Costas Pagondiotis "Mental Representations and the Argument from Hallucination"

At the first stage of the argument from hallucination mental particulars are introduced in relation to hallucinations, whereas at the second, generalizing step they are introduced in relation to veridical perception. The cogency of this argument rests crucially on the understanding of the nature and the kind of involvement of the postulated mental particulars. Robinson's phenomenal principle sets here the ground for the initiation of this discussion. In my paper, I will focus on the first stage of the argument from hallucination and I will examine two dominant ways mental particulars were conceived, namely as mental representations (see, for example, Hume's talk of "fleeting copies or representations of other existences") and as bare, non-representation-like mental entities. For reasons of brevity, I will focus mainly on the examination of the first conception, though critiques of the second conception will surface along the way.

The idea that hallucinations involve mental representations is often treated as part of the commonsensical understanding of the notion of hallucination. This is mainly because mental representations readily afford an account of perceptual error: hallucinations, illusions and other kinds of misperception are supposedly explained -in a unified way- as due to erroneous mental representations. However, when we attempt to spell out in detail the nature and function of mental representations, we come up against insurmountable problems. In the proposed paper, I will argue for three points:

1. The claim that mental representations are involved in hallucinatory experience is not supported by the phenomenology of hallucination. In particular, I will argue that no picture-like mental particulars are available at the phenomenological level. To this effect, I will introduce certain criteria which allow us to distinguish external pictures from other external objects which are not treated as representations. One such criterion that I will introduce is that the ability to identify something as a picture involves the ability to distinguish between representational bearer and representational content. I will show that these criteria do not apply to our hallucinatory or veridical experience. In response to this line of critique, the indirect realist may either hold that mental representations are quite unlike external pictures, but in that case he has to show in what way they make themselves manifest at the phenomenological level, or he may concede that mental representations are not accessible to the first person perspective.

2. If the indirect realist concedes this last point, he can still hold that mental representations are involved in hallucination and account for its content, but that the perceiver is aware only of the representational content and not of the representational bearer of the mental particulars as such. I will explore whether this position is compatible with the transparency of experience thesis, forcing, thus, the indirect realist to embrace a kind of intentionalism at the phenomenological - first-personal level. In this context, I will contrast the use of "mental representation" in indirect realism with that of intentionalism. On the other hand, the postulation of mental particulars functioning as representations inaccessible to the perceiver retains a kind of indirect realism, but at a third-personal level of approach. I will argue that the introduction of mental representations as an explanatory hypothesis cannot account for the intentionality of hallucination. In particular, if the indirect realist concedes that mental representations are not available as such to the perceiver, then he assumes that these representations can be defined independently of the user-interpreter. I will explore certain accounts of the referentiality of the postulated mental representations in terms of specific kinds of relations between the representational bearer and the representational content, and I will argue against the adequacy of these accounts. Thus, I will conclude that the postulation of internal particulars cannot account for the intentionality of hallucination.

3. Finally, I will briefly argue that hallucinations and the other kinds of perceptual error can be accounted for without recourse to mental representations. One could embrace direct realism and still leave room for the fallibility of perception. There are world-dependent and mind-dependent sources of perceptual error, such as the aspectuality of perception and its knowledge-ladenness, respectively. These sources create a variety of different kinds of perceptual error, some of which are repeatable and some not. I will propose that a possible mind-dependent source of hallucinatory experience is disruption of the sense of agency in imagination.