Making New Words: Morphological Derivation in English [Illustrated] 0198712367, 9780198712367

Citation preview

Making New Words

Making New Words Morphological Derivation in English R. M. W. Dixon Language and Culture Research Centre James Cook University

1

1 Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © R. M. W. Dixon 2014 The moral rights of the author have been asserted First Edition published in 2014 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2013956032 ISBN 978–0–19–871236–7 (hbk.) ISBN 978–0–19–871237–4 (pbk.) Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

For my pre-eminent multi-talent-ed ultra-percep-tive sasha-esque book-str-ess

Contents Preface  ix Time Frame, Abbreviations, and Conventions  xi List of Tables  xiv 1 Preliminaries  1 2 How to make new words  11 3 Criteria—affixes, semi-affixes, and compounds  45 4 What we describe  62 5 Un-expected, mis-judged, dis-organized counter-revolution : Making negative words  71 6 Ultra-patriotic, extra-special, pre-election pseudo-promise : Keeping the same word class  118 7 Be-feather, smart-en, em-power, beauti-fy, moral-ize : Making new verbs  178 8 Moon-like, death-ly, angr-y, mysteri-ous, memor-able : Making new adjectives  219 9 Eager-ness, bidd-ing, pay-ment, owner-ship, satisf-action : Making new nouns  296 10 Care-ful-ly, side-ways, home-wards, crab-fashion : Making new adverbs  365 11 Un-relent-less-ly de-west-ern-iz-ing non-mean-ing-ful infra-structures : Combining affixes  380

viii    contents 12 Envoi: The way of words  396 Appendix 1 List of adjective and verb semantic types, with sample members  400 Appendix 2 Alphabetical directory of adjectives and verbs in the list of semantic types  409 Glossary  428 References  435 Indexes  439 Index of derivational prefixes  440 Index of derivational suffixes  442 General index  445

Preface This book tells a story. It recounts the character of each of the two hundred or so affixes which serve to make new words in present-day English—where they come from, how they have developed, what they mean, and the manner in which they are used. We examine those affixes which go back to Old English or to some other Germanic language, those which come from a Romance language—generally Latin and/or French—and those which are of Greek origin, taken into English either directly or through a Romance conduit. Attention is paid to how and when each attained the status of a bona fide derivational affix. For every affix, there is a study of its meaning, or range of meanings. And an account of which sorts of words it may be attached to, what sort of words it creates, and the syntactic effect (if any). We note that some affixes feature in many derivations but are no longer used to make new ones, some have a degree of productivity, and others are highly productive. There is focus throughout on subtle contrasts of form and of meaning. Why do we say un-distinguished but in-distinguish-able (not *un-distinguish-able), why gold-en but silver-y (not *silver-en), why happi-ness but jealous-y (not *jealous-ness). Derivational affixes may have meanings which are similar but not quite identical. We contrast child-less and child-free, dis-place and mis-place, grac-ious and grace-ful, defend-er and defend-ant. What is being described here is the grammar of Standard British English, at the beginning of the twenty-first century. This is a dialect which has lost /r/ from the end of a syllable. Thus, for instance, inter-urban is pronounced /ˌintə-'ə:bən/ whereas in other dialects (predominantly in Scotland and the USA) it could be /ˌintər-'ə:rbən /, where the /r/ has one of a number of phonetic realizations. I have consulted very many sources, and made copious use of the best of them. Otto Jespersen’s linguistic achievements were outstanding. Part VI, Morphology (1942) of A Modern English grammar, on historical principles, included the first comprehensive account of derivational affixes (arranged by phonological form). The work was extended by Hans Marchand’s meticulous volume (second edition, 1969), The categories and types of present-day English word-formation (where the arrangement of affixes is alphabetical).

x    preface Substantial assistance was also afforded by the early account of Sweet (1891), Bauer and Huddleston’s long chapter on ‘Lexical word formation’ in the Huddleston and Pullum grammar (2002), the large grammar by Quirk, Greenbaum, Leech, and Svartvik (1985), volumes by Kruisinga (1932), Barnhart (1988), Adams (1973, 2001), Bauer (1983, 2001), Plag (1999), Bauer, Lieber, and Plag (2013), and very many more. I delved fruitfully into the treasure trove of both normal and nonce uses in H. L. Mencken’s epochal volumes The American language, an inquiry into the development of English in the United States (1936, 1945). Continuous use has been made of what is now called the Oxford English dictionary (OED) but was originally published, under the main editorship of James A. H. Murray, as A new English dictionary on historical principles, founded mainly on materials collected by the Philological Society (1888– 1933). The on-line version is an easily-accessible resource, with invaluable historical information and citations (although sometimes a little low on linguistic acumen). Reference has also been made to various smaller Oxford dictionaries, to A dictionary of the English language by Samuel Johnson (1755), to the second edition of The Random House dictionary of the English language, unabridged (Flexner 1987), to Collins Cobuild English dictionary for advanced learners (Sinclair 2001), and to a number of others. Conventional dictionaries can be a great help in the study of prefixes. When examining suffixes I have gladly utilized Gustav Muthmann’s outstanding piece of scholarship, his Reverse English dictionary, based on phonological and morphological principles (1999). I have also made judicious use of various corpus collections, including the International Computer Archive of Modern English, ICAME collection of English language corpora (Bergen: Norwegian Computing Centre for the Humanities), plus the Australian corpus of English (ACE), the LancasterOslo-Bergen (LOB) corpus and the Brown corpus of American texts. I have been circumspect in use of information from Google. As always, Alexandra Aikhenvald has been a fount of inspiration. During meals and elsewhen, over a period of several years, we have discussed matters affix by affix. And she has provided perceptive comments on the draft typescript. Laurie Bauer and Kate Burridge gave useful feedback on early drafts of several chapters. Hannah Sarvasy checked the final typescript, and made a number of most worthwhile suggestions.

Time Frame, Abbreviations, and Conventions OE stands for Old English (also known as Anglo-Saxon), spoken from the fifth to the twelfth century. This was the language of the Angles and Saxons, who took up residence in Celtic England during the latter part of the fifth century ad; the first manuscript records are from the sixth century. OE was a Germanic language, and it took in loans from another Germanic tongue, Old Norse, which was spoken by Viking invaders in the north of England. Christianity made its appearance in the seventh century, bringing with it ecclesiastical loans from Latin, such as deacon and hymn. ME stands for Middle English, the historical stage of the language which was spoken from the twelfth to the end of the fifteenth century. After the Norman conquest, in 1066, French was used for all official purposes although ME was spoken by the population at large. It is estimated that around 10,000 French words were borrowed into ME, about three-quarters of which are still in use today (Harley 2006: 257). The use of French diminished after the King of England lost his French possessions, at the beginning of the thirteenth century, and by the end of the fourteenth century, English had become the official language of government. There were also a considerable number of loans from Low Dutch during this period; for example, pickle. After 1500 we had Modern English; the period from 1500 to 1700 is sometimes distinguished as ‘early Modern English’. French was still considered a language of prestige, and it continued to supply loans, most originating in Latin. However, some learned terms were taken directly from Latin. There were also a number taken directly from Greek, in addition to Greek words which made their way into Latin, thence into French, and then English. In more modern times English took in loans from European languages such as Spanish and Italian, and from the languages of peoples whose lands had been absorbed into the British Empire. For example, pyjamas from Urdu, and boomerang from Dharuk (around Sydney in New South Wales). In many places, we quote the date of the first citation for a given word from the OED. It should be borne in mind that this is the earliest written

xii   time frame, abbreviations, and conventions attestation of that word. It is likely to have been in use for some time before this date. The historical stages of French are rather different from those of English: Old French from the ninth until about the middle of the fourteenth century, then Middle French from the middle of the fourteenth until the beginning of the seventeenth century. There was also Anglo-Norman, the variety spoken in England from the eleventh to the fourteenth century. In this book the actual stage or variety of French from which a loan came into English is only occasionally specified. In tables showing the origins of derivational affixes, the following abbreviations are used: Gmc Germanic, covering all Germanic sources, including Old English, Old Norse, other Scandinavian variates, and Dutch Rom Romance, covering all Romance languages Lat Latin Fr French Gk Greek

Other abbreviations A single letter is used for each of the three major word classes: N noun A adjective V verb The regular abbreviation NP indicates noun phrase. Core syntactic functions are accorded the standard abbreviations: S  intransitive subject   CS  copula subject A transitive subject    CC copula complement O  transitive object (The context will always make clear which sense of ‘A’ is intended.) The phonological form of a word is enclosed within slant brackets, as /'ʌpə / for upper. Primary stress is indicated by ' and secondary stress by ˌ.

time frame, abbreviations, and conventions    xiii An asterisk, *, indicates a form which is not acceptable. For instance: ‘One says in-frequent-ly rather than *un-frequent-ly’. In Chapter 11, ‘√’ indicates a root.

Conventions for hyphenation All forms recognized as derivational affixes are separated off by a hyphen in quoted examples; for example un-kind, joy-ous. It is not always the case that the orthographic form of a derived word includes all segments of the constituent forms. The following conventions are employed: (1) A ‘silent e’ at the end of a root is discarded. For example, sterile /'sterail/ plus -ize /-aiz/ is written as steril-ize /'steriˌl-aiz/. (2) If the vowel at the end of a root retains its value, then the hyphen is written after this. From mummy /'mʌmi/ and -ify /-ifai/ we get mummify /'mʌmi-fai/. (3) If the vowel at the end of a word is replaced by the initial vowel of the suffix, then the hyphen precedes the latter. For example, memory /'meməri/ and -ize /-aiz/ give memor-ize /'memər-aiz/, justify (/'dʒʌstifai/ and -ication /-ikeiʃən/ give justif-ication /ˌdʒʌstif-i'keiʃən/. Note that endings which are not recognized as derivational affixes, according to the criteria set out in Chapter 3, do not receive a hyphen. For instance, ock in hillock (not *hill-ock); see 4.1c.

List of Tables 2.1 Sample of derivational affixes  12 2.2 Parallel forms in -ity, -(i)ous, and -(i)ous-ness with dates of their first mention in the OED 32 2.3 Possibilities for double duty  36 2.4 Comparing double duty items with derivations  42 3.1 Derivational affixes and cognate free forms  55 4.1 Some analogic adaptations and the words they are modelled on  66 5.1 Alternative negated forms of lexemes in the sample text  72 5.2 Summary properties of negative prefixes  74–5 6.1 Number prefixes, with an example of each  120 6.2 Magnitude and degree prefixes  124 6.3 Forms of regular adjective quick, and irregulars bad and good 128 6.4 Location and degree prefixes  130 6.5 Comparison of some major senses of sub- and under- 131 6.6 Comparison of some major senses of super- and over- 134 6.7 Word-class effect with out-, in-, off-, on-, up-, and down- 148 6.8 Temporal prefixes  148 6.9 Sample instances of after- and post- 150 6.10 Ways of describing a deputy  163 7.1 Verbalizations based on nation names  216 8.1 Summary of properties of suffixes which create new adjectives  222–3 8.2 Occurrences of some adjective-deriving suffixes with semantic types of nouns  228–31 8.3 Contrasting the use of -ful and -less 259 8.4 Suffixes which contrast with -less 260 8.5 Type A of nation names with derived forms  266 8.6 Type B of names of nations and other places plus derived forms  269 8.7 Syntactic orientation of adjectives derived from verbs  279 8.8 Occurrence of some adjective-deriving suffixes with semantic types of verbs  280–1 9.1 Agentive nominalizations of varieties of phrasal verbs  310 9.2 Comparing -ance| -ence and -ancy| -ency nominalizations  341

list of tables    xv 1 0.1 10.2 11.1 11.2 11.3

Functions of adverbs derived from adjectival semantic types  371 Interaction of adverb and nominal derivation  373 Two derivational suffixes added to a noun base  383 Two derivational suffixes added to an adjective base  386 Two derivational suffixes added to a verb base  387

1 Preliminaries

1.1 Terminology 1 1.2 A little bit of syntax  6

1.3 Phonemic form and English orthography 8

1.1 Terminology 1.1a Spoken discourse typically consists—unless it is a monologue—of a number of utterances. John says this, Mary responds with that, Jane butts in with something, and then John makes a further comment. The utterances may occur in sequence, but they often overlap. Each utterance is made up of a string of sentences (generally determined by speakers’ pauses and intonation). Some of the sentences may be well-formed (especially in more measured discourse) while others may be part-formed (although still fully understandable to the others involved). For every one of the latter, if its speaker were given a transcript they would amend it to be well-formed. A written text—be it a book, a newspaper article, a letter, or anything else—is likely to be made up of paragraphs, and each paragraph of a number of sentences. (What is a paragraph and what is a sentence is to some extent at the whim of the writer—where they commence a new line, and where they use a full stop, or period.) A sentence is made up of one main clause and, optionally, a number of subordinate clauses. For instance: [The ex-gangster glorified his misdeeds]MAIN.CLAUSE, [when he was drunk]SUBORDINATE.CLAUSE A clause consists of a predicate and a number of arguments (more on this in 1.2). The predicate is realized by a verb phrase. This is glorified in the main

2    1 preliminaries clause just given, but it could be much longer, for example had been glorifying. Each argument is realized by a noun phrase; in the main clause here the subject argument is shown by noun phrase the ex-gangster, and the object argument by noun phrase his misdeeds. Each noun phrase has a head, which can be a pronoun or a lexical word, and may have one or more modifiers (such as grammatical forms the and his here, or adjectives, or a demonstrative). The heads of the noun phrases here are lexical words: ex-gangster and misdeeds. What we focus on in this book are lexical words and their composition. Each of these words can be analysed into a number of morphemes, with morpheme boundaries shown by hyphens. Thus: ex-gang-ster   mis-deed-s   glori-fi-ed A morpheme is the minimal unit of meaning of the language. Each of the prefixes ex- and mis-, suffixes -ster, -s, -(i)fi, and -ed, and roots gang, deed, and glory, has a meaning. (The fact that glory is a minimal bit of meaning is shown by the fact that there is no meaning attached to syllables within it—glo and ry.) 1.1b There are three types of morpheme:

• • •

Lexical roots (also called lexemes): gang, deed, and glory here. These are what get entered in the dictionary. Derivational affixes: ex-, mis-, and -fi here. Each of these is added to a root and forms a stem. Inflectional affixes: -s and -ed here. These are added to a stem and form a finished word.

A word is built up as follows: (a) Choose a root, which belongs to a particular word class. For instance, gang, or deed, or glory, which are all basically nouns. (Like many other nouns, all three also have secondary function as verbs.) (b) Optionally, apply a derivational affix, which forms a stem. This may belong to the same word class as the original root: root deed, which is here a noun add derivational prefix mis-, forming stem misdeed, which is also a noun

1.1 terminology   3 or it may belong to a different class: root glory, which is here a noun add derivational suffix -(i)fy, forming stem glori-fy, which is a verb There can be more than one derivational affix: root gang, which is here a noun add derivational suffix -ster, forming stem gang-ster, which is also a noun add derivational prefix ex-, forming stem ex-gang-ster, which is also a noun (c) Obligatorily, apply an inflectional affix which is appropriate to the word class of the stem, thus creating a word. Thus gang-ster and mis-deed, which are count nouns, must inflect for number: either plural, shown by -s, as in mis-deed-s, or singular, shown by zero suffix, as in ex-gang-ster. Mis-deed-s and ex-gang-ster are now finished nouns, which can each function as the head of a noun phrase. And glori-fy, which is a verb, must take a verbal inflection, here past tense -ed (the alternatives would be -ing, -s, or zero). Glori-fi-ed is now a finished verb, which can occur as the head of a verb phrase. The term ‘base’ is sometimes used for the form (root or stem) to which a derivational affix is added.

1.1c To be recognized as a morpheme, a form must either (1) occur as a free form, making up a complete word, or (2) occur, with the same meaning, in more than one word. For example, glory can occur as a free form, and also in glori-fy, glori-ous. Suffix -(i)fy recurs in scores of words, such as beauti-fy, pur-ify, solid-ify. Consider mollify and deify. Moll and de(i) do not occur outside these words and so cannot be recognized as morphemes. That is, mollify and deify have the status of unanalysable roots. They do end in ify but it is not in this instance a derivational suffix since there is no lexical root for it to be applied to. What happened is that English borrowed from French glory and glori-fy, beauty and beaut-ify, pure and pur-ify—enabling the recognition of -(i)fy as a morpheme here. It also borrowed mollify and deify but not the underlying roots. (See the further discussion in 2.5h–k.)

4    1 preliminaries Some linguists consider forms like moll and de(i)- to be morphemes of a special kind, often called ‘cran’ morphemes after the first part of cranberry, or ‘unique morphemes’ (Hockett 1958: 127–8). Under such an analysis there would be a profusion of ‘cran’ morphemes in English: fash in fashion, tremend in tremendous, and hundreds more. In fact, such elements do not have any independent significance. Meaning attaches to the whole word—mollify, deify, fashion, tremendous—not to just one part of it. If one allowed ‘cran’ forms to have the status of morphemes, this would weaken the value and explanatory power of morphological analysis. Cran /kran/ was discussed by Gleason (1955: 76–7) on the basis of a passing mention in Bloomfield (1933: 234–5). Ironically, cran is not in fact an example of a ‘cran’-­ morpheme. Krueger (1963) points out that ‘cranberries grow in swamps, low-lying bogs and marshes’ and ‘this is a favorite habitat of the long-legged crane’. From crane /krein/ plus berry /'beri/ came cranberry /'kranbəri/, joining other words which reduce diphthong /ei/ in a monosyllable to vowel /a/ when the word is lengthened; these include sane| sanity, chaste| chastity, shade| shadow, grain| granary. A better example of a ‘cran morpheme’ would be rasp as in raspberry. In the sixteenth century this fruit was known as raspis or raspis berry or rasp or raspberry. The name then became conventionalized as raspberry, with rasp not occurring outside this word. (Recently, a plural form rasps has re-emerged as a colloquial shortening of raspberries.)

1.1d There may be no derivational affixes at all, in which case the root makes up the stem all on its own. Or there can be one or two or many derivations involved, as in the classic long word anti-dis-establish-ment-arian-ism. Sometimes derivations must occur in a fixed order. One goes from gang to gang-ster to ex-gang-ster, not from gang to *ex-gang to ex-gang-ster, because there is no word *ex-gang. But at other times the order does not matter. Consider un-tru-th-ful. Each of three orders of application for derivational affixes is equally good:

root: true add un-: un-true add -th: un-tru-th add -ful: un-tru-th-ful

root: true add -th: tru-th add un-: un-tru-th add -ful: un-tru-th-ful

root: true add -th: tru-th add -ful: tru-th-ful add un-: un-tru-th-ful

The derivational suffixes must be added in the order in which they occur (that nearest the root first, and so on), but—for un-tru-th-ful—the un- prefix could be added at any stage, since all intermediate words are acceptable. Contrast this with un-faith-ful which must be derived faith → faith-ful → unfaith-ful, rather than faith → *un-faith → un-faith-ful, simply because *unfaith is not a word in English.

1.1 terminology   5 1.1e Some languages have multiple inflectional systems. In Latin, for instance, each noun stem must inflect for number and case, each adjective stem for gender, number, and case, and each verb stem for voice, mood, and tenseaspect, plus person and number of the subject argument. English has rather meagre inflections. They are:

•

For a count noun, a number system with two choices: singular, shown by zero ending, or plural, shown by orthographic -s. (A mass noun— referring to something which would not normally be counted, such as mud or rice—takes no inflection.) A possessive noun phrase, within a larger noun phrase, is marked by the genitive inflection, orthographic -s, suffixed to its final word. For instance: [[The King of Scotland]NP-’s hat]NP blew off. For an adjective, either plain form, with no ending, or comparative with -er, or superlative with -est. Alternatives to -er and -est are pre-head forms more and most. (See Dixon 2005b for details on the phonological conditioning for these alternatives.) For a regular verb, the choice is between past suffix which is -ed, or present, which is -s for a 3rd person singular subject (he| she laughs) and zero otherwise (I|  you|  we|  they laugh). There is also suffix -ing, which can be used with the auxiliary verb be (they are laughing) and has a variety of other grammatical uses. Irregular verbs also use -en in passive formation: compare regular verb had kick-ed with irregular had tak-en.

•

• •

In contrast to the paucity of inflections (all of which are suffixes), English has a profusion of derivational affixes—many suffixes and also many prefixes. We shall attempt in the pages which follow to discuss these in turn, for each one describing its origin, development, present-day meanings, use, and productivity. 1.1f The description just given of how a word is built up only mentioned one morphological process—affixation. This is indeed the main means of morphological derivation in English. But there are other ways of operating on a root or stem. For example, irregular (or ‘strong’) verbs in English may form past tense just by changing the vowel: present form take /teik/ and past took /tuk/. Or present sing /siŋ/, past sang /saŋ/, and passive form sung /sʌŋ/. And shifting the position of stress may, in a few instances, convert a noun or adjective into a verb (or vice versa), as for noun import /'impɔ:t/, and verb import /im'pɔ:t/, adjective frequent /'fri:kwənt/ and verb frequent /fri'kwent/.

6    1 preliminaries Morphological processes in other languages include changing tone, internal change involving vowels and consonants, subtraction, and reduplication. (See Sapir 1921: 51–81, Dixon 2010a: 138–49). 1.1g This book attempts a synchronic study of derivational affixes in the standard English of today. We examine the historical origin of each prefix and suffix, to see how this explains their patterns of present-day usage. But we do not take account of historical connections between roots which can no longer be regarded as derivations in the modern language. For instance, noun deed goes back to dæ´d or déd in Old English and was undoubtedly related to verb dón, which has become today’s do. But deed and do must be regarded as distinct lexemes in Modern English, and accorded separate dictionary entries. They have different meanings—for instance, the first thing I ‘do’ in the morning is shave, but I would not describe this as a ‘deed’. This is in contrast with, for instance, noun grow-th, whose meaning is fully inferable from the meanings of verb grow and nominalizing suffix -th. 1.1h There is one way of forming new words which is the opposite to that described above. Old English (OE) had an adjective grǣdig which developed into our modern greedy. By analogy with noun speed and adjective speedy, noun guilt and adjective guilt-y, and the like, the noun greed was created (around 1600). This is called back-formation. In essence, greedy was analysed as greed-y; that is, as being an adjective derived from noun greed by adding -y (see 8.2.8). Another instance commences with noun editor, borrowed from French in the middle of the seventeenth century. By analogy with verb conquer and noun conquer-or, verb investigate and noun investigat-or, together with other such pairs, the verb edit was back-formed from editor (now analysed as edit-or) at the end of the eighteenth century.

1.2 A little bit of syntax 1.2a There are two major clause types, according to the arguments which their predicates require:

•

a transitive clause has a transitive predicate which takes two core arguments:

1.2 a little bit of syntax   7 transitive subject, abbreviated as ‘A’ transitive object, abbreviated as ‘O’

•

an intransitive clause has an intransitive predicate which takes one core argument: intransitive subject, abbreviated as ‘S’

It is useful to show the function of an NP in its clause through an appropriate subscript. For example: [The ex-gangster]A [glorified]TRANSITIVE.PREDICATE [his misdeeds]O [The dog]S [barked]INTRANSITIVE.PREDICATE A, O, and S are core arguments, which must be stated, or else understood from the context. There may also be optional NPs, indicating time, place, beneficiary, etc. For instance, The dog barked for its master in the tavern late in the evening. 1.2b Some verbs are strictly transitive, which means that they only occur in transitive clauses, requiring both A and O core arguments. For instance like, promote, recognize, inform. Others are strictly intransitive, occurring in just intransitive clauses which have a single core argument in S function. These include arrive, chat, matter. In addition to these, there are quite a number of verbs which may be used in both intransitive and transitive clauses. It is important to distinguish two varieties of such ambitransitives:

•

S = O ambitransitives, where the S argument of the intransitive use corresponds to the O argument of the transitive use. For example, [The bomb]S exploded and [The expert]A exploded [the bomb]O. S = A ambitransitives, where the S argument of the intransitive use corresponds to the A argument of the transitive use. For example [The dog]S is drinking and [The dog]A is drinking [water]O.

•

1.2c There is also a minor clause type, copula clauses. Here the predicate consists just of a copula verb—typically be or become—and there are two core arguments: copula subject (CS) and copula complement (CC). For example: [My son]CS [is]COPULA.PREDICATE [a doctor]CC

8    1 preliminaries Whereas the predicate of a transitive or intransitive clause has referential meaning, the predicate of a copula clause marks a relation between CS and CC arguments. When the CC is a doctor there is a relation of i­dentity. Alternatively, the CC could be an adjective and the relation would be ­attribution—as in [My son]CS is [tall ]CC— or it could be a phrase indicating place—as in [My son]CS is [in the garden]CC—and the relation would be location. And so on. Sometimes the term ‘predicate’ is used for a combination of copula predicate and copula complement; for example, is a doctor would be termed the ‘nominal predicate’ of the clause. This is unhelpful. The CC is an argument, just like a CS, A, O, or S constituent.

1.3 Phonemic form and English orthography Parallel to the morpheme as the ‘minimal unit of meaning’, there is the phoneme as the ‘basic unit of sound distinction’. If one phoneme of a word is replaced by another, a new word is created. For example din and tin are different words, showing that d and t are contrastive phonemes in English. An ideal alphabet will have one letter for each phoneme. The orthography used for English is far from ideal. In many ways, the Roman alphabet, which is used to write English, is inappropriate for the language as it is spoken today. To properly describe Modern English, it is essential that relevant phonemic forms be quoted, rather than just their orthographic representations. Phonemic representations are enclosed within slant brackets, for example /plau/ for plough. Throughout this volume I have used the Everyman’s English pronouncing dictionary of Daniel Jones (DJ), generally preferring the editions wholly prepared by Daniel Jones (e.g. Jones 1956), but sometimes—especially concerning the nature of stress—using later editions, revised by A. C. Gimson (e.g. Jones 1977). The aim here is a phonemic representation, not a narrow phonetic one. A separate letter is used for each phoneme. In the case of some vowel phonemes, I employ easierto-type symbols: /a/ for a short low vowel, where DJ has /æ/; and /a:/ for a long low vowel, where DJ has /ɑ:/. Similarly for diphthongs; I use /eə/ where DJ has /ɛə/.

1.3a English orthography uses five vowel symbols, reflecting the five-vowel system in Latin. But every dialect of English has a larger number. I follow Daniel Jones in recognizing the following system of vowel phonemes for Standard British English:

1.3 phonemic form and english orthography   9

Short vowels /i/ as in sit /e/ as in get /a/ as in sat /ɔ/ as in hot /u/ as in put /ʌ/ as in up /ə/ as at the end of China

Long vowels /i:/ as in seat /a:/ /ɔ:/ /u:/

as in father as in saw as in too

/ə:/

as in burn

/ai/ /au/ /iə/ /uə/

as in fly as in how as in here as in poor

and the following diphthongs: /ei/ /ou/ /ɔi/ /eə/

as in day as in go as in boy as in there

1.3b The consonant symbols used to write English are also inadequate (although not quite so much so as the vowels). Letter g sometimes represents /g/ (as in girl, /gə:l/), other times /dʒ/ (as in gin, /dʒin/). Letter s is used for /s/ (as in basis, /'beisis/), and for /z/, (as in design /di'zain/). and also for /ʒ/ (as in decision, /di'siʒən/). Diagraph th is used both for voiced /ð/ (as in this, /ðis/) and for voiceless /θ/ (as in thin, /θin/). And so on. Some of the consonants of standard orthography do represent phonemes: p, b, t, d, k, m, n, l, r, f, v, s, z, h, w. The following phonemic symbols are used for the remainder: /g/ is always a dorso-velar voiced stop, as in gun, /gʌn/ /ð/, as in they, /ðei/ /θ/, as in thick, /θik/ /ʃ/, as in ship, /ʃip/ /tʃ/, as in chip, /tʃip/ /ʒ/, as in leisure, /'leʒə/ /dʒ/, as in jam, /dʒam/ , /ŋ/, as in sing, /siŋ/ /y/ (rather than DJ s /j/) as in yet, /yet/ 1.3c Lateral /l/ and nasal /n/ may function (rather like a vowel) as the nucleus of a syllable. This is shown by a short vertical line below the letter, as in bottle /'bɔtl̩/ and cotton /'kɔtn̩/. 1.3d Although not shown in the orthography, stress placement is an important feature of English phonology. Derivational processes may involve shift of stress.

10    1 preliminaries Each word has a primary stress, and the syllable on which it occurs must be marked if the word is of more than a single syllable. It is shown by writing superscript ' just before the first segment of the syllable—thus lateral /'latərəl/. There may also be a secondary stress, shown by subscript ˌ, as in hotdog /ˌhɔt'dɔg/. A few derivational affixes bear primary stress; for example, prefix c­ ounterand suffix -ese, while a number take secondary stress, including mis- and -esque. Others—such as -ous and -ian—affect the placement of primary stress within the stem to which they are attached. Most derivational affixes do not themselves take stress, nor do they affect stress in the base. 1.3e Many accounts of English grammar operate almost entirely in terms of the established orthography. However, the spoken register is the major mode for any living language and phonemic form should be described in the discussion of individual affixes, as I do in Chapters 5–10. The inadequacy of orthographic representation is shown by the fact that a single letter of the alphabet may represent different sounds. Consonantal instances were mentioned in 1.3b. In the examples in 1.3a, vowel symbol u is used for /u/ in put, and for /ʌ/ in up, and for /ə:/ in burn. Two words with quite different pronunciations and meanings are both written tear—verb /teə(r)/ ‘pull apart’ and noun /tiə(r)/ ‘drops of liquid that come from the eye when one weeps’. There are scores of similar examples. And, contrarywise, a given sound may have varying orthographic representation. To quote one of many instances, in British English /eə(r)/ is variously represented by ear, are, and ai (as in wear, stare, and stair). One semi-regular alternation is that between orthographic y at the end of a word and i in the middle, for /i/. Compare, for example: glory /'glɔri/, ­glori-fy /'glɔrifai/ and glori-fi-ed /'glɔrifaid/. I follow Daniel Jones in writing ‘(r)’ to indicate that—in Standard British English— the /r/ is pronounced when followed by a word beginning with a vowel (under certain grammatical conditions). Some dialects pronounce this /r/ in all circumstances.

2 How to make new words

2.1 What this book is about  11

2.6 Productivity 27

2.2 Things to explain  13

2.7 Prefixes and suffixes  33

2.3 Meanings 14

2.8 Double duty  35

2.4 Origins 18

2.9 Semantic types and semantic

2.5 Form and fusion  20

roles 43

2.1 What this book is about 2.1a English has three large open word classes (sometimes called ‘parts of speech’)—noun, verb, adjective—and one smaller one—adverb. And it has about 200 derivational affixes (90 or so prefixes and around 110 suffixes). Each of these produces new words. For example:

• • • • •

Adding -ness to an adjective may create an abstract noun; for example, sad-ness. Adding -ize to a noun may create a verb; for example, victim-ize. Adding -ive to a verb may create an adjective; for example, attract-ive. Adding -ly to an adjective may create an adverb; for example, clever-ly. Adding inter- to a verb, adjective, or noun may produce a new word of the same class; for example, verb inter-connect, adjective inter-national, and noun inter-action.

Table 2.1 illustrates derivational affixes in English, both those that change word class and those which make a new word—with different meaning—­ belonging to the same class. (Note that new words are not created from adverb roots.)

12    2 how to make new words Table 2.1  Sample of derivational affixes AFFIX APPLIES TO WORD CLASS

AFFIX FORMS A STEM OF WORD CLASS NOUN

VERB

ADJECTIVE

ADVERB

NOUN

mother-hood orphan-age art-ist

hospital-ize person-ify en-slave

father-ly fool-ish greed-y

sky-wards clock-wise edge-ways

VERB

happen-ing depend-ence dismiss-al

counter-attack out-bid re-write

harm-less use-ful accept-able

—

ADJECTIVE

honest-y eager-ness similar-ity

light-en solid-ify popular-ize

ultra-light post-modern old-ish

deep-ly five-fold

2.1b There is another way of making new words—by compounding; that is, simply combining two roots. The roots can be of the same class—nouns hen and house form the compound noun hen-house. Or they can belong to different word classes—colour (noun) plus blind (adjective) creates a compound adjective colour-blind, and hover (verb) plus craft (noun) gives the noun hovercraft. The meanings of the compounds just mentioned can more-or-less be inferred from the meanings of their parts: ‘house for hens’, ‘blind with regard to (some) colours’, ‘craft which hovers’. Other compounds have more idiosyncratic meanings. For example the verb cold-shoulder (from adjective cold and noun shoulder) means ‘act in an unfriendly way towards, ignore’, and adjective heavy-handed is ‘be overbearing (e.g. speak over-forcefully to), do something to too great a degree (e.g. put too much cinnamon on the trifle)’. Each compound is a lexical word, with an individual meaning, and requires its own dictionary entry. Compounds are essentially ad hoc. Only a limited number of principles can be recognized for the formation of compounds. In contrast, derivational affixes are in most cases productive, with regular meanings. Once a speaker of English is familiar with adjective sad and suffix -ness, they know what sad-ness means. There is no need for sad-ness to be entered in the dictionary. Some (but by no means all) dictionaries and grammars follow the OED in recognizing a set of ‘combining forms’. The OED used the label from the 1880s but did not provide criteria for recognition. Indeed, the only characterization it quotes is from Bloch and Trager (1942: 66): ‘In Latin and other languages, many words have a special combining form which appears only in compounds (or only in compounds and derivatives). . . . The foreign-learned part of the English vocabulary also shows a number of special

2.2 things to explain   13 combining forms; cf. electro-, combining form of electric, in such compounds as electro-­magnetic.’ Despite this, of the 2,234 combining forms given in the OED, quite a number are not of foreign origin (including half-, back-, -legger and -gate). ­Matthews (2007: 63–4) states: ‘Combining form: A form of a word, or a form related to or in meaning like a word, used only as an element in compounds: e.g. Anglo- in AngloAmerican or socio- in socio-economic . . .’ What is identified as a combining form varies immensely between sources; some are treated in this volume as affixes, some as analogic adaptations (see 4.2), others as parts of compounds. For the purposes of the present study, I have not found it necessary to recognize a category of ‘combining forms’.

2.1c The study of compounds falls outside the scope of the present book (although Chapter 3 does discuss criteria for distinguishing between a rootplus-root compound and a root-plus-affix derivation). Here we deal with prefixes and suffixes which create new words. Each of these derivational affixes is discussed in some detail. Using information provided here, the reader should be able to employ this whole array of prefixes and suffixes in making new words.

2.2 Things to explain English has a rich array of derivational affixes. Sometimes, there can be two affixes with similar meaning but different form, and one wonders why one is used in certain words, and the other in another set of words. 2.2a Negative prefixes un- and in- have similar meaning. Why is un- used in a certain word but in- with another that involves the same stem? For instance, why do we say in-determin-ate but un-determin-ed, and why in-cess-ant but un-ceas-ing? The brief reason is that in- tends to be used with words which include a derivational suffix that comes from a Romance language (here, -ate and -ant), while un- tends to be used when there is no such suffix. There is fuller discussion in 5.4−5. 2.2b Why can suffix -en be added to wood, forming adjective wood-en, but -en may not be added to metal? This -en suffix is used only with roots that are closed monosyllables (monosyllables ending with a consonant). Thus we get wood-en, lead-en, and g­ old-en but not *metal-en (metall-ic is used instead) or *silver-en, or ­*timber-en, See 8.2.10 and 8.2.17.

14    2 how to make new words 2.2c Why do we say heaven-ly and hell-ish, rather than *heaven-ish and *hell-ly? For a combination of reasons. Suffix -ly typically indicates a positive quality associated with the noun to which it is attached, an -ish a negative one. Also, most examples of -ish are with monosyllables, and -ly cannot be added to a root ending in -l. See 8.2.2, 8.2.4, and 8.2.6. 2.2d Prefaces mono-, uni-, and one- have essentially the same meaning. Why does one say mono-syllabic, uni-lateral, and one-sided (but not *mono-lateral, *mono-sided, *uni-syllabic, *uni-sided, *one-syllabic or *one-lateral )? The reason lies in the historical origin of prefixes and of roots. Syllabic goes back to Greek, and takes Greek prefix mono-, while lateral is from Latin and takes Romance prefix uni-, and Germanic prefix (or ‘semi-prefix’, see 3.3) one- is used with Germanic root side. See 6.2. 2.2e Why does one say de-nazi-fy and de-stalin-ize, but not *de-nazi-ze or ­*de-stalin-ify? Also, why is it the custom to say central-ize, with -ize, but ­glori-fy, with -(i)fy? The verb to describe making something German in character or language is German-ize, but with respect to French it is ­French-ify. Why? The major verbalizing suffixes in English are -(i)fy and -ize, of Romance and Greek origin respectively. Their use is phonologically conditioned. We find -ize with (inter alia) disyllabic nouns and adjectives ending in -n (for example, Stalin, German), and in -l (central). Suffix -(i)fy is used with disyllabic forms ending in a vowel (Nazi, glory) and with monosyllables (French). This is just a part of the picture; a full account is in Chapter 7. There are many further conundrums, which will be aired and resolved in the chapters that follow. 2.3 Meanings Which derivational affix may be attached to which root or stem depends on a number of factors, the main ones being: (1) The meanings of affix and of root or stem. (2) The historical origin of each—whether a Germanic form, inherited from Old English, or a form borrowed from a Romance language

2.3 meanings   15 (Latin and/or French), or from Greek, or both. (There are just a few affixes borrowed from another source; for example, -nik from Yiddish.) (3) The phonological form of the root or stem. (4) Conventions of usage which have grown up during the centuries of evolution of Modern English. (See for example, the discussion of negative prefixes un and in- in 5.4−5.) Each derivational affix has a meaning. It can carry a single sense, or there may be a fair semantic range. We can illustrate with a sample of derivational affixes which show variable swathes of meaning. 2.3a The prefix uni- always means ‘just one of’, as in uni-cellular ‘consisting of a single cell’, uni-dimensional ‘having just one dimension’, uni-directional ‘in just one direction’, and uni-cycle ‘cycle with only one wheel’. (See 6.2.) 2.3b Nominalizing suffix -er has three main meanings (see 9.3.1): (1) Describing the agent of an action described by the underlying root, as writ-er ‘someone who writes’, bak-er ‘someone who bakes’. (2) For an implement used in the action, as strain-er ‘a dish with perforations used to strain solid matter from a liquid’, and mow-er ‘machine with sharp blades used to mow a lawn’. (3) For a place where an activity (relating to the underlying verb) takes place. A din-er is ‘a place where one dines’ and a lock-er is ‘a small cupboard in which things can be locked, for safekeeping’. 2.3c Suffix -ment forms noun stems from verbs. These have a variety of kinds of reference (see 9.4.1−5, 9.4.8). (1) An activity or unit of activity: an arrange-ment is ‘things arranged in a particular way’, entertain-ment is ‘activities aimed at entertaining’, and commence-ment is ‘the act of commencing some undertaking’. (2) An attribute: bewilder-ment is ‘the state of mind of someone who is bewildered’, and enjoy-ment is ‘the pleasure a person experiences when enjoying something’. (3) The result of an activity: a judge-ment is the result of someone judging a matter; when someone is imprisoned they serve a term of imprison-ment.

16    2 how to make new words (4) The referent of the syntactic object of a verb: measure-ment is ‘the dimensions of an object obtained by measuring it’. (It can also be ‘the activity of measuring’.) A pay-ment is ‘that which is paid (usually, but not always, a sum of money) for certain goods or services’. (It can also be ‘the activity of paying’.) (5) A place where some activity happens. A settle-ment is ‘a place where people have settled, and built homes’, and an encamp-ment is ‘a location where people—often a group of soldiers, or gypsies—have set up a (generally, temporary) camp’. (A settle-ment can also be ‘an agreement to settle a disagreement’.) 2.3d Suffix -(e)ry is generally added to a noun and creates another noun with a different meaning. The meanings include (a fuller account is in 9.5.9): (1) Describing an attribute typically shown by the referent of the root noun. For example, devil-ry ‘wicked behaviour, such as one would expect from a devil’, and knave-ry ‘dishonest and crafty behaviour, characteristic of a knave’. (2) Describing the occupation of a person described by the root noon— dentist-ry is ‘the profession of being a dentist’ and cook-ery is ‘the practice of being a cook’. (3) For a collection of people or of things referred to by the root noun. These include yeoman-ry ‘a body of yeomen’, machin-ery ‘a collection of machines’. Jewell-ery is essentially ‘a collection of ornaments made by a jeweller, from jewels or imitation jewels’ and a piece of jewell-ery is used for a single necklace, bracelet, or pair of earrings, etc. (4) A place associated with the root noun. A pigg-ery is ‘an enclosure in which pigs are kept’. Bak-ery is ‘a place where baking takes place, a baker’s establishment’, and brew-ery ‘a place where brewing takes place, a brewer’s establishment’. These could be taken as related to verbs bake and brew, and/or to agentive nouns bak-er and brew-er. (5) The result of the activity described by the root (which is here a verb): a forg-ery is ‘a document which has been forged’ and robb-ery ‘the activity of robbing’. 2.3e For each affix, it is necessary to investigate which types of words it may be attached to, and what kind of words it creates.

2.3 meanings   17

• •

Prefix uni- may be added to a noun (uni-cycle) or, more often, to an ­adjective (uni-dimensional); it does not change word class. Suffix -er always derives a noun. It is generally added to a verb (for example, writ-er) but just occasionally to an adjective (northern-er ‘someone who comes from the northern parts of a land’) or to a noun (hatt-er ‘someone who makes or sells hats’). Suffix -ment also derives a noun. It is generally added to a verb (such as enjoy-ment). We also find odd-ment(s) ‘a collection of odd and unimportant objects’), from adjective odd, and better-ment ‘the act of making something better’, which is based on the comparative form of adjective good. Looking just at the modern language, it might appear that merriment ‘the activity of being merry’ is based on adjective merry; in fact it probably comes from the cognate verb to merry (now fallen into disuse). Suffix -(e)ry is almost always added to a noun and derives a noun with a different meaning. There are a few instances of its being added to a verb, including forg-ery and robb-ery, just mentioned, and to an adjective, such as brav-ery and gallant-ry.

•

•

2.3f It is important to distinguish between one derivational affix which has a range of related meanings and function, and two homonymous affixes which have the same form but quite different meanings and functions. For example:

•

There are two distinct suffixes with the form -en (both are of Germanic origin). One is added to monosyllabic nouns which describe some material, and derives an adjective ‘made of —, resembling —’; for example gold-en, wooll-en, ash-en. The other is added to adjectives and derives verbs; for instance, short-en, black-en. (Its possibilities of occurrence relate to the final segment of the adjective, and the semantic type it belongs to. See 2.6e, 7.3.2, and 8.2.10.). There are two negative suffixes un-. The first, which we can call un-1, was very common in OE and means ‘lacking a (valued) property’; it is used with adjectives, for example un-clean. The second prefix, un-2, goes back to OE and- ~ und-. It carries the meaning ‘reverse process or state’ with some transitive verbs (and adjectives derived from them), such as un-tie. (There is fuller discussion in 5.2 and 5.14−15.)

•

In the development of a language, what were originally distinct affixes— from different genetic sources—can merge. The present-day prefix mis- is a

18    2 how to make new words blending of OE mis-, as in mis-lead and mis-time, and French mes-, which has given rise to mis-adventure and mis-chance. In similar fashion, suffix un-1, from OE, and in- ~ im- ~ il- ~ ir-, from Latin and Old French, have almost identical meanings. Today, some words take one of these—un-certain rather than *in-certain—while another set of words take the other—in-curable rather than *un-curable. And some roots occur with both prefixes, showing a difference of meaning. For instance, in-evitable ‘bound to happen, nothing could stop it happening’, and unavoidable ‘will happen in these circumstances, but if some other course of action had earlier been followed, it might not have had to happen’. See the discussion in 5.4−5.

2.4 Origins Affixes come from two sources—local and foreign. A local affix is of Germanic origin. It was there, as an affix, in OE, and has percolated through to the modern language. 2.4a In OE, many adjectives formed abstract nouns with -ness, having the meaning ‘state of’. Alongside (casting OE words into modern spelling) bright, idle, thick, and drunken we had bright-ness, idle-ness, thick-ness, and drunken-ness, among many others. Further derivations were added in ME and later times, including sad-ness, happi-ness, and fit-ness. From the fourteenth century, -ness began to be added to adjectives borrowed from French, such as gentle-ness, eager-ness, and rude-ness. However, the majority of -ness derivations today are still with Germanic forms. (Many adjectives of Romance origin retain suffix -(i)ty, which is of Romance origin and has a similar meaning to -ness; for example, stupid-ity, modern-ity, similar-ity.) See 9.2.5−7. 2.4b Suffix -ish ‘having the characteristic of’ derives adjectives from common nouns and from proper names relating to nations and ethnic groups. OE included heathen-ish and churl-ish (OE form ceorl-isc from ceorl ‘man of low degree’; the root has almost dropped out of use but its -ish derivation remains strong). The suffix was added to further Germanic roots in ME and later times, including woman-ish, elf-ish (or elv-ish), fiend-ish, and hell-ish. It was also used with roots taken over from French. Fool was a borrowing

2.4 origins   19 from Old French into ME in the thirteenth century, and it soon acquired Germanic suffix -ish. Later came devil-ish, styl-ish, and others. From early times, -ish has been used with proper names, OE had Wahl ‘Wales’ and derived adjective We-lisc, which is today’s Welsh. Other early forms were Scott-ish, Brit-ish, and Dan-ish; to these were added—as the centuries advanced—Swed-ish, Jew-ish, Finn-ish, and many more. See 8.2.4, 8.3.2, and also 8.4.17. 2.4c A quite different story appertains for derivational affixes of foreign origin. These are not borrowed as affixes. What happens is that a number of roots are borrowed in bare form, and also with a particular affix attached. The latter form is then analysed within English, and the affix is identified within the borrowing language. 2.4d Prefix re- ‘do again’, which is predominantly added to a transitive verb, came down from Latin into French. ME borrowed words like relieve, rebel, and rehearse, for which no bare root was taken over. But there were also pairs such as charge and re-charge, cover and re-cover, figure and re-figure. It was on the basis of the latter that prefix re- was recognized in English, and came to be applied to Germanic as well as Romance roots—re-live, re-build, re-set, and dozens more. This prefix is now fully productive. See 6.9. 2.4e The semantic effects of nominalizing suffix -ment were summarized in 2.3c. Roots without and with -ment were borrowed from Old French into ME, including achieve and achieve-ment, judge and judg-ment, commence and commence-ment. By the fourteenth century, -ment was functioning as a suffix in English, being added to native roots as in amaze-ment, settle-ment, and merri-ment. This suffix’s application has burgeoned over the years. See 9.4.8. 2.4f Adjective-deriving suffix -ous (with variant forms -ious and -eous) ‘characterized by’ has its basis in Latin and came into ME from Old French through forms such as danger-ous, fam-ous, and riot-ous alongside plain root forms danger, fame, and riot. It was used a great deal from the fourteenth century, mainly added to further roots of Romance origin (for example poison-­ous). Suffix -ous has also been applied to some Germanic roots, including thunder-ous, wondr-ous, and murder-ous. Although there are scores of -ous derivations in general use, the suffix is not being utilized widely today to create new ones. See 8.2.14.

20    2 how to make new words 2.4g Most language development takes place surreptitiously. People just use language, while going about their daily lives, without thinking too much about what they are doing linguistically. This is how -ness and -ish, re- and -ment and -ous naturally became attached to new roots. But sometimes a different scenario pertains—a new affix is consciously brought into play, and placed with relish upon root after root. Mega- (relating to a Greek form meaning ‘great’) is an example of such an affix. It did not enter English though the normal channel of borrowing a number of affix-plus root/bare root pairs. Instead, words commencing with mega- were consciously created, in two rather different surges. Commencing in the 1870s (and continuing today), scientists add megato established nouns to describe something unusually big; for example meg-­allantoid ‘an animal with a large allantois (foetal membrane)’. Megais sometimes employed specifically for ‘a million’ (alongside kilo-, also of Greek origin, for ‘a thousand’), as in mega-dyne ‘a million dynes (units of force)’, and mega-byte ‘a million bytes (units of computer storage)’. Then mega- underwent a second birth, coming into general parlance to indicate something of stupendous size (and, therefore, importance). From mega-city in 1967, there was an outpouring of mega-’s in the 1970s and 1980s (after which the craze may have slackened a little). One heard of mega-bank, mega-deal, mega-event, mega-mall, even mega-resort. And the prefix indicates an attenuated characteristic in mega-bitch (from 1985), glossed by the OED as ‘an extremely malicious or treacherous woman’. 2.4h English is a Germanic language since its pronouns, demonstratives, inflections, and most frequent lexemes (including all the irregular— so-called ‘strong’—verbs) are Germanic. But much more than half of the less-­frequent lexemes are of Romance origin. And of the 200 or so derivational affixes discussed in this book, around 43 per cent come from a Romance source (with a further 18 per cent being from Greek).

2.5 Form and fusion 2.5a Some derivational affixes maintain a constant form and do not affect the form of a root or stem to which they are attached. For example, suffix -ness /-nəs/ (or /-nis/) is simply added to a root, the stress remaining on the root. Thus clever /'klevə/ plus -ness /-nəs/ gives clever-ness /'klevə-nəs/, and lazy /'leizi/ plus -ness /-nəs/ gives lazi-ness /'leizi-nəs/.

2.5 form and fusion   21 2.5b Prefix un-1 /ʌn-/, applying to adjectives, is also of this type. For example, un- /ʌn-/, plus clean /kli:n/ gives un-clean /ʌn-'kli:n/. Interestingly, there are some instances where the plain root has almost dropped out of use from Standard English (although it may continue in some regional dialects) but the un- form is still prevalent. For example, unwieldy ‘not easily handleable’; the base form wieldy ‘easily handleable’ was common until the seventeenth century. And un-kempt ‘untidy, not cared for’, which is much used today, comes from the archaic adjective kempt ‘(hair or wool) combed’, which was itself derived from the ME verb to kemb ‘to disentangle hair by running a comb through it’. 2.5c Suffix -ish is of the same character. Thus -ish /-iʃ/ added to hell, /hel/ simply gives hell-ish /'hel-iʃ/. Again, there are instances where the original root is scarcely used (or not used at all) today while the derivation thrives. Mawkish ‘nauseating, overly sentimental’ emanated from an old noun mawk ‘grub, maggot’. Raffish ‘attractive, in a vulgar way’ comes from an old form raff ‘worthless (people)’. This was in turn a shortening of rif and raf ‘every particle; things or people of small value’, a borrowing into OE from French (and this has given rise to present-day riff-raff ‘people who are of no consequence’). A further example is churlish, mentioned in 2.4b. If mawk and raff do not exist in the modern language outside mawkish and raffish, then they fail the criterion to be considered a morpheme, set out in 1.1c. Today, mawk and raff do not have independent significance. Meaning attaches just to the complete words, mawkish and raffish, which are thus to be treated as unanalysable roots. 2.5d Prefix re- (which is of Romance origin) has form /ˌri:-/. It is simply added to a root; the root retains primary stress, a secondary stress falling on /ˌri:-/. Thus re- /ˌri:/ plus float /flout/ gives refloat /ˌri:'flout/. There are many words in Modern English which begin with re but for which the form without the first syllable does not occur; for example, register, refrain, remember. The initial re in such words does go back to a prefix re- in Latin or French but it cannot be recognized as a prefix in present-day English. This re is an integral part of the root—it has the form /re/ and bears stress in register /'redʒistə/; it has the form /ri/ and is unstressed in refrain /ri'frein/, and it has the form /rə/ and is again unstressed in remember /rəm'embə/. Two words recover are written in the same way but have different form and meaning. Alongside unanalysable root recover /ri'kʌvə/ ‘regain normal state

22    2 how to make new words after illness or upset’, there is prefix re- /ˌri:-/ plus root cover /'kʌvə/ giving recover /ˌri:'kʌvə/ ‘cover again’. (Similar doublets include recount| re-count and reform| re-form.) 2.5e In contrast to the three affixes just discussed, there are a number which engender a change in form for some of the roots to which they are attached. Adjective-forming suffix -ous /-əs/ is of this type. (1) With some roots, -ous is simply added to the root (as happens with un-1, -ish and re-). For example fame /feim/ plus -ous /-əs/ produces fam-ous /'feim-əs/. Similarly for glori-ous /'glɔ:ri-əs/ and venom-ous /'venəm-əs/. (2) With a number of others, stress moves to the syllable preceding -ous (with concomitant change in vowel values). When -ous is attached to courage /'kʌridʒ/ we get courage-ous /kə'reidʒ-əs/. (3) With a further set, an unstressed vowel is omitted when -ous is added. For example, from monster /'mɔnstə(r)/ we get monstr-ous /'mɔnstr-əs/. (4) Or a final /i/ drops, as when from adultery /ə'dʌltəri/ we get adulter-ous /ə'dʌltər-əs/. (5) Or a final /f/ may be voiced before -ous, as in mischiev-ous /'mistʃiv-əs/ from mischief /'mistʃif/. 2.5f There are thus two techniques. For un-2, -ish, re-, and set (1) of roots with -ous, there is simply the addition of affix form to root form. For (2−5) with -ous, root and affix are essentially fused, to create a new word whose form could not be predicted (within the grammar of Modern English) from knowing the forms of root and affix. The nature of each instance of fusion may reflect the grammatical structure and phonological rules in the language from which affix and root were borrowed; or it may reflect an earlier stage of English. For example, the alternation between /f/ and /v/ in mischief /'mistʃif/ and mischiev-ous /'mistʃiv-əs/ goes back to an earlier situation in which [f ] and [v] were variants of a single consonant, [v] occurring word-medially and [f ] elsewhere. 2.5g A common suffix used to derive nouns from verbs is orthographically -ation, -ion, -ication, -tion, or -ition, and phonologically /-eiʃən/, /-ʃən/, /-ikeiʃən/, or /-iʃən/. It never involves simply addition, always some variety

2.5 form and fusion   23 of fusion. A sample of the extensive possibilities is (see 9.4.7 for a full account): (a) The nearest thing to simple addition is where the suffix just engenders stress shift. This happens with a form ending in -ize, /aiz/; for example, organize /'ɔ:gənaiz/ giving organiz-ation /ɔ:gənai'z-eiʃən/. (b) For a form ending in -ate /eit/, the final /t/ is replaced by /ʃən/, as in nominate /'nɔmineit/, nomin-ation /'nɔminei-ʃən/. (c) For a form ending in -ify /-ifai/, the final /ai/ is replaced by /-ikeiʃən/ and stress is shifted from first to third syllable, as in qualify /'kwɔlifai/, qualification /kwɔlif-i'keiʃən/. 2.5h For the examples of -ous and -((a)t)ion derivations just quoted, the unaffixed form does occur in present-day English. But this is not always so. English borrowed from French pairs of words with and without suffix -ous, such as glory and glorious, courage and courageous. It also borrowed French forms ending in -ous without also taking over the plain root. Within English, words such as tremendous and precious are not analysable. The criterion for being a morpheme, given in 1.1c, requires occurrence as a free form (making up a word on its own), or occurrence within at least two words. Tremend and prec occur each only in one word, tremendous and precious, and so cannot be recognized as morphemes. A similar scenario applies for the -((a)t)ion suffix. Organiz-ation and qualifi-­ cation can be analysed into root-plus-affix since we have plain forms organize and qualify. But this does not hold for region and fashion. Reg and fash do not occur outside these words which are, as a consequence, unanalysable. So far so good. But now consider the following paradigm: (1)

abstract noun ambit-ion nutrit-ion relig-ion superstit-ion

adjective ambit-ious nutrit-ious relig-ious superstit-ious

This is not an exhaustive list. Other pairs include: contagion| s, sedition| s, and faction| s.

What has happened here is that English has borrowed two (sometimes more) French forms based on the same root, although it has not borrowed the

24    2 how to make new words root. Ambit- occurs in more than one word and thus satisfies the criterion for being recognized as a morpheme. Since ambit- does not occur as a free form, we can call it a bound root, a type of root which only occurs with an affix. As a consequence, ambit-ion and ambit-ious are analysable into bound-rootplus-suffix. And similarly for bound roots nutrit-, relig-, and superst-. An important point is that -ous and -((a)t)ion had already been established as bona fide affixes, occurring with free roots. 2.5j A number of similar paradigms can be constructed, where bound roots may be recognized on the basis of occurrence in two or more words, each with an established affix. For example: (2)

verb stimul-ate particip-ate lubric-ate emigr-ate

concrete noun stimul-ant particip-ant lubric-ant emigr-ant

abstract noun stimul-ation particip-ation lubric-ation emigr-ation

Suffixes -ate, -ant, and -((a)t)ion each occur with free nouns (as in alienate, inhabit-ant, inform-ation) and on this basis we can recognize stimul-, ­particip-, lubric-, and emigr- as bound roots. The same suffixes occur in radi-ate| radi-ant| radi-ation and domin-ate|  domin-ant| domin-ation, where the -ant form is an adjective. Bound roots radi- and domin- may thus be recognized. Now consider: (3)

abstract noun exorc-ism antagon-ism hypnot-ism bapt-ism optim-ism sad-ism

concrete noun exorc-ist antagon-ist hypnot-ist bapt-ist optim-ist sad-ist

verb exorc-ize antagon-ize hypnot-ize bapt-ize optim-ize —

Suffixes -ism, -ist, and -ize are well-established, all occurring with free roots (for example, heathen-ism, balloon-ist, symbol-ize) and thus exorc-, antagon-, hypnot-, bapt-, optim-, and sad- should be recognized as bound roots. Other bound roots that occur just with -ism and -ist include pessim-, fasc-, soph-, hedon-, pragmat-, nepot-, athe-, monothe-, panthe-, and polythe-.

2.5 form and fusion   25 Plagiar-ism| plagiar-ist| plagiar-ize is an interesting set. The abstract noun was originally plagiary, from which plagiar-ist and plagiar-ize were naturally derived. In an unusual development, plagiary was replaced by plagiar-ism, entailing the recognition of plagiar- as a bound noun.

Another paradigm on similar lines is: (4)

adjective enorm-ous audac-ious feroc-ious sagac-ious

abstract noun enorm-ity audac-ity feroc-ity sagac-ity

Suffixes -ous and -ity are well-attested with free roots (for example, ­poison-ous, modern-ity) and we can thus recognize enorm-, audac-, feroc-, and sagac- as bound roots. Other examples occurring in this paradigm include: anonym-, anxi-, frivol-, loquaci-, magnanim-, mendaci-, precoci-, pugnaci-, pusillanim-, salaci-, unanim-, and voraci-.

2.5k In paradigms (1−4), each column involves a regular suffix. We can now examine the following: (5)

concrete noun diplomat democrat aristocrat plutocrat autocrat

abstract noun diplom-acy democr-acy aristocr-acy plutocr-acy autocr-acy

We could say that the first part is, in each instance, a bound root. But what of the last part? Suffix -acy is attested with free roots (for example, conspiracy, see 9.2.3). But there is no suffix -at with free roots. Yet -at does occur in the five words of the left-hand column, with similar semantic import in each. If it is to be recognized as a suffix, it is one which is limited to occurrence with bound roots. A more extreme paradigm is: (6)

adjective splendid squalid candid pallid

abstract noun splendour squalor candour pallor

26    2 how to make new words One could suggest analysis into bound roots splend-, squal-, cand-, and pallplus suffixes -id, and -our or -or (both pronounced as /-ə(r)/. The difficulty is that neither -id nor -o(u)r is a suffix occurring with free roots. For this analysis, two special suffixes, which only occur with bound roots, would have to be recognized. As at many places in linguistic analysis, it is hard to know when to draw the line. One possibility, which I rather favour, is to add a further criterion to that given in 1.1c—a form may only be recognized as an affix if it occurs with some free nouns. This would maintain the morphological analysis of paradigms (1−4), and the -acy column in (5), involving bound roots plus bona fide suffixes, but would lead to the words in paradigm (6), and in the -at column of (5), being considered not to be morphologically analysable. Decisions on such matters are very much a matter of opinion. (There is no definitive ‘right’ or ‘wrong’.) Other pairs of similar form include: fervid| fervour and torpid| torpor. Note that we get horrid, horror, and also horrify—this derived verb does include a suffix, -ify, which occurs with free roots. And also terror and terrify (but there is no adjective *terrid ).

2.5m Only a handful of derivational affixes take primary stress: Romance suffixes -aire (9.3.12a), -arium, and -orium (9.6.1), -eer (9.3.2), and -ese (8.3.7, 9.3.7), Greek suffixes -ology (9.5.8), -itis (9.5.10b), and -­ometer (9.7c), and Romance prefixes counter- and contra- (5.25−27). There is a special circumstance in which a derivational affix may take a primary stress (in addition to the primary stress within the stem). In 2.4g, two recent phases of use for prefix mega- were described. In its scientific sense, mega- is a subsidiary part of the word, and takes secondary stress. Thus megabyte /ˌmega'bait/ is about bytes with mega- acting as a numerical quantifier (there are a million bytes). Then there is the sense of mega- introduced in the 1960s, when it was deliberately employed to indicate outstanding size and importance. Suppose that several enterprises are being combined, and the word mega-business is used to describe the new colossus. It is mega- that is being focused upon (listeners know that businesses are being discussed), and here the prefix is likely to be accorded primary stress (in addition to the primary stress on the root), /'mega-'biznis/. 2.5n Some derivational prefixes (predominantly disyllabics) may take secondary stress. This was illustrated in 2.4d with re- /ˌri:-/; for example re-seal /ˌri:-'si:l/. Another example is anti- /ˌanti-/, as in anti-aircraft /ˌanti-'eəkra:ft/.

2.6 productivity   27 There are fewer examples of secondary stress among suffixes. They include many of Germanic origin, including -like (8.2.1), -free (8.2.29), -proof (8.2.40, 8.4.18), -most (8.2.42), -monger (9.3.13a), and -maker (9.3.13d). 2.5p If fusion is involved, a derivational suffix may affect the placement of primary stress within the stem to which it is attached. For example, we saw in 2.5e that courage /'kʌridʒ/ bears initial stress, but when -ous /-əs/ is suffixed to it, stress moves to the second syllable, whose vowel is strengthened: courage-ous /kə'reidʒ-əs/. The -((a)t)ion suffix has a complex phonology. As shown in 2.5g, it has the form /-eiʃən/ with organize /'ɔ:gənaiz/ and here stress shifts to the fourth syllable, which is made up of /z/ from the root and the initial /ei/ of the suffix: organiz-ation /ɔ:gənai'z-eiʃən/. The important point is that stress never goes on the main part of the suffix, /-ʃən/ (that which is shared by all the variant forms of the suffix). Indeed, it could not, for a syllable whose vowel is a short schwa /ə/ never takes primary stress. It seems that the suffixes which affect stress in the base are all of Romance origin. Besides -ous (8.2.14) and -((a)t)ion (9.4.7), they are -ary (8.2.39, 9.6.1), -al|- ial (8.2.19), -(at)ory (8.4.14), -i (8.3.6), -ity and -ability (9.2.1), -ic (8.2.16, 8.3.3, 8.4.2b), -ian (8.3.9, 9.3.7), -icide (9.5.10a), and -(i)ana (9.7a).

