CHAPTER 9 Semantics as a Mentalistic Enterprise

9.1 Introduction to part III

As observed in Chapter 4, meaning is the“holy grail”not only of linguistics, but also of philosophy, psychology, and
neuroscience—nottomention moredistantdomains such as cultural and literary theory. Understandinghowwe mean
and how we think is a vital issue for our intuitive sense of ourselves as human beings. For most people, meaning is
intuitivelythe central issue in the study of language—far more important than understanding details of word order or
morphology.

I think linguists have tried over the years to make a big deal out of how thestudy of language teachesus about human
nature. In the days when Deep Structure was claimed to be the key to meaning, such advertising was properly
seductive. But more recently, when the parade examples have concerned nitty-gritty phenomena like pronoun use and
word stress, the public has been less impressed. To be sure, there are important points to be made here about the
nature of learning and innateness—but what people really want to know about is still meaning. And generative
grammar on the whole has not made good on the promise so tantalizingly held forth inAspects.

Thereismoreatissuethanphilosophicaldiscussionsofhuman nature.As pointed outas long ago as Bar-Hillel(1970),
potential practical applications of linguistic theory such as machine translation are hobbled without an account of
meaning. Computational linguists, I am told, joke that every time they hire a theoretical linguist, their programs
become less effective. The problem is that fancy syntax alone isn't that much use for machine understanding.

Most of my own work for the past thirty years has been directed toward developing an account of meaning that is
compatible both with the psychological foundations of generative grammar and with the spirit of its formal
technology—thus parting company with mainstream concerns in generative grammar. In the