Naturalized epistemology

3. The role of observation

Naturalists disagree on many points, among them the role of observation in knowledge, the degree to which natural science is separable from epistemology, the role of normativity in epistemology, and the significance of scepticism (these latter three points are discussed in subsequent sections). One view about the role of observation in knowledge, that of Quine, is that ultimately the ‘evidence for’ or ‘test of’ our theory of the world is the prediction of the stimulations of exteroceptors. The components of our view that are most directly tested by these stimulations are observation sentences such as ‘It is raining’ or ‘There is a rabbit’. These are directly tested by stimulations because what each means is a set of stimulations that prompts assent to and a set that prompts dissent from the sentence. The test of our theory of the world is therefore its ability to predict the observation sentences that verbalize our stimulations (see Quine, W.V. §2).

But many naturalists reject Quine’s version of the empiricist claim that our theory rests on observation. One concern is that it is unclear how sensory stimulations can serve as evidence for a theory. It is relatively clear how a sentence can serve as evidence for a theory, so that we can see how observation sentences test theories, but stimulations are not sentences nor, for all we can see, any sort of propositional attitude. Moreover, the relationship between our stimulations and our assenting to observation sentences is causal, not evidential.

Another problem concerns the distinction between observation sentences (which are purported to have a special plausibility due to their intimacy with experience, so that they may serve to test theories) and the sentences in which theory is couched. Several philosophers (for example, N.R. Hanson 1968 and Paul Feyerabend 1986) have argued that there is no theory–observation distinction. There is no such thing as observation sentences whose meanings are determined by stimulations or the like. ‘Observation’ statements are just theoretical claims themselves, though ones we are accustomed to. So it will not do to say that our theory of the world rests on or is tested by non-theoretical observation statements. It is theory all the way down (see Observation §§3–4).