power. The system will govern even those transactions that fail to fit these criteria, and in any event, this
literature ignores the idea that there is a set of rights capable of constraining the design the law putatively
aiming to maximize overall welfare.

163

promises to do something in exchange for a promise by another to do something else. Promissory
obligation’s future-oriented perspective lies at the heart of contractual obligation’s identical orientation. And
the latter makes possible a huge range of market transactions otherwise virtually impossible to obtain.
The preceding remarks pave the way to an Instrumental Justification for contract law. If we assume
that a well-regulated market economy is justified (given the efficiency and autonomy gains), and that a legal
framework is practically indispensable for this kind of economy, and furthermore, that contract law or
something very much like it is a practically indispensable component of this framework, then it looks like
having a contract law is thereby justified. This kind of justification for contract law looks very much like the
one that normative economists favor, though it does include values other than efficiency—namely and
primarily, autonomy (or what Satz calls ‘freedom’ above). But one need not rely on a criterion of efficiency or
preference satisfaction to endorse it. One might rely exclusively on the fact that relatively and well-regulated
free markets, as compared to a Soviet-style, centralized allocation of resources, produce a tremendous amount
of freedom or autonomy.
There is something appealing about this Instrumental Justification. And any normative theory of
contract law that wholly ignores the important role that it plays in facilitating a market economy misses out
on something important. By the same token, the Instrumental Justification presented here does not deny that
promises play an important role in justifying contract law. To the contrary, promises play an essential justifying
role. It is precisely because of the unique structure of promissory commitment, which binds people to perform
for others in the future, that makes promising an essential piece of the puzzle. The Instrumental Justification
is also appealing because it helps respond to the two stock objections at the outset. First, in response to the
objection of why promissory morality justifies contract law, we need only point to the distinctive inter-temporal
structure of promissory obligation, namely the fact that promising involves a commitment to perform for
others over time, and that this very inter-temporal feature is what must be reinforced by law to ensure a
thriving market economy.
The Instrumental Justification also helps to answer the second stock objection: which promises ought
to be enforced. The answer, which will be developed in the next few sections, is that the promises that should
164

be enforceable are those promises that are involved in market transactions, or commercial promises for short.
Commercial promises help constitute markets in various goods. So given contract law’s aim—to facilitate
markets—it should come as no surprise that contract law should stand ready to enforce systematically
promises that accomplish this task. This provides a principled account of the promises that contract law
should primarily aim to enforce, one that makes sense in light of the Instrumental Justification.
In the next two sections, I will refine the notion of a commercial promise. For now let us consider
two worries that we might have about the Instrumental Justification. The first is less worrisome than the
second. First, notice that markets give rise to problems that offset their virtues to some extent. Some market
transactions may be mutually beneficial to those involved in the exchange yet may produce undesirable side-
effects—negative externalities—to third parties who were not involved in those voluntary exchanges. Other
markets are just inherently “noxious,” to borrow a term from Debra Satz, and should be very highly regulated
or prohibited entirely.
179
These noxious markets either undermine the conditions for agency or create social
problems that far outstrip the ability of any market-based mechanism for reform. Markets in human beings—
babies or adults—are paradigm examples of these noxious markets. Unlimited markets in child labor also
tend to undermine the agency of children by preventing them from developing capacities necessary to
navigate their lives in a liberal democracy (e.g., by creating incentives to prevent children from pursuing
education).
180
Other markets undermine vital public institutions: a market for votes would corrupt democracy
itself. We can multiply examples. But the point is that the justification of contract law becomes more
precarious once we realize that markets bring potential problems in addition to benefits.
But these observations do not undermine the Instrumental Justification. Indeed, assuming that the
state plays a legitimate role in mitigating these problems in order to maintain a well-regulated market, contract
law might play an important supporting role. This is especially clear in the case of noxious markets. Certain
contract law doctrines already embody this role. Contracts that are “contrary to public policy” or illegal are

179
See DEBRA SATZ, WHY SOME THINGS SHOULD NOT BE FOR SALE 91-112 (2010).
180
Id. at 157-69.
165

void.
181
So contracts in, say, sexual services or drug trades or other illicit markets will not be enforced given
their illegality as defined by other areas of law. Markets that depend for their existence on the exploitation of
vulnerable members of the public—where there is deeply unequal bargaining power—are limited by the
unconscionability doctrine, which holds that deeply unfair or exploitative contract terms are unenforceable.
182

Contracts containing racially discriminatory terms will not be enforced, given the illegality of racially
restrictive labor and real estate markets. Criminal law, consumer-protection law, anti-trust law, anti-
discrimination law, and property law all take the lead role in banning and regulating noxious markets, but
contract law surely plays an important role supplementing these other market-regulating areas of law. So the
first problem with the Instrumental Justification is not insurmountable, and indeed, yields a normative
resource—the goal of mitigating noxious markets and negative externalities—that might very well inform
contract law’s design.
But a second problem is deeper since it suggests that the Instrumental Justification is incomplete.
The problem is familiar to any instrumental justification for a social practice, which is that instrumental
justifications should respect the constraints imposed by justice and moral rights. To draw an analogy, it might
be the case, as H.L.A. Hart argued, that deterring crime plays a necessary role in justifying criminal law, but it
is not sufficient without explaining how deterring crime is constrained by rights, especially the rights of the
innocent not to be punished.
183
Otherwise the criminal-deterrence goal might end up “justifying” the
punishment of innocents despite this obvious injustice. Similarly, the Instrumental Justification provides only

181
See, e.g., Hurd v. Hodge, 334 U.S. 24, 34-35 (1948) (“The power of the federal courts to enforce the terms
of private agreements is at all times exercised subject to the restrictions and limitations of the public policy of
the United States as manifested in the Constitution, treaties, federal statutes, and applicable legal precedents.
Where the enforcement of private agreements would be violative of that policy, it is the obligation of courts
to refrain from such exertions of judicial power.”); see, e.g. Cal. Civ. Code § 1598 (“Where a contract has but a
single object, and such object is unlawful, whether in whole or in part, or wholly impossible of performance,
or so vaguely expressed as to be wholly unascertainable, the entire contract is void.”).
182
See, e.g., Williams v. Walker-Thomas Furniture Co., 350 F.2d 445 (D.C. Cir. 1965); Romney v. Fransiscan
Med. Group, 349 P.2d 32, 36 (Wash. Ct. App. 2015) (describing unconcionscable contracts as void).
183
This is not quite how Hart puts the point, which is more accurately characterized as insisting the need to
distinguish system-level questions of what the “general justifying aim” of a system of punishment is from the
allocational question of who should receive that punishment. See H.L.A. HART, PUNISHMENT AND
RESPONSIBILITY 1-27 (1968).
166

a system-level, market-based justification for enforcing promises. Facilitating markets by means of enforcing
promises must protect or respect certain rights to be justified.
But which rights? A host of rights will naturally constrain the operation of contract law in virtue of
constraining all human action: rights against harm, rights against unjust discrimination, and so on. But let us
focus on rights that are implicated every time one makes a contract. Given the Instrumental Justification,
every legally enforceable contract involves a promise’s being potentially legally enforceable; accordingly, a
natural thought is that promissory rights and their correlative obligations will play some constraining role. We
might put this in terms of a moral status that must be respect by the law: promisors and promisees, qua their
status as promisor and promisees, should be adequately respected. And since their status is defined in terms
of their holding certain promissory obligations and correlative promissory rights, respectively, the state should
presumably respect the legitimate expectations of promisors and promisees qua holders of promissory
obligations and promissory rights, respectively. If individuals are to be enlisted without manipulation into the
service of market-constituting public policy goals, then contract law should respect their expectations as
holders of promissory rights and obligations. This suggests that contract law should reflect at least to some
extent the norms governing promissory rights and obligations, which include norms governing their
formation, performance, and remedies for broken promises.
I acknowledge that this is far too quick. And I have no plans in this final chapter to develop detailed
views on how contract law’s design, at the level of doctrine, can best respect the rights of promise holders.
This is because I aim primarily to explain Contract as Commodified Promise, the view that explicates the
notion of a commercial promise and draws out its implications for contract theory. So I am content in simply
observing, however unsatisfyingly, that promisors and promisees are co-opted in the service of a public policy
goal of facilitating markets, and asserting that justifying this co-option is possibly only if contract law
adequately respects pre-existing rights of promisors as such. I leave to future work the questions of how
contract law must be designed to adequately respect promisors and promisees.
I want to focus on a different issue. The question of how contract law is to be designed is
complicated by the fact that primarily commercial promises are to be enforced. What exactly this means and
167

what implications this has for contract law is what I want to focus on in the rest of this Chapter. To lay the
groundwork, let me begin by exploring the notion of commodification, since I will use that notion to
explicate Contract as Commodified Promise.
3 Commodities, Commodification, and Why It Matters
Contract law should primarily stand ready to enforce commercial promises. This view makes sense
given the Instrumental Justification for contract law, namely, facilitating markets. As noted, this is not a novel
view. What is novel about Contract as Commodified Promise is that it represents an attempt to explicate the
notion of a commercial promise. No work has been done on this topic, to my knowledge, and perhaps with
good reason. Most commercial transactions look transparently so. If IBM enters into a merger agreement
with Widget Corp., we need no more precise concept of commerciality to tell us that this merger agreement is
commercial in nature. Examples can be multiplied. Add to this fact that, as I have acknowledged, some non-
commercial promises might very well be permissibly enforceable in contract law, the need for a more precise
account of commercial transactions looks less pressing still.
I nevertheless think that there is something to be gained for contract theory in taking a closer look at
the idea of a commercial promise. To see why, I will need to perform some groundwork. My claim will be
that, to explicate the nature of commercial promises, we should understand commerciality in terms of
commodification. That is, a promise is a commercial promise if and only if that promise is being treated as a
commodity. This section lays out working definitions of ‘commodity’ and ‘commodification,’ while also
sketching certain ideas in the literature about what can make commodification morally problematic, and
hence why the issue of commodification matters.
a. Commodities and Commodification
The relevant notion of a ‘commodity’ for our purposes is something close to what Elizabeth
Anderson has in mind when she writes, “Commodities are those things which are properly treated in
accordance with the norms of the modern market.”
184
I will instead offer the following:

184
Elizabeth Anderson, Is Women’s Labor a Commodity?, 19 PHIL. & PUB. AFFAIRS 71, 73 (1990).
168

Commodity. Commodities are those objects, actions, or other things, tangible or intangible,
which are usually treated in society as subject to market norms, including especially norms
permitting the purchase or sale of those things.

