Print This Post Print This Post

The second chapter of Experiments in Ethics (E in E) is entitled ‘The Case against Character’, and it focuses on a recent critique of virtue ethics due to Gilbert Harman, John Doris and some other philosophers. The inspiration for their attack on virtue ethics is a body of experimental work produced by ‘situationists’, members of an influential school of thought in social psychology.

One much discussed study in the situationist cannon is Darley and Batson’s ‘Good Samaritan’ study. Darley and Batson conducted a study of helping behaviour in which students at the Princeton Theological Seminary were asked to make individual presentations on the parable of the Good Samaritan. In order to get to the place where their presentations were scheduled the seminarians had to pass a research confederate of Darley and Batson. The confederate was slumped in a doorway looking distinctly in need of help. Unrushed seminarians helped the confederate 63% of the time. However, when seminarians were told beforehand that they were running slightly late for their presentation the helping rate dropped to an astonishingly low 10%. A variation in the situation, which appears insignificant, correlated with a dramatic change in behaviour amongst research subjects, and this appears inexplicable on the basis of appeals to character traits, short of positing the existence of some very unusual character traits in the majority of the population.

Appiah follows Doris in construing the situationist challenge to virtue ethics as an attack on the presupposition of globalism, which is a core conception of character that Appiah attributes to the majority of virtue ethicists, including Aristotle (E in E p.38). Globalism, according to Doris (2002, p. 22) – and Appiah follows Doris here – involves the postulation of character traits, with a high degree of cross-situational consistency, stability across time and ‘evaluative integration’. Evaluative integration occurs when the presence of certain traits probabilifies the presence of other traits. A strong form of evaluative integration is the ideal upheld in the ‘unity of the virtues thesis’. The claim that character traits typically exhibit cross-situational consistency is the component of globalism most directly challenged by situationists. Plainly the disposition to help others lacks cross-situational consistency, or at least it is not consistent across variations between hurried and unhurried situations.

Appiah accepts that if the situationist case is as evidentially well-founded as its proponents claim that it is, then globalism needs to be rejected. However, he argues that variants of virtue ethics can survive the demise of globalism, or at least he argues that virtue still matter in a world in which situationist lessons are learned. Even if we cannot hope to develop highly consistent character traits and even if we cannot hope to obtain a highly integrated character, the effort of attempting to behave in ways that exemplify virtues is one that is worth making. It makes out lives go better if we are a bit more virtuous, if we help others a bit more when we are in a hurry than we would if we were unconcerned about attempting to acquire the virtues, or so Appiah argues.

I am inclined to agree with Appiah’s conclusion in this chapter, however it seems to me that he has failed to address the situationist challenge in its strong form. Situationists, or at least hard-line situationists, such as Doris and Harman, do not simply claim that globalism is wrong and that cross-situationally consistent character traits are rare, they also offer situationist explanations of human behaviour which compete with character trait based explanations of human behaviour. They claim that explanations that appeal to aspects of the situations that people find themselves in generally provide us with much better explanations of behaviour than appeals to character traits. If they are right, then trying to be a bit more virtuous than we currently are won’t help change our behaviour outside of the context of narrowly construed situations. Trying to be more honest to our work colleagues won’t make us more honest in our dealings with strangers in public or with our family at home; at best it will make us more likely to be more honest in our dealings with our work colleagues in future.

Because they construe the character traits that humans possess as mostly lacking in cross-situational consistency, hard-line situationists, such as Harman, recommend that we cease worrying about trying to become more virtuous people and that we focus our efforts on building institutions that put us in situations where we will be prompted to behave in better ways, and that we avoid situations that will prompt us to behave in worse ways.
Appiah’s suggestion that we seek a middle way between virtue ethicists and situationists is sensible advice, but he has not argued for it properly and this is perhaps because he has not appreciated the extent of the situationist challenge.

Some further evidence for the conclusion that Appiah has not appreciated the extent of the situationist challenge appears at (E in E p. 45). According to him:

A situationist might well say that, as a prudential matter, we should, in fact, praise someone who does what is right or good – what a virtuous person would do – whether or not she did it out of a virtuous disposition. After all, psychological theory also suggests that praise, which is a form of reward, is likely to reinforce the behaviour.

