1. The Situationist Challenge

At the heart of the situationist challenge is the contention that we systematically underestimate the influence of the situation in explaining human behavior and overestimate the influence of purportedly fixed features of the agent. Apparently inconsequential and deliberatively irrelevant features of the situation predict and explain behavior, swamping the influence of deliberative agency. Crucially, these ‘situational inputs’ typically operate without the awareness of the agent. When experimental subjects are later asked about the influence of such factors on their decision making, a common response is outright denial.

Situationists support claims about the nonexistence of character by appeal to empirical studies showing that human dispositions to behave are time and again disappointingly unreliable and that such unreliable behavior is not consistent with the existence of robust traits of character. Subtle environmental variables inaccessible to conscious deliberation systemically interrupt the tendency of individuals to act in trait-congruent ways. As a result, situationists argue, to rely on the putative character traits of others is to invite disappointment and injury. To rely on one's own putative character traits is to expose others to analogous jeopardy.

Situationists maintain that our common practice of attributing good or bad character traits is empirically undersupported because individuals fail to act in trait-consistent ways when morally irrelevant environmental variables are introduced. Philosophical situationists appeal to landmark experiments in a long situationist experimental tradition in psychology. Below is a list of some seminal experiments of what became a significant research program:

• • Isen and Levin (Reference Isen and Levin1972) discovered that subjects who had just found a dime were 22 times more likely to help a woman who had dropped some papers than subjects who did not find a dime.

• • Haney and colleagues (Reference Haney, Banks and Zimbardo1973) describe Philip Zimbardo's infamous Stanford prison experiment where college students merely role-playing as guards commit acts of real cruelty and abuse.

• • Darley and Batson (Reference Darley and Batson1973) report that passersby in a hurry were six times less likely to help a person in significant distress than passersby not in a hurry.

• • Stanley Milgram (Reference Milgram1974) found that subjects would administer what they believed to be painful electric shocks at the polite request of an experimenter perceived as an authority.

• • Mathews and Canon (Reference Mathews and Canon1975) report that subjects were five times less likely to help an injured man in the presence of the sound of a loud lawn mower.

These experiments and many others in the same spirit are taken by situationists to be representative rather than aberrational. Experimenters manage to induce disappointing omissions as well as appalling behavior by manipulating seemingly trivial situational variables. Indeed, situationists emphasize how insubstantial the situational influences that produce moral failure can be. And to recapitulate: in many experiments these influences are both inaccessible to conscious awareness and deliberatively irrelevant from the point of view of the agent.

Situationists do not insist that the body of empirical evidence rules out the possibility that a person may act in the same way repeatedly when faced with a similar situation (Doris terms this tendency stability). Rather, they think that the empirical evidence rules out consistency, the reliable manifestation of trait-relevant behaviors across a diversity of trait-relevant eliciting conditions (Doris Reference Doris2002: 22). A philosophical situationist may grant the possibility of local traits, dispositions that are reliably manifested only within a narrow range of highly specified conditions—traits such as ‘dime-finding-dropped-paper compassionate’ and ‘sailing-in-rough-weather-with-one's-friends courageous’ (Doris Reference Doris2002: 115). But the supposition that human beings have consistent dispositions—dispositions that can be relied on across diverse types of situations—is claimed to be an untenable and discredited bit of folk psychology. The supposition's pervasiveness, and our disinclination to revise it even in the light of overwhelming counterevidence, is explained by the influence of the fundamental attribution error, the cognitive bias that leads us to favor explanations based on internal personality characteristics (Harman Reference Harman1999).

Philosophical situationists maintain that the situationist research program in psychology has important implications for normative ethics. Doris (Reference Doris2002) contends that virtue theory's recommendation to cultivate the virtues is one that very few (if any) of us can ever hope to achieve or even to approximate. Moreover, even more empirically modest forms of virtue ethics that exhort us to reflect on imagined virtuous exemplars for moral guidance are destined to alienate and to mislead since these exemplars are constituted so differently from ourselves. Harman (Reference Harman1999) argues further that traditional conceptions of character have deplorable effects, distorting discussions of law and public policy and leading to dangerous misunderstandings of other people.