Notice that unlike Anderson I offer a non-normative definition, according to which something
counts as a commodity in a society to the extent that it is typically treated in accordance with market norms.
So on Anderson’s account, in a society that allows chattel slavery, slaves are not commodities because people
are not properly treated in accordance with the norms governing purchase or sale of objects, whatever those
norms may be. By contrast, on my non-normative definition a slave in a slaveholding society is a commodity,
given that chattel slavery by definition involves treating slaves as subject to market norms, such as norms
governing purchase or sale. My aim in giving a non-normative construal of the term is just to distinguish
analytical and normative issues to the extent possible. The normative issues will be taken up in the next sub-
section.
Returning to the definition, its most important element is the idea of treating something as subject to
market norms. A complete list of extremely fine-grained market norms is certainly beyond the scope of this
paper, and as a descriptive matter, might also require in-depth, market-specific historical, sociological, and
anthropological spadework. But many of the norms that writers about commodification have in mind are
norms, social and legal, that permit or otherwise govern the following kinds of activities:
• Selling or attempting to sell something.
• Purchasing or attempting to purchase something.
• Brokering a sale or purchase or attempted to broker these activities.
• Participating in a quid pro quo exchange of something for something else or brokering the exchange.
• Lending or otherwise transferring it to someone else in a way that does not constitute a purchase or sale.
169

This list is surely not exhaustive.
185
The first three activities, which directly involve to the purchase or
sale of things (or attempts to do so), probably involve treating something as a commodity per se. There might
be unusual borderline or controversial cases, but for the most part, let us assume that treating something as
available for purchase or sale qualifies as treating that thing as subject to market norms; as treating that thing
as a commodity. I am less certain about the remaining two items on the list. Very often, but not always, a quid
pro quo exchange of value involves treating something as a commodity. We might imagine a favor economy,
where favors are performed only for favors, or a barter economy, where currency transfers are prohibited or
infeasible, and goods are transferred only for other goods. Each of these exchanges is plausibly interpreted as
activities that treat things as subject to market norms. Still, plenty of quid pro quo exchanges do not involve
commodification. An exchange of vows during a wedding ceremony likely does not contribute to the
judgment that the relationship is somehow being commodified. So as an indicator the idea of exchange is not
very strong.
I also have reservations about whether lending or otherwise treating something as available for
transfer involves treating it as a commodity. Perhaps so in many cases. But Italy can lend an antiquity to a
museum without anyone’s viewing the artifact as properly subject to market norms. And gifting objects is a
way of transferring an object that seems properly outside the domain of market norms. The conditions under
which a gift transfer is valid and proper are not primarily determined by market pricing mechanisms. And the
way one ought to value a gift does not depend entirely on, say, whether the gift has a high monetary value.
These above-listed examples can apply to individual acts and social practices. To the extent that I
accept an offer for employment as a junior lawyer, I am selling my labor and my employer is purchasing it.
Indeed, my employer can turn around and sell hours of my time at a considerable markup to our clients, a
markup rate which is set by ordinary supply-and-demand pricing mechanisms, and which in turn is a usual
way of setting the price for my services. Because lawyer’s labor-dedicated time is treated, in some ways, as
subject to market norms (they are purchased, subject to a market rates, as a proxy for labor on a client’s legal

185
There is considerable overlap here with the ways in which one might treat an object as a piece of property.
170

issues), and because this is a usual way of purchasing labor from lawyers, lawyer’s labor-dedicated time counts
as a commodity, on my usage of the term.
This brings us to commodification. As I will use the term, to commodify something is to treat it as a
though it were a commodity; to treat something as subject to market norms, whether or not it is usually
treated that way. This is also a non-normative account. Often, the term ‘commodification’ has a pejorative
ring to it in much the way that ‘discrimination’ does. But once again for clarity I want to separate analytical
from normative issues. Just as we can distinguish discrimination from wrongful discrimination, I want to distinguish
commodification from improper commodification.
186

Although I provide no precise recipe for determining whether a given act or practice involves
commodification, the above-listed examples once again provide useful guidance on the question, and I will
draw freely from the list in the remainder of this paper. Most of the time I will discuss transactions that
involve the purchase of sale of something—a kidney, a promissory right—and hence transactions to which
the notion of commodification plainly applies.
One last point. Things can be more or less commodified or more or less a commodity. A barrel of
crude oil is a quintessential commodity, available for purchase or sale in largely anonymous markets at prices
that are highly sensitive to market fluctuations. Other things—certain Greek antiquities, labor, and so on—
may be treated like commodities but only as a matter of degree, or not at all. I will have more to say about
this issue later on. The important point for now is to recognize that commodification can be a matter of
degree.
This analytical background aims to clarify what I mean when I talk about ‘commodities’ and
‘commodification.’ The definitions are stipulative, but they stipulate in a way that I think is largely compatible
with common usage, and do not seek to gain argumentative advantages by verbal fiat. Just as the important

186
I could avoid a normative account of commodification by defining it more narrowly, say, as treating
something as subject to market norms even though it is not usually treated this way. But I think that this is too
narrow: one’s labor is, in a sense, usually treated as subject to market norms in some ways, and I want to
preserve our ability to correctly observe that our labor is being commodified. In other words, I think it is
perfectly appropriate to say that our labor is commodified despite the regularity with which it occurs. So at
the risk of over-inclusion I will use the analytically stripped down account of commodification.
171

questions about discrimination are to a lesser extent analytical and to a greater extent about the conditions
under which discrimination against individuals is wrongful, what matters about commodification are the
normative issues, like whether and under what circumstances commodification is morally permissible or
problematic. I address these normative questions below.
b. Why Commodification Matters: the Moral Limits of Markets
The controversies over commodification—whether, for example, kidneys should be available for
purchase or sale—are not primarily analytical but normative controversies. The important questions are
whether certain things or acts or practices should be commodified, and if not, why not. When are things
properly or improperly subject to market norms? To get a grip on the kinds of answers given, it will be easier
to focus on the reasons given against commodifying certain things, acts, or practices. Let me clarify that my
aim in this sub-section is mostly descriptive: I aim to lay out certain forms of argument against commodification
that exist in the philosophical and legal literature, without necessarily endorsing any of them.
Two kinds of reasons have been offered against commodification, or more precisely, two kinds of
reasons have been given to explain what makes commodification improper when it is improper. Call the first
kind intrinsic reasons. These reasons focus on the ways in which commodifying something is incompatible
with the right way of valuing or respecting that thing. We can put the point in a few ways. Commodification:
expresses the wrong kind attitude towards the thing commodified; involves valuing the thing for the wrong kind of
reasons; or degrades or corrupts the target of commodification.
Michael Sandel elaborates on the idea of degradation or corruption. “The corruption argument,”
according to Sandel “focuses on the character of the goods themselves and the norms that should govern
them.”
187
He elaborates:
[The corruption argument] appeals not to consent but to the moral importance of the goods
at stake, the ones said to be degraded by market valuation and exchange. So to decide
whether college admission should be bought and sold, we have to debate the moral and civic
goods that colleges should pursue, and ask whether selling admission would damage those

187
MICHAEL SANDEL, WHAT MONEY CAN’T BUY 113 (2012).
172

goods. To decide whether to establish a market in babies up for adoption, we need to ask
what norms should govern the parent-child relationship, and ask whether buying and selling
children would undermine those norms.
188

This passage contains some elements illustrating what I call intrinsic reasons against
commodification. They are reasons that trade on the nature and value of the thing being commodified and/or
the attitudes of the parties doing the commodifying. Something is morally problematic, in and of itself and
apart from other results that the commodification will bring about.
Here is another example. A paradigm intrinsic critique targets chattel slavery, which involves
commodifying persons, which in turn treats persons as objects available for sale. Aside from the obvious fact
that slaves do not typically consent to sell themselves into slavery, intrinsic critiques go further and argue that
even if they were to try to consent this would be wrong, too. Slavery is incompatible with the right way of
valuing persons. Persons are owed basic respect and dignity, and chattel slavery deprives them of both. To
the extent that slaves are valued, they are valued for the wrong kinds of reasons: they are valued in the way
that livestock is valued, as being more or less useful as tools, or having more or less resale value in the market.
And chattel slavery obviously expresses the wrong kinds of attitudes towards persons insofar as they are
viewed primarily as objects.
All of these criticisms depend on a prior conception about the norms governing our behavior
towards persons and ideas about the right ways to value persons. But intrinsic critiques have been advanced
against commodifying reproductive labor, sex, the right to vote, and other things as well.
189
Again, each kind
of intrinsic argument will differ depending on the nature of the underlying object being commodified.

188
Id. at 112.
189
The practice of commercial surrogacy similarly involves, argues Elizabeth Anderson, treating women’s
bodies as means of production, and risks treating both women and their children as objects. The proper mode
of valuing children is by loving them, not assessing them defective as one would the shipment of widgets that
one has purchased. Similarly, the right way to value the right to vote is by exercising it intelligently, rather
than selling it on the open market. Selling one’s vote degrades its value.
173

These intrinsic critiques are often but not necessarily coupled with extrinsic reasons against
commodification, which object to undesirable states of affairs perpetuated or promoted by markets in
commodified objects. Consistent with extrinsic critiques, treating something as fit for market exchange might
not be intrinsically morally improper. Suppose that an individual sex-for-cash transaction between consenting
adults is not intrinsically morally wrong. Still, allowing the practice of prostitution to flourish—allowing a
robust market for sexual labor to develop—might perpetuate or promote: the exploitation of juveniles, the
illicit trafficking of humans, debased attitudes towards women that operate to oppress them, exploitation of
the poor who suffer as a result of distributive injustice, and so on. Permitting the growth and entrenchment
of markets to develop in women’s sexual labor threatens to facilitate these undesirable outcomes. And
morally speaking we should not be complicit in, or actively contribute to, practices—market-based or
otherwise—that contribute to or to create incentives to facilitate these kinds of harmful acts and injustices. So
we should not facilitate the commodification of women’s sexual labor.
190

The “corruption argument” that Sandel has in mind also contains elements that I would call extrinsic,
which just illustrates how intrinsic and extrinsic objections to commodification are not mutually exclusive and
sometimes difficult to disentangle. Sometimes commodification is morally problematic given that it tends to
cause important social norms to break down, sometimes in addition to the intrinsic problems with the
commodification at issue. Sandel goes on to explain the sense in which important non-market norms might
be undermined by market ones. He cites recent empirical work on the “commercialization effect,” which
suggests, for example, that offering to pay limited sums of money for certain goods and services—such as
blood or legal services for the underserved—are less successful in gaining participants willing to give their
blood or legal services, as compared to requesting donations of blood or pro bono legal services. Sandel takes
commercialization effects as evidence that commercializing certain valuable goods or services not only might,
under certain constraints, under-produce them (contrary to standard economic theory), offering payment for
services in lieu of requesting donations might even undermine norms governing altruistic giving. This and

