Introduction

Abstract

By making claims about obligations or ideals for which persons should strive, moral theories inevitably make assumptions about human abilities. This chapter argues that in the case of contract theories these include assumptions about social cognition: our abilities for perceiving and thinking about others. It introduces these assumptions by means of a discussion of the moral contract theories of T.M. Scanlon and David Gauthier.

1.1 Introduction

Moral theory is a normative discipline. Unlike psychology and other sciences that study how humans in fact behave, moral theory’s primary question is how we should act. This normative question cannot be answered by empirical findings alone: is does not imply ought. But findings on human behaviour are relevant in another way. It is widely accepted that morality must be compatible with our nature. Morality must be within the reach of actual persons. Is does not imply ought, but ought does imply can. By putting forward obligations or ideals for which persons should strive, moral theories make assumptions about human abilities. Psychology and associated fields that study human nature can provide information about whether such assumptions are likely to be correct or not. Assumptions can either conform to the findings in these fields, in which case they are empirically plausible , or contradict them in some way, and thus be empirically implausible. A moral theory is empirically plausible only if its assumptions about human abilities are in accordance with empirical findings.¹

Moral philosophers have always made substantial empirical assumptions, but these were seldom the object of systematic empirical scrutiny (Doris and Stich 2011). This situation has changed dramatically in the last two decades. Researchers in the interdisciplinary field of moral psychology , including both psychologists and philosophers, have begun to examine the plausibility of ideas about our moral functioning in the light of empirical findings. This has led to debates regarding the empirical plausibility of several important approaches in moral theory, including virtue theory (Casebeer 2003; Doris 2002; Harman 1999), utilitarianism (Greene and Baron 2001; Greene et al. 2008), and Kantianism (Berker 2009; Greene 2008; Mikhail 2008). One approach that has, despite its recent rise to prominence, received little attention from moral psychologists is social contract theory. It is my aim to change this.

With the publication of A Theory of Justice, John Rawls reintroduced the idea of the social contract into contemporary moral thought. Rawls uses this idea to determine which social arrangements are justified. The crucial idea is that social arrangements are justified if and only if they would be the object of agreement among appropriately situated persons who are choosing their terms of interaction. Rawls’s approach has had an enormous appeal, and the idea of the social contract plays an important role in the work of prominent theorists such as David Gauthier (1986), T. M. Scanlon (1998), Stephen Darwall (2006) and Derek Parfit (2011). While these theorists interpret the idea of the social contract in different ways, they all hold that whether an action is right or wrong depends on whether the action, or more typically the principles that permit the action, would be the object of agreement.

Contract theory shares some of the psychological assumptions of other moral theories. Like Kantians and utilitarians, contract theorists assume that persons can form moral judgments through reasoning. Also, like Kantians and rule-utilitarians, contract theorists typically assume that persons can subsume individual actions or arrangements under general rules or principles. However, contract theorists also make certain assumptions not shared by other theorists. This is particularly so with regard to our abilities for thinking about others and their perspectives, on which I will therefore concentrate.

Contract theorists hold that to judge whether an action or institutional arrangement is morally justified, one must determine whether it is in conformity with principles that would be the object of agreement. They thus assume that persons are able to discern the content of this hypothetical agreement. They thereby assume, I will argue, that persons are able to determine the acceptability of prin ciples from other perspectives than their own present point of view. This is one out of two assumptions on which my investigation regarding the empirical plausibility of contract theory will concentrate.

A second assumption is made, at least by some contract theorists, when explaining why we have reason to be moral; that is, why we have reason to perform actions and adhere to arrangements that are in conformity with principles that would be the object of agreement. Some contract theorists argue that it is in each individual’s interest to comply with such principles because otherwise one may be excluded from beneficial cooperative interactions. The assumption is that potential interaction partners can detect whether one can be trusted to comply or not, and as such will refrain from interacting cooperatively with persons who are not disposed to comply.

Both assumptions have been criticised. A persistent criticism of contract views is that, in contrast to what their defenders suggest, it is not clear what principles would in fact be the object of unanimous agreement (Braybrooke 1987; Gauthier 2003; Hare 1973). It is a well-known complaint against Rawls’s contract theory, for example, that it is not evident that persons situated in the original position would come to agree on his two first principles. Whereas some critics only argue that they would agree on alternative principles (e. g. Harsanyi 1975), other critics have argued that we are unable to discern what persons with wholly different perspectives from ourselves would come to agree on (e.g. Hare 1973). ² They thereby question the plausibility of the first assumption.

The second assumption, which is mainly made by David Gauthier, has proved to be particularly controversial (Buchanan 1990; Franssen 1994; Nelson 1988; Sayre-McCord 1989). While we may expect that persons who violate moral principles will be detected now and then, it is far from evident that the likelihood of detection is so high that it is typically advantageous to comply. Many have found Gauthier’s assumption that persons can be recognised as being untrustworthy even before they have committed a violation particularly unbelievable. In the words of one critic, “[p]eople cannot see inside each other’s heads and it is idle to examine models in which they can” (Binmore 1993, p. 138).

