Review of Philosophy and Psychology

Experimental and Cognitive Psychology
Published by


Interested in reviewing for this journal?
Editors on Publons
Top handling editors on Publons (Manuscripts handled)
Top reviewers on Publons (Manuscripts reviewed in last 12 months)
Endorsed by


  • I am happy to say that the author answered many important concerns in this new version of the paper, which now clearly states that it is not arguing against scientific personality psychology, unlike other situationist authors.

    I still fear that some readers might read the paper and conclude that personality psychology is unscientific and rests on folk intuitions alone. The author encourages this impression by stating repeatedly that the scientific study of personality is an interesting endeavour to be undertaken in the future (p. 3-4, p. 22). This statement seem to ignore the fact that a science of personality has been existing for more than half a century, in psychology, genetics, neuroscience, etc. (even though it is not, of course, above criticism). It would be nice to read one sentence recognizing this fact.

    Two points of concern:

    On p.3, the author argues that correlations between traits and individual actions are undetectable because they are not very strong, as shown by Mischel in 1968, and because we are bad at spotting correlations of any kind, as shown by Jennings et al. Mischel's work is now much too old to be convincing. A decade later, Nisbett found that correlations were actually 30% higher. I think a reference from the 2000s would be indispensable here. (And, if the author wishes to imply that personality traits are much less important than situational factors, it should be noted that correlation sizes between situations and behaviors are quite often not stronger than that - or else, prove the contrary).

    On the topic of the false consensus effect, p. 8-9, I think the author may have misunderstood Dawes and Mulford's critique. In the false consensus experiments, the subjects do not have the choice to make a prediction or not to make one. They have to answer the experimenter's question, even though they are given very little data to rely on, a fact which subjects can do nothing to change. Dawes and Mulford ask what the most rational option is, given those bad conditions: should people take into account their own decision in predicting others', or should they abstain from using that information? The second option means preferring to rely on no data at all. The first option means relying on just one data point. Dawes and Mulford conclude that the first option (use one data point) is more rational than the second (do not use the available data). It would, of course, be more rational to use better data, as the author points out. But the data is simply not there to be used, and the subjects have to make a decision (this is part of the experimental set-up). They make the best of a bad situation. Therefore, the authors should not so readily accuse human stupidity.

    Reviewed by
    Ongoing discussion
  • Explaining away intuitions about traits: why virtue ethics seems plausible (even if it isn't)

    This article belongs to the trend of research that challenges virtue ethics from the standpoint of situationist social psychology. Though he endorses the situationist thesis, the author does not defend it directly in this paper. Instead, he wishes to provide a "theory of error" that would explain why we use character traits to explain behaviors even though, according to situationists, such explanations have little or no explanatoy power. The article's main contribution consists in explaining how a series of well-known psychological biases might reinforce our belief in the existence and causal power of traits. Having thus, he thinks, explained away our beliefs about traits, the author concludes that situationism wins.

    The argument is straightforward and reads nicely, the topic is of clear interest to the readers of EROPP, but I have two important problems with the paper.

    Dogmatic and uncritical use of the psychological litterature

    The paper often relays questionable and controversial claims from social psychologists without acknowledging dissenting opinions and data (this actually starts at the first paragraph, where the author repeats the standard tale about Kitty Genovese; but this is unimportant). The fundamental attribution error is not "unchallenged" (p. 11; see Sabini, Seepmann and Stein's The really fundamental attribution error in social psychological research, Psychological inquiry, 2001). Neither is the false consensus effect (p.12; see Dawes and Mulford's extremely convincing argument, The false consensus effect and overconfidence: flaws in judgment or flaws in how we study judgments? Organizational behavior and human decision processes, 2001).

    Many assertions in the paper are presented without evidence. Who said that the fundamental attribution error was "innate" (p.12)? That "actual correlations between traits and individual actions are undetectable" (p.5)? This last one makes you wonder what personality psychologists have been busy doing in the last century. Of course, a simple PubMed search turns up a host of correlations between personality traits such as neuroticism or extraversion, with behaviors ranging from from asthma to Internet use to entering conversations with strangers, etc. Is this evidence irrelevant because it comes from the non-situationist side of social psychology?

    The argument, though clever, proves too much

    Most of the biases the author invoke (the availability bias, the base-rate neglect, the confirmation bias) apply to human reasoning in general, not just to reasoning about traits. They could be as easily invoked to mount a challenge against situationism. True, these 3 biases sometimes make us prone to endorse misguided, improbable or even stupid explanations of events. But this is true of all explanations: explanations of human behavior and explanations of natural events, situationist explanations as well as trait-based explanations.

    Take a (true) explanation of behaviors that a situationist may endorse: poverty tends to make people violent. I could try, following the author's reasoning, to debunk this explanation by saying: you see, we suffer from various biases that lead us to believe that violence and poverty are linked. We see a poor person hitting another; we conclude that poor people are violent - but that is just because of the availability bias (we reason by the last striking anecdote we know), and because of the base-rate neglect bias (we overestimate the probability that poor people hit one another because we don't take into account the fact that hitting behaviors are rare to begin with). We persist to believe that poverty tends to make people violent because we only consider confirmatory evidence (confirmation bias). Ergo, poverty has no effect on violence.

    Is this line of reasoning valid? Does it prove situationism wrong? Of course not. Just because we sometimes favor some explanations for bad reason, does not mean that a particular category of explanations (situationist or trait-based) is false.

    Reviewed by
    Ongoing discussion