FOR THE ESTABLISHING OF DERIVATIONAL RELATIONSHIP BETWEEN WORDS UNMARKED BY DERIVATIONAL MORPHEMES

1.1. In a recent article I have taken up the question of derivational relationship between derivationally unmarked words, extending it in another article to backderived words. Here I propose to conclude the argument by giving a survey of the criteria used and adding a few not yet dealt with. The criteria fall into two groups, those of content and those of external form. The criteria of the first group (2.–5.) are: semantic depen­dence, range of usage, semantic range, and semantic pattern. The most important is that of semantic dependence, as it is as often as not sufficient in itself to solve the question of derivational relationship while the other criteria have a more or less concomitant character. The criteria of the second group (6.–8.) are: phonetic shape, morphologic type, and stress. They will illustrate how external factors can help to indicate derivational relationship.

1.2. A derivational connection between words presupposes a correspondence both on the plane of expression (phonic form) and content. With derivationally unmarked words, as bridge sb and bridge vb, the first is automatically established by the lack of morphologic characterization which entails phonic identity. We thus are, in the case of homophonous or rather quasi-homophonous word pairs, only confronted with the problem of content. I say ‘quasi-homophonous’, as in many cases, e.g. conflict sb from conflict vb, we have at the same time phonological changes of vowel and/or consonant.

1.3. The term ‘derivationally unmarked’ is important. Many words are marked with regard to the class of words to which they belong, as is evident from the stress-distinguished type conflict sb from conflict vb. But this mark is not a derivational mark as is a suffix. The suffix -ize (e.g. westernize) not only tells us that the word formed with it is a verb, but at the same time that the verb is derived from the unsuffixed word western. Stress, however, in the above mentioned type, shows only categorial appurtenance without giving us immediate information as to which word is derived from the other.

Stress distinction is on the same level with distinction by absence or presence of voice in the case of words ending in a fricative, type belief from believe. The character of the fricative does not indicate whether the substantive is from the verb or the verb from the substantive. This derivational dependence is established by means of content and of semantic pattern. The voiced or voiceless character only places the respective verb in a certain word class, but it is not more than a categorial marker, not primarily connected with derivation. See below 8.1.–5.

1.4.1. It will be understood from the outset that there is no derivational connection between words if the words have no semantic features in common. This is clear for such obviously unrelated words as exact adj. and exact vb, handle sb and handle vb, exploit sb and exploit vb, defect sb and defect vb which are therefore out of consideration. There is no relationship between convict ‘prove or find guilty’ and convict ‘person serving a prison sentence’, though etymological ties do exist. The sb account in the majority of its meanings has no connection with the verb account which is chiefly used in constructions like account for, account to. The only connection is between account ‘render an account of’ and account sb, as in render an account to God, where the substantive is a deverbal derivation from the verb. The verb mind (do you mind, I do not mind) is today not connected with mind sb, nor is the verb matter (it does not matter) connected with the substantive matter. Between object sb (all senses) and object vb, project sb and project vb (the sense ‘plan’ is hardly alive with the verb) there exist no semantic ties.

1.4.2. A word may be a derivative in one sense and not in another. Both verb and substantive may follow separate trends of semantic development that are not necessarily parallelled by the other pair member. There is no connection between trust vb and trust ‘cartel’ though there is one between the verb and the substantive in various other senses. The substantive act (as in the act of a madman, or in the sense ‘simulated performance’) is a deverbal derivative (act like a madman, he is merely acting), but act as in Act of Parliament, act ‘division of a play’ is not, as it cannot be connected with any sense of the verb act. The sb care and the verb care are no longer closely related semantically. Only for the sense ‘charge, responsibility, look after’ is there a connection between verb and substantive. Sentences such as She cares for the children lead to a deverbal substantive care: the children are in her care, she takes care of the children, also to such recent compounds as caretaker and baby care. But other senses of the verb are not matched by substantives and vice versa. I don’t care, who cares? To care about a person have no substantive to go with, nor are care ‘watchful attention’ (as in with care, careful and careless), care ‘grief, anxiety’ (as in carefree, careworn) accompanied by verbal uses of care. In some cases, the verb taste derives a deverbal substantive (I have tasted the pleasures of life / I have had a taste of the pleasures of life). But taste ‘perceive by the sense of taste’ (I can neither taste or smell) is derived from the substantive taste. And taste as in the milk tastes sour is a derivative from the substantive taste ‘savour’. On the other hand, there is no verb for taste ‘sense of appropriateness’.