Maria Merritt (Reference Merritt2000) proposes a revisionary (to the Aristotelian) conception of virtue that openly acknowledges its ongoing dependence on particular social relationships and settings. Merritt maintains that acknowledging the ‘sustaining social contribution’ to character forces us to give up the thesis of ‘motivational self-sufficiency of character’: roughly, the idea that we are able to act consistently in the absence of relevant supportive social and institutional situational pressures. The best we can hope for is ‘stability supported more contingently by climates of social expectation’ (380). With this more modest goal in mind, we ought to ‘take an active, discriminating interest in the climates of social expectation we inhabit’ (381). Given what the situationist research tells us about the power of situational influence, the goal of consistency in the absence of social support is neither reasonable nor useful.

Doris's call for ‘scientific psychological realism’ (2002: 113) in ethics requires that we abandon theories that presuppose the achievability of reliable and robust dispositions. Doris avers that ‘many times. . .confidence in character is precisely what puts people at risk in morally dangerous situations’ (147). He presents the case of a flirtatious colleague who invites you to dinner: ‘If you take the lessons of situationism to heart, you avoid the dinner like the plague. . . .You do not doubt that you sincerely value fidelity; you simply doubt your ability to act in conformity with this value once the candles are lit and the wine begins to flow’ (147).

Doris's advice in the case of the flirtatious colleague amounts to a hedge. Although dinner might be socially pleasant and intellectually stimulating, he suggests that you should hedge the risk that you will tempted to infidelity by politely refusing the invitation. Doris thinks that we should be skeptical of the warrant of self-trust. His skepticism is grounded in disbelief in the possibility of robust traits (in this case, the trait of fidelity). Doris's advice in this case is very much in the spirit of Merritt's view that when the ‘sustaining social contribution’ to virtue is absent, we should not aim for or expect motivational self-sufficiency. In what follows, I argue that this attitude of hedging is endemic to situationism as a moral stance and that it is incompatible with forming relations of trust.

2. Hedging and Trust

The term of hedging refers to a strategy of insuring oneself against a loss. For example, a vegetable farmer may sell futures at the time of planting to protect herself in the event that crop prices fall precipitously when it comes time to harvest. Conversely, airlines purchase options on jet fuel to avert the risk of sharply higher prices in the future. Although the term itself is most commonly used in finance (futures, swaps, options, and derivatives are all used for hedging), this type of strategy is pervasive in many domains of life. At its core, hedging is a defensive posture in the face of uncertainty, a method of mitigating the severity of feared negative outcomes by expending resources to protect oneself in the event that those obtain.

Hedging is rational response to an unpredictable and sometimes inhospitable environment. A worker who fears his factory will be shuttered because of foreign competition may send out job applications to other plants even though he would prefer to stay where he is. A student who is uncertain about whether she will be accepted to her preferred college may invest resources applying to various ‘safety’ schools. Ceteris paribus, the higher the stakes, the more likely we are to hedge.

Hedging strategies also become more salient when we learn that relevant outcomes are influenced by chaotic or random processes. Suppose you learn that new research in meteorology indicates that near-term climatic conditions are more susceptible to unpredictable chaotic elements than previously thought, and that as a result the weather predictions on the nightly news are less reliable than previously imagined. What would be a rational response to such a modification in one's epistemic position? Along with assigning lower subjective probability to a weatherman's confident prediction that the weekend will be sunny, it would also make sense to pack a raincoat just in case the prediction proves wrong.

