Whorf and Vygotsky Revisited

Psychologists often speak about language and meaning in terms of individual "words" that label "things" in the world. Such expressions emerge not only in casual speech, but also pervade the scholarly literature. Influential works such as Roger Brown's (1958) Words and Things lend support to this tendency in their titles, if not always in their detailed argu ments. Speaking in this way perpetuates the illusion that language and speaking are pri marily about individual words, words con ceived of as sturdy self-sufficient forms with straightforward referential meanings, forms that we take out into the world to attach to various objects we encounter in our experi ence, much the way a hunter might take a snare out into the forest to capture rabbits (or other "natural kinds"). This view, in turn, strongly entails another, namely, that word meanings are merely derivative of experience, simply labeling and thus ultimately responding to objects, rather than having their own internal value and logic that can play a dynamic role in the constitution and conceptualization of experience. This deriva tive view of word meaning, now serving as a proxy for all of language, leads quite naturally to a view of human cognition in which lan guage plays a secondary, handmaiden role at best. In this way, our own "fashion of speaking" about language in terms of "words" shapes our scientific understanding of the significance of language for thought. Yet over a century of systematic linguistic research shows that this view of words and hence word meaning is unsustainable on sev eral counts. First, what qualifies as a word form can be difficult to determine and is often the product of analysis rather than its starting point, not only for scientists but also for chil dren learning language. Although this point can be difficult to appreciate for speakers of languages (such as English) that happen to have large numbers of minimal, freestanding forms, it is transparently obvious in other lan guages in which most forms are internally complex or bound to other forms. Second, many word forms do not actually refer to "things" (or "objects") but rather to events or properties, or even to nothing external at all, having essentially a grammatical function (e.g., English of). And those words that do refer to objects may do so in remarkably dif ferent ways, selecting this or that feature as the basis for reference (compare Quine, 1960). Finally, perhaps most importantly, even refer ential word meanings depend heavily on the