190
For a compelling example of this kind of argument against markets in women’s sexual labor, see DEBRA
SATZ, WHY SOME THINGS SHOULD NOT BE FOR SALE (2010) (Chapter 6).
174

similar work in social psychology suggests to Sandel that “[a]ltrusim, generosity, solidarity, and civic spirit are
not like commodities that are depleted with use,” but rather “[t]hey are more like muscles that develop and
grow stronger with exercise.”
191
But societies that allow market values and norms to encroach in all areas of
life cause vital non-market norms to atrophy.
In sum, writers have criticized other forms of commodification on intrinsic and extrinsic grounds,
and some forms of commodification might be morally worrisome for both intrinsic and extrinsic reasons;
intrinsic and extrinsic arguments against commodification often walk hand in hand. But I hasten to clarify a
few things: First, to commodify something is not necessarily morally problematic; only some forms of
commodification raise worries of the intrinsic or extrinsic variety. Second, debates about the proper legal
responses to commodification include more than simply banning the purchase or sale of certain things. As
Debra Satz reminds us, although markets in women’s sexual labor may have horrible intrinsic implications
and extrinsic consequences, it is an open and partly empirical question about the right way to respond to it.
192

Perhaps a highly regulated and unionized workforce of sex workers will, all-things-considered, be best.
Perhaps an outright ban on prostitution minimizes harms. The point is that we cannot simply infer that we
ought to declare unlawful certain forms of commodification and the markets that they give rise to simply
from the premise that those forms are morally problematic.
One more point of clarification. Commodification, as I briefly mentioned already, is not an all-or-
nothing enterprise. Something can be partially commodified. Margaret Jane Radin usefully coins the phrase
“incomplete commodification” to point out how.
193
The term, as she uses it, actually points to two different
phenomena. The first refers to how commodified understandings of certain transactions can coexist with
non-commodified understandings. To illustrate, recall the lawyer whose firm sells her labor an hour at a time.
This same lawyer might also understand her work as a calling and she might view herself as a member of a

191
MICHAEL SANDEL, WHAT MONEY CAN’T BUY 130 (2012).
192
DEBRA SATZ, WHY SOME THINGS SHOULD NOT BE FOR SALE 150 (2010) (“It is important to distinguish
between prostitution’s wrongness and the legal response that we are entitled to make to that wrongness. Even
if prostitution is wrong, we may not be justified in prohibiting it fi that prohibition makes the facts in virtue
of which it is wrong worse, or if it has too great a cost for other important values.”).
193
MARGARET JANE RADIN, CONTESTED COMMODITIES 102-14 (1996).
175

profession. She might wholly embrace her identity as a lawyer, an identity subject to norms of professionalism
that is defined in part by deeper norms—including ethical norms and norms governing one’s self-
conception—that at bottom are non-negotiable or otherwise subject to market norms. So being a lawyer or
laborer more generally can involve elements that are both commodified and non-commodified, viewed from
the perspective of the lawyer or laborer.
The second phenomenon indicating incomplete commodification is distinctively social or legal. Some
objects are more commodified than others. Consider books. These can be produced, sold, purchased, and
otherwise transferred with comparatively few legal restraints on these processes. I say “comparatively”
because other goods and services are subject to much greater regulation. Insurance laws, child labor laws,
food safety regulations, and so on all limit to some extent the ways in which insurance, labor, and food may
be produced and sold. Some of these regulations—food safety laws, for example—can be explained and
justified by reference to run-of-the-mill economic efficiency concerns. We can, for example, construe them as
forcing firms to internalize negative externalities or mitigating information asymmetries between food
producers and consumers, thereby curing market defects and promoting efficient allocation of resources.
But some regulations—and this is controversial, of course—are better understood as manifesting
social ambivalence about commodification in the first place. Debra Satz explains child labor laws in this way.
Some child labor might be justifiable in quite limited circumstances (some child actors, perhaps). But some
worry about the intrinsic nature of selling a child’s labor; that, since children lack full moral agency, this is
morally on par with treating a child like a beast of burden. Satz’s worries are partially of the extrinsic variety,
focusing on the cumulative harms to moral agency inflicted on children in societies with robust markets in
child labor. She argues, more extrinsically, that these societies create huge incentives for poor children to
work low-skill tasks at the expense of cultivating knowledge and becoming full moral agents able to function
in a democratic society. All of these attitudes may coexist in a society. This social ambivalence is expressed, at
least in some jurisdictions, through laws that highly regulate the production, solicitation, and sale of child
labor to limited circumstances as compared to the sale of adult labor. So if the commodification of adult labor
176

is a benchmark, then there is far less commodification of child labor, though both represent instances of
incomplete or partial commodification.
We can generalize. Something can be commodified more or less. Individuals can express ambivalence
towards the commodification of their own actions or the actions of others. Aspects of those actions may be
commodified even though they may co-exist with non-commodified aspects of those same actions. And
societies as a whole are ambivalent towards commodification in a way that often shows up in how certain
market activities are regulated. This will be important when unpacking Contract as Commodified Promise. I
will draw on the preceding background on commodification in what follows.
5 Contract as Commodified Promise, Introduced
With the preceding groundwork in mind, I now want to return to our main question: which promises
should be enforced in contract law? My answer is as follows:
Contract as Commodified Promise (CCP):
1. If a promise is commodified, then this is a strong pro tanto reason favoring its
enforceability in contract law.
2. If a promise is not commodified, or not properly commodified, then these are a
strong pro tanto reasons against its enforceability in contract law.
Before seeking to bolster these claims, let me first turn to a threshold question: what exactly does it
mean to say that a promise is commodified? In what follows, when a promise has been commodified, I have in
mind that a promissory right and its correlative obligation has been sold by the promisor to the promisee. Perhaps there are
other ways of commodifying a promise. But the purchase or sale of promissory rights is the most important
way, for our purposes, a promise can itself be commodified.
Despite this clarification, my suggestion initially sounds odd. If I promise to sell you a freight of
widgets for $10,000, what you have purchased is a freight of widgets and that is exactly what I have sold; the
widgets, not the promise, are the commodities. Similarly, if I promise to mow your lawn for $100, what you
have purchased my lawn-mowing labor. We can explain the purchase or sale without reference to the
underlying promise or associated promissory rights or obligations. Promising is needed to complete the
177

relevant speech act and effectuate the sale and transfer of ownership in the underlying good or service. But
the content of the sale itself does not include a promise; it includes the completed results of a promise.
But this is too quick. Insofar as transactions like these contain terms (implied or express or imputed
by law) that require a future action by one of the parties, then a purchase and sale of at least one promise is
involved. Return to the cases. Assume for the sake of argument that you have purchased a freight of widgets,
and that immediately pursuant to the terms of the agreement, you have ownership rights over the widgets. It
does not follow that you have possession of those widgets, which still need delivery either to your or someone
else. So there is an outstanding action that I still need to perform. Even if I simply promise to maintain
possession for the time being, this too involves a promised performance that you have purchased from me as part
of the overall deal. You obtain a right to my certain performance in the future. This is essentially a promissory
right. So too with the lawn. In this case, it is not plausible to insist that you have purchased from me nothing
more than a mowed lawn or labor that brings about this result, strictly speaking. I have not mowed your lawn
yet. You have instead purchased from me a right to a future performance that I cannot unilaterally back out
of; it is essentially a promissory right. Any time we have a right to a future performance that cannot be
unilaterally undone, and any time this right is purchased or sold, a promissory right or obligation has thereby
been commodified.
194

One more point before returning to CCP’s core claims. Promises, like anything else that might be
subject to market norms, can be more or less commodified. This was a point made in the previous sections.
The implications of this observation will be explored in greater detail in Section 6. For now, let me try to
explain CCP’s core claims in more detail.
a. Claim (1) of Contract as Commodified Promise

194
Someone might respond that at least some contracts effectuate nothing more than transfers of ownership
rights; that once certain writings are complete transfers of ownership over things are complete. Such a
document, however, is no longer governed by the law of contract but rather the law of property, including
especially the rules governing transfer of ownership. I take contract law’s proper domain to be properly and
primarily governing the enforceability of outstanding promissory rights and obligations that have been
commodified, not in already-completed transfers of ownership. In general, just because a document
effectuates a transfer of ownership rights does not entail that the document is primarily a contractual
document. The law of wills and trusts and estates begs to differ.
178

Let us begin with claim (1): If a promise is commodified, then this is a strong pro tanto reason
favoring enforceability in contract law. To break this down, let met analyze it in terms of [i] what it means for
a promissory right to be commodified, [ii] why proper commodification provides a strong reason for
enforceability, which is nevertheless [iii] only a pro tanto reason, rather than necessary or sufficient.
The first sub-issue—what does it mean for a promissory right to be commodified—has already been
discussed: to commodify a promise is to treat a promissory right or obligations as subject to market norms,
especially norms governing purchase or sale of things. I will not belabor the point. This brings us to sub-claim
[ii]: that the commodification of a promise provides a strong reason for enforceability. The support for this
claim flows from the Instrumental Justification. To see why, notice that if a promissory right is purchased or
sold in a transaction, then there is an outstanding performance due. This is just to say that the promisor has
sold to the promisee a right to a future performance of an action. But this is exactly the sort of transaction
that the Instrumental Justification seeks to facilitate in order to facilitate markets. Recall: legally enforcing
promises allows individuals to rely on more than merely moral assurances that a performance will actually
take place; it allocates to promisees a right to initiate legal action against delinquent promisors.
Now, this can be said of any promise that is backed by legal enforcement. The Instrumental
Justification, however, is centrally concerned with facilitating robust markets to secure the benefits therefrom.
To that end, the promises that should be enforced should be commercial promises. And to ensure that
promises will be commercial in nature, we should it helps that the promissory rights and obligations themselves are
purchased or sold, since the purchase or sale of anything involves a quintessential market transaction. Put
differently, perhaps the best way to ensure that there will be a tight link to the instrumental goal of market-
facilitation will be for the state to commit in advance to enforce nearly all properly commodified promises.
There is therefore a natural continuity or congruence between, on the one hand, the fairly common view that
contracts ought to enforce commercial promises primarily given its market-facilitating task, and my distinctive
view that we should understand commercial promises as those which involve the purchase or sale of
promissory rights.
179

