Book Review

[in Philosophical Psychology 15.2 (2002): 203-6]


Self-deception unmasked

Alfred R. Mele  

Princeton: Princeton University Press, 2001,

xi + 148 pp.



Al Mele has been as persistent as anyone in his pursuit of self-deception. He has taken it on in a series of papers over the past twenty years (notably Mele 1997) and at various places in previous books. The present book brings together his main ideas on the subject, and readers unfamiliar with its puzzles or Mele's approach to it will learn a lot. The cognoscenti will not only have their memories refreshed but will be treated to much that is new, including recent experimental work that Mele marshals in support of his deflationary account of "straight" self-deception. There is also a chapter devoted to the neglected case of "twisted" self-deception, in which the self-deceiver believes something he wishes not to be the case.

        Self-deception, as its name suggests, is often thought of as lying to oneself. This "agency view," built on the model of other-deception, gives rise to two puzzles, paradoxes even. It suggests that the self-deceiver comes to believe something contrary to what he already believes and somehow manages to bring this about intentionally. The "static" puzzle, as Mele calls it, is to explain how one can believe two contrary propositions simultaneously, and the "dynamic" puzzle is to explain how one can form a coherent plan to pull this off. Like many others who have written on the subject, this reviewer included (see, e.g., Johnston 1988, McLaughlin 1988, and Bach 1981), Mele rejects the agency view. His execution of the deflationary approach is distinguished by his heavy reliance on experimental work, regarding both motivated and unmotivated or "cold" cognition. His main theme is that the route to error in self-deception is continuous with other routes to cognitive error.

        As he did in previous work, Mele cites four processes that lead to motivated biasing: negative misinterpretation (downplaying counter-evidence), positive misinterpretation (treating evidence as more favorable than it is), selective focusing/attending, and selective evidence-gathering (pp. 26-27). No doubt these can all contribute to self-deception, as they can to more clearly cognitive phenomena, such as intellectual obstinance and resistance to new ideas, but the possibility of these processes is really part of the puzzle of self-deception itself, not part of its solution. If self-deception is puzzling, then so, for example, is selective evidence-gathering, which combines "hypersensitivity to evidence (and sources of evidence) for the desired state of affairs and blindness ... to contrary evidence (and sources thereof)" (p. 27). Mele goes on to highlight a new cognitive model, due to Friedrich (1993) and Trope & Lieberman (1996), which proposes that in testing hypotheses people adopt different "confidence thresholds" for accepting and for rejecting a given hypothesis. People do this because they differentially assess the cost of wrongly believing something and of wrongly disbelieving it, for example, that one's idea for a new product is brilliant. Costs and confidence thresholds affect both how hypotheses are evaluated and how far one is willing to go in testing them before making up one's mind.

        It seems to me that Mele relies too heavily on this model. For one thing, it does not (at least as he presents it) take subjective probability into account. The cost of being wrong about something, say that one's brakes will fail, may be very high, but if one takes this to be highly improbable, ordinarily one does not lower one's threshold enough to expect that they will fail. Also, he does not consider the type of case in which it is not necessary to make up one's mind, where withholding belief one way or the other is a feasible option. More fundamentally, the phenomenon posited by this model needs explanation in its own right, since having different confidence thresholds for accepting and for rejecting an hypothesis is itself an essential ingredient of self-deception. That is, part of what self-deception involves is lowering one's acceptance threshold for one proposition and raising it for another. So, in relying on this model Mele has not really shed light on self-deception so much as subsumed it under an empirically studied model. And neither he nor that model fully explains why threshold adjustment occurs on some occasions and not on others and to a greater degree on some matters than on others. Nor does he or the model explain why people don't see through their self-deception, especially when their noses are rubbed in it.

        Mele proposes a set of four elements as jointly comprising sufficient conditions for entering self-deception in acquiring a belief that p (he does not focus on what is involved in remaining self-deceived): the belief that p is false, data possessed by the person favor not-p over p, the person treats genuinely or apparently relevant data in a motivationally biased way, and this causes (in a nondeviant way) the belief that p. Note that satisfaction of the first two conditions is not necessary for the psychological process involved in self-deception: if it turns out that p is true or that the data objectively favor p over not-p but the person has little reason to believe this, that would not count as self-deception and yet the process would be the same. At any rate, Mele acknowledges that his four conditions are not necessary for self-deception. This is reasonable, given that his aim is merely to undermine the agency model.

        Unfortunately, Mele makes no attempt to show that there are no cases of beliefs that satisfy his conditions but do not count as cases of self-deception. Instead, he proceeds to argue against claims that certain other conditions are necessary for it. Even if he were right about that, this would not go to show that his account is superior, since he has not defended the necessity of his own conditions. Maybe there are other sets of sufficient conditions that are just as good as his (perhaps the notion of self-deception is too imprecise for there to be a set of singly necessary and jointly sufficient conditions).

        Mele specifically argues against a family of views according to which self-deception involves reducing tension or anxiety or keeping unpleasant thoughts out of mind. He grants that self-deception often involves such things but denies that it must. Yet meeting merely his conditions does not distinguish self-deception from other sorts of motivated belief. For example, it is not clear that someone who anxiously broods over a remote possibility or clings to a false hope is guilty of self-deception, even though their false belief that a certain possibility is realistic seems to satisfy Mele's conditions. Also, it might be argued that his conditions are satisfied by beliefs that are not generally regarded as self-deceptive, such as those involved in popular superstitions, religious and otherwise, and in other sorts of common nonsense. For example, many people evidently believe that unrepentant sinners will literally go to hell. Clearly this is not their own idea but, as they might acknowledge, what they were "brought up" to believe. They belong to a community of people who have long shared it, and they might realize that there are other communities with different beliefs also held not on the basis of evidence but because of upbringing. Does this belief still count as self-deception? Perhaps Mele would insist that it does.