2.6 Productivity 2.6a A derivational affix can be termed ‘productive’ if it is, from time to time, used with a new form to create a stem not previously encountered. This must happen naturally, in the regular course of language use, without any particular thought being given to the matter. The new usage may begin with just one speaker, or perhaps with several people each producing the new stem separately and at more-or-less the same time. This new employment of an established affix must then gradually and imperceptibly come into general use, as an accepted item in the inventory of the language. If just one person uses a derivational affix in a novel way—whether deliberately or naturally—this is simply anecdotal, not any sort of general development in the language. Suffix -en makes verbs out of adjectives. It is only used with adjectives from a certain semantic set, which includes the three main colour terms. And it is only added to roots ending in /p/, /t/, /k/, /f/, /s/, /ʃ/, /θ/ or /d/.

28    2 how to make new words One can say whit-en, black-en, and redd-en but not *yellow-en or *blu-en or *grey-en or *scarlet-en. (See 2.6e and 7.3.2.) Samuel Beckett often took artistic liberties with language. In his novel Watt (Beckett 1963: 36), one reads: Watt saw, in the grate, of the range, the ashes grey. But they turned pale red, when he covered the lamp, with his hat. . . . So Watt busied himself a little while, covering the lamp, less and less, more and more, with his hat, watching the ashes greyen, redden, greyen, redden, in the grate, of the range.

Becket was deliberately creating greyen, by analogy with redden. This produces an intriguing literary effect, but it does not make greyen into an acceptable word for everyday use. Now consider a new form being created not deliberately, as Beckett did, but naturally. A couple of years ago, entirely without thinking, I used appropriacy as the abstract noun corresponding to adjective appropriate, presumably by analogy with such pairs as accur-ate| accur-acy and intimate| intim-acy. Then friends told me that appropriacy is not a word in English. But surely it must be—appropriate, like accur-ate and intim-ate—is a Romance root, and -acy is a Romance suffix. I consulted the standard dictionaries and found, to my surprise, that there was no mention of appropriacy. The abstract noun is appropriate-ness, with Germanic suffix -ness. No other competent speaker employed appropriacy. I could have continued using it. If others had naturally (by unconscious imitation) joined me in this, it would have been the first new use of the -acy suffix (nowadays considered to be unproductive) for a couple of centuries. But such a happening would have been most unlikely. Instead, I conformed, and shamefacedly substituted appropriate-ness for each appropriacy in the draft I was writing. (And see 2.6h.) 2.6b The productivity of a derivational affix depends on the combination of a number of factors: (1) Its meaning, and the extent to which speakers feel a need to utilize that meaning. (2) Inherent semantic restrictions—some derivational affixes are restricted to use only with forms from certain semantic types. (3) Phonological restrictions—there are quite often constraints relating to the phonological nature of a root or stem which can accept a certain affix.

2.6 productivity   29 (4) Whether the affix is of Germanic or Romance (or some other) origin, and whether it is mainly confined to use with roots of the same origin, or has been—to a greater or lesser extent—generalized for use with roots of all sorts. (5) How well an affix fares in competition with other affix(es) which have similar function and meaning. It sometimes happens that one of a series of ‘rival’ affixes may be, as it were, in fashion for a while, with there being a preference for using it rather than the other(s). Then, a century or two later, habits may change, with a competitor now moving into favour. We can provide preliminary exemplification for these points. Many further examples are included in the chapters that follow. 2.6c To illustrate factor (1), consider two negative prefixes, both of Romance origin: non- and counter-. Speakers of English have a great deal of cause for employing non- ‘not a member of a specified class’. Suppose that a new word comes into use, sub-prime-monger, referring to people who sell sub-prime mortgages. One can immediately add non- and insist that a reputable bank manager is a nonsub-prime-monger. Non- is highly productive because there is considerable demand for it. Counter- ‘do the opposite of, be the opposite to’ is also productive, being available to create new words. For instance, a flurry of new forms came into circulation around the 1960s, including counter-terrorism, counter-coup, counter-intuitive, and counter-example. But it is much less used than non-, simply because it is not needed anything like so much. 2.6d Derivational suffixes -er and -ee are, to a certain extent, complementary. Among its other uses (see 2.3b), -er indicates someone who controls an activity, such as publish-er, employ-er, danc-er, swindl-er. The main function of -ee is to indicate someone affected by an action, such as employ-ee, appoint-ee, detain-ee; or someone exiting from a situation, such as escap-ee, divorc-ee, retir-ee, absent-ee. Why is it then that, on a dictionary count, agentive -er words are about twenty times as common as -ee words? One reason is that one often wants to describe someone as habitually doing something, much less often as having something habitually done to them. Someone who often swindles people is

30    2 how to make new words a swindl-er. But it is likely to be a different person they swindle each time. One is unlikely to encounter someone who habitually gets swindled (and who might be called a *swindl-ee). A murder-er shows a propensity for killing people. But since a person can die only once, the idea of a *murder-ee is nonsensical. (One could of course use it in a science fiction story about a race who have many lives.) Another factor is that many -er agentive derivations refer to actions which have non-human patients. For example: found-er (of a city or a business), hoard-er (of things), forg-er, fish-er, and dozens of others. All this illustrates factor (1) from 2.6b. Factor (2) also comes into play here. Suffix -er is of Germanic origin but is now highly productive and has been generalized to apply to forms of any genetic origin; for example ­observ-er (onto a Romance form). In contrast, suffix -ee emanates from a participial ending in French and is largely used with roots or stems of Romance origin. 2.6e As mentioned in 2.6a—when commenting on Samuel Beckett’s nonce use of greyen—application of the verbalizing suffix -en relates to factor (1), meanings of forms to which it is attached, and also factor (3), their phonological endings. First, -en is used with adjectives from the dimension and physical property semantic types (see 2.9 and the Appendix)—wid-en, deep-en, hard-en, light-en. It is seldom employed with human propensity adjectives (there is no *proud-en, *stupid-en, or *honest-en). The suffix is used with the three colour adjectives at the top of Berlin and Kay’s (1969) hierarchy—­whit-en, black-en, redd-en—but not usually with those lower down—there is no attestation for *scarlet-en or *violet-en, although pink-en has come into occasional use during the last century or so. Secondly, as stated in 2.6a, -en is restricted to use with roots ending in /p/, /t/, /k/, /f/, /s/, /ʃ/, /θ/, or /d/. Greyen is excluded (save as a literary affectation) on phonological and on semantic grounds. (There is a fuller discussion of this in 7.3.2.) 2.6f There are two productive derivational suffixes for deriving verbs, both of Romance origin. In 2.2e, we enquired why some forms take only -ify (fals-ify not *fals-ize, ugli-fy not *ugl-ize) and others only -ize (legal-ize not *legal-ify; patron-ize not *patron-ify). Here factor (3) comes into play—the reason is almost entirely phonological.

2.6 productivity   31 Simplifying somewhat, -(i)fy is used with monosyllabic roots and with disyllabic roots ending in a vowel, while -ize is employed with disyllabic roots ending in a consonant. There are other factors, including placement of stress. A large part of Chapter 7 is devoted to contrastive discussion of these two suffixes. 2.6g Three prefixes which have essentially the same meaning were mentioned in 2.2d. Mono- is of Greek, and uni- of Romance origin, while one- is a Germanic number word which also functions as a derivational prefix (or semi-prefix, see 3.3). The use of these prefixes illustrates factor (4), that the genetic origin of an affix may determine the types of forms it attaches to. Roots which come from Greek only take mono-; they include mono-mania, mono-syllabic, and mono-gamy (-gamy is recognized as a bound root since it also occurs in bi-gamy and poly-gamy, also with Greek prefixes). Most Romance roots take uni- (uni-polar, uni-dimensional) although some do take mono- (mono-­ lingual, mono-rail). For some roots either prefix is possible—mono-valent and uni-valent are both in use, with the same meaning, as are mono-cellular and uni-cellular. Germanic roots only take one-, not mono- or uni-; for example, a one-horse carriage, a one-legged man. Similarly, Romance and Greek roots by-andlarge shun one-. (There are just a few crossovers, such as one-dimensional alongside—and, indeed, much more common than—uni-dimensional.) 2.6h Suffixes -(a)cy and -ness, which both derive abstract nouns, were mentioned in the discussion—in 2.6a—of my nonce use of appropriacy. The Germanic suffix -ness was much used in Old English and by 1300 was being freely added to loans from French. In the fourteenth century there came Romance loans such as obstin-ate and obstin-acy, confeder-ate and confeder-acy. From analysis of such pairs, -acy became adopted as a suffix within English for deriving an abstract noun from an adjective of Romance origin ending in -ate. Early in the seventeenth century, intim-ate, accur-ate, and appropriate were borrowed from Romance sources. Quite soon, abstract nouns were created in two ways. First by adding -ness. And secondly by replacing the final -ate by -acy (this was a derivational process within English, applying to Romance forms). The interesting feature is that -ness was added to all of intim-ate, accurate, and appropriate whereas -acy forms were created only for the first two

32    2 how to make new words (the reasons for this are still to be understood). For a while one could say intim-ate-ness or intim-acy, accur-ate-ness or accur-acy. There was then a preference in favour of the -acy alternative—intim-ate-ness and accur-ateness are now quite uncommon, while accur-acy and intim-acy have high frequency. For appropriate there never was an -acy derivation, and the abstract noun here is appropriate-ness. This illustrates competition between affixes, factor (5) from 2.6b. The OED gives ‘intimateness = intimacy’. It states that, ‘more properly’, accurateness is ‘quality of a person’ while accuracy is ‘state of a thing’ (e.g. an experiment, or how a garden is kept). The Google electronic search-engine gives accur-acy as more than a thousand times more common than accur-ate-ness, suggesting that is it today used in both senses.

2.6j An interesting example of competition between suffixes—factor (5)— involves abstract-noun-forming derivations -ity and -ness. Consider the forms in Table 2.2. Suffix -ity became recognized as a derivation in English through pairs of Romance borrowings such as captive and captiv-ity, diverse and divers-ity. As mentioned in 2.4f, -ous came to be seen as a suffix through pairs such as advantage and advantage-ous, danger and danger-ous. These suffixes occur in columns (a) and (b) of Table 2.2,

Table 2.2  Parallel forms in -ity, -(i)ous, and -(i)ous-ness with dates of their first mention in the OED (a) abstract noun

(b) adjective

(c) abstract noun

audac-ity, 1432 /ɔ:'das-iti/

audac-ious, 1550 /ɔ:'dei-ʃəs/

audac-ious-ness, 1598 /ɔ:'dei-ʃəs-nəs/

vivac-ity, 1475 /vi'vas-iti/

vivac-ious, 1645 /vi'vei-ʃəs/

vivac-ious-ness, 1661 /vi'vei-ʃəs-nəs/

mendac-ity, 1540 /men'das-iti/

mendac-ious, 1616 /men'dei-ʃəs/

mendac-ious-ness, 1829 /men'dei-ʃəs-nəs/

tenac-ity, 1526 /te'nas-iti/

tenac-ious, 1607 /te'nei-ʃəs/

tenac-ious-ness, 1642 /te'nei-ʃəs-nəs/

pugnac-ity, 1605 /pʌg'nas-iti/

pugnac-ious, 1642 /pʌg'nei-ʃəs/

pugnac-ious-ness, 1681 /pʌg'nei-ʃəs-nəs/

feroc-ity, 1606 /fə'rɔs-iti/

feroc-ious, 1646 /fə'rou-ʃəs/

feroc-ious-ness, 1766 /fə'rou-ʃəs-nəs/

precoc-ity, 1606 /pri'kɔs-iti/

precoc-ious, 1650 /pri'kou-ʃəs/

precoc-ious-ness, 1681 /pri'kou-ʃəs-nəs/

2.7 prefixes and suffixes   33 but are here attached to bound roots, audac(i)-, vivac(i)-, mendac(i)-, tenac(i)-, pugnac(i)-, feroc(i)-, and precoc(i)-. Note that the bound roots, although with constant orthographic form, have phonological variants. Bound roots in the first five rows end in /as/ before -ity /-iti/ and in /eiʃ/ before -ous /-əs/, while those in the last two rows end in /ɔs/ and /ouʃ/ respectively. Now consider the historical scenario, examining—for each word—the date of first mention in the OED. Each form in column (a) appeared in English before the corresponding form in column (b), which was in turn before that in column (c). First of all, English took from Romance sources the abstract nouns in column (a). A little later it took in, from the same sources, corresponding adjectives in column (b). The Germanic suffix -ness gets applied to adjectives of all genetic origins and was added to forms in column (b) creating a second set of abstract nouns, in column (c). We now have competing abstract nouns, in columns (a) and (c), corresponding to the adjectives in column (b). Those in column (c) simply involve the addition of /-nəs/ to the adjective, whereas there is fair phonological difference between the bound roots in columns (a) and (b). Both the -ity and -ness abstract nouns are acceptable in present-day English but for six of them the -ity form, from column (a), is many times more common than the -ness form, in column (c)—that is, audac-ity, vivac-ity, mendac-ity, tenac-ity, pugnac-ity, and feroc-ity are the preferred forms. The exception is that precoc-ious-ness is slightly more common than precoc-ity. It seems that in most cases speakers of English, quite unconsciously, prefer a form of Romance origin with the Romance nominalizer -ity, rather than one with the highly-productive Germanic nominalizer -ness. Table 2.2 could be extended with sagac-ity| sagac-ious| sagac-ious-ness, vorac-ity| voracious| vorac-iou-sness, and poros-ity| por-ous| por-ous-ness. Comments on relevant frequency are based on the Google electronic search engine. There have been a number of publications on morphological productivity in English, many of them suggesting various kinds of statistical modelling. See the monographs by Plag (1999) and Bauer (2001) and further references to the literature therein.

2.7 Prefixes and suffixes It is instructive to compare the characteristics of the 90 or so derivational affixes which are prefixes and the roughly 110 which are suffixes. (Stress properties were outlined in 2.5m−p.)

34    2 how to make new words 2.7a Genetic origin

• •

Of prefixes, about 31 per cent are of Germanic origin, 42 per cent Romance, and 27 per cent Greek. Of suffixes, about 43 per cent are Germanic, 44 per cent Romance, 10 per cent Greek, and 3 per cent others.

2.7b Forms

• •

Prefixes are almost equally split between monosyllabic and disyllabic forms. Suffixes show more variation. About 70 per cent are monosyllabic, 24 per cent disyllabic, and two are trisyllabic (-ology in 9.5.8 and -atory in 8.4.14 and 9.6.1). There are also a handful of suffixes which just involve a final consonant, illustrated by insomnia-c, gif-t, heigh-t, ­warm-th, ­ten-th, Africa-n.

2.7c Functions

• •

Most prefixes do not change word class (or sub-class for a noun) but simply add a semantic element. Negators are all prefixes. Almost all suffixes serve to derive a stem of a different class or subclass from that of the underlying base.

Note also that two suffixes must always apply in a fixed order (a consequence of their class-changing nature), whereas there are pairs of prefixes which may occur in either order; see 11.3. 2.7d Generally, prefixes and suffixes are quite distinct. However, there is one prefix/suffix pair with essentially the same form and meaning. The Greek noun philós ‘desire’ developed into both a prefix and a suffix in that language. Through Latin, both have come down into English as rather marginal instances of derivational prefix and suffix. ME accepted early loans such as philosophy (literally ‘liking for or study of wisdom’) and philology (‘liking for or study of words’). These are not analysable within English. However, phil(o)- /fil(ou)-/ ‘liking for’ came to be adopted as a prefix, generally used in an ad hoc way. From 1616 we have

2.8 double duty   35 one instance of philo-mythology ‘liking for the study of myths’. The OED includes two examples of philo-pig. One, from 1828, describes someone who adores ‘all forms of pork, baked, roasted, toasted, boil’d or broil’d’. That from 2004 has a quite different tilt, being used for someone who considers that ‘those adorable porkers have the sweetest smiles’. In the mid-nineteenth century, -phile /-fail/ emerged as a suffix meaning ‘liking for’. We find Indo-phile ‘a lover or champion of the native inhabitants and interests of India’, Anglo-phile ‘a supporter or admirer of England and its people’, and so on. Associated with such designations for people, there are adjectives with -philic and abstract nouns with -philia—thus, Anglophilic, Anglo-philia. Suffix -phobe /-foub/ is the antonym of -phile. An Anglo-phobe is someone who is hostile to England and all things English; also -phobic, -phobia. However, there is here no corresponding prefix.

2.8 Double duty 2.8a We have described two varieties of derivational affix. One kind does not change word class. Re-elect is a verb, just like elect, while arch-villain and villain are both nouns, and super-clever and clever are both adjectives. The other kind of affix changes word class. For example, adding suffix -en to adjective wide derives verb wid-en, as in: (1)  The river becomes wideADJECTIVE just past the town (2)  The river wid-en-sVERB just past the town Now compare these with corresponding sentences involving narrow (the antonyms of wide): (3)  The river becomes narrowADJECTIVE just past the town (4)  The river narrow-sVERB just past the town We can say that narrow does ‘double duty’. It is basically an adjective (like wide) but has secondary function—in its bare form, without any derivational affix—as a verb. Let’s look an another example. From verb discover a noun is derived by suffixing -y, giving discover-y, as in: (5)  By sheer luck, he discover-edVERB an old Roman coin (6)  His lucky discover-yNOUN today was an old Roman coin

36    2 how to make new words Find has a meaning very similar to that of discover. Using find, we get: (7)  By sheer luck, he foundVERB an old Roman coin (8)  His lucky findNOUN today was an old Roman coin Find is another double duty word. Although primarily a verb (like discover), it may have secondary function—in bare form, without any derivational affix—as a noun. (A verb will of course, take the appropriate inflection—present tense -s in (2) and (4), regular past -ed in (5), and found, the irregular past form of find, in (7).) What I am calling ‘double duty’ is sometimes referred to as ‘conversion’ or as ‘zeroderivation’ (implying that just as wide adds -en to form a verb, so narrow adds a zero suffix to form a verb). ‘Conversion’ implies a process of change, while ‘zero-derivation’ invokes the notion of derivation by means of an affix (albeit one with zero form). ‘Double duty’ simply states that a given lexeme, in addition to its primary membership of one word class, has secondary function in another.

2.8b Each of the major word classes—noun, verb, and adjective—has some members which do double duty in one of the other classes, except that it appears that no verb does double duty as an adjective. In addition, we find a handful of prepositions which may also function as noun or as verb in limited circumstances. Table 2.3 exemplifies the possibilities.

Table 2.3  Possibilities for double duty PRIMARY WORD CLASS MEMBERSHIP:

ALSO DOES DOUBLE DUTY AS: NOUN

ADJECTIVE

VERB

ADVERB

NOUN

—

fat native fun

trap knife elbow

—

ADJECTIVE

human original potential

—

narrow clean blind

hard fast

VERB

walk delight cook

—

—

—

PREPOSITION

down in

—

up out

—

2.8 double duty   37 There are a few lexemes which have triple duty. For example, rival is basically a noun (for instance, He eliminated his rival in the struggle for power) but also functions as an adjective (They were bought out by a rival firm) and as a verb (Your chocolate cake cannot rival Mrs Green’s). We have seen, in (4), that adjective narrow may also function as a verb. In addition, it occurs as a noun, then usually in plural form, as in He expertly guided the boat through the narrows just past the town. 2.8c It is legitimate to enquire how we know what the primary function is for a double duty word. There are a number of factors. First, we have the intuitions of native speakers, which are reflected by the order in which word class memberships are stated in dictionaries. For walk we find ‘v, n’ (not ‘n, v’), and for knife ‘n, v’ (not ‘v, n’). And then there is frequency of use for each function; this does not provide a hard-and-fast criterion, but it does generally go along with intuitions. For example, walk is used as a verb much more often than as a noun, and knife is used as a noun far more frequently than as a verb. There is also a semantic basis. All members of a semantic set will share primary word class membership, but just a selection of them (often, some of those with the most generic meaning) will also do double duty in another word class. Consider the following set of words referring to sharp-pointed or sharp-bladed weapons: knife, spear, rapier, dagger, stiletto, scalpel, javelin, and sword. They all belong to the noun class, with only knife and spear being frequently found doing double duty as verbs. These are the two members of the set with the most general meaning—a scalpel is a kind of knife and a javelin is a kind of spear, and not vice versa. Knife and spear are also the most frequently used words from the set. Note also that their meanings as verbs are more restricted than their meanings as nouns; one may say both kill an enemy with a knife and knife an enemy, but corresponding to cut a cake with a knife it would not be felicitous to say *knife a cake. All of the semantic set throw, fling, chuck, and hurl are verbs. Throw has the most general meaning—flinging, chucking, and hurling are all special kinds of throwing. In keeping with this, throw is frequently found doing double duty as a noun (the other three have the potential for this, but take it up only rarely). Butter is sometimes used as a general term which also covers margarine. In keeping with this, butter also does double duty as a verb whereas margarine is only used as a noun.

38    2 how to make new words Brinton (2000: 91–3) has an excellent discussion concerning which class membership is primary for a double duty lexeme.

2.8d A verb with double duty will not undergo affixal derivation, and vice versa. The two techniques do interrelate, with some verbs in a given semantic set employing one practice, and the remainder the other. Consider, for example: victim, pauper, cripple, martyr. The first two add -ize (victim-ize, pauper-ize) to create a verb ‘make into a –’, while cripple and martyr simply do double duty, each also functioning as its own causative verb (for example, The blows from the cruel step-father’s cudgel crippled the child ). Consider another example. Verbs describing various kinds of vocation form an agentive nominative by adding derivational suffix -er—build-er, bak-er, repair-er, weav-er. None of these verbs undertakes double duty. The verb cook, in contrast, does; someone who undertakes cooking is a cook. Why does cook behave in a different way from build, bake, and so on? It may be because -er has another function with verb cook: a cooker is an apparatus with which one cooks (typically called a stove in American English), alongside other apparatus labels derived from verbs using suffix -er, such as mix-er, wring-er, and heat-er. In Chapters 7 and 9—on making new verbs and new nouns—discussion of double duty lexemes is integrated into the examination of derivational processes marked by affixes. As a preliminary to this, we can provide some comment on each of the cells in Table 2.3. 2.8e Prepositions as nouns and as verbs English has a goodly array of phrasal verbs, such as take over (a company), make up (a story), make up (a bed), put up (the price), bring out (a secret). For a number of rather significant phrasal verbs, the verbal component may be omitted leaving just the preposition, which then appears to be functioning as a verb. For example, one may hear: She upped the price (implying put up), or The scandal sheet outed the judge’s homosexuality (meaning brought out). Other instances of out used as a verb relate to its adverbial function, with a lexical verb being omitted when its identity is obvious from the rest of the sentence. For example, The butler outed the candle (put out, extinguished), The rapacious landlord outed the penniless tenant (put out, evicted), The boxer outed his opponent (knocked out). The possibilities for prepositions being used as nouns are rather limited and idiomatic. One can have a down on someone, meaning that they are

2.8 double duty   39 determined to do them down (affect them deleteriously). And an employee may have an in with the boss (being accorded insider status). A quite different matter is that some prepositions/adverbs have homonyms which function as adjectives. For example, multi-sensed adjective up is used in: She is up (out of bed), Time’s up, The game is up, The road is up, The computer system is up again, after having been down for hours. Adjective down (also with many senses) is in the last example and in: He’s down with the measles, She’s down in the dumps, We’re down to the last banana, It’s down to Margaret to decide.

2.8f Double duty for nouns It appears that there are only a handful of nouns which may also function as an adjective; it is hard to add to fat, native, and fun, given in Table 2.3. In contrast, very many nouns do double duty as verbs. However, note that only some (not all) members of a semantic type have this property. A sample is: Body parts, etc: head, face, eye, nose, neck, shoulder, elbow, hand, finger, knee, foot, bone, skin, tail, shadow, mind (But other body-part terms—for example, ear, nape, wrist—do not function as verbs. Note the wide diversity of meanings when undertaking double duty. For example, one shoulders a heavy load (or the blame), elbows one’s way through a crowd, hands a gift to a friend, bones a chicken, foots a bill, and so on.) Kin terms: mother, father Type of people: dwarf, boss, host, butcher, shepherd Fauna: fish, rat, wolf, horse, monkey, snake, parrot Artefacts: knife, spear, stone, cart, trap, brush, map, balance, frame, model, knot Noises/language: squeak, creak, whistle, question, name, joke, lie Buildings/locations: place, house, store, market, station Substances: water, fire, butter, milk Abstract: plan, puzzle, work, power, scorn, shame, bargain, shape, smell An account in terms of semantic function is in 7.5.1. 2.8g Double duty for adjectives Suffix -en was discussed in 2.6e. This derives a verb from an adjective subject to two conditions. One is semantic: -en is primarily used with adjectives from the dimension and physical property semantic types, and some from the

40    2 how to make new words colour type (see the Appendix). The other is phonological, -en being restricted to roots ending in /p/, /t/, /k/, /f/, /s/, /ʃ/, /θ/, or /d/. Some adjectives in an appropriate semantic type which do not end in an appropriate consonant make up for this lack by themselves doing double duty as verbs. This applies to narrow, shallow, yellow, cool, clean, clear. There are a number of other adjectives which also do duty as verbs (see 7.5.2). They include monosyllabic forms blind, tame, humble, wrong, and right. There are also a number of disyllabic adjectives with this property, which involve a shift of stress from the first syllable, when functioning as an adjective—for example perfect /'pə:fikt/—to the second syllable, when functioning as a verb—perfect /pə:'fekt/. Further examples are absent, adjective /'absənt/ and verb /ab'sent/, frequent, adjective /'fri:kwənt/, and verb /fri'kwent/. For these items, adjective and verb are spelt the same, despite differences in stress (and in second vowel). Quite a few adjectives can be head of a noun phrase, a slot generally filled by a noun, so that they are, in effect, doing double duty as nouns. In almost every instance, this is the result of shortening. If a certain adjective typically modifies a certain noun, then the noun may be omitted with the adjective taking on the full referential load. For example:

• • •

human (being), intellectual/professional/expert (person), private (soldier); medical (examination), social (function); red (wine), empty (bottle), original (painting), special (offer).

Some adjectives have the full possibilities of a noun, but others are more restricted. One can talk of a medical or an expert but when obvious does double duty as a noun it requires the definite article, the—we might hear She stated the obvious (implied: fact); one cannot say an obvious. The is also needed in She is fascinated by the curious| the unusual| the supernatural, which indicates a fascination with curious| unusual| supernatural phenomena. Adjective great may be used as a noun but only when the accompanying co-text or context makes clear what the person described is great at. For instance, we can say She is a tennis great or She is one of the greats of tennis, but not simply She is a| the great. One adjective which does double duty as a noun without it being thought of as the shortening of a noun phrase is potential ‘capable of happening or doing’. As an adjective it often refers to something undesirable—for

2.8 double duty   41 example, I can foresee potential problems with your plan. When used as a noun it invariably relates to something desirable, as in Your son shows tremendous potential (to achieve great things). As shown in Table 2.3, there are two common adjectives which do double duty as adverbs: hard and fast; see 10.5. 2.8h Double duty for verbs As already mentioned, it appears that no verb—in its bare form—does double duty as an adjective. Very many verbs can take part in a special construction: involving have a| take a| give a verb. For example She had a think about the offer, They took a look at the garden, He gave the mixture a stir. This is not here considered an instance of a verb having secondary function as a noun. (See Dixon 1991: 339−62 and 2005a: 459−83 for a full discussion of this construction in British English; note that American English differs rather markedly.) Leaving all this aside, there are many verbs which can do double duty as a noun. For example (see also 9.3−4): Agent: cook, cheat, guide, witness State: love, envy, delight, dislike, scare, shock, surprise Activity: run, walk, crawl, roll, swing, shake, dance, return, visit, stroke, wash, squeeze Unit of activity: smile, laugh, cry, fart, shout, bite Result of activity: wound, scratch, bite, break, plan Referent of syntactic object: plant, taste, smell, supply, pay, suspect Location: approach, exit The discussion here of double duty has been rather short. A great deal more detail will be found in Jespersen (1942: 83−134), Marchand (1969: 359−79), Adams (1973: 37−56), and Bauer and Huddleston (2002: 1658−64). Note that Jespersen uses the term ‘naked word’, Marchand and Adams employ ‘derivation by zero suffix/morpheme’, while Bauer and Huddleston prefer ‘conversion’. 2.8j It is interesting to compare double duty words with a coupling of verb plus derived noun. A selection is shown in Table 2.4. For the first seven rows, the items on left- and right-hand sides have similar meaning, and in the last row they belong to the same semantic domain.

42    2 how to make new words Table 2.4  Comparing double duty items with derivations VERB DOING DOUBLE DUTY AS NOUN

VERB

DERIVED NOUN

Gmc

show

Rom

demonstrate

demonstrat-ion

Gmc

burst

Rom

explode

explos-ion

Gmc

find

Rom

discover

discover-y

Rom

mention

Rom

announce

announce-ment

Rom

return

Rom

depart

depart-ure

NOUN DOING DOUBLE DUTY AS VERB Rom

copy

Rom

imitate

imitat-ion

Rom

question

Rom

enquire

enquir-y

Gmc

fear

Rom

prefer

prefer-ence

Rom = Romance; Gmc = Germanic

An interesting point is that some of the double duty items are Romance and some Germanic, whereas all the couplings on the right-hand side are Romance. We seldom (or never) get Romance on the left and Germanic on the right, or Germanic on both sides. 2.8k In 2.8g we mentioned disyllabic adjectives which do double duty as a verb with the same orthographic form but with stress on the first syllable for the adjective and on the second syllable for the verb (with consequential shift of vowel). There are many similar noun/verb pairs (see 7.2b), with the noun showing initial stress and the verb again having stress on the second syllable. For some of these the primary function is noun—for example, torment, rebel, conduct—and for others the verb sense is primary—insert, insult, suspect. There are also pairs marked by consonantal alternations, sometimes also with vowel shift (see 7.2c):

•

Final /f/ for noun or adjective, final /v/ for verb: calf| calve, thief| thieve, half| halve, belief| believe, life| live (all Germanic), grief| grieve, proof| prove (both Romance) Final /θ/ for noun, final /ð/ for verb: breath| breathe, cloth| clothe, mouth| mouth (all Germanic)

•

2.9 semantic types and semantic roles   43

• •

Final /s/ for noun (written as ‘c’), final /z/ for verb (written as ‘s’): advice|  advise, device| devise (both Romance) Final /s/ for noun, final /z/ for verb, but written the same: house (Germanic), abuse, excuse (both Romance)

And there are noun/verb pairs distinguished just by vowels; for example, blood| bleed, food| feed, writ| write, bit| bite, song| sing, abode| abide (all Germanic). See Jespersen (1942: 186−208) for a more comprehensive discussion.

2.9 Semantic types and semantic roles Words in the lexicon naturally group into a number of what are called ‘semantic types’, groups of words which share a common meaning component and grammatical features. For instance, adjectives such as big, small, broad, narrow, thin, and round belong to the dimension type. New, old, ancient, young, and modern make up the age type. Among verbs, the liking type includes like, hate, prefer, fear, value, and regret, while in the attention type we find see, hear, smell, show, find, witness, and many more. Discussion of the semantic scope and grammatical properties of adjective types—which serve to justify their recognition—is in Dixon (1977: 31–45, 1982: 15−34, 1991: 78−85, and 2005b: 84–93). Discussion of and justification for verb types is in Dixon (1991: 85−206, and 2005a: 93–206). Consideration of semantic types plays a major role in explaining the meanings and function of derivational affixes at many places in this volume. These include various techniques for creating negative words (in Chapter 5), making verbs (in Chapter 7), making adjectives (in Chapter 8), making several kinds of derived nouns (in Chapter 9), and deriving adverbs (in Chapter 10). The Appendix has a list of adjective and verb semantic types, with sample members of each (together with an alphabetical directory). An attempt has been made to include all those adjectives and verbs which have been quoted, for affix attachment, in this volume. Semantic types relating to the noun class have been less studied; there is a listing in 8.2, which helps to explain the scope of various adjectivalizing suffixes. As explained in 1.2, each clause has a predicate, and the nature of the predicate determines which arguments are needed to produce a coherent

44    2 how to make new words clause. Associated with each semantic type of verbs there are a number of ‘semantic roles’, linked to its argument slot(s). For example:

•

Verbs in the attention type require a Perceiver role, which is placed in A (transitive subject) function, and an ‘Impression’ role, in O (transitive object) function. Thus: [The dog]PERCEIVER:A [saw]ATTENTION.VERB [the snake]IMPRESSION:O

•

For the liking type the semantic roles are Experiencer (as A) and Stimulus (as O), as in: [The cop]EXPERIENCER:A [likes]LIKING.VERB [solving crimes]STIMULUS:O

An interesting feature of English is that there is also the annoying type of verbs (whose members include please, delight, annoy, and anger), with the same semantic roles as liking verbs, but paired with syntactic functions in the opposite way—Stimulus as A and Experiencer as O. For example: [Solving crimes]STIMULUS:A [pleases]ANNOYING,VERB [the cop]EXPERIENCER:O The similarities and differences between liking and annoying verbs are invoked in explanation at 8.4.11 and 9.4.1. The semantic roles associated with each verb type are given in the Appendix. Note that they are divided into Primary types, which refer directly to an activity or state, and Secondary verbs, which essentially provide modification for a (stated or understood) Primary verb. Primary verbs are subdivided into Primary-A, all of whose core arguments must be NPs, and Primary-B, for which at least one argument may either be an NP or a complement clause; for example, one can say either IA saw [the crime]O or IA saw [Tom stealing the jewels]O. The next chapter discusses criteria for deciding whether a form is a derivational affix, and contrasts root-plus-affix combinations with compounds.

3 Criteria—affixes, semi-affixes, and compounds

3.1 Compounds versus syntactic  constructions  45

3.2 How to tell it is an affix  47 3.3 Affixes and semi-affixes  54

3.1 Compounds versus syntactic constructions 3.1a Deciding whether a certain sequence is a compound (a morphological unit)—such as gingerbread or grandmother—or, instead, words in syntactic construction—such as wholemeal bread and grand piano—is one of the most difficult problems in linguistic analysis. Accounts of compounding in English provide an inventory of the types (noun plus noun forming a compound noun, as in gingerbread, noun plus adjective forming a compound adjective, as in headstrong, and so on) but typically devote little attention to criteria for determining whether a certain combination is or is not a compound. Easy-going seems always to be regarded as a compound, but I have found no source which identifies easy living as a compound. Why should this be? 3.1b One could suggest meaning as a criterion. The meaning of easy-going (‘calm and undemanding’) cannot be inferred from the meanings of its component words, whereas that of easy living can (it means living that is easy and comfortable, with no financial worries). It is true that many compounds are like easy-going in having a meaning all of their own; for example, cold-shoulder, heavy-handed (mentioned in 2.1b), bullseye, copycat, and cut-throat. But there are also many where the overall meaning is the sum of those of its parts—hen-house (‘house for hens’), colour-blind (‘blind as

46    3 criteria—affixes, semi-affixes, and compounds to colours’), hovercraft (‘craft that hovers’), hairbrush (‘brush used to brush hair’), left-hander (‘someone who uses their left hand for actions that most people would perform with the right hand’), and umpteen dozen more. Meaning is a sufficient criterion—a sequence with a non-predictable meaning will be a compound—but not a necessary one; some compounds do have a predictable meaning. 3.1c How about phonological criteria? Easy-going is shown in some dictionaries as having primary stress in the first syllable and secondary stress on the second, in others with the reverse, and in yet others with primary stress on both syllables. Easy living, if it is made up of two separate words, would presumably be said to have primary stress on both. But put these into a clause— She has an easy-going nature, She is in easy living circumstances. Surely these are likely to have exactly the same stress pattern? Stress is scarcely satisfactory as a criterion. 3.1d What about how they are written—easy-going with a hyphen and easy living without one? That is a significant point. In fact, easy-going is sometimes written with a hyphen and sometimes as one word, easygoing. However, we must note that English orthography is both inconsistent and fluid. Why do people write cannot as one word but may not as two? There is no rational reason, and no linguistic difference between these two combinations of modal verb and negator not. We sometimes find fieldwork written with a hyphen but sometimes without, fieldwork. Similarly for life-guard, lifeguard; wrong-doer, wrongdoer; lawn-mower, lawnmower, and many more. We find table talk alongside table-talk, town crier beside towncrier, and so on. Orthographic conventions are, of course, interesting, but scarcely suitable as analytic criteria. In any case, our aim is to describe language in its most immanent form—that is, speaking. 3.1e One relevant factor is how common a sequence of stems is. Easy-going has a fairly high textual frequency. But so does easy living, whether used as the head of an NP (I appreciate easy living) or as a modifier (He hopes for an easy living retirement). There was a 1937 film called Easy Living, one can subscribe to the Easy Living Magazine, shop at Easy Living Blinds or Easy Living Furniture. The number of times two forms are used together can be indicative of compound status, but scarcely criterial.

3.2 how to tell it is an affix   47 3.1f A syntactic criterion is employed by Payne and Huddleston (2002: 448–9)—the components of a syntactic construction are open to modification and coordination, whereas the components of a compound are not. One can say wholemeal sliced bread (where sliced modifies bread ), as an alternative to sliced wholemeal bread (where sliced modifies wholemeal bread ). And just the first element may be coordinated, as in wholemeal and multigrain bread. In contrast, when dealing with compounds one has to say gingerbread and shortbread, rather than *ginger- and short-bread. One can say easy beachside living and easy and carefree living, which could be taken to indicate that easy living is not a compound. This is perhaps the best criterion, but even it is not watertight. Grandmother and grandfather are clearly compounds, but one could say This weekend I’m going to visit my grand-mother and -father, coordinating the second elements. 3.1g Recognizing ‘what is a compound’ is a tricky matter, involving a combination of syntactic, semantic, and phonological criteria (plus some attention to frequency) together with a considerable degree of fine-tuning. A fuller investigation lies outside the scope of this study. We can now turn our attention to criteria for distinguishing a stem-plusaffix from a compound. This is perhaps a little less demanding, but still far from easy.

3.2 How to tell it is an affix 3.2a There are four necessary criteria for analysing a form as stem-plus-affix (recall that a root is a simple stem). These are conditions which must be satisfied for an affix to be recognized. We shall then mention four sufficient (but not necessary) conditions: anything which satisfies them will be an affix, but not every affix satisfies them. The first two conditions are that stem-plus-affix should be one grammatical word, and one phonological word. A grammatical word is established on grammatical criteria, and is a unit in the grammatical hierarchy ‘morpheme, grammatical word, phrase, clause, sentence’, while a phonological word is established on phonological criteria, and is a unit in the phonological hierarchy ‘phoneme, syllable, phonological word, intonation group’. The two units often coincide (which

48    3 criteria—affixes, semi-affixes, and compounds is why they are both called ‘word’) but do not always do so. For instance upside-down /'ʌpsaid-'daun/ is one grammatical word but many scholars would regard it as made up of two phonological words, each with its own primary stress. There is a full discussion of the criteria for and properties of grammatical word and phonological word in Dixon (2010b: 1–36). 3.2b A grammatical word is defined as follows for English:

•

It has as its base one or more lexical roots to which morphological processes (compounding, affixation, internal change, shift of stress, etc.) have applied. The resulting word has conventionalized coherence and meaning. The components always occur together, in a fixed order, with nothing intervening between them.

• •

Consider the phrase worthy of trust. A modifier can be applied to either (or both) of its component lexemes. We can add an adverb to the adjective worthy, giving absolutely worthy of trust, or an adjective to the noun trust, giving worthy of absolute trust. As an alternative to worthy of trust, one can use the adjective trust-worthy, where -worthy is a derivational suffix. Trust-worthy is a single grammatical word; it can be modified as a whole—absolutely ­trustworthy—but its parts may not be modified separately (one cannot say, for instance *absolute-trust-worthy or *trust-absolutely-worthy). The NP a clean, cool apartment includes two adjectives, Each can be ­modified—a clean, truly cool apartment or a truly clean, cool apartment (or a truly clean, truly cool apartment). Compare this with an ultra-clean apartment, where the derivational prefix ultra- creates the adjective ultra-clean. This is one grammatical word which can only be modified as a whole; one can say a truly ultra-clean apartment, but not *an ultra-truly-clean apartment. In just about every language, the morphemes within a word must occur in fixed order. Many languages permit considerable freedom for the order in which words occur in a phrase, or in a clause. English, however, has a pretty fixed word order. Just a little variation is permitted, enough to contrast with the fixed order of morphemes. For instance, both a clean, cool apartment or a cool, clean apartment are equally acceptable. In contrast, there is no alternative to an ultra-clean apartment (that is, one cannot say *a clean-ultra apartment).

3.2 how to tell it is an affix   49 Any criterion is likely to leak a little. Some affixes can be added to a complete phrase. For example, We appreciated her matter-of-fact-ness and He’s very public-school-ish, isn’t he? In each case, the derivational suffix forms one grammatical word with the last word of the phrase—/'matər əv='faktnəs/ and /'pʌblik 'sku:liʃ/. (Preposition of /əv/ is a proclitic to fact /fakt/, the clitic boundary being shown by ‘=’.) A root-plus-affix is always one grammatical word. The same criterion applies for compounds—each is one grammatical word. A leakage was illustrated, at the end of 3.1f, by sparsely occurring coordinations such as grand-mother and -father. Ordering constraints in English are rather strict, so that it is not easy to find examples where each of two orderings is acceptable. There is a degree of freedom in adverb placement—see 10.3—so that one can say either Truly John loves Mary or John truly loves Mary (but note that there is here a slight difference in pragmatic overtone). There can be several adjectival modifiers in an NP. If they belong to different semantic types, the unmarked order is pretty fixed (Dixon 1982: 24–6): value, dimension, physical property, speed, human propensity, age, colour (any deviation from this ordering has to be shown by a marked stress pattern). It is when we get two adjectives from the same semantic type that freedom of ordering emerges—for example, clean, cool or cool, clean (both from physical property), faithful, loyal or loyal, faithful (both from human propensity). But note that in such instances and is often inserted between the adjectives (still occurring in either order)—a faithful and loyal lieutenant or a loyal and faithful lieutenant.

3.2c The second necessary condition is that a root-plus-affix should constitute one phonological word. The main characteristic of a phonological word in English is that just one syllable bears primary stress. This is the normal situation for a derived stem. Primary stress is generally within the root, as in ante-natal /ˌanti-'neitl/, but sometimes on an affix, as in mono-syllable /'mɔnou-ˌsiləbl/. There may be a secondary stress—on the affix (especially if it is disyllabic), as in ante-natal, or on the root, if a prefix takes primary stress, as in mono-syllable. However, this is not criterial. In the great majority of cases, a derived word (root plus derivational affix) has one syllable with primary stress. In a minority of instances, there is also a syllable with secondary stress. In rather special circumstances, there may be two primary stresses in a root-plus-affix combination. When the prefix mega- (see 2.4g, 2.5m) is used in its scientific sense—as in mega-fauna /ˌmegə'-fɔ:nə/—or with a numeric value—as in mega-byte /ˌmega-'bait/—it behaves as a regular suffix. But on occasions when mega- is consciously employed, for special pragmatic

50    3 criteria—affixes, semi-affixes, and compounds effect, there is likely to be stress on both prefix and root; for example, megamachine /'megə'-mə'ʃi:n/. A further criterion is that phonological changes associated with the application of morphological processes always take place within a phonological word, not across word boundaries. An example, given in 2.5f, is final /f/ becoming voiced /v/ before -ous, as in mischiev-ous /'mistʃiv-əs/ from mischief /'mistʃif/. This condition is always satisfied for a stem-plus-derivational-affix assembly. Bloomfield (1933: 180) put forward what has become an oft-quoted criterion for a combination of roots being a compound; this concerned stress. He contrasted the syntactic sequence black bird, /'blak 'bə:d/, ‘any bird that is black’ (where each word has primary stress) with the compound blackbird, /'blakˌbə:d/, ‘a particular bird species, the common thrush, Turdus merula’ (which has only one primary stress, plus a secondary stress). (Note that this criterion, and the same example, were earlier given by Sweet (1891: 24−6).) In a compound consisting of two components, one bears primary and the other secondary stress. Whether primary stress falls on the first or on the second element varies a good deal. Some compounds are generally quoted with primary stress on the first component (blackbird and baby-sitter, for example), others with the stress on the second component (these include coldshoulder). But there is a great deal of variation (and controversy) concerning what dictionaries and grammar books say, and what speakers do. Like a root-plus-affix combination, a compound is generally both one grammatical word and one phonological word. The next two criteria for ‘what is an affix’ establish a clear distinction from compounds. Dictionaries and pronouncing dictionaries provide varied specifications for stress. For instance, some say that megafauna and megabyte have primary stress on the first syllable and secondary stress on the second or third, while others state the reverse. However, all maintain that there is just one primary stress in these words.

3.2d The third necessary condition for something being an affix is that it should have predictable semantic and/or syntactic effect with each stem to which it attaches. For instance, suffix -ness derives an abstract noun from an adjective. There is no need for a dictionary to include separate entries for sad-ness, lazi-ness, ugli-ness, or white-ness. It is sufficient for these derivations to be simply listed at the end of entries for sad, lazy, ugly, and white, with no meaning or word class stated since these are inferable from knowledge of the nature of suffix

3.2 how to tell it is an affix   51 -ness. Also for stupid-ity, modern-ity, and similar-ity, all involving suffix -ity, which has a similar effect to -ness. In contrast, the meanings of many compounds cannot be inferred from the meanings of the component roots and each of them must be accorded its own dictionary entry—hangdog, scapegoat, leapfrog, and so on. In fact, larger dictionaries do sometimes include entries for root-plus-derivational affix combinations. However, they are seldom consistent in doing so. For instance, in one dictionary I found enjoyment and excitement simply listed at the end of the enjoy and excite entries (with no meanings stated since these are predictable), but separate entries being provided for measurement and commencement (although these meanings are also predictable). A separate dictionary entry is especially likely when a prefix is involved, or when the affix has semantic effect (even when this is predictable) rather than just changing word class, as do -ness and -ity. A compound whose meaning is the sum of the meanings of its parts—henhouse, hairbrush, shipwreck, proofread, and many more—is almost always provided with an individual dictionary entry, although this is not strictly necessary.

3.2e The fourth and final necessary condition for a form to be recognized as a derivational affix is for there to be multiple occurrences of it. Sometimes there can be very many words including a certain affix. For example, suffix -ness is attested in more than 750 words in dictionaries, suffix -ish is in more than a hundred, verbalizer -en (as in black-en) is in about forty, suffix -ee (as in escap-ee) in more than twenty, and so on. The attentive reader may wish to draw attention to a form such as -ard| -art which in present-day English is found only in drunk-ard, lagg-ard, slugg-ard, and bragg-art. Are four occurrences sufficient to qualify -ard| -art as having the status of a derivational affix? Taken on its own, the answer is that this is a marginal case. But -ard| -art qualifies on the sufficient criterion that it is a bound form, only occurring in combination with free roots drunk, lag, slug, and brag. Looking now at compounds, a given root is likely to occur in only a few compounds, often only one. There are nouns pussycat and copycat and adjective/noun wildcat but no other common compound which includes cat as second element. The only compound with leap is leapfrog (frog also occurs as second element in bullfrog). And so on. There are some forms which occur in a considerable number of what are generally regarded as compounds, notably -maker and -man (and, to a lesser extent, -woman). In terms of the criteria employed here, there is a strong case for analysing these as derivational suffixes. See 3.3e−f.

52    3 criteria—affixes, semi-affixes, and compounds 3.2f Moving on now to sufficient conditions for something to be a derivational affix, the first of these is productivity. A derivational affix is productive if it is available for making new words. Suppose, for instance, that a new adjective should enter English: verital ‘always tells the absolute truth (no matter that it might cause offence)’. Speakers will, without demur, talk of un-verital ‘lacking the verital quality’ (as in He’s the most un-verital person I’ve ever met, telling lies seems to be his vocation) and they will create abstract noun verital-ness (for example, Mark my words, his absolute verital-ness is going to stand in the way of promotion). This would confirm the productive nature of prefix un-1 and suffix -ness. If a form is productive, this is a sufficient property for it to be considered an affix (and it will, in almost all cases, be supported by other properties). It is not, however, a necessary condition. There are many well-established derivational affixes, each occurring in dozens of words, which are no longer productive (save in a joking or affected literary manner). Some are of Romance origin; these include negative prefix in-| im-| il-| ir(as in in-sane, ir-religious, etc.), adjective-creating suffixes -ous (courage-ous, danger-ous) and -al (  fiction-al, univers-al ). Some are Germanic, productive in an earlier stage of the language but no longer so; for example adjectiveforming suffixes -en (as in wood-en, ash-en) and -ful. In contrast to -ful, Germanic suffix -less is freely used to make new words, as in My grandfather lives an email-less existence and How can you maintain that facebook-less-ness is a good thing, you’re missing so much. There are a fair number of adjective pairs where the positive member ends in -ful and the negative one in -less—for example, joy-ful| joy-less, shame-ful| shameless, merci-ful|  merc-iless, taste-ful|  taste-less. We also have scent-less but, at present, no scent-ful. One could dream about -ful once again becoming productive, with scent-ful being coined. And why stop there—what about defence-ful, track-ful, and child-ful, to stand alongside defence-less, trackless, and child-less? This is all fanciful. Although -ful features in about a hundred words, it is not now productive. It should be noted that there is no correlation between productivity, on the one hand, and broadness of meaning or frequency of use, on the other. The Romance-origin suffix -esque ‘having the style of, in a pleasing manner’ is shown on only half-a-dozen words in Muthmann’s (1999: 104) reverse dictionary; they include pictur-esque and statu-esque. But it is fully productive, as in the recent coinage Pinter-esque ‘in the (esteemed) style of playwright Harold Pinter’.

3.2 how to tell it is an affix   53 3.2g When two morphemes are combined together, this can just be a matter of addition—when each retains its basic form—or it may involve fusion— when a new word is created whose phonological form was not predictable from the forms of its components. This was discussed and illustrated in 2.5, especially 2.5f. Compounding invariably involves addition. Morphological derivation is sometimes just addition—from bewilder /bi'wildə/ and -ment /-mənt/ is formed bewilder-ment /bi'wildə-mənt/. But some instances of derivation involve fusion—solve /sɔlv/ plus -tion /-ʃən/ produces solu-tion /sə'lu:-ʃən/. If adding a form to a stem engenders fusion, then this is a sufficient (but not a necessary) reason for the form being regarded as a derivational affix. 3.2h A compound is a rather fixed affair; it has two components which belong together, each requiring the other. Contrast this with a derivational affix being added to a root or stem. There is here more fluidity—some derivational affixes can be added to a multi-word constituent (which may involve coordination). Alongside non-drinker we can have non-[beer-drinking] man, besides exchairman there is ex-[member of the board ], parallel to beard-ed one can say [[ pale and ashen]-face]-ed. Other derivational affixes which may attach to NPs include -ness and -ish (illustrated in 3.2b) plus anti-, counter-, -able, -er, -ism, -nik, -like, -ly, -y, -free, -less, -proof, -worthy, -maker, -style, and -fashion. Having scope over a multi-word constituent is a sufficient (but far from a necessary) criterion for recognizing affixal status. 3.2j The definition of a compound is that it involves two elements, each of which is a free root (or stem). That is, each may make up a whole word by itself. We can say leapfrog, or just leap, or just frog. In contrast, most derivational affixes cannot occur alone. In depend-ant, avoid-ance, mal-practice, and circum-navigate, the elements -ant, -ance, mal-, and circum- are bound forms. That is, they may not make up a word on their own. Not being able to function as a free form is the fourth sufficient condition for having the status of an affix. This is not, however, a necessary condition. There are a score or so derivational affixes which have the same (or very similar) form to an adjective or noun or preposition. Sometimes affix and free form have the same orthographic form but differ phonologically; other times they are also phonologically matching. An important consideration is, of course, whether they have parallel meanings. All this is discussed in 3.3.

54    3 criteria—affixes, semi-affixes, and compounds 3.2k It will be useful to summarize the conditions for a certain word being a stem-plus affix combination. necessary conditions (which should be satisfied): N1  One grammatical word N2  One phonological word N3  Predictable semantic and/or syntactic effect N4  Multiple occurrences sufficient conditions (if one of these is satisfied we do have an affix, but not all affixes satisfy these conditions): S1 Productive S2  Involves fusion S3  Can apply to multi-word constructions S4  Does not occur as a free form Note that N1 and N2 also apply for compounds, but the remaining criteria do not. Some affixes can occur as free forms in special circumstances. Suppose someone says, of John: This week, he’s anti-unions, anti-monarchy and anti-Christmas. A response might be: Oh, whatever you mention, he’s always anti, whereas Mary is always pro. A word may be abbreviated just to its prefix, as in She favours a maxi (i.e. maxilength dress); also semi (-detached house), vice (-captain), homo (-sexual). One sense of suffix -ism derives a noun referring to an ideological movement; for example, romantic-ism, femin-ism, evangelical-ism. They have been collectively referred to as ‘(-)isms’. Similarly, ‘(-)ologies’ is used as a cover term for nouns ending in -ology, referring to study of a branch of knowledge (psychology, numer-ology, philology, and many more). Indeed, there have been a number of books with ‘(-)ism’ or ‘(-)ology’, or both, in the title. They include -Isms and -ologies: All the movements, ideologies and doctrines that have shaped our world (Goldwag 2008) and Isms and Ologies: A guide to unorthodox and non-Christian beliefs (Kellett 1965).

3.3 Affixes and semi-affixes 3.3a A number of what appear to be derivational affixes each have the same— or very similar—form as a free lexeme. Some are generally regarded as affixes. The term ‘semi-affix’ has been used for others. And there is a further set which have generally been denied the status of affix, and simply regarded as parts of compounds. It will be useful to survey these systematically.

3.3 affixes and semi-affixes   55 3.3b We can first examine five adjective-deriving suffixes, and free form adjectives from which they derive historically. Each row of Table 3.1 provides comparative examples. The question to be posed is: should the words in the left-hand column be regarded as compounds—noun statesman plus adjective like producing compound statesman-like, and so on? We can reject this suggestion, and confirm -like, -free, -worthy, -proof, and -ful as derivational suffixes on a number of grounds. First, there are multiple occurrences of each, and all except -ful are productive. Secondly, each of these suffixes has a predictable semantic and/or syntactic effect. The corresponding entries in Table 3.1 were chosen with some care. For -like, -free, and -worthy, suffix and adjective do have similar semantic scope. For the others, there is only a degree of overlap. A child-proof gate is one that a child could not open (it is proof against a child’s being able to open it) whereas an idiot-proof camera is so simple that anyone of low intelligence would have no difficulty in using it. Quite a few derived forms with -ful may not be paraphrased with adjective full; they include need-ful, use-ful, and right-ful. Another useful criterion concerns the applicability of prefix un-1. Simple adjectives and many derived adjectives may be negated with un-1 (for example, un-happy, un-cooperative) but this prefix is only very seldom used with compounds—one says non-easy-going in preference to

Table 3.1  Derivational affixes and cognate free forms ADJECTIVE FORMED WITH DERIVATIONAL SUFFIX

SIMILAR MEANING USING FREE FORM WHICH IS COGNATE WITH DERIVATIONAL SUFFIX

(a)

in a statesman-like manner /'steistmən-laik/

in a manner like a statesman /laik ə='steistmən/

(b)

sugar-free products /'ʃugə-fri:/

products free of sugar /fri: əv='ʃugə/

(c)

a trust-worthy employee /'trʌstˌ-wə:ði/

an employee worthy of trust /'wə:ði əv=trʌst/

(d)

a weather-proof paint /'weðə-pru:f/

a paint that is proof against bad weather /pru:f ə'geinst bad 'weðə/

(e)

the reunion was joy-ful /'dʒɔi-fəl/

the reunion was full of joy /ful əv=dʒɔi/

56    3 criteria—affixes, semi-affixes, and compounds *un-easy-going. The ready acceptability of un-statesman-like, un-trustworthy, and un-faith-ful, for instance, helps confirm the affixal status of -like, -worthy, and -ful. Four of the suffixes in Table 3.1 have the same orthographic form as the corresponding free lexeme; only -ful and full differ. And those in rows (a−d) have the same phonological form. Alongside suffix -less (as in childless, speechless) there is the periphrastic comparative marker less (which is genetically unrelated). We can perceive some similarity of meaning (‘without’ and ‘of a smaller extent than’) but these two forms cannot be linked in the manner of Table 3.1. We also have two forms extra. The prefix extra- indicates ‘outside of’ (extra-marital, extra-terrestrial), a quite different meaning from that of free form extra ‘an additional quantity of’ (as in Give John an extra helping of pudding!). Note that both -less and extra- are productive. In summary, there are strong reasons for asserting the affixal status of the five suffixes in Table 3.1, plus -less and extra-. Marchand (1969) regards -ful, -less, and extra- as derivational affixes. He categorizes -like and -worthy as ‘semi-suffixes’, described as ‘elements which stand midway between full words and affixes’ (but with no criteria provided). There is no mention of -free or -proof. Jespersen (1942) does not mention -worthy or -free, but considers the other five to be suffixes.

3.3c We can now look at words whose first element has the same form as a preposition. Consider (this is just a sample, many more could be added in each row): under-clothes over-coat out-post in-shore up-stairs down-town off-shore after-noon

under-arm over-board out-house in-breeding up-beat down-pipe off-shoot after-birth

under-graduate over-achiever out-grow in-grown up-grade down-wind off-colour after-life

under-value over-estimate out-source in-flight up-coming down-load off-load after-thought

There is a tradition, among grammarians of English, for regarding all of these as compounds. That is, of denying the status of derivational prefixes to under-, over-, out-, in-, up-, down-, off-, and after-. These do, of course

3.3 affixes and semi-affixes   57 have the same form as prepositions. But -free, -worthy, -proof, -ful, -less, and extra- have the same form as adjectives, and they are considered acceptable as affixes. I do include them in the inventory of derivational prefixes. They are productive and each has a common semantic core. All are, of course, of Germanic origin. A number of them pattern closely with a corresponding prefix of Romance origin. Let us compare a small sample of words commencing with Romance prefix sub- and with Germanic prefix under-: sub-marine sub-soil sub-celestial sub-normal sub-standard

under-water under-floor under-world under-done under-weight

Surely under- has as strong a claim to be considered a derivational affix (rather than a part of innumerable compounds) as does sub-. Similarly for the other seven prefixes which have the same form as prepositions. In the list just given, sub- is used with Romance and under- with Germanic forms. But each prefix is fully productive and has come to be used with stems of different genetic hue. For example, sub- is appended to Germanic roots in sub-way and sub-let, while under- is used with Romance roots in undernourished, under-paid, and under-secretary. The two prefixes have parts of their meaning in common but do differ in a number of respects. Indeed, they may be used with the same root: sub-class is ‘a division within a larger class’, while under-class is ‘people of low social status’ (these two derivations relate to different senses of the noun class). There is further discussion and exemplification in 6.4.1. 3.3d There is another series of affix-like forms—those with numeric values such as many-, half-, one-, two-, and so on, plus all. Once again it is instructive to contrast with a corresponding Romance prefix: semi-conscious semi-detached semi-circle semi-nude semi-annual

half-awake half-timbered half-moon half-clothed half-yearly

58    3 criteria—affixes, semi-affixes, and compounds These prefixes have the same meaning; semi-, a Romance form, is predominantly used with Romance stems, while the Germanic form, half-, is employed almost exclusively with Germanic stems. The words in the bottom row are synonymous—semi-annual (Romance prefix with Romance stem) and half-yearly (Germanic prefix plus Germanic stem), both meaning ‘twice a year’. Surely this indicates that half- should be regarded as a bona fide derivational prefix, on a par with semi-. Greek prefix mono-, Romance prefix uni-, and Germanic one- were discussed and illustrated in 2.6g. Once again, forms tend to go together on genetic grounds. Compare uni-directional with one-way, uni-linear with onesided. Should not this qualify one-, like half-, to be included in the inventory of affixes? With higher numbers, evidence is sparser, mainly because the forms are less common. We find multi-lateral and many-sided, bi-lateral and two-sided, plus a few more. Of course, people may say things like fourteen-sided, for which there is no usable Romance equivalent. Half- and one- certainly satisfy the criteria to be considered prefixes, but a number like eight doesn’t. It is hard to know where to stop. Compoundpropounders would seize on this dilemma as reason for treating the whole caboodle as compounds. It is certainly something to think about. 3.3e There is a handful of other recurrent forms which we can attempt to rescue from the anonymity of compound-dom. English has more than a dozen affixes which create an agentive noun. Suffix -er (as in pack-er and preach-er) is highly productive and cannot occur as a free form. Suffix -ster (as in song-ster, gang-ster) occurs in just a few words and has limited productivity. But it too cannot occur as a free form and grammarians uniformly accept -ster as a derivational suffix. In contrast, -monger (as in cheese-monger, scandal-monger) and -wright (for example, wheel-wright, play-wright) tend to be given short shrift. Marchand (1969: 357–8) includes these as semi-suffixes (‘midway between full words and suffixes’) but other sources do not mention them. Monger and wright do exist as free forms, albeit of low frequency. But, as I have argued (on the basis of -free, -worthy, -proof, -ful, -less, and extra-), this should not be taken as a reason for excluding them from the roster of derivational affixes. A similar case can be made for recognizing -maker and -smith as derivational suffixes (as far as I am aware, they have thus far never been accorded this status). We find many words ending in -maker. Besides peace-maker, trouble-maker, merry-maker, mischief-maker, law-maker, hay-maker, and

3.3 affixes and semi-affixes   59 holiday-maker, the name of any piece of clothing, furniture, or equipment may have -maker added to it—dress-maker, shoe-maker, cabinet-maker, and so on. Suppose that some new object should be unveiled; say a folding totem pole. Imagine that it is given the name fotole; the people who produce it will at once be referred to as fotole-makers. Like some other derivational affixes, -maker can be added to a coordination. For example, He’s a [watch- and clock-]maker. This happens only very exceptionally with compounds (grand-mother and -father was mentioned in 3.1f  ). Like maker, the form smith functions as a free form, ‘a worker in metal’. And, like -maker, -smith shows many of the properties of a derivational suffix. It is used with the names of metals (tin-smith, copper-smith) and objects made from metal (lock-smith, gun-smith). Recently we have had word-smith ‘an expert in the use of words’. Also like -maker, -smith can be attached to a coordination: She’s a [gold- and silver-]smith. 3.3f Finally, there is the fascinating problem of how to deal with man /man/, when used with a modifier indicating a role or profession. There are three possibilities here. (1) There is a syntactic construction—an NP with man as its head, preceded by a modifier. For example, insurance man, customs man, removals man. In these combinations, man retains its canonical form /man/. (2) Man occurs as second element of a word, retains the form /man/, and takes secondary stress (this is typical of the second element in a compound). Dictionaries provide about twenty instances of this, including: caveman /'keivˌman/ chessman /'tʃesˌman/ snowman /'snouˌman/ mailman /'meilˌman/ con-man /'kɔnˌman/ superman /'syu:pəˌman/ (3) The same as (2), except that man here has a reduced vowel, /mən/ and bears no stress (this is typical of a monosyllabic suffix). There are well over a hundred of these, including: sea-man /'si:-mən/ police-man /'pli:s-mən/ post-man /'poust-mən/

French-man /'frentʃ-mən/ mad-man /'mad-mən/ horse-man /'hɔ:s-mən/

clergy-man /'klə:dʒi-mən/ spokes-man /'spouks-mən/ trades-man /'treidz-mən/

60    3 criteria—affixes, semi-affixes, and compounds An interesting point is that, to form the plural of words in sets (1) and (2), one replaces -man /-man/ by -men /-men/. However, the plural for words in set (3) does involve an orthographic change—from sea-man to sea-men, and so on—but the phonemic form remains the same. That is, both sea-man and sea-men are pronounced /'si:-mən/. The words in sets (2) and (3) are always regarded as compounds. But -man in items of set (3) has many of the characteristics of a derivational suffix—it has a predictable meaning, it has a fair degree of productivity, it lacks secondary stress, and—with vowel reduction from /a/ in the free form to /ə/ in the suffix—it involves a type of fusion. But if -man /-mən/ is to be considered a suffix for set (3), what about -man /man/ in set (2), where there is secondary stress and no vowel reduction? These are best considered as compounds. It is natural to enquire whether -woman behaves in the same way as -man. The free noun is /wumən/, and its vowels do not change when combined with other roots. That is, the vowel difference between sets (2) and (3) for -man does not exist for -woman. However, there is the same stress difference as with man. Woman bears secondary stress in words of type (2)—for example, superwoman /'syu:pəˌwumən/—but not in those of type (3)—such as spokeswoman /'spouks-wumən/. As for man, forms of type (2) with woman are best regarded as compounds, in contrast to those of type (3) where -woman has the status of a derivational suffix. Interestingly, -woman is substitutable for -man more readily in words from set (3) than in those from set (2). One is rather likely to hear police-woman, mad-woman, and French-woman, but much less likely to encounter snowwoman, con-woman, or cavewoman. 3.3g Forms examined in the last sections have become progressively more problematic. There are strong reasons for the five suffixes in 3.3b (all homonymous with adjectives) and the eight prefixes in 3.3c (all homonymous with prepositions) to be treated as derivational affixes on a par with -ify, un-, and -ness. Maybe Marchand’s term ‘semi-affix’ should be redeployed for the remainder: half-, one-, many-, two-, all-, -monger, -wright, -maker, -smith, -man, and -woman. They are certainly a sort of affix. Most of the time, in chapters which follow, they are simply called ‘affixes’.