So the Instrumental Justification provides natural support for claim (1), i.e., that if a promise is
commodified, then this is a strong pro tanto reason favoring its enforceability in contract law. But one might
think that this shows more than (1) claims. Why not adopt the stronger thesis, that commodifying a promise
involves a necessary or sufficient grounds for enforce that promise? The sufficiency claim is easy enough to
reject, since other doctrinal limitations—limitations that have little to do with the in-principle enforceability
of the promise—may preclude enforcement. For example, contracts in some jurisdictions have a writing
requirement, which prevents certain merely oral contracts from being enforced. One justification for
requiring a written document runs roughly as follows: certain contracts (contracts agreeing to the transfer of
real property, for instance) are prone to dispute. Given that the costs of reducing claims to writing are so low
compared to what are often high stakes, under circumstances likely to generate disputes absent a writing, the
state may reasonably require writings to defuse conflicts before they arise and to help resolve those conflicts
(if they arise anyway). But writing requirements, whatever their ultimate justification, prevent the enforcement
of certain commodified promises expressed via oral communication.
But why not allow commodification of promissory rights be a necessary condition on enforceability?
Although the thought is tempting, I want to leave open the possibility that some broken promises, even
though they are not commercial in nature, give rise to such great harms to promisees who quite reasonably
relied on those promises, that some area of law—maybe even contract law—should on occasion stand ready to
enforce them. Such cases, to the extent they exist, would show that promissory commodification is not
necessary.
b. Claim (2) of Contract as Commodified Promise: Improper Commodification of Promises
The second claim is a negative one. If a promise is not commodified, or not properly commodified,
then this is a strong pro tanto reason against its enforceability in contract law. Let me once again analyze the
claim into sub-issues: [i] why the non-commodified nature of some promises counts as a strong pro tanto
reason against legal enforcement, [ii] why some promissory rights are not properly commodified, [iii] why
improperly commodifying a promise counts as a strong reason against enforcement, [iv] which is pro tanto
rather than decisive.
180

The bulk of this sub-section involves claim [ii], while the next sub-section deals with [iii]. So let me
(relatively) quickly address sub-claims [i] and [iv], first. As for [i], the justification is simple: courts and their
supporting legal apparatus are costly mechanisms with limited resources supported by the public. To the
extent that a contract dispute arises, courts that have this jurisdiction should make sure that the dispute is
something that is within the purview of contract law and its core concerns. Given that contract law primarily
concerns the enforcement of commercial, and hence commodified promises, courts should only sparingly—if
at all—extend its jurisdiction to disputes arising from non-commodified promises. One might respond that
parties can finance their own day in public courts. This is sometimes true. But the court’s “resources” are not
limited to financial ones. Courts expend considerable cognitive and temporal resources, which must be
allocated carefully. This is so even in the event that we increase the numbers of judges and courts. This is not
to say that courts may never hear disputes involving non-commodified promises. The non-commodified
nature of the promise gives only a pro tanto reason against enforceability. But these cases should be considered
rarely.
Now let me address [iv]. This sub-claim is a caveat to the claim that improperly commodifying a
promise counts as a strong reason against enforcement. The issue raised by [iv] is why this reason is pro tanto
rather than decisive. Out of an abundance of caution, I follow Debra Satz in observing that just because some
form of commodification is improper, it does not follow that the law should ban it or otherwise attempt to
de-commodify it in some less extreme way. As I mentioned earlier, even if commodifying sexual services—
prostitution—is morally problematic, it does not follow that we should make prostitution illegal as opposed
to highly regulating it and, say, allowing sex workers to unionize. It might be all-things-considered worse to ban
prostitution. Similarly, it might be worse all-things-considered to decline to enforce a promise that has been
improperly commodified. This is why the fact that a promise is improperly commodified provides only a pro
tanto reason against enforcement.
Now let me try to motivate sub-claim [ii]: So which promises, if any, are not properly commodified,
and why? Which promises are commodified but ought not be? I will not provide a general account of the
conditions under which commodification is more or less morally or politically legitimate. That would be a
181

book unto itself. Instead, let me distinguish two kinds of reasons why commodifying promises might be
improper. The two kinds track the distinction, investigated in Section 4(b), between intrinsic and extrinsic
reasons against commodification. Recall that intrinsic reasons relate the way that commodifying something is in
itself inherently morally problematic, because it might involve, for example: expressing the wrong kind attitude towards
the thing commodified; valuing the thing for the wrong kind of reasons; or degrading or corrupting the target of
commodification or the parties performing the commodification. Extrinsic reasons against commodification
turn on the morally problematic outcomes that commodification tends to bring about, such as harms or
injustices.
My claim is that commodifying promissory rights can be criticized along these two dimensions.
Turning first to extrinsic reasons against commodifying a promise, consider an example of an extrinsically
improperly commodified promise. Suppose that someone promises to sell his kidney to someone else, and
suppose further that the commodification of kidneys is morally improper for intrinsic or extrinsic reasons or
both. Given that the commodification of kidneys is improper; given, that is, that treating kidneys like a commodity,
by attempting to sell them or purchase them is improper, this suggests that any promise that aims to facilitate
this purchase or sale is improper derivatively. So attempting to buy or sell a promise to sell a kidney will
likewise be improper. Promises that are improperly commodified for extrinsic reasons inherit their status as
improperly commodified from the improper commodification of the items in the underlying transaction.
By contrast, intrinsically improperly commodified promises are promises that should not be enforced
given the nature and value of certain promises themselves. More specifically, certain promises have a normative
character that is in tension with, or even incompatible with, their commodification, apart from the content of
an underlying promise. Let me give some examples to try to make this claim plausible:
Alice. Alice promises to help her friend (Amy) move to a new apartment. Amy, who now has
the promissory right to Alice’s performance, attempts to sell this right to Amy’s roommate,
such that Alice will help Amy’s roommate in lieu of Amy herself.
Bill. Bill promises to help a friend to help him move to a new apartment. Bill then pays a
third party to perform the move instead.
182

Cece. Cece pays professional movers $1,000.00 to help her move to a new apartment. The
movers thereafter sub-contract to some third party to perform the move.
I hope to use these cases to motivate the idea that commodifying some promises, apart from their
content, might be intrinsically morally problematic. By “apart from their content” I mean that, in some cases,
commodifying a promise to do x is morally problematic for reasons that do not necessarily depend on the
wrongfulness of commodifying x in the first place, such as in the kidney example mentioned earlier. In other
words, commodifying some promises is subject to intrinsic objections relating to norms that ought to govern
certain kinds of promising.
Let me discuss Alice first. The attempt to sell the promissory right involves commodifying that right.
The content of the right does not make the commodification problematic. Promises to help one move, I take it,
are valuable commodities that the moving industry and would-be movers depend on, and there is nothing
particularly bad about having such a market. So what makes Alice’s behavior morally problematic? A natural
thought is that there is deception in the air or something on par with deception. It is such an unusual thing to
do, to sell a friend’s promissory right, that Alice had no reason to expect. And if selling the right were a live
possibility, then Alice would not have agreed to help move.
This is a plausible explanation but hides a deeper account. Why would Alice agree to help Amy by
moving her, while reasonably decline to help her by either moving her or being liable to help the roommate
(with whom Alice is not a friend)? The sale does not appear to be within the scope of Amy’s presumed
authority.
I think that the nature of promises in the context of friendships has something to do with it. When I
promise a friend, not only is the content of the promise important, but the identity of the promisee and
promisor also matter. It matters that the promisor-promisee relationship obtains between one friend and
another. Part of the point of the promise in the context of friendships is not just to achieve some
instrumental end—moving from point A to point B—but also to reinforce and test the bonds of friendship,
or create those bonds where they do not exist previously. In this context, the prospect of buying or selling
promissory rights defeats this purpose, even if it would advance the goal of moving from point A to point B.
183

This is why I take Bill’s case to be unusual. Arguably, Bill has strictly fulfilled the content of the
promise: to help his friend move. But our conventional understanding of what that means for one friend to
“help” another to move involves more than simply paying someone else to help him move. Notice that it would
seem reasonable for Bill’s friend to ask why he paid movers rather than to come and aid in the physical labor;
it would seem reasonable to expect Bill to answer for his not coming in person to physically help move
furniture and such. Now, Bill might provide a really good answer. Perhaps Bill has the flu or has an
unanticipated job interview that day, and so on. But if Bill says that he simply preferred to play video games,
this is a bad reason to not show, reflecting poorly on how Bill understands friendships. It misses the point, in
some way, of asking a friend to help move, a point that goes beyond saving money in not hiring movers.
Friendship involves, to some extent, a disposition to want to spend time with one’s friend and to share in
burdens together. Commodifying a promise, in this context, is improper because it expresses an attempt to
bypass the burdens of friendship; indeed, it seems to reflect an attempt to buy one’s way out of or monetize
the burdens of friendship
I have assumed that Bill has satisfied the terms of the promise. This assumption is doubtful—but
even so this does not obviate the intrinsic commodification critique. To see why we might doubt that Bill has
fulfilled his promise, consider that, when a promise involves people in a close relationship, the friendship or
close relationship might create certain expectations about the content of the promise. Perhaps part of the
promise presupposed by Bill as promisor and his friend as promisee is that the right to performance includes
not only a right that the move takes place, but that Bill be the mover. This is one important way that certain
promises between members of close relationships may alter the content of the promise by default. So any
wrongness in Bill’s conduct can be simply explained in terms of his breaking the promise, without needing
recourse to commodification.
But even on this analysis the commodification is arguably improper. Bill’s paying someone else to
move simply reflects his way of compensating for breaking the promise. Arguably, Bill’s pre-emptive decision
to pay someone else to move, without consulting his friend beforehand (let’s say), suggests that Bill takes the
following attitude towards his promises: one may break a promise if one is willing to pay a price. But this is
184

not the proper attitude to take with respect towards one’s promissory obligations. When one makes a
promise, one must keep it. One may not unilaterally convert the content of a promise to do x into a promise
to do x or to pay some amount of money.
195
This notion—that it is okay to pay one’s way out of a promissory
commitment, takes the wrong kind of attitude towards one’s promise—hence is yet another way Bill’s treatment
of the promise might be criticized on intrinsic grounds. Attempting to view one’s promissory obligations as
subject to purchase and sale rather than outright performance reflects a degraded attitude towards one’s
status as a promise maker and towards promising more generally.
Contrast Alice and Bill with Cece. I see nothing particularly wrong with what the mover did in this
case. To be sure, if the sub-contractor is one that Cece studiously sought to avoid, then this raises red flags.
But this kind of thing is often to be expected in commercial contexts. Take a common experience. Initially,
you might take out a student loan with one lending institution, only to find out that your loan has been taken
over and managed by a different service provider. Your loan, your promise to repay certain debts, is bought
and sold like a commodity on secondary markets. But so long as the underlying content of the promise does
not change, the identity of the service provider is assumed to matter little.
Other writers have accept the idea hovering in the background of cases like Alice, Bill, and Cece—i.e.,
that some promises take on a different normative character depending on their role in close interpersonal
relationships.
196
Indeed, they argue that the natural home of promises lies in the context of close personal
relationships like friendships between Alice and Amy and Bill and his friend. They ground this claim in the
observation, for example, that we do not normally make promises with perfect strangers, and when we do,
the promises are usually contractual ones, backed by the prospect of legal enforcement. The thought is that
promises have a “natural” home in pre-existing, close, personal relationships between promisor and promisee.

