Chadha, Monima - No Speech, Never Mind!

11/02/2007

Philosophical Psychology Vol 20 No 5 October 2007

The primary goal of this paper is to refute Donald Davidson's old claim that only language users have beliefs. In the early part of the paper, author rehearses Davidson's discussion from 1982 onward, culminating in a summary of Davidson's main argument: (pg 644)

Premise 1 (P1): Only interpreters can have the concept of belief
Premise 2 (P2): Only creatures who have the concept of belief can have beliefs
Conclusion: Only interpreters can have beliefs

For Davidson, interpretation is necessary for the concept of belief because a belief is distinct from an incorrigible representation, and an interpreter needs to know this. Interpretation functions because it charitably assumes that the speaker means something true, but still may be wrong (or lying). To have the concept of belief, one must grasp the subjective/objective distinction. But how does one grasp such a thing if one isn't involved in the 'social activity of interpreting the utterances of others'? (pg 645). Davidson claims that other sorts of behavior modification are insufficient to show subjective/objective distinction-- it can only be done with language users in a social context-- only be done by interpreters. This is the argument for showing with P1 is necessary.

Author points out that Davidson's argument for P2 actually shows that P2 is sufficient for belief, not that it is necessary. Davidson admits this, but then challenges interlocutors to come up with another way that one can discover the subjective/objective distinction without having the concept of belief. Since none so far have answered the challenge, Davidson concludes it is necessary too. (pg 646)

Interlude: Davidson suggested in "Thought and Talk" that he had another argument for P1, using the opacity of substitution for beliefs. The rational explanation for the teleology of a belief is fine grained, using language only to get the distinctions. Non-linguistic creatures will fail to get these fine-grained intentional states into their systems of interpretation, so they will be rather un-belief-like. This has been challenged by other writers as being too rational-teleological and not functional-teleological. (pg 646-7)

Author will consider a different strategy: she will attempt to provide another way that non-linguistic creatures can have the subjective/objective distinction without having the concept of belief. This is primarily established using case studies of other primates. Author does the work to establish another kind of mental state, 'metacognition'. Metacognition differs from higher-order intentional states because it doesn't go upward in the intentional. This sort of metacognition monitors other mental states or processes and takes part in their modification, but does not have higher-order accessibility ('the concept of belief') (pg 649). The key here for the author is to position this metacognitive function in terms of monitoring and controlling information flow from perception and instincts or, roughly, desires. Author shows that chimpanzees and bonobos often will interpret the behavior of their conspecifics, putting them into the category of interpreters. (pg 650-1) Often primates will also engage in deception, which goes a significant way to showing they are making use of the subjective/objective distinction. So a non-linguistic animal might not have the concept of truth or falsity but still have metacognitive normative mechanism that amends false beliefs and affirms true ones, at least within certain contexts (they aren't context-free, like we think our rationality is!). (pg 653).

Author main point is that the interpretive capacities of some non-linguistic animals gives license to ascribe to them mental states, and they don't need the concept of belief to be interpreters. They need instead to have a context-dependent normative metacognitive ability to monitor and control their belief nets.