3.3 affixes and semi-affixes   61 There is further discussion of -monger, -wright, -maker, -smith, -man, and -woman in 9.3.13.

Before we embark on examination of individual derivational affixes, in Chapters 5–10, there is in Chapter 4 a brief discussion of what is included and what omitted, plus a description of ‘adaptations’ (things like swimathon and dancathon, based on marathon).

4 What we describe

4.1 Affixes which fall outside our scope 62

4.2 Adaptations 65 4.3 Plan of the following chapters  69

The following chapters provide profiles of around two hundred derivational affixes. Many of these are productive, and can be used to make new words in present-day English; for example, -ize, and -able. Some are no longer productive, although they were in an earlier stage of the language and—as a consequence—occur in a fair number of words, in which they are added to free forms; for example, -ful and -most. Others are of foreign origin and were never productive, but do occur in many words, in some of which they are added to a free form; these include -ous and -acy. Section 4.3 explains the organization of the remainder of this book, the syntactic and semantic principles according to which affixes are arranged into groups, so that each is discussed in conjunction with others which have similar meaning and function. Before that, Section 4.1 explains why a number of types of affix are ­excluded from our study. And Section 4.2 provides brief discussion of ‘adaptations’ (for instance, walkathon as an adaptation of marathon) which are an interesting example of word creation that is not quite the same thing as morphological derivation.

4.1 Affixes which fall outside our scope A number of affixes are marginal, in different ways. Some are only used in scientific terminology, a number occur only—or almost only—with bound

4.1 affixes which fall outside our scope   63 roots, and there are affixes which originally had wide use but are found in only one or two words in the modern language. 4.1a There is a profusion of affixes, based on Latin and Greek elements, which are used productively, but only in scientific naming terminology. We can mention three examples: (1) -acea /-eiʃə/, based on an ending in post-Classical Latin, is used in the names of classes of animals, including crustacea (crabs, lobsters, shrimps, and suchlike, having a hard shell), and cetacea (large marine animals such as whales and dolphins). (2) -on /-ɔn/, based on an ending in classical Greek, is used in physics for the names of subatomic particles (such as proton and neutron) and certain basic units (photon), and in chemistry for the names of certain gases and hydrocarbons (radon, freon). (3) -ol /-ɔl/, originally from Latin, is used in the names of some oils and oil-derived chemical compounds (such as benzol ) and related commercial products (Dettol, Lysol ). The ending occurs in alcohol and also in the names of particular alcohols (cholesterol, thymol ). It is also used in the names of hormones and other steroids (including cortisol and mestronol ). There are scores of such scientific affixes; with only a few exceptions, they are not used with free forms. All are listed—together with meanings and rules for their use—in handbooks of nomenclature for organic chemistry, inorganic chemistry, botany, zoology, geology, physics, and the like. Because of their specialist nature, and predominant occurrence just with bound forms, it would not be appropriate to include them in this volume. 4.1b In 2.5k there was discussion of at, id, and our (as in diplomat, splendid, and splendour), which occur only with bound roots of Romance origin, and are thus not to be treated as derivational suffixes. (Although we did mention that other linguists might employ more generous criteria.) There are other endings which occur almost exclusively with bound forms. These include: (1) In Romance languages, suffix -tude, /-tyu:d/, derives abstract nouns from adjectives. There are thirty or so English words ending in tude,

64    4 what we describe nearly all onto bound forms; for example, latitude, solitude, gratitude, attitude, fortitude. The only instances of tude being added to free forms are ineptitude and the rather rare quietude. (2) Noun-creating suffix -ison, /-isn/, was productive in Latin and Old French. This ending features in just a handful of Modern English words, including garrison, jettison, and venison. Only in comparison could it be said to be added to a free form. Neither ison nor tude can be regarded as a derivational suffix in English. 4.1c A few suffixes were productive in earlier stages of English, but most of the words they created have fallen into disuse so that really one cannot recognize them as suffixes in the modern language. We mention two: (1) The OE suffix -meal, /-mi:l/, meant ‘(a measure) taken one at a time’. There were a score or more words in OE and ME where -meal was added to a free noun, including footmeal ‘step by step’, cupmeal ‘a cupful at a time’, and lumpmeal ‘in lumps’. The only one that has survived is piecemeal, which nowadays means ‘done bit by bit, often not in a satisfactory manner’. Since meal occurs on a free form in only one word, it cannot really be recognized an a derivational suffix in present-day English. (2) The Germanic suffix -ock, /-ək/, occurred, with a diminutive meaning, in quite a few OE words. Several of these are still in use, but the form to which they are added is not a free form—for example, paddock, hassock, tussock, mattock. Exceptions are hillock ‘small hill or mound’, and bullock ‘castrated bull’. We do thus have two modern-day forms with ock added to a free noun, but the meaning of ock is different in the two cases, so that it should not today be recognized as a derivational suffix. 4.1d Quite a number of forms were considered for inclusion in this book, but excluded since they do not satisfy the criteria set out in 3.2. Those commencing a word (putative prefixes) include, quoting one sample word for each: ana (anabranch) cis (cisalpine) citra (citramontane)

crypto (cryptogram) macro (macropod) e (email) preter (pretergress) juxta (juxtapose) syn (synchronic)

4.2 adaptations   65 Those ending a word (putative suffixes) include: aroo/eroo (buckaroo) acious (fallacious) ane (urbane) archy (monarchy) een (colleen) gram (telegram) ice (jaundice)

icle (planticle) ismus (snobismus) oid (hominoid) ola (payola) ose (verbose) plex (complex) red (kindred)

right (upright) sake (namesake) scope (microscope) ton (singleton) trix (creatrix) up (booze-up) wort (liverwort)

Other linguists might well have different opinions concerning the admissibility of these forms as derivational affixes.

4.2 Adaptations A rather special process of word-formation came into play around the middle of the nineteenth century and burgeoned during the twentieth century. It involves making a new word by analogic adaptation from an existing one. 4.2a Landscape was borrowed about 1600 from landschap, a technical term in Dutch painting. New words have been created through replacing the initial element, land, with some other noun. An early adaptation (perhaps the first) is quoted by the OED, a letter from Thackeray dated 28 December 1856: A fairyland of frozen land, river and city-scape. The word landscape has since been the model for many analogic adaptations; the scape indicating ‘extended view of’. We find seascape, moonscape, cloudscape, mountainscape, sandscape, mindscape, and more (see Marchand 1969: 211). Table 4.1 provides a representative selection of further model words and three sample analogical adaptations based on each. 4.2b Cafeteria, ‘self-service restaurant’, was a loan from Spanish (around 1900), and has given rise to scores of names for self-service establishments. Kidnap (plus kidnapper, kidnapping) was coined in the late seventeenth century, as a compound of kid ‘child’ and verb nap, a variant of nab ‘catch or seize’; new verbs ending in nap mean ‘take illegally’. 4.2c Note that the first parts of model words landscape, cafeteria, and kidnap are all free nouns. This does not apply to the initial parts of model words

66    4 what we describe Table 4.1  Some analogic adaptations and the words they are modelled on MODEL WORD

ANALOGIC ADAPTATIONS

(1)

cafeteria

shoeteria, washateria, gasateria

(2)

kidnap

dognap, carnap, busnap

(3)

marathon

swimathon, dancathon, telethon

(4)

cavalcade

camelcade, autocade, tractorcade

(5)

hamburger

beefburger, cheeseburger, fishburger

(6)

bootlegger

booklegger, votelegger, meatlegger

(7)

cappuccino

mugaccino, cupaccino, babyccino

(8)

alcoholic

chocaholic, golfaholic, workaholic

(9)

Watergate

Hollywoodgate, Irangate, Monicagate

in rows (3) and (4), marathon (from Greek), and cavalcade (from Romance languages). Analogic adaptations ending in (a)thon refer to some activity extended over an unusually long period of time, and those ending in cade to a procession. 4.2d Hamburger, in row (5), is a shortening of hamburger steak, a patty of ground beef in a bread roll. The original form had morphological structure hamburg-er (named after the German city of Hamburg) but it has been ­reanalysed as ham-burger (although the meat used is generally beef rather than ham) and then -burger was analogically extended to other varieties of fillings within a roll. (The clipping burger would generally be understood as an abbreviation for the model word hamburger, not for any of the adaptations.) 4.2e A bootlegger is someone who makes or sells liquor illegally, so called because he used to hide a bottle in his knee-high boots. The second element, legger, now has the general meaning ‘selling something illegally’, for example (obscene) books, votes in an election, rationed foodstuffs. 4.2f Cappuccino, a recent loan from Italian, has become such a popular ingroup drink that it continues to spawn adaptations, some related to the vessel it is served in, and others to the type of user—babyccino is a very weak drink, but still with the distinctive froth on top, so that children can feel part of the party.

4.2 adaptations   67 4.2g The model word alcoholic is unusual in that it is an adjective derived from noun alcohol by adding suffix -ic; there are many analogic adaptations (meaning ‘addicted to’). It is interesting that whereas the model word ends in oholic, adaptations invariably use aholic. However, the difference in orthographic vowels—o and a—is not significant since each is pronounced as a schwa, /ə/; thus alcoholic /ˌalkə'hɔlik/, chocaholic /ˌtʃɔkə'hɔlik/. (Similar remarks apply for the alternation of e and a in cafeteria and washateria, from row (1) of Table 4.1; each is pronounced as /ə/.) 4.2h Watergate is the name of a building in Washington DC, where ­Richard Nixon’s nefarious break-in took place in 1972. The ending -gate was then adopted to identify any scandal. Hollywoodgate, in 1976, described the ­alleged forgery of a cheque and sparked off an investigation into corruption in the film industry. The Irangate scandal, in 1986, involved Ronald Reagan selling arms to Iran and using the proceeds to fund anti-Marxist Contra guerrillas in Nicaragua. When, in the late 1990s, Bill Clinton came close to losing his presidency over misdemeanours relating to Monica Lewinsky, the obvious label was Monicagate. 4.2j In the ‘model word’ column of Table 4.1, cafe, kid, and boot are free forms, as is ham (although only through innovative reanalysis); mar, caval, capp, and alc do not occur as free forms. All of the replacements for these initial elements, in the analogic adaptations, are free forms, and they are all nouns. In every instance, the adaptations belong to the same word class as the model word. This is a noun for most rows, but kidnap is a verb (with nominal form kidnapper). Alongside noun bootlegger we have bootleg, which is both adjective and noun (this carries over into the adaptations). Watergate is a proper name, and the adaptations substitute a proper name for the initial Water. 4.2k It might be asked why the second elements in these adaptations could not be regarded as derivational suffixes. The reason is that they do not ­behave like derivations, and they do not feel to native speakers like derivations. A derivational affix like -ize or -ness or inter- is simply added to a free form which has an appropriate meaning; there is no model word in the background. In contrast, a speaker using or creating an adaptation is always aware of the model word. Suppose there is a scandal in Brooklyn and a journalist refers to it as Brooklyngate. Ask them why, and the reply will be

68    4 what we describe ‘by analogy with Watergate’. They won’t say ‘by analogy with Monicagate’ since it is Watergate, not Monicagate, which is the model. Similarly for, say roofscape—‘by analogy to landscape’—or prawnburger—‘because it’s a sort of hamburger’ (not ‘because it is like a fishburger’). 4.2m Another significant characteristic is that, unlike affixes, the second part of an analogic adaptation seldom has its main vowel reduced, and sometimes bears primary stress following the stress profile of the model word. For example chocaholic /ˌtʃɔkə'hɔlik/, camelcade /ˌkaməl'keid/, washateria /ˌwaʃə'teriə/. As a rule, the adaptation has the same phonological form as the model— that is, the same number of syllables. For example the forms used to replace kid in kidnap are invariably monosyllabic, like kid. We hear dognap but something like alsatiannap is rather unlikely. Forms which appear in adaptations from cavalcade are invariably disyllabic, like caval. We get camelcade and autocade but would be unlikely to hear horsecade or carcade (there could be car-o-cade, adding an extra syllable to fit the phonological pattern). 4.2n It is perfectly possible that in the fullness of time some of these second elements of adaptations may develop into bona fide derivational affixes, with the original model word being forgotten about. But this has not happened yet. (Indeed, some of what are now established derivational affixes of foreign origin may well have developed in this way.) Other words used as models for analogic adaptations in recent times include arboretum, Argonaut, auditorium, automobile, broadcast, earthquake, escalator, fantastic, frankfurter, hijack, hippodrome, inflation, parachute (see 6.6.1), panorama, and telegram. (There are many examples in Marchand (1969: 211–14) and Mencken (1936: 176–80, 1945: 350–67), among other sources.) In a note at the end of 2.1, the notion ‘combining form’ was mentioned; it was apparently introduced by the OED. However, dictionaries vary greatly as to what is a ‘combining form’ and what is an ‘affix’. Indeed, the OED itself is inconsistent on the matter. This can be illustrated by its treatment of the final element of analogic adaptations. For instance, -scape, -nap, -burger, -legger, and -gate are listed as combining forms, while -teria, -cade, and -aholic are considered to be suffixes. Adaptations based on marathon involve athon after a first element which ends in a consonant (for example, talkathon) or thon after a vowel (moviethon). The OED describes -athon as ‘a combining form, barbarously extracted from marathon’, but states that -thon is a suffix, adding that it is a ‘variant of -athon, combining form’! Extensive use of analogic adaptations is a particular feature of present-day English. Most languages lack this feature, while others have it to a rather limited degree—see Aikhenvald (2007: 48–9).

4.3 plan of the following chapters   69 4.2p Adaptations should not be confused with blends, where a part of one word is combined with a part of another word (not the whole of the second word, as happens with adaptations). Examples include smog, referring to a mixture of smoke and fog; brunch, for a meal taken part-way between the normal times of breakfast and lunch, and eaten instead of both regular meals; Lewis Carroll’s chortle, describing a sound that is a combination of a chuckle and a snort; and Oxbridge, a term linking together Oxford and Cambridge Universities (in contrast to less ancient institutions).

4.3 Plan of the following chapters The following chapters describe, in turn, each of the two hundred or so forms recognized here as derivational affixes. The discussion is organized in terms of the forms, functions, and meanings of affixes. Prefixes are the focus of Chapters 5 and 6; in most cases they do not change word class (or subclass of a noun) but simply add a semantic element. Chapter 5 discusses the thirteen prefixes which create negative words. Most of these do not affect word class, but some senses of de-, dis-, and un-2 do serve to derive verbs from nouns and adjectives, and anti- can derive an adjective from a noun. There are about seventy-five further prefixes which —either exclusively or predominantly—do not change word class, and these are dealt with in Chapter 6. They are organized in terms of semantic fields—number, magnitude and degree, locationals, temporals, roles, repetition, and ‘not quite right’ (pseudo- and quasi-). At the end of this chapter, Section 6.11 deals with the half-dozen suffixes which do not change word class, each having a diminutive or endearing meaning (for example, puss-ikins). Chapter 7 surveys how verbs are created out of nouns and adjectives. There are basically just six verbalizing affixes, two prefixes (be- and en-| em-) and four suffixes (-en, -ate, -ize, and -(i)fy). The two last mentioned are highly productive and most of the chapter is devoted to examination of their phonological conditioning and meanings (including, how they are applied to proper names, as in American-ize and Israel-ify). In Chapter 8 there is detailed investigation of the more than thirty suffixes and one prefix (a-, as in a-shore) which derive adjectives. These are organized in terms of the types of bases they may be added to—common nouns (and a few adjectives), proper names, and verbs. (There is a degree of overlap.)

70    4 what we describe There are almost no examples of a verb being derived from a verb of a different type, and few of an adjective from an adjective of a different type. However, examples abound of one kind of noun being derived from another type. There are also, of course, processes for deriving nouns from verbs and adjectives. Chapter 9 discusses the more than fifty suffixes which derive new nouns. First, state nouns based on adjectives (such as kind-ness), followed by agentive nouns and the like derived from verbs and nouns (drumm-er, pian-ist). Then nouns which are derived from verbs to describe states, activities, ­results, locations, and so on (arrange-ment, assist-ance). This is followed by an account of suffixes which create abstract nouns from concrete nouns (mother-hood, friend-ship). The topic of Chapter 10 is the making of adverbs. There are half-a-dozen minor suffixes. For the major one, -ly, we study the types of adjectives it can attach to, and the functional possibilities of the derived adverbs. Finally, Chapter 11 surveys the ways in which several suffixes and/or several prefixes may be combined in a single word. Throughout Chapters 7, 9, and 10, there is discussion of adjectives and nouns doing double duty as verbs, verbs and adjectives doing double duty as nouns, and adjectives also functioning as adverbs. Particular attention is paid to looking at differences in meaning and function between derivational affixes which are semantically rather similar. For example, contrasting mann-ish with man-ly, sing-er with song-ster, domination with domin-ance.

5 Un-expected, mis-judged, dis-organized counter-revolution: Making negative words 5.1 Overview 71

5.14 (7) un-2 104

5.2 (1) un-1 76

5.15 Contrasting un-1 and un-2 107

5.3 (2) in-|  im-|  il-|  ir- 79

5.16 The use of un-2, dis-,

5.4 Competition between un-1 and in- 80 5.5 Which adjectives take un-1 and which take in-? 87

and de- 107 5.17 (8) mis- 108 5.18 Contrasting un-1 and mis- 110 5.19 Contrasting dis- and mis- 111

5.6 (3) non- 91

5.20 (9) mal- 112

5.7 Contrasting un-1 and in-

5.21 Contrasting mis- and mal- 113

with non- 93

5.22 (10) ex- 113

5.8 (4) a- ~ an- 95

5.23 Contrasting non- and ex- 114

5.9 Contrasting non- and a(n)- 96

5.24 (11) anti- 114

5.10 (5) de- 96

5.25 (12) counter- 115

5.11 (6) dis- 98

5.26 Contrasting anti- and

5.12 Contrasting un-1|  in- and dis- 101 5.13 Contrasting non- and dis- 104

counter- 116 5.27 (13) contra- 117

5.1 Overview 5.1a English has a wide range of negative prefixes. Indeed, it is not at all difficult to devise a short text in which every lexeme is in negative form: Inevitably, unavoidably, the unfortunate ex-bishop, now a nonchristian, must be disfrocked (or defrocked or unfrocked) since he both misbelieves that the antichrist is invincible and doesn’t

72    5 making negative words disbelieve in the unfettered impiety of unbelievers. He is undoubtedly disliked because he is so incontrovertibly unlikeable. It will thus be his misfortune to become indistinguishable from other undistinguished malcontent unchristian misfits. What is fascinating is that a single root may be used with two or more different negative prefixes. In the text just given we encounter seven lexical roots each used with two or three of the prefixes in-, un-, mis-, non-, dis-, and de-, as set out in Table 5.1 (the first line has two, etymologically cognate, forms of the root). The text includes further examples of three of these prefixes: invincible, impiety, incontrovertibly; undoubtedly, unfettered; and misfits. There are also instances of mal-, ex-, and anti-, in malcontent, ex-bishop, and antichrist. Not in the text are a(n)-, as in asexual, counter-, as in counter-attack, and contra-, as in contra-distinction. Table 5.1  Alternative negated forms of lexemes in the sample text inevitably

unavoidably unfortunate

misfortune

unchristian

non-christian

unfrock unbeliever unlikeable indistinguishable

disfrock misbelieve

defrock

disbelieve disliked

undistinguished

5.1b Of the basic list of thirteen negative prefixes, only three are of Germanic (Gmc) origin, and two have identical form today. There is un-1, used with adjectives to indicate the absence of a quality which is likely to be valued, as in un-clean and un-true; this comes from un- in OE. And there is un-2, used with transitive verbs typically to indicate the reverse of an action, as in un-tie; this reflects and- ~ und- in OE. (In the text, unfrock and unfettered use un-2 while the remaining un- derivations employ un-1.) Prefix mis- is also Germanic and has merged with mes- from French. The minor prefixes a(n)and anti- are of Greek (Gk) origin, with the eight remaining forms coming from Romance (Rom) languages—directly from Latin, or from Latin via French, or from French. 5.1c Only in-| im-| il-| ir- has a fusional profile, showing consonantal agreement with the initial segment of the stem to which it is attached. Negative

5.1 overview   73 prefix (4) in Table 5.2 has form an- before a vowel and a- before a consonant. Nine of the prefixes apply to a single word, whereas non-, ex-, anti-, and counter- can have scope over several words. Ex, anti, and counter also function, in limited circumstances, as free forms. Nine of the negative prefixes do not change word class. Each of de-, dis-, and un-2 can be added to a noun and form a transitive verb (for example: dis-frock, de-frock, and un-frock in our short text). Another sense of de- also involves addition to a noun, this time creating an intransitive verb (as in: The marines de-plane-d). Added to a noun, anti- can derive an adjective. Six suffixes each create a considerable number of negative words: un-1, in-| im-| il-| ir-, non-, dis-, mis-, and un-2. Of these, un-1, non-, and un-2 are highly productive, while mis- has a fair degree of productivity. Although the prefix in-| im-| il-| ir- occurs in a very large number of words today it is seldom used to derive new ones; dis- is also unlikely to take part in new formations. Four of the negative prefixes, with a narrower range of occurrence, do each carry the potential of being used in new combinations: de-, ex-, anti-, and counter-; prefixes a(n)- and mal- are really only productive in medical and scientific word coinings, while contra- (despite the British form contra-flow, from 1934) appears to have little productivity. Table 5.2 summarizes the origin, meaning, and scope of each prefix. In the second column, ‘PP’ indicates highly productive, ‘P’ mildly productive, ‘Sc’ productive mostly in scientific/medical terms, and ‘—’ scarcely productive. 5.1d There are a couple of other, very minor, negative prefixes. We find igjust in ig-noble ‘not noble’, taken via French from Latin ignoˉbilis. (It also commences unanalysable roots ignore and ignominy.) This could perhaps be regarded as a further variant of the in-| im-| il-| ir- prefix, here used before n to avoid an in-n sequence. (We do find in- before some words commencing with n-, such as in-numerate, but none are very common.) Noble is the adjective with the highest frequency to be negated in this manner. Then there is prefix ab-, which goes back to Latin preposition ab ‘from, away’. This commences a fair number of unanalysable roots (such as abscond, abhor, abstain), and ab-normal appears to be the only word in common use which has ab- as a prefix. Ab-normal may originally have meant just ‘not normal’ but has recently garnered negative overtones (deviating from the normal in an undesirable way); as a consequence, a-normal (employing negative prefix (4), a(n)-) is sometimes used to indicate ‘not normal’ with a neutral sense.

ORIGIN

FORM

MEANING

MAIN SCOPE

With many adjectives (and a few nouns); and lacking a (valued) quality, e.g. un-happy, im-proper with adverbs and nouns derived from them

(1)

PP Gmc

un-1

(2)

—

in|  -im-|  il-|  ir-

(3)

PP Rom

non-

not a member of a specified class, e.g. nonalcoholic, non-smoker

With many adjectives and nouns; and with complex constituents

(4)

Sc

Gk

a(n)-

not relating to a parameter, e.g. a-political, a-sexual

With a few adjectives and nouns

(i) get off a vehicle, e.g. de-bus

With a few nouns, deriving intransitive verbs

(5)

P

Rom

de-

(ii) deprive of, e.g. de-mast, de-frock, de-acid-ify, de-central-ize

With nouns (and a few adjectives), deriving transitive verbs; and with a few verbs

(i) deprive of, e.g. dis-mast, dis-frock

With a few nouns, deriving transitive verbs

(ii) lacking a (valued) quality, e.g. dis-honest, disadvantage, dis-like, dis-approval (of), dis-satisfi-ed

With some adjectives and nouns; with many transitive verbs, and with nouns and adjectives derived from them

(iii) reverse state or process, e.g. dis-connect, dis-embark-ation

With many transitive verbs; and with nouns derived from them

(i) deprive of, e.g. un-mast, un-frock

With a few nouns, deriving transitive verbs

(ii) reverse process or state, e.g. un-load, un-tie, un-veil

With many transitive verbs; and adjectives derived from them

(6)

(7)

Rom

— Rom

PP Gmc

dis-

un-2

74    5 making negative words

Table 5.2  Summary properties of negative prefixes

Table 5.2  (Continued) ORIGIN

FORM

MEANING

MAIN SCOPE

P

Gmc (& Rom) mis-

do wrongly, improperly, e.g. mis-pronounce, mis-believe

Mainly with transitive verbs; and with some nouns and adjectives, mostly derived from verbs

(9)

Sc

Rom

mal-

defective, inadequate, e.g. mal-practice

Primarily with activity nouns, also some adjectives and verbs

(10) P

Rom

ex-

former, e.g. ex-wife, ex-president

With some nouns and NPs

(11) P

Gk

anti-

alternative to, reverse of, opposed to, used against, With nouns, adjectives, and NPs e.g. anti-hero, anti-semitic

(12) P

Rom

counter-

do/be the opposite, e.g. counter-revolution

With some verbs, nouns and adjectives, and NPs

(13) —

Rom

contra-

opposed to, e.g. contra-indication

With a few verbs, nouns, and adjectives

5.1 overview   75

(8)

76    5 making negative words The Germanic prefix be- derives verbs from nouns and adjectives and is discussed in 7.3.1. It has a wide range of meaning, from the negative sense ‘deprive of’, as in be-head, to what is virtually the opposite meaning ‘adorn with’ as in be-jewel. The prefix arch- has a variety of senses; for example an arch-bishop is a superior rank of bishop. In one sense it is attached to a term describing someone of objectionable character, serving to emphasize their opprobriousness. For example, an arch-criminal is a greater menace to society than a plain criminal. We have arch-enemy, arch-terrorist, and arch-dolt, among others. The main discussion of arch- is in 6.8.1. 5.1e A couple of adjective-deriving suffixes (whose main discussion is in Chapter 8) have negative meaning: -free, as in a smoke-free environment, and -less as in a child-less couple. A number of adjectives derived from a noun or verb by means of suffix -ful form their negative through replacing the -ful by -less. For example: noun verb

power harm

adjectives power-ful power-less harm-ful harm-less

There are just a few examples of un-1 being used together with -ful. For example, un-merci-ful has a very similar meaning to merci-less. But there is a clear difference between un-law-ful and law-less: an un-law-ful action is something which the law does not allow, while a law-less person is someone who is always breaking the law. 5.1f We now discuss and contrast the thirteen negative prefixes, taking them in the order presented in Table 5.2.

5.2 (1) un-1 5.2a The original un- /ʌn-/ affix, written here as un-1, is the pre-eminent negative prefix in present-day English; it will be seen that un-1 occurs in six rows of Table 5.1, contrasting with in-, mis-, non- and dis-. This un- was very common in OE, but in fact rather few of the OE forms have come through into ME and Modern English; those which do include un-clean, un-fair, unripe, and un-wise. From the late fourteenth century on, un-1 has been used extensively to form new negative adjectives, in four ways:

5.2 (1) un-1   77 (i) Added to a simple adjective root, either Germanic as un-happy, un-fit, or Romance as un-equal, un-fair. (ii) Added to an adjective which is derived from a noun through one of a variety of suffixes, as un-luck-y, un-skil-ful, un-war-like, un-sister-ly, unnatur-al, un-sea-worthy, un-fortun-ate, and un-christ-ian. (iii) Added to an adjective which is derived from a verb, again through one of a number of suffixes, as un-impress-ive, un-break-able, un-like-able, un-satisf-actory, and un-tir-ing. (iv) Added to a noun. There are rather few of these, some involve a simple noun, as un-ease, and others a derived form, as un-believ-er, un-tru-th. There are many nonce forms. For example, in Lewis Carroll’s Through the looking glass, Humpty Dumpty explains to Alice that there are many more days on which you might get an un-birthday present than that for a birthday present. 5.2b Interestingly, there are about two hundred instances where the negative derived adjective, with un-1, came into the language a century or two before the corresponding positive adjective, without un-1. Un-bear-able was first noted in 1449 but bear-able only from 1550, un-sink-able in 1655 but sink-able not until 1865, un-deserv-ed in 1374 but deserv-ed only in 1552, and un-limit-ed in 1445 followed by limit-ed in 1551 (and see 8.4.7b). Note that some of these items are based on Germanic roots—such as bear and sink—and others on Romance roots—including deserve and limit. (These are quoted from Marchand 1966: 140 which includes many further examples.) Indeed, for a number of negative derived adjectives there is still no corresponding positive adjective; for instance, un-relent-ing, un-see-n, un-better-able, and (used of a ladies’ stocking) un-ladder-able. Some un-1 derivations involve verb plus preposition, with there frequently not being a positive correspondent; for example, un-hear-d-of, un-thought-of, un-cared-for, un-call-ed-for. There are many adverbs and nouns derived from adjectives in un-1— adverbs such as un-happi-ly, un-luck-i-ly, un-satisf-actori-ly, and nominalizations such as un-happi-ness, un-natural-ness, un-sea-worthy-ness, un-tru-th, untru-th-ful-ness, and un-amiab-ility. There seem to be no established verbs involving un-1. Jespersen (1942: 465) mentions that the noun un-employ-ment might appear to be based on a putative verb *unemploy but ‘really stands for *un-employ-ed-ment’, based on derived noun un-employ-ed.

78    5 making negative words 5.2c In some of the more common words with un-1, the prefix bears no stress; for example, un-happy /ʌn'hapi/, un-kind /ʌn'kaind/, and un-luck-y /ʌn'lʌki/. But in the majority of derivations, there is secondary stress on un-1 in addition to primary stress on the form which follows; for example un-fit /ˌʌn-'fit/, un-wise /ˌʌn-'waiz/ and un-satisf-actory /ˌʌn-ˌsatis'f-aktəri/. The fact that it generally does bear stress shows that un-1 is here regarded as a semantic operator. Indeed, in times past un-1 could be used to negate a complex NP; the OED quotes, from 1707, eggs, unbroken or crack'd, meaning ‘eggs, unbroken or uncracked’; but this would scarcely be acceptable today. 5.2d The meaning of un-1 with an adjective is that the referent of the noun which is modified by the adjective lacks the quality referred to by the adjective. Note that this does not have exactly the same import as the positive adjective negated by not. If it is not the case that someone is happy then it must be the case that they are not happy. But they may not necessarily be unhappy. This can be illustrated in: (1)

not happy happy un-happy

Happy and un-happy relate to opposite ends of a parameter of ‘happiness’. The scope of not happy is complementary to that of happy. But one can say: (2)  John isn’t happy and he isn’t un-happy either This relates to the blank space between happy and un-happy in the bottom row of (1). Un-1 is typically used with adjectives referring to qualities which are valued, the lack of such a quality being regarded as undesirable. (A rare exception to this generalization in un-objection-able.) The semantic import of the Germanic prefix un-1 is very similar to that of the Romance prefix in-|   im-|  il-|  ir-. Their meanings, and the lexical roots and stems with which each occurs, are discussed in detail below. Old-style telegrams were charged by the word with various tricks being used to create complex words in order to save money. The productiveness of

5.3 (2) in -/ im -/ il -/ ir -   79 un-1 is shown in the following from Evelyn Waugh (1936: 161): ‘Cables were soon arriving . . . Require earliest name life story photograph American nurse upblown . . . We replied: Nurse unupblown.’

5.3 (2) in-|  im-|  il-|  ir5.3a The prefix in-|  im-|  il-|  ir-, /in-/ ~ /im-/ ~ /i-/ (henceforth referred to by its basic form in-), was highly productive in Latin and Old French. It is assimilated to the initial consonant of the following form, becoming im- before b, p, or m and—in the orthography—ir- before r and il- before l. An in- form sometimes has secondary stress on the prefix. It is only likely to attract primary stress in exceptional circumstances, when the negative quality is being emphasized, as in a conversation such as: – You did say that John is modest, didn’t you? – No, I said that he is absolutely ím-módest /'im 'mɔdist/ Whereas in the normal pronunciation of immodest, /i'mɔdist/, there is a single /m/ (although written as mm to show that underlying in- is attached to a root commencing with m), in the emphatic /'im 'mɔdist/ the /m/ will be pronounced twice, indicating the speaker’s feeling that this is a prefix ending in /m/ attached to a root commencing with /m/. There are just a few in- words with main stress on the first syllable, often also showing reduction of the root vowels; for example: potent finite

/'poutənt/ im-potent /'im-pətənt/ /'fainait/ in-finite /'in-finit/

However, for these exceptional cases, both root and its in- derivative were taken over from French, together with their pronunciations and meanings. 5.3b From the thirteenth century on, there are many examples of both an adjective and its negative derivation being borrowed from French. For example: correct, in-correct regular, ir-regular

possible, im-possible legal, il-legal

80    5 making negative words These were then analysed within English, with recognition of prefix in- (with all its conditioned variants). By about 1450, in- had become a productive affix within English, being added to many adjectives of Romance origin. For example, the following had been borrowed from French: accurate

pecunious

replaceable

legible

Negative derivations were then formed on these within English by prefixing in-, giving: in-accurate im-pecunious

ir-replaceable

il-legible

These new formations followed exactly the same principle of assimilation as was applied in Latin or French: /im-/ before a bilabial consonant, and— orthographically—ir- before r and il- before l (although phonologically, the prefix is just /i/- before /r/ or /l/). There are just a few nouns which take in-; for example, in-digestion and in-experience. 5.3c It should be noted that there is also a locational prefix with the form in(and variant im- before a bilabial)—essentially a development from preposition in—as in in-coming, in-grown, im-migrate. This is discussed in 6.6.5b. Jespersen (1942: 471) mentions that there can be confusion between the two prefixes—in-habit-able may be taken to be a negation (‘not habitable’) or to relate to location (‘habitable in’). (Further examples are provided by Jespersen.)

5.4 Competition between un-1 and in5.4a From about 1450 on, new negative adjectives were being formed through the productive use of two prefixes, Germanic un-1 and Romance in-. Prefix un-1 was attached to adjectives of every sort, whether of Germanic or Romance origin; it never involved assimilation. In contrast, inwas added almost exclusively to Romance loans, and always replicated the assimilation the prefix exhibits within Romance languages. It appears that

5.4 competition between un-1 and in-   81 a speaker of English about 1500 must have had a sense of what was a loan, since it was only on these that in- derivations were formed. 5.4b Between about 1500 and 1700, there were scores of adjectives, almost all of Romance origin, for which two negative forms were in use, one with inand the other with un-1. (In contrast, non-loans had only un-1 negations.) Over the centuries, the situation was gradually rationalized; for most adjectives, just one negative form persisted with the other dropping out of use. We find (see also Jespersen 1933: 140–2): (i) Un-1 form continues into the present-day language, in- form ceases to be used. For most of these, the un-1 derivation was earlier, with an in- form being created later and used for a limited period. For example (quoting, from the OED, date of first mention for the persisting form, and first and last dates for those no longer in service): modern forms un-certain (c.1300) un-charit-able (1456) un-comfort-able (1592) un-pleas-ing (1480) un-utter-able (1586)

erstwhile competitors *in-certain (1491–1741) *in-charit-able (1496–1679) *in-comfort-able (1574–1655) *im-pleas-ing (1602–1613) *in-utter-able (1603–1859)

For a few, the in- form had been borrowed from Romance languages (rather than created in English); nevertheless, it was replaced by an un-1 derivation: un-imagin-able (1611) un-fortun-ate (1530)

*in-imagin-able (1533–1698) *in-fortun-ate (1390–1682)

One of the reasons for the abandonment of in-imagin-able may have been a preference not to have prefix in- attached to a form commencing with in- or im-. For example, we also say un-inhabit-able rather than *in-inhabit-able. However, there is the form in-imi-table which is now preferred to un-imitable (un-imit-able was first recorded in 1581, is noted by the OED as very common in the seventeenth century, but then fell out of use, the last quotation being from 1773).

82    5 making negative words (ii) In- form continues until the present day, un-1 form dropped out of general use (but may be retained in some regional dialects). For most of these, both root and in- negation were borrowed from Latin and/or French. They include: modern forms in-constant (1402) in-cur-able (1340) in-numer-able (1340) in-capable (c.1600) im-possible (c.1300)

erstwhile competitors *un-constant (1480–1757) *un-cur-able (1382–1775) *un-numer-able (c.1400–1611) *un-capable (1587–1806) *un-possible (1362–1866)

It is interesting that, in the fifteenth and sixteenth centuries, un-possible was much more common than im-possible but survives today only in some regional dialects. Under this heading come just a few Romance roots whose in- negation was created within English, including: in-active (1725) im-proper (1531)

*un-active (1591–1777) *un-proper (c.1380–1773)

(iii) There is a further set where both un-1 and in- forms are still in use, with essentially the same meaning. Generally, one form is preferred, although the other is also in currency:

•

un-1 form is today preferred over the in- form (which was created within English); for example:

un-controll-able (1577)

is preferred to

in-controll-able (1599)

•

in- form (which was created within English) is today preferred over the un-1 form, for example: in-advis-able (1870)

is preferred to

un-advis-able (1673)

•

in- form (which was borrowed entire from French, rather than being formed within English) is today preferred to the un-1 form; for example:

ir-redeem-able (1609) in-support-able (1530)

is preferred to

un-redeem-able (1584) un-support-able (1586)

5.4 competition between un-1 and in-   83 There are a fair number of further forms of all types. 5.4c Most often, when two prefixes with very similar meaning both survive with the same stem, the two derivations have developed a meaning difference. This applies for a number of in- and un-1 pairs. For example: in-art-ist-ic (1859) un-art-isti-c (1854)

‘not good art’ (could be said of a picture) ‘not interested in art’ (could be said of a person)

in-ed-ible (1822)

‘cannot be eaten’ (general statement about something, e.g. a type of mushroom) ‘not fit to be eaten’ (particular statement, could ­describe a badly-cooked meal)

un-eat-able (1611) in-evit-able (1430)

‘bound to happen, nothing could stop it happening’ un-avoid-able (1577) ‘will happen in these circumstances, but if some other course of action had earlier been followed, it might not have had to happen’ in-ept (1603) un-apt (1374)

‘foolish, lacking skill’ ‘not suitable in the circumstances’

There is also in-apt (1744), a more recent coining, which can mean either ‘foolish, lacking skill’ (like in-ept) or ‘not suitable in the circumstances’ (like un-apt). 5.4d There are a fair number of examples where two negative adjectives are formed from a given Romance borrowing, utilizing different prefixes. Most often, in- occurs in association with a Romance-origin suffix (such as -able| -ible)—or no suffix at all—and un-1 with a Germanic suffix (such as -ed or -ing). For example: in-fidelity (1509) in-digest-ible (1528) in-animate (1546) in-determin-ate (1391) in-cessant (1532) il-limit-able (1596) in-discriminate (1649) in-distinguish-able (1658)

un-faith-ful (1384) un-digest-ed (1528) un-animat-ed (1697) un-determin-ed (1442) un-ceas-ing (1382) un-limit-ed (1445) un-discriminat-ing (c.1800) un-distinguish-ed (1598)

84    5 making negative words Of these, infidelity, indigestible, inanimate, indeterminate, and incessant were borrowed in toto from Romance, while il-limitable, in-discriminate, and indistinguish-able were created in English from loan words limit, discriminate, and distinguish. It is instructive to observe the historical progression of some groups of adjectival formatives: digestible (c1386)

LOAN FROM ROMANCE

digest (1398)

LOAN FROM ROMANCE

indigestible (1528)

LOAN FROM ROMANCE digest-ed

PART PARTICIPLE Un-1 NEGATIVE FORM

un-digest-ed (1528)

Un-1 NEGATIVE FORM

un-digestible (1611–13)

From Latin adjective gratus, English borrowed an original adjective grate ‘pleasing’ (now obsolete), and from French it took gratitude and ingratitude. Hence we get, in present-day English: gratitude (1500)

LOAN FROM ROMANCE LOAN FROM ROMANCE DERIVATION IN ENGLISH Un-1 NEGATIVE FORM LOAN FROM ROMANCE

grate (1523) grate-ful (1552) un-grate-ful (1553) ingratitude (1568)

In these diagrams, a solid line indicates a derivation within English, and a broken line a cognate link between loans. 5.4e Romance prefix in- has similar meaning and scope to Germanic prefix un-1. Both describe the lack of a certain quality, the presence of which is regarded as desirable. As with un-1 forms, adverbs may be created from in- adjectives by the addition of -ly; for example, in-describab-ly, in-correctly, il-legal-ly, im-proper-ly. And there are many nouns which are effectively derived from an in- adjective through nominalizing suffixes of Romance

5.4 competition between un-1 and in -   85 origin, such as -ity, -ance| -ence, and -acy. Actually, in many cases both adjective and noun are in toto borrowings; these include: improper, impropriety irreverent, irreverence intolerant, intolerance

impatient, impatience imprudent, imprudence

For other nominals, there has been a degree of further derivation within English. This may happen in two ways: both adjective and noun borrowed from romance

negative forms created in english with prefix in-

active, activity significant, significance

in-active, in-activity in-significant, in-significance

positive and negative adjectives borrowed from romance

nominal forms created in english with suffix -acy

literate, illiterate

liter-acy, illiter-acy

In another instance there are three forms which are direct borrowings: adjectives flexible and inflexible and noun flexibility. The negative noun inflexibil-ity was then created within English. The most common nominalizing suffix on adjectives in English is -ness. This is used sparingly with non-native forms. Adjectives complete and incomplete were loans, with nominalizations complete-ness and incompleteness being created through adding a Germanic suffix to a Romance form. There are a number of interesting pairs with in- and un-1: adjective un-able un-equal un-just un-stable un-quiet

noun in-abil-ity in-equal-ity in-justice in-stabil-ity in-quiet-ude

Marchand (1969: 170) explains this by saying that ‘un- is much weaker’ with nouns than with adjectives. Jespersen (1942: 468) states: ‘pretty often un- is preferred before the shorter word, and in- before the longer word derived from it, which is generally also of a more learned nature’. But this is surely a

86    5 making negative words further example of the tendency, illustrated above, to employ prefix in- when there is a suffix of Romance origin. 5.4f There are a number of apparent in- derivations which do not have the expected meaning of the negative of the positive root. These include infamous (also showing unusual phonology), which means ‘notorious’ rather than ‘not famous’, and indifferent, which means ‘unconcerned with’ rather than ‘not different’. But all four forms—famous, infamous, different, and indifferent—are Romance loans; they carried with them from French both meaning and pronunciation. There are other negative adjectives which—after being introduced into English—developed a meaning that is apparently independent of the prevailing sense of the underlying positive adjective. (I) Pertinent means ‘appropriate or relevant to the matter under consideration’. Impertinent originally meant ‘not appropriate or relevant’ but then developed a more specialized sense ‘not appropriate or relevant to some social situation’; and from this developed the basic meaning today ‘impolite, disrespectful’. (II) The major meaning of adjective material is ‘of physical substance’. Immaterial originally indicated ‘not of physical substance’ but a further sense developed (from 1698) ‘of no consequence or importance’ and it is this that predominates today. Dr Johnson (1755) counselled: ‘This sense has crept into the conversation and writings of barbarians, but ought to be utterly rejected’. However, as the OED points out, it is simply the negation of a second sense of material, from 1475, meaning (to quote Johnson’s own definition for this sense) ‘important, momentous, essential’ (for example, in as for the more material faults of writing . . . from Dryden). We also find that the meaning of in-valu-able differs from ‘not valuable’. In fact, -able forms adjectives both from nouns (such as profit-able) and from verbs (for example, forget-able). It is likely that valu-able is based on the noun and in-valu-able on the verb. That is: noun value; one sense (c1340) ‘having significant worth or importance’ adjective valu-able (1589) ‘characterized by significant worth or importance’ verb value, one sense (1412) ‘estimate the worth of’ adjective in-valu-able (1576) ‘cannot have its worth estimated’

5.5 which adjectives take un-1 and which take in -    87 As mentioned before, un-1 is highly productive. But although there are hundreds of words formed with in-, the prefix appears to be no longer in general use to create new words. Marchand (1969: 170) dates the last new formations to the middle of the nineteenth century; they include irresponsive in 1846 and insubordinate in 1849. 5.5 Which adjectives take un-1 and which take in-? 5.5a Negative prefixes un-1 and in- have almost exactly the same meaning and conditions of use. They apply to adjectives (one sense of dis- serves as an alternative with just a few adjectives). But it is not the case that all adjectives accept a negative suffix. What is the basis for eligibility? We can first examine the behaviour of simple monomorphemic adjectives (including a few very common derivations such as luck-y, dirt-y, and health-y). There are three factors involved in determining whether they may be negated (by un-1 or in- or dis-). (i) As discussed in 2.9 (and see the Appendix), every word class can be divided into a number of ‘semantic types’ of lexemes. Each semantic type has a common meaning element and shared morphological and syntactic properties. First we need to consider each semantic type associated with the adjective class. (ii) And then we need to examine the semantic organization within each type, especially the kinds of semantic opposition involved. The following are particularly important:

•

antonym pairs—such as large and small—which are, effectively, names for the opposite poles of a parameter. They are used with respect to some implicit norm of comparison. For example, one can quite felicitously say The small elephant is large—the creature is small with respect to the elephant norm but large with respect to some wider norm (all animals? all movable objects?). Antonyms are gradable and typically occur in comparative constructions; if it is true that X is smaller than Y, then it follows that Y is larger than X. sets of complementary terms. Here, if one term is chosen it excludes the others. Suppose that a language has seven colour terms; knowing that some object is describable as ‘red’, we know that none of the other six terms are applicable to it.

•

88    5 making negative words (iii) We also need to examine the social value accorded to the quality described by each lexeme. 5.5b In essence, negative prefixes are not applied to adjectives which are part of an antonym pair or complement set, nor are they generally used in reference to a quality which has negative social overtones. In view of all this, adjectives from certain semantic types are likely to form negative derivatives and adjectives from other types not to do so. We can deal with the semantic types one at a time (expanding on the discussion in Dixon 1977: 31–5, 1982: 15–20). dimension. Most are in antonym pairs, and do not take negative prefixes. For example, big| little, large| small, wide| narrow, deep| shallow, long| short, and tall| short. There are also adjectives referring to shape such as square (also a noun) and round. Although these are not in antonym pairs, they do not form negatives. It seems that un-1 and in- (and dis-) do not apply to any dimension adjective. age. Again there are antonym pairs—new| old and young| old—with no negative forms. Ancient and modern are not quite antonyms; one can say un-modern, in relation to something which is not in accord with present-day ideas or fashions. colour. A complement set with no negative derivations. difficulty. The main terms—simple, complex, easy, hard, difficult, tough— are close to forming antonym pairs. If it is the case that X is easier than Y, then it is likely that Y is more difficult than X. The negative form un-easy does not relate to the primary sense of easy, ‘can be achieved without effort’, but to the human-propensity-type sense ‘relaxed in mind’ (as in I feel easy about it). There is also an adjective un-simple, but of marginal status. physical property. Again, basically antonym pairs which have no negative form—hard| soft, strong| weak, heavy|  light, wet| dry, hot| cold, fresh| stale, cheap| expensive. Dead and alive may form either an antonym pair or a complement set. We do find un-cool. However, this relates not to the sense of cool which is the opposite of warm (‘of quite low temperature’) but to the recently introduced meaning, ‘trendy, with-it’ (this is in effect a value sense).

5.5 which adjectives take un-1 and which take in -    89 There is a set of terms referring to corporeal well-being which do not exactly group into antonym pairs; they include healthy, well, sick, ill. Negatives are formed from those terms describing a quality which is valued: un-healthy and un-well. (Note that the homonymous adverb well does not take un-; it is related to the human propensity adverb good.) Beautiful and ugly are almost antonyms. The positive term beautiful is attested with a negative prefix, but un-beautiful is not of common occurrence. The adjective ripe has no antonymic opposite and so forms the common negative form un-ripe. Clean and dirty—in their central senses of referring to the presence or absence of dirt—do constitute an antonym pair. There is an adjective un-clean but this primarily relates to a secondary sense of the adjective ‘morally (or ritually) impure, does not behave or think in a socially acceptable manner’. (Since the form un-clean exists, it is also used with the meaning ‘dirty’; but it is unlikely that the un-1 derivation would have evolved if the adjective had been restricted to the ‘dirt-free’ sense.) value. Good and bad are antonyms, having no negative forms. For fear of causing offence, people sometimes prefer not to describe something as bad, saying instead that it is not good or not too good—but not *un-good, or even *non-good. Other adjectives from this semantic type which have positive overtones do form negatives—un-lovely, im-perfect, un-necessary, un-important, un-lucky. Odd, strange, and curious do not describe qualities which are valued, and— as would be expected—do not occur with a negative prefix. qualification. Since there are no antonyms or complements, and none of these adjectives have negative overtones, most do have negative forms: in-definite, im-probable, un-obvious, im-possible, un-usual, un-common, unlikely, un-certain, un-sure, in-correct, in-appropriate, and ab-normal (see 5.1d). There is today no commonly used negative based on sensible, which is a little surprising. (There is a word insensible ‘unconscious, numb’, which has a quite different meaning.) Right and wrong form a complement set—if something is not right it must be wrong; thus they do not form negatives. True and false are also—in their central meanings—complements, yet there is a derivation un-true. In fact this has basic reference not to the primary meaning of true ‘in accordance with facts/reality’ but to the secondary sense ‘loyal, sincere’ (as in He is a true

90    5 making negative words friend ). The primary meaning of un-true is ‘not loyal or sincere’. But since the word un-true exists, it is also used as a synonym for false (just as un-clean can be used to mean ‘dirty’). And there is a difference of degree—saying that a statement is ‘untrue’ sounds a little less damning than referring to it as ‘false’. similarity. Alongside similar, like, equal, consistent, and analogous there are dis-similar, un-like, un-equal, in-consistent, and (of lesser occurrence) unanalogous. These terms carry neither positive nor negative overtones. This semantic type also includes identical, which lacks a regular negative derivation. (Different and indifferent were mentioned in 5.4f.) human propensity. By and large, these adjectives do not enter into semantic oppositions. For example, sad is not the antonym of happy; if it is the case that X is happier than Y, this does not necessarily imply that Y is sadder than X. Some items from this semantic type describe a quality that is valued by society, and these form negatives—un-happy, un-ready, un-willing, un-kind, un-generous, dis-honest, un-occupied, and ignoble (see 5.1d). We would not expect to encounter negative derivations from others, describing a quality with negative connotations, such as angry, sad, mad, sorry, cruel, or lazy; there are no negative versions of these in general circulation today. A number of human propensity adjectives do refer to a valued quality but do not have generally-used negative forms in the present-day language (although they did in times past); they include fond, keen, glad, eager, and frank. There is no un-clever (although cleverness is surely considered a good thing) but maybe because clever and stupid are in antonymic opposition (if X is cleverer than Y, then Y is stupider than X). There are negative derivations from human propensity adjectives afraid and ashamed. Do these qualities have positive overtones? Yes—in a frightful situation it is sensible to be afraid and to behave accordingly (foolhardy to be unafraid and behave rashly). In similar fashion, adjective-deriving suffix -less generally attaches only to a noun referring to something considered valuable; it does attach to fear, since in a fearful situation it is appropriate to show fear (and courting danger to be fearless)—see 8.2.31. And after having performed some shameful act, it is right for the perpetrator to feel ashamed (bad to be un-ashamed).

5.6 (3) non-   91 5.5c Turning now to more complex adjectives, derived from verbs, we find that if the verb’s meaning has a mental component, then once again a negative prefix is only likely to be used when there is a positive connotation. For example, we find un-like-able, un-love-able, and un-enjoy-able, but not *unhate-able; there is un-impress-ive but not *un-weari-some, and un-pleas-ing but not *un-annoy-ing. For adjectives derived from verbs which just refer to an activity, a negative prefix may be used whether or not the result is desirable; for example, in-decis-ive, un-break-able, un-broke-n, un-spil-t, un-aided, un-oppos-ed, un-winn-able, and un-los-able. For adjectives derived from nouns, it seems that a negative prefix may be freely used if a speaker considers it appropriate to refer to the absence of a certain property, whether or not the property has positive overtones. We get (illustrating some of the major suffixes deriving adjectives from nouns): unprison-like and un-wind-y, alongside un-styl-ish, un-fashion-able, un-friendly, un-president-ial, and un-beard-ed. 5.6 (3) non5.6a The form non was a negative particle in Latin and French. It developed into a prefix in Law Latin (for example, non-payment) and from the fourteenth century on has been a robust formative in English, non- /nɔn-/. It is added to simple and complex adjectives and nouns (but not to verbs). The meaning is ‘not be a member of a certain class of people or things’, the class being signified by the adjective or noun to which non- is appended. Thus, a non-threatening stance belongs to the class of stances which are not threatening, a non-stop flight belongs to the class of flights which do not have any stops en route, a non-conformist is a member of the class of people who do not conform, and so on. Unlike un-1 and in-, the canonical use of non- carries no overtone of evaluation; simply a statement of class membership. The two prefixes are compared in 5.7. Except when it has been fully incorporated as part of a root—as in nonentity /nɔ'nentiti/—non- bears secondary stress, even in forms with a long history; for example non-payment /'nɔn'peimənt/. Indeed, it could be said to function as a separate phonological word within the grammatical word (having some of the characteristics of a proclitic). Allied to this, it is highly productive, being able to be used with a wide range of nouns and adjectives in an appropriate discourse context.

92    5 making negative words 5.6b The typical occurrences of non- can be listed. (I) with adjectives (a) With some simple adjectives, almost all of Romance origin; they include non-natural, non-standard, non-active, non-violent. (b) With adjectives derived from nouns (and some nouns doing double duty as adjectives); the nouns can be Romance or Germanic: – from nouns referring to substances, such as non-metal(-lic), nonwood(-en), non-gas-eous, non-fat, non-alcohol-ic (we might also include here non-fact-ual ); – from nouns referring to types of creature: non-human, non-heathen, non-adult; – from nouns referring to types of organization (or belief group): non-party, non-union, non-Moslem; – plus a few more based on proper names, as in non-Euclid-ian geometry. (c) With adjectives derived from verbs (either Romance or Germanic). Some use bare verbs, such as non-skid tyre, non-stick pan. Others are derived by one of the standard suffixes (-able| -ible, -ive, -ent, -((a)t)ion, -ing, and -ed ). Examples include non-break-able plate, non-effect-ive enrolment, nonexist-ent files, non-regulat-ion dress, non-bind-ing treaty, non-aspirat-ed consonant, and non-align-ed nation. (II) with nouns (which almost all derive from verbs, mostly of Romance origin) (a) referring to an agent, as non-smoker, non-student, non-believer, nonsubscriber, non-resident; (b) referring to a class of people, as non-kin, non-citizen, non-specialist; (c) referring to a speech act or artistic work; for example, non-sentence, non-answer, non-picture, non-fiction; (d) referring to an activity, attitude, or result. These are derived from verbs through a variety of suffixes—non-resist-ance, non-accept-ance, nonconform-ity, non-deliver-y, non-align-ment, non-admi-ssion. (III) with complex constituents. A company which does not make a profit is a non-[ profit-making] company, where non- is added to a derived adjective involving transitive verb and object noun. Similar examples are:

5.7 contrasting un-1 and in - with non -   93

non-[load-bearing] patio non-[office-holding] citizen

non-[gold-producing] strata non-[milk-drinking] pussycat

There is also the noun non-[slave-holder]. A non-[slave-grown] commodity is a commodity which slaves do not grow, the constituent to which non- is added involving subject plus transitive verb. And a non-[habit-forming] drug is a drug the use of which does not develop into a habit. Non- can be added to an NP as in non-[executive-committee-member], and his non-[understanding-of-my-dialect] irked me. The possibilities here are wide. One could easily imagine a dialogue which included the echoic his non-[the-ends-justifies-the-means] stance or her non-[holier-than-thou] attitude, in which non- is added to an familiar phrase (which had previously been included in the discourse without non-). As mentioned earlier, if a class of people or objects is delineated within a discourse, then a non- form can be used to describe people or things outside the class. Suppose that the entries for a poetry competition are placed by the judges in four piles, labelled ‘very good’, ‘good’, ‘poor’, and ‘awful’. One judge may suggest to the others: For a start, let’s dispose of all the non-[verygood] ones. 5.6c Thus, the potential productivity of non- is immense, greater even than that of un-1. Grammatically, it functions more like a syntactic modifier than like a morphological element. And this is reflected in its distinctive phonological character, always bearing (secondary) stress. There is an excellent account of the history of, and range of uses for, non- in Algeo (1971).

5.7 Contrasting un-1 and in- with non5.7a The use of non- is basically equivalent to that of not plus copula verb be. The class of things which are non-X is complementary with those which are X—if something ‘is not X’, then it is ‘is non-X’. This is quite different from prefixes un-1, in- (and dis-). Diagram (1) in 5.2d explains that unhappy is not the complement of happy, so that one can say John isn’t happy and he isn’t unhappy either. (However, not happy is the complement of happy, as would be non-happy if this were to be said.) The fact that non- creates a dichotomy makes it particularly suitable for use in scientific and legal work.

94    5 making negative words 5.7b The difference of meaning between un-1 and in-, on the one hand, and non-, on the other, is clearly brought out by adjectival minimal pairs. Un-1 or in- is attached to an adjective referring to a quality that is prized; the un-1 or in- word has a negative—often, damning or derogatory—meaning. To take an oft-quoted example, for anyone in the USA to refer to another citizen (or their behaviour) as un-American is about the worst insult there could be; the dictionary definition is ‘contrary to the ideals and values of the USA’. In contrast, non-American is a neutral term; if one talks of a non-American linguist or non-American company this is simply a statement of fact, without any overtones of good or bad opinion. A few more pairs are instructive:

•

An il-literate person within a literate society is to be despised as lacking an expected skill. In contrast, a non-literate people is simply a group who have never developed or adopted a writing system. An in-human act or an in-human person is likely to be cruel or brutal, going against the positive qualities associated with being a human being. A nonhuman is simply a creature who is not human—maybe an animal or an alien. An un-professional act is something which contravenes the ethical standards of the profession in question. A non-professional person is simply someone who does not follow a profession. A statement may be called un-repeatable if it is too rude for the narrator to enunciate in the presence of a certain set of listeners. A shop may advertise a bargain and refer to it as non-repeatable, meaning that they will not be able to offer it again. An un-charitable person or an un-charitable act lacks benevolence. However, a non-charitable organization is simply an organization which is not a charity. Ir-religious is often used of a person who is hostile to religion, and perhaps blasphemes it, whereas non-religious simply indicates that there is no connection to any religion, as in a non-religious building.

•

• • • •

There are some pairs of un-1| in- and non- derivations which in practice have very similar meanings—un-aligned and non-aligned (for example, a nation), in-essential (or un-essential) and non-essential, un-believer and non-believer. 5.7c There are a few in- derivations which lack the normal negative overtone. For example, in-credible can refer to something which is amazing, almost beyond belief (and often good), as in The wildflowers are incredible this year,

5.8 (4) a - ~ an -   95 and He received an incredible price for that house. A witness in court might give testimony considered to be non-credible, simply because the witness has been known often to lie in the past and what they say cannot be trusted in this instance. The non- derivation here shows a negative connotation (it is as if it carries the meaning one would expect in-credible to have). 5.7d A major difference between prefixes is that un-1 and in- forms essentially refer to a gradable property. One can say: He is very| highly un-­professional. However, non- derivations simply indicate membership of a class; as a ­consequence, one cannot say *He is very| highly non-professional. 5.7e Prefixes un-1 and in- are almost exclusively used with adjectives, and the comparisons with non- just quoted were adjectival. But non- can also be used quite freely with nouns. It is here that, since about the 1920s, a pejorative sense of non- has emerged—where Z is a noun, a non-Z is not just something which is not a Z, but rather something which should be a Z or ought to be a Z and simply does not come up to expectations. A non-event can be used to describe something that was heralded as highly significant but turns out to be a damp squib. A non-book may be used of something which looks like a book but in fact has little literary merit or useful content. (Algeo 1971 ­provides an illuminating discussion.)