195
To be sure, some promises might appropriately take disjunctive form at the outset (though even here I
have my doubts that this kind of perform-or-pay arrangement is wholly without tension in the context of
close relationships). But unilateral reformation by the promisor arguably involves breaking the promise. For
discussion of this point, see Seana Shiffrin, Must I Mean What You Think I Should Have Said?, 98 VA. L. REV.
159, 162-64 (2012).
196
See, e.g., Aditi Bagchi, Separating Contract and Promise, 38 FLA. ST. U. L. REV. 709 (2011); DORI KIMEL, FROM
PROMISE TO CONTRACT: TOWARDS A LIBERAL THEORY OF CONTRACT (2003).
185

Call promises made within the context of these relationships personal promises. The difficult question is what
any of this has to do with contract law. I will turn to this question towards the end of the next subsection.
c. Claim (2) of Contract as Commodified Promise (cont’d): Improper Commodification as Reason Against Enforcement
Let us take a step back. We have been discussing sub-claim [ii]: why some promissory rights are not
properly commodified, and I argued that there might be intrinsic and extrinsic reasons to oppose
commodification of promises. Now I take up [iii]: the sub-claim that, to the extent that some promissory
rights are not properly commodified, this provides a strong reason against enforcing those rights in contract
law. My argument will be indirect. I will suggest that the common law of contracts already does embody or
reflect the view that certain promises are improperly commodified, and moreover, insist that these legal
norms are worth preserving. In short, contract law already has morally sound doctrines that in effect support
[iii].
Let us begin with the distinction discussed already, between intrinsic and extrinsic reasons to oppose
commodifying certain promissory rights. Consider first extrinsic reasons to oppose commodifying promises,
and in particular, the kidney example mentioned earlier. Recall that the (presumed) fact that it is wrong (for
either intrinsic or extrinsic reasons) to commodify kidneys thereby makes it wrong to promise to sell or
purchase or otherwise engage in some act that promotes or perpetuates the commodification of kidneys. In
other words, certain promissory rights are improperly commodified because the underlying good or service
that is the subject matter of the promise is itself improperly commodified. In these cases, the promissory
right’s improper commodification is derivative; or, in the language of extrinsic and intrinsic reasons against
commodification, the reasons why the promissory right is improperly commodified are extrinsic to the
promise itself. Commodifying the promissory right to a kidney is improper given that commodifying the
kidney is itself improper.
How does this bear on contract law? To the extent that parties try to enforce promissory rights that are
bought or sold, contract law’s jurisdiction is potentially triggered, since there is a strong reason to enforce
commodified promissory rights (as I argued above). Now, contract law plays no direct role in prohibiting or
highly regulating the commodification of kidneys or other bodily organs. That job—determining whether to
186

ban a market outright or highly regulate it—is performed by other areas of law like criminal law. Indeed, the
official policy position of the United States is to strictly prohibit the sale of kidneys under its general
prohibition of the sale of bodily organs.
197
Contract law remains primarily about enforcing commercial,
commodified promises with the aim of facilitating markets. Courts should remain cognizant of contract law’s
subject-matter jurisdiction, so to speak.
198

But contract law is not wholly silent either. It contains doctrines that authorize courts to decline to
enforce certain promissory rights, in effect, that are commodified improperly for extrinsic reasons. The most
obvious doctrine is that contract terms will not be enforced to the extent that those terms are unlawful.
199

Another doctrine holds that contracts are void to the extent that they are contrary to public policy. Under
both doctrines, kidney sales contracts would not be enforceable because selling kidneys is unlawful. This
doctrine authorizes courts to decline enforcing promises, in effect, in cases where promises have been
commodified improperly—at least as judged by other areas of law. To the extent that this doctrine represents
a good feature of contract law worth preserving, it provides a passive way to ensure that courts do not play a
market-facilitating role with respect to certain noxious markets. This is one way in which improper
commodification of a promissory right, at least as judged by other areas of law, indirectly provides a strong
reason (and in the case of illegality, a decisive reason) against enforcing certain promissory rights.
To see why these doctrines are worth preserving, notice that there are independent justifications for
them grounded in the rule of law. These rationales can be explained in terms of incentives: the law generally,

197
See 24 U.S.C. § 274e(a) (“It shall be unlawful for any person to knowingly acquire, receive, or otherwise
transfer any human organ for valuable consideration for use in human transplantation if the transfer affects
interstate commerce.”); see also 24 U.S.C. §274e(c)(1) (“The term ‘human organ’ means the human (including
fetal) kidney, liver, heart, lung, pancreas, bone marrow, cornea, eye, bone, and skin or any subpart thereof and
any other human organ (or any subpart thereof, including that derived from a fetus) specified by the Secretary
of Health and Human Services by regulation.”).
198
I realize that this is a non-standard use of a phrase—subject-matter jurisdiction—that is typically used to
refer to the jurisdiction of courts, not substantive areas of law. I hope that this is not too distracting.
199
See, e.g., Cal. Civ. Code § 1598 (“Where a contract has but a single object, and such object is unlawful,
whether in whole or in part, or wholly impossible of performance, or so vaguely expressed as to be wholly
unascertainable, the entire contract is void.”); see also Cal. Civ. Code § 1599 (“Where a contract has several
distinct objects, of which one at least is lawful, and one at least is unlawful, in whole or in part, the contract is
void as to the latter and valid as to the rest.”).
187

and contract law in particular, should not be designed to as to create self-undermining incentives for unlawful
behavior. Recognizing and enforcing contracts according to which parties are required to engage in illegal
activity encourages people to form and follow through with unlawful contracts. The rule of law also requires
that the state should not place individuals under incompatible legal obligations. But contracts with illegal
terms do precisely this, since contractual obligations are legal obligations, and illegal contractual obligations
are legal obligations to perform illegal acts. And the state should avoid being complicit and dirtying its hands
by facilitating unlawful transactions or expressing approval thereof. The state, when it acts, should strive to
act with integrity.
None of these rule-of-law rationales is incompatible with my point. My point is that contract law
already has doctrines that authorizes courts to avoid enforcing certain attempts to commodify certain
promissory rights on the grounds, albeit derivatively, that those promissory rights should not be commodified.
Contract law’s role of facilitating markets by enforcing certain promises is circumscribed by the law’s
judgment that certain markets should not be facilitated or endorsed. This should suffice to show that
improperly commodified promissory rights provide a strong reason for non-enforcement, provided that
contract law plays some legitimate role in preventing, on extrinsic or derivative grounds, the improper
commodification of promissory rights.
More surprising, I think, is the thought that contract law contains some legal norms according to
which it will decline to enforce attempts to commodify promissory rights for intrinsic reasons, reasons relating
to the nature and value of promises themselves in certain contexts. To explore this thought, the balance of this
section will point out some legal doctrines that can be interpreted as reflecting the notion that contract law, in
at least some jurisdictions, already does decline to enforce certain promises for reasons intrinsic to promising
itself; that is, on the grounds that commodifying certain promises is incompatible with valuing them in the
right way.
My goal in this section and in the balance of this Chapter is modest. I endorsed the doctrine that
contracts should be void as either illegal or contrary to public policy. But in this section and the next, my aim
is simply to show how CCP—and in particular the claim that contract law should avoid enforcing promises
188

that are commodified improperly for intrinsic reasons—helps to interpret or make sense of certain feature of
United States common law doctrine. The next section will argue that we can make sense about recent
scholarly debates about remedies in light of CCP more generally.
The first doctrine worth considering concerns so-called donative promises, which are not
enforceable under United States contract law. The paradigm example involves a promise that the promisee
does not rely upon, but which both parties intend to be legally enforceable. These are essentially gift
promises, usually intended solely to benefit the promisee. For example, if I promise (even in writing) to give
you a car, full stop, or promise to give you a car in exchange for nominal sum that is a transparent ploy to
ensure that the exchange satisfies formally the common law doctrine of consideration, then courts might very
well decline to enforce that promise under the law of contract.
200
This doctrine walks hand-in-hand with the
doctrine of consideration, which holds that in general no contract is valid unless two parties engage in a quid
pro quo exchange of value, though for ease of exposition I will simply focus on the donative promise rule.
Legal scholar Melvin Eisenberg offers a justification for the rule against the enforcement of donative
promises in terms of commodification.
201
After observing that “much of the world of gift is driven by
affective considerations, like love, affection, friendship, gratitude and comradeship,”
202
Eisenberg asserts that
legally enforcing a donative promise would muddy the motivations that a promisor has for keeping her
promise, and that it would “never be clear to the promisee, or even to the promisor, whether the donative
promise that was made in spirit of love, friendship, affection, or the like, was also performed for those

200
See, e.g., Dougherty v. Salt, 125 N.E. 94 (N.Y. 1919) (declining to enforce gift promise to pay $3,000 at
promisor’s death or before to family member “for value received”); Fischer v. Union Trust Co., 101 N.W.
852 (Mich. 1904) (declining to enforce promise to pay two mortgages of value totally $8,000 in exchange for
one dollar); and Schnell v. Nell, 17 Ind. 29 (1861) (declining to enforce promise of $200 “in consideration of
one cent”). Notice that these cases would also flout the intent-to-be-legally-bound requirement, given that the
formal requirements for contract formation obtain, and that the parties manifestly intended to be legally
bound by their promissory commitments, yet the courts nevertheless declined to enforce them.
201
Melvin Aron Eisenberg, The World of Contract and the World of Gift, 85 CALIF. L. REV. 821, 847 (1997).
202
Id.
189

reasons, or instead performed to discharge a legal obligation or avoid a lawsuit.”
203
He claims that enforcing
this kind of promise “would have the effect of commodifying the gift relationship,”
204
adding
[G]ifts made pursuant to simple, affective donative promises would be seriously
impoverished, because at the point of the transfer, the promisor’s motives would be
invariably mixed. … [L]egal enforcement of simple, affective donative promises would move
the commodity rather than the relationship to the forefront, would essentially convert the
promise into a cash equivalent, and would submerge the affective relationship that the gift
was intended to totemize. Simple donative promises would be degraded into bills of
exchange, and the gifts made to keep such promises would be degraded into redemptions of
the bills. To protect a few promisees, and perhaps a few promisors, an enforceability regime
would cut off something very important in social life, and harm donative promisors, and
even donative promisees, as a class.
205

In this passage, Eisenberg here combines several ideas that are usefully separated for our purposes,
including but not limited to the following:
(a) Legally enforcing a donative promise contributes to its commodification.
(b) Commodifying a donative promise degrades it or otherwise fails to value it in the right way.
(c) Legally enforcing a donative promise degrades it or otherwise fails to value it in the right way.

