++EngLiSh DePaRtMeN++

The Description of English

A description of any language should contain an inventory of the building blocks; sounds and morphemes, roughly. It should also contain the rules for how those elements can be combined; phonotactic constraints, information about which differences between sounds are distinctive, how morphemes can be combined to form words, and how words can be combined to form phrases. In spite of the attention that the language has received, no complete description of English in this sense has yet been provided. To take but one example, even though there are many insightful descriptions of the English passive, the exact rules that allow for sentences such as This road has been walked on have not been provided. The view of a grammatical description just described coincides with the original conception of a ‘generative’ grammar. A generative grammar in that sense takes the building blocks of a language and ‘generates’ all and only the grammatical sentences of that language. Needless to say, no complete such grammar has been defined, not for English and not for any Associated with the question of what constitutes a description of English is the question of what such a description describes.

Traditionally, the object of description has been a variety of English referred to as the ‘standard.’ Many grammars of course aim not only to describe this variety, but also to prescribe it; to describe a variety which native speakers of English should aim to follow. Even though modern grammars of English such as Quirk et.al. (1985) and Huddleston and Pullum (2002) avoid prescriptivism, descriptions which aim also to prescribe are still prevalent, as witness the popularity of books such as Trask (2002). Descriptions of varieties of English other than the standard do, however, also have a long tradition. There are many good grammars of geographical dialects within Britain (for examples and references, see for instance Hughes and Trudgill 1980, Milroy and Milroy 1993), the US (e.g. Wolfram and Schilling-Estes 1998) and to some extent Australia and New Zealand (e.g. Burridge and Mulder 1998). See also Kortmann (this volume). Increasingly, varieties of English which have arisen in countries where English has not traditionally been the first language are also considered varieties in their own right and are described as such and not as examples of “English not used properly.” This has led to an area of study known as World Englishes (e.g. Trudgill and Hannah 2002). A description of a language, regardless of how one selects the particular variety, has to be based on data and a further issue involved in description is how to select these data. Although most descriptions rely on a mixture of types of data collection, a number of types can be distinguished. These are described in more detail in Meyer and Nelson, but given the direct way in which they impact on the relation between data and theory, we will discuss them briefly here. Each approach has advantages and disadvantages, and all of them involve some degree of idealization. An approach that has not been uncommon in descriptions and in theoretical work is introspection; the author of the description considers whether he or she would accept a particular pronunciation, a particular phrase or sentence and uses these judgments as a basis for the description. An advantage of this approach might be that a linguistically trained person can provide more subtle judgments, whereas non-trained native speakers might find it difficult to make the distinction between ‘is grammatical’ and ‘makes sense,’ a distinction which The disadvantages of this approach is crucial both for description and theory are, however, also obvious; even trained linguists might not have a good awareness of what they actually say.

There are examples of linguistic articles in which a construction is attested which is claimed in the description or in the analysis not to exist. The introspective approach is particularly dangerous in theoretical work within a particular framework, where the desire to provide a neat analysis within the favored theory may cloud the linguist’s native speaker intuitions. A more reliable way of collecting the data is to elicit grammaticality judgments from a group of native speakers or to get their judgments in a more subtle way through picture description tasks or similar processes. In an approach like this, a consensus view can emerge and peculiarities of individual speakers are ruled out. However, data collected in this way may deviate from naturally occurring data. The notion of a simple grammaticality judgment is not a straight forward one to most native speakers. If the speaker is aware of some high status standard which differs from their own variety, this may also interfere with their judgments, and in cases where there is no obvious standard, it may actually be difficult to get a definite judgment from native speakers. The use of corpora avoids many of the drawbacks identified with using native speaker judgments in that it allows wide-ranging studies of naturally occurring language. Especially with the existence of large-scale electronically available corpora, this has become an important tool for the study of all varieties of English (see McEnery and Gabrielatos) Biber et al. (1999) is an example of a corpus-based grammar of English.

There are of course drawbacks, especially in that the absence of a particular construction in a corpus cannot be taken as evidence that this construction is absent from the language. This is a familiar problem for those working on varieties for which there are no longer any native speakers, for whom corpus study is the only option. Similarly, constructions which would be described as ungrammatical by the vast majority of the language community may occur in corpora, say as speech errors, or in historical texts in the form of scribal errors. Most descriptions of English are based on the written language, though modern grammars do refer to alternative constructions which occur in the spoken language but which are infrequent in written form. Biber et al. (1999) is an exception in that it is partially based on spoken corpora. Miller and Weinert (1998) go one step further and describe spoken language as a separate variety with a partially different grammar from the spoken language (see also Miller).(giriums.blogspot.com)