II. Nouns converted from verbs (deverbal substantives).

The verb generally referring to an action, the converted noun may denote:

1) instance of the action, e.g. jump v — jumpn — ’sudden spring from the ground’; movev — moven — ‘a change of position’;

2) agent of the action, e.g. help v — helpn — ‘a person who helps’; it is of interest to mention that the deverbal personal nouns denoting the doer are mostly derogatory, e.g. bore v — boren — ‘a person that bores’; cheatv— cheatn — ‘a person who cheats’;

3) place of the action, e.g. drivev — driven — ‘a path or road along which one drives’; walk v — walkn — ‘a place for walking’;

4) object or result of the action, e.g. peelv— peeln — ‘the outer skin of fruit or potatoes taken off; findv—findи — ’something found,” esp. something valuable or pleasant’; etc.

For convenience the typical semantic relations as briefly described above may be graphically represented in the form of a diagram (see below, pp. 132-133).

In conclusion it is necessary to point out that in the case of polysemantic words one and the same member of a conversion pair, a verb or a noun, belongs to several of the above-mentioned groups making different derivational bases. For instance, the verb dustbelongs to Group 4 of Denominal verbs (deprivation of the object) when it means ‘remove dust from something’, and to Group 3 (acquisition or addition of the object) when it means ‘cover with powder’; the noun slideis referred to Group 3 of Deverbal substantives (place of the action) when denoting ‘a stretch of smooth ice or hard snow on which people slide’ and to Group 2 (agent of the action) when it refers to a part of an instrument or machine that slides, etc.

Denominal Verbs

Deverbal Substantives

It follows from the foregoing discussion that within conversion pairs one of the two words has a more complex semantic structure, hence the problem of the criteria of semantic derivation: which of the two words within a conversion pair is the derived member?

The first criterion makes use of the non-correspondence between the lexical meaning of the root-morpheme and the part-of-speech meaning of the stem in one of the two words making up a conversion pair. In cases like penn — penv, fathern— fatherv, etc. the noun is the name for a being or a concrete thing. Therefore, the lexical meaning of the root-morpheme corresponds to the part-of-speech meaning of the stem. This type of nouns is regarded as having a simple semantic structure.

The verbs pen, fatherdenote a process, therefore the part-of-speech meaning of their stems does not correspond to the lexical meaning of the roots which is of a substantival character. This distinction accounts for the complex character of the semantic structure of verbs of this type. It is natural to regard the semantically simple as the source of the semantically complex, hence we are justified in assuming that the verbs pen, fatherare derived from the corresponding nouns. This criterion is not universal being rather restricted in its application. It is reliable only when there is no doubt that the root-morpheme is of a substantival character or that it denotes a process, i.e. in cases like to father, to pen, a fall, a drive,etc. But there are a great many conversion pairs in which it is extremely difficult to exactly determine the semantic character of the root-morpheme, e.g. answerv— answern; matchv—matchn,etc. The non-correspondence criterion is inapplicable to such cases.

The second criterion involves a comparison of a conversion pair with analogous word-pairs making use of the synonymic sets, of which the words in question are members. For instance, in comparing conversion pairs like chatv— chatn;showv— shown; workv— workn, etc. with analogous synonymic word-pairs like converse — conversation; exhibit — exhibition; occupy — occupation; employ — employment,etc. we are led to conclude that the nouns chat, show, work,etc. are the derived

members. We are justified in arriving at this conclusion because the semantic relations in the case of chatv— chatn;showv— shown; workv— workn are similar to those between converse — conversation; exhibit — exhibition; employ — employment.Like the non-correspondence criterion the synonymity criterion is considerably restricted in its application. This is a relatively reliable criterion only for abstract words whose synonyms possess a complex morphological structure making it possible to draw a definite conclusion about the direction of semantic derivation. Besides, this criterion may be applied only to deverbal substantives (v -> n) and not to denominal verbs (n -> v).

Of more universal character is the criterion based on derivational relations within the word-cluster of which the converted words in question are members. It will be recalled that the stems of words making up a word-cluster enter into derivational relations of different degrees.¹ If the centre of the cluster is a verb, all derived words of the first degree of derivation have suffixes generally added to a verb-base (see fig. below. p. 135). The centre of a cluster being a noun, all the first-degree derivatives have suffixes generally added to a noun-base.

