The situationist critique of virtue ethics

Findings of modern empirical psychology have sparked new discussion in several fields. Among those is the philosophical debate about virtues. In the wake of said findings a new critique of virtue ethics has been put forward in the late 20th century: situationism.



    Aristotelian virtue ethics is still one of the predominant theories in ethics today. It can be distinguished from other moral theories like utilitarianism or Kantianism by its focus on moral character. Virtue ethics is concerned with ways to develop virtuous characters, that express virtues like honesty, bravery, generosity and others in action. The basic idea is that a virtuous person will act according to their virtues in all situations. Many philosophers have discussed and criticised this theory.[1]

    More recently, however, a new kind of critique drawing on empirical psychology has challenged the Aristotelian position. Inspired by famous studies in empirical psychology, situationism emerged as a philosophical position in the 1990s. Some of its main proponents are Gilbert Harman and John Doris, who in accordance with said findings in psychology (e.g. results from Milgram’s experiments) claim that there is virtually no empirical evidence for the existence of robust character traits in humans (except maybe for some very rare cases). According to the situationists the different situational factors ranging from the individual’s construal of a given situation to phenomena such as peer pressure have a much stronger influence on and therefore are better predictors of behaviour than is commonly accounted for at least in the Western tradition. This claim has various implications both for psychology and philosophy. The situationist doubt about the existence of robust character traits entails doubts about virtue ethics in general, since virtues, according to Doris, are defined as stable dispositions to behave according to the respective trait. Should situationism hold true, then virtue ethics loses the key component of its moral theory: the virtues.

    The goal of this essay is to evaluate the following thesis:

    The situationist critique, based on a large body of empirical evidence, exposes fallacies and negligence in the philosophical discussion of character traits. Virtue ethicists underestimate the influence of situational factors regarding (moral) behaviour and fall victim to the common fundamental attribution error. Virtue ethics must take this criticism seriously if it is to prevail in the face of results from empirical psychology.

    In order to do so a reconstruction of the key components of the situationist critique of virtue ethics is necessary. The first section of this article explains why a notion of character is necessary for any position in virtue ethics. The second section then discusses how the situationist critique of said notion of character problematizes virtue ethics as a whole. The section closes with an outlook on possible counterarguments from a virtue ethicist’s view that address the definition of character employed by situationist’s and at the same time illuminate a second time the importance of a notion of character for virtue ethics. In a last section one possible counterargument to the critique is presented and discussed shortly. The section closes with an evaluation of the thesis stated above.

    I: Character, virtues, and virtue ethics

    As the name ‘virtue ethics’ suggests, virtues are at the heart of any account of virtue ethics. Founder of the position is Aristotle. Virtues like for example honour are what make persons and the actions they choose good. An honest person will act honestly in respect to the situation they are faced with. Correspondingly, Aristotle defines virtues as behavioural dispositions, which can be learned or acquired and evolved into habits. (Aristotle, 81, 82) Virtues are further characterized as those dispositions that make a good character, which in turn allows a person to fulfil his or her function well. (Aristotle, 83) In short, virtues shape and determine how we deal with and act in the face of affects. They develop through habituation and are manifested in behaviour which are actions chosen using ‘phronesis’, which is practical wisdom. (Aristotle, 85) A good character is one that is virtuous. This characterization of virtues and the good person makes it evident that on a virtue ethicist’s view, a person in order to be good has to develop and facilitate dispositions which lead to ethically good conduct. These dispositions are the virtues which comprise character and the foundation of virtue ethics. Aristotle, according to Doris makes one further claim about the good character, namely that it is evaluatively integrated, meaning that where there is one virtue in a person, similar ones, or on a strict reading all others, are present as well. Doris calls this the inseparability thesis of the virtues. (Doris, 20) A good person on this view is judged by their character which is manifested in their behaviour. In this sense the notion of character is essential to virtue ethics. This is where the situationist critique applies. Before looking at the situationist critique it is important to note that a critical examination of the notion of character in the Nichomachean Ethics may allow for different interpretations and that situationist’s criticise one such interpretation which will be presented in the next section and critically discussed at the end of section II.

    II: The situationist position - Harman and Doris

    All situationist critique addresses the notion of character relevant to virtue ethics. In this section the arguments of two main proponents of situationism are discussed. First, Gilbert Harman’s approach is presented to lay the foundation for a more detailed discussion of situationism put forward by John Doris.

