Aune, Bruce - An Empiricist Epistemology Ch 5 "Observational Knowledge"

06/01/2007

Unpublished Manuscript

This is a chapter mainly focused on the foundations of knowledge, and dealing with brain-in-the-vat problems. Author eventually concludes that there is no empiricist proof against brain-in-vat (BIV) problems, and that the most recent answer to BIV problems given by Putnam fails as well, so this cannot be used as an argument against empiricism.

When dealing with prototypical empiricist foundations of knowledge, author claims there are two sources: observation and memory. This chapter deals with observation. The problem with observation is its fallibility: we have optical illusions, phantom sounds, hot things sometimes feel cold, etc. The commonplace reply to this is that we must be careful to question our observations, not to take them at face value, to examine their context, their sources, and so on (Locke). The philosophical problem with this reply is that it tries to correct empirical evidence with more empirical evidence-- a regress or vicious circle.

The answer proposed by people like Russell, Moore and Carnap was that there were immediately known 'sense-data' that we were infallible about that served as the inferential foundations of the rest of our knowledge. The problem with this response was that it eventually failed because it put a 'sensuous curtain' between perceivers and the real world. The real world becomes a Kantian 'thing-in-itself' that is unknowable. (pg 171) Arguments requiring the basis of empirical knowledge to be a non-inferential foundation go back to Aristotle, who showed an infinite regress if all our knowledge is inferential. (pg. 173)

However, author argues, we do not need sense-data for a non-inferential basis for our empirical knowledge. Author has developed a framework for what he calls 'imperfect' empirical knowledge, which is knowledge that could be false (ch 1). Working from this, we can accept that there is a non-inferential base from which empirical knowledge is made possible, but that non-inferential base can simply be (fallible) observational beliefs. Author tells a story of how, when we were young ,we took most of our observations at face value, particularly when they didn't conflict with the observations of others. As we became more critical, we began to form generalizations and theories, so our observations became subject to our theories. So the foundation for empirical knowledge is other empirical knowledge. No regress or vicious circle looms, since in all this 'knowledge' talk we are talking about 'imperfect' knowledge.

Author considers alternatives to the Russell/Carnap 'Foundationalism', that are also alternatives to his own theory. One such alternative is Bonjour's 'Coherentism', which asserts that knowledge is justified on the whole, and that particular beliefs are justified by their being able to fit (or not) into that coherent whole. Author claims this is too tough a standard to hold anyone to: everybody has gaps, even the scientific community does!

The problem with empirical knowledge as Hume has sketched it is that it is susceptible to external world skepticism, or a more modern version, BIV. If there were a good answer to this, we would like to hear it. Putnam suggests the answer of Semantic Externalism: the reference relation involves a connection to the actual things referenced ('meanings just aren't in the head!'). Putnam got grist for this argument by successfully arguing that a Turing machine might be thinking, but it would never be referencing if it couldn't sense the objects it was talking about. So the anti-BIV argument runs as follows:
1) if someone S is a BIV, the reference of his words would be electrical impulses, not actual things
2) therefore S's claim 'I am BIV' is not referring to actual things like brains and vats, so he cannot be describing himself as a BIV

Author argues this fails, since it doesn't alleviate our concern: either S has an 'exotic' meaning, or S is saying something false about a real person. Author considers other problems with Semantic Externalism. The major problem is that there is no clear story about how references relations come about. Putnam describes 'language entry and exit rules' that should correspond to behavior and experience, but this fails to capture the full extent of our ability to reference. Author claims that these simplistic rules sound a lot like the verificationist's claim that everything with meaning must be verifiable. If there is room for unobservables to be referenced (like H2O), then it seems there is still no good story about cases of genuine reference in Semantic Externalism. Since there is no clear-cut winner in this, BIV is still an issue for both empiricists and non-empiricists alike.