1 WHAT IS THE ROLE OF THE SELF IN SELF-DECEPTION? * RICHARD HOLTON The orthodox answer to my question is this: in a case of self-deception, the self acts to deceive itself. That is, the self is the author of its own deception. I want to explore an opposing idea here: that the self is rather the subject matter of the deception. That is, I want to explore the idea that self-deception is more concerned with the self’s deception about the self, than with the self’s deception by the self. The expression would thus be semantically comparable to expressions like ‘self-knowledge’ (which involves knowledge about the self) rather than to expressions like ‘self-control’ (which involves control by the self). 1 On this approach, what goes wrong, when we are self-deceived, is that we lack self-knowledge; or, more accurately, since one can lack knowledge without falling into error, what goes wrong is that we have false beliefs about ourselves. Not any kind of false belief about oneself; I am not self-deceived when I mistake my shoe size. Rather, self-deception requires false beliefs about the kind of subject matter that, were one to get it right, would constitute self-knowledge. It is an interesting fact about current English that, though we talk freely of self-knowledge, we have no common term to designate its absence. Seventeenth century writers talked of self-ignorance; but the term has fallen from use. I suggest that ‘self-deception’ is the nearest we have. Lack of self-knowledge is clearly central to many cases that we describe as cases of self-deception or self-delusion (following the OED I take the two as synonymous). For an illustration in a non-philosophical context, here is Cowper: How many self-deluded nymphs and swaines, Who dream they have a taste for fields and groves, Would find them hideous nurs’ries of the spleen, And crowd the road, impatient for the town 2 * An early version of this paper was given at a workshop on self-deception at Stirling. Thanks to the audience there, and to those who heard similar material at Aberdeen, Keele, Birmingham and Edinburgh. Special thanks to André Gallois, Eve Garrard, Rachana Kamtekar, Rae Langton, Al Mele, and Tim Williamson. 1 But note that I am not giving an argument about grammatical form. I am not saying that (1) x was self-deceived has the grammar of (2) x was deceived about x. rather than of (3) x was deceived by x, Evidence against that claim comes from observing the greater acceptability of (4) William was self-deceived about his mother than of (5) William was self-deceived by his analyst. 2 William Cowper, Task iii 316–19