Neither proponents nor critics of these assumptions of contract theory have, however, turned to empirical evidence to make their case. Given that both assumptions concern our psychological abilities, empirical studies may help to resolve controversies such as these. I will in particular be concerned with studies in the field of social cognition . Psychologists use this term to refer to capacities and processes involved in perceiving, interacting with, and thinking about others. Social cognition is a central topic of research in psychology, and is also pursued by economists, neuroscientists, linguists, computer scientists, and logicians. There is much empirical work available which, I shall show, sheds light on the plausibility of the two assumptions, and thereby on the empirical plausibility of the social contract approach towards morality.

With this investigation, I hope to contribute to the philosophical debate in three ways. First and foremost, my aim is to contribute to the development of social contract theory. A theory cannot be plausible as a moral theory if it is not empirically plausible—if its assumptions about human abilities are not in accordance with empirical science. Secondly, the investigation will be of relevance for moral theory more generally. Most if not all moral theories make assumptions about our social cognitive capacities . As a diverse range of empirical studies will be considered, the study will also provide information about the plausibility of other assumptions than just the two on which it concentrates. Finally, the investigation will be relevant for the field of philosophical moral psychology . Whereas moral psychologists have debated about what empirical evidence on character, reasoning, and emotions show about our moral functioning, social cognition has received less attention. Moral cognition and social cognition are closely related, however: moral thinking usually concerns others. This study will bring to the fore empirical work relevant for understanding our moral functioning that has not yet received sufficient attention from moral psychologists.

This first chapter will be mostly concerned with describing contemporary contract theory. This will allow me to further explain the nature of the above two assumptions for contract theory and their role in contract theory. My discussion of contemporary contract theory in the following section focuses on the theories of Rawls, Gauthier, and Scanlon, who are the key figures in this tradition. Although Rawls’s theory of justice is not itself an object of investigation, due to his influence on the moral contract theories of Gauthier and Scanlon it is helpful to take this theory as a starting point. In the third section I briefly return to the two assumptions; they will be further explicated in later chapters. The fourth section gives an overview of the book.

1.2 Contemporary Contract Theory

While the idea of a social contract can be found in the works of Thomas Hobbes , John Locke , and Jean-Jacques Rousseau , their use of it differs from that of contemporary contract theorists (D’Agostino et al. 2011; Kymlicka 1993). Classical contract theorists used the idea of agreement mainly to explain political obligation. The central idea was that if citizens have (or could have) given their consent to the establishment of their government or to the laws of their government, then they have an obligation to obey government and abide by its laws. Contemporary contract theorists, on the other hand, use the idea of agreement to identify which political and social arrangements are justified. The crucial idea here is not that agreement generates obligation, but that it reveals “what we have reason to do in our social and political relations” (Freeman 2007, p. 19).

After Rawls resurrected the idea of the social contract in the second half of the twentieth century, it was developed in different ways. Contract theorists are a diverse lot, including important thinkers such as Binmore (2005), Buchanan (1975), Darwall (2006), Freeman (2006), Gauthier (1986), Hampton (1993), Harsanyi (1977), Kavka (1986), Narveson (1988), Parfit (2011), Scanlon (1998), and Southwood (2010). It has become common to distinguish two strains in contemporary contract theory, one inspired by Hobbes and another more clearly influenced by Rousseau and Kant. This latter strain is often called contractualism , whereas the term contractarianism, although previously used to describe contract theory in general, is now typically reserved for the Hobbesian strain. While the central difference between these strains will play a role in the discussion below, I shall usually be concerned with contract theory in general. I will use the term ‘contract theorist’ to refer to thinkers of either strain who study and develop moral conceptions centred around the idea of agreement.

I shall in the following section explain contemporary contract theory further through a discussion of Rawls’s theory. I then turn to the contract theories of Gauthier and Scanlon, as these will be the focus of my investigation. Gauthier is the most influential defender of Hobbesian contractarianism while Scanlon is, after Rawls, the best known proponent of the sort that takes after Rousseau and Kant—he is also the one who coined the term ‘contractualism’ . A central difference between Rawls’s theory on the one hand and Gauthier’s and Scanlon’s on the other is that Rawls concentrates on political institutions, whereas Gauthier and Scanlon concentrate on interpersonal morality. Aspects of the theories of Gauthier and Scanlon most relevant for my purposes will be more extensively discussed in the later chapters.

1.2.1 Rawls

In A Theory of Justice, Rawls sets out to establish what it is for a society to be just. Society, Rawls says, can be thought of as an association of persons who in their relations to one another recognise certain rules of conduct as binding. These rules specify a system of cooperation with the purpose of advancing the good of those taking part in it. Society is thus “a cooperative venture for mutual advantage ” (p. 4). But while all members of society are better off through social cooperation than if each relied solely on his own efforts, society typically also includes a conflict of interest. Persons are not indifferent as to how the benefits produced by their collaboration are distributed: each prefers a larger to a smaller share of the cooperative surplus. Put differently, they are not indifferent as to how the system of rules that constitute the basic institutions of society distributes the cooperative surplus—regarding what Rawls calls ‘the basic structure of society’. A society therefore requires a set of principles which specify the basic structure of society. These principles constitute the society’s conception of justice. But having a conception of justice, which is common to all societies, does not yet make a society just, Rawls emphasises. A society can only be just if it has the appropriate conception of justice, one that is justifiable to its citizens. Rawls’s aim is to identify what principles constitute such a conception—to identify what he calls the principles of justice.

