И. В. Арнольд лексикология современного английского языка Издание

§ 1.4 TYPES OF LEXICAL UNITS
The term unit means one of the elements into which a whole may be divided or analysed and which possesses the basic properties of this
¹ See Chapter 8.
² A nonce-word is a word coined for one occasion, a situational neologism: (for the) nones — by misdivision from ME (for then) ones.
18
whole. The units of a vocabulary or lexical units are two-facet elements possessing form and meaning. The basic unit forming the bulk of the vocabulary is the word. Other units are morphemes that is parts of words, into which words may be analysed, and set expressions or groups of words into which words may be combined.
Words are the central elements of language system, they face both ways: they are the biggest units of morphology and the smallest of syntax", and what is more, they embody the main structural properties and functions of the language. Words can be separated in an utterance by other such units and can be used in isolation. Unlike words, morphemes cannot be divided into smaller meaningful units and are functioning in speech only as constituent parts of words. Words are thought of as representing integer concept, feeling or action or as having a single referent. The meaning of morphemes is more abstract and more general than that of words and at the same time they are less autonomous.
Set expressions are word groups consisting of two or more words whose combination is integrated so that they are introduced in speech, so to say, ready-made as units with a specialised meaning of the whole that is not understood as a mere sum total of the meanings of the elements.
In the spelling system of the language words are the smallest units of written discourse: they are marked off by solid spelling. The ability of an average speaker to segment any utterance into words is sustained by literacy. Yet it is a capacity only reinforced by education: it is well known that every speaker of any language is always able to break any utterance into words. The famous American linguist E. Sapir testified that even illiterate American Indians were perfectly capable of dictating to him — when asked to do so — texts in their own language “word by word”. The segmentation of a word into morphemes, on the other hand, presents sometimes difficulties even for trained linguists.
Many authors devoted a good deal of space to discussing which of the two: the word or the morpheme is to be regarded as the basic unit. Many American linguists (Ch. Hockett or Z. Harris, for instance) segmented an utterance into morphemes ignoring words. Soviet lexicologists proceed from the assumption that it is the word that is the basic unit, especially as all branches of linguistic knowledge and all levels of language have the word as their focal point. A convincing argumentation and an exhaustive review of literature is offered by A. A. Ufimtseva (1980).
If, however, we look now a little more closely into this problem, we shall see that the boundaries separating these three sets of units are sometimes fluid. Every living vocabulary is constantly changing adapting itself to the functions of communication in the changing world of those who use it. In this process the vocabulary changes not only quantitatively by creating new words from the already available corpus of morphemes and according to existing patterns but also qualitatively. In these qualitative changes new morphemic material and new word-building patterns come into being, and new names sometimes adapt features characteristic of other sets, those of groups of words, for instance.
19
Orthographic words are written as a sequence of letters bounded by spaces on a page. Yet, there exist in the English vocabulary lexical units that are not identical with orthographic words but equivalent to them. Almost any part of speech contains units indivisible either syntactically or in terms of meaning, or both, but graphically divided. A good example is furnished by complex prepositions: along with, as far as, in spite of, except for, due to, by means of, for the sake of, etc.
The same point may be illustrated by phrasal verbs, so numerous in English: bring up ‘to educate’, call on ‘to visit’, make up ‘to apply cosmetics’, ‘to reconcile after a disagreement’ and some other meanings, put off “to postpone’. The semantic unity of these verbs is manifest in the possibility to substitute them by orthographically single-word verbs. Though formally broken up, they function like words and they are integrated semantically so that their meaning cannot be inferred from their constituent elements. The same is true about phrasal verbs consisting of the verbs give, make, take and some others used with a noun instead of its homonymous verb alone: give a smile, make a promise, take a walk (cf. to smile, to promise, to walk).
Some further examples are furnished by compound nouns. Sometimes they are not joined by solid spelling or hyphenation but written separately, although in all other respects they do not differ from similar one-word nominations. By way of example let us take some terms for military ranks. The terms lieutenant-commander and lieutenant-colonel are hyphenated, whereas wing commander and flight lieutenant are written separately. Compare also such inconsistencies as all right and altogether, never mind and nevertheless.
All these are, if not words, then at least word equivalents because they are indivisible and fulfil the nominative, significative, communicative and pragmatic functions just as words do.
It is worth while dwelling for a moment on formulaic sentences which tend to be ready-made and are characterised by semantic unity and indivisibility: All right, Allow me, Nothing doing, Never mind, How do you do, Quite the contrary. They are learned as unanalysable wholes and can also be regarded as word equivalents.
To sum up: the vocabulary of a language is not homogeneous. If we view it as a kind of field, we shall see that its bulk, its central part is formed by lexical units possessing all the distinctive features of words, i.e. semantic, orthographic and morphological integrity as well as the capacity of being used in speech in isolation. The marginal elements of this field reveal only some of these features, and yet belong to this set too. Thus, phrasal verbs, complex prepositions, some compounds, phraseological units, formulaic expressions, etc. are divided in spelling but are in all other respects equivalent to words. Morphemes, on the other hand, a much smaller subset of the vocabulary, cannot be used as separate utterances and are less autonomous in other respects but otherwise also function as lexical items. The new term recently introduced in mathematics to describe sets with blurred boundaries seems expressive and worthy of
20
use in characterising a vocabulary — such sets are called fuzzy sets.¹