Now it is possible that concessive situationists would be willing go along with Appiah here, but hard-line ones will not. They will insist that there is no effective way of reinforcing behaviour that is effective across variations in situation. Praise unhurried seminarians all you like, when they exhibit helping behaviour, but you will not be able to change their behaviour when they are in the situation of being in a hurry. While Appiah may be able to find an accommodation between some versions of virtue ethics and some versions of situationism, it is not realistic to think that any version of virtue ethics that involves attempting to change our character traits can find an accommodation with variants of situationism that basically advise us to give up attributing character traits to people. Unfortunately, it is exactly these variants of situationism that Appiah is seeking to accommodate. A better way for him to deal with hard-line situationists would be to follow Sabini & Silver (2005) and others in disputing the evidential basis for hard-line situationism.


  1. 1. Posted by Jeff Huggins | March 17, 2008 6:54 pm

    I thoroughly enjoyed reading Appiah’s Chapter 2, and I also enjoyed reading Steve’s comments. Thanks to both of them. And thanks again to Ethics-Etc for hosting the group!

    There is so much of interest in Chapter 2 that it would take an essay to comment on everything of importance. But, instead, I’ll highlight four of Appiah’s statements or phrases to provide a foundation for my brief (after that) comments.

    On page 57, Appiah observes:

    “Perhaps our evolutionary histories have already “designed” some heuristics for us; perhaps, that is, our judgments are already guided by implicit moral heuristics.”

    Earlier, on page 53, Appiah uses a phrase that is, of course, necessary to consider in combination with the quote just mentioned, which is this three-word phrase:

    “We imperfect creatures …”

    On page 67, Appiah states:

    “Only a misguided theoretical parsimony would make us choose between considerations of character and considerations of consequence.”

    And, I think, it’s also helpful to note the following statement from page 72, which uses the word ‘conjunction’ again, which is a very important word (and on which I commented in my earlier comments on Chapter 1):

    “The conjunction of virtue ethics and situationism urges us to make it easier both to avoid doing what murderers do and to avoid being what murderers are.”

    These accurate comments (in my view) help set the stage for the matter. In addition to what Appiah wrote in his Chapter 1, and in addition to my earlier comments on Chapter 1, I would only add the following for now . . .

    Appiah mentions the notion of “standard” several times in Chapter 2. By what “standard(s)” do we measure these things, i.e., ethical/moral action and/or being? Of course, this question can be asked at several levels: What specific or narrow standard applies to (or guides) individual aspects of human behavior, to individual moral “heuristics”? (For example, what standard(s) applies, or should apply, to a notion of honesty or etc.?) Or, at a different level (a more macro level), what overarching standard sets the context for, and applies to, and connects, the individual “heuristics” and the entire aim of human morality?

    The answer to the latter question, I believe, is in what connects, what “intersects”, what is common between, what “bridges”, the different parts of the picture that Appiah discusses and that I describe in my earlier comments regarding Chapter 1. In other words, it is in the common aspects that are shared between (to oversimplify) nature and imagination, “is” and “ought”, experiment and thinking, and so forth. The foundation of the whole enterprise has to do with ongoing human survival and (given the way time works) human sustainability, achieved in the context of our existence as a social species, our vital relationships with the environment and other species, and our view of human equality (in important senses).

    At this point, I should also mention that I think of several books as being very complementary to Appiah’s thoughts and to the thoughts I mention here, in key respects: Gibbard’s “Wise Choices, Apt Feelings” and, at a different level, Maxwell’s “From Knowledge To Wisdom.”

    I hope this post is helpful in some way. Looking forward to Chapter 3.

  2. 2. Posted by S. Matthew Liao | March 18, 2008 12:46 am

    Thanks for the helpful summary and comments, Steve! I just have two small related points. I’m puzzled as to why virtue ethicists have to be ‘globalists,’ defined by Doris as having ‘consistent dispositions to respond across contexts under the guidance of a certain value’ (p. 38). It seems that a virtuous agent is one who would/should do what is appropriate for a particular circumstance, and not respond in the same way irrespective of the circumstances. For example, a virtuous agent shouldn’t be helpful to a would-be-murderer, whereas he should be helpful to a blind person crossing the street. In fact, as Appiah himself notes, one core element of any virtue ethics is that “the right thing to do is what a virtuous agent would do in the circumstances” (my italics) (p. 36). Relatedly then, if virtue ethicists do not have to be globalists, are the situationalist studies still that challenging to them?