Rawls uses the idea of the social contract for this purpose. According to Rawls, the principles of justice are those principles that persons who are jointly determining their terms of interaction would agree to. In his words, “[t]hey are the principles that free and rational persons concerned to further their own interests would accept in an initial position of equality as defining the fundamental terms of their association” (p. 10). Rawls thus uses the idea of agreement to fix both the content and the rationale of the principles of justice: the principles that would be agreed to are the principles of justice and they are so because they would be agreed to. As he puts it, the idea is to settle “the question of justification … by working out a problem of deliberation” (Rawls 1999, p. 16).

Rawls’s use of the notion of agreement in this way is typical for contemporary contract theory in general. Contract theorists take the act of agreement to indicate what reasons persons have—they take it to be reason-revealing (D’Agostino et al. 2011). As D’Agostino et al. (2011) put it, “if individuals are rational, what they agree to reflects the reasons they have”. If a principle or a social arrangement would be the object of unanimous agreement among rational persons, we know they have reason to accept it, and thus that it can be justified to them.

It is in part for this reason that Rawls does not rely on the idea of actual agreement but on that of hypothetical agreement. Principles of justice are principles that would be the object of agreement. Persons do not need to actually agree on a social arrangement to have reason to endorse it.

However, contemporary contract theories such as Rawls’s are hypothetical in another way that may be thought to contradict the purpose of revealing our reasons. As we saw, Rawls does not hold that principles are justified if they would be agreed upon by us in our actual circumstances, but if they would be agreed upon by free, rational and equal persons who are choosing their terms of interaction. In general, contract theories take arrangements to be justified if they would be the object of agreement among idealised parties in an idealised agreement situation. It may be wondered how the choices of idealised parties can reveal what principles or arrangements we as actual persons have reason to endorse.

One of the main reasons to idealise is, however, precisely to ensure that we have reason to endorse the agreement.³ Actual persons may be unaware or confused about considerations relevant for an agreement on the terms of interaction. They may lack relevant information, have false-beliefs, or be prone to bias. If these properties affect the content of their agreement, their agreement would not reveal those arrangements they have reason to endorse. For this reason contract theorists do not concentrate on an agreement between actual persons, but on an agreement between idealised representatives of actual persons.

Contract theorists typically assume the parties to the agreement to be both rational and to have all relevant information. Another idealisation that is usually made is that of mutual disinterestedness . The parties are concerned to further their own interest, and take no interest in the interests of others. This is in the first place to ensure that arrangements do not depend on feelings that may not be universal or stable. It is also to ensure that the agreement is not biased towards certain individuals. Some individuals may be more liked or more disliked than others. Furthermore, some individuals may assign more weight to the interests of certain others than vice versa. In order to avoid that individuals either profit or suffer unduly from such sentiments, contract theorists usually assume the parties to the agreement to have only self-regarding interests.

Rawls is well known for another idealisation that he makes with regard to the agreement situation. Rawls assumes that “no one knows his place in society, his class position or social status, nor does any know his fortune in the distribution of natural assets and abilities, his intelligence, strength, and the like” (1999, p. 11). The idea is that such knowledge may be expected to affect the sort of arrangements persons would prefer. Persons may use knowledge about their own situation to reach a more favourable agreement for themselves. In particular, more powerful parties may rely on the knowledge of their ‘threat advantage’ to extract better terms from those in worse positions (Freeman 2012). To avoid this happening, Rawls denies persons in the agreement situation all knowledge of particular facts—they are placed behind a veil of ignorance. The veil of ignorance ensures that the parties are impartial. In effect, the agreement situation is fair between all the parties to the agreement—the parties are in “an initial position of equality” (p. 11). This fairness of the agreement situation, Rawls assumes, transfers to the principles chosen in it.

The idea of the veil of ignorance has been received critically by other contract theorists. In particular, many have contested that the veil of ignorance is required to generate impartial principles (Parfit 2011). As we will see in the next two sections, neither Gauthier nor Scanlon follows Rawls in proposing the veil of ignorance as a condition for the agreement situation.

That other contract theorists diverge from Rawls with regard to the veil of ignorance makes clear that contract theorists may hold different views about the correct interpretation of the agreement situation. Indeed, as Rawls points out, the structure of contract theories can be divided in two parts: an interpretation of the agreement situation, and a set of arrangements that, it is claimed, would be agreed to by persons in that situation. The core aspects of Rawls’s interpretation of the agreement situation, which he calls the original position, have been described above. I will finish this section by briefly describing the relation between the first and the second part.