Proceeding from this regularity it is logical to conclude that if the first-degree derivatives have suffixes added to a noun-base, the centre of the cluster is a noun, and if they have suffixes added to a verb-base, it is a verb.² It is this regularity that the criterion of semantic derivation under discussion is based on. In the word-cluster handn— handv— handful — handy — handedthe derived words have suffixes added to the noun-base which makes it possible to conclude that the structural and semantic centre of the whole cluster is the noun hand.Consequently, we can assume that the verb handis semantically derived from the noun hand.Likewise, considering the derivatives within the word-cluster floatn—floatv— floatable — floater — floatation — floatingwe see that the centre is the verb to floatand conclude that the noun floatis the derived member in the conversion pair floatn— floatv. The derivational criterion is less restricted in its application than the other two described above. However, as this criterion necessarily involves consideration of a whole set of derivatives it can hardly be applied to word-clusters which have few derived words.

Of very wide application is the criterion of semantic derivation based on semantic relations within conversion pairs. It is natural to conclude that the existence within a conversion pair of atype of relations typical of, e.g., denominal verbs proves that the verb is the derived member. Likewise, a type of relations typical of deverbal substantives marks the noun as the derived member. For instance, the semantic relations between crowdn — crowdv are perceived as those of an object and an action characteristic of the object, which leads one to the ,

¹ See ‘Word-Formations’, § 6, p. 114.

² Information concerning the stems of the parts of speech the English suffixes are regularly added to may be found in “Exercises in Modern English Word-Building” by D. Vesnik and S. Khidekel, M., 1964.

conclusion that the verb crowdis the derived member; likewise, in the pair take v — taken the noun is the derived member, because the relations between the two words are those of an action and a result or an object of the action — type 4 relations of deverbal substantives, etc. This semantic criterion of inner derivation is one of the most important ones for determining the derived members within a conversion pair, for its application has almost no limitations.

To sum up, out of the four criteria considered above the most important are the derivational and the semantic criteria, for there are almost no limitations to their application. When applying the other two criteria, their limitations should be kept in mind. As a rule, the word under analysis should meet the requirements of the two basic criteria. In doubtful cases one of the remaining criteria should be resorted to. It may be of interest to point out that in case a word meets the requirements of the non-correspondence criterion no additional checking is necessary.

Of late a new criterion of semantic derivation for conversion pairs has been suggested.¹ It is based on the frequency of occurrence in various utterances of either of the two member-words related through conversion. According to this frequency criterion a lower frequency value testifies to the derived character of the word in question. The information about the frequency value of words although on a limited scale can be found in the available dictionaries of word-frequency with semantic counts.²

To give an illustration, according to M. West’s A General Service List of English Words, the frequency value of four verb — noun conversion pairs in correlative meanings taken at random is estimated as follows:

to answer (V = 63%) — answer (N =35%), to help (V = 61%) — help (N = 1%), to sample (V= 10%) — sample (N=90%), to joke (V=8%) — joke (N=82%).

By the frequency criterion of semantic derivation in the first two pairs the nouns (answerand help)are derived words (deverbal

¹ See H. О. Волкова. К вопросу о направлении производности при конверсии в парах имя — глагол (на материале современного английского языка). — Сб., Иностр. яз. в высшей школе, вып. 9. М., 1974.

² See ‘Fundamentals of English Lexicography’, § 5, p. 214.

substantive’s), in the other two pairs the verbs (to sampleand to joke)are converted from nouns (denominal verbs).

Of interest is also the transformational criterion of semantic derivation for conversion pairs suggested in linguistic literature not so long ago.¹ The procedure of the transformational criterion israther complicated, therefore only part of it as applied to deverbal substantives is described here.

The transformational procedure helping to determine the direction of semantic derivation in conversion pairs is the transformation of nominalisation (the nominalising transformation).² It is applied to a change of apredicative syntagma into a nominal syntagma.

By analogy with the transformation of predicative syntagmas like “The committee elected John” into the nominal syntagma “John’s election by the committee” or “the committee’s election of John” in which the derivational relationship of electand electionis that of a derived word (election)to its base (elect)the possibility of transformations like

Roy loves nature -> Roy’s love of nature ³ John visited his friend -> John’s visit to his friend She promised help -> her promise of help proves the derived character of the nouns love, visit, promise.Failure to apply the nominalising transformation indicates that the nouns cannot be regarded as derived from the corresponding verb base,

e.g. She bosses the establishment -> her boss of the establishment ⁴ I skinned the rabbit -> my skin of the rabbit He taxied home -> his taxi home