2. Semantic dependence (SD)

2.1. The word that for its analysis is dependent on the content of the other pair member is necessarily the derivative.

2.2. The verb saw must be derived from the stubstantive saw. Saw sb is satisfactorily defined as ‘a cutting instrument with a blade, having a continuous series of teeth on the edge’. That the instrument may be used for the action of sawing need not be included in the definition. On the other hand, the content analysis of the verb must necessarily include the semantic features of the substantive saw: saw vb ‘use a saw, cut with a saw’. The verb knife is naturally analysable as ‘wound with a knife’ whereas the substantive knife does not lean on any content features of the verb knife, which does not exist in the vocabulary of many speakers who commonly use the noun. A parallel case we have in telephone vb and telephone sb. For its analysis, the verb relies on the semantic features of the substantive telephone. Though seemingly parallel to telephone and saw the case of whistle sb (the name of the instrument) with regard to whistle vb is the reverse. The analysis of the verb does not call for any semantic features of whistle sb (the instrument). Whistling is aptly described by ‘forcing the breath through the teeth or compressed lips’ whereas the instrument whistle has for its explanation recourse to the content features of the verb:

whistle ‘instrument used for whistling’.

The criterion applies to a number of backderived verbs which are mentioned here for the sake of semantic similarity though they seem to contain a derivational marker. Television is not dependent on televise which is a much less common word, whereas televise is only explained by the content of television: ‘put on television’. Televise is the derived word.

3. Restriction of Usage (RU)

3.1. If one word has a smaller range of usage than its pair member, it must be considered the derivative. The following possibilities may arise.

3.1.1. One of the words is not generally accepted while the other is common. The substantive author is in general use while the verb is not. Author sb will therefore be considered the basis from which the verb is derived.

3.1.2. The preceding rule may be expressed in terms of frequency. One pair member is less common than the other and therefore less frequently used. The substantives father and author, for instance, are high frequency words whereas the verbs are used much less frequently. The word with the lesser degree of frequency will be regarded as the derivative. A few other relevant examples are the verbs knife, waterproof, copyright, anger.

3.1.3. If the use of one of the pair words is restricted to certain verb forms while the other occurs without any restriction, the latter is considered the basis. The verb neighbour occurs almost exclusively in ing-forms whereas the substantive is unrestricted use. The latter therefore is the basis. Amends, which is only found in the plural and only in phrase make amends, is the derivation while the verb amend is the basis. Similar considerations apply to the substantives thanks and travels. Thanks is only used in the plural, travel chiefly so (travel sg is ‘the principle of travelling, travelling in general’, but it cannot be used with a). There can be no doubt, therefore, that the verbs are the bases.

3.1.4. If one word is used only as a half serious word, it cannot be considered the basis. Though the verb burgle seems to be the basis of burglar, it is nevertheless the derivative as it is a semi-facetious word while burglar is not.

3.1.5. The verbs hunger and thirst are literary or poetic while the substantives are in colloquial use. The substantives therefore will be considered the bases.

4. Semantic Range (SR)

4.1. Of two homophonous words exhibiting similar sets of semantic features the one with the smaller field of reference is the derivative. In general terms this means that the more specific word is the derivative.