The species of hedging relevant to this discussion is a strategy designed to address a perceived lack of capacity for consistency. Here I refer specifically to Doris's sense of consistency as the reliable manifestation of trait-relevant behaviors across diverse trait-relevant eliciting conditions. Hedging is an appropriate response to a context in which one makes a determination that a person is not sufficiently consistent to be relied on. If Carmen is considering whether to tell Josephine about a new job offer, but she is uncertain whether Josephine is reliably discreet, Carmen may withhold the information even if she values Josephine's advice about whether to take the job. If Carmen does not think that she can count on Josephine's discretion after Josephine has had a few glasses of wine, she may forgo her counsel. She may tell another friend about the job even though she values his opinion less, simply because she trusts his discretion more.

3. Trust in the Face of Situationism

In contrast to hedging strategies, which essentially involve shielding oneself from the risk that others behave unreliably, trust essentially involves exposing oneself to the risk that the people one trusts will not pull through. Although there is much controversy regarding the appropriate motivation of the trustee in a trusting relationship, most commentators agree that trusting others paradigmatically requires that we:

1. (1) Render ourselves vulnerable to the people we trust. In particular, we render ourselves vulnerable to their betrayal of our trust, which elicits characteristic attitudinal responses (e.g., indignation).

2. (2) Be optimistic that others are competent to do what we trust them to do.

3. (3) Be optimistic that others are committed to doing what we trust them to do (conditions adapted from McLeod's Stanford Encyclopedia of Philosophy entry entitled ‘Trust’ [summer 2014 edition]).

Avoiding reliance and vulnerability by hatching elaborate contingency plans is incompatible with trust. Suppose Ingrid elicits a promise from Seth that he will not touch the last few macaroons in the refrigerator because she had already promised tonight's dinner guests her famous macaroons, for which they have expressed enthusiastic anticipation. If on the way home Ingrid picks up ice cream just in case Seth succumbs to the macaroons, this is a good indication that Ingrid does not trust Seth to do as he says he will. Her hedging indicates that either she is not sufficiently optimistic that Seth is competent to keep his promise (perhaps she thinks he is not strong-willed enough) or that she is not sufficiently optimistic that he is committed (perhaps she thinks he simply does not care enough about keeping his word).

For the most part we only trust people we deem competent and committed.Footnote ² Trust may or may not be warranted (well-grounded), depending on the trustworthiness of the trustee (formulation borrowed from McLeod's Reference McLeod and Zalta2014 Stanford Encyclopedia of Philosophy entry). Our estimation of the warrant for trust—that is, our estimation of whether the conditions for rational trust exist—is determined at least in part by our assessment of the reliability of the trustee. This assessment is grounded in our optimism or pessimism about the trusted party's competence and commitment to do what that party is trusted to do. A necessary condition of Ingrid trusting Seth is that she believe that Seth has a reliable disposition to resist temptation when the stakes are raised. Absent this belief, the conditions for trust do not obtain. If Ingrid is a situationist, she will not believe that individuals are capable of such reliable dispositions.

The situationist, pointing out that trust does not, and indeed cannot, require certainty, might reply by insisting that Ingrid's trust in Seth's reliability may indeed be warranted since in this kind of situation there is no reason to think that Seth is unreliable. In other words, the situationist may claim that Ingrid may rely on the stability of Seth's ‘local trait’. Ingrid may decide that Seth is sufficiently reliable in this narrow domain by reflecting on his track record: on many other occasions when she has relied on him on matters related to household entertaining, he has always pulled through. So, she thinks, she can rationally expect the stability of this admittedly narrow trait. The situationist may suggest that we generalize this strategy of relying only on local traits and that we revise our attitudes and practices so that they are compatible with situationism's denial of consistent dispositions. This revised set of attitudes and practices would be constitutive of ‘situationist trust’ or ‘s-trust’. The situationist might maintain that as it stands we are dangerously promiscuous in trusting others: our attitudes and practices relating to trust are both unwarranted and imprudent since they presuppose the existence of global traits. Trust and trustworthiness as we now think of them ought to be substituted with s-trust and corresponding s-trustworthiness. If A s-trusts B, then A counts on B only in a very narrow range of situations. Changes to the situation, even minor changes of no moral significance, vitiate the warrant of s-trust. Correspondingly, if B is s-trustworthy, she can be counted on only in a very narrow range of situation-types.