5.8 (4) a- ~ an5.8a Negative prefix a- /ei- / (before a consonant) ~ an- /an-/ (before a vowel), is of Greek origin, cognate with Germanic un-1 and with Romance in-. Many English loans from Greek (usually, via Romance languages) involve a- or an- as part of the root; for example, anarchy, apathy and atom. The prefix a(n)- acquired a degree of productivity in English from the nineteenth century. It means, roughly, ‘not relating to a parameter’, rather similar to the meaning of non-. For example:

• • •

a-political (1952)—a person, organization, or activity not related to politics a-tonal (1922)—music that is not related to tone; that is, not in any key or system of scales a-temporal (1870)—timeless, free from the limits of time, as in: Heaven is imagined as an a-temporal domain

96    5 making negative words 5.8b. Most a(n)- forms are adjectives (as above) but some are nouns; for example a-theist, a-theism (and derived adjective a-theist-ic). The initial abears primary stress in a-theism and a-theist, and also in words no longer recognized to involve a prefix such as apathy and anarchy, but in almost all productive derivations with a(n)- the prefix is not stressed. In 1869, Thomas H. Huxley coined agnostic, ‘someone who believes that nothing can be known outside human experience (including: it cannot be known whether or not there is a god)’, from the Greek loan gnostic, an adjective meaning ‘relating to knowledge, especially of a spiritual nature’. Today, negative prefix a(n)- has virtually no productivity in everyday style, but is used a good deal in medical and scientific terminology; for example a-placental ‘(an animal) lacking a placenta’ and a-caudal ‘tail-less’.

5.9 Contrasting non- and a(n)Non- and a(n)- have broadly similar meanings. Since there are so few a(n)derivations, it is not easy to find both prefixes on the same stem. Both asymmetrical and non-symmetrical are attested but it is difficult to discern a difference in meaning. However, there is one fully established minimal pair:

• •

a-sexual (1830) is used to describe a person who does not take part in sexual activity; that is, who is neither heterosexual nor homosexual. non-sexual (1836) is used to describe a type of activity not related to sex. For example, The choirmaster engaged in various activities with the boys, all of a non-sexual nature, and Reproduction may be divided into non-sexual and sexual.

Having described four prefixes which apply to adjectives, or to adjectives and nouns (un-1, in-, non- and a(n)-), we can now examine de-, which derives verbs from nouns and adjectives, and then investigate prefixes applying to transitive verbs: dis-, un-2 and mis-.

5.10 (5) de5.10a Prefix de- /di:-/ ultimately relates to Latin preposition de¯  ‘down from’. It is an inherent part of many Romance loans, including depend, descend,

5.10 (5) de -   97 deceive, and declare (also deflate, the converse of inflate). Early examples of de- functioning as a prefix in English include de-mast (1666) and de-truth (1647) but the prefix only really achieved productivity in the nineteenth century, following a period within French of using prefix dé- plus verbalizing suffix -izer (as in, for instance, dé-national-izer ‘denationalize’). Many de- derivations are of a fairly learned nature. This prefix seldom bears stress. 5.10b There are basically two senses. (i) Added to a noun referring to a vehicle, de- forms an intransitive verb ‘get off the vehicle’. There is a prefix en- ~ em- with converse meaning ‘get on the vehicle’ (7.3.3). Thus: em-bus (1915) de-bus (1915)

en-train (1881) de-train (1881)

em-plane (1923) de-plane (1923)

We also find de-rail (1850) to describe a train coming off the rails (or someone or something making it come off). There is a degree of productivity for the suffix with this meaning. The OED gives de-cart from India in 1860. Mencken (1945: 399) mentions that language purists in the USA tried to denigrate (and dispose of) de-train ‘by concocting satirical analogues’ such as dehorsecar and decanalboat. (ii) Creating a transitive verb on the basis of a noun, the verb meaning ‘deprive of ’. There are two subtypes, both quite productive: • Just add de- to the noun, as in de-plume, ‘strip the feathers off a bird’, de-forest (the land), de-bug (a computer program), de-frost (a refrigerator), de-gut (an animal), de-fang (a snake), de-cipher (a message), and de-throne (a monarch). (Many further examples are quoted in Mencken 1945: 399–400.) • Prefix de- plus a verbalizing suffix: – with suffix -ify, as in de-acid-ify, de-electri-fy, de-nazi-fy – with suffix -ate, as in de-oxygen-ate – with suffix -ize, as in de-militar-ize, de-magnet-ize, de-union-ize, depressur-ize, de-tribal-ize, de-Stalin-ize.

98    5 making negative words Some derivations are based on adjectives, including de-central-ize, de-christianize, de-urban-ize, de-material-ize. De- is also added to verbs, typically with one of the endings which were borrowed entire from Romance (rather than being created from an English noun), such as de-contaminate and de-segregate. 5.10c Nouns can be derived from verbs in de-; for example, the de-forest-ing of Amazonia, plus de-nazi-fi-cation, de-oxygen-at-ion, de-union-iz-ation, detribal-iz-ation. There are just a few other nouns with de-, such as de-emphasis (parallel to verb de-emphasize) and the medical term de-nutrition ‘treatment by deprivation of nourishment’. 5.11 (6) dis5.11a Prefix dis- /dis-/ is also of Romance origin. It is an indissoluble part of the root in many loans, such as distance, dispute, disaster, and disdain. Dis- became a productive prefix in English from the sixteenth century, generally used with Romance loans but occasionally with Germanic roots (most notable are dis-trust, dis-like, dis-believe, and dis-own). The prefix takes secondary stress when the first syllable of the stem to which it is attached does not bear stress, as dis-believe /ˌdisbi'li:v/. When the stem has initial stress, dis- generally also bears secondary stress—as in dis-prove /ˌdis'pruv/—but sometimes does not—as in dis-like /dis'laik/. One fascinating property is that prefix dis- always maintains its final sibilant as voiceless whereas, when dis constitutes the first syllable of an unanalysable root, the ‘s’ generally assimilates in voice to become /z/ before a vowel—as in disaster /di'za:stə/—and sometimes also before a voiced consonant—we find both /diz'dein/ and /dis'dein/ for disdain. 5.11b Negative prefix dis- basically has three distinct senses and uses. (i) Very similar to sense (ii) of de-. That is, creating a transitive verb on the basis of a noun, meaning ‘deprive of ’. Prefix un-2 also bears this sense. At various times, the three prefixes have been in competition on the same roots; these include: de-frock (1581) de-mast (1666) de-plume (c1420) de-throne (1609)

dis-frock (1837) dis-mast (1747) dis-plume (1480) dis-throne (1591)

un-frock (1644) un-mast (1611) un-plume (1587) un-throne (1611)

5.11 (6) dis-   99 (We also find dis-en-throne from 1606.) Today, the de- forms seem more common than those in dis-. Note that the verbs ‘remove feathers from’ are de-plume and dis-plume (rather than defeather and dis-feather) since they are based on French déplumer (from plume ‘feather’). Similarly de-frock and dis-frock ‘take away priestly status’ relate to French défraquer, and de-throne| dis-throne relate to French détrôner. However, mast is a Germanic root to which Romance-origin prefixes de- and diswere attached. It will be seen that the forms with Germanic prefix un-2 were in most cases coined at a slightly later date than the de- and dis- words. (5.16b discusses competition between the three prefixes with sense ‘deprive of’.) (In Old French des-, a development from Latin dis-, became de- before a consonant. However, in English both dis- and de- are found before stems beginning with any consonant or vowel.) Note that this sense of dis- is restricted to use with plain noun roots. Nowadays there are no dis- derivations similar to de-acid-ify or de-oxygen-ate or de-magnet-ize. (ii) The second sense of dis- is rather similar to the meanings of un-1 and in-. With adjectives and nouns (all of Romance origin), it refers to lack of a quality, normally something that is valued. Thus we get adjectives dishonest and dis-loyal and nouns dis-advantage and dis-courtesy. Each word class may be derived from the other, producing nouns dis-honest-y and disloyal-ty, adjectives dis-advantage-ous and dis-courte-ous. (Whether an adjective is negated by dis-, rather than un- or in-, relates to how it was treated in French.) Unlike un-1 and in-, dis- can be prefixed to certain transitive verbs which carry a positive connotation—with a similar meaning to that which un-1 and in- have with adjectives, of simple negation. For example, if you do not like something, then it is probable that you dis-like it. This sense is found with the following semantic types of verbs: using—dis-use obeying—dis-obey thinking—dis-believe speaking—dis-agree (with), dis-affirm, dis-confirm, dis-avow liking—dis-like, dis-approve, dis-favour, dis-respect, dis-trust, dis-honour annoying—dis-satisfy, dis-please making—dis-allow helping—dis-oblige

100    5 making negative words Almost all of these dis- forms have nominal and some have adjectival derivations. Nominalizations involve the following suffixes:

• • • •

-((a)t)ion—dis-satisf-action, dis-affirm-ation -al—dis-approv-al (of   ), dis-avow-al -ment—dis-agree-ment (with) -ence—dis-obedi-ence

plus dis-belie-f, dis-pleas-ure, dis-use, dis-like, etc. Derived adjectives use suffixes -able, -ful, and -ed; for example dis-agreeable, dis-respect-able, dis-respect-ful, dis-satisfi-ed. (iii) Dis- has a quite different sense with some transitive verbs of location and affect—it indicates reversal of a state or process. (It could be described as depriving something of the result of a process, thus having similarities to sense (i).) Connect refers to a process of joining two (or more) things together; dis-connect describes reversal of the process with the things being separated. There are also nominalizations, connect-ion and dis-connect-ion. A sample of the verbs taking this sense of dis- can conveniently be arranged with respect to the nominalizing suffixes they take:

• • • • •

-((a)t)ion—dis-organiz-ation, dis-locat-ion, dis-associat-ion, dis-embark-ation, dis-unifi-cation -y—dis-unit-y -ment—dis-arrange-ment, dis-establish-ment, dis-engage-ment, dis-inter-ment -ing—dis-mount-ing, dis-arm-ing, dis-colour-ing, dis-burden-ing, dis-yoke-ing verb having secondary function as a noun—dis-order The ‘reverse’ sense of dis- is also used with some verbs referring to social interaction; these can also be listed with their nominalizing suffixes:

• • •

-((a)t)ion—dis-invit-ation, dis-qualifi-cation -ment—dis-employ-ment -ance—dis-encumbr-ance

5.12 contrasting un-1/in- and dis -   101 5.11c Two verbs associated with possession are particularly fascinating— own and possess. Suppose it is the case that X owns| possesses Y. Then the negative forms of the verbs indicate reversal and deprivation respectively:

•

X dis-owns Y—X repudiates ownership of Y (as in an 1856 quotation from the OED: The prince . . . was . . . required to disown . . . the obligations contracted in his name) Z dispossesses Y of X—Z takes X out of the possession of Y (as in The bailiffs dispossessed Mr Freen of his dwelling house)

•

(Note that different kinds of NPs are appropriate as X in the various constructions here.) Other dis- derivations include dis-prove—if someone disproves X then they prove X not to be true (a variety of reversal). And if someone dis-infects X, they are removing possible causes of infection from X (a type of deprivation). If someone dis-continues an activity, they cease continuing with it (which is vaguely a kind of deprivation). 5.11d Prefix en- ~ em- may be used for a verbal derivation to denote getting into a position or state (em-bark, en-tangle) or gaining a status (en-throne). Dis- may be used with these to indicate reversal; see 7.3.3. Dis-em-bark and dis-en-throne have shorter forms dis-bark and dis-throne. Dis-en-tangle may not be reduced to *dis-tangle, but—using prefix un-2—one can say un-tangle as (preferred) alternative to un-en-tangle. (There are some, rather uncommon, derivations with disen~ disem- where no verb with just en- ~em- occurs; for example disem-burden.) 5.11e As with un-1 and in-, the verbs to which prefix dis- applies are almost exclusively transitive. One of the very few intransitives is appear ‘come into sight’, forming the reversal negative dis-appear ‘become out of sight’. The prefix dis- is scarcely productive today, being used mainly in nonce formations. 5.12 Contrasting un-1|  in- and dis5.12a One can refer to something as either un-plac-ed, using prefix un-1, or as dis-plac-ed, using dis-. These have different meanings, associated with distinct derivational paths:

102    5 making negative words 1 verb root 2 verb derived from 1 3 adjective derived from 2 4 adjective derived from 1 5 adjective derived from 4

place dis-place [dis-plac]-ed plac-ed un-[plac-ed]

‘put in a particular position’ ‘move from a particular position’ ‘been moved from a particular position’ ‘been put in a particular position’ ‘not been put in a particular position’

Un-1, with basic meaning ‘not’ applies to an adjective (not to a verb). It is here added to derived adjective plac-ed, deriving un-[  plac-ed ] ‘not been put in a particular position’ (with the implication that it is good to be put in a particular position). Prefix dis- ‘deprive of, reverse action’ is used with verb place to derive dis-place ‘move from or deprive of a particular position’. From this is formed adjective [dis-plac]-ed ‘been moved from a particular position’. There are many further pairs of (i) un-1 with an adjective derived from a verb, and (ii) an adjective derived from dis- with the same verb. They include: [dis-integrat]-ed ‘parts which had been joined together are now separated’ un-[integrat-ed] ‘parts have never been joined together’ [dis-invit]-ed un-[invit-ed]

‘was invited, but invitation was later withdrawn’ ‘never been invited’

[dis-ordain]-ed un-[ordain-ed]

‘had taken holy orders, and then been deprived of them’ ‘never taken holy orders’

[dis-crown]-ed

‘had been recognized as monarch, but the office was then taken away’ ‘never been officially recognized as monarch’

un-[crown-ed]

[dis-embark]-ed ‘had been taken on board a boat, and then put ashore’ un-[embark-ed] ‘never been taken on board’ Verbs yielding similar pairs include organize, order, locate, mount, encumber, colour. 5.12b There are a number of verbs whose dis- negative form is really only used in participial form; for example: [dis-us]-ed ‘had been used, but no longer being used’ un-[us-ed] ‘never been used’

5.12 contrasting un-1/in- and dis -   103 The verb dis-interest ‘free oneself of personal interest in’ is archaic. But it has left behind the adjective [dis-interest]-ed: [dis-interest]-ed un-[interest-ed]

‘lacking any personal interest (which could affect impartiality of judgement in some matter)’ ‘have no interest in some matter, couldn’t care less’

5.12c As stated above, with liking verbs, such as like, dis- simply has negative effect. An adjective can be derived from like through adding suffix -able. We thus get: 1 2 3

4 5

verb root verb derived from 1 adjective derived from 2

like dis-like

adjective derived from 1 adjective derived from 4

like-able

[dis-like]-ed

un-[like-able]

‘have good feelings towards’ ‘don’t have good feelings towards’ ‘someone/something that people do not have good feelings towards’ ‘easy for people to have good feelings towards’ ‘not easy for people to have good feelings towards’

5.12d There are some un-1| dis- pairs which are close to synonyms. For example, it is hard to discern any difference of meaning between un-[  pleas-ed] and [dis-pleas]-ed. 5.12e The discussion so far has contrasted un-1 and dis-. What about in-? There is the pair of adjectives in-harmoni-ous and dis-harmoni-ous and also nouns dis-harmony and in-harmony, which appear to show little difference in meaning. In Dr Johnson’s dictionary (1755) he gives just the noun with dis- and the adjective with in-; Johnson’s definitions are: disharmony inharmonious

noun, contrary to harmony adjective, unmusical, not sweet of sound

It is likely that noun dis-harmony (first attested 1602) was initially founded, based on the noun harmony (1532)—parallel to such dis- nouns as disadvantage and dis-comfort—with adjective dis-harmoni-ous (1659) being derived from this. In contrast, adjective in-harmoni-ous (1711) was based on adjective harmoni-ous (1638) and then the noun in-harmony (1799) was back-formed from this adjective.

104    5 making negative words 5.12f Prefixes un-1 and dis- can co-occur in the same word. Suppose a group of youths form a gang, which commits depredations. The police order them to dis-band; that is, to break up the gang. Someone might ask: Is the gang [dis-band ]-ed? And receive a reply: I’m sorry, but it seems that the gang is as yet un-[[dis-band ]-ed ].

5.13 Contrasting non- and disThere is an interesting minimal pair of nouns involving non- and dis-. As mentioned before, dis-appear is a rare example of an intransitive verb taking dis-. It enters into the following derivations: 1 2

verb verb derived from 1

3

noun derived from 2

4 5

noun derived from 1 noun derived from 4

appear dis-appear

‘come into sight’ ‘become out of sight after being in sight’ [dis-appear]-ance ‘act of becoming out of sight after being in sight’’ appear-ance ‘act of coming into sight’ non-[appear-ance] ‘failure of act of coming into sight to eventuate’

5.14 (7) un-2 5.14a This, the second un- /ʌn-/ prefix in modern-day English, is a development from OE and- ~ und-. It is cognate with Latin ante ‘before’ and Greek anti- ‘against’. The parent form in proto-Indo-European was anti- ‘against’, the locative singular inflection of ant- ‘front, forehead’ (Watkins 1985: 3). This prefix was common in OE, but only a few of these forms have come through in analysable form into the present-day language; they include unbind, un-tie, un-do, un-fold, and un-yoke. A number of modern roots include a relic of OE and- ~ und. For example, the OED gives answer as coming from prefix and- ‘against, in reply’ and *swarâ ‘affirmation, swearing’. The prefix was re-invigorated in ME times. Today, most derivations with un-2 show secondary stress on the prefix, although they may lose this stress in fast speech. 5.14b Un-2 has two distinct functions. Added to a noun it forms a transitive verb meaning ‘deprive of ’, like sense (ii) of de- and sense (i) of dis-. Added to a verb it typically has a reversative meaning, like sense (iii) of dis-.

5.14 (7) un-2   105 (i) un-2 may be added to one of a small group of nouns to form a transitive verb with the sense ‘deprive of’. Under sense (i) for dis- (in 5.11b) were given un-frock, un-plume, and un-throne (based on Romance roots) and un-mast (with a Germanic root); all may also take de- or dis- with indistinguishable meanings. We also find (with dates of first use, and of last use if now obsolete):

with Germanic root burden with Germanic root saddle with Germanic root horse with Romance root cloak with Romance root arm(s) with robe

de— — — — — (1841, rare)

dis1531 — 1859 1599–1677 1481 1590

un-2 1538 1382 1390 1598 1386 1598

(Robe was originally a Germanic root, but appears to have entered ME from Old French.) Saddle and arm each function as noun and as verb. Thus, unsaddle ‘take a saddle off’ and un-arm ‘remove arms from’ could be regarded as examples of the ‘deprive of’ sense (with a noun) or of the ‘reverse action’ sense, (ii) below (with a verb). It will be seen that there is a tendency to prefer un-2 with Germanic verbs and de- or dis- with those of Romance origin, but with a good deal of variation. There are also (with Romance roots) un-bishop ‘deprive of the office of bishop’ (1398) and un-pope which can mean either ‘deprive of the office of pope’ (1563) or ‘deprive [Rome] of the seat of the papacy’ (1655). (Disbishop and dis-pope are also attested, but not de-bishop or de-pope.) Like dis-, un-2 forms transitive verbs only with the plain form of a noun; it does not also create verbs using a derivational suffix, as de- does in de-acid-ify and the like. (ii) The most frequent meaning of un-2 is to indicate ‘reversal of a process’, particularly with verbs of the affect semantic type (and also with the general verb do).

•

from the wrap subtype of the affect semantic type: wrap, roof, clothe, dress, blindfold, sheathe, glove, wind, mask, load, ship, hand, harness, saddle (all Germanic); plus cover, veil, roll, mount (all Romance), etc. from the stretch subtype of affect: fold, coil, bend, twist, freeze, clench, twist (all Germanic), etc.

•

106    5 making negative words

•

from the build subtype of affect: tie, bind, yoke, button, lace, lock, bolt, knot, stick, glue, hitch, fetter, fasten (all Germanic), etc.

When we compare the roots to which un-2 is added and those taking dis-, both with the sense ‘reverse’, it will be seen that dis- is used almost exclusively with Romance loans (dis-yoke being an exception) and un-2 predominantly with Germanic roots. 5.14c Some of the verbs with un-2 are ambitransitive of type S = O. They may take un-2 in either their transitive or their intransitive value, as in The salesperson un-rolled the carpet The carpet un-rolled (by itself  ) The seaman un-coiled the rope The rope un-coiled (by itself  ) We also get The snake un-coiled and (with a metaphorical sense) The story un-folded. 5.14d The criterion for an affect verb to be used with un-2 is simple—it must describe an activity which can be reversed. One can veil a face and then un-veil it; one can butter a piece of bread but scarcely *un-butter it (it is not a straightforward matter to get all the butter off  ). One can bend a thin piece of metal and then un-bend it; one can stretch a piece of cloth but scarcely *un-stretch it (once stretched it is likely to stay stretched). 5.14e A number of other verbs have been used with un-2. For example, unmarry, ‘get divorced’; un-say, ‘deny’; un-teach ‘make students forget something, often by teaching the opposite’; un-charm ‘deprive of magical power, or of the power of fascination’; and un-wish, ‘revoke a wish’. None of these has moved into common everyday usage, save perhaps un-learn ‘try to forget something that you wish you hadn’t learnt (for example, a bad habit or false knowledge)’. There was a transitive verb nerve ‘give strength or courage to’. This is now seldom used save in the reflexive expression nerve oneself (to do something) ‘try to have courage’. But it has left a commonly used un-2 derivation, the transitive verb un-nerve ‘make worried or scared.’ 5.14f Un-2 is sometimes used redundantly, with a verb whose inherent meaning is to undo a process. For example, un-loose (1362) has a meaning no different from loose (c1225) ‘set free’. And un-thaw (1598) is a synonym of thaw

5.16 the use of un-2 , dis -, and de -   107 (c1000) ‘change from frozen to liquid state’. In these cases, the un-2 may perhaps add an element of intensity. Other redundant uses of un-2 include un-pick (for example, a lock). From noun cipher (1528) is derived transitive verb de-cipher (1545) ‘deprive (a message) of cipher’, by sense (ii) of prefix de-, and then, with the same meaning, un-de-cipher (1654), by redundant application of sense (iii) of un-2.

5.15 Contrasting un-1 and un-2 Un-1 ‘lack a (valued) quality’ applies to many adjectives (some of which are derived from verbs), and un-2 ‘deprive of, reverse action’ to a few nouns and to many transitive verbs. Adjectives involving un-1 and those involving un-2 have distinct derivations and meanings. Compare: verb educate adjective derived from this educat-ed adjective derived with un-1 un-[educat-ed]

‘instruct’ ‘having received instruction’ ‘never having received instruction’

verb tie verb derived with un-2 un-tie adjective derived from this [un-tie]-ed

‘fasten’ ‘remove fastening’ ‘having had fastening removed’

Some verbs may enter into both types of un- derivation, producing adjectives which are ambiguous. For example: with un-1 un-[cover-ed] ‘never been covered’ with un-2 [un-cover]-ed ‘was covered, but covering now removed’ Similarly for clothe, dress, roll, pack, load, fold, coil, and a fair few more.

5.16 The use of un-2, dis-, and de5.16a Although sense (iii) of dis- and sense (ii) of un-2 have similar semantic effect, ‘reverse process or state’, they seldom apply to the same verb. Dis- is most used with verbs of location (dis-locate, dis-arrange) and un-2 largely with verbs of affect (un-tie, un-mask). We do get both dis-arm (1481) and

108    5 making negative words un-arm (c1386)—with the same meaning, ‘take arms from’—which could relate either to verb arm (1205) ‘provide with weapons’ or to noun arms (1340) ‘weapons’. Since there are, as shown in 5.11b, several examples of both un-2 and dis- with nouns (mast, throne, etc.), to say that dis-arm and un-arm relate to noun arms would provide a tidy analysis—there are several nouns, but no (or very few) verbs, which take both prefixes. A number of minimal pairs with un-1 and dis- were quoted in 5.12a— un-[  plac-ed] and [dis-plac]-ed, un-[integrat-ed] and [dis-integrat]-ed, and so on. It is likely that un-2 is avoided as an alternative to dis- with verbs of location in order to prevent confusion between un-1 and un-2. (There are, as we have shown, some un-1| un-2 indeterminacies; but for the most efficient communication, these are best kept to a minimum.) 5.16b Sense (ii) of de-, sense (i) of dis-, and sense (i) of un-2 all mean ‘deprive’ and can be used interchangeably with some (but by no means all) nouns. This was illustrated by the arrays under (i) in 5.11b and under (i) in 5.14b. Bauer (2006: 191–3) provides an instructive comparison (with further example triplets). He demonstrates that, to indicate ‘deprive of’, dis- was the prefix most used from the fourteenth to the sixteenth centuries, while un- grew in popularity until it was the most-used prefix in the eighteenth century. Then, in the nineteenth century, dis- had a resurgence before, in the twentieth century both dis- and un- lost productivity. Meanwhile, de- had become steadily more frequent in new coinages, and is today the only productive prefix with this meaning.

5.17 (8) mis5.17a The Germanic prefix mis- /mis-/ occurred in OE; words from that period which have continued include mis-lead, mis-time, mis-do, and the noun derived from this, mis-deed. There was a prefix mes- in Old French with a similar meaning ‘do wrongly, do improperly’, and in ME this naturally coalesced with Germanic mis-. Thus, mischief is from Old French meschief and miscreant from mescreant. Other French loans entered English together with the mis-less form: mis-adventure and adventure, mis-chance and chance, and so on. Mis was often written as a separate word in OE, being sometimes hyphenated in ME times, and written as one word only from the sixteenth century.

5.17 (8) mis -   109 In early times it could modify a conjunction or disjunction. Jespersen (1942: 483) quotes from Chaucer, That shal he finde, that hir misdooth or seith (where mis- applies to both dooth and seith) and the OED quotes an example from 1614, For thou hast mis-said or done (again, mis- applies to both said and done). Mis- became prolific as a derivational affix in the seventeenth century, functioning as prefix to verbs and to just a few nouns. It retains a measure of productivity today. Prefix mis- generally bears secondary stress. When it precedes a stem commencing with s, it is generally the case that s is pronounced in both syllable-final and syllable-initial position, as in misspell /ˌmis'spel/. Like dis-, the final s of mis- is always pronounced as /s/—as in misapply /ˌmisə'plai/— whereas in an unanalysable root commencing with syllable mis the s is pronounced as /z/ before a vowel (assimilating in voice to it)—as in miserable /'mizərəbl/. 5.17b At one time mis- had a straightforward negative meaning with some verbs. Mis-remember could then mean ‘forget’ (that is, ‘not remember’); nowadays it bears the standard sense of mis-, ‘remember wrongly’. Mis-like was used for ‘don’t like’ but has now been replaced by dis-like, with sense (ii) of dis-. Mis-trust had a similar sense to dis-trust; both continue in use today with meanings that are hard to distinguish. There may be a difference of emphasis, mis-trust evincing a doubt as to whether someone can be trusted, and dis-trust indicating a total lack of trust. Mis-trust is exceptional. Most transitive verbs which take mis- bear the meaning ‘do wrongly, improperly’. Different shades of meaning are:

• • • • • • • • •

put in the wrong place: place, locate, lay, lodge, deliver, align put in wrong place with respect to each other: sort, file, deal, join, arrange, yoke, connect wrongly organize: organize, govern, manage, rule, plan, build, navigate wrong human (or other) as object: appoint, cast, choose wrong recipient: award, charge wrongly use: use, handle, apply, colour didn’t affect object properly: hit, field, fit, shape, represent, spend assign wrong token: value, number, date, address, caption, catalogue wrong mental process: understand, apprehend, conceive, believe, construe, calculate, hear

110    5 making negative words

•

language used wrongly: pronounce, articulate, recite, read, write, type, punctuate, spell, translate, interpret, copy, quote, state, assert, affirm, report, inform, explain, claim, advise, teach, educate, instruct

5.17c Only a few intransitive verbs accept mis-. A gun may mis-fire, a person may mis-behave. And a pregnant woman (or a scheme) may mis-carry. Many verbs with mis- form a nominalization, employing the usual derivational suffixes. Just a few examples of each are given here:

• • •

-((a)t)ion—mis-connect-ion, mis-appli-cation, mis-articulat-ion, etc. -ment—mis-place-ment, mis-manage-ment, mis-state-ment, etc. -ing—mis-understand-ing, mis-spell-ing, mis-colour-ing, etc.

There are also mis-deliver-y, mis-carri-age, mis-belief, mis-advice, and (with roots that do double duty as verb and as noun) mis-rule and mis-use. 5.17d Just a few nouns with mis- do not relate to a current verb. Mis-adventure (c1300) is a loan from French mes-aventure. Mis-giv-ing (1582) ‘feeling of worry or apprehension’ is from an archaic sense (1513) of mis-give ‘cause to create feeling of doubt or apprehension’. Miscegenation was coined in the USA about 1863 by David Goodman Croly and George Wakefield on the basis of Latin misce¯re ‘mix’ and genus ‘race’ (and so is only a mis- word by reanalysis). Among the very few mis- nouns which were derived directly from a noun are mis-trial (1628) ‘trial where something went wrong’, from trial (1577), and mis-fortune. The noun fortune (c1300) can refer to ‘chance, destiny, luck’, generally with the implication that this will be good; mis-fortune (1441) describes an adverse happening, where chance or destiny has worked out wrongly. 5.17e Some adjectives may be derived from verbs with mis-, all as participles. They include mis-shap-en, mis-cast, mis-pronounc-ed, mis-nam-ed, mishear-d, and mis-govern-ed. 5.18 Contrasting un-1 and misThe text at the beginning of this chapter includes adjective un-fortunate and noun mis-fortune. We find:

5.19 contrasting dis - and mis -   111

1 2 3 4

noun, borrowed from French (c1300) adjective, borrowed from Latin (1386) noun, derived from 1, with mis- (1441) adjective, derived from 2, with un-1 (1530)

fortune

‘chance, destiny’

fortunate

‘having good fortune’ ‘adverse chance, bad luck’ ‘lacking good fortune’

mis-fortune un-fortunate

(There is also the analogic formation mis-fortunate (1510), used in Scottish English, with the same meaning as un-fortunate.) Since mis- is still productive, further minimal pairs of adjectives (all derived from verbs) can easily be constructed—for example, un-[controll-ed] ‘not controlled’ and [mis-controll]-ed ‘controlled improperly’; un-[  prepar-ed] ‘not prepared’ and [mis-prepar]-ed ‘prepared in a wrong manner’. 5.19 Contrasting dis- and misThere are a considerable number of minimal pairs, involving transitive verbs, with dis- and mis-. For example: dis-place mis-place

‘move from particular position’ ‘put in wrong position’

dis-order mis-order

‘disturb arrangement of objects’ ‘put objects in wrong arrangement’

dis-join mis-join

‘separate things that had been connected’ ‘connect wrongly’

dis-colour mis-colour

‘spoil or remove the colour of’ ‘put the wrong colour on’

dis-use (typically in participial form, dis-use-d) mis-use

‘cease using’

dis-believe mis-believe

‘do not believe (a person or a doctrine)’ ‘believe that something is true when in fact it is not’

‘use wrongly’

Similar pairs are based on lodge, yoke, confirm, endow, invest, employ, and embark, among others.

112    5 making negative words As mentioned in 5.17b, mis-trust and dis-trust have similar meanings; a difference may be that dis-trust indicates an even lower level of confidence than does mis-trust. We can now move on to the final group of negative prefixes: mal-, ex-, anti-, counter-, and contra-. None of these changes word class. All occur with nouns, mal-, anti-, counter-, and contra- also with a few adjectives and verbs. Ex-, anti-, and counter- are like non- in that they can have a complete noun phrase as their scope. 5.20 (9) mal5.20a Prefix mal- /mal-/ comes from French prefix mal- ‘badly’. Early loans included mal-adventure (1447) ‘an unlucky undertaking’, now obsolete. Mal- was a productive prefix in English from the middle of the seventeenth century, with the meaning ‘defective, inadequate’. It occurs predominantly on activity nouns, which are often derived from verbs. At a later stage, participial adjectives, verbs, and agentive nouns may have been derived from mal- plus an activity noun. There is generally secondary stress on the prefix mal-, indicating that, like non-, it is thought of as an independent semantic element. From the eighteenth century, this prefix has been much used in coining medical terms, which is its predominant use today, for example, mal-position (of a foetus in the womb). However, it still does have a degree of productivity in everyday language. 5.20b Some of the most prominent mal- forms are: nouns mal-administration (1644) mal-government (1653–1880) mal-practice (1671) mal-conduct (1741) mal-formation (1800) mal-adjustment (1833) mal-organization (1841) mal-nutrition (1862) mal-adaptation (1877) mal-function (1939)

derivations (sometimes, back-formations) verb mal-administrate (1705) noun mal-practitioner (1800) adjective mal-formed (1817) adjective mal-adjusted (1886) adjective mal-organized (1862) adjective mal-nourished (1911) verb mal-function (1941)

5.22 (10) ex -   113 The verb mal-treat (1708) appears to have been borrowed from French maltraiter. Activity nominalization mal-treat-ment soon followed (1721). Malcontent (1581) was taken from Romance as both noun and adjective.

5.21 Contrasting mis- and malMis- applies primarily to transitive verbs and mal- to nouns, which are generally derived from verbs. We find: 1 2 3 4 5

verb, loan from Romance verb derived from 1 noun, loan from Romance noun derived from 2 noun derived from 3

govern (1297) mis-govern (c1400) govern-ment (1483) [mis-govern]-ment (1565) mal-[govern-ment] (1653–1880)

Thus, for a couple of hundred years from 1653, we had both [mis-govern]ment ‘governing inadequately’ and mal-[govern-ment] ‘inadequately governing’, with very similar meanings. Only mis-govern-ment has survived into present-day English. 5.22 (10) ex5.22a Prefix ex- /eks-/ goes back to Latin prefix ex- ‘out of’ and is a hidden element in many English roots, including example, experience, exclude. The prefix assimilated to the following segment in Latin (for example, in effervesce) but this has not come through. In this meaning it has a minor and scarcely productive sense in English, ‘out of’ as in ex-territorial and ex-orbital. 5.22b The standard sense is ‘used to be but is no longer, former’ and with this meaning the prefix is highly productive. It is almost always hyphenated in writing (and often bears secondary stress in pronunciation), indicating that it is regarded as a separate semantic element. The prefix occurs with: (i) Alienable (that is, affinal) kin or partners, such as ex-wife, ex-husband, ex-mother-in-law, ex-father-in-law, ex-girl-friend, ex-boy-friend. We find ex- making up a complete word, as abbreviation for ex-husband or ex-wife (or ex-partner, ex-boy-friend, ex-girl-friend).

114    5 making negative words (ii) Names of offices or occupations. These may be specific, as ex-mayor, ex-president. Or they can be generic, as ex-seaman, ex-student, and ex-smoker. 5.22c Ex- may modify a complete NP, as in ex-[secretary of the party], ex[senator from Queensland] and ex-[Australian ambassador to Ireland and the Vatican]. I have heard a man refer jocularly to his long-term de facto partner as his un-wife (that is, wife in all but ceremony). One wonders whether, should they break up, he would then refer to her as his ex-un-wife.

5.23 Contrasting non- and exAn ex-smoker is someone who used to smoke and then stopped. A nonsmoker is someone who does not now smoke; they may or may not have smoked in the past.

5.24 (11) anti5.24a Prefix anti- /anti-/ goes back to Greek anti- ‘opposite, against’ (often coming into English through Latin) and is an inherent part of such words as antipathy and antithesis. It has been a productive prefix in English since the second half of the sixteenth century. Anti- generally bears secondary stress (and is hyphenated in writing), showing that, like non- and ex-, it is regarded as a semantic modifier. Like them, it can be added to words and to phrases. 5.24b There are four meanings associated with anti- in the modern language: (i) ‘Alternative to’ or ‘rival of’. This occurs with nouns, as in anti-pope ‘elected as pope by one group of clerics in opposition to someone who was elected by another group and does currently hold the office’. (ii) ‘Reverse of’. This also occurs with nouns. For example, anti-climax ‘something unimportant occurring where something important (a climax) was expected’; anti-hero ‘someone with a central role (which would normally be assigned to a hero), but who lacks the admirable properties normally associated with a hero’; and anti-neutron ‘a particle with certain properties the reverse of those of a neutron’.

5.25 (12) counter -   115 (iii) ‘Opposed to, opponent of’. This may be added to nouns and adjectives:

• • •

added to a noun, deriving a noun (which may also function as an adjective). For example, anti-semite, anti-abolitionist added to a noun, deriving an adjective. For example, anti-slavery, anti-feminism, anti-abortion, anti-war added to an adjective, deriving an adjective. For example, anti- semitic, anti-feminist, anti-capitalist(ic)

Reed (1933: 14–16) provides an enlighting survey of this use and suggests that a ‘history of oppositions movements’ in the USA could be compiled from examination of coinages involving anti-. These include anti-federalist (from about 1788), anti-masonry (1827), anti-Texas-annexation (1846), and anti-secession (1865). The complement of this sense of anti- is pro- (discussed in 6.8.2). One can be either anti-abortion or pro-abortion. It is possible to encounter a double antiwithin an appropriate discourse context. Suppose someone asks Robin: Would you like to join my anti-abortion league? Horrified, Robin’s reply echoes the question, adding another anti-. He says: Certainly not, I’m anti-anti-abortion. Also in this sense, we do find anti used as a word on its own. For example, in He’s arguing for nuclear reactors but I’m taking the anti- position, the word anti- is short for anti-nuclear-reactors. (iv) ‘used against’

•

added to a noun, deriving an adjective. For example, anti-aircraft (gun), anti-wrinkle (cream). There are also many medical coinages, generally meaning ‘used against a particular disorder’; for example, antibacterial, anti-bilious, anti-catarrhal.

5.24c This prefix may be attached to a longish phrase; for example his anti[everything to do with the aristocracy] attitude, her anti-[love is what makes the world go round ] stance—which are sense (iii)— and an anti- [star wars protective shield ] missile—which is sense (iv). 5.25 (12) counter5.25a Prefix counter- /kauntə-/—from Old French contre- ~ countre‘against’—is an inherent part of such roots as counterfeit and countermand.

116    5 making negative words From the sixteenth century it has been a productive prefix in English. Counter- generally takes primary stress (and is hyphenated), the stem which follows then being marked for secondary stress. Counter- can be added to verbs, nouns and adjectives. 5.25b This prefix has two senses: (i) ‘Do the opposite of’. For example, if someone attacks, you may counter-attack (1916) them. A counter-revolution (1795), run by counterrevolutionaries (1799) aims to reverse the effects of a recent revolution. Counter-balance (1580) is to correct some imbalance by applying an opposite effect (this is both verb and noun). If someone makes a claim on you, one strategy is to make a counter-claim (1784) on them (this is also both noun and verb). (ii) ‘Be the opposite to’. A counter-culture (1970) operates with a set of values markedly different from (and often opposed to) those of mainstream society. A counter-attraction (1763) to X is an alternative attraction, of an opposite nature to X. Something is counter-productive (1959) if it produces the opposite effect to that desired. To countersign (1598) a document is to place one’s signature opposite a previous signature.

5.25c There are a fair number of further derivations, all variants of the semantic patterns just illustrated. And counter- can have a longer constituent as its scope, as in He belongs to the counter-[Chinese expansion] branch of the agency. It can also be doubled, as in: The enemy have a counter-terrorist arm, and we have responded with our counter-counter-terrorism wing.

5.26 Contrasting anti- and counterA term was needed to describe circular motion in the opposite direction to that of the hands of a clock. Counter-clockwise is attested from 1888 and anti-clockwise from 1898. Both terms are used today. It is reported that counter-clockwise is preferred in the USA and anti-clockwise in Britain.

5.27 (13) contra -   117 5.27 (13) contra5.27a Prefix contra- /kɔntrə-/ is from Latin contra- ‘against’, the same etymological origin as counter-. It is an integral part of such roots as contradict and contravene. The prefix—which carries a meaning ‘opposed to’ and generally takes secondary (sometimes primary) stress—has limited use in English. Examples include: ‘distinction made by contrasting with opposites’; for ­example, the OED gives, from 1874: You tend to produce a great capitalist in contradistinction to a number of small capitalists. There are also adjective contra-distinctive and verb contra-distinguish. contra-indication ‘an indication against the use of what would be (1623) the normal method of treatment’. There is also verb contra-indicate. contra-position (1581) ‘two things placed opposite to each other’. And there is a verb contra-pose. contra-flow (a British ‘going in a direction opposite to the normal one’. term from 1934) The Times of 9 April 1985 (quoted in the OED) provides a nice example: Resurfacing . . . has meant closing the northbound section and funnelling traffic into a contraflow system of two lanes each way on the southbound side. contra-distinction (1655)

5.27b There has been some substitution between counter- and contra-. For example, the 1987 Unabridged Random House Dictionary (but not the OED) mentions contra-clockwise. We can now move on, in the next chapter, to non-negative affixes which do not change word class.

6 Ultra-patriotic, extra-special, pre-election pseudo-promise: Keeping the same word class 6.1 Introduction 118

6.5 Within, between, and beyond  138

6.2 Number and the like: mono-,

6.6 Further locationals  142

uni-, one-, di-, bi-, two-, tri-,

6.7 Temporals  148

three-, hemi-, semi-, demi-, half-,

6.8 Roles  161

poly-, multi-, many-, pan-,

6.9 Repetition, re- 169

omni-, all- 120

6.10 Not quite right, pseudo-

6.3 Magnitude and degree  124 6.4 Above and below, and extensions to degree  129

and quasi- 170 6.11 Diminutive and endearing suffixes  171

6.1 Introduction 6.1a The great majority of suffixes always serve to change word class—­ between noun, verb, adjective, and adverb—or also change sub-class of noun (for example, music, music-ian, music-ian-ship). They are discussed in Chapters 7, 8, 9, and 10. In contrast, prefixes seldom change word class or sub-class, instead just adding a semantic element. Negative prefixes were dealt with in the last chapter. We saw there four exceptions to the not-changing-word-class principle: some (but not all) senses of de-, dis-, and un-2 derive a transitive verb from a noun or adjective, and anti‑ may derive an adjective from a noun. This chapter is concerned with the remaining prefixes. Most of them ­always maintain word class and sub-class. However, the six Germanic prefixes out-, in-, off-, on-, up-, and down- (which have the same form as

6.1 introduction   119 prepositions/adverbs) maintain word class in some instances, and change it in others—noun to adjective or verb, verb to noun or adjective. They are discussed in 6.6.5; with their word class effect summarized in Table 6.7. Four other prefixes generally maintain word class but may also d ­ erive an adjective from a noun. They are after- and post- in 6.7.1d, pre- in 6.7.3b, and pro- in 6.8.2 (this behaves like its complement anti-, discussed in 5.24b). There are just a few suffixes which do not change word class—­diminutives and endearments, which are discussed here in 6.11. And feminine agents which are most conveniently discussed with other agentive nouns in 9.3.9–10. 6.1b Suffix -ish may be added to a noun and derives an adjective which describes a quality associated with the referent of that noun. For example, mul-ish ‘behaving in an obstinate way (as a mule does)’. There is a second sense of -ish whereby it is added to an adjective and (like the prefixes described in this chapter) does not change word class. Instead, it moderates or attenuates the reference of the adjective. For example, a yellow-ish garment is a sort-of yellow colour but not what one thinks of as prototypical yellow, and a clever-ish boy shows some sparks of intelligence but is not what one would call clever through and through. A full discussion of -ish is in 8.2.4, 8.3.2, and 8.4.17. 6.1c Prefixes in English generally bear secondary stress, with primary stress being on the form to which they are attached. Sometimes a prefix may also take primary stress; this is more common for nouns than for verbs. (Full discussions of stress will be found in Jespersen 1942: 464–533, and also Marchand 1969: 129–208.) Each prefix generally has a fixed form. When a Greek or Romance prefix, which has been taken over as a prefix in English, is also reflected in unanalysable words, there may be a different phonological form and stress pattern. Compare (and see also the discussion of re- in 2.5d): words with segmentable prefixes mono-lingual /ˌmɔnou-'liŋgwəl/ bi-lateral /ˌbai-'latərəl/ post-date /ˌpoust-'deit/ pre-judge /ˌpri:-'dʒʌdʒ/

unanalysable words monopoly /mə'nɔpəli/ biscuit /'biskit/ posture /'pɔstʃə/ predict /pri'dikt/

120    6 keeping the same word class 6.1d The prefixes discussed in this chapter are arranged according to their semantic effect. First, those referring to number and the like (Section 6.2), then magnitude and degree (6.3), various kinds of locationals and temporals (6.4–7), roles (6.8), repetition (6.9), then those describing something which is not quite right (6.10). Finally, suffixes indicating diminutive and endearment (6.11). 6.2 Number and the like: mono-, uni-, one-, di-, bi-, two-, tri-, three-, hemi-, semi-, demi-, half-, poly-, multi-, many-, pan-, omni-, all6.2a English has a plentiful stock of prefixes (and semi-prefixes) referring to numbers, plus ‘many’, ‘half’, and ‘all’. These apply to nouns and adjectives (and adverbs derived from them). For each value there is a prefix from Greek, one from Romance languages, and one Germanic. They are illustrated in Table 6.1. The Greek and Romance prefixes occur in many unanalysable loans (where the second portion does not occur as a word on its own). These include monologue, monarch, universe, unison, diphthong, dilemma, biceps, biscuit, trilogy, triple, polygamy, polyglot, multifarious, multiply, panoply, pantheon, omnivorous, omnifarious. Table 6.1 shows eleven prefixes of Greek and Romance origin. (Note that tri- relates to both languages; tri-syllabic and trilogy are of Greek origin while tri-focal and triple are Romance.) They became productive in English from the sixteenth century on. All—except for pan- and omni-—are extensively used in scientific naming, but each prefix also marks a derivational process in everyday vocabulary. Table 6.1  Number prefixes, with an example of each GREEK mono- /mɔnəu-/ e.g. mono-mania di- /dai-/ e.g. di-polar tri- /trai-/ e.g. tri-syllabic hemi- /hemi-/ e.g. hemi-sphere poly- /pɔli-/ e.g. poly-tonality pan- /pan-/ e.g. pan-tropical

ROMANCE uni- /yuni:-/ e.g. uni-lateral bi- /bai-/ e.g. bi-labial tri- /trai-/ e.g. tri-focal semi- /semi-/ e.g. semi-circle multi- /mʌlti-/ e.g. multi-national omni- /ɔmni-/ e.g. omni-present

GERMANIC one- /wʌn-/ e.g. one-legged two- /tu:-/ e.g. two-faced three- /θri:-/ e.g. three-score half- /ha:f-/ e.g. half-moon many- /meni-/ e.g. many-sided all- /ɔ:l-/ e.g. all-time

6.2  number and the like: mono -, uni -, one -, di -, bi -, two -   121 6.2b In the examples provided in Table 6.1, prefixes of Greek origin are used with roots of Greek origin, Romance with Romance, and Germanic with Germanic. Across the language as a whole, however, things are not quite so tidy. It is the case that Greek roots almost always take Greek prefixes. But we also find prefixes of Greek origin used with Romance forms; for example, mono-consonantal, bi-plane, pan-African. There are quite a few roots for which Greek and Romance prefixes are both acceptable. For the following pairs, each member is commonly used, and there is basically no difference in meaning: (1) mono-dimensional mono-directional mono-cellular mono-valent mono-cycle

uni-dimensional uni-directional uni-cellular uni-valent uni-cycle

(Note that the first four rows involve Romance roots, while cycle is of Greek origin.) However, there are quite a few words for which no alternative is possible. Someone knowing just one language is mono-lingual, rather than *uni-­ lingual, and something with just one side is uni-lateral, not *mono-lateral (both lingual and lateral are Romance forms). 6.2c Generally, if a root or stem occurs with prefixes from more than one row of Table 6.1, they will all come from the same column. One says biweekly ‘every two weeks’, tri-weekly ‘every three weeks’, semi-weekly ‘twice a week (literally, every half week)’, never *di-weekly, *hemi-weekly. Similarly for -monthly, and -annually. Syllable is a word from Greek, so that we naturally get mono-| di-| tri-| polysyllabic. This series of prefixes is extended to other linguistic units. That is, we get mono-| di-| tri-| poly- with -vocalic, -consonantal, (although both are Romance forms) and -morphemic (of Greek origin, although it entered English through French). 6.2d The words in the two columns of (1) were created at about the same time, around the second half of the nineteenth century. At that time, both Greek and Romance prefix series seem to have had good productivity.

122    6 keeping the same word class Today there are signs of a preference for Romance prefixes, with these tending to replace the corresponding Greek ones. For example, di-syllabic, dimorphemic, and the like are now being employed. They are not yet more common than bi-syllabic and bi-morphemic, but they do seem to be gaining ground on them. Recently created words show a marked preference for Romance prefixes. Someone who indulges in intimate acts with people of both sexes is said to be bi-sexual (never *di-sexual). Since about 1860, bi-polar and di-polar have been alternative (and, effectively, equivalent) descriptions for something having two poles, at opposite extremities. In the 1970s a term came into use to describe a variety of mental illness characterized by mood swings; this is called a bi-polar (not *di-polar) disorder. 6.2e In the ‘half’ row in Table 6.1, semi- comes from Latin. There is a further alternative, demi- from French. It occurs in a number of French loans such as demi-lustre, and also with Germanic forms as in demi-god ‘a being who is partly divine and partly human’. When a name was needed for a musical note which is half the length of a quaver, semi-quaver was introduced, towards the end of the sixteenth century. In the seventeenth century, there was some competition from demi-­ quaver, but semi-quaver prevailed. Then demi-semi-quaver was coined for a note half the length of a semi-quaver. And, for a note half that length again, the magnificent derivation hemi-demi-semi-quaver was created in 1853, utilizing Greek, French, and Latin prefixes for ‘half’. 6.2f Let us now consider the Germanic forms in the last column of Table 6.1. These were originally just free words and have, over the last few centuries, taken on a second role as semi-prefixes. In the great majority of instances a Germanic root only takes a Germanic prefix, and Germanic prefixes are only used with Germanic roots. We can usefully repeat here part of the array from 3.3d, illustrating Romance and Germanic roots with similar meaning, each taking an appropriate prefix: semi-conscious semi-nude semi-annual

half-awake half-clothed half-yearly

6.2  number and the like: mono -, uni -, one -, di -, bi -, two -   123 Romance form lateral is used with Romance prefixes and Germanic stem sided with Germanic ones. It might be thought that uni-lateral has a similar meaning to one-sided, bi-lateral to two-sided, and so on. In fact there is a difference, which can be brought out by considering the different kinds of noun phrase head likely to follow the Romance and the Germanic derivations respectively. Here is a sample: uni-lateral decision bi-lateral agreement tri-lateral negotiations

one-sided view of the matter two-sided tool three-sided table

Note that the number derivations may not be switched between the entries on each row—one cannot felicitously say bi-lateral tool or two-sided agreement. Very roughly, the -lateral terms refer to interaction between people or groups of people, while the -sided terms relate to some item, which may be physical (tool, table) or conceptual (view, or report, account, etc.). (Note that this is a rather rough-and-ready, tentative analysis; much more work is required on the topic.) 6.2g There can be mixing, with prefix and stem coming from different genetic groups. For example, alongside multi-lateral and many-sided (both from about 1600) there has recently come into being multi-sided (it is as yet only about one-fifth as common as the two longer-established forms). And we find both multi-coloured and many-coloured, which do have about the same meaning. When describing the size of a building, it is common to count upwards as follows: single-storeyed two-storeyed three-storeyed multi-storeyed This employs Romance prefix multi-, Germanic semi-prefixes two- and three-, and Romance form single- (one-storeyed would be a less-preferred alternative). A regular pot pourri. 6.2h Of the prefixes in the lowest row of Table 6.1, Romance omni- and Germanic all- are closest in meaning, indicating ‘in every way (or time)’.

124    6 keeping the same word class For example, one might describe an absolute monarch or a deity as either omni-potent or all-powerful, the only real difference being that the first is in a high-flown and the second in a down-to-earth style. In contrast, Greek prefix pan- relates to the whole of something; for example, pan-Christian links all component sects of this religion, and a pan-European treaty is one signed by all (or at least most) of the countries of Europe. Table 6.1 could be continued with rows for numbers higher than three: tetra-| quad(ri)-| four-, penta-| quinta-| five-, and so on. The higher-number Greek and Romance prefix forms occur in some scarcely analysable English words, such as tetrahedron, quadrant, pentagon, quintet. They are used productively in scientific nomenclature, and in just a few everyday words, such as quadri-lateral, quadri-­ centennial. The Germanic equivalents (four, five, etc.) show less cause than the lower numbers to be considered semi-­affixes, rather than just numbers entering into compounds such as four-eyes (for someone wearing spectacles) and five-legged. (As at many places in linguistic analysis, it is hard to know where to draw the line.) See 3.3d. Bauer and Huddleston (2002: 1719-20) have an informative account of numerical elements.

6.3 Magnitude and degree We now discuss four prefixes from Greek and three from Romance sources, plus the Germanic form well- (in 6.3.5). It is instructive to match the Greek and Romance prefixes, in Table 6.2. However, it must be borne in mind that here, and in similar tables throughout this chapter, items within each row have some similarity of meaning but also significant semantic and functional differences. None of the first four prefixes (mega-, micro-, maxi-, and mini-) came into English by the normal route; that is, through borrowing root and prefixplus-root pairs. Instead they were each consciously ‘manufactured’. For mega- and micro- this involved taking over prefixes from Greek to be prefixes within modern-day English, while mini- and maxi- were created on the

Table 6.2  Magnitude and degree prefixes GREEK mega- /megə-/ micro- /maikrəu-/ hyper- /haipə-/ hypo- /haipou-/

ROMANCE maxi- /maksi-/ mini- /mini-/ ultra- /ʌltrə-/

OUTLINE MEANING significantly large (of their kind) significantly small (of their kind) more than normal less than normal

6.3.2 maxi - and mini -   125 basis of Romance loans (miniature and minimum, and maximum). Each of the four is used with nouns.

6.3.1 mega- and microThere was discussion in 2.4g of how mega- was consciously adopted from Greek by scientists, in the late nineteenth century, being added to an existing noun to describe something that was uncommonly big; for example megabacterium. Then, from the 1960s, it came into popular usage for something of a size to be boasted of, such as mega-festival and mega-project. Micro occurred in microcosm (from the fourteenth century), a loan that was unanalysable in English. It took on the character of a prefix in the seventeenth century, through coinings such as micro-scope, an instrument for perceiving very small objects, and micro-phone, an instrument for strengthening scarcely audible sounds. In the nineteenth century its major use mirrored that of mega, when it was much employed in scientific naming; for instance micro-cephalic ‘having a small head’. In physics today, mega- is used for ‘a million’ and micro- for ‘a millionth’; for example mega-amp ‘a million amps’ and micro-amp ‘a millionth of an amp (a unit of electric current)’. Micro- did not quite match the explosion of use of mega-, from the 1960s, for something of fabulously large size. This is doubtless because being extremely small is, generally, less notable than being excessively large. However, having a tiny size may be regarded as desirable where electronic and other machines are concerned, so that we do get micro-circuit, micro-chip, and so on.

6.3.2 maxi- and miniMiniature was taken from Italian in the sixteenth century, and minimum from Latin in the seventeenth. These were the basis for the creation of minias a prefix in English, perhaps first in mini-bus from about 1850, said to be a shortening of miniature omnibus. (A competing name, a century later, was micro-bus.) However, it wasn’t until well into the twentieth century that mini- flowered as a productive prefix, with nouns and adjectives. Maximum had been in the language from the early eighteenth century and it spawned prefix maxi-, the antonym of mini-. The two Romance prefixes are typically used to describe vehicles (mini-taxi), appliances (mini-camera),

126    6 keeping the same word class ladies’ clothing (maxi-coat), political units (maxi-state), and we even encounter maxi-secret ‘something to be kept really hush-hush’. Plain mini is often used as an abbreviation for a mini-car (in Britain) and for a mini-skirt, one which is unusually short. There is a degree of competition, and indeed substitutability, between Greek prefixes mega- and micro-, and Romance ones maxi- and mini-. It is the Greek pair which are most used in scientific terminology but all four are productive in everyday speech. 6.3.3 hyper- and hypo6.3.3a Scientific nomenclature tends to make more use of Greek than of Romance prefixes. In keeping with this, both hyper- (originally meaning ‘over, beyond’) and hypo- (‘under’) have, since the nineteenth century, been much employed for coining new terms in music, mathematics, chemistry, and physiology. For example, hypercapnia ‘an abnormally high concentration of carbon dioxide in the blood’ and hypoxic ‘deficient in oxygen’. Around 1600, English borrowed (via French) both Greek adjective critical and derivation hyper-critical, thus establishing hyper- as a prefix in the language, with the meaning ‘more than is desirable’. However, it was not until the nineteenth century that hyper- really took off, as a prefix to nouns and adjectives, giving hyper-sensitive, hyper-tension, hyper-fit, and so on. Hyper-market, ‘combined supermarket and department store’ is a straight loan from French, without any overtones of undesirability. Recently, hyper- has been accorded a new use. It is employed in computing to describe establishing a link (a hyper-link) between two documents, thus creating a hyper-text, all parts of which are of easy access. Besides its use with adjectives and nouns, just illustrated, hyper- is sometimes used with verbs; for example hyper-ventilate ‘breathe more rapidly than is beneficial’. 6.3.3b In contrast, hypo- is today little used outside scientific language. ­Hyper-critical, ‘critical to an unacceptable extent’, is a common word, whereas one scarcely hears hypo-critical ‘not critical enough’. And hyperactive, ‘more active than is considered suitable’, has a quite high frequency, but the use of hypo-active is pretty well confined to physiology (for example, describing the poor condition of the thyroid gland).

6.3.5 well -   127 Like almost all non-Germanic affixes, hyper and hypo are found in many loans which are not analysable within English. They include hyperbole, hypocrite, and hypothesis. 6.3.4 ultraThis prefix comes from Latin preposition ultra ‘beyond, on the far side of’, and it shows this spatial sense in the first occurrence noted by the OED, from 1551—ultra-equinoctials ‘people who live on the other side of the equator’. (At that time equinoctial line, or just equinox, were used for the equator, where day and night should be of equal extent.) It was two and a half centuries later that ultra- became regularly used as a prefix in English and new coinages then abounded—ultra-patriotic, ­ultra-fashionable, ultra-benevolent, ultra-orthodox, and many more. It usually means ‘extremely’; for example, saying that something is ultra-clean implies that there is no possibility of any germs remaining there. If someone is described as being ultra-conservative, it implies that they show conservative traits in attenuated form. This may be judged a good or a bad thing depending on the predilections of speaker and hearer. But, typically, when someone uses ultra- they mean it to have positive overtones. To say that a person is ultra-critical implies that criticism is applied where it should be, in every possible way, and to good effect. Compare with hypercritical which is used of someone being unnecessarily overcritical, in a manner which is unhelpful and unpleasant. (And see 6.4.4.) Ultra- is of considerable utility in a sales pitch. A TV advert for a new computer called Ultrabook extolled its virtues as ultra-fast, ultra-thin, ultra-stylish. Although used predominantly with adjectives, ultra- does occur with a few nouns (ultra-Calvinist, ultra-Calvinism). Unlike hyper, it is little used in scientific nomenclature; but see supersonic, mentioned in 6.4.6. 6.3.5 well6.3.5a The Germanic form well has a wide range of functions. It can be an ‘introducer’, it is an adjective meaning ‘healthy’, it is the adverb corresponding to good, it is a component of compounds, and—what is relevant for this chapter—well- also functions as a prefix. It will be useful to briefly survey these various uses.

128    6 keeping the same word class 6.3.5b Well may be an ‘introducer’, initiating discourse, with varying meanings depending on the intonation employed. For example: Well, he’s won after all (surprise), Well, it’s true (confirmation), W-e-l-l (drawn out), maybe we shouldn’t go (hedging), Very well, you may go (affirmation), or a neutral introducer, as in the response of (2). 6.3.5c Well can be an adjective, then generally meaning ‘healthy’, as underlined in: (2)  Question. Is he well yet? Response. Well, he’s not very well now, but he’s getting better and soon he’ll be fully well. 6.3.5d. Well is the adverb corresponding to adjective good (where the regular form would have been *good-ly). It is instructive to compare—in Table 6.3— the forms of good with those of its antonym bad and a sample adjective with regular forms, quick. Whereas the comparative adverb can be more quick-ly or quick-er, worse or more bad-ly, in the final column it must be better, never *more well. (Note that, as seen in (2), better is also the comparative for well ‘healthy’; however, this adjective has no superlative or adverb form.) As an adverb, well can modify a verb, as in The minister spoke well. Or it may modify an adjective (typically, the participial form of a verb) as in The minister was well spoken and The well spoken minister convinced a number of doubtful delegates. 6.3.5e Well also occurs in a number of compounds, whose meanings cannot be inferred from the meanings of the components. These include well-being, well-off, well-heeled, well-met, and well-taken. 6.3.5f And well- /wel-/ functions as a prefix meaning ‘in an approved manner’; for example, well-deserved, well-earned, well-assured, well-accustomed, Table 6.3  Forms of regular adjective quick, and irregulars bad and good ADJECTIVE COMPARATIVE SUPERLATIVE

quick quick-er quick-est

bad worse worst

good better best

ADVERB COMPARATIVE

quick-ly more quick-ly (or quick-er)

bad-ly worse (or more bad-ly)

well better (never *more well)

6.4 above and below, and extensions to degree   129 well-fooled. In all of these derivations it is added to the participial form of a verb (functioning as an adjective), very similar to the adverbial use of well in well-spoken, well-behaved, and well-adjusted. The analytic question then arises: how can we distinguish between well- as a prefix and well as an adverb? In both instances the meaning of the whole is predictable from the meanings of parts. There are, however, three syntactic tests. For well functioning as an adverb, modifying an adjective (which is the participial form of a verb): (1) Well can be replaced by its antonym adverb badly. Alongside the well behaved boy we have the badly behaved boy. (2) Well can be replaced by the comparative adverbial form better, as in John was a better behaved boy than his cousin. (3) The sentence can be rephrased, with well as an adverb modifying the underlying verb, as in The boy behaved well. Behave is intransitive. For a transitive verb, test (3) involves using the verb in passive form. For instance, We saw some well woven scarves, alongside The scarves are woven well. (And, for the other two tests: We saw some badly woven scarves, The scarves are better woven than the mats.) A derived form involving prefix well- fails these tests. One day, I had the top bid at an auction for a certain painting, and was congratulated by the auctioneer: Well-bought, sir. The fact that well- is here a prefix, rather than an adverb, is shown by the inadmissibility of a sceptical friend saying *Badly-bought, or *Better-bought than the picture you acquired ten minutes ago, or *The picture was bought well. It is unusual to recognize a prefix well-. (Note that the arguments advanced here undoubtedly apply to earlier stages of the language.) The second edition of the Random House Dictionary provides, at the bottom of relevant pages, long lists of words commencing with each of 27 ‘prefixes’, one of which is well- (Flexner 1987: xxxii). Of the more than 1,500 words with well- (pp. 2157–9) some involve adverbial and others prefixal use. These can be distinguished by applying tests (1–3).

