Psycholinguists have devised experiments to try to learn how children and adults are able to acquire words so easily. You might think that the learning of new words is a simple matter of association: someone points at something and says “flurge” and you learn that that something is called a flurge. This may be the way that we learn some words, but surely not the way we learn the majority of words in our mental lexicons. For one thing, not everything for which we have a word can be pointed at.

And even if someone points and says a word, it is often not clear from the context what exactly is being pointed out. Psycholinguists sometimes call this the Gavagai problem, following a scenario first discussed by the philosopher W.O. Quine. To summarize:

Picture yourself on a safari with a guide who does not speak English. All of a sudden, a large brown rabbit runs across a field some distance from you. The guide points and says “gavagai!” What does he mean?

One possibility is, of course, that he’s giving you his word for ‘rabbit’. But why couldn’t he be saying something like “There goes a rabbit running across the field”? or perhaps “a brown one,” or “Watch out!,” or even “Those are really tasty!”? How do you know?

In other words, there may be so much going on in our immediate environment that an act of pointing while saying a word, phrase, or sentence will not determine clearly what the speaker intends his utterance to refer to.

Besides, we are rarely in a situation in which someone is actively instructing us about the meanings of words; although parents may point to things in a picture book and name them for a child, or school children may be asked to memorize a list of vocabulary words, we learn most words without explicit instruction and seemingly with very little exposure. Although we do not know nearly enough about this subject, there are several things that we do know about how word learning occurs.

First, it is believed that both children and adults are able to do what the psycholinguist Susan Carey has called fast mapping (Carey 1978). Fast mapping is the ability to pick up new words on the basis of a few random exposures to them. In one experiment, Carey showed that children who were casually exposed to a new color name chromium during an unrelated activity (following instructions to pick up trays of various colors) were able to absorb the word and recall it even six weeks later. Experiments have shown that adults exhibit this fast mapping ability as well; while the ability to learn linguistic rules (say, of syntax or phonology) is thought to decline after puberty, the ability to learn new words remains robust.

Psycholinguists have proposed a number of other strategies that both children and adults seem to use in learning new words.2 One might be called the Lexical Contrast Principle. For example, in an experiment similar to yours, children were asked to point to the zorch (or some other made-up word), and what they invariably did was to point out the unfamiliar object. According to the Lexical Contrast Principle, the language learner will always assume that a new word refers to something that does not already have a name.

A second word learning strategy might be called the Whole Object Principle. In the experimental condition described above, when subjects are presented with the word zorch and an unnamed object, they will assume the whole unnamed object to be a zorch. They will not assume that zorch refers to a part of the object, to its color or shape, or to a superordinate category of objects to which it might belong.

A related strategy might be dubbed the Mutual Exclusivity Principle. In the second experiment above, there are only familiar objects for which subjects already have names. When asked to point out the plitz, experimental subjects typically do one of two things: they might first look around the room for something else that might be called a plitz, or they might assume that the word plitz refers to a part of one of the familiar objects or a special type of one of them. Subjects, in other words, will assume that if an object already has a word for it, the word plitz cannot be synonymous with those words.

These experiments are of course not just hypothetical. Paul Bloom, Susan Carey, and many other psycholinguists have conducted them both with children of various ages and with adults, and have obtained the results described above. What is perhaps most astonishing about their results is that their experimental subjects often remember the words they’ve been exposed to when they are retested weeks after the original experiment. But maybe we should not be surprised by this: how otherwise could we have learned 60,000 words by the time we’re 18?

Children not only learn individual words, but they learn the rules that allow us to create and understand new words. Indeed, there is evidence that English-speaking children as young as 18- to 24-months old are able to create new compound words (that is, words like wind mill or dog bed) and to turn nouns into verbs, a process which is called conversion. Not too long after this, children will begin to use prefixes and suffixes, both for inflection and lexeme formation. We know that they have learned the rules when they produce words that are novel and therefore that they could not have learned from the language spoken around them.