For example, if what Merritt (Reference Merritt2000: 374) calls the ‘sustaining social contribution’ is absent, then all bets are off. Substituting trust with s-trust would necessitate significant changes to our reactive attitudes. For example, we would not respond with resentment when others cave in to situational pressure, and we would not feel guilty, ashamed, or diminished when we ourselves fail to come through in such circumstances. We might be disappointed in the sense that we can be disappointed when our lottery number doesn't come up, but we would not be let down in our normative expectations. Rather, we would regard people who fail to pull through with what Strawson dubbed the ‘objective attitude’; and turn our minds toward avoiding reliance on them or, alternatively, to ways of inducing them to behave in the ways we need. Conversely, when others achieve reliability, we will chalk it up to situational support rather than to grit or to conscientiousness.

The situationist might welcome such revision. Indeed, the situationist may claim that she is simply dissuading us from being naïve or Pollyannaish in our trust of others and is thereby sparing us from disappointment and bitterness. She might identify in us a hazardous tendency to ‘spread’ trust across domains, stemming from our susceptibility to the fundamental attribution error. She might caution us that from the fact that a colleague is reliably punctual when it comes to attending department meetings you should not conclude that he will be reliably punctual in carpooling your children to day care. Similarly, from the fact that your new lover has proved loyal in the face of temptation you should not infer that he will be true when passions subside. Because of the nature of our relationships with others, the evidence we have of their trustworthiness is often domain-specific. It is simply a matter of prudence, the situationist would maintain, to rein in our tendency to trust across domains.

But it is crucial to notice that the kind of trust compatible with situationist's denial of consistency—s-trust—goes far beyond prudence and domain-specificity. Doris dubs ‘the queasy realization’ the thought that ‘had things been just a little different’ (e.g., had I found a dime or been in a hurry), I might have behaved very differently (118, italics in original). The only remedy for this queasiness (and it is only a partial cure) is the exclusive reliance on local traits: ‘Person evaluations referencing local traits require confidence only that people will behave similarly in iterated trials of highly similar situations, and this confidence may survive the queasy realization’ (2002: 118, italics in original). But the criterion of similarity of situation is often very eccentric. The sorts of circumstance that undermine dispositions include trivialities such as finding a dime, hearing the noise of a lawn mower, or smelling baked bread—aspects of the situation that the agent herself may not even be aware of, and were she aware of them, she would deem them irrelevant to her deliberation.

If the situationist characterization is correct, then there seems to be no antecedent way to know the disjunction of circumstances in which local traits will manifest themselves.Footnote ³ Perhaps Seth's reliability with regard to not embarrassing Ingrid will falter because temptation happens to come in the form of delectable macaroons. Or perhaps because of his glucose levels at that particular time of day or because one of the guests reminds him of his mother, the situation-type will have been altered. Surely, however, genuine trust is not compatible with a similarly long, idiosyncratic list of provisos many of which will be unknown to the potential trustor or to the trustee. There are a multitude of possible situation-types within even a very narrow pre-determined domain of trust. Thus, s-trust and prudently domain-specific trust are not equivalent. Even very carefully circumscribed trust requires belief in the trustee's capacity for cross-situational consistency.

Situationists explain the apparent consistency in people's behavior by appeal to the supposition that people frequently find themselves in similar situations (Harman [Reference Harman1999] argues further that our tendency to attribute behavioral consistency to inner traits of character is an instance of the fundamental attribution error). This allows the situationist to account for the fact that we do not seem to act in random, uncharacteristic, or unpredictable ways. And the explanation is reasonable as far as it goes: we may make a simple inductive inference from behavioral regularity in the past to behavioral regularity in the future.Footnote ⁴ But the distinctive state of mind required for trust necessitates more than just the expectation of similarity of behavior grounded in anticipated similarity of situation. In order to trust someone, you must have the normative expectation that this person will act in a trustworthy way in the face of nonexcusing alterations to situation.