6.4 Above and below, and extensions to degree We now look at a set of prefixes which originated from prepositions indicating location. They still maintain a locational sense, but the main four (sub-,

130    6 keeping the same word class Table 6.4  Location and degree prefixes ROMANCE

GERMANIC

MAIN MEANINGS

sub- /sʌb-/ also infra- /infrə-/

under- /ʌndə(r)-/

below, part, subsidiary importance, less than normal

super- /su:pə-/ also supra- /su:prə-/, and sur- /sə:(r)-/

over- /ouvə(r)-/

above, greater importance, more than normal, to a high degree

under-, super-, and over-) have expanded to a wide range of other meanings. It is again useful to indicate, in Table 6.4, correspondences between prefixes of different origins, here Romance and Germanic. We deal in turn with each row in the table, and then, in 6.4.9, consider epi-, which is from Greek. 6.4.1 under- and sub6.4.1a Under- was a preposition in Old English, and also a prefix there; under-go and under-lie have come through into the modern language (although with modified meanings). It has continued to be highly productive, being attached to words of both Germanic and foreign origin. 6.4.1b The Latin preposition sub could be followed by a noun in ablative case—when its meanings included ‘under’, ‘at the bottom of’, and ‘be under the power of’—or one in accusative case—then the meaning was ‘to or along the underside of’ or ‘come under the power of’. Early loans included subject, substance, and sublime—which are not analysable in English—but also sub-dean, sub-alpine, and sub-join, alongside dean, alpine, and join, establishing sub- as an English prefix. It has continued to be highly productive, but the great majority of stems to which it is added are of Romance origin. 6.4.1c It is fascinating to compare the range of meanings for sub- with that for under-, in Table 6.5. Note that this covers only a fraction of the hundreds of words commencing with these prefixes (a full comparison would run to scores of pages). The word class of the derived form is shown by: A, adjective; N, noun; and V, verb. There are, of course, many further derivations related to the sample words in the table. For example, alongside noun sub-class we have verb sub-class-ify and abstract noun sub-class-if-ication. Together with adjective under-develop-ed go verb under-develop and noun under-develop-ment.

6.4.1 under - and sub -   131 Table 6.5  Comparison of some major senses of sub- and under(1)

PHYSICALLY BELOW

A sub-marine A sub-celestial N sub-soil

A under-water A under-foot N under-side N under-world V under-line

(2)

LOWER PART

N sub-script

N under-carriage N under-growth

(3)

CLOTHING

(4)

SUBSIDIARY IMPORTANCE

N sub-dean N sub-librarian N sub-title N sub-machine-gun

(5)

PART WITHIN

N sub-committee N sub-class N sub-routine N sub-culture V sub-divide V sub-let

(6)

LESS THAN QUANTITY

A sub-zero A sub-atomic A sub-tropical N sub-teen

A under-age A under-weight A under-populated N under-achiever V under-bid V under-pay V under-value V under-let

(7)

LESS THAN QUALITY

A sub-conscious A sub-audible A sub-human A sub-prime

A under-fed A under-developed A under-used A under-privileged

N under-clothes N under-pants N under-secretary N under-graduate N under-class

Almost all the forms to which sub- is added, in Table 6.5, are Romance; only let and teen being Germanic. The word teens was used in the seventeenth century for ‘years from thirteen to nineteen’, but teenager, for ‘a person in their teens’, did not arrive until the 1920s. It became shortened to teen and then arose sub-teen ‘of an age just below the teens’. In contrast, almost half the forms in the table which take prefix under- are non-Germanic. Carriage, secretary, graduate, class, age, populated, achiever, value, developed, and privileged all have Romance origin. Both prefixes attach to all three major word classes, but it is interesting to note that under- is on a higher proportion of verbs than sub-. Indeed,

132    6 keeping the same word class under- can be prefixed to a wide variety of verbs (generally relating in some way to a quantity); besides those in the table there are under-employ, undercharge, under-spend, under-estimate, under-state, and a myriad more. 6.4.1d We can now briefly comment on some of the meanings expressed in the sub- and under- columns of Table 6.5. In row (1), adjectives under-water and sub-marine have essentially the same meaning; only the latter has had its function extended to be a noun describing a vessel which can travel underwater. Most names for body parts and for items of clothing are Germanic and thus take under-. Besides those in the table we get under-arm, under-belly, under-shirt, under-skirt (also adjective under-hand, which has a metaphorical meaning ‘in a secret and deceitful way’). Garment was borrowed into ME from Old French in the fourteenth century and under-garment created by 1530; this was before sub- had become well established as a prefix in English. Row (4) ‘subsidiary importance’ has a varied range. The two prefixes have identical effect in sub-dean, sub-librarian, and under-secretary. Since secretary is a Romance form, why not prefer sub-secretary? (Could it be to avoid unseemly-­felt alliteration?) Under-graduate is used of a student before they have graduated, and bears a temporal sense. Indeed, the complementary term (in some countries) is post-graduate, involving temporal prefix post- (see 6.7.1). Then there is sub-machine-gun. The first piece of gunnery that fired automatically (rather than having to be re-loaded for each shot), the machine gun, was a bulky affair. When a light and portable version was produced, which could be held in the hand and fired from the shoulder, sub-machinegun seemed an apposite designation. 6.4.1e There are a few forms which are included in both columns, but with different meanings. Under-class refers to people who are socially and financially inferior. Subclass can have this meaning, but more often it is used just of a class within a class (with no implication of relative worth); for example: Transitive verbs are a sub-class within the major word class Verb. If a property is let to someone, they sign a lease. Suppose that they have to go away for a time, they may sub-let it to another person, who now signs a sub-lease with the person who signed the original lease. Under-let has a totally different sense, indicating that a property could not be let out to an acceptable extent: The agent told me I’d be able to get a good return on my investment, but it has been so under-let that I haven’t received enough to pay the mortgage.

6.4.3 over - and super -   133 6.4.2 infraThe rather minor prefix infra- also comes from a Latin preposition, infra¯ ‘below’. Its major use is in scientific nomenclature, for example, infra-sternal ‘beneath the sternum (breast-bone)’. Infra- pairs with ultra- when added to sonic, indicating audibility (see 6.4.6), and with colour terms, indicating visibility. We have infra-red ‘radiation with a wavelength greater than that at the red end of the visible spectrum’ and ultra-violet ‘radiation with a wavelength less than that at the violet end of the visible spectrum’. In everyday usage, the prefix was used (in the 1920s) to create infra-structure ‘component parts and their organization’ or ‘facilities’. It is now scarcely productive.

6.4.3 over- and super6.4.3a Over- (like under-) has been both a preposition and a prefix from OE until the present day. OE derivations which are still with us (with little change of meaning) include over-flow and over-shadow. The prefix is highly productive, with every type of word. 6.4.3b Super- is based on Latin super, which functioned as both adverb and preposition, meaning ‘over, above’. Loans into ME included superlative and superfluous, which are not analysable in English. There were also such borrowings as super-human, super-natural, and super-abundance, alongside human, natural, and abundance. These established super- as a prefix by around 1600, although its productivity did not really flourish until the nineteenth century, both in scientific and everyday coinings. 6.4.3c Table 6.6 presents a selection of the hundreds of words derived through super- and over-. Rows (1–7) have similar headings to those of Table 6.5, for sub- and under-, but the semantic correspondences are only approximate. As with sub-, most of the forms to which super- is added are Romance, the only non-Romance items in rows (1)–(7) in the table being boss and highway. And, similar to under-, prefix over- can be added to stems of every genetic hue. Rows (1)–(7) of the last column of Table 6.6 features five of the Romance forms identified for Table 6.5 (age, populated, achiever, value, and developed ), plus pass and powering. There are a good number of instances of under- and over- attached to the same form, especially in row (6)—under-weight and over-weight,

134    6 keeping the same word class Table 6.6  Comparison of some major senses of super- and over(1)

PHYSICALLY ABOVE

A super-terrestrial

A over-head N over-pass V over-hang

(2)

UPPER PART

N super-script

(3)

CLOTHING

(4)

GREATER IMPORTANCE

N super-boss N super-criminal

(5)

HIGHER ENTITY

N super-class N super-structure

(6)

MORE THAN QUANTITY

(7)

MORE THAN QUALITY

A super-natural A super-human N super-man N super-market N super-highway

A over-fed A over-developed A over-used A over-powering

(8)

TO A HIGH DEGREE

A super-sensitive A super-active A super-critical A super-curious A super-friendly A super-anxious A super-excited

A over-sensitive A over-active A over-critical A over-curious A over-friendly A over-anxious A over-excited

N over-coat N over-shoes N over-lord

A over-age A over-weight A over-populated N over-dose N over-achiever V over-pay V over-value

under-populated and over-populated, and so on. Note also under-garment and over-garment. Paired with sub-class ‘class within a class’ we get super-class ‘a class of classes’ (in biology, a unit between the defined entities phylum and class). Alongside sub-structure ‘the foundations of a building, or organization’, there is super-structure ‘the part of a building above its foundations, the framework of an organization, superimposed on its foundations’. Compare with infra-structure (mentioned in 6.4.2) ‘component parts of an organization; facilities (for example, of a school or military base)’.

6.4.4 contrasting over - and hyper -   135 6.4.3d Row (8) of Table 6.6 does not correspond to anything in Table 6.5 (for sub- and under-). Prefixes super- and over- have a special, highly productive use. They can be added to a wide range of adjectives, of every genetic origin. (Of those included in row (8), friendly is Germanic and critical is of Greek origin, with the other five being Romance; many more could be added, of each type.) Basically, super- and over- can be added to the same set of adjectives, but with contrasting meanings:

• •

super- simply indicates ‘to a high degree’ over- carries the implication ‘too much, more than is desirable’

For example, if you move into a new house and describe the neighbours as super-friendly, this implies that their behaviour is welcome and useful. But if instead you say that they are over-friendly, it implies that they act in a way that you find intrusive and unacceptable. If a child is superexcited about some coming event, you smile at such youthful enthusiasm. Whereas if they are described as over-excited, this is not really a good thing—they may act in a silly way and probably won’t be able to get to sleep at night. 6.4.4 Contrasting over- and hyperIt is interesting that the Germanic prefix over- and Greek hyper- share the same basic meaning: ‘too much, more than is desirable’. Many adjectives may take either prefix, while others appear to be restricted to one. Germanic forms generally take only over-. For example, over-careful, over-long, overclever, over-clean, over-cold, over-sweet. Adjectives of Greek origin generally only take hyper-; for instance hyper-hilarious, hyper-neurotic. The same applies to some Romance adjectives, such as hyper-alert. But dozens of Romance adjectives may occur with either prefix. For example: hyper-sensitive hyper-active hyper-critical

over-sensitive over-active over-critical

There is a contrast: typically, hyper- indicates a greater extent (almost a pathological degree) of too-much-ness than over-. If someone is over-active

136    6 keeping the same word class you may just tell them to calm down and try to take things more easily. But if they are hyper-active this is a more serious matter, maybe they should go and see a doctor (there could be something amiss with a gland). Suppose that a friend is over-critical of something you have written or done. You need not get too offended, just mention to them that your way of doing things is different from theirs. But if someone else (they would not be a friend) is hyper-critical, they are behaving in an intimidatory way, which makes you feel rather uncomfortable. 6.4.5 Contrasting super- and ultraWhen used with adjectives, the two Romance prefixes super- and ultra(6.3.4) have similar meaning: ‘extremely, to a high degree’. Potentially, each can be used with a wide range of adjectives. But there do appear to be preferences. The compendious Random House Dictionary provides, in addition to the standard entries, lists of words commencing with each of 27 prefixes. That for ultra- includes 100 adjectives and that for super- more than 350 (Flexner 1987: 2050–1, 1907–12). Just 32 adjectives are in both lists. We must be aware that ultra- could potentially be used with all 350-plus super- forms, and super- with all of those given for ultra-. But the Random House lists do point to significant tendencies, including the following:.

•

Some adjectives referring to physical properties, dimension, and speed occur with both prefixes (cool, dense, fast, pure, sharp, soft); others are only in the ultra- list (dry, heavy, hot, large, low, slow, small, wide, lightweight); none are just in the super- inventory. Human propensity terms tend to be confined to super-. They include arrogant, bold, excited, generous, indignant, intelligent, lenient, obedient, offensive, optimistic, polite, prudent, severe, strict, wise, and zealous (plus lucky and fortunate). We do however get respectable, rich, serious, and sophisticated in both lists. Religious, royalist, orthodox, and realist (referring to belief and the like) are only noted with ultra-. Precise is in both lists but related terms are confined to the super- listing: accurate, complex, elaborate, plus efficient, ideal, formal, and informal. The super- list also includes terms referring to value: dandy, elegant, handsome, and sweet (plus obese and ugly).

• • • •

6.4.6 five prefixes with sonic   137 And then there are adjectives noted by Random House with both prefixes. These include careful, civilized, confident, critical, human, modern, patriotic, powerful, rational, secret, and smart. What can the difference be? It does indeed appear to be rather slight and may well differ from adjective to adjective, and according to circumstance of use. But, overall, it seems that ultra- indicates a higher degree of the given quality than does super-. Someone who is ultra-careful exerts even more care than someone who is super-careful. And an ultra-patriotic person is even more devoted to their homeland than a neighbour who is just super-patriotic.

6.4.6 Five prefixes with sonic Prefixes which apply to sonic relate to two features of sound—its speed and its intensity. 6.4.6a The speed of sound (in dry air at 20°C) is 1,236 kilometres per hour (or, for pre-metric folk, 768 miles per hour). Three derived adjectives relate to this: sub-sonic, slower than the speed of sound super-sonic, faster than the speed of sound hyper-sonic, greater than five times the speed of sound Note that Romance prefixes sub- and super- simply indicate ‘less than’ and ‘more than’. Typically for a Greek prefix, hyper- has a more technical sense: ‘greater than five times’. 6.4.6b Sound is audible to human ears if it has a frequency of between about 20 and about 20,000 hertz. A different pair of prefixes describes deviation from this range: infra-sonic, sound with an intensity of less than 20 hertz (below the range of human audibility) ultra-sonic, sound with an intensity of more than 20,000 hertz (above the range of human audibility) Things are never totally tidy in human language. Ultra-sonic is occasionally used with the meaning assigned here to super-sonic, and vice versa. But the meanings just given are the most common and accepted.

138    6 keeping the same word class 6.4.7 supraThis prefix emanates from supra¯ which (like super) functioned as both adverb and preposition in Latin, meaning ‘over, above, on top of’. As Marchand (1969: 197) puts it, prefix supra- is ‘in most of its functions the weaker rival of super-’. It is employed in some anatomical terms, such as supra-renal ‘above (or on) the kidneys’. We find supra-national ‘transcending national limits’ and a few odd coinings by inventive writers, such as a supra-dialectal literary norm. 6.4.8 surLatin super- reduced to sur- in French and came into English through direct loans such as surfeit and surplice (which are not analysable in English). Surcame to be recognized as a prefix through borrowings such as charge and sur-charge ‘an additional charge’. Sometimes a government imposes, on its wealthier citizens, a sur-tax, so-called to distinguish it from an earlier supertax. An election or two later, a new super-tax may be unveiled, so-called to distinguish it from the old sur-tax.

6.4.9 epiSection 6.4 has so far dealt only with Romance and Germanic prefixes. Greek had epi- ‘upon, over’ which occurs in loans (many through Romance) that are not analysable in English—they include epidemic, epileptic, and epilogue. Epi- /epi-/ does function as a prefix in English, mainly in fairly technical terms such as epi-glottis ‘a cartilage which covers the glottis during swallowing’ and epi-centre ‘the place on the earth’s surface directly above the true centre of an earthquake’.

6.5 Within, between, and beyond The three Romance prefixes intra-, inter- and extra- nicely complement each other. Compare: intra-galactic travel ‘travel within a galaxy’ inter-galactic travel ‘travel between galaxies’ extra-galactic travel ‘travel outside our galaxy’

6.5.3 extra -   139 6.5.1 intraLatin intra¯ ‘within, inside’ is the source for the English prefix intra- /intrə-/ ‘within’, used predominantly with adjectives. This developed in the late nineteenth century and is used widely in science; for instance intra-molecular, intra-cellular. There is a smallish number of derivations in general use— intra-mural athletics is restricted to participants from a single college, and intra-state commerce describes trading entirely within a state.

6.5.2 interPrefix inter- /intə-/ came into English much earlier than—and is far more common than—intra-, occurring with adjectives, verbs, and nouns. It relates to Latin preposition inter meaning ‘between (in space or time), among, amid’. Loans, via French, included intercede, interject, and interpose (which are not analysable in English), together with such as inter-change, and inter-lace, enabling recognition of the prefix. From the fifteenth century, inter- was added to Germanic roots—first inter-mingle and then inter-mix, inter-marry, inter-link, and many more. This prefix has two meanings. The first is simply ‘between’, as in inter-­ continental flights and to inter-leave pages in a book. More frequently, interindicates mutual and reciprocal relations, as in adjective inter-dependent, verbs inter-breed, inter-connect, inter-mingle, and noun inter-action. And also, of course, inter-net for the global computer network linking many smaller networks and thereby allowing them to communicate with each other.

6.5.3 extra6.5.3a Whereas there is no free form corresponding to intra- or inter-, alongside prefix extra- /ekstrə-/ we also have lexeme extra, which functions as adjective and as noun. Latin had adverb and preposition extra¯ ‘outside’, which in the classical language appeared as prefix only in extra¯-ordinar̄ius, giving rise to French extraordinaire and (by 1460) English extraordinary. There are further instances of the prefix in later stages of Latin, borrowed as extra-mundane ‘beyond the material world’ and extra-mural ‘outside the walls (e.g. of a city or college)’.

140    6 keeping the same word class 6.5.3b Extra emerged as a free form in English towards the end of the eighteenth century (and at about the same time in French). One idea (oft repeated) is that it was a shortening of extraordinary| extraordinaire, but Marchand (1969: 165–6) rightly casts doubt on this. The lexeme functions primarily as an adjective meaning ‘additional’—It’s going to be cold tonight, so I’ve put an extra blanket on the bed, and If I do extra work I expect extra pay for it. The noun extra has several meanings; for example, as in I started off as an extra on the film and then they made me the star, and When you purchase a new car, be mindful of what you’re going to have to pay for all the extras! 6.5.3c Our concern here is with prefix extra-. It begins with a locational sense ‘outside of, beyond’, as in extra-galactic civilizations, extra-marital affairs, and, in a school, extra-curricular activities (e.g. sports, orchestra). We can analyse extra-ordinary as ‘beyond the ordinary’. The prefix has recently been extended to adjectives, then meaning ‘beyond the normal reference, a high degree of the quality’. For example, extrabright colours, extra-wide road, extra-strong peppermints, and extra-special occasion. In this sense, extra‑ is immensely productive; it can be used with a very wide range of adjectives. 6.5.3d Many grammarians, while regarding extra‑ as a prefix in extra-­ terrestrial and extra-marital, maintain that it is an adverb (and not a prefix) in extra-bright, extra-long, and the like. The main criterion for regarding extra‑ to be a prefix when it comes before an adjective is that we have here a single grammatical word. This is shown by the fact that nothing can intervene between prefix and adjective in extrabright, extra-careful, and so on. (That is, one cannot say *extra blindingly bright, or *extra very bright or *extra assiduously careful or *extra really careful.) And the semantics fits. The meaning of extra-clever ‘beyond being simply clever’ is not discordant with locational senses (where extra‑ is acknowledged to be a prefix) such as extra-galactic ‘beyond the limits of the galaxy’ and extra-marital ‘beyond the confines of a marriage’.

6.5.4 Contrasting ultra-, super-, extra- and very(-) 6.5.4a In 6.4.5 we noted that ultra‑ and super‑ have, when used with adjectives, rather similar meanings. A third Romance prefix, extra‑, has much the

6.5.4 contrasting ultra -, super -, extra - and very(-)   141 same import: ‘extremely, to a high degree’. We can add very /veri/, also of Romance origin, which is generally regarded as an adverb that may modify an adjective. Let us try all four of these with careful (any of several score other adjectives could equally well have been chosen): 1 2 3 4

very careful extra-careful super-careful ultra-careful

/ˌveri'keəfəl/ /ˌekstrə'keəfəl/ /ˌsu:pə'keəfəl/ /ˌʌltrə'keəfəl/ or /'ʌltrə'keəfəl/

All four magnify the quality of being careful, and it seems that they do so in the order listed—very indicates some intensification, extra‑ a bit more, super‑ even more, and ultra‑ states that it would really be impossible to be more careful. 6.5.4b A question now presents itself. If extra‑, super‑, and ultra‑ are analysed as prefixes, why should very not be too? Why indeed. Very has an interesting history. It came into ME from Old French verai (modern vrai) ‘true’. Very was first used as an adjective ‘truly, the actual one’ as in This is the very book I had been searching for. It soon came to function as intensifier to a manner adverb—The statue is very nicely carved. And then to modify an adjective as in very careful. The phonological representations in 1–4 are as these forms would be said in flowing speech. For each, primary stress is on the first syllable of careful /'keəfəl/, and secondary stress on the first syllable of very /ˌveri/, extra‑ /ˌekstrə/, super‑ /ˌsu:pə/, and ultra‑ /ˌʌltrə/. Each of these four could be emphasized, and then take primary stress. This is most likely for ultra‑ (as shown in 4), which tends to be used contrastively. The point is that each of 1–4 is, in normal pronunciation, one phonological word. It is also one grammatical word, shown by the fact that nothing can intrude between ultra‑, super‑, extra‑ and the following adjective. And the same applies for very. We can compare very with adverbs which typically modify an adjective. In each instance, something can intervene between them and the adjective: really (rather) silly, truly (quite) stupid, quite (ruthlessly) aggressive, rather (stupidly) naive, simply (frighteningly) immature. This is not possible for very. That is, it is not acceptable to say *very ruthlessly aggressive or *very fastidiously careful.

142    6 keeping the same word class Really, truly, simply, rather, and quite also bear primary stress; they must be regarded as free forms. Not so very. It satisfies all the criteria for being recognized as an affix (one phonological word, one grammatical word, and so on—see 3.2). Maybe the term ‘semi-prefix’ would be ­appropriate here. Every account of English ever written has pigeon-holed very as an adverb. To now suggest that it be accorded the status of a prefix may seem scandalous. Well, my analysis can be accepted, on its merit. Or rejected, on the basis of tradition.

6.6 Further locationals We can now examine twelve further locational prefixes. Four make up pairs of different genetic origin (Greek/Romance for peri‑| circum‑, Romance/ Germanic for trans‑| th(o)rough) but most are singletons. 6.6.1 paraPrefix para‑ /parə‑/ goes back to para ‘beside’, which was both a preposition and a prefix in Greek. A number of loans are unanalysable in English, including parallel, literally (in Greek) ‘one beside another’, paralysis ‘beside being “loosened” or disabled’, and paranoia ‘beside mind’. In present-day English, para‑ is used with nouns and adjectives and has a variety of meanings. Para-language (and para-linguistic) refer to ancillary features of vocal communication, such as timbre of voice, facial expressions and hand gestures. In similar fashion, para-normal refers to ghosts and things beyond what everyone is normally aware of. Similarly for paraphysical, para-military. The prefix has a different sense in para-medic, someone who has limited medical skills, and can work beside a doctor or render assistance until a doctor comes. A similar meaning pertains in para-professional, para-legal. And para-typhoid, a disease with some symptoms similar to typhoid but being overall milder. Over the last couple of millennia, there have been various contraptions to slow the rate of fall of a heavy object. The name para-chute was coined in French at the end of the eighteenth century, the chute meaning ‘fall’. Adaptations based on this include para-brake, a parachute acting as a brake to assist a plane in landing, para-troop(er), soldiers using parachutes (and a further sense of para-medic, a medic in the paratroops).

6.6.3 peri - and circum -   143 6.6.2 teleWords in English commencing with tele‑ /teli‑/ have been coined using the Greek prefix tele‑ ‘far of f ’, added to predominantly Greek and Latin forms. Tele‑ is, like para‑, much used in scientific naming. The name tele-scope was created in the early seventeenth century, and words such as tele-pathy and tele-kinesis followed a couple of hundred years later. Tele‑ only really took on the status of a prefix, ‘at a distance’, with the coining of tele-phone and then tele-vision. These two nouns have led to many adaptations. Tele-lecture, tele-conference, tele-marketing, and tele-shopping all involve a telephone, while tele-cast (television broadcast), tele-drama, tele-camera relate to television. Verb tele-vise is a back-formation. And, a hundred years after photo-genic (1839), there came the adjective tele-genic ‘comes out well on television’.

6.6.3 peri- and circumPrefix peri‑ /peri‑/ relates to Greek adverb and preposition peri ‘around, about’ while circum‑ /se:rkəm‑/ is from Latin adverb and preposition circum, with very similar meaning. In English, peri(‑) and circum(‑) have slightly different import and are never attached to the same form. For example, the circumference of some area is the measure around its perimeter. Typically for these genetic types, peri‑ is the prefix most used in scientific nomenclature. There are many well-established loans that are not analysable in English, such as periphery, peripatetic, and perimeter. From the nineteenth century, peri‑ was used productively to derive specialist adjectives, including peri-natal ‘relating to the time just before and just after birth’ and peri-glacial ‘happening on the margins of a glacier’. Circum also commences many borrowings which are not analysable in English, such as circumcise (etymologically ‘cut around’), circumscribe (‘write—that is, draw a line—around’), and circumspect (‘look around’). Then there were loans such as locution, navigate, and polar, followed by circum-locution ‘talking in a roundabout way’, circum-navigate ‘navigate right round something (typically the earth)’, and circum-polar ‘around or near a pole’, establishing circum‑ as an English prefix with nouns, verbs, and adjectives (almost all Romance forms). It is much used in astronomy with the sense ‘revolving around’—circum-lunar, circum-solar, circum-stellar.

144    6 keeping the same word class 6.6.4 th(o)rough- and trans6.6.4a From its earliest stages, OE had preposition and adverb Þurh ‘through, by means of’. In later OE this developed into Þuruh, which was progenitor for modern forms through and thorough. For many centuries, prepositions/adverbs through and thorough were alternatives, with essentially the same meaning. The OED quotes, from 1684, An hole thorough the bottom of the Vessel and, from 1594 (Shakespeare’s Lucrece), Through the length of times he stands disgraced. (Through could have been used in place of thorough in the first example, and vice versa for the second.) In 1803 Wordsworth wrote O’er hilly path and open Strath, We’ll wander Scotland thorough (where now one would say through). There was then specialization, with thorough used just as adjective ‘in a careful and detailed manner’ (The detective made a thorough investigation of the crime-scene) and through as preposition ‘from one end to the other’ (He wandered through the meadows) and adverb (She cried the night through). 6.6.4b From an early stage each also functioned as a prefix. Thorough‑ /θʌrə‑/ is found in thoroughfare, from the fourteenth century, on down to thoroughgoing, from the nineteenth. These are not really segmentable in Modern English, and thorough‑ is no longer a productive prefix. However, prefix through‑ /θru:‑/—with very similar meaning to the preposition/adverb, ‘from one end to the other’—is used in several ways, generally attached to a noun. A through-draft comes in one window and goes out of another. One can move unimpeded along a through-street, since traffic on crossstreets has to give way. A through-lounge extends from front to back of a house. The advent of travel by rail, steamship, bus, and air has led to many new derivations. On a through-train one does not have to change, a through-­ passenger is going all the way and has bought a through-ticket. Also with ‑bus, ‑plane, ‑carriage, ‑journey, ‑flight, ‑traffic, and no doubt more. 6.6.4c Prefix trans‑ /trans‑/ has the meaning ‘across, beyond, from one place or profile to another’, complementary to that of through‑ ‘from one end to the other’. It comes from the Latin preposition trans ‘over, across, on or to the other side of’. (Trans‑ is mostly used with adjectives and verbs, through‑ with nouns.) Early loans included translate, transparent, and transfer. Trans‑ was recognized as a prefix in English through early borrowings such as marine

6.6.5 out-, in-, off-, on-, up -, and down -   145 and trans-marine, figure and trans-figure. It was soon added to Germanic forms—trans-plant is recorded from 1440. Verbs tranship (trans-ship) and trans-locate are, like trans-plant, ‘move from one place to another’. Adjectives trans-border and trans-Alpine indicate ‘across a line or range’, while trans-continental and trans-Atlantic are ‘across an area’. Trans-literate ‘change from one orthography to another’ follows the example of (the unanalysable) transmute ‘change one type of thing into another (notoriously, change lead into gold)’. And there is a predilection which has only recently been named: trans-sexual ‘wanting to change to be of the opposite sex’. 6.6.5 out-, in-, off-, on-, up-, and downA number of Germanic forms have, from OE or early ME times, functioned as preposition, adverb, and also derivational prefix. We discuss just the main functions and meanings of each prefix. Most of the occurrences of these prefixes do not change word class. But, for every one of them, there are some instances of word-class shift. These are mentioned for each prefix, and summarized in Table 6.7, at 6.6.5f. 6.6.5a Prefix out‑ /aut‑/ has two main senses. The first relates to the meaning of the adverb outside or out. The prefix may be added to a noun—an outbuilding is outside the main building, an out-patient lives outside a hospital and just calls in for treatment, an out-law is a criminal who operates outside the law. Or it may be attached to a verb—to out-stretch one’s hand is to stretch it out. A few examples of out‑ (in its first sense) effect a change in word class. If water or laughter bursts [verb] out, we can say that there is an out-burst [noun] of water or laughter. Similarly, if water pours [verb] out, we have an out-pour or out-pouring [noun]. Someone who is cast [verb] out of a society becomes an out-cast [noun]. If something is not within the doors [noun] of a house, it is out-doors [adjective]. And if something is spread [verb] out, then it is out-spread [adjective]. In its second sense, out‑ is attached to a verb and means ‘do more than [the referent of the object noun phrase]’. Mary out-bid John at the auction means that Mary bid more than John did. In this sense, out‑ may be added to hundreds of verbs—out-run, out-shoot, out-weigh, out-number, out-drink,

146    6 keeping the same word class out-live, and so forth. Plus, of course, the general word out-do (Mary out-did John in everything). The second sense always involves out‑ being prefixed to a verb and deriving another verb, but the argument structure is adjusted. This is discussed further in 7.1c. 6.6.5b The meaning of prefix in‑ /in‑/ corresponds just to the first sense of out‑. For example, noun in-land, land in from a coast, and verb in-lay, to lay something just inside something else so that their edges are level (for example, pieces of pearl as ornamentation on a posh piece of furniture). Like out‑, it can either maintain or change word class. Some derivations parallel those for out‑, including in-doors and in-patient. Note also in-tray, tray for documents just come in, and out-tray, tray for documents ready to be sent out. Noun in-put refers to what is put [verb] in and out-put to what is put out. However, there are some interesting differences. Out-house is a noun, describing an out-building used as a toilet, whereas in-house is an adjective—an in-house magazine is produced just for circulation within a company (within its ‘house’). Immigrate ‘migrate to’—and emigrate ‘migrate from’—are of a quite different genetic hue, being reflections of Latin immigra¯re ‘move into’ and e¯migra¯re ‘move from’ respectively.

6.6.5c Prefix off‑ /ɔ:f‑/ ‘away from’ may be used with verbs, adjectives, and nouns and has meaning ‘away from’, similar to that of the preposition/ adverb. A truck is loaded [verb] when a load [noun] is placed in it; once the destination is reached, the loan will be off-loaded [verb] off the truck. With a colour adjective, off‑ indicates ‘not the typical shade of that colour’. For example, off-yellow may be used for a rather wishy-washy colour that resembles yellow. And the more general off-colour describes a person looking or feeling unwell. Added to nouns we get off-shore ‘away from the shore’, off-season ‘away from the ideal season’ and off-Broadway ‘away from the main theatre district, on Broadway’. Each of these off-derivatives can function as noun or as adjective. However, some derivations from a noun are only adjectival—a musical note being off-key, and a policeman being off-duty. 6.6.5d Prefix on‑ /ɔn‑/ is less common than off‑; its meaning is always close to that of adverb/preposition on. There are a number of adjectives, derived

6.6.5 out-, in-, off-, on-, up -, and down -   147 from nouns, which pair the two prefixes. We can contrast a star’s on-screen performance with their off-screen persona; similarly with stage, course, track, and site. In cricket the side [noun] of the field where the batsman’s legs stand is the on-side [noun] (alternatively, the leg-side), with the other side being the off-side. We also get on-duty and on-shore, the complements of off-duty and off-shore. There are a few other on‑ derivations. For instance, an on-looker [noun] is a person who looks [verb] on. If some activity is going [verb] on at the present time, it can be described as on-going [adjective]. 6.6.5e Prefixes up‑ /ʌp-/ and down- /daun-/ retain the meanings of the corresponding prepositions/adverbs. Much of the time they pattern together. Added to noun river we get adjectives up-river and down-river. Similarly with stream, hill, wind, draught, train, stairs, and town. With noun trend we get nouns up-trend and down-trend; similarly with turn and swing. The two prefixes may also be added to verbs (not changing word class); one can improve a profit forecast by up-grading it, or weaken it by down-grading (and from these we get nouns up-grade and down-grade). Some verbs only allow one of the prefixes, because of their meaning. One may only up-root [verb] a plant—pull up its roots [noun]—and rain can only create a down-pour [noun], as in pours [verb] down. Other derivations are also confined to just one prefix. If you travel upcountry [adverb] you are going to a part of the country [noun] which is up and away from the centre of population. An up-rising [noun] occurs when people rise [verb] up against their rulers. This may lead to the down-fall [noun] of a dictator—in an entirely metaphorical sense, he falls [verb] down from his position of power. As for all Germanic—and other—prefixes, there are also compounds beginning with the same forms (they are compounds since their meaning cannot be inferred from the meanings of the component parts); for instance, upshot, upset, uphold, down-home, and downhearted. And we find totally different meanings for upright ‘absolutely vertical or totally law-biding and honest’ and downright, which may intensify an adjective (rather like very or really), as in It’s downright false, or a noun, as in It’s a downright falsehood. Note that downright is only used to emphasize an unwelcome property (this being the indirect contribution of the down component).

148    6 keeping the same word class Table 6.7  Word-class effect with out-, in-, off-, on-, up-, and downDERIVING: FROM

NOUN

NOUN

out-patient in-land off-season on-side up-trend down-turn out-burst in-put

VERB

on-looker up-rising down-pour

VERB

up-root out-stretch in-lay off-load up-grade down-grade

ADJECTIVE

ADJECTIVE out-doors in-house off-screen on-stage up-river down-hill out-spread

on-going up-lifted down-trodden off-yellow

6.6.5f It is useful to summarize the word-class maintaining and changing possibilities for these six prefixes in Table 6.7. (Most of the examples repeat some just given, a few are added.) There appear to be few verbs derived from nouns, and nothing derived from an adjective except for off- with colour terms. A fully exhaustive search might uncover further possibilities, but this table provides an indication of the possibilities. 6.7 Temporals The main prefixes with a temporal meaning are set out in Table 6.8; all of them also have a locational sense. As before, it is useful to arrange them in Romance and Germanic columns; prefixes in a given row have some similarity of meaning. Table 6.8  Temporal prefixes ROMANCE

GERMANIC

OUTLINE MEANING

post- /poust-/

after- /a:ftə-/

behind; following

pre- /pri:-/, ante- /anti-/

fore- /fɔ:-/

(in) front of, in advance, preceding

retro- /retrou-/

back- /bak-/

back; in the past

mid- /mid-/

in a central position (with respect to space or time)

6.7.1 after - and post -   149 Three purely temporal prefixes of Greek origin, proto‑, paleo‑, and neo‑, have limited productivity in English, and are discussed in 6.7.7. There was a prefix pro‑ ‘before, ahead, in front’ in Greek which is reflected in such unanalysable English words as prophet and prologue (this is quite different from the Romance prefix pro‑ ‘instead of, in favour of’, discussed in 6.8.2). Enclitic (first recorded in 1656) and proclitic (1846) are Greek terms which entered English via Latin. In the first edition of his classic textbook Morphology (1946), Eugene A. Nida created clitic. On the basis of this back-formation, one might wish to recognize pro‑ and en‑ as prefixes in English.

6.7.1 after- and post6.7.1a OE had adverb after referring to time, preposition after which could relate to time or space, and also derivational prefix after‑. Today, the prefix can have spatial import ‘behind’—for example after-brain is used for the metencephalon, the hind portion of the brain—but more commonly its meaning is temporal, ‘following on from’. Since ME times, after‑ has been used for portions of a ship ‘relating to rear part, towards the stern’: after-deck, after-cabin, after-mast, after-bulkhead, after-hatches, and a number more.

6.7.1b Prefix post‑ reflects the Latin adverb/preposition post ‘behind, after’. There were early loans such as posterity and postpone, but the prefix did not become commonly used until the end of the sixteenth century, with derivations such as post-predicament (a technical term in logic, describing relations considered after Aristotle’s basic categories or ‘predicaments’) and postdate. Like after-, prefix post- has a spatial meaning—as in post-abdomen ‘the rear portion of the abdomen (especially in insects and crustaceans)’—but more commonly its meaning is temporal, ‘a period of time following after’. 6.7.1c It is instructive to compare the two prefixes, to try to understand which is used when, and why. To assist with this, Table 6.9 presents some sample instances, showing the word class of each derivation by A, N, or V. Four general comments can be made. The first is a general tendency for Germanic prefix after‑ to occur with Germanic roots and Romance prefix post‑ with Romance roots (although all other possibilities are attested). We can also note that the great majority of after‑ derivations are nouns, plus a few adjectives. In contrast, most post‑ derivations are adjectives, together with a few nouns (and just one or two verbs).

150    6 keeping the same word class Table 6.9  Sample instances of after- and post1

N

after-brain

Rom

N

post-abdomen

Rom

2

N

after-birth

Gmc

A

post-natal

Rom

3

N

after-care

Gmc

A

post-operative

Rom

4

N

after-life

Gmc

A

post-colonial

Rom

5

N

after-effect

Rom

A

post-traumatic

Gk

6

N

after-thought

Gmc

V

post-date

Rom

7

N

after-party

Rom

A

post-electoral

Rom

8

N

after-word

Gmc

A

post-doctoral

Rom

9

N

after-heat

Gmc

V

post-heat

Gmc

10

A

after-school

Gk

A

post-school

Gk

11

A

after-lunch

Gmc

A

post-lunch

Gmc

12

N, A

after-war

Rom

A, N

post-war

Rom

The third observation relates to orientation. For after‑, there are basically two possibilities: (1)  after-X refers to an X which follows something else. For example, afterlife is life after death, an after-effect is a delayed effect of some happening (for example, gas poisoning), an after-thought is an additional thought that occurs to someone after a discussion had been considered concluded, an after-word is a few words at the end of a book (a bit like a postscript), and after-heat is the heat generated by a nuclear reactor after it has been shut down (or by some other generator of heat). (2)  after-X refers to something which follows an X. For example, afterbirth is the placenta and foetal membrane which are expelled after a birth. Also after-lunch, and after-school. After-war is generally used in sense (2), referring to the period following a war. But it is sometimes employed in sense (1), for a war which follows something else (often, an earlier war). The OED quotes (from 1981): The two world wars brought about not only the most fundamental changes in the political systems of the world but also the afterwars in Korea, Vietnam, and other areas of the world. Some schools provide care, after classes have finished, for children whose parents are still at work. This may be called after-care—using after‑ in sense

6.7.1 after - and post -   151 (1)—or after-school care—this is sense (2). Similarly, a party following the conclusion of some event such as a show may be called either an afterparty—sense (1)—or an after-show party—sense (2). Whereas the majority of after-derivations are of type (1), we find that type (2) predominates for post‑. Post-electoral euphoria is something that follows an election, and post-traumatic stress eventuates some time later than the trauma. Similarly for post-natal, post-operative, post-colonial, postdate, post-doctoral, post-lunch, post-school, and post-war. The verb post-heat is an exception, using sense (1); the Random House Dictionary defines it as ‘to heat (a metal piece, as a weld), after working it, so as to relieve stresses’. Fourthly, we can compare the typical referential scope of the two prefixes. Most instances of post‑ relate to a period of time or some quality that extends over such a period—the post-Elizabethan age, the post-­apartheid period, postChristmas lassitude in sales, the post-cold-war atmosphere, post-independence problems. In contrast, after‑ derivations tend to refer to a happening at some specific time—an after-shock from an earthquake, the after-glow when the sun has just gone down, an after-show party (or after-party). In the first eight rows of Table 6.9, after‑ and post‑ are added to different forms, in rows 9–12 to the same forms. Distinctions of meaning here illustrate the semantic tendencies just outlined. We have already explained the difference between noun after-heat and verb post-heat. After-school and post-school are both adjectives, but with entirely distinct meanings. After-school refers to some happening after the end of a school day—after-school activities or after-school care. But postschool refers to a period of time following the end of formal schooling— post-school life, post-school education. After-lunch and post-lunch are to a great extent interchangeable. Nevertheless, after‑ is more likely for some specific activity (after-lunch speech, after-lunch drive) and post‑ for something less demarcated (the OED illustrates with post-lunch siesta and post-lunch nap). After-war and post-war show similar possibilities; either can modify period, or conditions, or map of Europe. But to celebrate an outbreak of peace one can only have an afterwar party, not a *post-war party (if this were said, one might infer that the party went on for years). 6.7.1d Most additions of after‑ and pre‑ do not affect word class. This applies to those in rows 1–9 of Table 6.9. Note that noun after-heat is based on noun heat and verb after-heat on verb heat. However after-school, post-school,

152    6 keeping the same word class after-lunch, and post-lunch are adjectives derived from nouns. (Both afterwar and post-war do double duty as noun and as adjective.)

6.7.2 fore6.7.2a It is reasonable to enquire why there is no derivational prefix before‑, alongside after‑. In fact the prefix parallel to after‑ is fore‑. OE had adverb/ preposition fo¯ r or fo¯ re ‘before’. By combining this with prefix be‑ ‘by about’, the complex adverb/preposition befo¯ ran ‘before’ was created. It is the longer form which has evolved into modern word before, while fore‑ became a productive derivational prefix. 6.7.2b A handful of forms inherited from OE are scarcely analysable today—foreboding, forestall, and forebears. However, dozens are, and they can usefully be organized in terms of meaning and word class. (1) Locational sense ‘front, front part of, in front of’, used with nouns:

•

Body part terms, including head (fore-head is the front part of the head), lock (of hair), tooth, arm, finger, skin, and also leg, foot, and paw (for four-legged creatures). And, like after‑, for parts of ships: deck, mast, sail, yard, and others. Terms referring to the environment and buildings, including shore (fore-shore is what lies between low and high tide marks, or what is between water’s edge and cultivation), land, ground, edge, court, stage. We also get fore-name, the name that precedes a family name (John in John Smith) and fore-word, a few words at the front of a book, introducing it. In ME, fore-man described someone who goes in front, and leads. It gradually lost this sense and came to refer to a leader in a different sense, someone who is in charge of a group of workmen, or who chairs a jury. Fore-woman followed in the early eighteenth century and then, in the anti-chauvinist landslide of the 1970s, we got fore-person.

•

• •

(2) Temporal sense ‘do at some time in advance’, used with verbs. If you fore-warn someone you warn them that something will happen before it actually happens; a soothsayer fore-tells what will happen in the future. Similarly with see, ordain, and taste.

6.7.3 pre - and ante -   153 (3) Temporal sense, used with adjectives and nouns. Some of these are derived from verbs mentioned under (2)—adjectives fore-warned, fore-told, nouns fore-warning, fore-sight, fore-taste, and so on. For some adjectives and nouns there is no corresponding verb with fore‑; they include forethought, fore-play, fore-runner, fore-cited. These derivations have varying meaning. For some it is similar to the verbs in (2)—a fore-warning is given prior to the time at which it applies, fore-thought is thinking applied in advance; similarly for fore-told, foretaste, fore-sight. But fore-play describes caresses of arousal that take place over a period of time right up to coition. Fore-noon is the period up until midday. And if something is fore-cited this does not mean that it has been cited before it would be expected to be, rather that it was cited sometime before the present. 6.7.2c It will be seen from the examples quoted that fore‑ is used mainly with Germanic stems. Of the thirty or so forms which take fore‑ just mentioned, only court, stage, person, ordain, and cite are of Romance origin, and each of them was taken over early, in ME times. (See also 6.7.4.)

6.7.3 pre- and ante6.7.3a Prefixes pre‑ and ante‑ go back to Latin prepositions/adverbs—prae ‘front part of, ahead of’, which required ablative case on a following noun, and ante ‘before, in front of’, which took accusative. They both also functioned as prefixes, with similar but by no means identical meaning. The two prefixes in English show a degree of semantic overlap. They are sometimes in competition (and are in a number of cases interchangeable). It appears that ante‑ is gradually receding, losing ground to its rival. Note that Latin prae is cognate with Germanic fore‑, both emanating from proto-IndoEuropean *per(i) ‘forward, through, in front, before’.

6.7.3b Romance forms commencing with pre were plentiful in ME; they include pretend, prejudice, pregnant, and precise (none analysable in English). Loan doublets such as destine and pre-destine, suppose and pre-suppose established pre‑ as a productive prefix in ME times. To begin with, pre‑ was typically added to Romance forms but, as it came to be used more and more, every kind of stem fell under its purview. We can survey the range of meanings.

154    6 keeping the same word class (1) Locational sense ‘front part of, in front of’, with a noun or adjective. In contrast to fore‑ and ante‑, there are relatively few of these. A pre-molar is at the front of the set of molar teeth, the space in front of an altar is described as pre-altar, anything which comes before a consonant is pre-consonantal, and a set of grammatical items dubbed pre-positions are positioned before a noun phrase. (2) Temporal sense with verbs ‘do at some time in advance’. If you pay someone for a piece of work before they have actually done it, you prepay them. If someone arranges the form their funeral should take, and the music to be played, they are pre-arranging it. Similarly with establish, determine, conceive, heat, cook, record, pack, and many more. Verb pre-exist has a different meaning, ‘happened at some time in the past’, as in I believe that, in an earlier incarnation, my soul pre-existed in an antelope. (3) Temporal sense ‘existing in advance’ with adjectives and nouns. Many of these are derived from verbs listed under (2)—nouns pre-payment and pre-conception, adjectives pre-established, pre-cooked, pre-recorded, and pre-packed. In addition, we have the euphemistic description pre-owned for second-hand goods, and pre-condition for a prerequisite which must be satisfied, as in A pre-condition for the renewal of peace talks is repatriation of all prisoners. (4) The most frequent use of pre‑ is added to adjectives, relating to a period of time before some event or epoch, or a quality associated with such a period. For instance, pre-Christian times, pre-scientific ideas, pre-industrial society, pre-menstrual tension, pre-trial discussions, the pre-dawn chorus of birds. In this sense it parallels post‑ (6.7.1b–c); by and large, the two prefixes can be used with the same stems—pre‑ or post‑ with colonial, Elizabethan, marital, trial, and literally hundreds more. As with post‑, there are derivations of this type in which pre‑ is added to a noun, and derives an adjective; for example, pre-birth nausea, pre-­ performance nerves. (5) With just a few stems, pre‑ can have a superlative sense. Someone who is pre-eminent in their profession is the most eminent of all. Similarly with pre-dominate. (And their nominalizations, pre-eminence and pre-dominance.)

6.7.3 pre - and ante -   155 Interestingly, one of the uses of prefix prae‑ in Latin was to add an intensive sense to an adjective. There were altus ‘high’, prae-altus ‘very high’; frı¯gidus ‘cold’, prae-frı¯gidus ‘very cold’; and a good number more.

6.7.3c Derivational prefixes with quite different meanings may have much the same semantic effect. Compare sub‑ ‘less than’ (6.4.1) and pre‑ ‘before’ attached to teen ‘someone aged from thirteen to nineteen’: sub-teen ‘someone whose age is less than thirteen’ pre-teen ‘someone in the period before the age of thirteen’ The two prefixes approach from different semantic directions and produce what are effectively synonyms. 6.7.3d Prefix ante‑ came into English much later than its Romance congener pre‑. From the early seventeenth century, we find it added to existing Romance loans, as in ante-chamber. There are two main meanings: (1) Locational, used with nouns, ‘a small introductory place (typically, situated in front of the main place)’. Thus, an ante-room is a small room leading into a larger room. In this sense, ante‑ is used with many terms for parts of churches (which are invariably of Romance origin): chapel, choir, nave, and so on. (2) Temporal sense ‘during a period preceding’ with adjectives. For example, ante-mundane ‘before the creation of the world’, ante-classical ‘before the classical period’. 6.7.3e Negative prefix anti‑ /anti‑/ ‘alternative to, reverse of, opposed to, used against’, which is of Greek origin, was discussed in 5.24. In many dialects of English, it has the same pronunciation as Romance prefix ante‑ /anti‑/. The two are distinct (to a discerning eye) in written form, but are undistinguished when spoken. The discourse context will generally make clear which is intended—but not always. Suppose that John and Mary were both destined by their parents to go to college. But John rebels, leaves school and for some years works on a construction site before eventually deciding to study further. Nowadays, John refers to the interregnum as his anti-college years /ˌanti-'kɔlidʒ 'yə:z/.

156    6 keeping the same word class But Mary went to college straight after high school and refers to her teens as the ante-college years /ˌanti-kɔlidʒ'yə:z/. To an outsider, there could well be confusion. This homophony is certainly one reason leading to the gradual decline of ante‑. Alongside post-natal, for the period after childbirth, both pre-natal and ante-natal are in common use to refer to medical care in the period before a birth. There is no danger of ambiguity here—*anti-natal ‘opposed to birth’ just does not make sense. But alongside post-war, one is only likely to encounter pre-war. Ante-war /ˌanti-'wɔ:/ is a homophone with anti-war / ˌanti-'wɔ:/, and since anti‑ is the more common suffix (and lots of people are anti-war) it is the negator reading which would prevail. Thus, ante-war is avoided in favour of pre-war. 6.7.3f We can now investigate the possibility of pre‑ and ante‑ being added to the same stem. In fact both occur with fix, but with entirely different meanings. Noun ante-fix was coined in the nineteenth century as a technical term in architecture, describing an ornamental addition to eaves or cornices to conceal a joint. Pre-fix was borrowed from French, in the early fifteenth century, as a verb with temporal meaning ‘fix or arrange or determine in advance’. It then took on a locative sense ‘put in front of’, as in After the conferral, I can pre-fix Sir to my name. Use in grammar then followed (reflecting Latin prefixum, French pré-fixe), as both noun and verb. Although pre-fix (literally ‘fixed in front’) can be analysed in present-day English, this does not apply to suffix or to affix (both related to forms in Modern Latin and in French).

With other stems, pre‑ and ante‑ have essentially the same import. Verb antedate has two main senses—either assign an earlier date than the actual one (He ante-dated the document, to pretend that it had been written before charges were laid against him), or precede (The rise of science in China ante-dated its emergence in Europe). Pre-date has the same two senses. (The OED has one instance of fore-date in the second sense, but this derivation is not in common usage.) Pre-natal and ante-natal are, as just mentioned, also substitutable synonyms, as are pre‑ and ante‑ derivations with nuptial, prandial, and mundane. In any competition there is likely to be a winner and here it is pre‑ which is out in front. This is most pronounced in American English, where ante‑ forms are seldom used (despite being listed in dictionaries such as Webster’s

6.7.4 contrasting pre - and ante - with fore -   157 and Random House). The only ante form in common currency is the Latinism (not analysable in English) antebellum relating to the period before the American Civil War. Ante-date came into English around 1600 and pre-date did not follow until 250 years later. But usage is no respecter of history: pre-date is today ten times more frequent. 6.7.4 Contrasting pre- and ante- with foreRomance prefixes pre‑ and ante‑ and Germanic fore‑ show differences in function and meaning, but also some similarities. 6.7.4a Whereas pre‑ has only limited locational use, both fore‑ and ante‑ ­exhibit a fair amount. There is, however, little overlap. Fore‑ ‘front, front part of, in front of’ is used with mostly Germanic nouns referring to body parts ( fore-arm), the environment ( fore-ground ), buildings ( fore-stage) and a few more. Ante‑ ‘a small introductory place’ appears with Romance words (ante-chamber). We do find both ante-court for the first of a series of courts, and fore-court, which can have the same meaning or may refer to the front part of a court. (And see 6.7.5c.) 6.7.4b There is significant overlap concerning the temporal senses fore‑ and pre‑. The same basic meanings attach to fore-warn and pre-warn. Similarly when these prefixes are used with judge, ordain, taste, and cited (all Romance) or tell, see, shadow, show, doom (like warn, all Germanic). With some stems, either prefix could be employed but only one is in common usage. Why do we generally say pre-dawn and fore-noon (rather than fore-dawn and pre-noon)? English has simply settled into this pattern of speaking, as languages do. There is one root which takes both prefixes but with very different meanings. The Germanic verb cast was taken into ME from Old Norse. Its basic meaning was ‘throw’ (as in cast a line when fishing, cast ashore after a shipwreck, cast off clothing) but there was a wide range of senses. One referred to mental activity ‘consider, plan’ (perhaps ‘throw the mind’) and in the early 1400s this led to verb (and later noun) fore-cast ‘think of beforehand, predict’. Another sense of cast, from the early 1500s, described heating some substance (such as metal or glass) until it became liquid and then pouring it into

158    6 keeping the same word class a mould where it would harden into the desired shape. Four hundred years later, pre-cast was coined, for something (metal, concrete) being cast in advance, before being placed in its intended position. 6.7.5 retro-, back-, and hind6.7.5a Prefix retro‑ relates to Latin adverb retro¯  ‘backwards, back, behind’. It has never been very much used in English and is productive mainly in anatomical terminology, such as retro-lingual ‘behind the base of the tongue’. An early borrowing, about 1400, was retrograde indicating ‘go in the opposite direction to that expected, in space or time’—an army fleeing from defeat may make a retrograde march back to base, and changing something for the worse is a retrograde step. This word is unanalysable in English (it cannot be related to verb grade, a later loan with a different meaning). Later non-segmentable loans, from the seventeenth century, included retrospect and retrogression. Active, action, and act were all borrowed from French in ME times. Prefix retro‑ came into effect with the introduction, two or three hundred years later, of retro-active, retro-action, and retro-act. The meaning can be ‘decided now but applying at some time in the past’—as in We have just been awarded a retro-active pay rise, effective two months ago—or ‘in a backwards direction (opposing earlier motion)’—They made a retro-migration back to the original homeland. When, in the space age, an auxiliary rocket engine was attached to a spacecraft, which would fire in the opposite direction to the normal engines and slow the craft down for re-entry, it was called a retro-rocket. In Latin, post‑ was the complement of ante‑ for spatial reference where no movement was involved, and retro‑ the complement of ante‑ where there was movement (Michèle Fruyt, personal communication).

6.7.5b Germanic form back has functioned as both noun and adverb from ME days. It also takes part in many compounds. And back‑ is a prefix with recurrent meanings. The locational sense is ‘at the rear of’. A back-yard is the yard behind a building, a back-room is a room in the rear part of a building, and back-woods is used of pristine land, typically wooded, at ‘the back of beyond’. There is also a temporal sense, relating to time in the past. Back-pay is remuneration for work already done. A back-log describes tasks which should

6.7.6 mid -   159 have been attended to in the past. Verb back-track means going back over a decision or undertaking made in the past and re-assessing it. 6.7.5c Prefix retro- has limited use with Romance forms and back‑ is mainly applied to Germanic roots. There appears to be no direct overlap. However, a sentiment may sometimes be expressed in either way. Back-date has recently been innovated as an alternative to pre-date and ante-date. To say that a payrise is back-dated is similar to saying that it is retro-active or retrospective. In the game of tennis, back‑ contrasts with fore‑ in two ways. In a fore-hand stroke the palm (front) of the hand is towards the net, and in a back-hand stroke the back part is. Also, the front part of a tennis court is known as the fore-court, as opposed to the back-court. (In basketball the terms are front-court and back-court.)

6.7.5d Germanic prefix hind‑ /haind‑/ ‘back part of’ is found with just a few Germanic roots. Basically, it is the complement of fore‑ in its sense (1) ‘front, front part of’ with body part terms (6.7.2b). Alongside fore-leg and fore-quarters there are hind-leg and hind-quarters. Other derivations include hind-brain, hind-wing, hind-gut. And there is hind-sight, ‘being able to understand things in retrospect, which were not clear at the time’; compare this with fore-sight, ‘ being able to anticipate what is going to happen in the future’. (Preposition behind was created, in OE times, from bi- ‘by’ and hinder ‘back, behind’.)

6.7.6 midPreposition mid ‘with’ in OE also functioned as a prefix. Only one of its derivations has survived today—midwife, literally ‘with woman’. Modern prefix mid‑ is a different matter, being a development from the non-cognate OE adjective midd(el) ‘middle’. This remains an adjective in its long form, and has developed into a monosyllabic prefix mid‑ ‘in the central portion of, with respect to space or time’. In its locational sense, mid‑ may relate to any area or mass: we get midland, mid-town, mid-ocean, mid-stream, mid-air, mid-brain. It can be added to a proper name—mid-Atlantic refers to the central part of an expanse of ocean. Compare this with the use of adjective middle in the Middle Atlantic, describing the central part of the Atlantic coast of the USA (which is not an area), and the states which adjoin it: New York, New Jersey, and Pennsylvania.

160    6 keeping the same word class The temporal sense is even more pervasive. Mid‑ can be prefixed to any noun that may refer to a period of time: mid-day, mid-night, mid-summer, mid-season, mid-century, mid-’twenties, mid-race, mid-stride, mid-flight, midcareer. People sometimes have a mid-life crisis in their middle age—life here describes a period of time and thus takes mid‑, while age has numerical reference and may only be modified by adjective middle. A further use of mid‑ is with colours. For something to be described as mid-green indicates that it is in the central part of the range of hues which could be designated as green. 6.7.7 proto-, paleo-, and neoWe now consider three prefixes which have exclusively temporal meanings. They go back to Greek but have sometimes entered English through a Romance conduit. 6.7.7a Prefix proto‑ /proutəu‑/ is a continuation of Greek prefix pro¯ to‑ ‘first, chief’. It was used as a prefix in English from about 1600, added to a lexeme of Greek origin in proto-type, and to Romance forms in proto-rebel and proto-traitor. It came into wide use in the nineteenth century, for scientific labels such as proto-galaxy and linguistic names such as proto-Semitic. An early meaning ‘leader of a group’, as in proto-rebel and proto-traitor has dropped out of general use, together with these words. Proto-type exhibits the modern meaning ‘a preliminary stage of’. (Interestingly, the meaning of proto-typical ‘absolutely typical’ does not relate to that of proto-type in the way that typical does to type.) A proto-literate society is at an early stage of acquiring literacy (or else it employs a primitive form of writing). A proto-galaxy describes a concentration of gas and dust which may evolve into a galaxy. Proto-Indo-European is the putative ancestor from which all attested Indo-European languages are thought to have evolved. Similarly for proto-Germanic (a descendant of proto-Indo-European), proto-­ Dravidian, proto-Austronesian, and so on. There is no written record of a proto-language; it is a reconstruction, by systematically comparing forms across m ­ odern languages. The reconstruction includes proto-­phonemes, proto-lexemes, and sometimes even proto-word-order. 6.7.7b For prefix paleo‑ (or palaeo‑) /paliou‑/ the Greek source was prefix palaio‑ ‘old’. Like proto (in protozoa, protoderm), paleo features in many

6.8.1 arch -   161 scientific terms constructed on the Greek model, such as paleocene and paleozoic. Productivity is almost exclusively in scientific terms. The meaning is, basically, ‘an early stage of’—a paleo-pole is where the Earth’s magnetic pole was situated at some time in the past, a paleo-botanist studies extinct and fossil plants, and so on. 6.7.7c Alongside pale(a)o‑ is prefix neo‑ /ni:əu‑/, from Greek neo‑ ‘new’. The two forms pattern together in scientific terms such as paleolithic and neolithic, referring to the earlier and later stages, respectively, of the Stone Age. Neo‑ is used as a productive prefix in such derivations as neo-Devonian, for the latter part of the Devonian geological era (similarly, paleo-Devonian is used for the earlier part). Since the middle of the nineteenth century, the use of neo‑ has far outstripped that of its antonym paleo‑. It may describe the revival of some practice—neo-classical architecture is a modern style employing Greek and Roman motifs. Or it can indicate a modernized version of a doctrine, as in neo-Darwinian. Most insidiously, there is the coinage neo-Nazi for a revival of Hitlerian dogma. 6.8 Roles We now consider eleven prefixes, of varied origins, which do not have locational or temporal reference but instead relate to roles, in a wide sense of this term. 6.8.1 archThe prefix arch‑, which is added to nouns in English, emanates from Greek noun arkhos ‘chief’. It came into English mainly through the prefix arche‑ in Old French, hence its pronunciation /a:tʃ‑/. However, we do get /a:k‑/, reflecting the original Greek, in arch-angel /ˌa:'keindʒəl/. Ecclesiastical derivations were already present in Old English, the prefix then signalling ‘principal, superior in hierarchy’, as in arch-bishop and archdeacon. It was then naturally extended, in ME and more modern times, to arch-priest, arch-duke, and the like. A further sense evolved from the sixteenth century and is today highly productive. Added to a noun referring to a person with a negative q ­ uality (either in general, or relative to the speaker) it indicates a magnification of

162    6 keeping the same word class such ­quality—arch-traitor, arch-enemy, arch-foe, arch-villain, arch-rogue, arch-seducer. In essence, arch‑ means ‘to a high degree’, and is specialized either to describe an upper rank within a clerical or aristocratic hierarchy, or the amplification of some undesirable characteristic. There was, in earlier times, a further sense ‘original’. Around 1600, English adopted archetype /ˌa:kitaip/ (the spelling and pronunciation reflecting Latin archetypus ‘original’), at first meaning ‘original variety’. This has shifted to the prevalent, presentday meaning ‘ideal or characteristic example of’.