203
The conventional justification for this doctrine involves so-called process problems. The worry is that
someone could simply accuse another person of making a promise to give something valuable as a gift, and
that it would be too difficult or risky to empower people to make these false accusations. Another less-
remarked-upon justification is that this rule represents an anti-circumvention device. Often these faux
contracts or gift promises are made when the promisor is at or near death. But the law of wills, including all
the requisite formalities, should govern post-death transfers of property. Enforcing these promises would
threaten to set up a clash between those named pursuant to properly executed wills and those promised
property via non-will contracts, or create incentives for would-be devisees to try to circumvent the formalities
of the law of wills. The anti-donative-promises rule, in other words, aims to keep the jurisdictional lines
between contract and the law of wills relatively clear.
204
Melvin Aron Eisenberg, The World of Contract and the World of Gift, 85 CALIF. L. REV. 821, 847 (1997).
205
Id.
190

(d) Legally enforcing a donative promise renders the motives for keeping a promise obscure,
whereas it should be clear that keeping a promise should be motived by moral considerations,
love, affection, or something else.
Claim (a) suggests that the mere enforcement of a promise via contract law tends to commodify that
promise. This makes some sense if contract law’s goal is primarily to facilitate markets. The association might
taint other, non-commercial promises, if enforced through contract law. Enforcing non-commercial promises
would do so in accordance with contract law norms that were designed with facilitating markets in mind, not
respecting non-commercial promissory commitments such as those that exist between friends. And recall our
working understanding of commodifying something: treating that thing as subject to market norms. Surely
contract law counts as a set of market norms, given its Instrumental Justification. So it turns out that the very
act of enforcing a promise in contract law involves commodifying that promise to some extent. But the
question would still remain, when should contract law enforce a promissory right or view that right as properly
within its purview? So even if Eisenberg is correct about (a), as I am inclined to think he is, the point he is
making goes beyond it.
It is really (b), the claim that commodifying donative promises, and what this commodification does to
the nature and value of that promise, which performs all the work. Here it is useful to recall the cases of Alice
and Bill and to subject them to an Eisenbergian interpretation. Part of the problem with Amy’s attempt to sell
the right to help Alice move, and part of the problem with Bill’s attempt to buy his way out of the promissory
obligation, lies in the fact that both maneuvers, in Eisenberg’s language, “would move the commodity rather
than the relationship to the forefront, would essentially convert the promise into a cash equivalent, and would
submerge the affective relationship that the gift was intended to totemize.”
206
Rather than making clear that
the promise was made “in [the] spirit of love, friendship, affection, or the like,” we might now wonder
whether Amy made the promise to help Alice move in order to turn around and sell it for some quick cash.
As for Bill, his problem is more fundamental, since his actions are compatible with the attitude of viewing
moral obligations as fungible with cash. It is not legal enforceability per se that brings about this corruption of

206
Melvin Aron Eisenberg, The World of Contract and the World of Gift, 85 CALIF. L. REV. 821, 848 (1997).
191

personal promises and one’s attitude towards promises more generally, it is the commodification (this is also
true in the case of Amy and Alice). We can understand these claims without bringing legal enforceability into
the equation. And the commodification is morally problematic inherently, for reasons intrinsic to private
promises.
Others make points similar to Eisenberg, though they implicitly emphasize claims (c) and (d), and
moreover, make no mention of commodifying promissory rights as playing any important role in that
explanation.
207
So, like Eisenberg, they emphasize the wrong things; they focus on the way that legal
enforceability per se rather than commodification threatens to crowd out or corrupt or degrade personal
promises and the attitudes of those who make and keep them.
208
Notably, Aditi Bagchi places legal
enforceability front and center.
209
According to Bagchi, “[t]he moral character of a [personal] promise
depends on the fact that it is not only freely made but also freely kept.”
210
When a promisor makes a legally
binding commitment, however, “she simultaneously creates a second sufficient reason—liability in the case of
breach—the first reason does no work, or there is no way for the independent sufficiency of the first reason
to manifest itself objectively.”
211
That is, promisees cannot be sure that a contractual promise is kept for any
other reason than simply because the promisor sought to avoid legal liability.
Bagchi also argues that the different background relationships that promises and contract presuppose
give rise to different expectations about the appropriate norms that might govern promise making and,
importantly for contract law, norms that govern in cases of non-performance. Bagchi argues that mere
promises may be both more demanding in some ways and less demanding in others. To illustrate, she asks us
to suppose that a friend has promised to pick us up at the airport, only to discover that we are in a wheelchair

207
See generally DORI KIMEL, FROM PROMISE TO CONTRACT: TOWARDS A LIBERAL THEORY OF CONTRACT
(2003); Aditi Bagchi, Separating Contract and Promise, 38 FLA. ST. U. L. REV. 709 (2011).
208
DORI KIMEL, FROM PROMISE TO CONTRACT: TOWARDS A LIBERAL THEORY OF CONTRACT 58 (2003)
(“Enforceability is built into contract in a way that no equivalent source of reassurance is built into
promise.”); see CHARLES FRIED, FROM PROMISE TO CONTRACT: A THEORY OF CONTRACTUAL
OBLIGATION (1981).
209
Aditi Bagchi, Separating Contract and Promise, 38 FLA. ST. U. L. REV. 709, 710 (2011).
210
Id. at 710.
211
Id. at 712.
192

(having suffered an accident during a ski trip).
212
A friend may be reasonably expected to work with us to find
a way to get home, even if this will be more difficult than expected.
213
A friend may also, to embellish
Bagchi’s example, carry more of the burden by carrying most of the luggage and assisting us in stowing the
wheelchair in our vehicle, and so on. Contrast this with a taxi service operating at arm’s length. Depending on
the terms of service and established commercial practice, the service might reasonably decline service
altogether, or demand additional payment—even though both the friend and service made promises. By the
same token, friendship may be less demanding in some ways. It may be inappropriate to demand payment
from friends for broken promises, even though this is perfectly reasonable in cases involving arm’s length
transactions.
Bagchi’s views are nuanced and I no doubt fail to do them full justice here. And her views as I have
characterized them seem in line with the analyses of Alice, Bill, and Cece cases discussed in the prior section.
But notice that, even on Bagchi’s explanation, legal enforceability need not take a central role in the explanation.
What matters first and foremost is the social context in which the promise is made, and in particular, whether
an attempt has been made to commodify a personal promise. If the promise is a private one between those in
close personal relationships, then treating these personal relationships in accordance with market norms—
regardless of whether those norms are legal or not—simply subjects those promises to the wrong kind of
norms and expectations. It is only once we understand contract law as providing essentially market norms
that, secondarily, we see why it would be inappropriate to subject personal promises to legal enforcement. But
the propriety of commodifying promissory rights is a necessary part of that explanation. We should not
enforce certain private promises via contract law because doing so in itself would be tantamount to improper
commodification of promissory rights—generated by private promises—that ought not be commodified, and
only secondarily for reasons relating to threat of sanction and the like. Or at least as an interpretive matter this
appears to be the implicit justification for the court’s declining to enforce certain promises that otherwise
satisfy contract law’s formalities.

212
Id. at 720-24.
213
Aditi Bagchi, Separating Contract and Promise, 38 FLA. ST. U. L. REV. 709, 722-24 (2011).
193

To bolter this point, that legal enforceability of private promises should be avoided due to the fact
that it tends to commodify them, notice that the prospect for legal enforcement does not tend to “corrupt”
or “degrade” all close intimate commitments. Consider marriage. Marriage is typically regarded as a legal
status reserved only for the closest and most intimate relationships, and moreover, is taken by its participants
as a way of enhancing rather than degrading one’s relationship with a lover. But this is not something we
would expect if legal enforceability itself degrades or corrupts the underlying relationship. So the idea,
expressed in (c) and (d), of whether legal enforceability or recognition itself degrades intrinsically valuable
relationships or the reasons with staying in those commitments looks tenuous.
The upshot is this. Writers like Eisenberg, Kimel, and Bagchi are correct to suggest that some
personal promises should not be enforced in law, or more minimally, to suggest that as an interpretive matter
that United States contract law has reasons for declining to enforce certain promises even when those
promises appear to satisfy certain formalities. But I think that these authors exaggerate the threat posed by
legal enforceability itself to the value of certain promises. The marriage claim bolster this point. Eisenberg comes
closest to the mark, since he recognizes the connection between legal enforceability and commodification,
pointing out that legal enforceability might tend to commodify private promises, and thereby treat those
promises improperly. But even Eisenberg does not go far enough. It is really the intrinsic reasons against
commodifying personal promissory rights, with or without the prospect of legal enforceability, which is the
problem. The problem is treating personal promissory rights in a way that is inappropriate given the role that
they play in close interpersonal relationships. Commodifying these rights and obligations treats them as
fungible assets suitable for purchase or sale. That is not the right way to think about or treat personal
promises. If this is correct, then the various ways that law tries to make room for or accommodate personal
promises—including by declining to enforce them, via the doctrines governing donative promises or some
other mechanisms—reflects a surprising constraint on the promises that contract law will enforce. This
constraint, which reflects a controversial ideology concerning the moral limits of commodification, lies at the
heart of contract law.
194

Let me take a step back to locate where we are in the dialectic. I have been trying to provide support
for the sub-claim [ii] of claim (2) of Contract as Commodified Promise: that, to the extent that some
promissory rights are not properly commodified, this provides a strong reason against enforcing those rights
in contract law. There were extrinsic reasons and intrinsic reasons certain promises should not be
commodified, and in turn, enforceable in law. Extrinsically, promissory rights might be improper to buy and
sell due to the underlying improper commodification that they facilitate. Contract law already does, and
should, decline to enforce these promissory rights and obligations, via the doctrines of illegality and voidness
for reasons of public policy. Promissory rights that seek to secure sales or purchases of, say, kidneys should
not be enforced according to these doctrines. But the primary role of determining what should and should
not be commodified comes from elsewhere in the law.
I then turned to the less obvious claim that there are certain reasons intrinsic to certain promissory
rights and obligations that make them improperly commodified, and more importantly, that this provides a
strong reason against enforcing these promises. I pointed out certain doctrines, like the donative promises
rule, which arguably already prevent for intrinsic reasons commodifying certain personal promises. The state
should not allow itself to be party to attempts to degrade or corrupt personal promises, or allow this sphere
of interpersonal morality to become usurped by market norms. Alternatively, the law must at least make room
for non-market norms to flourish. As an interpretive matter, the law reflects the view that the law should
strive not to enforce personal promises, especially those that arise in the context of, and are partially
constitutive of, close, personal, and intimate relationships.
6 Contracts as Commodified Promises Meets the Contract Remedies Debates
In addition to providing an answer to the issue of which promises should be enforced, Contract as
Commodified Promise offers a new lens through which to view existing debates in contract theory. More
specifically, I think that the view that contracts involve commodified promises can be brought to bear on
ongoing debates about the proper remedies for breach of contract. More specifically still, I think that
understanding contracts as commodified promises provides normative resources that Robust Promise
Theories, a subset of Promise Theories, can use in these debates. Although I will not attempt to resolve any of
195

them, viewing them through the lens of promissory commodification can bring a new and potentially fruitful
perspective to these disagreements.
Recall Robust Promise Theory. It holds that the common law of contract fails to adequately respect
promissory rights qua promissory rights. For example, according to Seana Shiffrin contract law (roughly) fails
to deter or express appropriate condemnation towards opportunistic breaches. This because there is a strong
default rule of contract law that awards expectation damages, even if the promisor breaks his contract in
order to secure a slightly better deal. There is also a strict rule against awards of punitive damages, which
prevent victims of broken contracts from suing in a way to express condemnation. Nor can specific
performance be demanded, save for exceptional circumstances. Finally, to add insult to injury, the promisee-
victim may be forced to try to minimize even the compensatory damages owed by the promisor, give the so-
called duty to mitigate damages. As we also saw in Chapter One, Shiffrin has criticized this collection of
doctrines on the grounds that the rationales proffered in their support fundamentally disrespects promissory
morality or otherwise fails to take or express the right attitude towards it.
214
However these complaints are
cashed out (and I canvassed a number of possibilities in Chapter One), Robust Promise Theory favors
reforming the common law of contract to better harmonize with the demands of promissory morality.
Whatever one makes of these concerns, notice that Contract as Commodified Promise provides new
normative resources for the Robust Promise Theorist. Ironically, their criticisms would not be that contract
law fails to respect promises, but rather that contract law as currently in place either fails to respect the
commodified nature of promissory rights or perversely encourages improper commodification of promissory
rights. That is, Shiffrin’s concerns can be bolstered by concerns that the law does not commodify promises
enough or that it threatens to over-commodify promissory rights and obligations.
These argumentative strategies are of course incompatible but they might give rise to a dilemma for
defenders of status quo remedial regime in the common law of contract. Begin with first horn of the dilemma:
the idea that contract law fails to adequately respect the nature of promissory rights qua commodities. This