Rawls claims that persons in the original position would agree on two principles of justice. The first principle guarantees equal liberties for all. The second principle states that economic and social inequalities are only justified when they are attached to offices open to all and to the advantage of the worst off. While the exact arguments for these principles go beyond the purpose of the present text, it is worth noting that Rawls takes the principles to be selected through a process of individual rational choice: each person in the original position chooses this set of principles because it is the best way to secure her interests as a citizen in society.⁴

We may expect that when two contract theorists differ with regard to the first part of their theories, the interpretation of the agreement situation, they will also differ in the second part, the content of the agreement. In addition, we may expect that when two contract theorists have a similar interpretation of the agreement situation, they will arrive at similar conclusions with respect to the content of agreement as well. In line with this, Rawls has suggested that the above principles are ‘deduced’ from his original position. With respect to Rawls’s theory, however, some authors have questioned whether the relation between the first and the second part is this straightforward (Hare 1973; Harsanyi 1975). While concurring with Rawls’s description of the original position, John Harsanyi has argued that parties would agree on the principle of average utility rather than the two principles of justice. But there may be a way to resolve the link between the first and the second part in this case. Although Harsanyi and Rawls accept a similar description of the original position, they hold different views with respect to the method of reasoning employed by the parties.⁵ If we add this element to the first part of their respective theories, they no longer derive a different agreement from the same agreement situation.

1.2.2 Gauthier

After Rawls’s Theory of Justice, David Gauthier’s Morals by Agreement (1986) is undoubtedly the next milestone in the development of contemporary contract theory. While clearly influenced by Rawls, Gauthier’s use of the idea of the social contract diverges from Rawls in striking ways. The first difference concerns the scope of his theory. Unlike Rawls, Gauthier sets out to defend not only a political contract theory but also an ethical theory (Vallentyne 1991a). Whereas Rawls aims to give us a conception of social justice , Gauthier aims to generate a complete conception of morality on the basis of the idea of agreement.

There is a second way in which Gauthier’s theory is more ambitious than Rawls’s. Rawls, as I have just mentioned, takes the principles of justice to be the object of rational choice from the original position. That does not mean that the principles are based in rationality alone, however. Rawls characterises the original position in such a way that, in his words, “the principles that would be chosen, whatever they turn out to be, are acceptable from a moral point of view” (p. 104). This is most clearly so with respect to the veil of ignorance : its function is to ensure that principles of justice are impartial with respect to social and natural inequalities, among other things. Rawls’s original position is thus not free from moral presuppositions, and the principles of justice are consequently not derived from assumptions about rationality alone.

Gauthier rejects Rawls’s approach . On his view, moral theory should resolve what he calls the “foundational crisis” of morality:

From the standpoint of the agent, moral considerations present themselves as constraining his choices and action, in ways independent of his desires, aims, and interests. […] And so we ask, what reason can a person have for recognizing and accepting a constraint that is independent of his desires and interests? He may agree that such a constraint would be morally justified; he would have reasons for accepting it if he had a reason for accepting morality. But what justifies paying attention to morality, rather than dismissing it as an appendage of outworn beliefs? (Gauthier 1991, p. 16)

Gauthier believes it is insufficient for a moral theory to only show what morality requires, as this would not show persons to have reason to be concerned with morality in the first place. He believes a moral theory must convincingly address what is usually called the ‘Why be Moral?’ question. To answer this question, Gauthier aims to show morality is an effective way to further one’s non-moral aims and interests. Or as he puts it, “to generate morality as a set of rational principles for choice” (Gauthier 1986, p. 5).

This difference in aim has important consequences for Gauthier’s interpretation of the agreement situation. In order to ensure that the morality agreed upon by the parties has a rational foundation, there is no place for a veil of ignorance that prevents the parties from being partial to their own interests. Persons in the agreement situation must have knowledge of their identities or they would be unable to see which arrangements would be rational for them to accept.

Having this knowledge obviously affects what arrangements the parties come to agree on. Whereas parties in the original position occupy, due to their uncertainty, a literally identical position with regard to the basic structure of society, parties in Gauthier’s agreement situation are very much aware that certain arrangements are better for them than others. They thus favour different agreements. Gauthier claims that in order to settle on an agreement that is rationally acceptable for each, the parties engage in a bargaining process with one another in which each tries to get the best bargain available.

Gauthier envisions this bargaining process as follows. It begins from what he calls the initial bargaining position . Each of the bargainers has a starting point, which is supposed to represent “what she brings to the bargaining table” (1986, p. 130). Gauthier thinks of the initial bargaining position as involving a certain level of well-being or utility for each of the bargainers. They join the bargaining table in order to increase this initial bargaining position with the fruits of cooperation. What they have to figure out is, again, how to distribute the gains of cooperation—the cooperative surplus. Gauthier takes this procedure to involve two steps. First, each party advances a claim: the distribution that he would like others to agree to. As the claims made will typically be incompatible, concessions will have to be made until a set of mutually compatible claims is reached. This is the second step.

How do the bargainers reach an agreement? Gauthier argues that as each bargainer wants as much as possible of the cooperative surplus, in the first step of the bargaining process each claims as large a portion as possible—each person puts forward that agreement under which she does best. The size of this claim depends on that person’s natural endowment. As Gauthier writes in an earlier paper, “[i]f we compare the well-being which accrues to the naturally intelligent, strong, and enterprising, under that arrangement maximally beneficial to such persons, with the well-being which accrues to the naturally dull, weak, and lazy, under that arrangement maximally beneficial to them, we shall find the former to be greater” (Gauthier 1990, p. 163). Indeed, as he goes on to point out, the ‘naturally gifted’ may do better even under those arrangements maximally beneficial to the ‘naturally deprived’, as obtaining these may require larger rewards for the gifted.