Modern English vocabulary is exceedingly rich in conversion pairs. As a way of forming words conversion is extremely productive and new conversion pairs make their appearance in fiction, newspaper articles and in the process of oral communication in all spheres of human activity gradually forcing their way into the existing vocabulary and into the dictionaries as well. New conversion pairs are created on the analogy of those already in the word-stock on the semantic patterns described above as types of semantic relations. Conversion is highly productive in the formation of verbs, especially from compound nouns. 20th century new words include a great many verbs formed by conversion, e.g. to motor — ‘travel by car’; to phone — ‘use the telephone’; to wire — ’send a telegram’; to microfilm — ‘produce a microfilm of; to tear-gas — ‘to use tear-gas’; to fire-bomb — ‘drop fire-bombs’; to spearhead — ‘act as a spearhead for’; to blueprint — ‘work out, outline’, etc. A diachronic survey of the present-day stock of conversion pairs reveals, however, that not all of them have been created on the semantic patterns just referred to. Some of them arose as a result of the disappear-

¹ See П. А. Соболева. О трансформационном анализе словообразовательных отношений. — Сб. Трансформационный метод в структурной лингвистике. М., 1964.

² See ‘Methods and Procedures of Lexicological Analysis’, § 5, p. 251. ³ The sign -> shows the possibility of transformation.

⁴ The sign ->denotes the impossibility of transformation.

ance of inflections in the course of the historical development of the English language due to which two words of different parts of speech, e.g. a verb and a noun, coincided in pronunciation. This is the case with such word-pairs, for instance, as loven (OE. lufu) — lovev (OE. lufian); workn (OE. wēōrc) — workv (OE. wyrcan); answern (OE. andswaru) — answerv (OE. andswarian) and many others. For this reason certain linguists consider it necessary to distinguish between homonymous word-pairs which appeared as a result of the loss of inflections and those formed by conversion. The term conversion is applied then only to cases like doctorn— doctorv;briefa— brief vthat came into being after the disappearance of inflections, word-pairs like workn— work v being regarded exclusively as cases of homonymy.¹

Other linguists share Prof. Smirnitsky’s views concerning discrimination between conversion as a derivational means and as a type of word-building relations between words in Modern English. Synchronically in Modern English there is no difference at all between cases like taxin — taxi v and cases like loven— lovevfrom the point of view of their morphological structure and the word-building system of the language. In either case the only difference between the two words is that of the paradigm: the historical background is here irrelevant. It should be emphatically stressed at this point that the present-day derivative correlations within conversion pairs do not necessarily coincide with the etymological relationship. For instance, in the word-pair awen — awev the noun is the source, of derivation both diachronically and synchronically, but it is quite different with the pair mouldv— mouldn: historically the verb is the derived member, whereas it is the other way round from the angle of Modern English (cf. the derivatives mouldable, moulding, moulderwhich have suffixes added to verb-bases).

A diachronic semantic analysis of a conversion pair reveals that in the course of time the semantic structure of the base may acquire a new meaning or several meanings under the influence of the meanings of the converted word. This semantic process has been termed reconversion in linguistic literature.² There is an essential difference between conversion and reconversion: being a way of forming words conversion leads to a numerical enlargement of the English vocabulary, whereas reconversion only brings about a new meaning correlated with one of the meanings of the converted word. Research has shown that reconversion

¹ Because of the regular character of semantic correlation within such word-pairs as well as within conversion pairs formed on the semantic patterns I. P. Ivanova intro-

duces the notion of patterned homonymy. She points out that conversion is one of the sources of homonymy, there are also other sources such as coincidence in sound-form of words of different parts of speech, borrowing two words of different parts of speech in the same phonetic shape, and some others. (See И. П. Иванова. О морфологической характеристике слова в современном английском языке. — Сб. : Проблемы морфологического строя германских языков. М., 1963; see also I. Arnold. The English Word. M., 1973, ch. VIII.)

² See П. М. Каращук. Реконверсия и ее роль в развитии семантических структур соотносящихся по конверсии слов. — Сб. “Словообразование и его место в курсе обучения иностранному языку”, вып. I. Владивосток, 1973.

1.37

only operates with denominal verbs and deverbal nouns. As an illustration the conversion pair smoken— smokev may be cited. According to the Oxford English Dictionary some of the meanings of the two words are:

SMOKE я

1. the visible volatile product given off by burning or smouldering substances (1000)¹ c) the act of smoke coming out into a room instead of passing up the chimney (1715)

SMOKE v