214
Seana Shiffrin, Must I Mean What You Think I Should Have Said?, 98 VA. L. REV. 159, 162-64 (2012); see also
Seana Shiffrin, The Divergence of Promise and Contract, 120 HARV. L. REV. 708 (2007).
196

horn needs an additional, contestable assumption to get going: that if contract law does recognize, implicitly
or explicitly, that promissory rights are treated as commodities, and if all commodities are forms of property (this is
the new assumption), then contract law’s remedial regime protects these proprietary rights very weakly as
compared to other proprietary rights. The Robust Promise Theorist could then insist that we ought bring
greater consistency in how the legal system protects property rights, given background demands that the law
ought to act with systemic consistency.
To see why, assume for the sake of argument that all commodities—all things available for purchase
or sale—are a form of property. Now notice the standard civil remedies available for violations of property
rights. Protection of property rights—both tangible and intangible—are associated with strong forms of
injunctive relief. If someone is trespassing on land, intentionally or not, a court order permitting the forcible
removal of the trespasser is a remedy available as a matter of course. When intangible property rights are at
issue injunctive relief is similarly available: websites that empower pirates to illicitly trade copyrighted works
can be forced to close, and illicitly printed books might literally be burned, despite otherwise very strong free
speech protections. Moreover, property rights, when they are intentionally violated, can be enforced with
punitive damages.
But if contractual rights are, as we have assumed for the sake of argument, simply a form of property
rights, then they are treated incongruously in common law jurisdictions. Working backwards, punitive
damages are not available in contract, even for intentional or opportunistic breaches of contract, in which
someone breaches a contract in order to pursue a better opportunity. Injunctive relief (specific performance)
for non-performance of contractual obligations is rare; the standard remedy is expectation damages. Again,
this might seem incongruous if promissory rights were themselves a form of property right; it is tantamount
to allowing the breaching party to appropriate the promisee’s property right without that person’s consent.
Of course these features of contract law are peculiar to (most) common law jurisdictions. Civil law
jurisdictions, at least formally, appear more receptive to both regular issuance of specific performance (i.e.,
injunctive relief to enforce contractual obligations) and punitive damages. So if Contract as Commodified
Promise is correct, and if commodities are a form of property (necessarily), then this provides some prima facie
197

case in favor of adopting a civil law approach to contract law remedies. Doing so better respects property
rights and their consistent treatment under law.
But as they say, one person’s modus ponens is another’s modus tollens. All of this suggests a response to
the prima facie argument for reforming contract law. The rules against punitive damages, and the rule
disfavoring specific performance, might be legal rules that implicitly indicate that contractual promises,
though commodities fit for purchase or sale, are either not property rights or, even if they are, are at best
partial or incomplete commodities. Indeed, perhaps the common law remedial regime is itself designed to
prevent improper commodification. Recall from Chapter One a concern motivating weak enforcement of
specific performance and non-enforcement of punitive damages. Routinely allowing plaintiffs to sue to force
individuals to perform certain personal services contracts, against their will when defendants are willing to pay
instead, looks like forced labor. This, in turn, treads too closely to the paradigm example of improper
commodification, chattel slavery. More generally, when proprietary rights in another person’s labor is
concerned, there is good reason for the remedies used to enforce those rights should not be too rigorous.
Contract law should not be used as an instrument to enlist others in de facto slavery or indentured servitude, or
anywhere close. The commodification that contract law indulges, the fact that labor is turned into expectation
damages or reliance damages in lieu of actual court-enforced labor, actually aims to prevent improper
commodification rather than to undermine proprietary rights.
But Robust Promise Theorists can also use the notion of improper commodification as a resource to
argue for certain other reforms. Recall that perhaps Robust Promise Theory worries about the law’s
promoting or expressing the wrong kind of attitude towards promissory morality; that by transforming certain
contractual obligations into cash equivalences or by declining to allow plaintiffs to sue promisors for punitive
damages or (rarely) specific performance, this encourages the attitude that promissory obligations are
themselves things properly bought and sold. Now, in my view, I do not see anything particularly wrong with
this position, so long as personal promises are not at issue. There is, as already discussed, a certain kind of
impropriety in trying to buy and sell or otherwise commodify personal promises. But as long as it is clear that
198

the underlying transaction is properly commodified, I see little problem in encouraging this attitude towards
those promises.
Still, there is room to respond. We can now re-interpret the Robust Promise Theorist’s complaint in
terms of a view that universalizes personal promises, treating all promises as if they were personal promises.
The commodification of promises in general is inherently morally problematic. And encouraging the attitude
that promissory rights are things properly capable of being purchased or sold is in turn inherently
problematic. Let me quickly add that I doubt that Robust Promise Theorists would accept this
recharaterization of their view, or even that they would accept it as an independent way of bolstering their
position. The view looks radical. Contract law as a whole, to the extent that it is primarily concerned with
enforcing commercial (and hence commodified) promises, looks like it is engaged in an inherently morally
problematic enterprise. But the commodification critique resonates with the kinds of concerns that Seana
Shiffrin raises in her work.
Again, it is not my present aim to settle the dispute between those who endorse a more robust
enforcement regime and those who think that the current array of remedies works just fine. Instead, I want to
show merely that CCP provides a new lens through which to view the debate and perhaps to make it more
tractable: that the debate can be fairly characterized in terms of a debate about the commodification of
promises. And even if we do not re-characterize the debate, the view that contracts do and should involve
commodified promises primarily might bring new resources to those engaged in those debates. Those who
tend to favor stronger protection of contractual rights thereby, without necessarily intending to do so, in
effect support greater commodification of promises—i.e., treating promises even more like paradigm real or
personal property or even intellectual property. They also implicitly assume that all commodities are a form of
property. By contrast, those who resist changing the status quo often, and without perhaps realizing it, defend a
regime that holds that promises are only very partially commodified. And this might be a good thing.
7 Conclusion: An Objection
This has been a tentative sketch, perhaps sketched too quickly, of a view that attempts to marry an
instrumental, market-based justification for contract law and a promise-based one. The Instrumental
199

Justification for contract law holds that contract law aims to facilitate markets by enforcing promises. This
answers one of the main stock objections to Promise Theory: why enforce promises? The answer is: because
doing so facilitates market economies. The Instrumental Justification also points towards an answer to the
second stock objection: which promises should be enforced? The answer: primarily commercial promises.
Much of this Chapter has taken great pains to explicate the idea of a commercial promise in terms of
a commodified promise, and in doing so, has enabled me to articulate the core doctrines of Contract as
Commodified Promise. The doctrines are twofold. First, the fact that a promise has been commodified
provides a pro tanto reason favoring legal enforceability in contract law; and second, the fact that a promise has
not been commodified, or has been improperly commodified, provide pro tanto reasons against enforceability.
Although I present CCP as a view that holds independently of any particular jurisdiction’s contract law, some
of the justification for these views draws from existing doctrines in United States law that I regard as generally
good or worth preserving. The argument is thus, in part, interpretive: part of the support for the doctrine
comes from its ability to make sense of aspects of legal doctrine that I take to be justifiable.
But I want to conclude with an objection, one that might have occurred to the reader much earlier.
Why should the state make any judgment at all about which promises are or are not properly commodified?
True, perhaps contract law should not facilitate the purchase or sale of kidneys, or any other illicit trade
banned by some other area of law like criminal law. But set aside such extrinsic cases where the law takes a
position on whether to enforce promises. Focus on the intrinsic reasons against enforceability. Why not hold
a much more modest view of CCP, according to which contract law takes no independent view on whether
or not promises are properly intrinsically commodified. If friends or family want to corrupt their relationships
or promises by commodifying them, isn’t it their business and no one else’s? And if it is solely their business,
this calls for a major revision of CCP: first, the fact that a promise has been commodified is a pro tanto reason
in favor of enforcing it under contract law, and second, the fact that a promise has not been commodified is a
200

pro tanto reason against enforcement. This formulation effectively strips ‘properly’ from the formulation
entirely.
215

I have no general objection to this view. I see it as another plausible version of CCP. Nevertheless, to
say that I have no general objection is not to say that I necessarily endorse it. My version of CCP obviously
holds that there is a place for contract law to weigh in on the enforceability of certain promises in virtue of the
intrinsic problems faced with that enforcement. The other version of CCP provides a more hands-off role for
contract law. But recall that, in pursuing the market-based Instrumental Justification, one must respect
promisor and promisees statuses as such; one must respect, accordingly, the demands and expectations of
promissory morality. Some of these demands and expectations, it turns out, relate to the question of the
circumstances under which promissory rights and obligations are permissibly bought or sold, even if the promisor
and promisees legally consent to the purchase or sale. There might be strong overriding reasons to make courts
available for promissory disputes that arise in the context of friendships; recall that I claim only that the fact
that a promise is improperly commodified provides a strong reason against commodification. But to say that
contract law should say nothing about the matter, in all cases, seems too strong.
Contract law, given its commercial justification, should enforce first and foremost commercial
promises and screen out non-commercial ones. But it should also be sensitive to promissory morality and
claims that certain promises ought not be in the commercial realm even when the parties themselves want
them to be. Otherwise there is some risk, however slight, that noxious markets in improperly commodified
promises will develop. And this is a core part of CCP. But this is surely not my last word on developing the
view. I am open to revision. And I am eager to investigate, in future work, the practical implications of
Contract as Commodified Promise.