6.8.2 proPrefix pro‑ /prou‑/ reflects Latin preposition and prefix pro¯ (‑) ‘in place of, on behalf of’. Latin derivations included pro¯ -consul ‘someone who acts in place of a consul (for example, as commander of an army, or governor of a province)’. Coinages on the Latin pattern, with pro‑ added to a title and meaning ‘in place of, as deputy or substitute for’, entered English from the late sixteenth century—pro-rector, pro-regent, and the like. Lately, pro‑ has come to indicate simply a rung on a hierarchy. One facet of the global inflation of titles has led, in some universities, to the dean of a faculty being re-named ProVice-Chancellor. This is an office below that of Deputy Vice-Chancellor, which is directly below Vice-Chancellor. It is the DVC who would substitute in the VC’s absence, not a PVC. A rather different use of pro‑ dates from the early nineteenth century. Added to a noun, it forms an adjective meaning ‘in favour of’. Beginning with derivations such as pro-popery and pro-slavery, the prefix is fully productive in this sense. The 1987 Random House Dictionary lists more than 430 instances, from pro-abolition, through pro-democracy and pro-farmer, to pro-Zionist (and every corpus search throws up more). This sense of pro‑ is the complement of sense (iii) of anti‑ ‘opposed to, opponent of’, in 5.24b. At the beginning of 6.7 there was mention of English forms (such as prophet) commencing with a different pro, this reflecting Greek prefix pro‑ ‘before, ahead, in front of’.

6.8.3 vicePrefix vice‑ /vais‑/ goes back to Latin preposition vice ‘instead of’, which was in turn based on the noun vicis ‘change, alternation’. Loans came from

6.8.4 ways of describing a deputy   163 French—vice-chancellor in the fifteenth century, vice-admiral and vice-president in the sixteenth, and so on. These were at first pronounced with vis‑, as in the French of that time, later replaced by vice to model Latin (save for viscount which remained). It is today mildly productive with titles of office, meaning ‘deputy for, assistant to’—vice-chairman, vice-captain, vice-principal. 6.8.4 Ways of describing a deputy We saw, in 6.4.1c, that both sub‑ and under‑ can be used to describe a person or thing which is of subsidiary importance, mentioning the frequently encountered derivations sub-librarian and under-secretary. A similar sense attaches to both pro‑ and vice‑. It is natural to enquire how these alternatives are deployed. Table 6.10 summarizes the possibilities for the four prefixes with twelve titles. Information in this table has been obtained from a number of dictionaries and corpus collections. Not all forms are equally common—as indicated in 6.4.1c, sub-librarian and under-secretary are very common, while under-librarian and sub-secretary are rather rare. There may well be further possibilities. Nevertheless this table presents a useful overview. It is interesting that no title scores in all four columns. Most have three or two yes’s, but with bishop we only get under‑ and with president only vice‑. As mentioned in 6.8.2, modern universities require labels for the burgeoning levels of administrators (hired instead of teaching and research faculty) Table 6.10  Ways of describing a deputy sub-

under-

pro-

vice-

captain

yes

yes

—

yes

lieutenant

yes

yes

—

yes

principal

yes

yes

—

yes

consul

yes

—

yes

yes

rector

yes

—

yes

yes

chancellor

—

yes

yes

yes

librarian

yes

yes

—

—

secretary

yes

yes

—

—

chairman

yes

—

—

yes

director

yes

—

—

yes

bishop

—

yes

—

—

president

—

—

—

yes

164    6 keeping the same word class and to fill this need we find the double-prefixed Pro-Vice-Chancellor (with pro‑ accorded a new sense). There are additional terms for a deputy. Sous-chef, sous-lieutenant, and a few more are complete loans from French, but they are felt to be foreign; sous‑ has not evolved to be a prefix which is added to new titles within English. Instead of a prefix one can, of course, employ an adjective. All twelve nouns in Table 6.10 may be modified by deputy or associate or assistant. Note that an adjective can be placed before a prefix but not vice versa. One could say Deputy Under-secretary, and Deputy Pro-Vice-Chancellor but— despite the fact that Deputy Secretary and Deputy Vice-Chancellor are titles—not *Under-deputy secretary or *Pro-deputy Vice-Chancellor.

6.8.5 stepGermanic prefix step‑ /step‑/ refers to a relation through re-marriage—your step-mother is your father’s new wife, and you are her step-child. Derivations step-father, step-mother, step-child, step-son, and step-daughter all came through from OE, while step-brother and step-sister were added in ME times. Step‑ can be added to any term of relationship—step-grandchild, stepauntie, step-family, and so on. In OE, ste¯op-cild meant ‘orphan’, and the prefix is related to OE verb a¯ stı¯epan ‘bereave’.

6.8.6 byPrefix by‑ (or bye‑) is a grammaticalization of English preposition by (as in He lives by the lake). It was already a prefix in OE but our modern forms go back no further than ME. By‑ can mean ‘subsidiary’. A by-election, for one or just a few seats, is held between general elections for all seats. A by-product is something which is produced—or happens—secondarily to something else (Rum is a by-product in sugar manufacture). By-play, on the stage, is some talking or action on the side, secondary to the main theme. Jespersen (1942: 532) describes prefix archi‑ (as in archi-phoneme) as ‘a byform of arch‑’ (see 6.8.1). Another sense of by‑ is ‘out of the way’. In his 1755 dictionary, Samuel Johnson defines by-coffeehouse as ‘a coffeehouse in an obscure place’,

6.8.8 meta -   165 quoting from Addison in The Spectator number 403 (12 June 1712): I afterwards entered a by-coffeehouse, that stood at the upper end of a narrow lane, where I met with a conjuror. A more modern coinage is by-pass for a road that enables traffic to pass by (and not drive through) a town. The prefix is used with nouns, and appears still to be productive. 6.8.7 coLatin preposition com ‘with’ developed into a prefix, with forms co‑ before a vowel or h or gn, com‑ before b, p, or m, and con‑ elsewhere. Many Latin forms made their way into English—convene, conspire, combine, compare, coagulate in ME times, coincide, cohere, and condone in early Modern English. Executor and co-executor were Latin loans into ME, and they established co‑ /kou‑/ as a prefix. From about 1600 it has been used productively with all types of word. On verbs and adjectives the meaning is ‘jointly, together’, as in co-exist, co-organize, co-existent, co-equal. From 1643 we have co-worker with a narrow meaning ‘working in collaboration’, or a wider one ‘working in the same industry’ (or an even wider one, as in Co-workers of the world unite!). Conspire is from Latin conspı¯ra¯re ‘agree, plot’, etymologically ‘with-breathe (breathe together)’. And now we append the English prefix—co-conspirators are people who jointly plan some nefarious activity. Co‑ can be added to many of the titles in Table 6.10, the meaning being ‘share responsibility’—bishop, chairman, director, librarian, principal, rector, secretary, and also pilot, editor. A sporting team may have co-captains, who provide leadership in alternate matches. From the 1920s there have been co-stars in moving pictures. And when a school ceased to be uni-sexual it embraced co-education, becoming co-educational.

6.8.8 metaThe Greek preposition meta ‘with, after’ is reflected in metaphor, metabolic, metathesis and a number of other forms, not analysable within English. In the nineteenth century, meta‑ /metə‑/ came to be used as a prefix in English (applied to nouns and adjectives) through deliberate coinages, predominantly scientific. It bears a range of meanings, including ‘at the back of, beyond’, as in meta-thorax ‘the rear part of the thorax (in an insect)’ and meta-carpus ‘bones between the wrist (carpus) and the fingers’.

166    6 keeping the same word class One recurrent modern use is that a meta-X is an ‘X describing an X’. A meta-language is the medium through which linguists describe the structure of a human language, meta-philosophy is a philosophical discussion of the nature of philosophical tenets, meta-pragmatics is the awareness by a speaker of the pragmatic functions of a speech-act. The OED illustrates meta-theory with a 1929 quotation from a mathematician: With each theory here is associated a meta-theory in which we reason intuitively about the theory. In similar vein, astronomers have described the complete system of galaxies as the meta-galaxy (literally, a galaxy of galaxies). 6.8.9 auto- and selfPrefix auto‑, of Greek origin, and Germanic self‑ have similar meanings. 6.8.9a In origin and development, auto‑ is rather similar to meta‑. It relates to Greek adjective auto ‘self, on one’s own’, which is reflected in unanalysable loans such as autopsy, autochthon, and autonomy. The English prefix auto‑ /ɔ:tou‑/ was introduced through coinages from about 1800. Typical of Greek prefixes, it finds wide use in scientific terminology meaning ‘selfacting, self-propelling’. Derivations in more general use include auto-ignition, referring to fuel which spontaneously ignites, under certain conditions (plus verb auto-ignite and adjective auto-igniting), a camera with autofocus, and a refrigerator with auto-defrost. When it relates to humans (and animals), auto‑ means ‘do it about oneself’, as in auto-biography, or ‘do it to oneself ’, as in auto-hypnosis ‘self-induced hypnosis’ and auto-cannibalism ‘consumption of part of an organism or cell by itself ’. In the 1870s, automobile was borrowed as a whole from French (rather than being created within English), and is sometimes reduced to just auto. Automobile has spawned a number of adaptations—autobus, autocab, autotruck, and even autocade (see 4.2). Automaton and automatic come from Greek auto-matos ‘self-moving’. Automatic is sometimes abbreviated to auto‑, as in auto-pilot (automatic pilot), or to a full word auto, as in Turn the dial from manual to auto! 6.8.9b The OE word spelled sylf or self or seolf has evolved into modern-day self, which can be used as adjective or as noun. It features in a number of

6.8.9 auto - and self -   167 compounds, where the meaning of the whole cannot be inferred from the meaning of the parts; for example, self-made. Since at least Shakespeare’s time, self‑ /self‑/ has functioned as a prefix. The 1987 Unabridged Random House Dictionary lists almost 500 derivations, in which the meaning of self‑ is fully predictable. Most are adjectives, but there are also a fair number of nouns and some verbs. We can usefully recognize six senses. (1) A person is perceived by others as having a certain quality with respect to themself. This can be an inherent property, as when we say that someone is self-assured or that they have (or lack) self-assurance. Other a­ djective/ noun pairs involve confident| confidence, aware| awareness, motivated| motivation, and important| importance. We can add adjectives such as ­absorbed and sufficient, and nouns including belief, esteem, and respect. Or the quality can relate to the person’s attitude towards themself. We can again have both adjective and noun—He is self-interested and He is motivated by self-interest both indicate that his main interest is in how something could be to his advantage. Others involve righteous| righteousness, and adjectives conscious and centred. A further sub-type of sense (1) concerns how a person treats themself. They can be said to be self-controlled or to exercise self-control. Further adjective/noun pairs involve disciplined| discipline, critical| criticism, doubting| doubt, pitying| pity, denying| denial, and indulgent| indulgence. (2) A person does something to themself. They may be self-employed, or self-proclaimed as an emperor, or an injury may be self-inflicted or an envelope self-addressed. These are just adjectives. We also get verb/adjective/ noun triplets to which self‑ may be added: educate| educated| education, teach| taught| tuition, appoint| appointed| appointment, examine| examined/ examination. Verb/noun pairs include analyse| analysis and hypnotize| hypnosis. (3) A person has to do something for themself which they might expect someone else to do for them. In a self-service place one has to serve oneself; in a self-catering apartment there is no catering provided. Alongside these we have a further adjective, self-funding, and the noun self-defence. (4) Something is made by a person about themself. This is illustrated by nouns self-portrait, self-biography, and self-description. (5) A thing does something by itself without outside intervention. A camera may be self-focusing, or self-loading or self-winding (or we can use

168    6 keeping the same word class the corresponding verbs). Similarly with self-adhesive, self-tuning, and self-defrosting. Within the same sense there is self-destruct| destructive| destruction. (6) A quality is entirely included within something, without outside intervention. A name may be self-explanatory, a prophecy self-fulfilling, a statement self-evident or self-contradictory, a plan self-defeating, a colony self-governing, and an apartment self-contained.

6.8.9c For a limited number of self‑ derivations, there is a synonym using prefix auto‑ (not the auto which is an abbreviation of automobile and automatic). This applies for all those listed under sense (5), a thing which does something by itself without outside intervention. Both self-defrost and autodefrost are in general circulation. Similarly with focusing, loading, winding, tuning, ignite| igniting| ignition, and destruct| destructive| destruction. From sense (3) we get both self-hypnosis and (perhaps because it has a Greek base) auto-hypnosis. Both self‑ and auto‑ are also used in sense (4), something made by a person about themself. Auto-portrait and self-portrait is each well-attested. Auto-biography and self-biography were introduced more-or-less simultaneously, at the very end of the eighteenth century. However, self-biography lasted for only a short while, being usurped by auto-biography (perhaps in part because this also has a Greek base).

6.8.10 hetero- and homoAlthough we find heterogeneous and homogeneous (not analysable in English) from the early seventeenth century, prefixes hetero‑ /het(ə)rou‑/ ‘different’ and homo‑ /houmou‑/ ‘the same’ were only introduced through scientific coinages during the nineteenth century. They reflect Greek lexemes heteros ‘other, different’ and homos ‘same’ and in English are attached almost exclusively to adjectives and nouns. The two prefixes have complementary meanings, and are typically used with the same stem. For instance homo-cyclic ‘a cyclic compound which only uses carbon atoms’ as opposed to hetero-cyclic ‘a cyclic compound where one of the ring members is not carbon’, and homo-chromatic ‘all of, or r­ elating to, one colour’ versus hetero-chromatic ‘of or relating to several colours’.

6.9 repetition, re -   169 The pair of prefixes came into public purview in the 1890s with the coining of homo-sexual, describing attraction to people of one’s own sex, and its complement hetero-sexual. They can be abbreviated to just homo and hetero. The opprobrium sometimes attached to homosexuality has let to derivations such as homo-phobia, homo-phobic, homo-phobe, and also homo-phile ‘someone who supports the rights of homosexuals’. 6.9 Repetition, reLatin prefix re‑ ‘back, again’ is a part of many loans into ME and later stages of the language, which are unanalysable in English; among them are record, repent, and reward. There were also doublets such as assemble and re-assemble, enter and re-enter, enabling the recognition of prefix re‑ /ri:‑/ by late ME times. (See also 2.4d.) New derivations in English involved many Romance stems (such as re-consider) and an increasing number of Germanic ones—re-write, re-build, re-fill, re-think, and so on. Prefix re‑ applies almost exclusively to verbs and nominalizations based on them—re-arrange and re-arrangement, re-activate and re-activation. It has one major meaning and a number of subsidiary ones. (1) Used with a transitive verb, re‑ ‘again’ indicates that the action referred to by the verb applies a second (or further) time to the same ‘object’. There are a very large number of examples of this sense. One can re-elect a president, re-stock a larder, re-charge a battery, re-publish an out-of-print book, re-open a museum, re-affirm a belief, re-read a favourite book, refinance a failing company, re-enact an ancient ceremony, re-commence an interrupted project, and so on and so forth. Sometimes the use of re‑ implies the instigation of change, as when someone re-thinks a policy, or re-educates the populace (to try and get them to abandon their traditional ideas in favour of some new philosophy of the rulers). (2) The prefix can refer to a return to some previous state. One may live in a place all one’s life, move away, and then re-enter it. Something (or someone) could always have been visible, then disappear, and then re-appear (note that this is an intransitive verb). Money which has been lent can be re-paid. (3) It can describe a change in location or time. Facilities may be relocated, people re-settled, resources re-distributed, a meeting re-scheduled.

170    6 keeping the same word class (4) It is added to some S = O ambitransitive verbs used intransitively, such as re-train and re-locate. And there is also re-marry. We can say just John re-married or, if he had the same bride as previously, Rev. Smith re-married John and Mary or John and Mary got re-married (but one would prefer John married Mary again, rather than *John remarried Mary). As mentioned in the discussion at 2.5d, there are minimal pairs involving an unanalysable loan, in which re is /ri/, and a prefixed derivation, in which re‑ is /ri:‑/. For example recount /'rikaunt/ ‘tell, narrate’, and re-count /ˌri: -'kaunt/. ‘count again’. Also reform| re-form, recover| re-cover, release| re-lease, restore| re-store (and see Marchand 1969: 191).

6.10 Not quite right, pseudo- and quasiThe final pair of prefixes consists of Greek pseudo‑ and Latin quasi‑. They are superficially similar but in fact exhibit a significant meaning difference. 6.10a Pseudo‑ /psu:dou‑/ goes back to Greek pseudos ‘falsehood, lie’. It became established as a prefix in English from about 1600 with coinages such as pseudo-Catholic ‘falsely pretending to be a Catholic’, and is now rather common with both nouns and adjectives. Like other Greek prefixes, it is widely used in science, typically indicating a close or deceptive resemblance. For instance, pseudo-paralysis for an inability to move some part of the body which might be thought to be due to paralysis but in fact has another cause. And pseudo-pregnancy describes a physical condition which looks like pregnancy but is not. In everyday usage the most common meaning is ‘pretending to be’ or, as Samuel Johnson (1755) put it: ‘false or counterfeit’. A pseudo-intellectual acts like one but is in fact devoid of scholarship. Pseudo-science covers such pursuits as astrology, which have no scientific basis as the term ‘science’ is generally understood. A totalitarian country where voting takes place but there is only one party on the ballot paper is a pseudo-democracy. 6.10b Latin quasi ‘as if, a sort of’ is the source for English prefix quasi‑, /kwa:zi‑/ or /kweizai‑/, ‘a bit like, but not fully’. There were occasional uses from the seventeenth century, and these increased dramatically from around 1800. Today, the prefix is highly productive with nouns, adjectives, and adverbs.

6.11 diminutive and endearing suffixes   171 Quasi-art has some artistic qualities, but these are not developed enough for it to be called true art. A quasi-military uniform looks at first like the real thing, but closer inspection reveals that it relates to no known army. The OED quotes from The Church Times of 2003: He prefers to describe humanism, Marxism, and nationalism as quasi-religious (perhaps because of the religious-like fervour they arouse among their adherents). The conventions of a tribal society are sometimes described as quasi-legal institutions since they have something of the function of legal entities in major nations without partaking of the formality. Another use of quasilegal was to describe the status of cyder during Prohibition in the USA; being alcoholic it was officially under a ban, but this was not really enforced. 6.10c Pseudo-X describes pretending to be like X, or being similar to X, but in neither case being X. Quasi-X is used for having some characteristics of X but not being a full X. This can be illustrated with contrasting pairs:

•

quasi-cripple, someone who has some small thing wrong with them (say, missing two fingers from one hand) but not really so serious as to justify the label ‘cripple’ pseudo-cripple, someone who has nothing at all wrong with them but pretends to be a cripple (perhaps, so that they can take part in the para-Olympics)

• • •

quasi-christian, someone who subscribes to some of the tenets of Christianity, but not the most significant ones pseudo-christian, an atheist who puts up the front of being a christian (maybe, to gain access to a community in which non-christians are regarded as devils)

6.11 Diminutive and endearing suffixes We round out this chapter with discussion of a handful of suffixes which may maintain the same word class. By comparison with other languages, English has rather limited morphological means for creating diminutives (and none at all for augmentatives). As is often the case, a diminutive meaning may shade into a simply affectionate one. There are a few endings with diminutive meaning which are not productive. From French we had (e)rel, as in cockerel ‘young cock’. It has a disparaging

172    6 keeping the same word class sense in such forms as wastrel and scoundrel. The Germanic diminutive en is found in chicken, kitten, and maiden. (Hillock was mentioned in 4.1c.)

6.11.1 -let Suffix ‑let /‑lit/ comes from French and is added to nouns. One, rather minor, meaning describes an ornament associated with some part of the body, such as arm-let, wrist-let, and ankl(e-l)et. The major sense is diminutive: tart was borrowed by 1400, followed a few years later by tart-let ‘a little tart’. By the sixteenth century, ‑let had become a productive suffix in English. It can refer to a young (and, therefore, small) animal. Eagl(e-l)et was borrowed from French in the early sixteenth century and soon after we find owl-(l)et, and (in 1880) pig-let, both based on Germanic roots. The suffix is fully productive (especially in American English) and is mostly applied to inanimates, including book-let, note-let, drop-let, stream-let, cloud-let, flatlet, and ring-let (of hair). Attached to titles, ‑let can indicate disdain—king-let for a minor or petty king (held somewhat in contempt). A derivation with an entirely positive overtone is star-let, for a young actor considered likely to develop into a star. (It says something about prevailing attitudes that this term is applied almost exclusively to young female—not male—actors.)

6.11.2 -ling Suffix ‑ling /‑liŋ/ was common in OE, its meaning then being ‘associated with’. For example rǽp-ling ‘prisoner’ (from ráp ‘rope’), together with our modern hire-ling and under-ling. The term earth-ling—a staple of presentday science-fiction—goes back to 1593, referring then to the inhabitants of earth as contrasted with denizens of heaven. The diminutive sense developed in ME, with names of animals: cod-ling, duck-ling, go(o)s(e)-ling. The ‘associated with’ and diminutive senses combine in nest-ling ‘a bird too young to leave its nest’ and seed-ling, ‘a young plant raised from seed’. When used with titles, ‑ling implies ‘a minor and poor example of’. The OED defines king-ling as ‘a little or petty king (less contemptuous than kinglet)’. Prince-ling may describe someone who rules over a tiny domain which is far from rich (alternatively, it can be used of a young prince).

6.11.3 - y ~ -ey ~ -ie ~ -ee, -sy , and the echo technique   173 Most instances of ‑ling are attached to nouns. In days gone by, it also derived a few nouns from adjectives or verbs, as with weak-ling, for someone physically or morally weak, starve-ling, for an under-nourished person or animal, and change-ling, for a child secretly substituted for another (all three are from the sixteenth century). 6.11.3 -y ~ -ey ~ -ie ~ -ee, -sy, and the echo technique 6.11.3a Suffix ‑y ~ ‑ey ~ ‑ie ~ ‑ee has variant spellings—one sees both Charl-ie and Charl-ey, both Bobb-y and Bobb-ie, both bootie and bootee (for a small soft shoe worn by a baby) and so on (we will just call it ‑y). It always has the form /‑i:/ and can only be added to a monosyllable. Note that there are four other suffixes ‑y /‑:i/—adjectivalizer ‑y (see 8.2.8, 8.4.13) and three nominalizers ‑y (see 9.2.2, 9.4.11, and 9.3.13a).

X-y may indicate a small token of X, or a loving attitude towards X, or both. A small dog may be referred to as dogg‑y if you are fond of it and it is relatively small. Dogg‑y is much more likely to be used of a pet corgi or pekinese, less likely (because of its size and character) of a pet Alsatian. Similarly with puss‑y, pigg‑y, mous‑ie, frogg‑y, bird‑ie. One of the earliest derivations (in the fifteenth century) was bab‑y from babe. This was soon followed by kin terms. Suffix ‑y may be added to mam or mum and to dad (not to mother or father), to gran (not to grandmother) and to aunt. Recently, ‑y has been used in shortenings such as book-ie for bookmaker, post-ie for postman, mov-ies for moving pictures, comm-ie for communist, bolsh-ie for bolshevik. This is particularly common in Australia—one may give a Chriss-ie (Christmas) press-ie (present) to a truck-ie (truck-driver) from Tass-ie (Tasmania). From the eighteenth century, this suffix was extended to use with monosyllabic abbreviations of proper names. The entry for Thursday, 14 October 1773 in The journal of a tour to the Hebrides (1775) by James Boswell begins: When Dr Johnson awakened this morning, he called, ‘Lanky!’ having, I suppose, been thinking of Langton, but corrected himself instantly, and cried, ‘Bozzy!’ He has a way of contracting the names of his friends. Goldsmith feels himself an important man, as to be displeased at it. I remember one day, when Tom Davies was telling that Dr Johnson said, ‘We are all in labour for a name of Goldy’s play,’ Goldsmith cried, ‘I have often desired him not to call me Goldy.’

174    6 keeping the same word class Over the last couple of centuries, ‑y has been added to short monosyllabic forms of many forenames. For example: long form

monosyllabic short form(s)

adding -y

Richard

Dick Rick Rob Bob Bess Mike Kate

Dick-y Rick-y Robb-y Bobb-y Bess-ie Mik-ey Kat-ey

Robert Elizabeth Michael Katherine

There is a phonological restriction—suffix ‑y is scarcely ever added to a form ending in a vowel, such as May /mei/, Di /dai/, Lea /li:/. However, it is acceptable onto a monosyllabic ending in /u/ (which here has the function of a semi-vowel /w/). That is, we can say Joe-y /'dʒou-i:/ (or /'dʒow-i:) and Sue-y /'su:‑i:/ (or /'suw‑i:/). The suffix is not very readily attached to forms ending in ‑l; for example, Val and Paul. Although aunt-ie is very common, *uncl-ie is impossible (one would have to miss off the final l and say unc-ie). Note that ‑y can follow a consonant cluster, including those that contain an l, so long as it is not the final element, for example Alf-y. (Final ‑r also disallows ‑y; from Patricia we get Patt-y but not *Patr-y.) Sometimes, there is no stand-alone short form, with ‑y just being added to the first (or sometimes the second) syllable of the full name. For example, And-y from Andrew, Lind-y from Linda, Mont-y from Montgomery, Mandy from Amanda. In Benj-ie, from Benjamin, the j from the beginning of the second syllable of the full name is retained. 6.11.3b A by-form of ‑y is ‑sy /‑si:/. It is likely to indicate a greater degree of affection than plain ‑y: a dog-sy may be more doted over than a dogg-y. The -sy form may only be added to a monosyllable which ends in a single ­consonant—one can say mum-sy but not *aunt-sy. And it appears to be mostly (although not exclusively) used on nouns with female reference— there is Bet-sy alongside Bett-y, Pat-sy as an alternative to Patt-y. (One would not really say *dad-sy, on the pattern of mum-sy). When ‑sy is attached to the name of a person who the speaker clearly does not adore, it can indicate ironic disrespect. Someone who is not an admirer

6.11.3 - y ~ -ey ~ -ie ~ -ee, -sy , and the echo technique   175 of the ideas of Sigmund Freud could say, to a Freudophile friend: And what would your Freud-sy say about that? We also find ‑sy as a by-form of adjectivalizing suffix ‑y; see 8.2.8d. 6.11.3c There is also the informal echo technique for creating endearments. Starting with a monosyllabic form to which ‑y /‑i:/ is added, this is repeated, with the initial consonant(s) being replaced by either p or w. For example: George-y Gwen-y Joan-y Samm-y Mike-y Bobb-y Brian-y Paul-y

/'dʒɔ:dʒi:/ /'gweni;/ /'dʒouni:/ /'sami:/ /'maiki:/ /'bobi:/ /'braiəni:/ /'pɔ:li:/

Georgy-Porgy Gweny-Penny Joany-Poany Sammy-Pammy Mikey-Pikey Bobby-Wobby Briany-Wiany Pauly-Wauly

/'dʒɔ:dʒi:-ˌpɔ:dʒi:/ /'gweni-ˌpeni;/ /'dʒouni:-ˌpouni:/ /'sami:-pami:/ /'maiki:-ˌpaiki:/ /'bobi:-ˌwobi:/ /'braiəni:-ˌwaiəni:/ /'pɔ:li:-ˌwɔ:li:/

The echo component generally begins with p unless the initial consonant(s) are, or include, b or p, in which case the initial consonant is w. That is, if the first part commences with a bilabial stop, the second part cannot also do so. Words with an initial labio-dental fricative, f or v, appear to be an intermediate category, accepting either p or w in the echo repetition. From Fred we can have either Freddy-peddy or Freddy-weddy, and from Val either Vally-pally or Vally-wally. (Sometimes initial w can be used in place of p with names that do not commence with p, b, f, or v—Mikey-wikey instead of Mikey-pikey. But initial p is never used after a name starting with p or b.) It is not the case that these echo forms are based on the ‑y derivations discussed in 6.11.3a. Names which do not yield a ‑y form can be echoed; for example, Paul and Val. In essence, the echo technique can apply to any name. Even those beginning or ending in a vowel—from Alan we get Ally-pally and from May there is May-pay. The echo-formation rule is straightforward. Take the first (C1(C2)) VC3(C4) of a name, add /i:/, then repeat it, dropping the initial C1 or C1C2 and adding initial p or w. If there is no C3(C4), then no /i:/ is added. There is one more thing: an echo may also be formed on an ‑sy derivation, as in Katesy-patesy (or Katesy-watesy) and mumsy-pumsy (or mumsy-wumsy).

176    6 keeping the same word class 6.11.4 -(i)kin(s) Suffix ‑kin(s) made its way into ME from Middle Dutch and is preserved in many family names, including Tom-kin(s), based on Thomas, Haw-kin(s), based on Henry, and Wil-kin(s), based on William. It survives as a derivation to a common noun in lamb-kin, used either of a cute little lamb or a sweet young person. Today, ‑ikins /‑ikins/ carries a sense of endearment, similar to ‑y, ‑sy, and the echo technique. One simply appends ‑ikins to any monosyllabic name: Dick‑ikins, Bob-ikins, Kate-ikins, Paul-ikins, Sue-ikins, May-ikins. Since this derivation applies to forms which do not readily accept -y (such as Paul and May), it should not be considered as the addition of ‑kins after ‑y, but rather a separate suffix ‑ikins. There are thus a number of techniques for showing endearment. From Jane we can get Jane-y, Jane-sy, Janey-painy (or Janey-wainey), Janesypainsy (or Jainsy-wainsy), and Jane-ikins. And perhaps also Janes(y)-ikins. The ‑sy forms do seem to imply a greater degree of affection than the plain ‑y ones. Beyond that, I leave it to others to suggest a hierarchy of fondness. 6.11.5 -er(s) This suffix had an informal origin. It is a mark of in-group bonhomie, endearing in a kind of macho way. In 1875, upper class youths from Rugby school (later moving on to Oxford University) introduced the affectation of adding ‑er(s) /‑ə:(z)/ to the initial (C1(C2))VC3 of any noun. The OED quotes the following from the Daily Chronicle of 25 March 1904 (I have added the original words in brackets): ‘An undergraduate left his bedd-er [bedroom] in the morning to eat his brekk-er [breakfast] in his sitt-er [sitting-room]; later he attended a leck-er [lecture], and in the afternoon he might . . . take some . . . ekk-ers [exercise].’ In late nineteenth-century England, there were two varieties of football. Rugby football (which originated from the school of that name) naturally came to be called Rugg-er. (But only in England; aficionados of Rugby Union in Australia do not use this abbreviation.) Association Football (played with a round ball) also merited a soubriquet. The initial a of Association was omitted and an ‑er derivation based on the new initial syllable. Hence was created Socc‑er, a word which does have wide distribution throughout the English-speaking world.

6.11.5 - er(s)   177 This playful shortening, using ‑er(s), was also applied to proper names, and indeed continues to the present day amongst Englishmen who relate to the upper class or thereabouts. For example, it pervades cricket broadcasting. Commentator Brian Johnston was Jonn-ers and Jonathan Agnew is so well known as Agg-ers that a recent book was titled Aggers’ Ashes: the inside story of England’s 2011 Ashes triumph. Having in the last two chapters examined derivational affixes which, by and large, do not change word class, we can now turn to those that do, starting with affixes which create verbs.

7 Be-feather, smart-en, em-power, beauti-fy, moral-ize: Making new verbs 7.1 Introduction 178 7.2 Double duty and internal variation 180 7.3 The affixes  183 7.4 Phonological conditioning  192

7.5 The semantics of productive verbalization 199 7.6 The ‘how’ and the ‘why’  206 7.7 Deriving verbs from proper names 214

7.1 Introduction 7.1a English has fewer techniques for creating new verbs than it has for deriving adjectives and nouns. Nevertheless, the ways it has for making verbs are fascinating. Why do we say em-bitter, making a verb from adjective bitter by adding prefix em-, but sweet-en, now using suffix -en? Why does one hear motor-ize, making a verb from noun motor by adding -ize, a suffix which goes back to Greek, but electr-ify with suffix -(i)fy, which is of Latin origin? The verb to describe making something German in language or character is German-ize, but with respect to French it is French-ify. Once more, why? Quite a few nouns do double duty as verbs, without the need for any derivational affix. Why do we have to add -(i)fy to noun class, forming verb class-ify whereas list can be used in its bare form for verb function? Similarly, for ‘put the food in a store’ one just says store the food (with store doing double duty as noun and as verb) whereas for ‘put the patient in a hospital’, -ize must be added to hospital, giving hospital-ize the patient.

7.1 introduction   179 This chapter attempts to put forward principled answers to these and similar questions. We will see that the motivating factors are a combination of phonological form, semantics, and genetic origin. 7.1b Many languages have derivational processes for forming a transitive verb stem from an intransitive verb root, and vice versa. English has scarcely any. What it does have is a large number of verbs which can be used either in a transitive clause, with transitive subject (A) and transitive object (O) arguments, or in an intransitive clause, with an intransitive subject (S) argument; see 1.2. Some of these ambitransitive verbs are of type S = A; for example, MaryS is knitting and MaryA is knitting [a scarf   ]O, JohnS ran yesterday and JohnA ran [two kilometres]O yesterday. Others are of type S = O; for example, [The balloon]S burst and [The child ]A burst [the balloon]O. 7.1c There is one derivation in English which applies to transitive verbs and changes the nature of the O argument. This involves the prefixation of out-. Compare (all constituents in parentheses are optional): (1) JohnA bid $100O (for the vase) (at the auction) (2) MaryA bid $110O (for the vase) (at the auction) (3) MaryA out-bid JohnO (by $10) (for the vase) (at the auction) In the plain transitive sentences, (1) and (2), the noun phrase in A function refers to the person who bid, and that in O function to the amount bid. In the out- derivation, (3), the person who made the higher bid is referent of the A noun phrase and the one who made the lower bid is referent of the O noun phrase. As noted in 6.6.5a, there are many other verbs which can take out- in this argument-changing sense. If Mary ran further than John, then she out-ran him. If John drank eight beers and Mary only six then he out-drank her. And also shoot, weigh, bark, number, live, catch, and so on. (The Unabridged Random House Dictionary lists more than 300 such verbs—Flexner 1987: 1373−8.) The general verb do may also take prefix out-. If Mary out-did John in school, it means that she was overall more successful than him. (However, in this instance it is not feasible to ‘explain’ the out-derivation in exactly the way that we did for out-bid in (1−3).) 7.1d Ending le occurs on a number of words with a variety of meanings, such that it cannot really be recognized as a suffix. For instance, spark (basically a

180    7 making new verbs noun) and sparkle (basically a verb), spit (basically a verb) and spittle (noun) ‘what one spits’. Within the context of this chapter, suck and suckle are a fascinating (and unique) pair of verbs. Suck is an ambitransitive of type S = A and suckle an ambitransitive of type S = O. Moving from one to the other involves reorientation of argument roles, as seen in: (4) (a) [The baby]A is sucking its [mother’s milk]O  (b)  [The baby]S is sucking (5) (a) [The mother]A is suckling [the baby]O  (b)  [The baby]S is suckling The distribution of arguments can be diagrammed: transitive suck intransitive suck transitive suckle intransitive suckle

baby mother/mother’s milk A O S O A S

7.2 Double duty and internal variation 7.2a Languages with a relatively skimpy morphology tend to have a number of lexemes doing ‘double duty’ in two (or perhaps more) word classes (see 2.8). Old English had a rich morphology and made little use of double duty. Changes in ME included loss of much of the morphological apparatus on nouns and on verbs, and the appearance of a number of lexemes which are primarily nouns or adjectives but have a secondary function as verbs. Later in this chapter (in 7.5.1−2) we survey the range of meanings conveyed by the large number of nouns (and the few adjectives) which also do double duty as verbs. They are of Germanic and Romance origin, in about equal numbers. We can here give just a few examples:

• • • •

OE had noun blóstm and verb blóstmian, both becoming blossom, functioning as noun and as verb. OE had noun brycg and verb brycgian, both becoming bridge, again a word taking on double duty as noun and as verb. ME dance, noun and verb, was borrowed from Old French noun dance (or danse) and verb dancer (or danser). ME blame, noun and verb, was borrowed from Old French noun blâme and verb blâmer.

7.2 double duty and internal variation   181 7.2b There are thus hundreds of lexemes in English which have exactly the same form when used as noun or as verb, or as adjective or verb. But there are some for which a difference has been created. English may be light on morphology, but it does have one device not found in French—contrastive stress placement. The preferred placement of stress in English is on the first syllable of a root, and this is particularly strong in the case of nouns and adjectives. French has invariable stress on the final syllable. Some dozens of disyllabic words were borrowed from French but a stress difference was fashioned—it falls on the second syllable of the verb (as in French), but on the first syllable of the noun or adjective (as is most common in English). For example:

• • • •

noun insult /'insʌlt/ from Old French insult verb insult /in'sʌlt/ from Old French insulter noun decrease /'di:kri:s/ from Old French decreis verb decrease /di'kri:s/ from Old French decreistere

Once this pattern became established, it was naturally extended. Nouns fragment and escort were borrowed (and accorded initial stress) in the 1500s. A century or two later, verbs were created from these nouns simply by shifting stress to the second syllable. Adjective frequent /'fri:kwənt/ and verb frequent /fri'kwent/ were adopted from Latin. Note that sometimes noun and verb are identical save for stress (as shown for insult). Other times there may be reduction of an unstressed vowel (as is typical in English). The initial stressed syllable of verb decrease has a long vowel /i:/; when stress falls on the second syllable, the vowel is shortened to /i/. Noun suspect is /'sʌspeckt/, but in the verb /səs'pekt/ the vowel of the first syllable, now unstressed, is reduced to a schwa /ə/. Also estimate and separate. There are also a few trisyllabic double duty lexemes (all Romance loans) where noun and verb are written in the same way, and both bear initial stress but the noun has a schwa [ə] in the final syllable, corresponding to a noncentral vowel in the verb. For example, verb document /'dɔkyument/, noun document /'dɔkyumənt/, and also compliment, complement, implement, estimate, graduate, isolate. 7.2c The various components of a language system are continually interacting. A change in phonology, for instance, is likely to engender some shift in the grammar, or in the forms of lexemes.

182    7 making new verbs In OE, [f ] and [v] were alternative pronunciations of a single phoneme. Voiced [v] occurred between voiced sounds, and voiceless [f ] elsewhere. During the development of ME, /f/ and /v/ emerged as contrasting phonemes. In verb believe, the v|  f phoneme came between two vowels (voiced sounds) and was pronounced as [v]. In noun belief, the same v|  f phoneme was at the end of the word and was accorded voiceless pronunciation /f/. Then the final e was dropped from believe (it is still written, but no longer pronounced—a ‘silent e’), and the voiced sound [v] now came at the end of a word, contrasting with [f ]—believe /bə'li:v/, and belief /bə'li:f/. Ergo, we have separate phonemes /v/ and /f/, which serve to establish a verb/noun pair with internal variation, believe|  belief. This process established similar verb/noun pairs based on Romance loans into early ME, including prove| proof, relieve| relief, grieve| grief (and the verb/adjective pair save| safe). The same sort of change applied for s and z, which were also linked as a single phoneme in OE and developed into separate phonemes within ME. The noun written as hus in OE and ME became present-day house /haus/. The related verb was written as husian in OE and huse in ME. Since the s came between vowels it was voiced, and stayed voiced when the final e ceased to be pronounced, giving modern verb house /hauz/. There are a number of verb/noun pairs distinguished by final /z/ versus /s/, most being due to Romance loans into ME. Sometimes verb and noun are spelled the same, as with house and excuse. For others, c is written for the /s/ in the noun and s for the /z/ in the verb. Thus, noun advice /əd'vais/ and verb advise /əd'vaiz/; also device| devise. (Interestingly, noun practice and verb practise are written differently but pronounced the same, /'praktis/.) There was also a single lamino-dental fricative phoneme in OE, pronounced as voiced [ð] between voiced sounds and as voiceless [θ] elsewhere. These also developed into separate phonemes, /ð/ and /θ/, although both are written as th. For example, OE noun cláð became ME and modern cloth /klɔ:θ/, while OE verb cláðian became ME cloathe and then modern verb clothe /klouð/. Noun and verb are distinguished by final /θ/ versus /ð/, and also a vowel difference. The same applies for bath|    bathe, breath|    breathe, wreath| wreathe, and sheath| sheathe. And also for mouth and teeth, although here noun and verb are written the same (with no ‘silent e’ on the verb). Note that all final /θ/ versus /ð/ pairs are Germanic, since these sounds are not found in Latin or French.

7.3 the affixes   183 There are also a few noun/verb pairs distinguished by final t| d, the contrast being taken over from Romance. For instance noun descent /di'sent/ and verb descend /di'send/; also ascent| ascend and extent| extend. 7.2d And we find a number of Germanic noun/verb pairs distinguished by vowel quality; they reflect alternations at various stages in the history of the language. There is /i/ in the noun and /ai/ in the verb for bit /bit/, bite /bait/ and the reverse for life /laif/, live /liv/. (Note that life|  live also shows the f| v difference described just above.) Other pairs include seat| sit, shot| shoot, song| sing, and abode| abide. There are detailed and insightful discussions of double duty—variously called ‘zero derivation’ and ‘conversion’—and of internal variation—sometimes called ‘phonological modification’—in Jespersen (1942: 83–134, 186–208), Marchand (1969: 359−79). Clark and Clark (1979), and Bauer and Huddleston (2002: 1638−44).

7.3 The affixes Three of the negative prefixes described in Chapter 5 can derive verbs from nouns and adjectives (each of them may also preserve word class):

• •

Romance prefix de-, in its sense (i) ‘get off a vehicle’, derives intransitive verbs from nouns, e.g. de-plane. See 5.10. Sense (ii) of de-, sense (i) of another Romance prefix dis-, and sense (i) of Germanic prefix un-2 all mean ‘deprive of’ and derive transitive verbs from nouns (and a few adjectives). For example, de-mast, dismast, un-mast. See 5.10−16.

Then there are six affixes whose sole function is to derive verbs from nouns and adjectives. There are two prefixes—Germanic be- and Romance en-/ em-. And four suffixes—Germanic -en, Romance -ate, and -(i)fy, plus -ize which is ultimately of Greek origin. Of these, only -(i)fy and -ize are fully productive. Most of this chapter (7.4–7) deals with -(i)fy and -ize, plus double duty lexemes. We look at the phonological forms of bases to which -(i)fy and -ize may be added, and the semantics involved, plus some explanation for why things are the way they are. The final section (7.7) examines how, using these two derivational suffixes, verbs are created from proper names. We explain why

184    7 making new verbs people say de-nazi-ify, with deverbal suffix -(i)fy, and de-stalin-ize, with -ize (rather than *de-nazi-ize and *de-stalin-ify). Before getting into that, we provide accounts of verbalizers be-, -en, en-| em- and -ate (which have low productivity and do not take stress or affect stress placement in the base) plus an introduction to the two highly productive suffixes -(i)fy and -ize. 7.3.1 Prefix beGermanic prefix be- /bi-/ has the same ultimate origin as modern-day prefix by- (discussed in 6.8.6)—OE preposition be or bı¯  ‘by, along, about, concerning’. It was already a verbalizing prefix in OE. Over the centuries, be- has had many senses. There are today just odd relics of some, and only one still has any measure of productivity. Typically, verbs derived by be- are used a good deal in past participle form, with -ed, functioning as adjectives; for example, the be-feather-ed savage. (See 8.2.12d.) There are a number of examples of be- deriving a transitive from an intransitive verb, as in be-wail ‘wail about something’ and be-moan ‘moan over something’. The few adjectives taking be- have negative overtones—if a sailing boat is be-calm-ed, it lacks wind that would enable it to move; if a person is be-little-ed they are humiliated and made to appear little. One sense of the OE prefix be- was ‘remove’, as in be-niman ‘deprive of’ from niman ‘take’. In the sixteenth century, a bishop could be be-sleeve-ed. The only residue of this sense we have today is be-head meaning ‘remove a person’s head, or a vegetable’s top, etc.’. In some instances, be- was attached to a transitive verb and appeared just to add a measure of intensity. For instance be-friend had more-or-less the same meaning as the old verb friend, ‘make friends with’. Similarly for bewitch and verb witch, ‘apply witchcraft to’, be-devil and verb devil ‘treat as a devil does’, be-grudge and verb grudge. The plain verbs friend, witch, and devil then dropped out of use, leaving be-friend, be-witch, and be-devil as the apparent verbalizations of nouns friend, witch, and devil. Verbs grudge and be-grudge both continue, with very similar meanings. The one sense of be- which still has a limited degree of productivity appears only to have developed in the sixteenth century. It means ‘cover with, adorn with’. Someone may comment that the duchess is be-jewell-ed, that her husband is be-whisker-ed, and that the judge is be-wigged. The OED

7.3.2 suffix- en   185 quotes the Manchester Guardian of 1958 describing A sulky adolescent and his be-jean-ed girl-friend. This was at the period when jeans were changing from being just rough-wearing workman’s apparel to becoming an article of fashion, something to be noted. There is today a habit among some youths of sporting metal studs, not only in ear lobes but also in nose, lip, even tongue. One might say, of someone thus accoutred, that they are mightily be-studd-ed.

7.3.2 Suffix-en 7.3.2a The other verb-deriving process which is of Germanic origin involves the suffix which is written -en. This is usually pronounced as a vocalic nasal /- n̩/, as in broad-en /'brɔ:d- n̩/, but is /-ən/ after /k/ or /p/, as in weak-en /'wi:k-ən/. The modern suffix can in part be traced back to an OE process for forming verbs from nouns and adjectives by adding suffix -an or -ian or -nian. For example (see also verbs blóstm-ian ‘blossom’ and brycg-ian ‘bridge’, mentioned in 7.2a, plus hus-ian ‘house’ and cláð-ian ‘clothe’ from 7.2c): noun adjective

slite ‘tear’ lufu ‘love’ scearp ‘sharp’ faest ‘fast, fixed’

verb

slīt-an ‘tear’ luf-ian ‘love’ scyrp-en ‘sharpen’ faest-ian ‘fasten’

7.3.2b During later ME and early Modern English times, -en developed into a recurrent means for deriving ambitransitive verbs from adjectives. Used intransitively they mean ‘become’, as in The river deep-en-s just below the village; used transitively they have a causative meaning ‘make’, as in He deepen-ed the grave. A fascinating feature of verbalizing suffix -en is that its use is determined by a number of interconnecting factors: (1) It applies only to monosyllables. (2) It is used almost exclusively with Germanic roots. There are just a few examples of -en with roots of Romance origin—moist-en and quiet-en. The first two factors are closely linked. A high proportion of Germanic roots are monosyllabic, and vice versa.

186    7 making new verbs (3) It is used almost exclusively with adjectives. (4) It is used with many adjectives from the semantic types dimension (wid-en, short-en, thick-en, fatt-en, etc.) and physical property (hard-en, rough-en, sweet-en, etc.). It is used with one speed adjective (quick-en), with the three main colour terms (black-en, whit-en, redd-en) and, marginally, with colour adjective pink, giving pink-en. The suffix is scarcely attested with adjectives from the human propensity type. It is not used, for example with proud or rude or kind. We do find madden, sadd-en, and gladd-en, but these are almost exclusively used in transitive clauses. Note that many human propensity adjectives are disyllabic (for example, afraid and stupid ). We also find -en with two irregular comparative forms: wors-en and less-en (less is the original comparative of little). Better, the antonym of worse, does not take -en since it has two syllables (and ends in a vowel). (5) The suffix is only added to a root ending in a voiceless stop (/p/, /t/, /k/), or /d/, or a voiceless fricative (/f/, /s/, /ʃ/, /θ/). Some examples are:

• • • • • • • •

final /p/: cheap-en, damp-en, deep-en, rip-en, sharp-en, steep-en final /t/: bright-en, fatt-en, flatt-en, light-en, quiet-en, short-en, smart-en, soft-en, straight-en, sweet-en, tight-en, whit-en final /k/: black-en, dark-en, quick-en, sick-en, slack-en, thick-en, weak-en final /d/: broad-en, dead-en, hard-en, madd-en, redd-en, sadd-en, wid-en final /f/: deaf-en, rough-en, stiff-en final /s/: coars-en, less-en, loos-en, wors-en; and fast-en /fa:s-n̩/, moisten /mɔis-n̩/ final /ʃ/: fresh-en final /θ/: smooth-en (and length-en, strength-en—see below).

In OE, only a very small number of words ended in /b/ or /g/ and the few monosyllabic modern adjectives with these final consonants (big, smug, glib) were borrowed from Old Norse or Dutch in post-OE times. As described in 7.2c, OE had just one series of fricative phonemes, with voiceless variants occurring word-finally (these are the modern voiceless fricative phonemes); hence the absence of suitable adjectives ending in a voiced fricative.

7.3.2 suffix- en   187 What is notable is that-en is never added to an adjective ending in a nasal, or /l/, or /r/ (in dialects where this occurs) or a vowel—for example, not to slim, clean, brown, green, keen, cool, noble, dull, cruel, poor, blue, free, low, slow, shy. As a consequence of the monosyllable and final-consonant conditions, we can say wid-en but not *narrow-en, deep-en but not *shallow-en, sweet-en but not *sour-en, quiet-en but not *noisy-en, light-en but not *heavy-en, quick-en but not *slow-en, fatt-en and thick-en but not *thin-en, moist-en but not *dry-en. There is an ingenious ‘way out’ for some adjectives ending in /ŋ/ or a vowel. Derivations *long-en, *strong-en, and *high-en are not a­ dmissible. What happens here is that a verb is formed from the corresponding nominalization, which does end in an appropriate consonant—length-en, strength-en, height-en. Despite this intricate set of conditions on the use of -en in times past, the suffix is no longer productive today. 7.3.2c As relics of an earlier period, there are threat-en, heart-en, and frighten, verbalizations with -en based on nouns. Each may only be used transitively. Threat, threat-en, heart, and fright go back to OE, while heart-en and fright-en were coined during the early stages of Modern English, when -en derivations were in vogue. 7.3.2d It is reasonable to enquire, concerning those adjectives which do not satisfy the conditions for taking -en, how may verbs be made out of them? There are a number of possibilities: (i) The adjective simply does double duty as a verb. This applies for— among others—clean, cool, dry, empty, narrow, slow, and, to a lesser extent, shallow and sour. And see 7.5.2. (ii) There is a verbal derivation involving the Romance-origin prefix en-/ em-, as in en-large, en-noble, and en-feeble. This is discussed in the next section. (iii) Some adjectives form transitive verbs through suffix -(i)fy or -ize; for example fals-ify, prett(y)-ify, pur-ify, simpl-ify, solid-ify, tranquill-ize. See 7.3.5−6 and 7.5.2. (iv) We simply have to use the adjective in a periphrastic frame: become heavy and make (something) heavy. Also for afraid, clever, happy, honest, lucky, noisy, poor, stupid, and more.

188    7 making new verbs This discussion of -en has been somewhat simplified. For fuller details see Sweet (1891: 467−8), Marchand (1969: 271−3), and especially Jespersen (1942: 351−9) and further references therein.

7.3.3 Prefix en-| em7.3.3a We can now consider verbalizing prefixes of Romance origin. Prefix en- eventually reflects Latin prefix in-. Loans from French typically have form en- /en-/, or em- /em-/ before /b/ or /p/ (as in em-bed, em-power). ME took over encroach, enquire, endorse, envelop (which are not analysable within English) and also large, chain, and tomb, followed by en-large, enchain, and en-tomb, establishing en- as an English prefix. There is a little variation between en-| em- and in-| im-, which reflects the original Latin vowel (and between pronunciations with initial /e/ and with /i/). For example, enquire (which is more common in Britain) and inquire (preferred in the USA), em-bed and im-bed. The nominalization corresponding to verb en-crust is in-crust-ation. In a few instances, en- and in- have created distinct lexemes, as with ensure, a general term relating to making sure that something happens, and insure, describing a procedure for providing financial benefit in the wake of some unfortunate event. 7.3.3b There are a number of instances of en- with a verb when it seems to have just an intensifying effect; for example, en-wrap, en-wreathe. Most occurrences are with nouns and adjectives. As mentioned in 7.3.2d, a few adjectives take en-| em- with a causative meaning ‘make’. We have ambitransitive en-large, and transitives en-noble and en-feeble. If you en-able someone to do something, this means that you make them able to do it. This sense also applies with nouns power, rage, and slave, giving transitive verbs em-power ‘make have power’, en-rage ‘make get into a rage’, and en-slave ‘make become a slave’. Also en-courage, en-trust. The verb en-joy, based on noun joy, is of particular interest. One would expect its meaning to be similar to those of em-power and en-rage; that is, ‘make have joy’. Indeed, en-joy did once have such a meaning. The OED quotes, from the late fifteenth century: For to gladde and enjoye the people (meaning, make the people glad and having joy). But this sense fell out of use, to be replaced by the present one ‘have joy about’, as in She en-joy-ed the dinner. There is one sense in which en-| em- may retain a degree of productivity. If one says ‘en-X YO’ (where X is a noun, Y is a noun phrase, and en-X is

7.3.4  competition between suffix - en and prefix en -/ em -   189 a transitive verb), this may mean either ‘put an X around or upon Y’, as in en-circle, en-shroud, en-crown, or ‘put Y in or on an X’, as in en-cage, en-case, en-tomb, en-trap en-snare, en-code, en-throne. A more abstract instance of this sense is en-danger ‘put in danger’. En-camp is an ambitransitive verb, indicating either ‘put (the soldiers) in a camp’ or ‘set up a camp’. Derivations like en-train and em-bus ‘get on/put on a train/bus’ are the opposites of de-train and de-bus ‘get off/take off a train/ bus’, sense (i) of negative prefix de- in 5.10b. Prefix en-| em- often co-occurs with suffix -ment; see 11.4a. In 5.11b we noted that disen-throne is an alternative to dis-throne.

7.3.4  Competition between suffix -en and prefix en-| emGermanic suffix -en and Romance prefix en-| em- may both be used with adjectives, and have similar meanings. They expanded their ranges of use at about the same period, in late ME and early Modern English times. At a similar stage, a number of adjectives were undertaking double duty as verbs (probably, more than today). With only a few exceptions, suffix -en was restricted to Germanic roots, but some of these also accepted prefix en-| em-. For example, rich was an adjective in OE, it also functioned as a verb in the fourteenth century (this use is now obsolete), formed derivation en-rich at about the same time, and finally rich-en about 1800. En-rich is now the most common derivation, meaning ‘make rich in terms of having lots of money’. But there are other senses of rich—when applied to food it can indicate a superfluity of, for example, butter or cream. Verbal derivation rich-en is typically used in relation to this meaning. Germanic adjective bold formed verbal derivations bold-en and em-bold and then—with both prefix and suffix—em-bold-en; it is this which has survived. Em-bright and bright-en were both in currency in the sixteenth and seventeenth centuries, but only the -en form has survived. Similarly for endark and dark-en. Derived from adjective light, we find verbs en-light (now very rare), lighten and—with both prefix and suffix—en-light-en. There has here been specialization. Light-en is used for ‘make lighter’ in both the senses ‘not heavy’ and ‘not dark’, whereas en-light-en generally refers to intellectual or spiritual insight, as in Doctor Smith enlightened us as to the causes of gout.

190    7 making new verbs 7.3.5 Suffix -ate ME created a goodly number of verbs ending in ate by adaptation from Latin participles ending in atus. Thus we got create, debate, translate, copulate, and many more, not analysable within English. Others were borrowed as plain adjective or noun and also as -ate verbalization—urine was followed by urin-ate, we got alien and then alien-ate, liquid and then liquid-ate, and so on. This enabled the recognition of -ate /-eit/ ‘make’ as a verbalizing suffix in English. It has been used just a little, predominantly on Romance forms; for example, activ-ate ‘make active’. If you motiv-ate someone you provide them with a motive for undertaking some endeavour. When a piece of music is orchestra-ate-d, parts are written out to be played by the individual members of an orchestra (and the verb has been extended to other kinds of organization—­one can orchestr-ate a rebellion, or an intricate trade agreement). When a verb came to be needed, in the midnineteenth century, to describe the insertion of the hyphen (a Greek root), hyphen-ize was first used, but was soon replaced by the now standard hyphen-ate. A couple of hundred years ago, a number of -ate verbs were invented on a rather playful basis. For example bibulate, for constant drinking; this was based on the now archaic verb bibble ‘keep drinking’ (and motivated by Latin verb bibere ‘drink’). The most appealing is still with us—titivate ‘make small adjustments to smarten something up’. Goodness knows where this came from; probably, some jovial person just invented it. There are two other suffixes -ate /-eit/, one deriving adjectives from nouns (as in ­affection-ate) and the other deriving nouns (as in consul-ate and doctor-ate). See 8.2.44 and 9.5.6.

7.3.6 Suffixes -ize and -(i)fy The remaining two verb-deriving affixes, -ize and -(i)fy, each has wide productivity. The remainder of this chapter will be devoted to a comparison between them, and also of nouns and adjectives doing double duty as verbs. We shall survey the phonological, semantic, and genetic factors motivating their use. At the end of the chapter, we examine how proper names may be verbalized. Forms less rebarbative than de-nazi-fy and de-stalin-ize can be

7.3.6  suffixes - ize and - (i)fy   191 constructed on the names of recent national leaders. One naturally says de-reagan-ize (rather than *de-reagan-ify) and de-bush-ify (not *de-bushize). But what about Kennedy? Neither an -ize nor an -(i)fy derivation seems to come naturally here. The relevant factors are explored in 7.7b. Introductions to the two suffixes are now provided, as a preliminary. 7.3.6a Greek formed both intransitive and transitive verbs with suffix -izein, which was adapted into Latin as -isare and then into French as -iser. The English suffix is pronounced /-aiz/ and written often as -ize but sometimes, in Britain, as -ise (reflecting the French spelling). The first loan into ME may have been baptize, followed by many other verbs based on Greek roots (scandal-ize, canon-ize, agon-ize, symbol-ize) and also on Romance ones (author-ize, patron-ize, civil-ize, and so on). 7.3.6b In contrast, suffix -(i)fy /-(i)fai/ is of wholly Romance origin. Latin verb facere ‘do, make’ gave rise to derivational suffix -ificare ‘make’, which increased in use during later stages of the language. This led to French suffix -fier and borrowings into English such as glor-ify, and pur-ify. Once the suffix became established in English, it was applied to Romance roots, as in beaut-ify, object-ify, intens-ify, and also Germanic ones, such as ugli-fy and prett -ify. 7.3.6c Suffix -(i)fy (but not -ize) has a special function. It may be used for jocular effect within a fairly intimate register. Someone may say I’ve just dead-ifi-ed the cockroach (where kill might have seemed too brutal a word to employ). In that instance, -(i)fy does derive a verb from an adjective. More often it is just added to an existing verb, indicating a low level of seriousness—They’re always argu-ify-ing about nothing, or Your turn to cook-ify tonight. This is further discussed at the end of 7.5.1 (and see Jespersen 1942: 452). The Grand Duke, in W. S. Gilbert’s comic opera of that name (Act 2) uses the established person-ifi-ed and then echoes it with a nonce coinage: But bless my heart and soul alive, it’s impudence personified I’ve come here to be matrimonially matrimonified Recent useful and helpful discussions of suffixes -ize and -(i)fy include Lieber (1998, 2004) and Plag (1999). Mencken (1936: 193, 1945: 400−2) provides fine exemplification for the productive use of -ize in American English.

192    7 making new verbs 7.4 Phonological conditioning 7.4.1  Phonological conditioning of -(i)fy This suffix is sometimes quoted with basic form -ify, other times as -fy. It is often added to a root ending in /i/ and it is then an arbitrary matter to say which /i/ is retained and which omitted—should it be glory-fy /'glɔ:riˌfai/ or glor-ify /'glɔ:r-iˌfai/? To handle this and other circumstances, it is convenient to refer to the suffix as -(i)fy. (We follow the analysis glory-fy /'glɔ:ri-ˌfai/.) A word ending in -(i)fy /-(i)fai/ must have (with only very occasional and short-lived exceptions):

• • •

the antepenultimate syllable bearing primary stress; the penultimate syllable, which is unstressed, typically having nucleus /i/, although this can—unusually—be some other vowel; the final syllable, /ˌfai/ bearing secondary stress.

For example, beauti-fy /'byu:ti-ˌfai/. There may be one or more unstressed syllables preceding the antepenult as in electri-fy /i'lektri-ˌfai/. Many -(i)fy derivations may accept negative prefix de-, which is unstressed; thus, declass-ify /di:'kla:s-iˌfai/ and de-electri-fy /di:i'lektri-ˌfai/. We can summarize the types of root to which -(i)fy is added. Note that the great majority are of Romance origin. (i) Disyllabic roots, with initial stress, ending in /i/; the suffix then has form /-ˌfai/. For example, glori-fy /'glɔ:ri-ˌfai/ and beauti-fy. Some are most used in past participle form; for example, ladi-fied, gentri-fied, citi-fied, mummi-fied. A fair number have a consonant cluster before the /i/; these include ugli-fy, countri-fy, cockni-fy, dandi-fy, yankee-fy, monki-fy. (ii) Other disyllabic forms with initial stress, ending in a vowel other than /i/; the suffix again has form /-fai/. There is only a small number of established -fy derivations of this form (although more can readily be created). Argu-fy /'a:gyu-ˌfai/ dates from 1751. During the time when Negro was a non-taboo word, we had Negro-fy, as in a quotation of 1859 from the OED: A ring in the nose, aided by all the brown unguent in the world, would have been powerless to negrofy the bold Saxon outline of his features.

