particular so as to solve the Proble mof 2 ( multiple instances in working me mory of a single long-ter m
memory type)?- How does the brain instantiate learning, such that pieces assembled in workingmemory can be installed into
long-ter m me mory? - How does one-trial learning differ fro mslow, many-trial learning?
- How is innate specialized f-knowledge instantiated in the brain, and how does it guide behavior and learning?
- How is innate specialized f-knowledge coded on the genome, and how does it guide brain development,
including the characteristic localizations of different functions?
Oneofthethings I tried tostress about these questions as theycameup is thattheyare notparticular tolanguage.The
same questions are posed by the visual system and the formulation of action—for any organis mthatfinds its way
around its environment and behaves with any degree of complexity. Language just happens to be a domain where
some of the questions have been more a matter of open debate than elsewhere.
In reconciling linguistic theory with a mentalistic instantiation, it has been important here to abandon the idea that
linguistic entities in the brain aresymbolsorrepresentations. We have instead been able to treat the msi mply asstructures
built of discrete combinatorial units. In the case of phonology and syntax this was not too painful. But in the case of
semantics it was necessary to spend much of Chapter 10 working out an alternative to the mystery of intentionality:
how it is that language seems to refer to the world despite the fact that all the machinery instantiating language is
trapped in our brains. However, again the issues prove not to be particular to language; rather they pertain as well to
vision and other perceptual/cognitive modalities—in any organis mthat perceivesthe physicalworldroughlyas we do.
It is just that this particular way of framing the issues traditionally comes to us from philosophy of language.
A more technical question about memory arises from the discussion of semi-regular patterns and inheritance
hierarchies in Chapter 6. In the present framework (as well as some related frameworks), a great deal of theoretical
weight has come to rest on the idea that the taxonomic relations and redundancies among items somehow allow them
to“cost less.”For example, in some sense the idiomtake advantage ofought to be in some sense“cheaper”thantake
umbrage at, becauseadvantageis an independent word andumbrageoccurs only in this idiom. Similarly, the irregular past
tensebroughtshould be“cheaper”thanwentbecause (a)broughtbears some phonological similaritytobringwhereaswent
bears no relation togo, and (b)fought, taught, bought, sought, andcaughtshare the same distant relation to their present
tenses. Yet standard