215
I thank Steve Bero for this thoughtful worry.
201

Works Cited
Anderson, Elizabeth. Is Women’s Labor a Commodity?, 19 Phil. & Pub. Affairs 71 (1990)
Atiyah, P.S. Promises, Morals, and the Law (1981)
Bagchi, Aditi. Separating Contract and Promise, 38 Fla. St. L. Rev. 709 (2011)
Barnett, Randy E. Contract is Not Promise; Contract Is Consent, 45 Suffolk L. Rev. 647 (2012)
Bix, Brian H. Contract Law: Rules, Theory, and Context (2012).
Chang, Ruth. Commitments, Reasons, and the Will, in 8 Oxford Studies in Metaethics 74 (2013)
Corbin, Arthur L. Corbin on Contracts § 1204 (2006)
Craswell, Richard. Contract Law, Default Rules, and the Philosophy of Promising, 88 Mich. L. Rev. 489 (1989)
Eisenberg, Melvin A. The World of Contract and the World of Gift, 85 Calif. L. Rev. 821 (1997).
Enoch, David. Giving Practical Reasons, 11 Phil. Imprint 1 (2011)
Farnsworth, E. Allan. Contracts § 12.18 (4th ed. 2004)
Finnis, John. Natural Law and Natural Rights (1980)
Fried, Charles. Contract as Promise Thirty Years On, 45 Suffolk. U. Law Rev. 961 (2012)
Fried, Charles. Convergence of Contract and Promise, 120 Harv. L. Rev. F. 1 (2007)
Fried, Charles. Contract as Promise: A Theory of Contractual Obligation (1981)
Graves, Jack. Penalty Clauses as Remedies: Exploring Comparative Approaches to Enforceability, 29 Touro L.
Rev. 681 (2013)
Hart, H.L.A. Punishment and Responsibility (1968)
Heuer, Ulrike. Promising – Part 2, 7 Phil. Compass 842 (2012)
Hohfeld, W.N. Fundamental Legal Conceptions (1919)
Hume, David. A Treatise on Human Nature (1978) (eds. L.A. Selby-Bigge, P.H. Nidditch)
Kaplow, Louis & Steven Shavell. Fairness Versus Welfare (2006)
Keating, Gregory C. Strict Liability Wrongs, in Philosophical Foundations of the Law of Torts 292, 296 (2014)
Kimel, Dori. From Promise to Contract: Towards a Liberal Theory of Contract (2003)
Kolodny, Niko and R. Jay Wallace, Promises and Practices Revisited, 31 Phil. & Pub. Affairs 119 (2003)
Kraus, Jody. The Correspondence of Contract and Promise, 109 Colum. L. Rev. 1603 (2009).
Liberman, Alida. Book Review, 124 Ethics 201 (2013) (reviewing David Owens, Reshaping the Normative
Landscape (2012))
Linzer, Peter. On the Amorality of Contract Remedies — Efficacy, Equity, and the Second Restatement, 81
Colum. L. Rev. 111 (1981)
202

Markovits, Daniel & Alan Schwartz, The Myth of Efficient Breach: New Defenses of the Expectation
Interest, 97 Va. L. Rev. 1939 (2011)
Marmor, Andrei. An Institutional Conception of Authority, Phil. & Pub. Affairs 238, 238 (2011)
Marmor, Andrei. Social Conventions (2009).
Menetrez, Frank. Consequentialism, Promissory Obligation, and the Theory of Efficient Breach, 47 UCLA
L. Rev. 859 (2000)
Moore, G.E. Principia Ethica (1903)
Murphy, Liam, Contract and Promise, 120 Harv. L. Rev. F. 10 (2007)
Oman, Nathan. Markets as the Moral Foundation for Contract, 98 Iowa L. Rev. 183 (2012)
Oman, Nathan. Specific Performance and the Thirteenth Amendment, 93 Minn. L. Rev. 2020 (2009)
Owens, David. Does a Promise Transfer a Right? in Philosophical Foundations Contract Law (Gregory Klass, et
al., eds. 2014).
Owens, David. Shaping the Normative Landscape (2012)
Pratt, Michael G. Contract: Not Promise, 35 Fla. St. U. L. Rev. 801 (2007-2008).
Rawls, John. A Theory of Justice (1999)
Rawls, John. Two Concepts of Rules, 64 Phil. Rev. 3 (1955)
Raz, Joseph. Is There a Reason to Keep a Promise?, in Philosophical Foundations of Contract Law (Gregory Klass,
et. al., ed. 2014)
Raz, Joseph. The Authority of Law (2d ed. 2009)
Raz, Joseph. Practical Reason and Norms (1999)
Raz, Joseph. Morality of Freedom (1986)
Raz, Joseph. Promises in Morality and Law, 95 Harv. L. Rev. 916 (1982).
Raz, Joseph. Promises and Obligations, in Law, Morality, and Society: Essays in Honor of H.L.A. Hart (P.M.S.
Hacker and Joseph Raz, eds. 1977)
Ripstein, Arthur. Force and Freedom: Kant’s Legal and Political Philosophy (2009)
Radin, Margaret Jane. Contested Commodities (1996)
Rosati, Connie. The Importance of Self-Promises, in Promises and Agreements: Philosophical Essays (Hanoch
Sheinman, ed. 2011)
Rowan, Solène. Reflections on the Introduction of Punitive Damages for Breach of Contract, 30 Oxford J. of
L. Studies (2010)
Sandel, Michael. What Money Can’t Buy (2012)
Satz, Debra. Why Some Things Should Not Be For Sale (2010)
203

Scanlon, T.M. Promises and Contracts, The Theory of Contract Law: New Essays (Peter Benson, ed., 2001)
Scanlon, T.M. What We Owe to Each Other (1998)
Schwartz, Alan & Robert Scott, Contract Theory and the Limits of Contract Law, 113 Yale L.J. 541 (2003)
Shiffrin, Seana. Must I Mean What You Think I Should Have Said?, 98 Va. L. Rev. 159 (2012)
Shiffrin, Seana. Immoral, Conflicting, and Redundant Promises, in Reasons and Recognition: Essays on the
Philosophy of T.M. Scanlon (R. Jay Wallace, et al., eds. 2011)
Shiffrin, Seana. Promising, Intimate Relationships, and Conventionalism, 117 Phil. Rev. 481 (2008)
Shiffrin, Seana. The Divergence of Contract and Promise, 120 Harv. L. Rev. 708 (2007)
Simpson, Laurence P. Punitive Damages for Breach of Contract, 20 Ohio St. L. J. 284 (1959)
Smith, Stephen A. Contract Theory (2004)
Southwood, Nicholas & Daniel Friedrich, Promises Beyond Assurance, 144 Phil. Studies 261 (2009)
Thaler, Richard H. & Cass R. Sunstein, Nudge (2008)
Thompson, Judith Jarvis. The Realm of Rights (1990)
Watson, Gary. Promises, Reasons, and Normative Powers, in Reasons for Action (David Sobel and Steven Wall,
eds., 2009)
Wenar, Leif. Rights, Stanford Encyclopedia of Philosophy (2015), http://plato.stanford.edu/entries/rights/

Cases Cited
Applied Equipment Corp. v. Litton Saudi Arabia Ltd., 869 P.2d 454, 515 (Cal. 1994)
Dougherty v. Salt, 125 N.E. 94 (N.Y. 1919)
Fischer v. Union Trust Co., 101 N.W. 852 (Mich. 1904)
Hurd v. Hodge, 334 U.S. 24 (1948)
Pac. Mut. Life Ins. Co. v. Haslip, 499 US 1 (1991)
Romney v. Fransiscan Med. Group, 349 P.2d 32 (Wash. Ct. App. 2015)
Schnell v. Nell, 17 Ind. 29 (1861)
Vincent v. Lake Erie Trans. Co., 124 N.W. 221 (Minn. 1910)
Whiten v. Pilot Ins. Co., [2002] 1 S.C.R. 595
Williams v. Walker-Thomas Furniture Co., 350 F.2d 445 (D.C. Cir. 1965)

Statutes Cited
Cal. Civ. Code § 1598
204

Cal. Civ. Code § 1599
24 U.S.C. § 274e(a)
24 U.S.C. §274e(c)(1)

Abstract (if available)

Abstract Promise Theories of contract law aim to justify and explain certain central features of that branch of private law, while still allowing room to criticize and call for the reform of less central features, especially when those features are at odds with the demands of promissory morality. This dissertation argues that the plausibility of Promise Theory depends on which theory of promissory obligation that the Promise Theorist presupposes. To bolster this claim, the dissertation evaluates several leading philosophical theories of promissory obligation and points to difficulties they face in grounding a Promise Theory contract law. The final chapter turns to more fundamental challenges facing Promise Theory, which is that they do not explain why the law should concern itself with enforcing promises, or which promises should be enforced if at all. I argue that Promise Theory must humble itself by relying on an instrumental view about the point of a system of contracts, which is to maintain a market economy. In explaining what this implies, I argue that contracts involve commodifying promises, and that the moral worries raised by commodifying things generally and commodifying promises in particular inform and illuminate otherwise puzzling features of the common law of contract, such as the doctrine of consideration and the non-enforceability of donative promises.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Responsibility and the emotional structure of relationships

PDF

On being bound: law, authority, and the politics of obligation

PDF

Describing authority

PDF

The mental states first theory of promising

PDF

Units of agency in ethics

PDF

Aggregating complaints

PDF

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

PDF

Rethinking reductive realism in ethics

PDF

Responding to harm

PDF

A perceptual model of evaluative knowledge

PDF

Aggregating happiness: seeking and identifying a single plausible unifying theory

PDF

On law enforcement officers’ liability to defensive harm

PDF

Belief as credal plan

PDF

The constitution of action

PDF

A deontological explanation of accessibilism

PDF

Process-oriented rationality

PDF

Getting our act(s) together: the normativity of diachronic choice

PDF

The factors present in an outperforming charter middle school: a case study focusing on promising practices, school leadership, and cultural norms

PDF

Telling each other what to do: on imperative language

PDF

Enabling human-building communication to promote pro-environmental behavior in office buildings

Asset Metadata

Creator Encarnacion, Erik (author)

Core Title Contract, from promise to commodity

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Philosophy

Publication Date 07/14/2016

Defense Date 05/02/2016

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

Tag agency theory,commodity,consideration,Contract,interest theory,Law,normative powers,OAI-PMH Harvest,Promise,property,punitive damages

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Marmor, Andrei (committee chair), Keating, Gregory C. (committee member), Quong, Jonathan (committee member), Stone, Rebecca (committee member), Watson, Gary (committee member)

Creator Email eencarna@usc.edu,erik.encarnacion@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-268302

Unique identifier UC11279506

Identifier etd-Encarnacio-4548.pdf (filename),usctheses-c40-268302 (legacy record id)

Legacy Identifier etd-Encarnacio-4548.pdf

Dmrecord 268302

Document Type Dissertation

Format application/pdf (imt)

Rights Encarnacion, Erik

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