In the second step the bargainers make concessions on their claims. Given their rationality and mutual disinterestedness , rational bargainers seek to minimise their concessions. Each will find it unacceptable to make a relatively larger concession than others, Gauthier claims, unless doing so is necessary for reaching agreement. According to Gauthier, this directs the bargainers to that agreement in which the greatest concession that any bargainer has to make with respect to his ideal agreement is the smallest. Gauthier calls this the Principle of Minimax Relative Concession .⁶ This MRC principle serves as a rational starting point for further agreement on the terms of cooperation (Hampton 1991).

Gauthier points out that the norms and practices that these contractors would agree to may not satisfy all our moral intuitions . In stark contrast with Rawls’s original position, bargaining power plays a crucial role in Gauthier’s agreement situation. A naturally talented person will put forward a larger claim than a naturally deprived person and will thus, after both have conceded a proportion of their claim, end up with a larger share of the cooperative surplus. A society based on the MRC principle may thus include much larger inequalities in wealth, responsibility, and power than a society based on Rawls’s principles of justice. Moreover, agents without bargaining power may not obtain any rights under this account. As Gauthier puts it, “[a]nimals, the unborn, the congenitally handicapped and defective, fall beyond the pale of a morality tied to mutuality” (1986, p. 268).

There is, however, also an important part of our common sense morality that does fit the MRC principle, Gauthier claims:

Many of our actual moral principles and practices are in effect applications of the requirements of minimax relative concession to particular contexts. We may suppose that promise-keeping, truth-telling, fairdealing, are to be defended by showing that adherence to them permits persons to co-operate in ways that may be expected to equalize, at least roughly, the relative benefits afforded by interaction. These are among the core practices of the morality that we may commend to each individual by showing that it commands his rational agreement. (Gauthier 1986, p. 156)

Gauthier holds that rational bargainers will come to agree on a crucial part of the practices and norms that make up our common sense morality. Given that the bargainers are idealised representatives of us, we also have reason to agree to these practices and norms.

However, by showing that there is a morality that it is rational to agree to, Gauthier has not yet shown that it is rational to be moral. For this to be the case, it must also be rational to comply with the agreement made. This is not evidently the case. While mutual cooperation is beneficial, it typically also involves costs for individuals. It is rational to agree to cooperate with another when these costs are exceeded by the fruits of cooperation. But if other parties to an agreement have already done their part, or if one’s own behaviour cannot be observed by others, complying may not be rational. While it is certainly rational for me to agree to help another if he first helps me, it may not be rational to return the favour when the time comes. Gauthier argues that cooperative activities often have the structure of the Prisoner’s Dilemma , in which defecting is almost universally taken to be the rational choice.

This Problem of Compliance, as it is sometimes called, is a thorny problem for contract theories that seek to derive morality from rationality. If moral requirements can contradict the demands of rationality, the contract theorist’s claim that it is rational to be moral would clearly be false. Moreover, in that case we may not expect that rational bargainers in an agreement situation would come to agree on anything at all: why make an agreement with others if they cannot be expected to comply with its terms? Without a solution to this problem, then, the project of generating morality from rationality fails. As Gauthier puts it himself, in that case “we must conclude that a rational morality is a chimera” (1986, p. 158).

Gauthier believes he has a solution to the Problem of Compliance . He claims that it is rational to adopt moral norms as constraints on the pursuit of one’s self-interest. When persons commit themselves to complying with the requirements of morality, they can gain each other’s trust and cooperate to their mutual interest, Gauthier argues (Cudd 2012). The idea behind this argument is that having the disposition to comply with moral norms affects how others respond to one. Gauthier assumes that a person’s interaction partners can to a certain extent detect whether she is morally committed or not; he supposes that persons are what he calls translucent towards one another. Due to one’s translucency , adopting moral constraint may be expected to have a positive effect on one’s cooperative opportunities and is therefore, Gauthier claims, rational. I discuss this argument extensively in Chapter 6.

Note that the Problem of Compliance does not arise for contract theorists such as Rawls who uses the idea of agreement only to explicate what morality or justice requires. With regard to situations in which it is not in one’s individual interest to comply with the hypothetical agreement, such a theorist can consistently hold that even if compliance is not rationally required it is morally required. That is not to say, however, that a tension between self-interest and morality may not pose a problem for such a theorist. Many contract theorists, and certainly Rawls, are concerned with the stability of their moral conceptions. They suppose persons who live in a society governed by their conception will in general be motivated to comply with it; otherwise the conception would not be suited for its function. A large tension between morality and self-interest may reduce compliance to the point that a conception is insufficiently stable. I briefly return to this later (§ 6.1).

