Although philosophy and literature have sometimes seemed at odds, they are brought into close proximity when we consider the logical structure of autobiography. Autobiography is often motivated by a desire for self-knowledge or self-justification; it is also a kind of writing notoriously prone to self-deception. The very notion of self-deception, however, is frequently seen to be strangely paradoxical. It is epistemically paradoxical because it seems to require that the very same person believes what he/she does not believe. It is morally paradoxical because it seems to require that, with respect to a particular action, the very same person can on occasion be both blameworthy and innocent. In both areas, in short, self-deception seems to insist that one is a deceiver and deceived at the same time.
My objective in this essay is to outline a plausible way of resolving these paradoxes, and in the process to elucidate certain features of the autobiographical activity. Beginning with some concrete examples of deception in autobiography, I shall then go on to raise various possible skeptical objections to seeing these as instances of self-deception, including a philosophical skepticism that is motivated by the apparent paradoxicality of the concept of self-deception. In turn, after considering some attempted resolutions of these paradoxes, I shall then attempt to develop an alternative approach, one which draws upon insights from cognitive psychology and recent theories of narrative. As a way of further locating my discussion at the crossroads of various disciplines, the autobiographies which I shall use as test cases are those of three philosophers who seem to have had a problematic childhood: Jean-Jacques Rousseau, John Stuart Mill, and Jean-Paul Sartre.
Consider first the case of Jean-Jacques Rousseau. Toward the end of Book 6 of the Confessions Rousseau gives an account of his decision not to resume his affair with Madame de Larnage (245-47). He adduces various reasons for his action on this occasion: remorse caused by his unfaithfulness to his beloved "Maman," Madame de Warens; anxiety about the exposure of his absurd pretence at being an exiled English Jacobite called Dudding, notwithstanding his innocence of the English language; a certain loss of the "early vigour" of his attachment to Madame de Larnage; a concern about the possibility of her family being hostile to their liaison; and a fanciful fear that he might fall in love with Madame de Larnage's daughter and seduce her (an unlikely event given his self-confessed lack of enterprise in his dealings with the opposite sex). At the end of all this Rousseau decides to break off with Madame de Larnage ("with a few sighs I admit") and return to Maman. He congratulates himself: "I have a right to think well of myself. I am capable of putting duty before pleasure" (246). On his return from Montpellier to Chambery, however, he discovers himself to have been supplanted in Madame de Warens's affections by a young man called Wintzenried, of whom Jean-Jacques has many unkind things to say in the succeeding pages.
Both internal and external evidence show that this account in the Confessions is riddled with falsehood. Correspondence from Rousseau, together with other documents, indicate that he had long ago been supplanted in Maman's affections by Wintzenried and that he was well aware of this. Rather than going back to renew his old life, he was returning to a situation that he already knew to be degrading. In the Confessions, however, he represents his discovery of the affair with Wintzenried as occurring after his return from Montpellier. A hostile critic like Lester Crocker consequently sees this passage as "sheer fraud" (120). Yet perhaps a more charitable critic like Peter France is closer to the truth when he judges Rousseau as self-deceived (94).
My second example is John Stuart Mill's Autobiography. The first paragraph of Mill's work self-depreciatingly states: "I do not for a moment imagine that any part of what I have to relate, can be interesting to the public as a narrative, or as being connected with myself. …