Descriptive linguistics – Part 1
Descriptive linguistics – Part 2

Thus, if we compare John worked and The king of England opened the par­liamentary session, then The king of England may be the extension of John, while opened the parliamentary session is the extension of worked.

The notion of extension is extremely important for Wells, since, in terms of his theory, to define immediate constituents means actually pointing out extensions that make up a certain sequence of morphemes. It is the principle of extension that allows to break the sentence The king of England opened the parliamentary session into The king of England + opened the parlia­mentary session. It is impossible to divide the sentence into The king + of England opened the parliamentary session, since of England opened the parliamentary session cannot be regarded as extension of anything shorter.

Principles and methods of descriptive linguistics are systemically high­lighted by Harris in Methods in Structural Linguistics published in Chicago in 1951. Harris sees the aim of descriptive analysis in studying the structure of the sentence in terms of morphemic classes and their positions. Accord­ing to Harris, the sentence is a segment of speech produced by one speaker and separated from all the preceding and following speech with a pause. The scholar insists on eliminating the distinction between morphology and syntax. It should be added that, though Harris differentiates between mor­phological and syntactic criteria, these terms acquire in his book a specific meaning, since Harris on principle ignores the difference between a word, its part, a word combination and a sentence. To overlook this difference is characteristic, to a greater or lesser extent, of all descriptive linguists.

Thus, the morphological structure of a sentence, according to Harris, is a sequence of the most general morphemic classes and constructions (the term “construction” stands here either for a word or for a phrase). It is typical of Harris to formalize syntactic studies making use of mathematic symbols. For example, classes of mutually interchangeable forms are described by other linguists as forms or morphemic sequences occurring in similar envi­ronment. Harris, in his turn, expresses this in the following way: if A and В occur in environment C – D, they are members of the same class. Harris also distinguishes between morphemic classes: class N occurring before -s (plu­ral inflection) and after the and adjectives; class V before -ed, -ing and after should, will, etc.; class A occurring between the and N; class D used between the and A but never occurring between the and N (e.g. rather and very). 

Harris believes that having defined elements and their classes, one may describe any speech segment as combination of these elements, applying the same method on all the levels – phonological, morphological and syntactic.

Differences between morphological and syntactic criteria, used to dis­tinguish between various morphemes, are analyzed by Harris as follows. Morphological criteria are necessary to consider the immediate environ­ment of morphemes, e.g. -ly in largely. But these criteria are not enough. It appears necessary to take into account the “broad environment” of a mor­pheme, i.e. the whole sentence, which means application of the syntactic criterion. It is on the ground of these criteria that larger morphemic classes are distinguished. For example, for English largely and manly the immedi­ate environment is -ly. However, largely is classified as a qualitative ad­verb, whereas manly is an adjective. To reveal this distinction between these seemingly identical cases, it is necessary to consider their positions and re­lations with the environment.

For the English language, the final aim of the analysis is to bring to light such syntactic patterns as Noun + Noun, Noun + Verb + Preposition, Noun + Verb + Preposition + Noun, etc. Consequently, descriptive linguistics deems such notions as “sentence parts”, “subject”, “predicate” as meaningless and refuses to operate with them, which leaves the notion “sentence” useless as well. Harris does not explicate the methodology of distributive analysis, but it may obviously be reduced to the following stages: 1) segmenting of a sentence into components, and 2) comparing the components and referring similar components to groups.

Descriptive linguistics – Part 4