  3. 3. Posted by Constantine Sandis | March 18, 2008 3:29 am

    I completely agree with Matthew that a virtue ethicist can be a contextualist or indeed even a particularist (and certainly many particularists are happy to embrace virtue ethics). What is ‘globalist’ in all virtue ethics is rules about DISPOSITIONS, there need not be any rules about ACTIONS save that which says ‘always act from a virtous disposition’. But we should not expect global dispositional patterns to be mirrored by global behavioural behavioural patterns.
    If a glass is fragile it is likely to break when dropped on marble, but not when dropped on a duvet. The fact that it isn’t liable to break doesn’t only shows that it isn’t *excessively* fragile. Likewise the disposition of honesty won’t manifest itself in cases such as that of the virtuous person interrogated by a Nazi officer. It would be *excessively truthful* (i.e. in excess of the golden mean) to tell the officer that you are hiding people in your basement.
    The seminarian cases seem to me to be compatible with the (dare I say intuitive) view that most of us are only very weakly virtuous, our dispositions triggered more frequently in felicitous situations.

  4. 4. Posted by Fiona Woollard | March 18, 2008 10:20 am

    I agree with Matthew and Constantine. In summarising the argument, Steve said that the difference in behaviour between the rushed and unrushed seminarians “appears inexplicable on the basis of appeals to character traits, short of positing the existence of some very unusual character traits in the majority of the population.” I do not think that the character trait of being less likely to stop and help someone when in a hurry is unusual – in fact it is just what I would predict. The same goes for the people who were more helpful to others when wafted with the smell of baking bread – this smell puts you in a good mood and I know that I am more likely to do supererogatory acts if I am in a good mood.

    I’m interested in the relationship between the situationalist claim that appeals to situation explain/predict behaviour better than appeals to character and the claim that there is no point in trying to develop a good character. If there is a high degree of uniformity of character amongst people (i.e. most people have similar faults, similar virtues and similar tendencies to respond to irrelevant features of the situation), then we may be able to predict what will happen in a given situation with a fair degree of accuracy without knowing much about the character of the agent. Yet this does not show that character is irrelevant. To show that developing virtue is pointless, we would also need to show that no amount of effort can improve upon this standard human character and make us fall into the unpredictably good set. Of course, the claim that we can only make improvements in very specific areas (I can make myself more likely to be honest to shopkeepers but not to be honest to my family) makes this seem a pretty daunting task.

  5. 5. Posted by Fiona Woollard | March 18, 2008 10:21 am

    Sorry, I should have said “I can make myself more likely to be honest to shopkeepers but this will not make me more likely to be honest to my family”

  6. 6. Posted by Steve Clarke | March 18, 2008 11:40 am

    Thanks for all your comments folks.

    Matthew: Appiah doesn’t claim that virtue ethicists have to be globalists – the claim is that most of them are, in fact, globalists. I’m not sure if this is true or not. How challenging situationist studies are for virtue ethicists who are not globalists depends on what lessons you think follow from these studies (this is disputed) and what virtue ethicists are committed to (I find it hard to generalise about virtue ethicists). So your second question does not have an easy answer.

    Fiona: I agree with the gist of your comments about the utility of explanatory appeals to character in circumstances where situations are explanatory. Most people do seem to be surprised by the results of the Darley-Batson experiment. We may be able to explain these results by appeal to the existence of character traits. For example, we can say that people have a great need to ‘save face’ and are therefore unlikely to be willing to lose face by turning up late for a presentation, even when confronted by a moral emergency. The problem is that this is a character trait claim that most people do not seem to recognise in themselves. So when we accept it we are out of kilter with folk psychology (as is the situationist).

