<philosophy of mind> to have access to knowledge that has the properties of knowledge in the ordinary sense, but is not necessarily accessible to consciousness or dependent on warrant or justification.
Cognize is a technical term Chomsky introduces (1980, p. 69) to denote a relation a subject S has to S' knowledge. Just as S can be said to know that p, S can be said to cognize that q, where p and q stand for propositions representing some state of affairs or another. As such, cognising is said to differ very little from knowing in the ordinary sense. But there are some important features of cognising that set it off from the standard conception of what it is to know something.
Perhaps the most salient feature of cognising is that it is a relation primarily -- though apparently not exclusively -- associated with unconscious or tacit knowledge. As Chomsky describes it, cognising "has the structure and character of knowledge" in that it is a matter of knowing-that (see knowledge that}, which is to say it is propositional and may involve belief (1980, pp. 70; 93-94; 1986, p. 269). What distinguishes cognising per se from ordinary knowing is that in many cases, what is cognised is inaccessible to consciousness (1980, p. 70; 1986, p. 269). However, it would seem that cases of explicit or conscious knowing entail cognising as well. Chomsky states, for example, that when we know that p, we cognize that p (1986, p. 265). What he means is that specific facts that are explicitly known may derive from rules and principles that are (presumably) unconsciously cognised (1986, p. 265). As an example, Chomsky cites the case of a person who, having unconscious knowledge of the binding principles, through deduction (or a similar process) can determine whether or not a pronoun and common noun encountered in the same sentence are coreferring (1986, p. 270). Cognising would thus seem to undergird at least some instances of explicit "knowing that" in important ways.
Despite the apparent fact that what is implicitly cognised often issues in conscious knowledge, it is hypothesised that at least some of what is cognised is inaccessible to consciousness in principle. The prototypical example of knowledge that is both cognised and is claimed to be inaccessible to consciousness in principle is a native speaker's knowledge of grammar. Chomsky is careful to point out that to the extent that cognised knowledge is tacit or implicit, it is not the equivalent of knowledge which is conscious but that the holder cannot articulate (1986, p. 271). But although cognising is largely, if not primarily, associated with cognitive structures inaccessible to consciousness, the line between the cognised cognitive unconscious and consciously accessible knowledge is, according to Chomsky, not inviolable. Rather, what is cognised can produce conscious knowledge, as is illustrated by the example of the binding principles' affording explicit knowledge of coreference, described just above. Cognised knowledge is thus not held to be inferentially insulated from ordinary, i.e., conscious, beliefs.
One feature of cognising that sets it apart from many ordinary concepts of knowledge is that what is cognised is held not to consist in warranted or justified belief. Instead, Chomsky holds that cognised knowledge counts as a case of caused belief (1980, pp. 93-95). What this means is that cognised knowledge is caused by triggering experiences, which in at least some instances interact with innate principles to produce specific instances of knowing that p, where p is a cognised content of some sort. A prime example of caused knowledge is of course knowledge of grammar, which Chomsky has long hypothesised to arise from the interaction of an innate universal grammar and the triggering experiences afforded by exposure to a particular language. In Chomsky's example, a native English speaker, when exposed to evidence that the phrase "each other" is a reciprocal, will now that this is the case by virtue of the triggering effect such evidence has on the innate principle of opacity (1980, p. 94). but such knowledge does not, on Chomsky's account, stand as justified or warranted; as he puts it, "[e]ven where there is a triggering experience, this does not supply "good reasons" in any useful sense of the term"; (1980, p. 96).
Although the term "cognize" was introduced to dispel certain ambiguities in Chomsky's theory of the nature of human knowledge, some ambiguities seem to remain. It is not always clear, for example, exactly how or even whether cognising differs from knowledge as such. Matters are not helped by Chomsky's statement, after the term is introduced, that he will simply use "know" to mean "cognize" (1980, p. 70), or by his other statements that appear to conflate the two terms.
Nevertheless, the concept does appear to have two main functions. The first is to de-link knowledge from any necessary dependence on justification or warrant, thus forestalling possible empiricist objections to a theory of knowledge grounded, as Chomsky's is, in content nativism. The second is to provide a covering term encompassing two kinds of knowledge -- explicit and implicit -- that otherwise might be treated as highly dissimilar and possibly unrelated (or in the case of implicit knowledge, simply dismissed outright).
One plausible interpretation of "cognising" then, is that it denotes an epistemological relation that is
a) neutral in regard to whether or not what is known (or "cognised") is known explicitly, and
b) can be fixed independently of justification or warrant.
References
Chomsky, N. (1980). Rules and Representations. New York, Columbia University Press.
Chomsky, N. (1986). Knowledge of Language. New York, Praeger.
See also tacit knowledge, implicit memory, rules.
Daniel Barbiero <barbiero@enigma.com
Chris Eliasmith - [Dictionary of Philosophy of Mind] Homepage
Try this search on OneLook / Google