From the first chapter of Ray Jackendoff's 2007 book Language, Consciousness, Culture: Essays on Mental Structure, some key comments about linguistics and thinking:

... It is my impression that of all the cognitive sciences, only linguistics has systematically and explicitly investigated the content of mental structures that underlie a human capacity. The rest of cognitive neuroscience has for the most part made do with relatively rudimentary notions of mental structure, exploring more intensely issues of neural localization and/or the "horizontal" capacities of working memory, attention, learning, and the like. ... (with the exception, Jackendoff notes, of vision, music, and some complex actions)

...

... The neurons deep inside the brain that are responsible for cognition have no privileged access to the "real world"; they interact only with other neurons. Contact with the "real world" is established only through long chains of connection leading eventually to sensory and motor neurons. If this is the hardware on which mental capacities "run," then mental capacities too are necessarily limited in their contact with the "real world." They are sensitive to the outside environment only insofar as they are connected through functional (or computational) links to the sensory and motor capacities. ...

...

Some neuroscientists say we are beyond this stage of inquiry, that we don't need to talk about "symbols in the head" anymore. I firmly disagree. We know that language is organized into speech sounds and that speech sounds are only the first step in analyzing linguistic structure. As far as I know, there exist absolutely no attempts to account for even this trivial degree of linguistic complexity in neural terms, and speech sounds only scratch the surface. In my opinion, it is the height of scientific irresponsibility to totally dismiss linguistic theory, claiming that some toy system (say a computational neural network) will eventually scale up to the full complexity of language. A linguist who made comparably ignorant claims about the brain would be a laughingstock. End of sermon. ...

...

... In both language and vision, if we want to figure out how the brain works, it behooves us to try to understand what functions the mind has to compute. A proposed theory of neural behavior is incomplete if it does not offer genuine solutions to the problems of combinatoriality, structural hierarchy, and binding among structures.

It is not that these problems are particular to language. It is just that linguistic theory focuses on these problems and builds on them in a way that theories of other "vertical" faculties of mind usually have not. Part of the message of this book is that these properties recur in other faculties, should we care to look for them. Sixty years ago, nearly everyone thought that language was perfectly transparent and hardly complex at all and many nonlinguists, even some in psychology and neuroscience, still think so). Since then we have learned that not only is language far more complex than we ever would have dreamed, but so is every other aspect of the mind/brain that has been investigated.

To sum up: Pretty much all cognitive neuroscientists agree in rejecting dualism; ultimately the mind must run in the brain, and there are no mental properties that are causally independent of brain events. However, to insist that neural accounts have absolute priority, that they somehow have a greater reality or are "more scientific" than functional accounts, to me has a chilling effect on inquiry. It seems to me that in the practice of research, the relationship between neural and functional accounts ought to be a two-way street: what we know about each dimension of the problem ought to enrich our study of the other. ...

_{(cf Mental Bandwidth Boosters (1999-06-26), Higher Level Language (2007-08-17), Babel-17 (2015-10-20), Patterns in the Mind (2024-05-12), ...) - ^z - 2026-01-21}