This is true even if you anticipate that these alterations are relatively unlikely. If you think that your babysitter will fail to take good care of your child if her friends come to visit the house, then you do not really trust your babysitter even if you think it is unlikely that her friends will stop by. For example, you may think that it is just as unlikely that your babysitter will have an aneurism as that your babysitter's friends will visit (perhaps she is an au pair whose friends live in a faraway country). But if she fails to take reasonable care of your child because of the distraction of her friends, you will be indignant; if she fails because of an aneurism, moral indignation has no place. Failing to take adequate care of your child in the former circumstance is a strong indication that your babysitter lacks sufficient commitment to do what she is trusted to do. Her neglect is therefore expressive of moral indifference, and you will respond with indignation. In contrast, suffering from an aneurism constitutes a clear excusing condition and so you will not interpret her failure as expressive of ill will or moral indifference. These aspects of the trusting relationship are not based in predictability. Reliance based on predictability only allows you to trust your babysitter in the minimal sense that you trust your household appliances.

4. Trust, Trustworthiness, and the Reactive Attitudes

Let us return to Ingrid, Seth, and the macaroons. Were Ingrid a situationist, how should she feel when Seth's reliability is undermined by an esoteric situational pressure resulting in a failure to do what he is trusted to do? Should she feel betrayed? She may well feel disappointed in her expectation that Seth did not pull through, especially given his track record of rarely letting her down when it comes to public embarrassment. But if Ingrid adopts wholeheartedly the situationist stance on character, she will not have the normative expectation that Seth will be consistent since she will not believe that Seth has the capacity for reliable dispositions across different types of situations. Moreover, she will be aware that she does not have dependable ways of individuating and identifying the wide array of situation types that may destabilize Seth. Consequently, feelings of betrayal would seem misplaced.

A distinctive mark of genuine trust is that it renders us vulnerable in a very particular way. In contexts where we trust another and we are let down, we do not feel mere disappointment in the way that we feel disappointment when a furnace gives up the ghost or a car breaks down. We feel let down in a distinctive way: hurt, resentful, betrayed. Conversely, when a person we trust comes through for us, especially in the face of trying circumstances, we feel something more than happiness. We feel grateful, perhaps even touched. These considerations begin to give a sense of the intrinsic value of trust. Trust is far more than a means of realizing efficiencies in social interaction. As Richard Holton puts it, ‘trusting someone is one way of treating them as a person’ (1994: 67). As such, trust plays an essential role in cultivating and sustaining relationships of love, respect, and intimacy. Holton describes the case of two people who begin rock climbing together:

Just as the value of trust goes far beyond social coordination, betrayal of trust goes beyond mere disappointment. When our trust is betrayed, we often feel that we have been made a fool of. That we have allowed someone to make a fool of us can be, in extreme cases, degrading and even humiliating. Only other persons can have this kind of power over us. When I trust the sturdiness of a bench and it turns out that my trust is misplaced, I may well feel like a fool. But I will not feel that I have been made a fool of unless my trust in the bench derives from my trust in a person who led me to sit on the bench.

Holton (Reference Harman1994) notices that this network of responses form part of what P. F. Strawson (Reference Strawson1974) dubbed the reactive attitudes. Behind the specific attitudes of gratitude and resentment, moral esteem and moral indignation, stands a more general attitude that Strawson calls the participant attitude and Holton calls the participant stance. Readiness to feel gratitude or resentment is partially constitutive of that stance although we might not always take up the appropriate attitude even when it is deserved. We might, for instance, forswear resentment. But forswearing resentment is different from having no readiness to feel it. The thoroughgoing adoption of situationism, I argue, undercuts all such readiness.