        Mele does not confront the fact that self-deception is a process that involves not just the formation of a certain belief but ongoing resistance to the emergence of the truth. After all, self-deception is about matters that matter, about issues that keep coming to mind. Self-deception is not the one-shot affair that Mele's account allows--the self-deceiver must keep at it. In downplaying the importance of reducing tension or anxiety and of keeping conflicting thoughts at bay (pp. 52-56), Mele supposes that coming to believe a certain thing in a certain way suffices for self-deception. In seems to me, however, that self-deception is distinguished from mere motivated believing, such as that involved in tension-free dogmatism or bigotry, by the fact that it involves repeated resistance to the truth. Insofar as one of the roles of believing a given proposition p is to dispose one to think that p whenever the thought of p occurs, there is more to self-deception than believing that p in a motivationally biased way: one has to resist the ongoing rational tendency, whenever the thought of p occurs, to think that not-p. On Mele's account self-deception is too easy.

        A related point is that in focusing on what it is to be self-deceived in believing some particular proposition, Mele neglects the fact that this commonly involves self-deception or at least some degree of irrationality about other propositions. Self-deception weaves a web of distorted belief strengthened by a fabric of biased evidence gathering and assessment and of selective attention and inattention. For example, if you already believe self-deceptively that nobody likes you, you might take a nice gesture by someone as an attempt to get something from you. Instead of regarding their good deed as an expression of good feeling, you take it as evidence of an ulterior motive. So it is unfortunate that Mele focuses on entering self-deception and neglects what is involved in remaining self-deceived. For that is where the dynamics of self-deception really come into play. Becoming self-deceived about something would not be all that significant if people did not tend to stay self-deceived about it. Mele's attention to attention mainly concerns its role in initially becoming self-deceived, specifically in regard to evidence, and neglects the role of keeping counter-evidence out of mind in sustaining self-deception and that of flooding one's mind, in the face of residual doubts, with thoughts about favorable evidence.

        Mele eliminates one plausible necessary condition on self-deception. Contrary to common opinion, it is not necessary that what the self-deceiver believes be something that he wants to be the case: people are sometimes self-deceived about something they wish not to be the case. This is what Mele calls "twisted" self-deception, as illustrated by the jealous husband who believes on flimsy grounds that his wife is having an affair. If, like "straight" self-deception, the twisted version involves motivated belief and handling of evidence, it poses a special puzzle, since its motivation is hard to understand. That is, why would anyone deceive himself into believing something that he would rather not be so? Mele considers the sorts of emotion--fear, anxiety, insecurity, as well as jealousy--that might provide the requisite motivation, but unfortunately his discussion gets side-tracked on the issue of whether the emotions in question involve desire, either constitutively or as an effect. For some reason he supposes that emotion cannot itself be motivational, as when he considers whether "emotion, and not motivation, can play a biasing role" (p. 115). But it seems that certain emotions are inherently motivational, e.g. fear, anger, and love. And, as suggested above, he downplays the role of attention and its control. In his example, it seems that the jealous husband would not fear that his wife is carrying on unless he couldn't keep from thinking about this possibility. Also, there is a general question about motivation, applying both to straight and twisted self-deception, that Mele does not take up: under what conditions is one motivated to engage in self-deception? Mele is mainly concerned with what self-deception involves when it occurs, but not with why it occurs when it does. After all, people who are self-deceived about one thing are not deceived about all sorts of other things that they could be equally well-motivated to deceive themselves about.

        For anyone unfamiliar with the literature on self-deception, including Mele's own contributions to it, his debunking of the agency view is clear and compelling. By the time he disposes of it, viewing self-deception on the model of intentionally deceiving someone else will seem about as plausible as understanding self-help on the model of helping someone else. But nowadays few philosophers suppose that deceiving oneself is like intentionally deceiving someone else. Even so, thanks to its stress on the cognitive side and its treatment of the case of twisted self-deception, there is much that is of interest in this book, even to those well-read on self-deception and disabused of its paradoxes.




Bach, K. (1981). An analysis of self-deception. Philosophy and Phenomenological Research, 41, 351-370.

Friedrich , J. (1993). Primary error detection and minimization (PEDMIN) strategies in social cognition: a reinterpretation of confirmation bias phenomena. Psychological Review, 100, 298-319.

Johnston , M. 1988). Self-deception and the nature of mind. In B. McLaughlin  & A. Rorty (Eds.) Perspectives on self-deception (pp. 63-91). Berkeley: University of California Press.

McLaughlin, B. 1988). Explaining self-deception. In B. McLaughlin  & A. Rorty (Eds.) Perspectives on self-deception (pp. 29-62). Berkeley: University of California Press.

Mele , A. (1997). Real self-deception (including peer commentary and author's response). Behavioral and Brain Sciences, 20, 91-134.

Trope, Y. & A. Lieberman  (1996). Social hypothesis testing: cognitive and motivational mechanisms. In E. Higgins and A. Kruglanski (Eds.) Social psychology: handbook of basic principles (pp. 239-270). New York: Guilford Press.


Kent Bach

San Francisco State University

Department of Philosophy

1600 Holloway Avenue

San Francisco, CA 94132 USA