Thought Insertion and Immunity to Error Through Misidentification

JOHN CAMPBELL (1999) has recently maintained that the phenomenon of thought insertion as it is manifested in schizophrenic patients should be described as a case in which the subject is introspectively aware of a certain thought and yet she is wrong in identifying whose thought it is. Hence, according to Campbell, the phenomenon of thought insertion might be taken as a counterexample to the view that introspection-based mental self-ascriptions are logically immune to error through misidentification (IEM). Thus, if Campbell is right, it would not be true that when the subject makes a mental self-ascription on the basis of introspective awareness of a given mental state, there is no possible world in which she could be wrong as to whether it is really she who has that mental state.

Notice the interesting interdisciplinary implications of Campbell's project: on the one hand, a fairly precise notion elaborated in philosophy such as IEM (and the related notion of error through misidentification [EM]) is used to describe a characteristic symptom of schizophrenia. ¹ On the other hand, such a phenomenon, described in the way proposed, is taken to be a possible counterexample to a sort of philosophical dogma such as IEM of introspection-based noninferential mental self-ascriptions.

In the first section of the paper, I will point out the characteristic features of EM and explain logical immunity to error through misidentification of introspection-based mental self-ascriptions; in the second section I will consider the case of thought insertion in more detail and show why, after all, it is not a counterexample to the view that introspection-based mental self-ascriptions are logically IEM. Finally, I will offer a redescription of the phenomenon of thought insertion.

Error Through Misidentification and Logical Immunity to Error Through Misidentification

For our purposes, it is enough to characterize error through misidentification as a kind of error that can affect both physical and mental self-ascriptions, as well as judgments about other objects and individuals, once these judgments are causally and rationally grounded on a belief in an identification component. For instance, the judgment "My hair is blowing in the wind" can be based on the observation of myself on a shop window. In such an event, by forming the judgment, I also acquire a related structure of beliefs (that I need not entertain in a process of conscious inference, but that are such that I should have the conceptual resources to entertain them) such as "That person's hair is blowing in the wind," where I am that person. Now, if a mistake occurs in the latter identification component, then EM will affect the resulting I judgment. Hence, a judgment is liable to EM if and only if its grounds contain an identification component and it is affected by EM if and only if that identification component is wrong. ²

For our purposes it is enough to characterize IEM as follows:

A singular judgment of the form [A is F], made on certain grounds, is IEM if and only if the subject who is making the judgment on those bases cannot be wrong as to whether it is A who is (or appears to be) F, because her judgment is not rationally and causally based on any belief in an identification component. ³

For instance, if the subject judges (e.g., "I am in pain") on introspective awareness of her pain, her judgment cannot be wrong (at least) as to whether it is she herself who is in pain. The question is why such a self-ascription is IEM and, moreover, is so in any logically possible world. The distinction between de facto and logical IEM is due to Sydney Shoemaker (1968) and was originally understood as follows ⁴ : some self-ascriptions such as bodily ones made on the basis of somatic proprioception are IEM in this world; they are IEM if they are made on the basis of information received...