    In his article, Harman claims that moral thought often commits what psychologists call the ‘fundamental attribution error’. This bias, according to Harman is found in most of Western culture but is certainly not limited to that tradition. It describes the tendency to attribute events to persons rather than situations. (Harman, 315) Everyday examples are phrases like ‘of course he did X, after all, he is a nice person’ or ‘what else can you expect from a person this ambitious’. The idea is that we predict behaviour based on assumptions of character. We expect people to behave in certain ways because of character traits (which is the term psychologists use to refer to virtues and vices) we apply to them and explain behaviours and events in the same way. That moral intuitions like this can be flawed he infers from the common flaws found with intuitions in the domain of physics. Many studies (for example those done by Nisbet and Wilson) show that lay people think that objects dropped from a moving aircraft fall straight down while in reality that is not correct because the object keeps the momentum and will fall at a point much further away. (Harman, 315)

    Harman now claims that the same applies for moral intuitions due to the fundamental attribution error. What drives his argument is his second claim that there is no empirical evidence which supports the existence of character traits in the relevant sense. These two claims are the foundation of the situationist critique. For if we accept these premises then we have to look elsewhere for explanations of behaviour. The situationist’s proposal is that situational factors have to take on the explanatory role in this regard. (Harman, 317-318)

    As elaborated in section I and according to Harman’s reading, Aristotle in the Nichomachean Ethics characterizes character traits, or virtues as he calls them, as stable and broad dispositions that reliably issue behaviour relevant to the character trait in question. To challenge this notion is Harman’s goal. He does so by alluding to various studies in empirical psychology. Due to the factor of time, I will limit myself to the discussion of only one of these examples, namely the (in)famous Milgram experiments. I will forego a detailed discussion of the experiment’s setting for the same reason. In Milgram’s experiments participants were instructed to quiz another person and to deliver electrical shocks for false responses. Unbeknownst to the participants the person who received the shocks was an actor. The shocks varied from low voltages, which the participants received themselves to have an idea of its effect, to extremely high voltages which could seriously damage the receiver. The expectation was that most people would stop at 150 volts, labelled as ‘Very Strong Shock’. The results however revealed that of the 40 participants in an early and typical study all went beyond that point and 26 administered the highest voltage shock. (Harman, 321-322) Experimenters, participants and other interviewees alike did not expect these results. Even after the ‘learner’ became unconscious shocks were still administered by the participants under the supervision of the instructor. Harman says, that according to virtue ethics we have to conclude that more than half of the participants lacked virtues such as compassion to an extent which would render them at least sociopathic in lay terms. According to the situationist and most probably common sense this cannot be the correct interpretation.

    John Doris in his book Lack of Character further develops the situationist account first proposed to philosophy by Gilbert Harman. At the heart of the situationist critique is the aforementioned character notion of virtue ethics. Doris calls it the globalist notion of character. According to him, it consists of three main theses. The globalist notion of character claims that character traits are reliably manifested in behaviour (1) cross-situationally, (2) temporally and that they are (3) evaluatively integrated, meaning that they occur in relation to other similar traits. (Doris, 22-23). The three theses that Doris proposes directly oppose the former. They lay the base for his situationist account. Doris claims (i) that behavioural variance is owed to situational differences rather than differences in character, (ii) that empirical experiments question the attribution of robust character traits and (iii) that personality is seldom evaluatively integrated. (Doris, 24-25) He discusses a multitude of psychological studies in support of these claims, none of which I can discuss in detail. Only one I want to highlight because it shows how even in everyday situations smallest situational factors can have a significant impact on behaviour. The study in question was done by Isen and Levin in 1972 and aimed to measure mood effects on behaviour. They created a situation that allowed them to grasp the impact that the psychological reward of finding a dime in a phone booth (which they had previously placed there) has on helping behaviour. What they found was that independent of expectations regarding to character of the people in the study, those who found a dime beforehand were more likely to help a person in distress than those who had not. (Doris, 30-32) These findings lead Doris to suggest that robust character traits, are if existent at all, very rare. In response the situationists turn to local traits. The dismissal of robust character traits does not mean that behaviour is completely random or solely determined by situational factors. The situationist grants the existence of local traits which allow us to predict behaviour in similar situations. We can for example predict that a person who has not cheated in previous tests is more likely to act that way in similar situations than a person who has. This means that cross-situational consistency is proportional to situational similarity. (Doris, 62-63)

    What Doris concludes is that personality can have an impact on behaviour but only indirectly. Character traits allow us to choose situations and environments that fit with our expectations. This is also the way in which Doris wants ethical conduct or reflection to occur. Instead of trying to work on character we should focus on the kinds of situations we put ourselves in and how we can shape our environment to facilitate ethically good conduct. (Doris, 146) Intuitive examples one can think of might be that the crime rates are generally lower in wealthy and stable communities while they are likely to be higher in areas where there is poverty. If the situationist’s argumentation holds, it poses a serious threat to the moral psychology adopted in virtue ethics. For if we deny the existence of robust traits in the relevant sense, then the explanatory power of virtues and with it the appeal of virtue ethics fades.