Trusting is one way of treating someone as the sort of person who is capable of having dispositions that can be relied upon. Judgments that bear on whether to take up the participant stance toward a particular individual are in part judgments about this capacity. This should not be taken to imply that this stance is not an all-or-nothing affair. Even when you trust someone, you need not trust that person in every way. It is possible, in fact very common, to trust a person in one sphere without trusting him or her in others. The adoption of situationism, however, infects all trust because situationism is a thesis not about the vulnerability of particular types of people in particular types of situations, but a thesis about people's capacity for consistency in general.

The thoroughgoing situationist will withdraw from the reactive attitudes relating to trust. Consequently, she will not suffer the indignity of being made a fool of, since she will not put herself at risk of betrayal because of another's lack of consistency. Although such withdrawal might be sensible in specified contexts (when one is dealing with addicts or with young children, for instance), it is a different matter entirely to generalize this stance to all aspects of life as a part of a comprehensive disavowal of confidence in character. As a result, I think that situationists misapprehend the ultimate implications of their own theory. Doris claims quite explicitly that situationist ethics does not require a substantial reworking of ordinary ethical thought: ‘My skepticism is not radically revisionary—generally problematizing ethical thought—but conservatively revisionary—problematizing only particular, and dispensable, features of ethical thought associated with characterological moral psychology’ (2002: 108). Doris maintains that we can be justified in expressing a full range of reactive attitudes including guilt, indignation, and recrimination as well as moral esteem while safely disposing of notions of character (129). According to him, we must simply delimit our moral evaluations to an individual's motives and forswear assessment of a person's character.

But much guilt, indignation, and recrimination as well as feelings of moral esteem and admiration are rooted in our sense that individuals (including ourselves) ought be worthy of the trust that we place in them, which in turn is rooted in the normative expectation of consistency. Situationism makes relying on others without hedging seem like a risky, naïve, or even negligent strategy. As a result, the situationist will resort to hedging strategies of various kinds. For example, a situationist may try to put into place safeguards to prevent others from acting in undesirable ways by manipulating aspects of the situation. Or she could lower the stakes for herself by bringing herself to care less about the significance of lapses. Finally, she may simply adopt the standing policy to ‘always have a Plan B’. All of these strategies serve as prophylaxis to vulnerability and disappointment. But while situationism insulates us from the injury of disappointing moral lapses, it systematically erodes our optimism that others are competent and committed to acting in ways that are compatible with a belief in their trustworthiness.

Philosophical situationists may object that I have made too much of their claim that situationist ethics need not be radically revisionary. After all, situationists do call for eliminating moral discourse that appeals to character predicates, an approach that Doris himself allows is revisionary not only for Aristotelian virtue ethics but also for Kantian, contractualist, and even some strains of consequentialist ethics (Doris Reference Doris2002: 107). If the situationist is ready to give up on honesty, courage, and compassion, then perhaps she will not object to my contention that she must give up on trustworthiness as well. So the situationist could reply that while my argument goes through, it lacks any real dialectical force against situationism.

The problem with this approach is that it fails to recognize the way in which trustworthiness is conceptually distinct from widely acknowledged virtues such as courage or compassion. I will not here address the question of whether trustworthiness is a virtue (see Potter [Reference Potter2002] for an answer in the affirmative and Jones [Reference Jones2012] for an answer in the negative). What I have to say below does not presuppose either kind of view. Doris, citing Thomson (Reference Potter1996), suggests that the situationist may still characterize an action as virtuous if all she means by this is that the person acted with certain motives and beliefs in a certain circumstance: ‘A just action involves the belief that the desired state of affairs is equitable, say, and a courageous action involves the belief that performing the action is more important than safety—but this need not involve reference to persisting dispositions’ (116). The reason situationist ethics is not radically revisionary, then, is that we can still praise individuals for courageous actions and still blame individuals for cowardly actions even if we no longer attribute the character traits of courage or cowardice to them.