  7. 7. Posted by Alex Grossini | March 19, 2008 11:25 am

    the comment #3 is very useful. there’s a hiatus between what we are and what we should/should like to be.

    i just can’t understand the point: what’s so different in “trying to be virtuous” and trying to >behave in better ways, and (…) avoid situations that will prompt us to behave in worse ways

  8. 8. Posted by Roman Altshuler | March 21, 2008 11:54 am

    I think Fiona’s point is an extremely important one, particularly if we accept a view on which virtue is a kind of knowledge, which involves recognizing the morally salient features of a situation and acting on them (I am thinking of McDowell here). There is nothing particularly strange about the fact that people who are in a hurry are more likely to miss salient features, for example, or fail to recognize those features as overriding (given that being in a hurry tends to involve a commitment of cognitive resources to the hurrying). Similarly, being in a good mood might make us less self-centered and so more likely to notice the morally salient features of situations.

    What this comes down to, in a way, is whether the situationists have a theory. Showing, for example, that people tend to be more helpful after finding a dime or when inhaling a pleasant aroma does not constitute a theory, but only a set of observations. Of course there is something like a theory offered: that minor features of situations are good predictors of our behavior. But this isn’t really a theory; it’s just a collection of exceptions cited to show the falsity of another theory. What other theory? The theory that our character traits are global. But virtue ethics obviously doesn’t claim that our character traits are global; only that the virtuous person’s character traits are global (not in the sense that a virtuous person is always honest, but in the sense that a virtuous person is always honest when honesty is called for).

    And this is why I cannot understand why situationism is any kind of threat to virtue ethics. Sure, if the situationists showed that we cannot have consistent character traits, that would be a threat. But all the experiments show, as far as I can tell, is that most people don’t exhibit consistent character traits. And this should surprise no one. If I am honest at home but dishonest at work, this does not show that I lack the ability to be consistently honest. What it shows, rather, is that I have reasons to be honest at home and dishonest at work (for example, I want my coworkers to have a good opinion of me, and so tend to hide unfavorable features from them; but I want my family to know the real me). That is: if my aim is not to be virtuous, but to be liked or understood, it follows that I will not exhibit consistent virtuous character traits. But that says absolutely nothing about what would happen if I made honesty itself my aim.

    If situationists actually offer a theory, on the other hand (e.g., we tend to be more ethically responsive when we are happy; we tend to be less ethically responsive when our cognitive resources are deployed elsewhere), then this is precisely the sort of thing someone who set out to become virtuous would want to take into account. Appiah mentions something like this (“Reading about these experiments will only remind her that she will often be tempted to avoid doing what she ought to do.” (49)), but I think the important feature of the experiments is not just that they show us what sorts of things tempt us to avoid doing what we ought, but that they show us the sorts of environmental factors that lead us not to notice that we ought to do something at all.

    I think this is what Appiah is getting at, really. Note the following quotes: “If experimental psychology shows that people cannot have the sorts of character traits that the virtue theorist has identified…” (47); “If we’re not capable of being virtuous…” (56); “Even if I’m persuaded by situationist moral psychology… that I cannot be a prudent or just or courageous or compassionate person through and through…” (69). The “if” in each of these is crucial: nobody should be persuaded by the fact that we generally do not exhibit consistent virtue, that we cannot do so. What we do need to consider, on the other hand, is that environmental factors do play a role in enabling virtuous behavior.

    One last point: I found the discussion on p. 43 very interesting. Appiah raises the possibility that we do not mention environmental factors when explaining our behavior because the explanation calls for reasons; but he immediately points out that this is not why we don’t mention those factors: we don’t mention them because we are not aware of them. I doubt that this is the whole story. Rather, what strikes me as more likely, is that we are more likely to be aware of features that serve as reasons than ones that don’t. “I helped because the person needed help” is right; “I helped because the person needed help and there was a pleasant aroma” is also right. But since the aroma does not provide a reason, I might simply not notice it. What the aroma might do is make me more responsive to ethical reasons, which is again something I am likely to miss because without the aroma, the reasons wouldn’t have been salient to me, and so I would lack resources for a contrastive explanation. But this just shows that our introspection tends to be normatively guided (it doesn’t show that we cannot become more attuned to features that do make us more ethically responsive).

Post a comment

Name: (required)

Email Address: (required) (will not be published)



(Spamcheck Enabled)

This work is licensed under a Attribution-NonCommercial-NoDerivs 3.0.