Epistemology

5. The scope of knowledge

A third perennial question in philosophy regards the scope or extent of our knowledge. What kinds of things can we know, and what kinds of things are impossible to know? Questions about the scope or extent of our knowledge are closely related to sceptical arguments, which purport to show that we know less than we commonly think we do. Some sceptical arguments purport to show that we can know nothing at all. Others are more restricted in their scope, purporting to show that we can have no knowledge of the world beyond appearances, for example, or that we have no moral knowledge, or no religious knowledge.

Traditionally, epistemology has often engaged in a ‘project of vindication’ – that is, the project of trying to show, on the sceptic’s own terms, that we do have knowledge. Contemporary epistemology, however, has largely abandoned this project. For example, it is now widely agreed that one cannot show that one has knowledge in a way that is non-circular, or in a way that does not assume the very knowledge that the sceptic denies that we have. Plausibly, however, this is a function of rhetorical context rather than some epistemic deficit. In other words, so long as one is engaged in a kind of debate with the sceptic, it is rhetorically inappropriate to assume the very knowledge at issue in order to make one’s case. But that in no way implies that one does not have the knowledge at issue.

For reasons such as these, contemporary epistemology has largely given up the ‘project of vindication’ in favour of a ‘project of explanation’. That is, contemporary epistemology generally assumes that knowledge is widespread, and undertakes the project of explaining the nature and value of the knowledge, how it is possible to have knowledge in various domains, and how knowledge is related to other important phenomena. However, contemporary epistemology continues to engage sceptical arguments in the context of this explanatory project. The idea is that sceptical arguments function well as heuristic devices for rooting out plausible but ultimately problematic assumptions about the requirements for knowledge.

To illustrate, consider the following sceptical argument, versions of which have received attention throughout the history of philosophy. The argument begins with two seemingly innocent assumptions: that knowledge in general must be grounded in good evidence, and that our beliefs about the mind-independent world are grounded in perceptual appearances. The argument then proceeds to show that there is no good inference from the way things appear in perception to the way things are, and concludes that therefore our beliefs about the world do not qualify as knowledge. Here is the argument stated more formally.

Therefore,

Therefore,Why accept premise 3 of the argument? There are various considerations that speak in favour of premise 3, including the idea that appearances can be misleading. Moreover, any attempt to verify that appearances are not misleading must presumably depend on the evidence of further appearances, and so we are threatened with either an infinite regress or a circle.

From the point of view of contemporary epistemology, the sceptical argument presents a kind of puzzle. The puzzle arises because each of the argument’s premises seem independently plausible, but together they seem to entail a highly implausible conclusion – that we have no knowledge of the mind-independent world.

Various solutions have been offered in response. One kind of ‘solution’ is to accept the argument’s reasoning, and to recant the idea that we have knowledge of a mind-independent reality. Although this kind of response has been common in the history of philosophy (cf. Hume, Berkeley and Kant), contemporary epistemology has generally challenged the argument’s reasoning rather than accepted its conclusion. For example, coherentists have argued that knowledge is consistent with some kinds of evidential circularity, and infinitists have argued that knowledge is consistent with some kinds of evidential regress (see Coherence theory of knowledge and justification; Epistemic infinitism; Foundationalism).

A different kind of response points out that 4 does not strictly follow from 2 and 3. Rather, the argument here depends on the implicit assumption that good evidence requires a good inference. Put differently, it depends on the assumption that evidential relations must be inferential relations. Several contemporary theories of perceptual evidence challenge that assumption. For example, internalist theories posit synthetic a priori evidential (but non-inferential) relations between appearances and reality. Externalist theories, in contrast, posit contingent evidential (but non-inferential) relations (see Internalism and externalism in epistemology). Versions of the latter approach include reliabilism and virtue epistemology. Safety theories and sensitivity theories can also be understood as positing non-inferential evidence relations (see Modal epistemology). Alternatively, they can be understood as challenging premise 1 of the sceptical argument – that all knowledge must be grounded in evidence. Finally, disjunctivism also denies that perceptual knowledge involves an inference from experience to world. On one interpretation, these theories posit an ontological relation rather than an inferential relation between world-involving experience and mind-independent world. (See also Scepticism; Pyrrhonism.)