7.4.1  phonological conditioning of - (i)fy   193 (Interestingly, there was an alternative contemporary verbalization, Negrify, which has the more common vowel /i/ before the -fy.) (iii) Added to monosyllables; the suffix then has form /-iˌfai/. For example, class-ify /'kla:s-iˌfai/; also gas-ify, jok-ify, speech-ify, fish-ify. The suffix can be added to monosyllables ending in a consonant cluster, as fals-ify, French-ify, sand-ify. There is a shift in vowel quality between type /taip/ and typ-ify /'tip-iˌfai/. There appear to be no established instances of -(i)fy being added to a monosyllabic word ending in a vowel. However, these can readily be produced. For instance, if a tyrannical leader called Ho wished to impose his political ideas on his people, the verb to describe this would be Ho-ify /'hou-iˌfai/. (iv) There are some instances of -ify added to a disyllabic noun or adjective root which bears initial stress and ends in a consonant. In the -ify form, primary stress must be shifted to the second syllable of the root so as to fall in the antepenultimate syllable of the word (a well-nigh inconvertible requirement for -(i)fy derivations). The stress shift may involve vowel change. For example: acid solid person

/'asid/ /'sɔlid/ /'pə:sn/

acid-ify /ə'sid-iˌfai/ solid-ify /sə'lid-iˌfai/ person-ify /pə:'sɔn-iˌfai/

Also humid-ify, subject-ify, and object-ify. Diverse, /dai'və:s/ has stress on the second syllable, and this is retained in divers-ify, /dai'və:si-'fai/. Similarly for intense/intens-ify. And in W. S. Gilbert’s couplet (7.3.6c), matrimonial /ˌmatri'mounyəl/ gives matrimon-ifi-ed /ˌmatri'mouny-iˌfai-d/, retaining the stress placement. (v) Verbs have been created from a number of adjectives ending in -ic by omission of the final /k/—electri-fy /i'lektri-ˌfai/ from electric /i'lektrik/, syllab-ify from syllabic, and de-tox-ify from toxic. As already mentioned, many -(i)fy verbs may take negative prefix de-; one can say de-glori-fy, de-myth-ify, de-class-ify. Indeed, some -(i)fy derivations

194    7 making new verbs are only used with de-; we have mentioned de-toxi-fy. Many (perhaps all) -(i)fy verbs form nominalizations with -(i)cation; for example, glori-fi-­ cation, class-ifi-cation, pur-ifi-cation, quanti-fi-cation. Primary stress goes on the -a- /ei/ of-cation, with the original primary stress now becoming secondary stress. Thus glori-fy /'glɔ:ri-ˌfai/ yields glori-fi-cation /ˌglɔ:ri-fi'-keiʃən/. In this derivation there are two unstressed syllables between secondary and primary stresses. And these -cation nominalizations may combine with de-, which is here always unstressed, as in de-class-ifi-cation /di:ˌklas-ifi'-keiʃən/, and de-nazi-fi-cation. (See 9.4.7.)

7.4.2  Phonological conditioning of -ize Like those ending in -(i)fy, words ending in -ize have a preferred shape:

• • •

the antepenultimate syllable should bear primary stress; the underlying stem (without -ize) should end in a consonant (generally, just a single consonant); the final syllable, ending in /ˌaiz/, bears secondary stress.

For example, moral-ize /'mɔrəˌl-aiz/. As with -(i)fy, there may be an unstressed syllable before the antepenult; for example material-ize /mə'tiriəˌlaiz/. The antepenultimate stress rule is less rigid for -ize than for -(i)fy. Primary stress may go on to the fourth syllable from the end, with two unstressed syllables intervening between this and secondary stress on the final /aiz/, as in institutional-ize /insti'tju:ʃənəˌl-aiz/. The types of words to which -ize may be added can be summarized as follows. (As with -(i)fy, the great majority are of Romance origin.) (vi) A polysyllabic noun or adjective ending in a liquid (/l/ or /r/) or in a nasal (/m/, /n/ or /ŋ/). Quite a number of these have final -al (which was in Latin dissimilated to -ar after an l) or -an, showing their Romance origins. For example:

• •

ending in -l: national-ize, spiritual-ize, scandal-ize, capital-ize, legal-ize, special-ize, cannibal-ize, symbol-ize, tranquill-ize, and many more; ending in -r: familiar-ize, circular-ize, popular-ize, character-ize, bowdlerize, pressur-ize, secular-ize, and many more;

7.4.2  phonological conditioning of - ize   195

• • •

ending in -n: urban-ize, christian-ize, Belgian-ize, union-ize, section-ize, carbon-ize, patron-ize, revolution-ize, Latin-ize, and many more; ending in -m: victim-ize, solemn-ize, item-ize, atom-ize, and some others; ending in -ŋ: very few commonly occurring words. But this type can readily be formed; for example, Keating-ize (based on the name of a former Australian Prime Minister, Paul Keating).

A few words ending in a liquid change vowel quality when -ize is added. For example, satire /'sataiə(r)/, satir-ize /'satəˌr-aiz/, and sterile /'sterail/, steril-ize /'steriˌl-aiz/. We noted that adding -(i)fy may induce shift of stress to adhere to the template for this derivation, as in /'sɔlid/ to /sə'lid-iˌfai/. This does not happen with -ize. Although most derivations have penultimate stress, if stress occurs in the underlying stem on a syllable which would be earlier than the antepenult, then it remains. As mentioned before, we then get two unstressed syllables between the primary stress and the secondary stress on final /aiz/, as in American-ize, /ə'merikəˌn-aiz/ and also in monophthong-ize /'mɔnəfθɔŋˌg-aiz/ (note the insertion of /g/ here). (vii) A polysyllabic noun or adjective ending in -ic, /ik/. There are two ways in which a verbalization with -ize is achieved, both producing a word with the canonical stress profile. One involves dropping the -ic, the other retaining it. (vii-a) Omit the -ic syllable, and add -ize. This applies when the original root, to which -ic was added, was disyllabic and had initial stress. Adjective-deriving suffix -ic generally places primary stress on the syllable immediately preceding it; see 8.2.16. When -ic is omitted and -ize added, stress reverts to the initial syllable of the root. For example: 1. 2.

3.

Disyllabic root with initial stress Add -(t)ic, stress second syllable (the syllable immediately preceding -ic) Replace -ic by -ize, stress reverts to initial syllable (the antepenult in the word)

dogma /'dɔgmə/ dogma-tic /dɔg'mat-ik/

magnet /'magnət/ magnet-ic / mag'net-ik/

dogmat-ize /'dɔgməˌt-aiz/

magnet-ize /'magnəˌt-aiz/

196    7 making new verbs There are quite a few others of this type, including drama| drama-tic| dramatize, system| system-atic| systemat-ize, trauma| trauma-tic| traumat-ize. A similar pattern is found in hypnotic /hip'nɔtik/, hypnotize /'hipnəˌtaiz/, but here there is no simple underlying noun (hypnosis does not fit the pattern). A variant is: democrat /'deməkrat/

democrat-ic / /ˌdemə'krat-ik/

democrat-ize /di'mɔkrəˌt-aiz/

The noun democrat bears initial stress. In adjective democrat-ic, primary stress moves to the third syllable, the one immediately preceding -ic (as required by this suffix). In democrat-ize, stress moves to the second syllable, the antepenult, to fit the preferred stress pattern of -ize forms. A similar pattern is found in anaesthetic /ˌanis'θetik/, anaesthet-ize /a'ni:sθiˌt-aiz/ (there is here an irregular relation between noun anaesthesia and adjective anaesthetic). (vii-b) A number of -ic forms have no corresponding noun, and then the -ic syllable is retained, -ize being added after it (with the /k/ of -ic /-ik/ being palatalized to /s/ before /-aiz/). For example: romantic /rə'mantik/ italic /i'talik/

romantic-ize /rə'mantiˌs-aiz/ italic-ize /i'taliˌs-aiz/

A number of adjectives ending in -ic have no underlying nominal root and consist only of two syllables. The -ic is retained (again becoming /is/ before /-aiz/) and stress remains on the initial syllable, thus satisfying the stress requirements for both -ic and -ize. For example: public

/'pʌblik/

public-ize /'pʌbliˌs-aiz/

The adjective politic /'pɔlitik/ does not have stress on the syllable preceding -ic; but it is not an adjective derived from a noun by adding -ic. Suffix -ize is added to politic (retaining the -ic), giving politic-ize /pɔ'litiˌs-aiz/ with primary stress moving to the antepenult, the preferred stress pattern for -ize derivations. The behaviour of derivations based on poet /'pouit/ is interesting. Since the noun is disyllabic, we get adjective poet-ic with second syllable stressed,

7.4.2  phonological conditioning of - ize   197 /pou'et-ik/. It falls into set (vii-a) so that the derived verb should be poet-ize /'poueˌt-aiz/. This is attested, from 1581. But from 1804 a competitor arose, poetic-ize /pou'et-iˌs-aiz/, an irregular exemplar of set (vii-b). Both verbalizations are in use today. (viii) A noun of at least three syllables ending in /i/. These often refer to a speech act or have abstract reference. The typical profile for this set is a trisyllabic root, with primary stress on the initial syllable (which is the antepenult), ending in orthographic -y (which in almost all cases relates to French orthographic -ie). The final -y /i/ is replaced by -ize /-aiz/, which bears secondary stress. For example: memory harmony agony eulogy

/'meməri/ /'ha:məni/ /'agəni/ /'yu:lədʒi/

memor-ize harmon-ize agon-ize eulog-ize

/'meməˌr-aiz/ /'ha:məˌn-aiz/ /'agəˌn-aiz/ /'yu:ləˌdʒ-aiz/

Others of this ilk include energy, elegy, botany, rhapsody, sympathy, synchrony, plagiary, fantasy, jeopardy. There are also some with more concrete meaning, such as subsidy, scrutiny, colony, deputy. Also in this set are stems of four syllables the last of which has nucleus /i/. Stress goes on the second syllable (which is the antepenult) and is retained in the -ize derivation, this again conforming to the preferred stress pattern for -ize forms. For example: economy /i'kɔnəmi/ econom-ize /i'kɔnəˌm-aiz/ monopoly /mə'nɔpəli/ monopol-ize /mə'nɔpəˌl-aiz/ Also apology, anthology, mythology, philosophy and (with final orthographic e /i/, relating to its ultimate source in Greek) epitome. For category, dictionaries often give pronunciation /'katigəri/, with four syllables, carrying over into categor-ize /'katigəˌr-aiz/. This has a non-­ canonical profile, with stress on the pre-antepenult. Such pronunciations may apply in some circumstances, particularly in the USA. But a very common pronunciation is as /'katigri/, ommitting the /ə/ and thus having three syllables, which then forms categor-ize /'katigˌr-aiz/, a derivation which does show the canonical stress pattern. Exactly the same remarks apply for military and militar-ize.

198    7 making new verbs Theory is a most interesting word. It may be pronounced with three syllables /'θi:əri/, or with two, /θi:ri/. From the disyllabic form one would expect the verbalization to be theor-ify /'θi:ri-ˌfai/, in set (i) of the -(i)fy discussion above. From the trisyllabic form we would expect to derive theor-ize /'θi:əˌraiz/. This is the generally occurring form and it is likely that—irrespective of how the underlying noun theory is treated—theor-ize is normally pronounced with three syllables, including unstressed /ə/ between the initial stressed and the final unstressed syllables; that is /'θi:əˌr-aiz/ rather than /'θi:ˌr-aiz/. When one plays around with -(i)fy derivations, those for most items from set (viii) sound absolutely unacceptable—*botan-ify, *eulog-ify, *jeopard-ify, and so on. In contrast, theor-ify doesn’t sound too bad. This verbalization is in fact attested, but only with marginal occurrence. However, it appears that the disyllabic pronunciation of theory is gradually taking over. In time, this may lead to theor-ify becoming more common, at the expense of theor-ize. (ix) A polysyllabic noun or adjective ending in d, t, or s. In contrast to sets (vi−viii), this type of -ize derivation is not at all common. There are a few disyllabic roots ending in d which take -ize; for example, standard /'standəd/, standard-ize /'standəˌd-aiz/. Also with hybrid, bastard, method, liquid, Soviet. Legitimate /li'dʒitimit/ may add -ize, giving legitimat-ize (attested from 1791) /li'dʒitiməˌt-aiz/. This has major stress on the pre-antepenult and does not conform to the preferred stress pattern for -ize derivations. As a consequence, a shorter form arose (from 1848), legitim-ize /li'dʒitiˌm-aiz/, which does show the canonical stress pattern. Both are in use today, although ­legitim-ize is gaining in frequency over its older rival. Nouns synthesis and emphasis drop their final -is when taking -ize, in order to produce a form with the preferred stress pattern—synthesis /'sinθəsis/, synthes-ize /'sinθəˌs-aiz/. There is one totally irregular -ize derivation—propagand-ize /ˌprɔpə'gand -aiz/. Here stress is kept on the same syllable as in the noun propaganda /ˌprɔpə'ganda/. The resulting -ize form has stress on the penultimate, rather than on antepenultimate and final syllables. The proper derivation should be something like /prɔ'pagənˌdaiz/. (Try saying this aloud; it is scarcely recognizable as related to propaganda /ˌprɔpə'ganda/.) It was noted that many of the words to which -(i)fy is added end in a consonant cluster—monosyllables such as ugl-ify, countr-ify, fals-ify, sand-ify, and

7.5.1  verbs from nouns   199 disyllables such as object-ify, intens-ify. In contrast, -ize is added almost exclusively to forms ending in a single consonant. The main exceptions are categ(o)rize /'katigˌr-aiz/ and milit(a)r-ize. Plus final -rd—in bastard-ize, standard-ize, jeopard-ize—for dialects which preserve an /r/ pronunciation here. Nominalizations in -ation may be formed from many -ize verbalizations; for example union-iz-ation /ˌyunyən-ai'z-eiʃən/. Just like -(i)fi-cation nominalizations, there is primary stress on the penultimate syllable, preceded by two unstressed syllables and then secondary stress on the fifth syllable from the end. And, as with -(i)fy, the negative prefix de-, which is unstressed, can be added to many -ize verbalizations and also to many subsequent -iz-ation nominalizations. For example de-material-ize /di:-mə'tiəriəˌl-aiz/ and dematerial-iz-ation /di:-məˌtiəriəl-ai'z-eiʃən/. This last commences with two unstressed syllables, followed by one with secondary stress—the same as that showing primary stress for the underlying noun material /mə'tiəriəl/— then two unstressed syllables followed by /'zei/ with primary stress and finally unstressed /ʃən/. 7.5 The semantics of productive verbalization 7.5.1  Verbs from nouns Looking first at nouns, the great majority of those verbalized through -(i)fy or -ize (or suffix -ate or prefix en-| em-) are of Romance origin. In addition, many nouns do double duty as verbs without any change in form; they include nouns of Germanic and of Romance origin, in about equal numbers. The relations of meaning between a noun and a verb derived from it cover a fair semantic range. Some of the recurrent meaning relations will now be summarized. The great majority of derived verbs are transitive; it is convenient to use A to indicate transitive subject and O for transitive object. A few are intransitive; S indicates intransitive subject. P indicates a peripheral (non-core) argument. X is used to indicate the underlying noun. (I) ­Meanings of nouns doing double duty as verbs (and some with internal variation) (a) A behaves towards O as an X would. For example, Mary nursed John indicates that Mary behaved towards John as a nurse would to a patient.

200    7 making new verbs Others include bully, butcher, tailor, doctor, boss, host, pilot, referee, shepherd, parrot, dog. A classic pair showing a difference within this type is mother (a child), used to describe someone looking after a child in the way that a mother would and father (a child), which refers to providing (through sexual intercourse) the semen which leads to the conception of a child. (b) A treats O as if O were an X—baby, fool. (c) A makes O into an X—cripple, martyr, heap, group, list, knot, cash. Related to these is verb panic ‘(make) get into a panic’. (d) A puts O in/on X—air (the clothes), table, market, store, can, seat. A subtype of this is where X is used to transport O, as in cart, ship. (e) A puts X in/on O—water, sugar, butter, grease, poison, label, muzzle (dog), man (a ship), bridge (a gap), enamel (a plate). (f) A gets X from O—milk (cow), core (apple), skin (animal), stone (cherry), weed (the garden), tax (citizens). (g) A uses X on O in the way that X is generally used—bayonet, club, knife, axe, iron, ski, brush, comb, mop. Note that there is a second verb based on stone: ‘throw stones at’, as in stone the christians; compare with ‘take stone out of the middle of’ as in stone the cherries. A subtype involves a body part as X, as in elbow (someone out of the way), finger (the precious diamond necklace), and eye (some interesting object). (h) A assigns X to O—price, name, title, date, time, limit, average, total. (i) S (an animal) produces X (a child)—(a cow) calves, (a sheep) lambs, (a mare) foals. (j) S picks X—blackberry, as in she is blackberrying. This list is by no means exhaustive. There are many minor patterns. A small sample is:

• • • • •

shop for P, ‘look in shops for P’ puzzle over P, ‘think about P, which is a puzzle’ mistake O for P, ‘think O is P but this is a mistake’ chance to meet O, ‘meet O by chance’ fish for compliments, ‘try to attract compliments as an angler does fish’

7.5.1  verbs from nouns   201

• •

gesture to someone ‘make a gesture to someone to attract their attention to something’ (motion to someone is similar) soldier on, ‘continue on with a task despite difficulties, as a soldier might continue on in battle’ (a metaphorical sense)

In 2.8f there is a sample of nouns which undertake double duty as verbs, organized according to the semantics of the nouns.

(II)  Meanings of verbs derived from nouns through suffix -ize Turning now to verbs derived from nouns with -ize, we find there are a few which accord with (a–e) above. They include: (a) A behaves towards O as an X would—burglar-ize (a house), patronize, poet(ic)-ize, bowdler-ize (a salacious literary work, in the way that Dr Thomas Bowdler expurgated Shakespeare ‘for family reading’). (b) A treats O as if O were an X—a rather metaphorical use of lion-ize (a celebrity) ‘treat as special and unusual (as a lion is among animals)’. (c) A makes O into an X—victim-ize, pauper-ize, liquid-ize, carbon-ize, crystal-ize, magnet-ize, union-ize, colon-ize, dramat-ize, novel-ize, anthologize, fantas-ize. (d) A puts O in/on X—hospital-ize (a patient), computer-ize (a set of data), memor-ize (some fact). (e) A puts X in/on O—macadam-ize (in the UK) or bitumen-ize (in Australia) a road. However, most -ize derivations—and also almost all those employing -(i)fy —do not correspond to the recurrent meaning relations just listed for double duty lexemes. Many -ize and -(i)fy verbs relate to affecting the inherent nature of something. Indeed, this can be seen by comparing the examples quoted under (e). One can water the plants, sugar the strawberries, man the boats, or bridge a gap, but in each case the added thing remains distinct from that to which it is added. With the -ize derivation macadam-ize, the macadam layers become an indissoluble part of the road. Further semantic relations between a noun and its -ize derivation can now be listed: (k) A provides an X of/to O to/for P—summar-ize, subsid-ize, eulog-ize. A couple of intransitive verbs also belong here. If S offers an apology to P, then S apolog-izes to P; and if S expresses sympathy for P, then

202    7 making new verbs S sympath-izes with P. And eleg-ize may be used either intransitively or transitively, Also in this set are energ-ize, harmon-ize, militar-ize, and motor-ize—if an army is motor-ized, it is equipped with motor vehicles. (l) A subjects O to X. If A applies pressure to O, then A pressur-izes O. If A subjects O to criticism/critical remarks, then A critic-izes O. Similarly for satir-ize, scandal-ize, scrutin-ize, jeopard-ize. Traumat-iz-ing O implies doing something such that O goes into a trauma. Agon-ize is an intransitive verb; if S agon-izes over P, then S subjects themself to agony concerning P. (m) A represents X of/for O. For example, if A is a symbol of O, then A symbol-izes O, as in The fall of the Berlin Wall symbol-iz-ed (that is, represented a symbol of) the end of the Cold War. Also character-ize, epitom-ize, signal-ize. (n) S achieves X. If someone econom-izes on household expenditure then they achieve economy in household expenditure. Also in this set is: A achieves X over O—if someone monopol-izes the cardboard box market, then they achieve a monopoly over the cardboard box market. (o) S follows the pursuit of X—botan-ize, philosoph-ize, theor-ize, and so on. Moral-ize ‘indulge in moral judgements’ is a tangential member of set (o). Note that the verb moral-ize is typically used in a disparaging way, implying that the moralizer is acting in a self-righteous and perhaps hypocritical manner. This is an example of the general principle that every derived form has its own character and meaning, seldom being simply the association of the meaning of the underlying root or stem plus the semantics of the derivational process involved. As with double duty items, the listing just produced is not exhaustive. For instance, it does not include (among others):

• •

deput-ize for ‘act as deputy for’ woman-ize ‘pursue women for sexual purposes (in a manner considered socially unacceptable)’.

And the verb cannibal-ize (for example, a machine) ‘use parts of one machine to repair another’ appears to have only glancing semantic relation to the underlying noun cannibal.

7.5.1  verbs from nouns   203 (III)  Meanings of verbs derived from nouns through suffix -(i)fy Verbs derived with -(i)fy typically mean ‘(make) take on the character of X’. If someone who comes from the country is said to be citi-fied, this implies that they have assumed some of the habits typical of city life. Preach-ify can be used to describe someone who talks as if preaching—their way of talking is reminiscent of the way a preacher delivers a sermon—or else to describe a sermon or series of sermons which are tedious; in both senses the verb has negative connotations. There are fewer derivations with -(i)fy than with -ize (or of double duty forms). Some do fit into the sets presented above. (c) A makes O into an X—mummi-fy, acid-ify, lique-fy, gas-ify, fish-ify (d) A puts O in/on X—class-ify, cod-ify (k) A provides X to/for O—glori-fy, beauti-fy, electri-fy, dirti-fy (m) A represents X of/for O—exempl-ify, sign-ify A further set of verbs created by -(i)fy is similar to (c) but has a significant difference: (p) A makes O have (some of   ) the characteristics of X—Yankee-fy, Cockney-fy, countri-fy, citi-fy, ladi-fy, gentri-fy, dandi-fy, yuppi-fy, monki-fy (or monkey-fy), object-ify, person-ify. (Some of these are most often used in passive participle form, e.g. ladi-fi-ed.) In many circumstances, -(i)fy and -ize have the same meaning, their use being conditioned by the phonological nature of the form which is to be verbalized, as set out in 7.4.1–2. Thus, one can only gas-ify ‘change into gas’, not *gas-ize, and only colon-ize ‘make into a colony’, not *colon-ify. But beyond these instances, there is a tendency for a verb with -(i)fy to have a meaning ‘something like X’ rather than just plain ‘X’. If someone jokes, they are likely to be telling jokes, or making a joke out of something, whereas if they jok-ify they may just be chatting in a jokey sort of way to liven up a conversation. If one says that someone has dirtied an object, there is an implication that it should be kept clean and is now dirty which is not at all a good thing. But if someone is said to have dirt-ifi-ed it, then it is likely that they made it a little bit mucky but this does not really matter at all. A particularly exotic coinage is hungr-ifi-ed. Saying I feel a bit hungr-ified can mean that I am just a trifle peckish, as opposed to the pragmatically

204    7 making new verbs powerful I’m hungry. (Note though that all of these recent coinages have a wide range of meaning. Hungr-ifi-ed can also be used for ‘very hungry’, and dirti-fy for ‘do the dirty on, play a dirty trick on’.) As mentioned in 7.3.6c, these are jocular—often termed ‘slang’ or ­‘colloquial’—uses of the -(i)fy derivation. They may be based on nouns or adjectives, and even on verbs. Perhaps the most established of the latter is argu-fy (first documented in 1751)—this can refer to a disagreement about something, less serious than a full-blown argument. Alternatively, argu-fy can be used to describe arguing endlessly and tediously. Similarly, speech-ify (this from a noun) can refer to adopting an arrogant speech-making-like tone in the course of everyday conversation; or else to be continuously making formal speeches of a long and boring nature. In its jocular or informal sense, -(i)fy is increasingly productive. One can farm-ify a piece of land, which means undertake a bit of farm-type work on it. One can hom-ify a temporary residence, which might involve setting up some familiar photographs to make it seem more like home. List-ify can be used for doing a bit of listing, perhaps not too systematically or seriously. An object can be gold-ifi-ed or silk-ifi-ed a bit. And so on and on. 7.5.2  Verbs from adjectives A verb derived from an adjective, X, will mean either ‘S becomes X’ or ‘A makes O become X’. As described in 7.3.2, suffix -en derives verbs from simple adjectives under a number of conditions, including that they are monosyllabic and end in /p/, /t/, /k/, /d/, /f/, /s/, /ʃ/, or /θ/. These can generally be used either intransitively or transitively—for example The odds on the favourite short-en-ed or The bookmakers short-en-ed the odds on the favourite. For some of those that lack a final consonant to which -en may be added, the adjective simply does double duty as a verb—narrow, slow, cool, dry, smooth, calm, empty, and (with a limited sense) thin. Brave only forms a transitive verb with the meaning ‘be brave to do something’, as in He braved the snow to attend that lecture ‘he was brave to go through the snow to attend that lecture’. As verbs, clean, blind, tame, and free are also restricted to transitive function. We noted in 7.3.2–4 that a few adjectives of Romance origin form a verb by adding Romance prefix en-| em-; thus, en-large, en-noble, en-feeble, em-bitter. An unusual feature of English is the opposition of adjectives beautiful and ugly, and nouns beauty and ugli-ness. The roots are noun beauty

7.5.2  verbs from adjectives   205 (a loan from Old French) and adjective ugly (from Old Norse). Verbs ending in -(i)fy are typically formed from disyllabic nouns ending in /i/, hence beauti-fy (attested from 1526). By analogy, verb ugli-fy (1576) was created from adjective ugly. Much later, verb pretti-fy (1850) was derived from Germanic adjective pretty. Other verbs ending in -(i)fy which relate to adjectives were borrowed entire from French—fals-ify, pur-ify, simpl-ify, and solid-ify. Note that most verbs derived from adjectives through -(i)fy are only used transitively. (Intens-ify is an exception, having both transitivity values.) Suffix -ize is extensively used to derive verbs from adjectives which are synchronically polymorphemic; for example fiction-al-ize ‘make ­become fictional/a work of fiction’, west-ern-ize ‘make become western/have ­characteristics of the west’; also margin-al-ize, christ-ian-ize, circul-arize, spirit-ual-ize, America-n-ize. In addition,-ize is used with adjectives which, while not being analysable within the modern language, include an ­adjectivizing suffix of Romance origin—legal-ize, popular-ize, equal-ize, rational-ize, urban-ize, and many more. These -ize verbalizations are like -(i)fy derivations from adjectives (but unlike-en forms) in that they are almost all restricted to transitive use. An exception is material-ize, which is typically (but not exclusively) intransitive, as in A lucky opportunity ­material-iz-ed for her. Since the meanings of verbs derived from adjectives by means of -ize are varied, it is useful to classify them as relating to:

• • • • • • • • •

quantity—equal-ize, decimal-ize, metric-ize location—internal-ize, external-ize, local-ize, global-ize, circular-ize scope—popular-ize, particular-ize, special-ize, general-ize, marginalize, personal-ize social conventions—legal-ize, legitim(at)-ize, criminal-ize, westernize, urban-ize, democrat-ize, vulgar-ize conforming to norm—regular-ize, standard-ize, rational-ize, formal-ize religion—spiritual-ize, pagan-ize, secular-ize, solemn-ize expression—fictional-ize, dramat-ize, verbal-ize, material-ize nation/language—American-ize, German-ize person/ideology—de-stalin-ize, christian-ize

We also find -(i)fy verbalizations for the last two headings—French-ify, denazi-fy and so on. Whether -ize or -(i)fy is used for a verbalization based on

206    7 making new verbs a proper name is basically determined on the basis of phonology. This is discussed in 7.7. There are, as might be expected, a few idiosyncratic -ize verbs derived from adjectives. They include:

• •

Familiar-ize. This is typically used reflexively; if one familiar-ize-s oneself with something, one makes oneself familiar with it. Penal-ize. It could be said that adjective penal means ‘involving punishment’ and verb penal-ize ‘impose punishment’, but the types of punishment differ significantly. A penal sentence is imprisonment and a penal institution is a prison. But to penal-ize someone is not to send them to prison, rather to impose a penalty, as in The teacher penal-iz-ed him ten marks for not handing in the final assignment on time.

7.6 The ‘how’ and the ‘why’ Whether a noun or adjective does double duty as a verb or forms a derived verb—and, if so, whether by -(i)fy, by -ize, (or by one of the other processes described in 7.3)—depends on three factors:

• • •

the meaning of the noun or adjective. See the list of meanings associated with double duty, -(i)fy, -ize, and -en, enunciated in 7.5; the phonological form of the noun or adjective; its etymology; some roots were borrowed from Romance languages but their verbal derivations developed within English; they include cod-ify and popular-ize. For others, both root and verbalization were borrowed, as with pure and pur-ify, liquid and liqu-efy, harmony and harmon-ize, solemn and solemn-ize.

We first summarize the phonological conditioning factors, described in 7.4.1–2. And then put forward a number of explanations, while examining contrasting verbalizations. 7.6.1  Integrated summary of phonological conditioning The examples of double duty listed under (I) in 7.5.1 were almost entirely monosyllabic and disyllabic forms. They included two words which are

7.6.1  phonological conditioning   207 typically shown as trisyllabic in dictionaries but which are in fact frequently pronounced as disyllables: referee /re'fri:/ and average /'avridʒ/. There is also the verb blackberry, a compound with one monosyllabic and one disyllabic component. Other nominal compounds can also be used as verbs; for example, These hens free-range. Overall, nouns and adjectives which do double duty as verbs are typically Germanic or Romance forms which have one or two syllables, or compounds involving such roots. In 7.4.1 the phonological conditioning of -(i)fy was summarized, and in 7.4.2 that of -ize. (Recall that both are used predominantly with Romance roots.) We can now integrate these conditioning environments. It was noted that -(i)fy verbs require primary stress on the antepenultimate syllable and that -ize derivations prefer this. Since -ize /-ˌaiz/ is a single syllable, this suffix is prototypically attached to a disyllabic form with initial stress—legal /'li:gəl/ and legal-ize /'li:gəˌl-aiz/. When added to a trisyllabic form, with initial stress, ending in -y /i/, -ize replaces the -y—fantasy /'fantəsi/, fantas-ize /'fantəˌs-aiz/. The other suffix has two forms. It is disyllabic -ify /-iˌfai/ when following a monosyllabic root—code /'koud/, cod-ify /'koud-iˌfai/. And it is monosyllabic -fy /ˌfai/ when added to a disyllabic root ending in /i/—beauty /'byu:ti/, beauti-fy /'byu:ti-ˌfai/. Now to summarize the phonological conditioning for -(i)fy and -ize. (Item numbers (i), (ii), . . . from 7.4.1–2 are cross-referenced here.) (α) monosyllabic stems, (iii).

•

Can only take -ify, as in gas-ify.

(β) polysyllabic stems ending in a vowel —disyllabic root, with initial stress, ending in /i/, (i).

•

Add -fy /fai/, as in countri-fy /'kʌntri-ˌfai/.

—root of three or more syllables, with antepenultimate stress, ending in /i/ (generally relating to orthographic -ie in French), (viii).

•

Replace final /i/ by -ize /-aiz/, as in sympathy /'simpəθi/, sympath-ize /'simpəˌθ-aiz/.

208    7 making new verbs These can be trisyllabic nouns, with initial stress—such as sympathy—or quadrisyllabic forms with primary stress on the second syllable—as apology /ə'pɔlədʒi/, apolog-ize /ə'pɔləˌdʒ-aiz/. Two disyllabic forms which bear initial stress and end in a vowel other than /i/ were mentioned under (ii) in 7.4.1; suffix -fy is attested in Negrofy and argu-fy. In fact, verbs formed from nouns or adjectives ending in a vowel other than /i/ are very scarce. How proper names ending in a vowel are treated is discussed in 7.7. (γ) polysyllabic stems ending in a liquid (/l/, /r/) or in a nasal (/n/, /m/ or /ŋ/), (vi).

•

Add-ize. Placement of primary stress in the underlying form is maintained. If the underlying form is disyllabic, with initial stress, the preferred antepenultimate stress pattern for an -ize derivation is achieved, as in victim /'viktim/, victim-ize /'viktiˌm-aiz/. The underlying form may have more than two syllables but show penultimate stress; then the preferred stress pattern will again eventuate, as in editorial /ˌedi'tɔ:riəl/, ­editorial-ize /ˌedi'tɔ:riəˌl-aiz/. Or it may have primary stress on an earlier syllable, as in decimal /'desiməl/, decimal-ize /'desiməˌl-aiz/; this results in a non-preferred pattern, with primary stress on the pre-antepenult, separated by two unstressed syllables from secondary stress on the final /ˌlaiz/.

The corpus assembled for this study includes one disyllabic form ending in /n/ which takes -ify rather than the expected -ize—person /'pə:sn̩/, personify /pə'sɔn-iˌfai/, (iv). Note that primary stress has here had to shift to the second syllable to satisfy the penultimate stress condition on -(i)fy verbalizations. This is discussed in 7.6.2a. (δ) polysyllabic stems ending in -ic /ik/ —original root (without -ic) is disyllabic and bears initial stress, (vii-a).

•

Replace final -ic /ik/ by -ize /-aiz/ as in drama /'dra:mə/, dramat-ic /drə'mat-ik/, dramat-ize /'draməˌt-aiz/. If the original root had more than two syllables with initial stress, then -ize again replaces -ic and

7.6.1  phonological conditioning   209 stress goes on the antepenult, different from its position in the underlying root. Under (vii-a) this was illustrated for democrat /'deməkrat/, democrat-ic /ˌdemə'krat-ik/, democrat-ize /di'mɔkrəˌt-aiz/. —if there is no underlying root (without -ic), (vii-b).

•

Add -ize /-aiz/ after -ic (which assimilates to /is/ before /aiz/), as in critic /'kritik/, critic-ize /'kritiˌs-aiz/. The preferred stress pattern pertains.

There are just a few adjectives ending in -ic, which replace the final -ic with -ify. These were set out under (v)—electri-fy /i'lektri-ˌfai/ from electric /i'lektrik/, syllab-ify from syllabic, and de-tox-ify from toxic. In each instance the -ic form has penultimate stress and so replacing final /ik/ with /-iˌfai/—or final /k/ with /-ˌfai/—results in a verbalization with antepenultimate stress. It seems that all three of these verbalizations entered English from Mediaeval or Post-Mediaeval Latin, so that the use here of -(i)fy—rather than the expected -ize—has an etymological explanation. (ε) polysyllabic stem ending in a stop or fricative. There are a fairly small number of nouns and adjectives of this type which may be verbalized. —disyllabic stem, with primary stress on second syllable, (iv).

•

Add -ify. Stress is retained on the second syllable, satisfying the antepenultimate requirement for an -(i)fy verbalization. For example subject /səb'dʒekt/, subject-ify /səb'dʒekt-iˌfai/; also object-ify, intens-ify, divers-ify.

Note that these roots could not take -ize /-aiz/ unless stress were moved to the initial syllable, something of which there is no instance. Note also that some of the words end in a consonant cluster. We remarked that -ify is often added after a consonant cluster but that -ize scarcely ever is. —disyllabic stem, with primary stress on initial syllable, (iv) and (ix). There appear here to be two competing mechanisms:

•

Add -ify and move primary stress one to the right, as in humid /'hyu:mid/, humid-ify /hyu:'mid-iˌfai/. Also acid-ify, solid-ify.

210    7 making new verbs

•

Add -ize and retain stress, as in liquid /'likwid/, liquid-ize /'likwiˌd-aiz/. Also standard-ize, hybrid-ize, bastard-ize, method-ize.

The question then is: why do some verbs of this type take -ify and others -ize? The answer appears to be etymological. Humid-ify, acid-ify, and solid-ify were borrowed entire from French. But for liquid, standard, hybrid, bastard, and method, it was just the root which was borrowed from French (or Latin); creation of a verb by adding -ize happened at a later stage within English. This indicates that -ize is the productive verbalization for disyllabic roots ending in a stop and bearing stress on the initial syllable. In summary, we have seen that:

•

-(i)fy is a productive verbalizer with:

– monosyllabic forms; these take -ify /-iˌfai/; – disyllabic forms, with initial stress, ending in /i/ (and probably also those ending in other vowels); these take -fy /-ˌfai/; – disyllabic forms ending in a stop or fricative, with stress on the second syllable; these take -ify /-iˌfai/.

•

-ize is a productive verbalizer with:

– forms ending with a liquid or nasal; – disyllabic forms ending in a stop or fricative, if stress is on the initial syllable; – forms ending in -ic /ik/; – forms of three or more syllables ending in -y /i/ (corresponding to French orthographic -ie); here -ize /-aiz/ replaces final y /i/. There are a number of exceptions to these general rules, some of which will be explained in the next section.

7.6.2  Some explanations and contrasts 7.6.2a Why does the verbalization based on person /'pə:sn̩/ involve -ify, giving person-ify /pə:'sɔniˌfai/, which requires stress to shift to the second syllable? Why could not person take suffix -ize, as do patron, human, and woman? The answer lies in the semantics of the verbalization. When a verb was first coined from person (the earliest written record is from 1728), it had

7.6.2  some explanations and contrasts   211 the meaning ‘attribute human characteristics to’. This belongs to set (p) of 7.5.1, ‘A makes O have (some of   ) the characteristics of X’. Set (p) relates only to suffix -(i)fy, not to -ize. Thus, the verb created around the 1720s was person-ify, not person-ize. Later, the verb took on a second sense, ‘exemplify, represent’, which falls into set (m) of 7.5.1, that relates to both -(i)fy and -ize. If this second sense had come first, then on semantic grounds either -(i)fy or -ize could have been used, and it is likely that -ize would have been chosen since person has the appropriate phonological form for an -ize derivation. But the ‘characteristics’ sense was prior, and this serves to explain why the verb created was person-ify. (Note that there is another verb, based on the derived adjective person-al, and here -ize is the suffix used, giving person-al-ize ‘make become personal’.) 7.6.2b There are three major components of music—rhythm, melody, and harmony. The verb harmon-ize ‘create harmony, make harmonic’ is of frequent occurrence whereas melod-ize and rhythm-ize (although they have been used) are not in common currency. Why? The likely explanation is pragmatic. Music consists basically of melody (a succession of notes) and rhythm (a pattern of beats or pulses). To these is added harmony (combinations of different notes sounded simultaneously). Given melody and rhythm, one can then harmon-ize. One would less often want to talk of melod-iz-ing or rhythm-iz-ing. 7.6.2c Let us now look at a quite different semantic domain, the names of trade and professional people. Some are derived from verbs—bake-r from bake, teach-er from teach. Others are not. Of these, tailor, butcher, doctor, nurse, shepherd, and engineer are basically nouns which also do double duty as verbs. Why cannot milliner and haberdasher also function as verbs? Perhaps because they are too long; it was noted in 7.6.1 that the great majority of nouns which do double duty have just one or two syllables. But if doctor can be a verb, why can’t dentist? And if butcher can be, why not grocer? (All these nouns are disyllabic.) We can note that the noun burglar (a sort of trade) does form verb ­burglar-ize. This throws up another question—why do we have burglar-ize when there is a perfectly good monomorphemic verb burgle? Here an answer can be provided. The activity noun burglary (c.1200) and agent noun burglar (c.1268) came from Romance. For more than six hundred years there was no commonly used verb to describe the activity. Then, more or less

212    7 making new verbs simultaneously, two were created—verbalization burglar-ize is first attested (in the USA) in 1871, and the back-formation verb burgle in 1872. As mentioned in 7.5.1, -ize derivations typically describe ‘affecting the inherent nature of something’. In keeping with this, verb burgle generally refers just to the stealing of goods, whereas burglar-ize is likely to be employed if there is accompanying violence, with the house being ransacked or trashed. Note that burglar has an appropriate form to take -ize—a disyllabic noun with initial stress, ending in r. Dentist, ending in a consonant cluster, does not qualify, which is probably why there is no dentist-ize. But grocer and ­lawyer have similar form to burglar. There seems no reason why, in the absence of a double duty function, there should not be verbalizations lawyer-ize and grocer-ize, parallel to burglar-ize—save that language is to some extent an artefact of human whim, rather than a mechanical device. 7.6.2d Many of the explanations for ‘why this verbalization?’ are phonological. In 2.2e we enquired why one says central-ize, with -ize, but glorify, with -(i)fy. Simply because a disyllabic form with initial stress, ending in /l/ takes -ize, whereas a disyllabic noun, again with initial stress, ending in /i/ takes -(i)fy. And why should noun scrutiny form a derived verb scrutin-ize whereas noun review does double duty as a verb This is because the form of scrutiny is appropriate for verbalization with -ize whereas that of review /rə'vyu:/, with stress on the second syllable and final /u:/, is not appropriate for verbalization by either -ize or -(i)fy. Typically, a Germanic noun undertakes double duty, while a Romance loan of similar meaning will take -ize. For example: verb = noun shop cripple

verb = noun plus -ize hospital-ize victim-ize

As mentioned before, there are a fair number of examples where a verbalization with -(i)fy was borrowed from Romance, together with the underlying noun or adjective. These include: just and justify pure and purify liquid and liquefy

false and falsify simple and simplify terror and terrify

7.6.2  some explanations and contrasts   213 Now just, false, and pure are monosyllables. If a verb were to be derived from them within English it could only be by -(i)fy and would have exactly the same form as the borrowed verb. Similarly for simple, which is disyllabic (the second syllable consisting of syllabic l ) but could only form verb simpl-ify. However, liquid and terror are disyllabic. They could—and do—form a verb by the addition of -ize: liquid-ize and terror-ize. These have quite different characteristics:

•

Liquid-ize (1837) has virtually identical meaning to liquefy (1547) and may differ only in terms of context of use. The Collins Cobuild English dictionary states that liquid-ize is used for making something liquid ‘through an electrical appliance’, yet many people use liquefy for this. Current Oxford English dictionaries state that liquefy is mainly used in chemistry, but in fact the usage is much wider. It seems that liquefy and liquid-ize have essentially the same meaning and are generally interchangeable. (There is also the verb liquid-ate, 1670, ‘put an end to’, which is based on Mediaeval Latin liquidare. Note that liquefy comes from French liquéfier.) Terrify (1578), based on Latin terrificare, means ‘fill with terror, frighten’. Much later, a new verb terror-ize (1823) was created. At first it meant much the same as terrify (like liquefy and liquid-ize) but by the 1850s a second sense emerged: ‘rule by terror, maintain power through a rule of terror’.

•

7.6.2e There are a number of examples of a noun forming a verb with -ify and an adjective derived from the noun forming a verb with -ize. For example: noun myth verb person

verb myth-ify verb-ify person-ify

adjective myth-ic verb-al person-al

verb myth-ic-ize verb-al-ize person-al-ize

Myth-ify (1873) and mythic-ize (1840) have very similar meanings ‘create a myth about someone or something, treat as a myth’. There is also the noun myth-ology (c.1420) from which is derived verb myth-olog-ize (1603) ‘follow the pursuit of mythology, interpret with respect to mythology’ (along the lines of philosoph-ize, botan-ize, and theor-ize). Verb-ify (1813) means

214    7 making new verbs ‘convert (for example, a noun) into a verb’; ver-bal-ize (1609) also has this sense, and can mean besides ‘express in words’ (or ‘express in too many words, be verbose’). Person-ify and person-al-ize were discussed before. 7.6.2f Some apparent redundancies have an indirect etymological explanation. Why do we have both syllab-ify /si'labiˌfai/ and syllab-ize /'siləˌbaiz/ from syllable /'siləbl/? Well, syllable (c1384) and syllabize (1656) were taken from Romance, and then syllab-iz-ation (1926) was created from syllabize. But, before this, syllabification (1838) had been adopted from Mediaeval Latin, and finally syllabify (1926) was back-formed from this. 7.6.2g There are also examples of -ize applying to two forms of the same word. Middle English took from Latin colony (1382) and at a later stage derived from this colon-ize (1622) ‘make (some territory) into a colony’. Then came adjective colonial (1776) ‘relating to a colony’ and, based on this, colonial-ize (1864) ‘(make) take on the characteristics of someone living in a colony’. 7.6.2h Suffixes -(i)fy and especially -ize are highly productive. What more natural than that they should take over from a suffix which is no longer fully productive, such as -ate? Adoption of domestic-ate (1639) from Mediaeval Latin was closely followed by the creation within English of domestic-ize (1656). Domestic-ate is still far and away the more popular but domestic-ize continues on. It is only slightly fanciful to predict that domestic-ize is likely to gradually take over in the centuries ahead.

7.7 Deriving verbs from proper names 7.7a Why is it normal and natural to use French-ify (rather than French -ize) to describe translating something into the French language or making someone or something take on characteristics of the culture of France, but German-ize (rather than German-ify) with respect to German language and culture? The answer is phonological. French is a monosyllabic word, which takes -ify; German is a disyllabic word with initial stress ending in -n, naturally taking -ize. In each derivation, primary stress falls on the antepenult. (German-ify, with stress moved to the second syllable, and French-ize have been used, but very sparsely.)

7.7 deriving verbs from proper names   215 One can only readily verbalize a noun or adjective if it has an appropriate phonological form, as set out in 7.4.1–2 and summarized in 7.6.1. In addition to a name for each nation, there is also a noun describing an inhabitant, and an adjective describing the language or culture (the last two may have the same form). As set out in Table 7.1, a verb may be derived from one of these items, sometimes from more than one, depending on their phonological form. (Note that the table is not exhaustive; there are more examples for almost every set.) In set A of Table 7.1, adjectives French and Welsh and inhabitant names Dane and Scot are all monosyllabic forms and so take -ify. Danish, Scottish, and Denmark are disyllabic but end in a sibilant or /k/, which are not well attested with -(i)fy or with -ize. Compounds such as Scotland, Frenchman, and Welshman sound less felicitous when verbalized (with -ize) than do Scot-ify, French-ify and Welsh-ify. (France and Wales are monosyllables and could take -ify, but verbs based on the corresponding adjectives, French-ify and Welsh-ify, are preferred. Franc-ify is attested but it is many many times less frequent than French-ify.) America, in set B, is trisyllabic, ending in a vowel, with stress on the second syllable. Mexico is trisyllabic with initial stress and final /ou/ while Germany is the same but with final /i/ (not related to French orthographic ie). None of these have an appropriate form to take a verbalizer, but their inhabitant/ adjective forms all end in -an and readily verbalize with -ize. Language names in set C are disyllabic with stress on the second syllable, an appropriate form to be verbalized with -ify—Nepal /ni'pɔ:l/, Nepal-ify /ni'pɔ:l-iˌfai/; Japan /dʒə'pan/, Japan-ify /dʒə'pan-iˌfai/. A verb can be derived through -(i)fy either from nation name Turkey or from inhabitant name Turk, in set D; the derivations Turki-fy and Turk-ify fall together as /'tə:kiˌfai/. Two verbalizations are also possible for sets E and F, but with different forms. In set E, nation name Israel /'izreil/ bears initial stress and forms verb Israel-ize /'izreiˌl-aiz/, while inhabitant/adjective form Israeli /iz'reili/ has stress on the second syllable and gives verb Israeli-fy /iz'reili-ˌfai/. For set F, the use of -(i)fy and -ize are reversed with respect to set E. I­ nhabitant/ adjective form Russian takes -ize to form Russian-ize (first attested in 1831), while nation name Russia /'rʌʃə/ loses final /ə/ and then adds -ify to derive verb Russ-ify /'rʌʃ-iˌfai/. The OED ascribes the first written instance of this to Queen Victoria, in 1865: Good Alice seems quite Russified. Both verbs are in use today, with Russ-ify generally being p ­ referred over Russian-ize.

NATION

VERBALIZATION

INHABITANT

VERBALIZATION

ADJECTIVE

VERBALIZATION

A1

France

—

Frenchman

—

French

French-ify

A2

Wales

—

Welshman

—

Welsh

Welsh-ify

A3

Denmark

—

Dane

Dan-ify

Danish

—

A4

Scotland

—

Scot

Scot-ify

Scottish

—

B1

America

—

American

American-ize

American

American-ize

B2

Mexico

—

Mexican

Mexican-ize

Mexican

Mexican-ize

B3

Germany

—

German

German-ize

German

German-ize

C1

Nepal

Nepal-ify

Nepalese

—

Nepalese

—

C2

Japan

Japan-ify

Japanese

—

Japanese

—

D

Turkey

Turki-fy

Turk

Turk-ify

Turkish

—

E

Israel

Israel-ize

Israeli

Israeli-fy

Israeli

Israeli-fy

F

Russia

Russi-fy

Russian

Russian-ize

Russian

Russian-ize

216    7 making new verbs

Table 7.1  Verbalizations based on nation names

7.7 deriving verbs from proper names   217 7.7b Verbs are also formed from the names of prominent people and ideologies. Bush-ify means ‘make more closely conform to the ideas set forth by President Bush’. We earlier mentioned de-nazi-fy and de-stalin-ize which mean ‘divest of the insidious influence of Nazis/Stalin’. Here there is only one form as a possible candidate for verbalization, so whether there is a felicitous-sounding verb depends on the phonological form of the name. (Proper names generally do not do double duty as verbs.) For example (note that negative prefix de- may optionally be added to any of these):

• •

monosyllabic names take -ify—Bush-ify, Blair-ify, Ford-ify, Burns-ify. names with penultimate stress ending in /i/ take -fy. These can be disyllabic, as Trotsky /'trɒtski/, Trotski-fy /'trɒtski:-ˌfai/, and (de-)nazify. Or they can be longer, as Disraeli /diz'reili/, Disraeli-fy /diz'reili-ˌfai/, Mussolini /ˌmusə'li:ni/, Mussolini-fy /ˌmusə'li:niˌ-fai/. disyllabic names, with initial stress, ending in a liquid, a nasal, /d/ or /t/ take -ize—Carter-ize, (de-)hitler-ize, Reagan-ize, (de-)stalin-ize, Gladston-ize, Whitlam-ize, Howard-ize, (de-)soviet-ize.

•

Names with a different phonological profile are less easily verbalized. For example:

•

From disyllabic names Isaacs and Adams can be formed Isaac-ize and Adam-ize. The final s is here dropped, since -ize does not normally follow a sibilant or a consonant cluster. But note that the monosyllabic name Burns retains the final s in Burns-ify, since -ify may follow a consonant cluster. Consider the name of the long-serving Australian Prime Minister, Menzies /'menziz/. One could say Menzi-fy /'menzi-ˌfai/, dropping the s and adding-fy, or Menzies-ize /'menziˌz-aiz/. Both appear possible, but both are less felicitous than prototypical verbalizations such as Bush-ify and Clinton-ize. Castro /'kastrou/ has penultimate stress and ends in a vowel. Castro-fy is possible (similar to Negro-fy) but Castro-ify—with final /ou/ replaced by /-iˌfai/—sounds much better. Roosevelt is a disyllabic form with initial stress ending in a cluster. It is because of the cluster (disliked by -ize) that Roosevelt-ize does not sound really good (perhaps just acceptable).

• • •

218    7 making new verbs

•

Kennedy /'kenədi:/ is trisyllabic with initial stress, ending in /i/. Neither verbalizing suffix goes well with this. One could use -fy, which would involve shifting stress to the second syllable. Because of the stress difference Kennedi-fy /ke'nidi-fai/ does not sound all that similar to Kennedy /'kenədi:/. It is not possible to replace the final -y /i/ by -ize / aiz/—saying Kennedize, as in subsidy/subsidize—since it does not relate to orthographic ie in French. One could simply add -ize at the end of the whole name, Kennedy-ize /'kenədi:ˌ-aiz/ but this sounds rather awkward (it has a non-preferred stress pattern). The fact is that the phonological form of the name Kennedy is such that it cannot very easily be verbalized. The name of the preceding president, Eisenhower /'aizənˌhauə/, fares only a little better. An -ize derivation, Eisenhower-ize, is awkward and scarcely acceptable. For an -(i)fy verbalization, primary stress must go on the antepenultimate syllable, giving Eisenhower-ify /ˌaizən'hauəriˌfai/. This is recognizable as related to Eisenhower /'aizənˌhauə/, since there has simply been interchange of primary and secondary stresses between first and third syllables.

•

7.7c In conclusion, we saw that the principles for deriving verbs from common nouns and adjectives depend on a combination of etymology, semantics, and phonological form. Just the phonological factors carry over into the verbalization of proper names. The study of how verbs may be formed from names confirms these basic principles inferred from investigation of common nouns and adjectives. (There are some exceptions to these principles, but overall a fairly small number.) We can now move on, in the next chapter, to the rather more extensive means for creating new adjectives. This chapter is a thorough revision of my paper ‘Deriving verbs in English’, published in Linguistic Sciences 30: 31–52, 2008.

8 Moon-like, death-ly, angr-y, mysteri-ous, memor-able  : Making new adjectives

8.1 Introduction 219 8.2 Adjectives derived from common nouns and from adjectives  224

8.3 Adjectives derived from proper names 265 8.4 Adjectives derived from verbs  278

8.1 Introduction 8.1a In English, as in many other languages—though not in all—there are a fairly small number of simple (monomorphemic) adjectives. These are augmented by a large number of derived forms. This is an account of how adjectives are created from nouns, verbs, and adjectives (plus a few prepositions), by the addition of a derivational affix. Each affix has its own range of meaning and is limited in the roots it may be added to. This is, in part, determined by the phonological forms of roots, by the semantic types they belong to, and by whether they are of Germanic or Romance (or Greek) origin. For example, we need to explain why an adjective can be formed from gold by adding suffix -en, producing gold-en, but not from silver (today there is no *silver-en). Why there are adjectives heaven-ly and hell-ish, but not *heavenish or *hell-ly. People think of -ful and -less as having complementary meanings; from fear are derived both fear-ful and fear-less. Yet these two suffixes do not always operate in parallel; we have brain-less but not *brain-ful, and forget-ful but not *forget-less.

220    8 making new adjectives There are a number of nouns which may form an adjective in two or more ways, involving different suffixes which have distinct meanings. What is the difference between mann-ish, man-ly and man-like, between wood-y, ­wood-en and wood-ed, between yellow-ish and yellow-y, between master-ly and m ­ aster-ful, and between child-less and child-free? These and other semantic pairs will be discussed and explained. 8.1b The affixes discussed in Chapters 5 and 6 generally do not change word class (or sub-class) membership, being concerned just with a semantic addition, as in un-happy, ultra-violet, quasi-scientific. About fifty of them may be added to an adjective, creating another adjective, as in the three examples just given. Just a handful of the affixes described in Chapters 5 and 6 have a secondary function in which they do change word class. Anti- ‘used, against’ was discussed in 5.24 and pro- ‘in favour of’ in 6.8.2. From nouns aircraft and democracy are formed adjectives anti-aircraft (gun) and pro-democracy (movement). Six derivational prefixes which are related to prepositions are discussed in 6.6.5. Most of their occurrences do not change word class, but there are just a few instances of each which do. Adjectives created from nouns include out-doors, in-house, off-screen, on-stage, up-river, and down-hill, while from verbs we get out-spread, on-going, up-lifted, and down-trodden. In Chapter 7, attention is paid to nouns and adjectives undertaking double duty as verbs, and in Chapter 9 to verbs and a few adjectives doing double duty as nouns (and see the general discussion in 2.8). There is little to say on this matter in the present chapter. There are no instances of a form which is basically a verb doing double duty as an adjective. Just a handful of lexemes function as noun and as adjective with opinions varying as to which word class membership is primary—fat, native, fun. 8.1c There are a number of forms which create a limited number of adjectives and are borderline candidates for the status of derivational suffix. Some of the contenders are:

•

Happy ‘over-eager to be involved with’, has a degree of productivity but falls short on the criteria for being an affix which were set out in 3.2. One hears the trigger-happy gangster and the gossip-happy anthropologist and so on. Such formations are probably best regarded as compounds.

8.1 introduction   221

•

Right occurs attached to a number of prepositions, but with a different meaning for each. Compare upright ‘vertical stance, honourable behaviour’, downright ‘emphatically unpleasant’ (as in downright hostility), outright ‘direct, absolute’ (as in outright lie, outright victory), and forthright ‘straightforward, decisive’ (as in forthright manner). Others on the borderline are crazy (and mad ), which can be added to boy, girl, cat, horse, sports, football, and rock-music; sick, which can be used with the name of a vehicle (car, bus, boat, and so on) and with home; and friendly, which can be attached to child, user, and reader. In each instance, extensions are possible just to new nouns of the same type, as is typical for compounds.

•

8.1d There are a handful of minor Germanic suffixes which essentially derive adverbs; some of these may do double duty and also function as an adjective in modifying a noun. For example: a home-wards journey, a side-ways movement, in a clock-wise direction, a ten-fold increase, a holiday-style mood. They are discussed in 10.6. 8.1e This chapter deals with 31 suffixes and one prefix (a-) which serve to create adjectives from nouns and verbs plus—in a couple of instances—from prepositions. Six of them also derive an adjective from an adjective. Their properties are summarized in Table 8.1. As in Table 5.2, ‘PP’ indicates highly productive, ‘P’ mildly productive, ‘Sc’ used a good deal in scientific terms, ‘(joc)’ used in a nonce formations of a jocular nature. And a ‘—’ in this column represents scarcely (or not at all) productive. The ‘origin’ column shows whether the affix is a Germanic retention, or a loan from Romance languages, or a form which goes back to Greek (generally entering English through a Romance conduit). Suffix (25) -i is found in both Semitic and Indo-Iranian languages. If a suffix has a cognate free form, this is indicated by an *. The chapter has three parts. Section 8.2 is concerned with affixes which make adjectives from a common noun (CN), and sometimes also from an adjective. Section 8.3 considers adjectives based on a proper name (PN). Then in Section 8.4 we look at adjectives derived from verbs. Some affixes with a wide range of functions feature twice—two of those from 8.2 are also in 8.3, and eight from 8.2 also in 8.4. (There is just a little further overlap, mentioned within the individual entries.)

Table 8.1  Summary of properties of suffixes which create new adjectives FORM

USED WITH

MEANING

(1)

PP

Gmc

-like*

CN, MW

similar to, characteristic of

(2)

P

Gmc

-ly

CN, A, MW

desirable quality associated with

(3)

PP

Gmc

-ish

CN, A, PN, PREP, V, MW

with CN: quality (often negative) associated with; with A: a bit, approximately; with PN: associated with

(4)

P

Rom

-esque

CN, A, PN

having the style of, in a pleasing manner

(5)

PP

Gmc

-y, -sy, -ity/-ety

CN, A, V, MW

with CN, A: characterized by; feels/behaves like a; is engrossed in; with V: liable to

(6)

—

Gmc

-en

CN

made from, has the appearance of being made of

(7)

PP

Gmc

-ed

CN, MW

provided with something which is not an inherent part

(8)

Sc

Rom

-ous, -ious, -eous,

CN

characterized by (predominantly with abstract nouns)

(9)

P, Sc

Gk, Rom

-ic, -(a)tic, -ical

CN, PN, V

(quality) associated with

(10)

—

Rom

-al, -ar

CN

belonging to, relating to

(11)

—

Gmc

-ful*

CN, A, V

with CN: showing; characterized by; with V: likely to do; likely to be verb-ed

(12)

(joc)

Gmc

-some

CN, A, V

with CN: apt to undertake; apt to be a/have; with V: apt to (cause), likely to do (with a negative quality)

(13)

PP

Gmc

-free*

CN, MW

lacking something considered as undesirable

(14)

PP

Gmc

-less

CN, V, MW

with CN: lacking something one would expect (and, usually, hope) to have; with V: lacking something

(15)

PP

Rom

-able(*), -ible

CN, V & phrasal Vs

with CN: suitable for, characterized by; showing; with V: can be verb-ed; suitable to be verb-ed; likely to be verb-ed

222    8 making new adjectives

ORIGIN

Table 8.1 (continued) ORIGIN

FORM

USED WITH

MEANING

P, Sc

Rom

-ary

CN

is a, is a part of, is like a

(17)

PP

Gmc

-proof*

CN, V, MW

with CN: keeps out; with V: to stop—happening

(18)

—

Gmc

-ern

compass points

in a location in

(19)

—

Gmc

-most

CN, PREP

furthest in a particular direction

(20)

P

Gmc

-ward(s)

CN, PREP

in the direction of

(21)

—

Rom

-ate

CN

full of

(22)

Sc

Rom

-ine, -ene

CN, PN

property associated with

(23)

(joc)

Ger

a-

CN, V

with CN: located on, at, in; with V: in a state of

(24)

P

Gk, Rom

-ite

PN

person associated with a certain place, doctrine associated with a certain person

(25)

P, Sc

IE, Semitic

-i

PN

person or quality associated with a certain place

(26)

P

Rom

-ese

PN, CN

person or quality associated with a certain place

(27)

PP, Sc

Rom

-(a)n

PN, CN

quality associated (predominantly) with a person

(28)

PP, Sc

Rom

-ian, -ean

PN, CN

quality associated (predominantly) with a person

(29)

(P)

Rom

-((a/i)t)ive

V, CN

liable to do, can do

(30)

—

Rom

-(at)ory

V

does, tending to do, serves to do

(31)

—

Rom

-ent, -ant

V

having the quality of verb-ing

(32)

P

Gmc

-worthy*

V, CN, MW

be appropriate to be verb-ed

Note: Used with: CN, common noun, A, adjective, PN, proper noun, V, verb, PREP, preposition, MW, multi-word combination. * indicates cognate free form.

8.1 introduction   223

(16)

224    8 making new adjectives 8.1f Of the 32 adjectivalizing affixes, (26) -ese takes primary stress while (1) -like, (4) -esque, (13) -free, (17) -proof, and (32) -worthy generally take secondary stress, and (19) -most sometimes does so. Seven suffixes may affect stress placement in the base: (8) -ous, (9) -ic, (10) -al, (16) -ary, (25) -i, (28) -ian, and (30) -(at)ory. The remaining affixes neither take stress themselves nor affect stress placement in the base. 8.1g Some productive suffixes (all of them Germanic in origin) can be added to a multi-word construction—indicated by ‘MW’ in the ‘used with’ column of Table 8.1. Most often they will be attached to a fairly familiar collocation, but need not be restricted to such. A suffix may be added to the last word of a coordination, as with It’s [dirt and dust]-free; The warrior is [shield and sword ]-less; He experienced a distinctly [bread-and-water]-y diet in the kidnappers’ hideout; She’s both [trust and credit]-worthy. Or, a suffix may be added to a complex NP, as in She put on a distinctly [queen of all she surveys]ish air. Of Romance-origin suffixes, -able is the most thoroughly integrated into English. It is add-able to a phrasal verb (which is one lexical unit), either after the second element (put-off-able) or, more recently, after both (put-able off-able). 8.1h Between entries on individual affixes are interposed short sections on contrast. These have a variety of functions, sometimes several at once. Comparison of affixes may assist in best explaining the meaning of each; it may illustrate phonological constraints on usage, and it may illuminate correlations between base and affix in terms of genetic origin. Some of the contrasts are of a general nature, illustrating productivity factors, while others just discuss a couple of critical words.