But trustworthiness is inherently dispositional: ‘reference to persisting dispositions’ is mandatory, not optional. It is part of the concept of trustworthiness that I can be trustworthy even if I am never called on to display my trustworthiness. While it is true that trustworthiness can be expressed in action, it does not make sense to characterize particular actions as ‘trustworthy’ without assuming in the background dispositional reliability. The same goes, mutatis mutandis, for untrustworthiness. Contrariwise, it does make sense to talk of the ‘one-off’ generous, courageous, or compassionate action that is unconnected to a persisting disposition. This is not, of course, to deny that a person may be trustworthy in one domain without being trustworthy in another. What moral psychologists refer to as ‘three-place trustworthiness’—S is trustworthy with respect to A in D—explicitly builds in domain-specificity. I may trust you when it comes to professional matters but not when it comes to matters of the heart. Or I may think that my friend's discretion can be relied on unless he has been drinking. But I cannot sensibly say, ‘I trust him unless some item on a long esoteric list of situational pressures, many of which neither he nor I can identify antecedently, is brought to bear on him’. This would not be recognizable trust or trustworthiness.

None of this is to deny that we may encounter ‘one-off’ betrayals of trust just as we may encounter unexpected bursts of compassion or disappointing lapses of courage. One can feel let down by someone who is generally trustworthy, just as one can feel let down by a person of moral courage who has an uncharacteristic moment of spinelessness. Moreover, the correct description of a particular act as a betrayal of trust does not depend on any description of the long-standing character traits of the betrayer. Nonetheless, we typically do not trust others unless we think it is reasonable to hold them to the standard of doing what they are trusted to do across a wide range of situations. It is only against this background of expectation that judgments of trustworthiness together with the full range of reactive attitudes make sense. If you start to lose faith in a person's competence to meet the normative expectation of actions consistent with good will, then you will eschew reliance by hatching contingency plans or by trying to manipulate or coerce. These kinds of strategies constitute withdrawal (not always immediate) from the reactive attitudes, resulting in suspension of the characteristic attitudinal response to the perception of another person's good will, ill will, or indifference.

The situationist outlook has radically morally revisionary implications because of its fundamental incompatibility with self and interpersonal trust, essential dimensions of our moral lives. It is a dialectical felicity that Doris's own conception of the distinction between ‘revisionary’ and ‘radically revisionary’ is also couched in the language of the reactive attitudes:

Doris's own assessment of situationism's moral revisionism is focused on the viability of doing away with the globalist moral emotion of shame. He does not consider the implications of situationist moral psychology for human relationships predicated on trust and belief in trustworthiness. But substantive reflection on what it means to live one's life as a situationist must address the viability of the situationist outlook in light of the central importance of relationships of trust to our moral lives. As I have argued, full embrace of situationism results in progressive withdrawal from the reactive attitudes constitutive of the stance of trust. Moreover, all of the reactive attitudes that rely on trust are under analogous threat. These include those mentioned by Doris—‘anger, resentment, approbation, and admiration’ as well as others—indignation, hurt feelings, gratitude, reciprocal love, and forgiveness. Once we start seeing others as lacking the competence to carry off what they are trusted to do cross-situationally, this imperils the very coherence of holding them to a normative expectation of good will. This is because whether an agent succeeds in performing an action expressive of good will (or avoiding an action expressive of ill will or indifference) is hostage to a host of situational pressures not antecedently identifiable and so not amenable to the seek/avoid strategy Doris describes in the case of the flirtatious colleague.Footnote ⁵

So where does this leave situationism? If the situationist acknowledges both the indispensability of trust as well as the fundamental incompatibility of trust with the situationist outlook, she might suggest that we attempt to quarantine our attitudes and practices relating to trust from situationist critique. The situationist may counsel that we stop thinking like situationists when it comes to our trusting relationships. She might argue that even if the psychological evidence about the capacity for consistency vitiates the warrant for trust, trust is nonetheless psychologically desirable because of its role in sustaining human relationships and because of the respect it expresses. Even if trust is predicated on an illusion, it is an illusion that is deeply valuable. Situationists might find consolation in the way that some commentators have described trust as a moral virtue. Thomas Hurka, for example, maintains that trust is the virtue of ‘believing beyond one's evidence in the virtues of particular people, such as one's friends or family’ (2001: 108).

