Semantic criteria: what speakers do with words

Even more important is what speakers and writers do with language. When they produce utterances, they perform actions. They act to produce sounds or marks on paper, but the purpose of producing the sounds (in many situations) is to draw the attention of their audience to some entity and to say something about it, to predicate a property of it. Examples of acts – let us use the generally accepted term ‘speech acts’ – are making statements, asking questions and issuing commands (in the broadest sense). These speech acts are prominent in and central to human communication and are allotted grammatical resources in every language. Other acts are not so prominent but are no less central to human communication and relate directly to the different parts of speech.

Two such speech acts are referring to entities and predicating properties of them. In English, the class of nouns, established on formal criteria, contains words denoting entities, and nouns enter into noun phrases, the units that speakers use when referring to entities. This is not to say that every occurrence of a noun phrase is used by a speaker to refer to something; nor is the difference between nouns and other word classes connected solely with referring; nonetheless, speakers require noun phrases in order to refer, and noun phrases can be used to refer only because they contain nouns.

The notion of predication as a speech act is prevalent in traditional grammar and is expressed in the formula of ‘someone saying something about a person or thing’. Predication has been largely ignored in discussions of speech acts, perhaps because it is always part of a larger act, making a statement or asking a question or issuing a command. In English, verbs, including BE, signal the performance of a predication.

Whether adjectives and adverbs are associated with a speech act is a question that has not received much discussion. It is interesting, however, that in traditional grammar adjectives are also labelled ‘modifiers’, a label which reflects the function of these words in clauses. Speakers and writers use verbs to make an assertion about something, and the assertion involves assigning a property to that something. They use adjectives not to make an assertion but merely to add to whatever information is carried by the head noun in a given noun phrase.

Explaining the different word classes or parts of speech in terms of speech acts offers a solution to one difficulty with the traditional definitions; the class of things is so wide that it can be treated as including events; even properties, which are said to be referred to by adjectives, can be thought of as things. In contrast, different speech acts correspond to different word classes. The speech-act explanation also provides a connection between word classes in different languages. Assuming that basic communicative acts such as referring and predicating are recognized by speakers of different languages (communication between speakers of different languages would otherwise be impossible), the words classed as nouns in descriptions of, say, Russian, and the words classed as nouns in descriptions of, say, English, have in common that speakers pick words from those classes when referring. Similarly, speakers and writers pick what are called verbs when predicating, adjectives when adding to the information carried by a noun (that is, when they perform the speech act of modifying) and adverbs when they add to the information carried by a verb or an adjective.

Linguists nowa-days use the term ‘word classes’ and not the traditional term ‘parts of speech’. ‘Word classes’ is neat and self-explanatory but is associated with the idea of words pinned down on the page or in the transcript of speech. ‘Parts of speech’ is not self-explanatory, but it does have the merit of reminding us that we are dealing not with dead text but with speakers and writers doing things with language.