4.2. The content of the substantive butcher appears also in the verb butcher, but with restricting depreciative features, involving a particular field of reference. The substantive enters into the verb with a narrow specification of its content, butcher being emotionally motivated as ‘slaughter of animals’. The verb butcher therefore is derived from the substantive butcher. The verb cheat does not mean ‘be a cheat, i.e. an habitual cheater’, so the substantive cannot be the basis. A convert is ‘a person that has been convened to a religious or other belief’. The verb convert does not only mean ‘make into a convert’ but has many more content features. Therefore the substantive must be considered a derivative from the verb. An invert is simply a ‘sexually inverted person’ whereas the verb has a wide range of meanings though none that connects it with the substantive. We will therefore say that the substantive is derived from the verb with a special meaning. The case of pervert is almost exactly parallel. The substantive cannot be the basis as the verb has several semantic features beside that of ‘make into a pervert’. The verbal phrase run about cannot be derived from the substantive runabout as it does not mean ‘be a runabout (= roving person, light wagon, roadster)’, but is much wider in meaning. The criterion is applicable in general to postparticle combinations, as the substantives are more specific in meaning than the verbal phrases.

5. Semantic pattern (SP)

5.1. Certain words have characteristic meanings which mark them as derivatives. If, for instance, the content analysis of a substantive follows one of certain typical patterns which connect it with a homophonous verb, it must be considered a deverbal derivative. Examples are the personal substantives cheat and convert. Cheat is naturally analysed as ‘one who cheats (habitually)’, convert as ‘one who has been converted’. Substantives so analysable are deverbal derivatives. Smaller sense groups not mentioned there are ‘distance covered by –’, as in drop, fall, run; ‘the typical qualities or properties of something as determined by –’, found with the five verbs feel, look, smell, taste, touch. The cough, comparable to G der Husten (the article as in the plague, the cholera), to denote the affection in general, was, acc. to OED, common down to 1600. Today we have a similar type of deverbal substantives, used in plural form, to denote nervous fits: the gripes ‘spasms of intestinal pain’ 1601, the fidgets 1674, the creeps 1862, the jumps 1879, the jitters 1931.

5.2. On the other hand, there are certain sense groups characteristic of desubstantival verbs. The verb baby is analysed as ‘treat as a baby’, the verb father as ‘act as father’, bridge as ‘connect by means of a bridge’, herd as ‘form into a herd’. Substantives so analysable must be regarded as derivatives from the respective verbs.

6. Phonetic shape (PS)

A certain phonetic shape may put a word in a definite word class. Most words ending in [šən], as -atlon, -action, -ension, -ition, -otion, can only occur as substantives. Our inference from this fact therefore will be that verbs showing these phonetic features are derived from substantives: probation, station, ration, mention, pension, fraction, traction, acquisition, audition, condition, motion. Another characteristic substantival ending is [mənt], as in compliment, document, implement: the verbs will be considered derived from substantives in [mənt], unless other criteria exclude such an analysis. The criterion of stress excludes torment and ferment. As desubstantival verbs do not shift their stress (though trisyllabic words change their [ə] to [ε]: [mənt] sb / [mənt] vb), we cannot consider the verbs derived from the substantives, an analysis which is also excluded by analysis of content for ferment (the verb cannot be analysed as ‘treat with ferment’ or the like), though torment vb could be analysed as ‘subject to torment’. But the criterion of semantic pattern (torment sb falls under one of the headings of deverbal substantives, either ‘state of being tormented’ or ‘particular instance of tormenting’) will induce, us to regard the substantive torment as the derivative. It will be noted that only primary substantives in -ment (document, etc.) have [ə] in the final syllable whereas the vowel of the final syllable with the derived deverbal substantives is unshifted [ε]: ferment, torment. The vowel [ε] in the final syllable of the substantives torment and ferment, therefore, would be an additional external criterion for their derived character. The majority of polysyllabic words in [tšər, tšur] spelled -ture, or [ed], spelled -ade, are substantives. A verb therefore with one of these endings will be considered a derivative: capture, fracture, gesture, puncture / barricade, cascade, crusade, serenade.

7. Morphologic type (MT)

Morphologic type is indicative of the primary or derived character of composite words. The overwhelming number of substantives as against verbs with the types blacklist (adj/sb), snowball (sb/sb), sideslip (sb/dev sb) entitles us to the conclusion that the types are basically nominal ones and the verbs so structured are derived.