But there is a marked difference between believing in the face of incomplete evidence, and believing despite the evidence. Believing in the human capacity for consistency despite evidence to the contrary is irrational, imprudent, and sometimes even negligent. Katherine Hawley puts it well: ‘We owe it to ourselves and to others to place our trust appropriately, where such appropriateness depends at least in part on the trustworthiness of the trustee; this is especially important where we make decisions about trust which affect third parties’ (2012: 15). If indeed the consistency presupposed by global traits is as rare as the situationist believes it to be, then it would be very difficult to take up the stance of trust while knowing this. And even if we could accomplish this feat of cognitive and emotional quarantine, it would be foolhardy and irresponsible. Of course, we might try to convince ourselves that we are in special circumstances: situationists do not rule out the possibility that there may be exceptional individuals capable of cross-situational consistency. But we would have to start thinking about rational trust as something truly rare, and we would have to have confidence that we are able to identify those exceptional individuals who are worthy of it.

A very different avenue open to situationists might be to revisit their characterization of ‘local traits’, rendering them significantly less narrow and eccentric.Footnote ⁶ If the domain of local traits were knowable ahead of time, then the situationist could attempt to organize her life around their contours, strategically avoiding situational pressures that threaten responsiveness to trust. But vexing problems still lurk. First, it is not always possible to avoid situations where local traits will be unreliable, so a responsible situationist may still have to eschew reliance on others and abjure their reliance on her. Second, although a substantial reconceptualization of local traits might help to answer my charge of radical ethical revisionism, I think it would constitute a major revision to situationism, significantly diminishing its distinctiveness and moral significance. Ultimately, the situationist faces a dilemma: either she must accept the radically morally revisionary implications of the situationist stance or character, or else she must radically revise situationism itself.

There is one last avenue open to the situationist. She might claim that the context of interpersonal trust is exceptional, and that it lies outside of the domain of the situationist's denial of consistency. To explain this anomaly, she might invoke what Philip Pettit (Reference Pettit1995) dubs the ‘trust-responsiveness mechanism’ whereby individuals are motivated to show themselves to be reliable when trusted because they desire the good opinion of others. If this mechanism were robust enough, it would do much to mitigate the queasiness of Doris's ‘queasy realization’. But it would also represent a major revision of the situationist thesis since it would count as a way to neutralize the vaunted ‘power of the situation’ and to achieve a morally significant measure of cross-situational consistency. More important, thinking about trust this way would also require significant revision to the reactive attitudes. Pettit notes that trust-responsiveness (as opposed to the trait of trustworthiness) ‘counts by most peoples’ lights, not as a desirable feature for which they need to strive, but rather as a disposition—a neutral or even shameful disposition—that it is hard to shed’ (203). Consistency through trust-responsiveness does not merit gratitude or moral esteem, and the malfunction of the trust-responsiveness mechanism does not warrant feelings of betrayal.

Trusting others requires being in a distinctive state of mind (see Holton Reference Holton1994: 63). When possessed of this state of mind, we are empowered to think and act in trustful ways. But a theoretical commitment that most others are constitutionally incapable of consistency, the position urged by situationist ethics, makes such a state of mind unattainable.

Even worse, a person's commitment to do what she is trusted to do will wither as her faith in her competency declines. One's commitment to carry out an undertaking depends on the hope that one has the required capacity. Such hope is not reasonable for situationists who are honest with themselves. One would be remiss to invite the reliance of others if one did not think that one had the capacity to shoulder such reliance responsibly. Consequently, a situationist who is responsible will abjure the trust of others, a stance that excludes her from a broad swath of meaningful human engagement.