Pinker - Learnability and cognition : the acquisition of argument structure

Series Foreword

This series in learning, development, and conceptual change will include state-of-the-art reference works, seminal book-length monographs, and texts on the development of concepts and mental structures. It will span learning in all domains of knowledge, from syntax to geometry to the social world, and will be concerned with all phases of development, from infancy through adulthood.

The series intends to engage such fundamental questions as

The nature and limits of learning and maturation: the influence of the environment, of initial structures, and of maturational changes in the nervous system on human development; learnability theory; the problem of induction; domain-specific constraints on development.

The nature of conceptual change: conceptual organization and conceptual change in child development, in the acquisition of expertise, and in the history of science.

Lila Gleitman

Susan Carey

Elissa Newport

Elizabeth Spelke

The Secret Life of Verbs: A Preface to the New Edition

Learnability and Cognition is a technical book on how people learn, understand, and use verbs in sentences. Though it was written for students and researchers in linguistics and cognitive science, this monograph contains the seeds of the popular books that I have written in the years since. Learnability and Cognition features not just busy John, long-suffering Mary, and the rest of the dramatic personae of university-press books in linguistics, but also lyrics from Bob Dylan, wisecracks by Dolly Parton, and strange usages in TV ads, sitcoms, hobbyist magazines, and unguarded emails, together with a recipe for making litmus paper from cabbage juice.

The books subject matter, too, would prefigure my later books on big themes in the study of human nature. The phenomena that whipped me into a multiyear obsessional frenzy when I wrote this book in the 1980s struck me as rich enough two decades later to inspire a book that would become a New York Times bestseller, The Stuff of Thought. In Stuff, I compared the discovery of conceptual structure (which in this book serves as the solution to a paradox in verb learning) to Alices stumbling down a rabbit hole and finding a phantasmagoric underworld. In figuring out how children master contrasting verbs like fill and pour or give and carry, I had to ponder the human concepts of causation, space, time, matter, and purpose. The syntax of verbs, I discovered, interacts with the stuff of thought, and thereby inspired not just the title of my popular book but its subtitle: Language as a Window into Human Nature.

If my past experience is a guide, you are not reading these words right now. I have written forewords or afterwords to new editions of four of my other books, and in the years since they appeared they have led to not a single citation or piece of correspondence. So in this essay I will not try to update every section with research summaries that no one will ever read. Clearly I would not allow Learnability and Cognition to be republished if I did not think that its analyses were still relevant to todays cognitive scientists (no one has ever gotten rich from a book on the acquisition of argument structure), and I will simply point to some general literatures that continue the lines of thought that I introduced here.

Foremost among them is, of course, The Stuff of Thought. Chapter 2, Down the Rabbit Hole, summarizes the argument in this book without most of the linguistic and theoretical detail, and brings it up to date with citations of new cross-linguistic surveys and corpus analyses. (Other than the Child Language Data Exchange System, large corpora of adult speech were not available at the time that L&C was written, which was the era of the ARPANET, the 1,200-band modem, and the ten-megabyte hard drive.) The meticulous analyses of the English verb system by Beth Levin and Malka Rappaport Hovav which I relied on so heavily in these pages have been updated in their 2005 book Argument Realization, and have been implemented by Martha Palmer in the online lexicon VerbNet (http://verbs.colorado.edu/~mpalmer/projects/verbnet.html).

The Stuff of Thought also has a chapter called Fifty Thousand Innate Concepts (and Other Radical Theories of Language and Thought), which defends the very idea of conceptual structure (something that L&C did not do) against alternatives such as Jerry Fodors claim that verb meanings are atomic and innate, the claim of radical pragmatics that word meanings are infinitely fluid and evanescent, and the claim of linguistic determinism that conceptual structure is a product of language acquisition rather than one of its essential enablers. And chapter 4 of Stuff, Cleaving the Air, makes good on the promise that lexical semantics can be a window into our deepest concepts by probing our conceptions of matter, space, time, and causality and how they are expressed in language. Among other things, Stuff expands on the theories of two of the semanticists whose work I depended on in these pages, Ray Jackendoff and Len Talmy. Both of them have since come out with their own masterworks: Jackendoff wrote Semantic Structures, Meaning and the Lexicon, and A Users Guide to Thought and Meaning; Talmy compiled his lifes work into the two-volume set Toward a Cognitive Semantics.

In retrospect, it was wise of me to remain more-or-less agnostic in my choice of grammatical formalisms, because the reigning grammatical theory of the time, Noam Chomskys Government and Binding, has been discarded by its erstwhile proponents, and its successor, the Minimalist Program, has offered little new insight into the relationship between lexical semantics and argument structure (or, Jackendoff and I have argued, into anything else). My use of a diluted version of Joan Bresnans Lexical Functional Grammar has allowed the analyses in L&C to be upward-compatible with current versions of her theory, as well as with other plausible contemporary theories of grammar such as Ray Jackendoffs Parallel Architecture, the descendants of Generalized Phrase Structure Grammar, and the various versions of Construction Grammar such as those developed by Ronald Langacker, Adele Goldberg, and William Croft. Indeed, my notion of the thematic core of an argument structure, which delineates the conflation class of verbs compatible with that argument structure, is very close to the idea of a construction meaning invoked by theories of construction grammar.

A frequently asked question about the relationship between the two approaches is whether my lexical rules (which map between related semantic structures, and hence, according to the theory, related argument structures) are needed at all. Couldnt the meaning of a verb be mapped directly onto its argument structure (or construction), so that verbs with two or more related meanings would simply map onto two or more constructions? The answer is that yes, lexical rules are needed, and no, meaning-to-construction mappings are not enough. The reason is that lexical rules govern the distribution of morphological roots among related meanings, in particular, whether a root may be carried over intact from one construction to another (as in Hand me the book and Hand the book to me), must be modified morphologically (as in He will hand the book over to me and The book will be handed over to me), or are not eligible to appear in the related construction at all (as in She killed/*died the spider and The spider died/*killed). Mappings that connect meanings to constructions directly, while saying nothing about the lexical roots that express each meaning, cannot account for these differences.

I have been pleased to see that the original phenomena that inspired this bookchildrens uses of verbs in sets of related argument structuresare being studied by developmental psycholinguists with increasing precision and ingenuity. Jess Gropens experiments and transcript analyses of the locative and dative were published in