Monday, 27 October 2014

Robert Audi against doxasticism

I caught a talk by Robert Audi at the end of last week (the sort of thing that just seems to happen in Durham). Since I am giving a talk on a related subject a little later in the term I will try to summarise very briefly the motivations for some of what philosophers think about in this area in pretty basic terms (not that my own understanding rises much above the basic).

First, though, his abstract.

Doxasticism: Belief and the Information - Responsiveness of Mind
Our beliefs are a map of our world. They shape our hopes, direct our desires and intentions, and structure our values. They are as multifarious as their propositional contents, but their objects are not limited to propositions. They may be directly about the world of our experience—thus causally connected with external objects—or, directly or indirectly, may concern elements internal to the mind. Beliefs also differ psychologically: in strength, depth, influence on behavior, and accessibility to consciousness. They differ normatively in how well-grounded they are in rationality and in justifiedness. Philosophers have written widely and informatively about belief, but there remains an aspect of the topic that needs further analysis. It concerns the conditions under which an information-bearing state—say a perception or the recalling of some past event—yields belief. This paper opens with a distinction between belief and a psychological condition easily conflated with belief, illustrates the tendency of philosophers to overlook this distinction, and offers a positive conception of the mind’s information-responsiveness that does not require as much belief-formation—doxastic uptake, if you like—as has been commonly supposed to be produced by perception and other experiences. This conception is clarified by a partial sketch of the natural economy of mind. The paper then considers whether the economical view proposed requires abandoning the venerable belief-desire conception of intentional action, and, in the concluding section, suggests some ways in which intellectual responsibility is both clarified and extended by the overall work of the paper.

I am sitting in a Wetherspoons pub in Durham which took a little finding. I had to look. So wandering around with my eyes open was an active project. I did that. But what did I gain as a result so as to know where the pub is? Well, on one approach, I gained a series of visual (and other) experiences and they told me where the pub is. (Readers may or may not find that phrase fishy.) But, surely, the thought goes, they can only do that if they have some sort of ‘tellable’ content, some sort of representation of how the world is nearby. Hence, experiences are representational states, representing how the world is. What sort of state is a representational state? Or, what is such a state like? Well one candidate is belief. And hence the idea that seeing is believing: one of Audi’s targets.

But that is to go a little quick. Another intuitive way of thinking about experience is that it is of how the world seems. We experience the appearance of things. That is not yet to say the terrible thing that Descartes seems to want us to say: we experience only the appearance and not the reality of things. (That way sceptical madness lies. It may instead be, however, that in the appearance, we take in how things are. Appearances need not be misleading.) But it is enough to suggest that seeing isn’t necessarily believing. And hence the middle period McDowell suggests that experiences are invitations to believe. Still, on McDowell’s then picture, the invitations are themselves representational states. They just don’t have the force of beliefs. But if endorsed, the resultant belief just has the very same content as the experience, the invitation to believe. (A further question: what sort of thing are beliefs? What are they like? Well as Sellars suggests in the Myth of Jones, we might think of them as akin to explicit linguistic judgements. Hence experiences as (invitations to form) beliefs as linguistic judgements.)

If I heard him correctly, Audi thinks something similar to that. Criticising overly intellectual philosophy of both action (requiring, eg., a concept of intention of any agent acting on an intention, including small children with no abstract concepts) and of linguistic meaning (any communicating speaker has to have nested communicative intentions, a la Grice), he argued that there was much less belief around than some philosophers assumed. In fact, in many circumstances, a disposition to believe is all that is available and all we need. As a concession, he suggested we might call a disposition to believe a ‘virtual belief’.

There is a different direction we might take if the idea of experiences telling us anything seems fishy (following Austin and Travis). It is to load all the belief-like content into a response made by a subject to their experiences and deny that the experiences themselves are representational states. They are barely states at all. Audi might have believed that but, when I asked him, did not seem to and suggested that the visual ‘percept’ carried a kind of content, just not a belief-like one. Sub-belief-like content could then be anchored in or carried by the experience (well this is how I understood him).

That left me surprised by one of his other key belief-reducing moves. The intention-in-action of entering a shop to buy a paper, or of playing a complex piece of music, could be carried, he suggested, by a ‘script’. These were not structures of beliefs but involved a kind of responsiveness to the situation. In the music case, this eliminated the worry that one somehow has to anticipate every note one will later play.

Now it seems to me that a subject still needs to ‘entertain’ or conceive of playing a piece of music in some way if their playing is not extemporising though I agree it seems wrong to think that the whole piece of music comes before their mind in some sort of explicit belief-structured anticipation. But Audi suggested that this conception might be carried by an indexical ‘this piece of music coming…’.

Such a response seems more plausible in the playing than before (though even in the playing, the later playing is still to come but may be interrupted). How, in the latter case, can they point to the as yet non-existent piece? (The obvious resource unavailable to Audi is an object-independent descriptive thought with no indexical component. That seems belief laden, however.) But I wonder whether what seems to me to be merely a superficial plausibility stems from a difference between the pragmatics of conversation and the metaphysics of thought. The pianist might say ‘the next piece of music...’ and play it. Or she might say that and then, distracted, accidentally play the wrong piece. The hearer is in no position to distinguish the cases but the pianist is. In the latter case she did not play the piece she was thinking of. Thought and utterance-as-heard come apart.

I was surprised that beliefs were eliminated in the perception case in favour of something representational and present (the precepts of visual experience) but by something not yet present in the case of intentional action. It did not seem symmetric.

(Noone of the importance as Robert Audi needs a reference from me. But I was struck by the way in which, in a spirit of Let’s do the whole thing right here in the barn! he volunteered to return a couple of hours later to talk to any graduate students with further interest in this or any other topic. Above and beyond!)