Another important area of language development involves learning the meanings of words, a process that continues throughout one’s lifetime. As we mentioned earlier, children’s first words tend to be things in their immediate environments that adults talk about to them, as well as some words involved in social interaction. But if we think about what is ultimately involved in “knowing” what a word means, we cannot fail to be impressed with the complexity of such knowledge. We must, for example, figure out what other entities we encounter can be labeled with the same word. The family pet is doggie, but what else can I use this word for? As we have seen, even very young children have the idea that labels must be extended to other objects, although they initially may overgeneralize or undergeneralize word meanings. We must also learn that words occur at different levels of abstraction. For example, the family pet is a doggie but it is also an animal. We must learn, however, that while all doggies are animals, not all animals are doggies and other “non-doggies” can also be animals. Furthermore, to use a word appropriately, we must also know how to pronounce it, what part of speech it belongs to, what other words it can occur with, what grammatical endings can be attached to it, and whether it is regular or irregular in its morphology. We also come to know which words may have more than one meaning, which are synonyms, and which are antonyms. The task seems difficult enough even for simple nouns with clear and concrete referents, but that much more difficult for actions, abstractions, and function words.

A great deal of research has been devoted to the acquisition of word meanings and how we arrive at more-or-less shared meanings for words that we have acquired under a wide range of individual circumstances. While all researchers recognize that learning word meanings requires interaction with adults, they differ in their perspectives on how active and independent children are in forming their own hypotheses. Some researchers assume that children approach the task with certain conceptual constraints that help them narrow down their guesses in the very early stages of acquisition. Clark (2003) outlines several of these proposed constraints. One says that children will assume you are naming a whole object, rather than just part of an object, when it is pointed out to them. Doggie is the whole animal, not just its tail or its paw. Another constraint says that children will isolate an object from its surroundings, so that doggie means the animal but does not include the rug it is lying on or the bone in its mouth. Another such constraint proposes that young children favor basic categories (apple as opposed to fruit or Macintosh) when they are learning new words. It has also been proposed that if very young children already know a term for an object, they will not accept another one. So, if they know dog, they will not accept animal or poodle as alternative labels for objects they have classified as dogs.

Another view of early word acquisition focuses on the fact that words often have meaning components, such as “female” (queen, princess, maiden, girl) or “relative” (brother, mother, father, sister) and suggests that children acquire word meanings by acquiring such semantic features bit by bit to build into word meanings. Thus, a child might initially class together all things with the feature “round” and overgeneralize the label ball for all such objects. Once the features are learned that distinguish such objects from one another, the child can assign more appropriate labels. One widely studied semantic feature in child language acquisition is “cause,” which occurs in verbs such as break and bend when they are used as transitive verbs. I broke the glass means “I caused the glass to break.” Children often seem to recognize this semantic component and overextend it to verbs that do not contain it in the adult language, as in I falled it.

Clark raises questions about the explanatory value of the above constraints and strategies for learning word meanings and argues for a much more prominent role for pragmatics in understanding how children acquire such knowledge. That is, she focuses heavily on the contexts in which adults and children interact to provide word-learning opportunities. She proposes that young children make two pragmatic assumptions about word meanings that guide them in their semantic development (2003, pp. 143–144). The first, conventionality, assumes that there are agreed-upon terms in the language community, that there is reliability and predictability in the association of a term with a meaning. The second, contrast, assumes that two different forms will not be used for the same meaning; that is, if the forms of words are different, their meaning will be thought of as different as well. Using these two assumptions, combined with their previous word knowledge and their assumptions about what the adult intends to say in that context, children problem-solve their way to a meaning. For example, Clark (2003, p. 147) cites a study in which four- and five-year-olds were asked to pick out a “chromium” tray from a set of differently colored trays. They were told to pick the chromium one, “not the red one.” From this instruction, the children were able to infer that chromium was a color, and based on their prior knowledge of color terms, they were able to choose the one color that was unfamiliar to them. Interestingly, several weeks afterwards, the children remembered that chromium was a color and used it appropriately.

Children also make use of the context in which a word appears to infer its meaning. That means, of course, that different children may start off with very different ideas about what a word means and only gradually converge on a common meaning after a great deal of exposure in many different contexts. The adjective small, for example, will mean different things depending upon what the frame of reference is. A baby elephant can be small relative to its mother, and a dime can be small relative to a quarter. It is therefore difficult to say exactly when a child has learned the meaning of a word. There may be times when usage of the word is appropriate for the context and other times when it isn’t while the child gathers enough information to flesh out its full meaning. Some “errors” may show themselves quite late, such as when a four-year-old says “Don’t let me do it” for “Don’t make me do it.” According to Clark (2003), these meanings are fine-tuned gradually through adult correction, either directly or indirectly, as well as by the child’s attention to adult usage.

Clark (2003) outlines a concept of meaning acquisition in which children develop pragmatic skills at a very early age, enabling them to glean meanings from the conversations in which the words are used. In the next section we take a closer look at what those pragmatic skills consist of in young children and how they develop as children mature.