8. Criterion of Stress (St)

8.1. With composites, stress is sometimes indicative of derivational relationship between a substantive and a verb. Compound verbs with locative particles for a first member have the basic stress pattern middle stress/ heavy stress, as outlive, underestimate whereas compound substantives have the heavy stress on the first element, as outhouse, undercurrent. A verb therefore which conforms to this latter stress pattern is characterized as a desubstantival verb. Cases in point are such verbs as outlaw, outline, outfit, understudy ‘act as understudy’ (for which OED gives the wrong stressing ‘/’).

8.2. To a certain extent, the criterion of stress applies to verbs of French and/ or Latin origin which are monemes in English, but are etymologically analysable as ‘prefix + verb’ in Latin or French, as in conflict sb from conflict vb.

8.2.1. If the substantive is distinguished from the verb by stress, it must be considered a derivative from the verb, unless content criteria preclude such an analysis. This follows from the rule that desubstantival verbs retain the stress pattern of the underlying substantive (as in focus ‘bring into focus’ vb from focus sb). This postulate would apply to abstract ‘epitome’, compound, compress, concert ‘agreement’, conduct, confines, conflict, conscript, consort, construct, contest, contract BE (AE has forestress with both sb and vb), contrast, convert, digest, in sense ‘arrange methodically’ vb, ‘methodically arranged compen­dium’ sb, discount, escort, export, extract, ferment, import, impress, increase, insert, insult, intercept, invert, invite, perfume, permit, pervert, present ‘gift’, proceeds, produce, progress, protest, rebel, recess, regress, retail BE, survey, torment, transfer, transform, transplant, transport.

The substantives are all analysable as deverbal derivatives according to one of the sense groups typical of deverbal substantives.

8.2.2. The rule formulated above states nothing about endstressed nouns, (as nuance, cement, crusade, festoon) from which verbs are zero-derived. All denominal verbs (verbs derived from substantives or adjectives) retain the entire phonetic pattern of the respective nouns. Nor does our rule imply that substantives derived from end-stressed verbs should shift their stress. The tendency toward homologic stressing is observed with most deverbal substantives: advance, approach, attack, attempt are stressed like the corresponding verbs. Our rule does not apply either to the few adjectives which would seem to be derivationally connected with verbs: absent, present, subject cannot be regarded as derivatives from verbs. Nor can the verbs be derivatives from the adjectives. Deadjectival verbs can only follow the semantic patterns ‘make –’, ‘become –’, ‘be –’ into which the verbs absent and subject can be made to fit only with some semantic or syntactic restrictions. The verb absent occurs only in combinations with self-pronouns (absent oneself). Subject fits in only in sense ‘make liable to’ from subject adj. ‘liable to’ (as in subject to temptation). The verb present has no semantic connection with the adjective present for any speaker of present-day English. That the verbs fall outside the homologic stress and vowel pattern of deadjectival verbs, as already mentioned, that the verb absent oneself was coined on the analogy of present oneself rather than as a derivative from the adjective absent shall be added only for the sake of completeness.

8.2.3. The reason for the stress distinction was probably the awareness of educated speakers of the composite character of the verbs in French and/or Latin. In deriving substantives from the verbs, they followed the native pattern of words for which there existed both a verb and a substantive, differentiated only by stress, preparticle words of the type overthrow vb / overthrow sb, postparticle words of the type black out vb / blackout sb. A stress-distinctive pattern has developed chiefly with disyllabic words beginning with con-, trans-, pro-. It will be difficult to tell why others similarly structured, as detain, have not developed. But then we do not know either why some real prefixal verbs have developed stress-distinguished substantival derivatives (re- (a refill, etc.), mis- (a misprint, etc.), and inter-(an interchange), etc.) while others have not (dis-; un-; de-). At any rate we notice that of the many zero-derivatives from verbs of the detain type which were made in earlier stages of the language almost none have survived.

Date: 2016-04-22; view: 794