Self-deception

1. Intentional self-deception and partitioning strategies

Philosophers who deem self-deception conceptually or psychologically impossible typically contend that deceiving, by definition, is an intentional activity and that putative self-deceivers, therefore, must intentionally deceive themselves. In various time-lag scenarios, intentionally deceiving oneself clearly is possible. Paul, a pudgy prankster who has intentionally deceived others about their weight by secretly adjusting their bathroom scales, now intends to deceive himself into believing that he is lighter than he in fact is. Cognizant of his forgetfulness, the prankster adjusts his own scales for this purpose and counts on eventually forgetting that it was done. In this way, Paul may intentionally bring it about that he believes himself to be lighter than he is. In problematic cases, there is no such time lag. Rather, at some point during agents’ intentional efforts to deceive themselves, they succeed in so doing. Some contend that such cases are impossible and that self-deception, therefore, is nonexistent. (These theorists suppose that self-deception is essentially intentional and involves no time lag.) Others defend the possibility of such cases. Yet others argue that intentional self-deception is remote from garden-variety self-deception and that even if the hard cases are impossible, that leaves the reality of self-deception intact.

Some philosophers propose to account for the possibility of self-deception by postulating a partitioned mind. Partitioning strategies range from postulating full-blown sub-agents with their own motives and intentions to employing modest distinctions among kinds or levels of awareness (see Modularity of mind; Cognitive architecture). In general, partitioners accept a strict interpersonal model of self-deception and seek to locate the requisite divisions within a single mind. A seriously divided mind may contain both a deceiving agent and a deceived subject. The deceiver (D) may believe that p while intentionally causing the subject (S) to believe that ∼p. To the extent to which a divided mind is analogous to a pair (or group) of persons, static problems seem manageable. As you may believe that p while I believe that ∼p, D may believe that p while S believes that ∼p. Dynamic problems would be equally manageable. The basic idea of a representative dynamic problem has two major components. First, given that self-deceivers intentionally deceive themselves, they must intentionally follow a plan in so doing. Second, setting aside time-lag scenarios, any attempt at such plan-following would be self-defeating: just as I will not deceive you during a temporal span t, if you know what I am up to throughout t, you will not deceive yourself during t, if you know what you are up to throughout t. In a radically divided mind, S may have no idea what D is up to, and the problem vanishes.

Radical partitioners are often charged with substituting a species of ‘other-deception’ for self-deception. The criticism is that no human beings both deceive and are deceived by themselves on the radical view; rather, a sub-agent is the deceiver and something else is deceived. Further, the mental condition of radically partitioned persons strikes some theorists as disturbingly similar to multiple personality disorder. Surely, ordinary self-deceivers need not be so fragmented? Arguably, more modest proposals can account for self-deception.

Less radical partitioners eschew appealing to internal agents. Normally, however, their partitioning hypotheses are designed specifically to handle static problems, and dynamic counterparts remain. It has been proposed, for example, that a person may simultaneously believe that p and believe that ∼p, because the true belief may be possessed only unconsciously and therefore need not pre-empt the simultaneous holding of a contrary belief (see Unconscious Mental States). However, this does not answer the charge that intentional processes logically required for self-deception are necessarily self-defeating.