1.2.3 Scanlon

Gauthier, we have just seen, argues that a person has reason to treat others in accordance with principles that would be the object of agreement because it is in her own interest to do so. As he writes, “[t]he basic concern with agreement is to justify to oneself the constraints that adherence to moral principles requires” (Gauthier 2003, p. 168). As the title of his major work What We Owe To Each Other suggests, Scanlon holds a rather different view. Besides being motivated by their own interests, Scanlon takes persons to also be moved by a certain form of respect for others. This leads to a very different contract theory of interpersonal morality.

Scanlon’s theory concentrates on the question of what it is for actions to be wrong. Put roughly, Scanlon holds that an action is wrong if and only if it is unjustifiable to others. More precisely, “an act is wrong if its performance under the circumstances would be disallowed by any set of principles for the general regulation of behaviour that no one could reasonably reject as a basis for informed, unforced general agreement” (p. 153). Note that this is different from saying that actions are wrong because they are unjustifiable; rather, it is to say that properties that make an action wrong are those properties that make the action unjustifiable. Or as Ashford and Mulgan (2012) put it, “[w]hat wrong acts have in common is that they cannot be justified to others”.

The idea of an agreement situation is less explicit in Scanlon’s contract theory than in that of Gauthier and Rawls. It is clearly there in the background, however. Moral principles are justified, Scanlon claims, if they would be the object of agreement among parties who seek principles that no one could reasonably reject as a basis for informed unforced agreement. Following Rawls and Gauthier, Scanlon takes the parties to the agreement situation to be informed, free, and mutually disinterested; and like Gauthier, he drops the veil of ignorance . Unlike both Rawls and Gauthier, however, Scanlon does not take the parties to be first and foremost concerned to advance their own interests. Instead, as the above description shows, Scanlon assumes that the parties are motivated to find an agreement that is acceptable to other parties who are similarly motivated.⁷

This ‘motivational basis’, as he sometimes calls it, is the distinctive aspect of Scanlon’s theory. Scanlon attributes this motivation to the parties in the agreement situation because he takes it to be the most plausible interpretation of our concern with morality. As he writes, “those who are concerned with morality look for principles for application to their imperfect world which they could not reasonably reject, and which others in this world, who are not now moved by the desire for agreement, could not reasonably reject should they come to be so moved” (1982, p. 227). Acting only in ways that can be justified to others on the basis of terms they cannot reasonably reject is on Scanlon’s view what morality is about; it is the ‘subject matter of morality’. As he writes, “when we address our minds to a question of right and wrong, what we are trying to decide is, first and foremost, whether certain principles are ones that no one, if suitably motivated, could reasonably reject” (1998, p. 189).

Whereas Gauthier speaks of the rational acceptability of principles or arrangements, Scanlon speaks of their reasonable acceptability or, more frequently, reasonable rejectability. Whether a principle is reasonably rejectable or not depends on whether and what objections persons may pose against the principle. If the general acceptance of a given principle allowing some type of action imposes certain burdens on me, I have an objection to it. This does not yet mean I may reasonably reject the principle. Whether this is the case, depends on the burdens imposed by alternative principles that would replace the principle if it were rejected. Let’s assume that in some particular case the only alternative principle is one that disallows the type of action allowed by the principle that burdens me. If this alternative principle imposes an even greater burden on anyone else, I may not reasonably reject the principle under consideration, Scanlon claims. Indeed, if I am reasonable, I would redraw my objection once I see that the other has a stronger objection to the alternative (Ashford and Mulgan 2012). On the other hand, if the alternative principle would not impose larger burdens on anyone else, I may reasonably reject the principle under consideration. This means that it would be wrong for others to perform actions governed by the principle.

It is worth noting that Scanlon holds that whether a principle imposes burdens on a person or not depends not on that person’s preferences, but on “what people have reason to want” (p. 204). This includes reasons associated with their well-being, but goes further than just that. It includes reasons to avoid harm, to have our interests taken into account, to control our own bodies, to want outcomes to depend on our choices, to be treated fairly, to be able to rely on the assurances of others, and to give special attention to our own projects, friends, and family (Frei 2009; Scanlon 1998, 2003). Note that Scanlon differs in this respect from Gauthier, who takes reasons to depend on individual preferences alone.

That Scanlon attributes a moral motivation to the parties in his contract situation makes clear that he does not share Gauthier’s concern with the ‘Why be Moral?’ question. Indeed, Scanlon thinks that, given that the great majority of people are in fact moved by moral concerns, moral theorists do not need to justify morality itself (Freeman 1991). What is needed, on his view, is “a fuller explanation of the reasons for action that moral conclusions supply” (Scanlon 1998, p. 148). In particular, Scanlon thinks a moral theory must explain why moral reasons almost always take precedence over other reasons, and why we take it to be so important that people are morally motivated.

Scanlon does this in the first place by relating his account to the value of living with others on terms of mutual respect.⁸ As he writes, the “ideal of acting in accord with principles that others (similarly motivated) could not reasonably reject is meant to characterize the relation with others the value and appeal of which underlies our reasons to do what morality requires” (p. 162). This relation, Scanlon says, is “worth seeking for its own sake”. We have reason to want to stand in a relation of mutual respect with others. Given that respecting them requires treating them only in ways that can be justified to them, we have reason to do so.

