acceptability (n.)

The extent to which linguistic DATA would be judged by NATIVE-SPEAKERS to be possible in their language. An acceptable UTTERANCE is one whose use would be considered permissible or normal. In practice, deciding on the acceptability of an utterance may be full of difficulties. Native-speakers often disagree as to whether an utterance is normal, or even possible. One reason for this is that INTUITIONS differ because of variations in regional and social backgrounds, age, personal preferences, and so on. An utterance may be normal in one DIALECT, but unacceptable in another, e.g. I ain’t, I be, I am. Much also depends on the extent to which people have been brought up to believe that certain forms of LANGUAGE are ‘correct’ and others are ‘wrong’: many do not accept as desirable those sentences which the PRESCRIPTIVE approach to GRAMMAR would criticize, such as I will go tomorrow (for I shall go . . .), or This is the man I spoke to (for . . . to whom I spoke). To a LINGUIST, all such utterances are acceptable, in so far as a section of the community uses them consistently in speech or writing. The analytic problem is to determine which sections of the community use which utterances on which occasions. Within a DIALECT, an utterance may be acceptable in one CONTEXT but unacceptable in another.

Linguistics has devised several techniques for investigating the acceptability of linguistic data. These usually take the form of experiments in which native-speakers are asked to evaluate sets of utterances containing those language features over whose acceptability there is some doubt (acceptability tests). It is necessary to have some such agreed techniques for judging acceptability as, especially in speech, very many utterances are produced whose status as sentences is open to question. In one sample of data, someone said, I think it’s the money they’re charging is one thing. The job of the linguist is to determine whether this was a mistake on the speaker’s part, or whether this is a regular feature of a speech SYSTEM; if the latter, then whether this feature is idiosyncratic, or characteristic of some social group; and so on. Such investigations by their nature are inevitably large-scale, involving many INFORMANTS and sentence patterns; they are therefore very time-consuming, and are not often carried out. An utterance which is considered unacceptable is marked by an asterisk; if marginally acceptable, usually by a question mark, as follows:

*the wall was arrived before

?the wall was arrived before by the army sent by the king

These conventions are also used to indicate ungrammatical or marginally grammatical sentences. In linguistic theory, though, the difference between the acceptability and the GRAMMATICALITY of a sentence is important. A sentence may be grammatically correct, according to the RULES of the grammar of a language, but none the less unacceptable, for a variety of other reasons. For example, owing to the repeated application of a rule, the internal structure of a sentence may become too complex, exceeding the processing abilities of the speaker: these PERFORMANCE limitations are illustrated in such cases of multiple EMBEDDING as This is the malt that the rat that the cat killed ate, which is much less acceptable than This is the malt that the rat ate, despite the fact that the same grammatical operations have been used. In GENERATIVE linguistic theory, variations in acceptability are analyzed in terms of performance; grammaticality, by contrast, is a matter of COMPETENCE.