My essay examines the question of whether the phenomenon of thought insertion puts any pressure on the well-known principle that our first-personal way of knowing about thoughts rules out the possibility of misidentifying the subject of those thoughts. Prima facie, it seems that that a subject who reports an experience of thought insertion knows what she is thinking in a characteristically first-personal way but is wrong about who is thinking the thought (Campbell 1999). If this is right, then it suggests our ordinary first-person access to thoughts may not really be immune to errors of misidentification.
The underlying issue here is not just about the possibility of immunity to a certain kind of mistake but rather about the nature and structure of conscious experience. Immunity to error invites us to reflect on the structure of an experience of thought insertion, to understand what that experience is like. Is it structurally similar to more familiar experiences of thoughts or is it radically different and, if so, how? One reason it is important to get clear about the nature of a typical experience of thought insertion is that, as Jordi Fernandez has argued in his recent book, 'in order to understand why a subject with thought insertion has the kind of experience that she tries to express by saying things like "I have such-and-such a thought but it is not my thought", we first need to have some grasp of what the experience might be like.' (2013, pg. 147)
Many theoretical approaches claim that experiences of thought insertion are missing some feature that we find in ordinary experiences of thinking. For instance, it is often said that these experiences lack a 'sense of agency' (Stephens and Graham 2000; Carruthers 2012), or 'action-awareness' (Peacocke 2008), or the subject's endorsement of a thought content (Bortolotti and Broome 2009; Pickard 2010). What these all views suggest is that an experience of thought insertion is different in kind from an ordinary experience, that lacking some component makes it fundamentally different from ordinary experience.
At the workshop, I'll propose that an experience of thought insertion is structurally identical to ordinary conscious experience; it is a state of awareness of oneself thinking a particular thought. Thus, rather than missing some element or feature found in ordinary conscious experience, I shall claim that the experience of thought insertion involves something more, a kind of phenomenological overlay that alienates the person from her state of awareness.
This proposal is analogous to the account developed by Ellis and Young (1990; cf. Stone and Young 1997) to explain the Capgras delusion (recently discussed by Max Coltheart on this blog) and I think it also fits very nicely with prediction error theory (i.e., the anomalous experiential overlay is plausibly thought of as an aberrant prediction error signal - also discussed by Max Coltheart and Phil Corlett).
Finally, I think the picture, unlike others, helps us understand why subjects of thought insertion often express ambivalence to their experiences of inserted thoughts.