Metaphysics of knowledge

DOI: 10.4324/9780415249126-P066-1
Version: v1,  Published online: 2017
Retrieved October 19, 2018, from

1. The nature of knowledge

A useful distinction to draw is that between ontology, which concerns what types of things exist, and metaphysics, which concerns the ultimate nature of things (Block 1980). In short, ontology is about what there is and metaphysics is about what it is like. Sceptics challenge commonsensical knowledge of the external world, whereas followers of G.E. Moore, for example Pryor (2000; 2004), dogmatically maintain the existence of such knowledge (see Scepticism). That is a dispute over ontological commitment to knowledge. Indeed, even if knowledge abounds, there is a further issue of providing an exact measure of knowledge so as to determine how much is known by an agent at a particular time (Treanor 2012). Epistemologists also disagree about whether having a certain modal profile is what makes knowledge the state it is. Such disagreement is a metaphysical dispute about what states of knowledge have in common by virtue of which they are knowledge. The following will focus on metaphysical questions about knowledge.

The traditional view is that knowledge is reducible to true belief plus something else. Historically, the third component in addition to truth and belief was assumed to be justification in the sense of available reasons on which the belief is based (see Justification, epistemic). However, Gettier (1963) demonstrated the insufficiency of justified true belief for knowledge, following which a number of remedies were proposed (see Gettier problem). Lehrer and Paxson (1969) insisted that the justification on which the belief is based not be defeated, but increasingly complex attempts at spelling out such a condition only resulted in more complex Gettier counterexamples. What unites these defeasibility analyses is a conception of justification necessary for knowledge as being reflectively accessible. According to these views, the metaphysical nature of knowledge is internal in just that sense. Others opted for a conception of justification as external to the agent’s perspective. For example, Goldman (1979; 1986) took a belief to be justified only if produced by a reliable cognitive process such as visual perception, where reliability of a process type is a matter of a preponderance of true beliefs, irrespective of the agent’s access to any truth ratios (see reliabilism). Such reliabilist justification is deemed necessary for knowledge, and so the metaphysical nature of knowledge is external in just that sense. If a belief is reliably produced, its truth would seem to be non-accidental, but even this process reliabilist view was confronted with Gettier-style counterexamples (Shope 1983).

The foregoing attempts at analysing knowledge as a conjunction of true belief and some internal or external condition can be seen as an exercise in either conceptual or metaphysical analysis (see Conceptual analysis). Epistemologists often fail to clarify which of the two they have in mind. But any reductive analysis of the concept knowledge has metaphysical consequences for the state of knowledge as the extension of that concept. Suppose on the one hand that Gettier has shown that the concepts knowledge and justified true belief are distinct, but on the other hand that the state of knowledge is strictly identical with the state of justified, true belief. This would imply that those two concepts are necessarily coextensive despite their distinctness, which sounds like a bizarre metaphysical coincidence, as Williamson (2000: 29) and Sosa (2015: 16) put it.

If any attempt to decompose knowledge into a motley of true belief plus something else proves elusive, perhaps a better tack is to turn the tables. On Williamson's knowledge-first approach (1995; 2000), knowledge does not a priori factor into individually necessary and jointly sufficient conditions (see Williamson, Timothy). In particular, knowledge is no compound of a mental state and a non-mental condition, but rather the most general factive propositional attitude. That anyway is the default position, according to Williamson, but as Cassam (2009) and Fricker (2009) noted, such a claim is in need of positive argument. Interestingly, as Nagel (2013) observes, developmental and social psychologists have classified knowledge as a mental state. Maybe knowledge is even a natural kind, on a par with other natural kinds from the special sciences, subject to psychological and philosophical study, as Kornblith (2007) claims (see naturalized epistemology). Williamson (2000: 39) does emphasize how factive attitudes matter to us as their distinctive value consist in having an essence that involves a matching between mind and world. In particular, only knowledge justifies belief in that only knowledge constitutes evidence and evidence is what justifies belief. Since believing is a non-factive attitude, believing involves no matching relation. Indeed, knowing is irreducible to any amplified believing. Believing should instead be analysed in terms of knowing: to believe p is to treat p as if p is known. Thus, knowledge sets the gold standard for belief. The more believing approximates to knowing, the more appropriate believing becomes. Belief aims at knowledge, and belief that fails to hit the target is a kind of botched knowing.

Hossack (2007: 2011) also holds that knowledge is fundamental, rather than being reducible to something else or constituted by something else. But he rejects Williamson’s claim that knowledge is an attitude towards a proposition. On his view, knowledge is a mental primitive – that is, a fundamental relation between mind and world, rather than a relation to representations of the world. More precisely, Hossack takes knowledge to be a fundamental relation to facts, where facts are not to be identified with true thoughts or propositions, as in Frege (1918) (see also Frege, Gottlieb (1848–1925)). Facts are, on his view, property instantiations.

A third view is defended by Holton (unpublished manuscript) according to which knowledge and true beliefs take facts, understood as true propositions, as their objects, while false beliefs take propositions as their object. Holton’s view helps respond to the objections Williamson (2000: 44–8) levels against the view that knowledge and belief take different objects. Thus (*) sounds perfectly proper:

(*) I always believed that you were a good friend; now I know it.

But if belief and knowledge take different objects then the anaphoric ‘it’ cannot inherit its referent from the that-clause. A potential downside is that Holton is committed to a disjunctive account of propositions whereby the true propositions – that is, the facts – are ontologically prior to the false propositions – that is, the non-facts.
Citing this article:
Kallestrup, Jesper. The nature of knowledge. Metaphysics of knowledge, 2017, doi:10.4324/9780415249126-P066-1. Routledge Encyclopedia of Philosophy, Taylor and Francis,
Copyright © 1998-2018 Routledge.

Related Articles