    Section III: in defence of virtue ethics

    But not all hope is lost for virtue ethics, for even if we have to assume that character is necessary for it, there are still ways to evade situationist critique. In this last paragraph, I will present Rachana Kamtekar’s counterargument to Doris’ critique to show that a) a notion of character is in fact essential for virtue ethics but that b) this is not necessarily its downfall. Kamtekar focusses on Doris’ claims that our dispositions are narrow (local traits), rather than broad and that these dispositions are not evaluatively integrated. Her thesis is that the notion of character situationists criticise have little to do with the actual notion of character in traditional virtue ethics. (Kamtekar, 460) I will now give an overview of her argument to support this point.

    Her argument can be broken down into four steps. First, she claims that the character traits situationists challenge are conceived of by them as “independently functioning dispositions to behave in stereotypical ways.” (Kamtekar, 460) They are isolated from how people reason. To her it comes as no surprise that phenomena like these do not shape our behaviour. Second, she states that the traditional notion of character in virtue ethics is holistic and inclusive of reason. Character on this view is a more or less consistent and integrated set of motivations, desires, beliefs, goals and values. This corresponds to the paraphrased points in section I. It is a person’s character as a whole that explains behaviour not independent traits. Third, what makes a person virtuous is that her character is organized in a way that there is no conflict between the values and beliefs that it comprises. Fourth, the right organization of character is achieved by practical reason and results in cross-situational consistency. (Kamtekar, 460) Behavioural inconsistency is accounted for in Kamtekar’s view not by absence of character traits (like Doris concludes) but by cognitive and motivational difficulties of practical reason.[2]

    On Kamtekar’s reading Aristotle defines virtue as “a disposition to act and feel in particular ways in response to rational considerations […] expressed in our decisions, which are determined through rational deliberation.” (Kamtekar, 479) It is important to note that these do not necessitate particular kinds of behaviour, but they are merely tendencies that nudge us to behave as we do. Furthermore, they are not sufficient for explaining behaviour and Aristotle does not claim they are. This leads us to Kamtekar’s next point, namely that it is ultimately the belief of what a good life of the respective person is and in what ways she thinks they are to achieve, that explain her behaviour. Finally, the Aristotelian virtues do not have to be as broad as they are portrayed by situationists. Besides virtues like honesty that come to mind and seem to be very broad, Aristotle mentions and discusses a long list of virtues in the Nichomachean Ethics and since the book has most likely not been passed down in its entirety, we can suspect that the list goes on. Even of the virtues discussed, there are some that are quite narrow. Aristotle mentions among others, magnificence and magnanimity which are both distinct respectively from generosity and pride. (Kamtekar, 479-480) These indicate that virtues can be fairly narrow and in that sense very similar to local traits. In addition, virtues like honour do not decide on concrete actions but as Kamtekar noted are guidelines for our practical reason. Between having a virtue and acting according to it in a specific situation is a complex interplay of different means, desires and deliberation. To achieve reliable manifestation of these virtues in behaviour takes time and practice and may be very rare. Kamtekar’s argument shows how, while the notion of character is necessary for virtue ethics, the kind of globalist character notion situationists critique is not what Aristotle had in mind. Therefore, the situationist’s critique does not render virtue ethics useless. In fact, an account like Kamtekar’s readily accommodates deliberation about situational factors.

    Kamtekar’s position manages to evade the situationist critique. She admits to the importance of situational factors and incorporates deliberation about them in her model of character. The thesis does not hold true for her account. However, the shift in focus, philosophers like Doris demand, namely to situational factors, may in fact benefit the discussion about virtue ethics and especially: inform our day to day decisions in a new way.



    [1] For a more detailled discussion of virtue ethics see:

    [2] This Doris discards calling it an intellectualist account of virtue (put forward by Mc Dowell among others) which evades the situationist critique because it accounts for virtues without relying on its manifestation in behaviour but in doing so loses the practical nature that makes virtue ethics appealing in the first place.



    Aristoteles (2006): Nikomachische Ethik (U. Wolf, Hrsg.; Bd. 55651). Rowohlt-Taschenbuch-Verl.

    Harman, Gilbert (199): “Moral Philosophy Meets Social Psychology: Virtue Ethics and the Fundamental Attribution Error.” Proceedings of the Aristotelian Society, vol. 99, 1999, pp. 315–331. JSTOR,

    Doris, John (2002): Lack of Character: Personality and Moral Behavior. Cambridge: Cambridge University Press. doi:10.1017/CBO9781139878364.

    Kamtekar, Rachana (2004): “Situationism and Virtue Ethics on the Content of Our Character.” Ethics, vol. 114, no. 3, 2004, pp. 458–491. JSTOR,