Scanlon ties the ideal of justifiability also to the more familiar value of human life. The distinctive value of human life, he says, lies for an important part in the capacity to assess reasons and justification (1998, p. 105). Appreciating the value of human life must therefore involve recognising and respecting this rational capacity. To respond properly to this value, Scanlon proposes, we must treat rational others only in ways that would be allowed by principles that they could not reasonably reject.

Clearly, Scanlon has a rather different contract theory than Gauthier. Whereas Gauthier uses the idea of agreement to generate a mutually beneficial morality from non-moral premises, Scanlon uses it to explicate certain moral ideals. As I mentioned before, writers on contract theory have come to take this difference to be so crucial that they place Gauthier and Scanlon in different contract theory traditions; Gauthier is the best known proponent of the more Hobbesian contractarianism , and Scanlon of the more Kantian contractualism (e.g. Darwall 2003).

However, the distinction should not be taken too strictly. As D’Agostino et al. (2011) write, “there is often as much difference within these two approaches as between them”. Moreover, there is much that they agree on. Both Gauthier and Scanlon use the idea of hypothetical agreement to explicate the content as well as the rationale of morality. They both conceive of morality as a system of rules that enables persons to have valuable relations with each other. And they both take moral persons to be motivated to justify themselves on terms that others have reason to accept. But whereas Gauthier characterises the relations that morality enables first and foremost by mutual advantage , Scanlon concentrates on mutual respect. And whereas Gauthier concentrates on the instrumental use of the disposition to justify oneself to others, Scanlon takes moral agents to have an intrinsic desire to justify themselves to others (Ashford and Mulgan 2012).

The above discussion concerns the first part of Scanlon’s contract theory, the interpretation of the agreement situation. What does the second part of his theory look like? Unlike Rawls and Gauthier, Scanlon does not identify one or two crucial major principles that subsequent arrangements must satisfy. To the contrary, he says that there is “an indefinite number” of valid moral principles (1998, p. 201). This reveals that Scanlon has a different focus than Gauthier and Rawls. Scanlon uses the idea of hypothetical agreement not in the first place to provide a justification of a particular conception of justice or morality. Instead, he seeks to “characterize the method of reasoning through which we arrive at judgments of right and wrong” (p. 2). Indeed, as I understand Scanlon, he takes the idea of hypothetical agreement to play a central role in actual moral thought. Neither Rawls nor Gauthier seems to make this more descriptive claim .⁹

1.2.4 Conclusions

Despite several important differences, contract theorists are united by one fundamental idea: that moral principles are justified if and only if everyone has reason to agree to them. Contract theorists use the idea of hypothetical agreement as an instrument to identify these principles. The most fundamental difference among contract theorists concerns the question of what sort of reasons persons have for accepting or rejecting moral principles, which is reflected in their different interpretations of the agreement situation.

1.3 Two Assumptions of Contract Theory

As I mentioned in the introduction, my investigation of the empirical plausibility of contract theory concentrates on two psychological assumptions made by contract theorists. Drawing on the above discussion, I will in this section describe these assumptions in somewhat more detail.

In order to avoid confusion, it may be helpful to briefly touch on a type of psychological assumption that my investigation will not concern. The above discussion mentioned several assumptions regarding the parties to the agreement situation. Parties are assumed to be rational, to be well informed, and to be mutually disinterested. They are supposed to be motivated to find an agreement that is best for themselves, or to find an agreement that none can reasonably reject. As I mentioned before, these assumptions are idealisations that are introduced to ensure that conclusions of the agreement situation have normative force for us. They are not meant to be generally true of us.¹⁰

Contract theorists are, however, committed to certain assumptions about human agents. I started this chapter by briefly mentioning the ought-implies-can principle. This principle implies that by proposing certain moral norms, contract theorists assume persons can comply with them. But the principle can be applied more generally. Contract theorists defend a particular type of moral conception that they claim we have reason to adopt.¹¹ They thereby suppose that persons can adopt this conception—that they can become contractarian moral agents .¹²

This empirical assumption includes a cognitive as well as a conative requirement. For agents to be able to rationally adopt a moral conception, they must first of all be able to understand the justification provided for the conception. They must be able to understand that the principles that the conception consists of would be the object of agreement. Contract theorists are committed to the assumption, or so I will claim in the next chapter, that agents are able to work out what would be the outcome of the agreement situation that they propose. Moreover, both Gauthier and Scanlon assume in addition that agents can use this idea to ‘test’ whether actions or norms are justified. Indeed, as we saw above, Scanlon seems to assume we already do so when we engage in moral reasoning. I will call the assumption that we can apply the idea of hypothetical agreement for moral evaluation the Practicability Assumption.

Adopting a moral conception involves being motivated to comply with its demands. By supposing that agents can rationally adopt a given moral conception, contract theorists thus suppose a conative requirement is satisfied: that agents can, or even will typically, be moved by its demands. Rawls, Scanlon, and Gauthier all make this assumption explicitly: they all argue their conceptions of morality fit human motivation. Rawls claims that persons in the original position would choose his conception of justice in part because they take it to be more stable than alternative conceptions such as utilitarianism: the contractors recognise that citizens will, due to their sense of justice, generally be motivated to observe its requirements. Similarly, Scanlon assumes persons are motivated to justify themselves to others in a way that fits his moral conception.

