Steven Pinker

 about  books  research  teaching  articles  lectures & media  photos 

Department of Psychology
Harvard University

Research

Bibliomics Science cover

BIBLIOMICS

Erez Lieberman Aiden and Jean-Baptiste Michel, working with a team at Google, have developed a tool (ngrams.googlelabs.com) that searches a half-trillion word corpus of five million books, a significant fraction of all the books published since the fifteenth century. We are using this resource to study the vocabulary and grammar of English and how it has changed over the past several centuries.

ANALOGICAL REMINDING

This project, conducted with Harvard graduate student James Lee, seeks to explore a cognitive mechanism underlying the formation of new conceptual metaphors (like the use of words for motion as a way of talking about changes of state, or the use of terms for war in discussing arguments). With the use of spontaneous and elicited remindings, we are seeing whether people naturally think about one idea if it is related to another via shared conceptual structure, not just shared sensory features.

INDIRECT SPEECH

James Lee and I (in collaboration with Martin Nowak and Jennifer Kan are trying to understand why people so often veil their threats, bribes, solicitations, and sexual come-ons in innuendo rather than blurting them out explicitly. Using both questionnaires and in vivo methods, we are testing the possibilities that indirect speech is perceived as more considerate, is aimed at a virtual audience, introduces uncertainty, or keeps shared knowledge between the participants from becoming common knowledge.

Click here to participate in a web-based study on indirect speech.

Inflectional Morphology

INFLECTIONAL MORPHOLOGY

This project focused on inflections such as the past tense and plural to understand the psychology of language. The key idea is that irregular forms such as come-came and mouse-mice are unpredictable and have to be memorized, whereas regular past tense forms such as walk-walked and rat-rats are rule-governed and can be generated on the fly, particularly to new forms such as mosh-moshed and blog-blogs. Because the two forms are matched for meaning, length, and complexity, they can be used to contrast the two mental mechanisms underlying language: memory (for simple words) and combinatorial grammar (for complex words, phrases, and sentences). We have studied how past tense forms develop in children, how they are processed in speech production, how they vary (and don't vary) across languages, how they change in history, how they are impaired in neurological disorders, and how their details can be explained by linguistic theory. Click here for a list of publications from this project.

Click here to participate in a set of web-based experiments on meaning, sound, and inflectional forms of nouns and verbs, conducted in collaboration with Harvard graduate student Yi-Ting Huang.

Twins

THE TWINS STUDY AT HARVARD

This project, conducted in collaboration with Jennifer Ganger, gathers longitudinal data on the development of words, sentences, and past-tense forms in a large sample of monozygotic (identical) and dizygotic (fraternal) twins. If milestones of language development appear in closer synchrony in identical twins (who share all their genes) than in fraternal twins (who share half their genes, among those that vary), it would suggest that language development is paced in part by a genetically influenced maturational timetable. Click here for the Twins Study at Harvard web site.

neuroimaging of Inflection

NEUROIMAGING AND INTRACRANIAL RECORDING OF LANGUAGE PROCESSES

In collaboration with Ned Sahin and Eric Halgren of the University of California, San Diego, I have been using functional magnetic resonance imaging (fMRI) and intracranial encephalography (iEEG) with epilepsy patients to investigate the neural organization of language. For example, we compare patterns of brain activity when people repeat words and when they generate forms with regular (he walked), irregular (he ran), and zero-marked (they walk) inflections. The goal is to characterize the neural processes involved in reading words, pronouncing words, choosing appropriate grammatical features, and implementing the features as suffixes or irregular forms. Click here for a recent conference poster reporting preliminary results.

THEORETICAL PAPERS

Now and again I write theoretical papers on topics in cognitive science, including language acquisition, the evolution of language, verb learning, visual cognition, cognitive architecture, and the nature of concepts. Click here for a list.