Gauthier makes a stronger claim. Both Scanlon and Rawls suppose that persons accept certain moral ideals, and they rely on these to explain why persons would be motivated to comply with their moral conceptions. Gauthier, on the other hand, claims that agents can be motivated to comply with his moral conception solely on the basis of their non-moral interests. On Gauthier’s view, being morally committed is an effective way to further one’s non-moral interests. As I mentioned before, the idea behind this claim is that being morally committed may be expected to have a positive effect on how others respond to one. Persons are translucent, Gauthier assumes, and may therefore expect to do better for themselves by being moral. I will call this assumption the Translucency Assumption .

While Rawls and Scanlon do not make the Translucency Assumption, its plausibility is also relevant for their theories. As I shall explain later on (§ 6.3), Gauthier’s argument for the rationality of moral commitment is not essentially related to his particular moral conception. Roughly put, the idea is that it is advantageous to comply with norms that others expect one to follow, provided that the norms are mutually beneficial and not unfair. This idea can in principle also be combined with other moral conceptions than Gauthier’s, such as Rawls’s or Scanlon’s.

Both the Practicability Assumption and the Translucency Assumption are empirical assumptions. Both assumptions state that persons have certain abilities. The Practicability Assumption states that persons have the abilities required to work out what others would agree to under certain conditions. The Translucency Assumption states that persons have the abilities required to recognise whether others are morally disposed or not. Both assumptions thus regard persons as having particular psychological abilities. More precisely, they state that we have social cognitive abilities: abilities to perceive and think about others. Empirical findings on our social cognitive abilities can thus reveal to what extent these assumptions are plausible.

My investigation regarding the empirical plausibility of contract theory will concentrate on the Practicability Assumption and the Translucency Assumption. There are three main reasons for doing so. The first is that they are among the most crucial psychological assumptions in contemporary contract theory. If the Translucency Assumption would turn out to be false, Gauthier would have no response to the Compliance Problem. This would call into question his project of grounding morality in rationality. And if the Practicability Assumption would turn out to be false, the whole social contract approach may be in jeopardy. If persons do not have the capacities to work out what would be agreed upon in a hypothetical agreement situation, they would be unable to reliably derive moral principles on the basis of it. In that case, we would not be able to use the contract theorist’s procedure for moral justification. This may also affect our reasons for accepting the arguments contract theorists give for certain moral principles—even if we would have an idea of what would be adopted by persons in a given agreement situation, we would know that our social cognitive abilities are such that we may not trust our judgment. Contract theorists may in that case still have an interesting idea of what needs to be the case for principles to be justified, but may not give us a moral conception that we can rationally adopt for living our lives.

The second reason for concentrating on these two is that it is far from clear that they are indeed plausible. As I mentioned in the introduction, philosophers have expressed their doubts about both of these assumptions. A quick glance at the empirical literature may strengthen such doubts. Psychological studies reveal that though we have capacities for thinking about others, or ‘mindreading’ capacities, we are far from excellent at it. As the social psychologist Nicholas Epley (2008) writes in a recent overview of the empirical findings, “people are fairly impressive mind readers in some instances and undeniably terrible in others” (p. 1456).

The third reason for concentrating on these two assumptions is that they both concern social cognitive capacities. As I mentioned in the first section, while there has been much research regarding social cognition in the past few decades, there has been little research regarding the implications of such findings for moral psychology and moral theory. Character, reasoning, and emotion, on the other hand, about which contract theorists may also make assumptions, have received more attention. An investigation into these two assumptions is therefore opportune.¹³

1.4 Overview of the Book

My investigation has two parts. The first part concerns the Practicability Assumption, the second part the Translucency Assumption. The parts can be read independently from one another.

The first chapter of Part I explains why contract theorists are committed to the Practicability Assumption. It will also put forward the claim that they thereby give the ability to consider other perspectives than one’s own a crucial role in moral thinking. The two subsequent chapters consider what empirical findings on our capacity for perspective-taking show about the plausibility of the Practicability Assumption. Whereas Chapter 3 is mainly concerned with the role of perspective-taking in our moral thinking, Chapter 4 is concerned with how good we are at perspective-taking. In Chapter 5 I will discuss what we may conclude about the plausibility of the Practicability Assumption on the basis of the findings presented in the earlier chapters.

Part II starts with a chapter that explains in more detail why contract theorists may make the Translucency Assumption. The chapter introduces three challenges that have been posed against the Translucency Assumption, which are dealt with in the subsequent chapters. Chapter 7 considers first the challenge that people are not translucent at all, and then turns to the challenge that people are not sufficiently translucent for it to be rational to be moral. Chapter 8 discusses whether persons may not do better as opportunists who only act in accordance with morality most of the time, without actually committing themselves to morality. Finally, Chapter 9 considers what implications the empirical findings of the previous chapters have for Gauthier’s argument for the rationality of being moral.

Chapter 10 presents my conclusions regarding the question of how empirically plausible moral contract theory is in the light of the findings on social cognition. I finish with some practical advice for those attracted to a contractarian conception of morality.