8.2 Adjectives derived from common nouns and from adjectives There are about two dozen affixes which create adjectival stems from common noun roots; six of them also apply to adjective roots. As a basis for description and explanation, it is useful to roughly classify the nouns of English into a number of semantic types. (Just a representative set of exemplars is given for each.) (A) Body parts, other parts and related states: nose, chin, tooth, beard, hair, leg, brain, mind, adenoids, liver, fat, pimple, bone, skeleton,

8.2 derived from common nouns and adjectives   225 tear, blood, faeces, breath, venom, fur, feather, root, anaemia, bronchitis, diabetes, asthma, gout, fever, disease, pain, hormone (B) Description of types and groups of people —age-group and sex terms: man, woman, lady, gentleman, child, boy, girl, baby —role or profession: king, president, chief, soldier, scholar, merchant, poet, artist, secretary, editor, aristocrat, despot, nomad, heathen, ghost, god, devil, fiend, monster, satan, elf, burglar, thief, master, workman, statesman, scholar —social role: friend, enemy, rival, neighbour, companion, guest, host, democrat, amateur —groups: family, audience, club, department, army, tribe, nation, republic, empire, colony (C) Kin: mother, father, child, son, daughter, husband, wife, aunt, uncle, sister, brother (D) Fauna (Non-human animates): monkey, sheep, horse, wolf, kitten, swine, hog, goat, mule, mouse, cat, snake, fish, hawk, owl, spider, wasp, ant, mosquito, louse, vermin (E) Flora: wood, tree, flower, leaf, fruit, grass, seed, oak (F) Nature: nature —elements and materials: air, fire, spark, smoke, water, vapour, ice, foam, silk, wool, leather, wax, stone, pebble, metal, gold, silver, lead, acid —consumables: food, meat, milk, fruit, sugar, cheese, flour, treacle, crumb, calorie, poison —geography: hill, mountain, cavern, precipice, plain, valley, desert, river, lake, sea, ocean, land, island, country, sand, dirt, dust, stone, pebble, earth —celestial: sun, moon, star, shade, shadow, sky, world —miscellaneous locations: universe, globe, home, heaven, hell —weather: weather, wind, rain, drizzle, thunder, snow, mist, fog, damp, draught —season: season, summer, winter, spring, autumn (G) Property, buildings and institutions: place, property, farm, village, school, shop, prison, house, room, door, window, partition (H) Artefacts: hat, shirt, chair, cushion, bed, pillow, table, car, plane, axe, radio, book, saddle, string, shackle, gun, sword, shield

226    8 making new adjectives (J) Noise and language —noise: noise, squeak, creak, whisper, sound —language and music: language, dialect, name, word, noun, vowel, syllable, diphthong, rhythm, melody, harmony —speech acts: conversation, statement, question, answer, reply, argument, apology, epigram, joke, story (K) Science and art: science, chemistry, history, politics, literature, fiction, poetry, geography, philosophy, logic, grammar, art, painting, picture, statue, sculpture, photograph, dance, opera, sport, superstition, religion (L) Activity and experience —activity: action, deed, activity, experience, event, business, trade, industry, problem, strategy, experiment, trick, trial, party, quarrel, war, murder, suicide, funeral, marriage, adventure, accident, dream, nightmare —general: motion, rest, state, situation, weather, temperature, weight, space —experience: life, death, origin, spirit, peace, liberty, luck, fancy, chance, risk, fate, failure, victory, defeat, disaster, riot, chaos —cause and the like: cause, effect, result, reason, proof, evidence, purpose, function, method, condition —mental knowledge: idea, concept, opinion, belief, faith, memory, thought, doubt, problem, puzzle —fact: fact, sense, nonsense, truth, error, mistake —physical: effort, force, power, energy —vision: vision, view, scene, sight, illusion (M) Society: custom, habit, practice, law, style, mode, fashion, trend, manners, duty, character, emblem, glory, fame, environment (N) Human and other qualities —deriving human propensity adjective: • feel x: emotion, anger, sympathy, joy, fear, regret, fury, envy, shame, passion, affection, worry, lethargy, melancholy • show x: greed, caution, courage, respect, skill, mercy, care, prejudice, temperament, conscience, scorn, irony, ambition, honour, humour, nostalgia • have x: mood, wit, guilt, grace, virtue, wish, influence, advantage

8.2 derived from common nouns and adjectives   227 • considered to be x: delight, mystery, hazard, danger, trouble, advantage, fright, inspiration, pleasure • quality of a person: fool, idiot, prude, snob, gossip, rascal, rogue, nuisance, traitor, villain, whore, cheat, coward, miser, parasite, freak, lout, dolt, nuisance, brute, hero, monster —deriving dimension adjective: size, dimension —deriving age adjective: age, youth —deriving colour adjective: colour, gold, silver —deriving value adjective: beauty, worth, value, fault —deriving physical property adjective: flavour, smell, taste, comfort —deriving corporeal adjective: hunger, thirst, health —deriving speed adjective: speed, haste (P) Measure, money, part, type and group —measure: number, excess, addition —money: money, wealth, price, cost, expense, profit, debt, tax, rent —part: part, piece, grain, drop, page, essence, episode, period —type: type, sort, kind, example —arrangements: bunch, crop, heap, harvest, pair, pile, stock (Q) Orientation, space and shape —orientation: front, back, top, bottom, end, edge —space and shape: space, shape, circle, square, slope, fork, hook, parabola —compass points: north, south, east, west, north-east, etc. (R) Time: time, day, night, year, moment, instant, pause, occasion, early, late Note that nouns in A–J have concrete reference (according a widish denotation to this term), and those in K–R have an abstract meaning. Adjective roots in English may be grouped into a number of semantic types; these are listed and illustrated in the Appendix. We need here just to pick out the Colour type, which shows special properties: (S) Colour: black, white, red, yellow, green, blue, pink (T) Other underived adjectives: big, bald, sick, dead, alone, old, good, fast, wrong, right

Table 8.2  Occurrences of some adjective-deriving suffixes with semantic types of nouns (a non-exhaustive listing) (1) -like

(2) -ly

(3) -ish

(5) -y

(7) -ed

(8) -ous

(9) -ic

(10) -al/-ical

(11) -ful

(13) -free

(14) -less

(15) -able

A

nose-like leg-like bone-like fever-like

—

liver-ish fever-ish

tooth-y hair-y brain-y pimpl-y blood-y furr-y gout-y

beard-ed feather-ed [plus many compounds such as bow-legged]

venom-ous

anaem-ic bronchit-ic diabet-ic asthma-tic

adenoid-al faec-al skelet-al hormon-al

tear-ful pain-ful

fat-free fever-free disease-free

chin-less hair-less leg-less brain-less mind-less breath-less pain-less

—

B

man-like lady-like gentleman-like child-like king-like god-like workman-like statesman-like

man-ly gentleman-ly king-ly ghost-ly god-ly master-ly scholar-ly friend-ly neighbour-ly

mann-ish child-ish boy-ish heathen-ish devil-ish fiend-ish elf-ish amateur-ish

—

—

monstr-ous

poet-ic artist-ic aristocrat-ic despot-ic nomad-ic satan-ic

president-ial secretar-ial editor-ial department-al trib-al nation-al

master-ful

man-free child-free

god-less friend-less companion-less

—

C

mother-like aunt-like uncle-like brother-like

mother-ly father-ly sister-ly brother-ly

—

—

—

—

—

—

—

—

father-less mother-less child-less wife-less

—

D

monkey-like sheep-like wolf-like kitten-like mule-like hawk-like spider-like

—

sheep-ish wolf-ish kitten-ish hogg-ish mul-ish owl-ish wasp-ish

horse-y mouse-y mosquito-y louse-y

—

vermin-ous

—

—

—

ant-free louse-free

horse-less fish-less

—

Table 8.2 (continued) (1) -like

(2) -ly

(3) -ish

(5) -y

(7) -ed

(8) -ous

(9) -ic

(10) -al/-ical

(11) -ful

(13) -free

(14) -less

(15) -able

E

tree-like flower-like leaf-like fruit-like

—

—

wood-y flower-y leaf-y fruit-y

wood-ed leaf-ed

—

—

—

—

grass-free

flower-less grass-less seed-less

grass-able

F

foam-like milk-like poison-like moon-like snow-like spring-like

crumb-ly earth-ly world-ly home-ly heaven-ly

silk-ish hell-ish summer-ish winter-ish

silk-y water-y milk-y hill-y sunn-y wind-y thunder-y summer-y

ice-d pebble-d

vapor-ous poison-ous mountain-ous cavern-ous precipit-ous thundr-ous

metall-ic acid-ic calori-fic

natur-al glob-al univers-al season-al

—

smoke-free ice-free sugar-free

sugar-less flour-less moon-less wind-less

season-able

G

prison-like window-like

—

—

room-y

partition-ed

—

—

—

—

shop-free

—

partition-able

H

chair-like cushion-like car-like

—

book-ish

cushion-y pillow-y

saddle-d string-ed

—

—

—

—

shackle-free gun-free

hat-less shirt-less

—

J

noun-like story-like

—

—

noise-y whisper-y word-y joke-y

—

—

syllab-ic rhythm-ic melod-ic apologe-tic epigram-atic

dialect-al diphthong-al conversation-al

—

—

noise-less sound-less name-less

question-able answer-able

K

—

—

—

art-y sport-y

—

superstiti-ous religi-ous

scient-ific histor-ic geograph-ic photograph-ic

chem-ic-al histor-ic-al fiction-al logic-al

—

dance-free religion-free

—

—

Table 8.2 (continued) (1) -like

(2) -ly

(3) -ish

(5) -y

(7) -ed

(8) -ous

(9) -ic

(10) -al/-ical

(11) -ful

(13) -free

(14) -less

(15) -able

L

business-like party-like war-like dream-like life-like death-like

death-ly

nightmar-ish

trick-y luck-y chance-y risk-y

spirit-ed

industri-ous murder-ous adventur-ous victori-ous disastr-ous riot-ous

problem-atic chao-tic strateg-ic energe-tic

industri-al experiment-al suicid-al funer-eal origin-al fat-al idea-l fact-ual nonsens-ic-al function-al

peace-ful purpose-ful thought-ful doubt-ful truth-ful force-ful power-ful

quarrel-free murder-free nightmarefree fancy-free

dream-less motion-less rest-less life-less spirit-less luck-less function-less thought-less doubt-less effort-less power-less

action-able marriage-able reason-able knowledge-able memor-able

M

—

manner-ly

styl-ish mod-ish

trend-y

—

glori-ous fam-ous

character-istic emblem-atic

habit-ual practic-al environ-ment-al

law-ful duti-ful

—

law-less

fashion-able

N

—

coward-ly miser-ly

fool-ish prud-ish snobb-ish rogu-ish whor-ish freak-ish lout-ish dolt-ish

angr-y greed-y mood-y guilt-y gossip-y rascal-y silver-y fault-y smell-y hungr-y speed-y

age-d

envi-ous cauti-ous courage-ous ambiti-ous grac-ious humour-ous virtu-ous advantage-ous mysteri-ous danger-ous traitor-ous

sympathe-tic melanchol-ic honor-if-ic nostalg-ic idiot-ic hero-ic

emotion-al prejudic-al temperament-al influent-ial inspiration-al X-dimension-al

joy-ful fear-ful regret-ful respect-ful skil-ful merci-ful scorn-ful grace-ful wish-ful fright-ful youth-ful colour-ful beauti-ful

trouble-free idiot-free gossip-free fault-free

joy-less fear-less shame-less merci-less humour-less grace-less age-less colour-less taste-less comfort-less

honour-able pleasur-able size-able valu-able comfort-able

Table 8.2 (continued) (1) -like

(2) -ly

(3) -ish

(5) -y

(7) -ed

(8) -ous

(9) -ic

(10) -al/-ical

(11) -ful

(13) -free

(14) -less

(15) -able

P

grain-like page-like

cost-ly

—

wealth-y price-y

money-ed

—

episod-ic period-ic

addition-al

—

debt-free tax-free rent-free

price-less

profit-able tax-able

Q

circle-like hook-like

shape-ly north-er-ly

—

—

hook-ed fork-ed

spac-ious

parabol-ic

—

—

—

end-less shape-less

—

R

—

time-ly (dai-ly) (night-ly)

early-ish lat-ish

—

—

—

—

paus-al occasion-al

—

—

—

—

S

—

—

black-ish whit-ish yellow-ish green-ish

yellow-y green-y pink-y

—

—

—

—

—

—

—

—

T

—

sick-ly dead-ly good-ly

bigg-ish bald-ish old-ish

—

—

—

—

—

wrong-ful right-ful

—

—

—

232    8 making new adjectives Table 8.2 illustrates, for twelve of the most common adjective-deriving suffixes, which semantic types of common nouns and adjectives each occurs with. Note that the information in this table is far from exhaustive; many more words could be included in a good number of the cells, a few more examples in others. We now examine, one by one, affixes (1)–(23) from Table 8.1, as they derive adjectives from common nouns (and some also from adjectives). Note that two suffixes whose major use is with verbs may also be added to a few verbs—see the discussion of (29) -(at)ive in 8.4.6 and of (32) -worthy in 8.4.19. Suffix (27) -(a)n is mostly used on proper names, but is also found added to a few common nouns (for example, republic-an, suburb-an)—see 8.3.8.

8.2.1 (1) -like 8.2.1a The Germanic suffix -like /-ˌlaik/ does not affect stress on the root to which it is attached, and itself generally takes secondary stress. It developed from the free form like ‘similar to’ which itself evolved in ME times from OE lı¯c ‘body, form, appearance’. Originally the second element of a compound, -like developed affixal status in the early days of Modern English. From the mid-fifteenth century we have chieftain-like and devil-like; since then it has steadily gathered momentum as a suffix. It has a straightforward meaning, very close to that of adjective like, ‘similar to, characteristic of’, and carries no overtones that this is either a desirable or an undesirable quality. The suffix -like is highly productive, and may be added to almost every concrete noun in types A–J. Indeed, a criterion for a form being a concrete noun could be that it co-occurs with -like. It may also be added to a phrase, as in He behaves in a distinctly [soon to be boss]like manner. Just a sprinkling of abstract nouns may also take -like, including activity nouns such as business, party, war, funeral, and dream (type L), and nouns referring to parts such as grain and page (P), or to shapes such as circle and hook (Q). One can also refer to a life-like resemblance (of a statue or a painting) or to a death-like silence. This suffix is not comfortable with most abstract nouns; for instance, one would scarcely say *delight-like or *truth-like.

8.2.2 (2) - ly   233 8.2.1b A homonymous suffix -like may in colloquial usage be added to an adjective with the meaning ‘sort of’; for instance She’s old-like|  beautiful-like ‘She’s sort-of old/beautiful’. Marchand (1969: 356) mentions that it may also be added to a phrase, such as [in a hurry]-like and [out of her mind]-like.

8.2.2 (2) -ly 8.2.2a Germanic suffix -ly /-li/ goes back to the same OE root as -like—OE noun lı¯c ‘body, form, appearance’—but developed by a different route. In OE we find woruld-lic ‘world-ly’, and -ly consolidated its function as a derivational suffix within ME, with woman-ly, man-ly, shape-ly, state-ly, among others. During the early centuries of Modern English, it was also added to Romance loans, as in prince-ly, manner-ly. This suffix appears not to be applied to nouns ending in /l/ or in /i/. One can say gentleman-ly but not, nowadays, *lady-ly; aunt-ly is attested (­Jespersen 1942: 407) but *uncle-ly would be unacceptable. (Daily appears to be an exception; it is a development from OE dæglich, founded on dæg ‘day’.) 8.2.2b The adjective-deriving suffix -ly must be distinguished from the grammatically quite different homonym, -ly, which derives adverbs from adjectives (as in clever-ly, slow-ly); this is discussed in 10.2–4. A difference is that the adverb-deriving suffix may follow /l/ or /i/, as in easi-ly and loyal-ly. The two suffixes cannot co-occur; that is, an adjective formed with -ly may not add a further -ly to create an adverb. For example, one cannot derive adverbs *neighbour-ly-ly or *miser-ly-ly, but must instead say something like in a neighbour-ly| miser-ly manner (see Chapter 10). 8.2.2c The adjective-deriving suffix -ly has limited productivity today. It occurs with many nouns referring to varieties of people such as man-ly and mother-ly (types B and C) and it may be applicable with new nouns relating to these types. It generally indicates a desirable quality associated with that class of person—father-ly advice is likely to be the sort of sensible guidance a good father should provide, and sister-ly care would be inferred to be kind and loving. The suffix also has a positive sense in shape-ly ‘of a pleasing shape’ and time-ly ‘at an appropriate time’. Heaven-ly refers to ‘a delightful quality, such as one would expect to encounter in heaven’.

234    8 making new adjectives We do find -ly added to nouns which have a negative meaning, and here it just indicates ‘in a manner one would expect of such a person’—coward-ly behaviour ‘behaving as one would expect a coward to’; note also cost-ly mistake ‘mistake that carries a high cost’. There is a set of modifying forms ending in -ly based on time words—such as dai-ly and night-ly—but these are probably best regarded as adverbs which have secondary function as adjectives. See 10.2b, 10.4a. Just a few adjectives may form a further adjective through the addition of -ly. Someone who is often alone (or who has a lone existence), and does not like this, is described as lone-ly; someone who is often sick and affected by this may be termed sick-ly. The adjective good-ly only relates to quantity, but still retains the positive sense of good—a good-ly amount is a quantity which is considered good in the circumstances. A dead-ly blow with a dead-ly weapon (or a dead-ly disease) will render the recipient dead. In contrast death-ly, from noun death, indicates something reminiscent of a dead person, as in She has a death-ly pallor. Just occasionally, -ly may be added to an NP; Jespersen (1942: 407) quotes in a [   fellow-creature]-ly way from G. B. Shaw’s 1911 play The Doctor’s Dilemma. 8.2.3 Contrasting -like and -ly Adjectival suffixes -like and -ly have developed by different historical paths. There is a little overlap in the types of nouns they can occur with and a number of minimal pairs which illustrate their differences in meaning. For example (and see Jespersen 1942: 406–7):

•

brother-like behaviour could refer to any of the ways in which one brother behaves towards another, either good or bad—taunts and blows, or encouragement and protection. brother-ly behaviour will refer to one brother looking after another, in an entirely positive way.

• • •

gentleman-like appearance refers to looking like a gentleman (which may or may not be regarded as a good thing by the speaker). gentleman-ly manner refers to behaviour traditionally considered courteous; for instance, opening the door for a lady.

Note that, for phonological reasons, people do not nowadays say *lady-ly. Thus, lady-like does double duty, corresponding to both gentleman-like and gentleman-ly.

8.2.4 (3) -ish   235

•

god-like is used with respect to some outward quality, such as appearance and demeanour. It could carry either positive or negative overtones (for example, The new manager, with his god-like manner, is just impossible to talk to). god-ly is most likely to refer to inherent qualities, these being held to be a good thing; for example, A god-ly life is characterized by religious devotion.

•

Sometimes there is a measure of substitutability; both death-like and death-ly may modify silence, hush, appearance, and pallor. One could specify that death-like states ‘as if there were death’ while death-ly implies ‘with some of the features of death’, but these characterizations blur into one another. 8.2.4 (3) -ish 8.2.4a The Germanic suffix -ish /-iʃ/ is a descendant of OE suffix -isc. This was used predominantly for deriving adjectives from proper names (such as Den-isc ‘Danish’), discussed in 8.3.2. However, -isc was added to some common nouns, for instance hæ´ðen-isc ‘heathen-ish’, cild-isc ‘child-ish’ (see 2.4b). This usage expanded from ME times on. The suffix has high productivity. But, although -ish can potentially be added to a root of any length, the majority of common instances are with monosyllabic roots. Indeed, if added to a disyllabic noun, the derivation may omit an unstressed vowel in order to stay disyllabic. For example liver, /'livə/ plus -ish, /-iʃ/ gives liver-ish, which may be pronounced as either /'livəriʃ/ or /'livriʃ/. Similarly for fever, summer, winter, devil. 8.2.4b The suffix often indicates ‘showing a negative quality associated with’ (Sweet 1891: 465 characterizes it as generally expressing contempt). For example, from names of animals are derived adjectives which are basically derogatory—wasp-ish sense of humour ‘sharp and biting’, wolf-ish appetite ‘eating quickly, perhaps tearing food apart’, sheep-ish demeanour ‘meek and docile’. Similarly with swine, hog, goat, mule, and others. The suffix is typically used with nouns which themselves have a derogatory or negative reference, then describing the concomitant quality— fool-ish, rogu-ish, whor-ish, brut-ish, heathen-ish, snobb-ish, fever-ish, nightmar-ish.

236    8 making new adjectives We find book-ish used of someone who spends a great deal of time in study, but when this derivation is used it does tend to carry the implication that some of the time might be more appropriately spent on other things. This suffix is also used with preposition/adverb up; the derived adjective, upp-ish, is used of someone who is stuck up and arrogant. There are, however, some uses which carry a quite positive sense, as stylish ‘in the current style’. Alongside manner-ly ‘showing good manners’ we might expect the positive adjective derived from style to involve suffix -ly. However, as noted in 8.2.2, adjective-deriving suffix -ly cannot be added to a form ending in /l/. This may, at least partly, explain why -ish is here used to create a word with no derogatory overtones. Colour adjectives have been used with -ish since the end of the fourteenth century, this use being extended to many other adjectives over the next couple of hundred years. Something is redd-ish if it shows a trace of red and bigg-ish if it is just a little bigger than the norm. We find old-ish ‘not really old but advancing towards that state’, and hungry-ish ‘getting to feel a bit hungry’. Similarly, with nouns summer and winter (but not with spring and autumn or fall); if the weather is summer-ish it is what might be expected in summer, but occurs at some other time of the year. With numbers and some time expressions the sense of the suffix is ‘approximately’—She’s forty-ish, Come at five-ish, He arrived late-ish. And it can be added to a few verbs, as in tickl-ish; see 8.4.17. Suffix -ish can be added to a phrase—I’ll be back at [  four o’clock]-ish. The derogatory sense may still prevail—He’s very [ public school ]-ish.

8.2.5 Contrasting -like and -ish If you wish to compare someone with an important person whom you do not esteem, then -ish is the appropriate suffix to use, as in He has a distinctly [Vladimir Putin]-ish air. But if the comparison is with a person who is admired, then -like is the suffix to use: He shows a rather [Franklin D. Roosevelt]-like attitude. Since nightmare has a negative meaning one naturally derives an adjective with -ish, as in a nightmar-ish experience. Dream does not hold bad connotations and the corresponding adjective based on it is dream-like; if one hears Our vacation had a dream-like quality this means that it was scarcely like real life (the overtone is undoubtedly positive).

8.2.7 (4) - esque   237 8.2.6 Contrasting -ly and -ish These two suffixes are in a sense complementary, -ly often indicating a positive quality associated with the noun to which it is attached and -ish a negative one. Thus, for instance, god-ly and devil-ish. Also heaven-ly and hell-ish (but note that -ly does not attach to words ending in /l/). These suffixes may be added to the same root, nicely exemplifying their difference in meaning:

• •

man-ly behaviour (used of a man) ‘behaving as a man is expected to act; for example, strong and courageous’. mann-ish behaviour (used of a woman) ‘behaving in a way considered appropriate for a man but not for a woman, perhaps in terms of dress or stance or activity’.

We also have man-like, which could be used to refer to an alien who looks somewhat like a man without actually being one. In days gone by, the noun man had two senses: (a) human being, and (b) male human being. Interestingly, man-like could relate to either sense, but both man-ly and mann-ish only to (b). It is instructive to compare the common use of -ly and -ish with certain types of noun: -ly man-ly — king-ly

-ish mann-ish child-ish —

also with: woman, mother, etc. also with: boy, girl, baby, etc. also with: chief, etc.

Child-ish is typically used of an adult behaving in a way which is considered inappropriate for them and is reminiscent of how a child acts. There are no positive characteristics associated with children which could give rise to *child-ly. In contrast, any behaviour expected of a king is likely to be considered good, hence king-ly but not *king-ish. 8.2.7 (4) -esque The Romance-origin suffix -esque /-esk/ always takes secondary stress. French loans, beginning from the sixteenth century, included grotesque, which cannot be analysed within English, and pictur-esque, which can be. Instances with common nouns are generally full loans from French. The suffix has become mildly productive in English when added to the proper

238    8 making new adjectives names of people. It may also be added to adjectives derived by -(a)n; for example, Africa-n-esque. It indicates ‘having the style of, in a pleasing manner’. Statu-esque, used of a person, indicates that they hold themselves with grace and dignity, like a statue; to describe a scene as pictur-esque suggests that it is attractive, not unlike a high-quality painting. Added to names it states ‘in the distinct and highly respected style associated with’; thus, Pinter-esque and Garbo-esque. 8.2.8 (5) -y, -sy, -ity| -ety 8.2.8a Suffix -y /-i/ is the continuation of the high-frequency adjectiveforming suffix -ig in OE. There were blo¯ d-ig ‘blood-y’, grǣd-ig ‘greed-y’ and many others. More were added in ME and Modern English. It is much used with nouns, and also with verbs of some semantic types (this is discussed in 8.4.13). Interestingly, this suffix is not added to roots ending in /ai/ or /i/—we can say elbow-y but not *knee-y, wasp-y but not *bee-y, legg-y but not *thigh-y. However, there is one noun ending in /ei/ which takes -y, clay-ey. 8.2.8b The most common sense is ‘characterized by’, used with nouns from a wide variety of semantic types. The major impression one gets of a hair-y person is their hair, of a tooth-y grin the teeth (not the mouth). Similarly for fogg-y day, sand-y terrain, angr-y response, greed-y boy, and louse-y hair. The item in question can be a body or other part (brain-y girl, leaf-y foliage), an annoying insect (mosquito-y campsite), a natural feature (starr-y night, hill-y country), a noise or speech act (squeak-y voice, joke-y manner), or an abstract noun as in mood-y, greed-y, hungr-y, risk-y, fault-y, and trick-y. A wealth-y person is someone characterized by wealth. The feature of note for a price-y item is its high price. Another sense is ‘looks like a, feels like a, behaves like a’. A person may be called mouse-y if they are as insignificant as a small rodent; a person’s hair maybe be called mouse-y if it has the typical light brown colour of a mouse. A pillow-y carpet feels soft to walk on, almost like a pillow. A further sense is ‘engrossed in’, as in sport-y and art-y. If a person is described as horse-y this may indicate that they devote their life to training or breeding or riding horses. (Or, in a quite different sense of the suffix, they may look like a

8.2.8 (5) - y , - sy , - ity /- ety   239 horse.) We also find -y added to some colour adjectives, such as green-y ‘with a touch of green’. 8.2.8c The suffix has typically been added to monosyllabic or disyllabic nouns. In the latter instance, the resulting adjective generally also has just two syllables; thus hunger /'hʌŋgə/ yields hungr-y /'hʌŋgri/. A schwa, /ə/, can also be omitted from flower-y, water-y, thunder-y, summer-y. This suffix is getting more and more productive. It is being used with words of any length, often in jocular vein, as in He’s a windbagg-y liberal, and I’m feeling really Hamlett-y: to link or not to link. It can also be added to a phrase, such as She’s an [open air]-y kind of girl, and It’s a [   fish and chips]-y sort of cafe. Note that there are several homonymous suffixes /-i/. In 6.11.3 we had discussion of -y ~ -ey ~ -ie ~ -ee, which derives nouns and indicates a loving attitude, as in Jane-y and puss-y. 9.2.2 and 9.4.11 deal with the suffix -y which derives abstract nouns from adjectives and verbs; for instance, difficult-y, injur-y, and discover-y. (See also 9.3.13a.) 8.2.8d In 6.11.3, we noted that the ‘endearing’ suffix -y ~ -ey ~ -ie has a by-form -sy /-si/—a dogg-sy may be doted over more than a dog. In congruent fashion, adjective-deriving suffix -y also has an -sy variant. This may originally have been -y added to plural -s—as in trick-sy, from the sixteenth century—but is now a suffix in its own right, generally with a somewhat dismissive meaning. Adjective art-y may be used of a person or object which either has real artistic merit, or which purports to have. In contrast, art-sy just about always indicates an unfounded pretension to art. This can be exacerbated by employing rhyming couplets. For example, in increasing order of mockery: artsy-craftsy, artsy-tartsy, arts-fartsy. However, folk-sy just relates to the ways of simple folk, without affectation; it can be accorded either positive overtones (‘friendly’) or negative ones (‘unsophisticated’). 8.2.8e Suffix -ity or -ety, /-əti/, another variant of -y, is used only a little in today’s Standard English. Whereas adjective itch-y describes someone with an itchy patch on their skin, itch-ity can refer to feeling disconcerted by an obligation: She felt itch-ity at the prospect of meeting John’s parents for the

240    8 making new adjectives first time. Upp-ity, created in the late nineteenth century, has a similar meaning to the earlier upp-ish, ‘stuck up, arrogant’. In some circles there is a technique for swearing without actually using swear words—just put blank (as noun) or blank-ety (as adjective or adverb) in place of each of them, as in I’m going to blankety crush those blankety little blanks who stole my blankety car.

8.2.9 Contrasting -ish and -y There are some words which are found with both -y and -ish, showing a subtle but significant difference in meaning. With most colour adjectives, one can use either suffix; for example:

• •

yellow-ish ‘showing a tinge of yellow’ yellow-y ‘not quite yellow but similar to it’

Interestingly, black and white take -ish but scarcely -y. Something either is or is not black or white so that it makes little sense to say ‘not quite black (or white) but similar to it’ (one would instead say dark grey or light grey). Summer and winter may also take either suffix; for example:

• •

summer-ish ‘weather showing some characteristics of summer’ summer-y ‘weather reminiscent of summer, but occurring in another season or (perhaps within a cold “summer”)’

One hears summer-y dress ‘appropriate for wearing in summer’, but not *summer-ish dress. And summer-y day ‘day with weather typical of summer’, but not *summer-ish day. Another noun forming adjectives with both suffixes is dog. One speaks of a dogg-y smell but of dogg-ish affection. There is, however, overlap of use. (Malkiel 1977 provides a useful comparison of -y and -ish.)

8.2.10 (6) -en 8.2.10a The Germanic suffix -en ‘made from’ is realized as /-ən/ after p and k (consonants which cannot be followed by a syllabic nasal) and is typically

8.2.10 (6) - en   241 reduced to a syllabic nasal /-n̩/ elsewhere; compare silk-en /'silk-ən/ with wood-en /'wud-n̩/. OE had a commonly-occurring suffix -en, which derived adjectives from nouns (also engendering umlaut of the root vowel); for instance gyld-en ‘golden’, wyll-en ‘woollen’, leðr-en ‘made of leather’ and stǣn-en ‘made of stone’, from gold, wull, sta¯n, and leðer. A fair proportion of these fell out of use (those that remained dropping the mutation), but many more adjectival derivatives were created in ME and early Modern English times. 8.2.10b In 7.3.2, we described another -en suffix, deriving verbs from adjectives (for example, deep-en, wid-en). This reflects OE suffix -an ~-ian ~ -nian and is etymologically distinct from the adjective-deriving suffix -en. As was illustrated in 7.3.2, there is a phonological restriction on the verb-deriving suffix -en. It may only be added to a monosyllabic root ending in a voiceless fricative (/f/, /θ/, /s/, /ʃ/), a voiceless stop (/p/, /t/, /k/), or /d/. A fascinating observation is that, in early Modern English times, the adjective-deriving suffix -en developed almost the same phonological constraints as the verb-deriving suffix -en. It was used on monosyllabic roots, after voiceless stops—hemp-en, oat-en, wheat-en, silk-en, milk-en, oak-en—after /d/— gold-en, lead-en, wood-en—and after voiceless fricatives—earth-en, wax-en, flax-en, ash-en, larch-en. The one difference from verb-deriving -en was that the adjective-deriving suffix was also used after /l/ and /r/—wooll-en, hair-en, leather-n. (Full information is in Jespersen 1942: 346-9). It is not used after vowels, nasals, /b/, or /g/. We can see that whereas in OE times an adjective could be made from sta¯n ‘stone’ by adding -en, this was no longer possible. 8.2.10c Adjective-deriving suffix -en is no longer productive and the use of its derivations has been steadily dwindling. Some of those just listed are no longer in use (for example, milk-en, larch-en, hair-en) and others have limited employment (leather-n). Turning now to semantics, the suffix is added to nouns referring to materials, with the primary meaning ‘made from’, as in wood-en, wax-en, earth-en, gold-en. However, this can be extended to ‘has the appearance of being made of’—a gold-en jug may glitter like gold, a wood-en expression on someone’s face may appear to lack life, as if carved from wood. Lead-en may describe an object which is so heavy it might be made of lead, or it can be used of a dull and heavy sky, or of a poorly executed translation.

242    8 making new adjectives 8.2.11 Contrasting -y and -en Why, in present-day English, do we have to use silver-y as the correspondent of gold-en, rather than *silvern? In OE and early Modern English times there was an exception to the principle that -en should only be added to monosyllables. A few disyllabic roots ending in -er formed adjectives by simply adding -n, creating the ending -ern /-ərn/ (reduced to /-ən/ in those dialects which have omitted /r/ at the end of a syllable). OE did have silfr-en ‘made of silver’ and this came into early Modern English as silvern. But, with the general decline in use of adjectival -en derivations, silvern dropped out of use during the nineteenth century. (The only -ern form to survive today—and that only just—appears to be leathern.) So gold-en continued but silvern fell by the wayside, replaced by a derivation using the increasingly productive -y: silver-y. The newer gold-y is also possible for ‘looks like gold’ but gold-en is usually preferred, even though this is ambiguous between the ‘made of ’ and ‘looks like’ senses (these would generally be distinguished by the context). There are a few nouns which may easily accept both suffixes; in each case, the -en derivation is much older. For example:

•

wooll-y ‘includes a component of wool, or covered with wool (woolly lambs); or looks or feels like wool’; there is also wooll-y thinking ‘muddled thinking, reminiscent of tousled wool’ wooll-en ‘made of wool (or a mixture of wool and artificial fibres)’

•

Garments made of wool are often called just woollens (woollen is here doing double duty as a noun). Note that especially warm ones may be described as winter woollies; this involves a different -y suffix, that of endearment; see 6.11.3.

• •

silk-y ‘made of silk, or feels like silk’ silk-en ‘made of silk, or looks like silk’

The overlap is here perhaps greater. We find, for instance, both silk-y hair and silk-en hair, both silk-y voice and silk-en voice. 8.2.12 (7) -ed 8.2.12a OE had a derivational suffix -ed(e), added to nouns and deriving adjectives; for example hring ‘ring, link’ and hring-ed ‘made up of

8.2.12 (7) - ed   243 rings or links’. The suffix has been productive at all historical stages of English. It basically means ‘provided with something which is not an inherent part’. Not every man has a beard and one who has is said to be beard-ed, a horse may be saddle-d, an instrument string-ed, a drink ice-d, a person money-ed, a tongue fork-ed, a conversation or a person spirit-ed, a room partition-ed, a pathway pebble-d. It is normal to have a straight back, two eyes, five fingers, and a head covered with hair; deviations from the norm may be described by a compound adjective with -ed on the second, body part, component—hunch-back-ed, one-eye-d, two-finger-ed, and bald-head-ed. Beginning in OE times, there have been an enormous number of such double-barrelled derived adjectives, including red-head-ed, pale-face-d, cruel-heart-ed, half-witt-ed, short-winded, good-nature-d, three-corner-ed. 8.2.12b The -ed suffix which is added to nouns should be clearly distinguished from the participial ending -ed ‘in a state of’ on verbs—as in delight-ed, interest-ed (Marchand 1969: 264 suggests that the two suffixes probably had a common origin in the distant past, going back to proto-Indo-European -to.) There is similarity of meaning and also of phonological form. The verb suffix is: /-id/ after /t/ or /d/, otherwise /-t/ after a voiceless and /-d/ after a voiced sound. The noun suffix differs only in that /-id/ can be used after a few nouns ending in a voiced stop or affricate. Legg-ed is most often pronounced /'leg-id/—rather than /legd/—in a form such as three-legged stool. Aged can be pronounced /'eidʒid/, although in a compound such as well-aged /'eidʒd/ would be used. 8.2.12c In some adjectival derivations, a root-final fricative may become voiced before -ed. For example leaf plus -ed can be written as leaf-ed or as leaved and pronounced either as /li:ft/ or as /li:vd/; similarly for hoof, mouth, tooth. 8.2.12d The suffix can be used with any type of noun which has an appropriate meaning. And it may be added to an appropriate coordination, as in [[quick and sharp]-sight]-ed or [[  pale and ashen]-face]-ed. In addition, two adjectives derived with -ed may share a modifier, as in feeble-mind-ed and bodi-ed (Jespersen 1942: 429). The Germanic prefix be- (see 7.3.1) can be added to adjectives derived through suffix -ed, and may then insinuate a derogatory tone. If someone

244    8 making new adjectives says She is be-jewell-ed, the implication is that the adornment is perhaps a little excessive. Hearing The barrister came into the room be-wigg-ed, one gets the impression that the speaker considers the barrister to look mildly ridiculous in this get-up. Noun wretch goes back to OE. In ME times, -ed was added, deriving adjective wretched, a somewhat unusual example of -ed semantics.

8.2.13 Contrasting -y, -en and -ed The noun wood has a range of meanings—it can refer (a) to the material in the trunk or branch of a tree, or (b) to pieces of a tree used to make a fire or an artefact, or (c) to a collection of growing trees. Wood is unusual in that it can form adjectives with -y, -en, or -ed. Interestingly, each of these derivations employs a different sense of the noun. Prototypical meanings are:

• • •

wood-y: sense (a), ‘having the texture or scent of wood’, as in wood-y vegetable or wood-y fragrance wood-en: sense (b), ‘made of wood’, as in wood-en bookshelves wood-ed: sense (c), ‘abounding in trees’, as in wood-ed valley

8.2.14 (8) -ous, -ious, -eous 8.2.14a Suffix -ous /-əs/ came into ME through Romance loans and reflects Latin derivational suffix -o¯ sus. Latin examples include co¯ pia ‘plenty, abundance’, co¯ pi-o¯ sus ‘copious’, and glo¯ ria ‘glory’, glo¯ ri-o¯ sus glorious’. Borrowings from French such as jealous, serious, and hideous were not analysable in English, but we had pairs such as courage and courage-ous, adventure and adventur-ous which established -ous as a derivational suffix in English. Further adjectives were created, within English, from Romance loans; for example poison-ous and advantage-ous. And -ous was added to a smallish number of Germanic roots, such as murder-ous, thunder-ous, and wondr-ous. Some -ous adjectives were borrowed from French together with a corresponding abstract noun ending in -ion. For example, ME accepted religion and religious, ambition and ambitious, superstition and superstitious. Nouns caution and infection were also borrowed. Adjectives relating to these— cautious and infectious—were then created within English, by analogy with pairs such as religion and religious.

8.2.14 (8) - ous, -ious, -eous   245 One fascinating adjective is right-eous, based on adjective right. This goes back to root riht ‘right’ plus suffix -wis ‘-wise’ in OE. The entirely Germanic combination was gradually transformed, along the lines: right-wis > right-wos > right-eous, then being re-analysed as if it were formed with the Romance suffix -(e)ous (and on an adjective rather than on a noun). Another etymology of note is that of wondr-ous. In ME the genitive form of Germanic noun wonder was wonder-s ‘of wonder’, and this was re-­interpreted as wondr-ous, by analogy with such Romance loans as marvell-ous. 8.2.14b The suffix has varying phonological effects (already outlined in 2.5e). It is simply added to some roots: venom /'venəm/ plus -ous /-əs/ gives venom-ous /'venəm-əs/. Stress moves to the syllable preceding -ous (with concomitant changes in vowel value) in the derivation of courage-ous /kə'reidʒəs/ from courage /'kʌridʒ/; similarly for mysteri-ous, advantage-ous, and industri-ous. An unstressed vowel may be omitted, as in the derivation of monstr-ous /'mɔnstrəs/ from monster /'mɔnstə(r)/; similarly in wondr-ous. There is optional reduction in thunder-ous, adulter-ous, murder-ous, dangerous. The final /i/ drops from adultery in forming adulter-ous. And—as with -ed—a final /f/ may be voiced before -ous, as in mischiev-ous from mischief; this is undoubtedly a relic of the time when [f] and [v] were variants of one phoneme, [f] occurring finally and [v] medially. 8.2.14c After roots ending in -y /-i/, we get either -i-ous (as in mysteri-ous, envi-ous) or else -e-ous (as in pite-ous, plente-ous), but both endings are pronounced /-i-əs/. A number of derivations end in orthographic -ious, retaining the i from Latin, but in fact ci represents /ʃ/, as in grace /greis/, graci-ous /'greiʃ-əs/, and space /speis/, spaci-ous /'speiʃ-əs/. However, some roots ending in a consonant take -ious or -eous (rather than just -ous), suggesting that these are alternative forms of the suffix; for instance uproar-ious, labor-ious, gas-eous. 8.2.14d The predominant use of -ous (with meaning ‘characterized by’) is with abstract nouns of Romance origin, from types K–N; they include ­victori-ous, glori-ous, fam-ous, and envi-ous, plus adjectives such as a­ mbiti-ous and cauti-ous from the -i-on/-i-ous pairs just mentioned (more are given in Table 8.2). One common employment of this suffix is with a noun referring to something bad, then deriving an adjective describing the unwelcome quality

246    8 making new adjectives associated with it. For example, ruin-ous gambling ‘likely to lead to ruin’, monstr-ous behaviour ‘as a monster would act’, plus danger-ous, hazard-ous, villain-ous, adulter-ous, venom-ous, poison-ous. (Covet-ous is a rare example of an adjective in -ous being related to a verb; both words were borrowed into ME from Old French. Whereas covet indicates a plain desire for something, covet-ous implies an undesirable obsession.) However, there are examples of -ous being added to a noun with neutral or positive connotations, then producing an adjective with an agreeable sense, as with graci-ous and spaci-ous (plus such scarcely analysable forms as tenacious, vivacious, audacious). This suffix is unlikely to be added to new roots in everyday language, and appears to be not really productive there. It is, however, much used in scientific nomenclature.

8.2.15 Contrasting -y and -ous Germanic suffix -y and Romance -ous have similar meanings ‘characterized by’. Why do some nouns take -y and other -ous? For example: noun anger mood gossip smell

adjective angr-y mood-y gossip-y smell-y

noun fury grace humour danger

adjective furi-ous graci-ous humour-ous danger-ous

The main factor is genetic—the four nouns in this array which take -y are Germanic and those which take -ous are of Romance origin. There is also the phonological factor: -y cannot be used with a noun ending in /i/, such as fury and envy. Looking back at Table 8.2, -y is used with many concrete nouns (brain-y, louse-y, fruit-y) because these are mostly Germanic forms. Suffix -ous is mainly used with Romance loans, which include many abstract meanings. We can compare the entries for semantic types L, M, and N in the -y and -ous columns of Table 8.2:

•

All the -ous forms are Romance, except for murder-ous where -ous is added to the Germanic root murder. This was probably inspired by adjective mordeux ‘murder-ous’ in French.

8.2.16 (9) - ic , - (a)tic   247

•

The majority of nouns taking -y are inherited from OE. However, the Germanic suffix -y was extended to apply to Romance nouns trick, chance, risk, and fault; the -y adjective was created within English anything between one and three hundred years after the corresponding noun was borrowed from French. The same applies to rascal-y, which may have been influenced by rascaille in French.

Hill is Germanic and mountain is Romance. Hence the derived adjectives are hill-y and mountain-ous. But notice that if a tract of country is characterized as mountain-ous, the implication is that this is not a good thing (in keeping with the ‘unwelcome quality’ sense of -ous, mentioned in 8.2.14d). Compare with hill-y which can carry an appreciative overtone. The Germanic root thunder is perhaps unique in taking both suffixes.

•

First, there is adjective thunder-y—parallel to wind-y, sunn-y, rain-y, and the like—with the meaning ‘characterized by thunder’. For example, Thundery weather is forecast for next week. We also find thunder-ous, meaning ‘like thunder’. The negative sense of -ous may apply here; a thunderous explosion is likely to be frightening. But this derivation need not carry any negative nuance; every performer welcomes thunder-ous applause, applause which sounds as loud as thunder. (One would not hear *Thunder-y applause.)

•

8.2.16 (9) -ic, -(a)tic 8.2.16a Suffix -ic /-ik/ has its origins in Greek -ikos. For example, Greek authent-icos ‘of first-hand authority, original’ gave rise to Late Latin ­authent-icus, Old French autent-ique, and Modern English authentic. ME accepted this and a number of similar adjectival loans from Old French, such as lunatic and rustic (which cannot be analysed in English) and ­fantas-tic (alongside fantasy). These ended in -ic and gave rise to the introduction of -ic as a suffix deriving adjectives from nouns of Romance origin, such as poet-ic, histor-ic, hero-ic. Frequently, noun and adjective both reflect Romance/Greek forms, so that -ic is not simply added to a noun root, as in amnesia| amnes-ic, arthritis| arthrit-ic, democracy| democrat-ic. The suffix has alternative form -(a)tic, mostly on nouns reflecting roots with a final -ma in Greek (coming into English through a Romance channel).

248    8 making new adjectives For instance, problem-atic, system-atic, diagram-atic, enigma-tic, trauma-tic, dogma-tic, and drama-tic. The Latin noun opera formed its adjective operatic by analogy with drama| drama-tic. Primary stress generally goes on the syllable immediately preceding -ic, with some consequential vowel changes, as in atom /'atəm/, atom-ic /ə'tɔm-ik/. The suffix derives adjectives from proper names (for example, Iceland-ic, Byron-ic, discussed in 8.3.3) and from a few verbs (such as horrific; see 8.4.2b). It occurs with quite a number of common nouns, including concrete nouns relating to illness, types of people, and materials (types A–B and F) and abstract nouns (J–N). It is hard to find many examples of this suffix being added to a noun which is not of Romance origin. We do get satan-ic, in the eighteenth century, based on Satan, a noun which was present in OE but comes, via Latin and Greek, from a Hebrew root meaning ‘adversary’. And, in the seventeenth century, adjective bard-ic was created from bard, a loan into ME from Gaelic. (And see Jespersen 1942: 393.) 8.2.16b The suffix -ic simply indicates a quality associated with the noun—a bronchit-ic person is a person with bronchitis, a scientif-ic problem is a problem in science, an aristocrat-ic attitude is the attitude characteristic of an aristocrat, an idiot-ic response is the sort of response one would expect from an idiot. Whether a noun takes -ic depends in large part on its genetic provenance. Alongside poet| poet-ic we find anaemia| ­anaem-ic, geography| geograph-ic, energy| energe-tic, analysis| analy-tic, epigram| epigram-­atic. In each case, the modern words reflect the original forms in Greek and/or Latin and/or French. Today, the suffix is scarcely productive in everyday usage. However, it is much employed in scientific names—see the example from the OED given at the end of the entry for (16) -ary in 8.2.39. An agentive noun can be derived from a noun or adjective by adding -ist, and then an adjective from this by appending -ic. For example: noun art, noun art-ist ‘someone who does art’, adjective art-ist-ic ‘in the manner of an artist’. Also capital-ist-ic, fatal-ist-ic, real-ist-ic, and so on. The interesting way in which verbalizing suffix -ize is added to an -ic forms is discussed in 7.4.2. In 11.2.1 there are examples of -ic added after verbalizer -(i)fy.

8.2.18 contrasting - ous and - ic   249 8.2.17 Contrasting -en and -ic I go out to buy a new bookcase and the salesperson enquires about the material. Would I prefer a wood-en bookcase or a metall-ic one? Why the different suffixes? Noun metal was a Romance loan into ME about 1230 and adjective ­metall-ic (directly reflecting Latin metallicus and Greek metallikús) followed a couple of hundred years later. Wudu ‘wood’ was a commonly occurring noun in OE but (unlike gold, stone, wool ) appears not to have formed an adjective with suffix -en. An adjective relating to wood was eventually created in the sixteenth century. Being a Germanic root, it did not employ -ic, but instead -en, the Germanic suffix available for use with monosyllabic Germanic roots, ending in one of a number of consonants, which did include /d/; see 8.2.10.

8.2.18 Contrasting -ous and -ic Suffixes -ous and -ic are both of Romance origin, used predominantly with Romance roots. Why do some Romance nouns form an adjective with -ous, and others with -ic? Why do we get poison-ous and acid-ic, rather than ­*poison-ic and *acid-ous? Why monstr-ous but despot-ic, riot-ous but ­chao-tic, humour-ous but sympathe-tic? Essentially, it is just the ways the dice rolled. A particular suffix began to be used with a certain noun, and that combination came into general use. Nothing else was needed. However, there may sometimes have been a semantic element involved. An -ic derivation is often used to refer to something specific—as in an ­hero-ic act of bravery, an idiot-ic idea, an honor-if-ic title—and an -ous adjective for reference to something more general—scandal-ous conduct, traitor-ous behaviour, a grac-ious demeanour. There are just a couple of nouns which may take both suffixes. Based on noun number, we get adjectives numer-ic (or numer-ic-al) which has a specific denotation—for example, Put a numeric value on it!—and numer-ous, which has the general meaning ‘many’. Corresponding to noun barbar-ity there are adjectives barbar-ous and barbar-ic. Many dictionaries treat these as synonyms and there is indeed a

250    8 making new adjectives good deal of overlap between them. Nevertheless, the specific/general distinction appears to play some role. One reads of a (specific) barbaric act but of (general) barbarous treatment of prisoners. Scientific naming has its own problems, and may pursue opportunistic solutions to them. In some compounds iron is divalent and in others it is trivalent. Based on the Latin noun ferrum ‘iron’, an adjective could be created using either -ous or -ic. In this instance, both have been employed (both from the 1860s). Ferr-ous indicates a valency of two, as in ferr-ous oxide, FeO, and ferr-ic a valency of three, as in ferr-ic oxide, Fe2O3.

8.2.19 (10) -al, -ar 8.2.19a There are two derivational suffixes -al, of Romance origin. One forms nouns from verbs (for example, recit-al, dismiss-al, surviv-al, arriv-al) and is discussed in 9.4.13. The suffix under consideration here creates adjectives from nouns and reflects Latin ending -alis. Loans into ME from French included mortal, royal, and special, which were not analysable in English (the noun species was not borrowed until 300 years later). There were also nouns nature and eternity, together with adjectives natur-al and etern-al. These established -al as an adjective-deriving suffix in English, and it was steadily applied to many further Romance nouns, including nation-al, occasion-al, influent-ial. This suffix has been added to very few Germanic roots, perhaps the most common being tid-al (first recorded in 1808). It might be thought that bridal involves Romance suffix -al being added to Germanic form bride. In fact, the etymology is more exotic. OE had a compound brid-ealo involving roots ‘bride’ and ‘ale’, with the meaning ‘wedding feast’. In the fifteenth century, this was re-interpreted as adjective brid-al, as if it involved derivational suffix -al, by analogy with the many Romance loans which did. The canonical form of this suffix is -al, realized as /-əl/ or often simply as syllabic /l/—trib-al can be /traibəl/ or /traibl̩/. With a word ending in syllabic /n/, such as nation /neiʃn̩/, there are a variety of possible pronunciations for the -al form, nation-al, involving syllabic value for /n/ or /l/ or both (see Jones 1956: 320)—/'naʃnəl/ or /'naʃn̩l/ or/'naʃənl̩/ or /'naʃn̩l̩/. There are further forms of the suffix: -ual /-yuəl/ and -ial or -eal /-iəl/, as in fact-ual, concept-ual, grad-ual, editor-ial, baron-ial, funer-eal. In these derivations we get orthographic t and d realized as /tʃ/ and /dʒ/—/kən'septʃuel/, /'gradʒuəl/. And in president-ial we get ti realized as /ʃ/ so that the suffix is just /-(ə)l/—/ˌprezi'denʃəl/ or /ˌprezi'denʃl̩/.

8.2.20 contrasting -ic and -ic-al   251 Like -ous and -ic, the suffix -al is firmly based on the original Latin forms of nouns and derived adjectives. There is sometimes a shift of primary stress to the syllable before the suffix and this may carry with it vowel changes, as in accident /'aksidənt/, accident-al /aksi'dentəl/. Other derivations retain stress and vowel qualities; for example, diphthong-al and univers-al. Some keep stress placement but change vowels, such as nature /'neitʃə/, natur-al /'natʃrəl/. (See Bauer 1983: 117–19 for a full account of stress factors.) If a word included two l’s, Latin dissimilated one to r, and this shows up in variant form -ar after some roots including an l; for example, pol-ar and line-ar. 8.2.19b The suffix may indicate ‘belonging to’—a fiction-al character is someone belonging to fiction, skelet-al fragments indicates bits from a skeleton. Or it can mean ‘relating to’—an environment-al disaster is a disaster relating to the environment, a dialect-al difference is a difference relating to dialects. Suffix -al is added to descriptions of kinds of people (type B from 8.2), as in minister-ial office and secretar-ial workload. And to nature, colony, tribe, universe, globe, and season—a univers-al generalization is held to apply everywhere (in the universe?), a trib-al council is the council of a tribe. It applies to nouns describing types of activity—experiment-al design. And to words referring to human qualities—emotion-al response, influent-ial friend. There are a few further types, illustrated in Table 8.2. The adjective-deriving suffix -al is probably no longer productive in everyday discourse. The word environment-al has come into greater use recently but, accordingly to the OED, it dates back to the 1880s (and may have been coined on the analogy of experiment-al, which is from the fifteenth century and temperament-al, from the seventeenth).

8.2.20 Contrasting -ic and -ic-al Alongside adjectives derived with -ic (relating to Latin -icus) and those derived with -al (Latin -alis), there was in the sixteenth century a flood of loans in -ical (relating to Late Latin -icalis), essentially combining these suffixes. A fair number of roots took both -ic and -ical, with the same or very similar meanings. In the sixteenth century there were domest-ic and domest-ic-al, grammat-ic and grammat-ic-al, metall-ic and metall-ic-al. There was competition and, as Marchand (1969: 342) puts it: ‘there can be only one survivor’. By the end of the nineteenth century, only domest-ic, grammat-ical, and metall-ic remained in use.

252    8 making new adjectives But some pairs did survive, and to do so they needed to establish a difference in meaning, often subtle but still discernible. We can compare: geograph-ic fact philosoph-ic point histor-ic decision electr-ic shock

geograph-ic-al study philosoph-ic-al discussion histor-ic-al film electr-ic-al energy

The -ic adjectives describe a type of fact, point, decision, and shock (reminiscent of the ‘specific’ profile of -ic when contrasted with -ous, in 8.2.18). In contrast, the -ic-al derivations indicate manner or nature—a study in terms of geography, a discussion in the manner of philosophy, a film in the character of history, energy supplied by electricity. In the same vein, class-ic-al refers to a manner of architecture, painting, music, etc.—the class-ic-al era. This contrasts with class-ic which describes a particular art-work or performance as being of outstanding worth; for example, The recital was of class-ic-al music but it certainly was not of class-ic quality. There are some -ic| -ical pairs whose meanings have radically diverged. Econom-ic refers to the discipline of economics—for instance econom-ic theory, econom-ic reform—whereas econom-ical indicates a low rate of spending or use—econom-ical housekeeping, econom-ical fuel consumption. Adverbs can be formed from many adjectives by adding -ly—see 10.4. Interestingly, the -ly is almost always added after -ic-al, only very seldom straight after -ic; today one hears domest-ic-al-ly, histor-ic-al-ly, electr-ical-ly, class-ic-al-ly rather than *domest-ic-ly, *histor-ic-ly, *electr-ic-ly, *class-ic-ly. Kaunisto (1999, 2004, 2007) provides detailed empirical studies of a number of -ic| -ic-al pairs.

8.2.21 Contrasting -ous and -al Industry is unusual in that it forms an adjective with both -ous and -al. However, these relate to different senses of the noun:

•

an industri-ous student is someone characterized by industry, in the sense of ‘hard work’

8.2.22 (11) - ful   253

•

an industri-al site is a place relating to industry in the sense of ‘organized manufacturing’.

8.2.22 (11) -ful 8.2.22a OE created suffix -fu(l)—out of adjective full—and this was added to abstract nouns to derive adjectives. For example, synn-full ‘sin-ful’, hyhtfull ‘joy-ful’, scam-full ‘shame-ful’. Many of these continued, and more were created within ME and Modern English, both from Germanic roots (  fearful, harm-ful ) and from Romance loans (  faith-ful, grace-ful ). From ME times the suffix was also added to some verbs; see 8.4.4. Suffix -ful can be either /-fəl/ or else /-fl̩/, with a final syllabic lateral; thus pain-ful, /'pein-fəl/ or /'pein-fl̩/. It is pretty well confined to abstract nouns (types L–N) and has the meaning ‘showing’, as in joy-ful, thought-ful, forceful, skil-ful, colour-ful. There are just a few instances with adjectives, such as wrong-ful and right-ful. The suffix may sometimes be glossed as ‘characterized by’ as in peace-ful and (occurring here with a concrete noun) tear-ful. 8.2.22b Many adjectives created with -ful take the negative prefix un-. A number may also occur with -less (see 8.2.31). Sometimes un-root-ful and root-less appear to have very close meanings; for example un-merci-ful and merci-less. Other times there is a clear difference in meaning—one talks of an un-law-ful action, something which the law does not allow, but of a lawless person, someone who is always breaking the law. Some -ful derivations don’t take un-, the -less form being preferred; people generally say care-less rather than *un-care-ful, and similarly for colour, purpose, and power. However, un-skil-ful is preferred to skill-less. 8.2.22c The adjective-deriving suffix -ful occurs on a goodly number of nouns but appears to be no longer productive. It must be distinguished from a quite different suffix -ful, which is productive and can be added to any noun referring to a container or building, deriving a noun with meaning ‘a full container/building (of   )’, as in cup-ful. OE did have the compound handfull, but this second -ful developed into a suffix only in ME times (spoon-ful, house-ful, mouth-ful ) and later (arm-ful, plate-ful, spade-ful, room-ful ). The noun-deriving -ful may be pronounced either as /-ful/ or as /-fəl/, being then homophonous with the adjective-deriving suffix; but, unlike that suffix, it is unlikely to be reduced to /-fl̩/ with a final syllabic /-l̩/.

254    8 making new adjectives 8.2.23 Contrasting -ly and -ful Suffixes -ly and -ful basically occur with different types of noun. However, they may both be added to master:

• •

master-ly ‘in the manner characteristic of a master, with fine competence’ (this has distinctly positive overtones, as is common for -ly derivations) master-ful ‘showing the qualities of a master’ this may be used in a positive way (something like ‘leading by example’) or with a negative nuance (‘dominating’)

8.2.24 Contrasting -ous and -ful There are a number of nouns which can take either the Germanic suffix -ful, ‘showing’, or the Romance-origin suffix -ous, ‘characterized by’. They include both Germanic roots (such as wonder and right) and also some of Romance origin (pity, grace, bounty). Compare, for example:

• •

grac-ious behaviour, which may be courteous or benevolent grace-ful movements, which may be flowing and elegant

For right, different senses of the word are involved in the two derivations:

• •

right-eous indignation/crusade—characteristic of what is considered proper and appropriate right-ful ownership/position—showing a just claim

(It was mentioned in 8.2.14a that right-eous arose from re-analysis of OE right-wise.) Other pairs show varied types of difference. Wondr-ous means ‘remarkable and excellent’, a stronger approbation than wonder-ful. Bount-eous and bounti-ful have similar meanings (both seem rather archaic), as do pit-eous (also a little archaic) and piti-ful (this form is widely used). 8.2.25 (12) -some Adjective-forming suffix -sum was active in OE; for instance, from lang ‘long’ was derived lang-sum ‘prolonged, lasting, tedious’ and from wynn ‘joy’ there was wynn-sum ‘pleasant, delightful’. The latter has survived into the modern

8.2.27 contrasting - ous and - some   255 language as winsome, although the original root has been discarded. The suffix -some /-səm/ (sometimes reducing to /-sm̩̩/) remained productive in ME times and for a while thereafter but always had a fairly low frequency. It derives adjectives from verbs (for example, loath-some) as well as from nouns. Some roots do double duty as nouns and as verbs and either could be taken as the basis for adjectivization (these include trouble, quarrel, and venture); they are discussed in 8.4.11. The meaning is ‘apt to cause’ (as worrisome) or ‘apt to’ (meddle-some). A few roots are unequivocally nouns and here the meaning is ‘apt to undertake’, as in adventure-some, or ‘apt to be a/have’, as in burden-some, flavour-some. (There are also a few derivations from adjectives, such as lone-some.) During the development of English, -some has gradually come to be used less and less; it is no longer productive, save perhaps in nonce and jocular coinings. There is a quite different formative -some, from OE indefinite form sum ‘some’. It is added to lower numbers and forms a noun referring to a cohesive group of that number of people: a foursome is four people playing golf together.

8.2.26 Contrasting -ly and -some The adjective lone (as a modifier) alternating with alone (as copula complement) forms both lone-ly and lone-some, with very similar meanings. They can both be applied to a person, indicating that they are alone and feel unhappy because of this, or to a place, indicating that anyone there would experience this feeling. A difference may be in genre: lone-some is less common and sounds a little more high-flown and old-fashioned, whereas lone-ly is an everyday term.

8.2.27 Contrasting -ous and -some Romance roots trouble, venture, and adventure form adjectives with both Romance suffix -ous and Germanic -some. Each has a perceptible semantic resonance as shown by typical collocations: troubl-ous times trouble-some client adventure-some youth adventur-ous undertaking venture-some investor ventur-ous speculation

256    8 making new adjectives The -some derivations frequently refer to characteristics of a person (or of a thing, as when a king exclaims: this troublesome crown), whereas the -ous forms often relate to an activity or an era. There is however, a great deal of overlap, each form being found with both senses. The more overlap of meaning, the more likely it is that one form will become more frequent and eventually eclipse the other. Today, adventurous and trouble-some are far more common than adventure-some and troubl-ous respectively. (The adjectives based on venture both retain a low frequency.)

8.2.28 Contrasting -ful and some The Germanic word awe originally meant ‘fear, dread’ (this sense is now obsolete for the noun). The adjective aw(e)-ful—dating from OE times— was based on this. The original meaning of aw-ful ‘characterized by dread’ gradually shifted to the present-day sense ‘unpleasant, really bad’. The meaning of the noun shifted to ‘dread, touched with veneration’ (often applied to a god) and then to ‘wonder, perhaps tinged with trepidation’; it is from this that, in the late fourteenth century, adjective awe-some was created, meaning ‘apt to induce wonder, remarkable’.

8.2.29 (13) -free The Germanic suffix -free /-ˌfri:/ always bears secondary stress. It would have commenced as the recurrent second element of compounds, but developed into a suffix about a thousand years ago; OE had such useful derivations as gafol-fre¯o ‘exempt from tax or tribute’ and toll-fre¯o ‘exempt from toll’. The meaning of the suffix is similar to that of adjective free (of), indicating ‘lacking something considered as undesirable’. Thus fat-free meal, mosquito-free field location, smoke-free environment, tax-free allowance. The suffix can be added to any noun which refers to something thought of as undesirable. Contrary-wise, if -free is used with a noun, one can assume that the speaker regards the referent of the noun as something to be avoided in the context. Feminists may welcome a man-free bar, people on a diet look for sugar-free products, lazy schoolchildren may long for a grammar-free syllabus. One hears on the news that a holiday weekend has been fatality-free. In recent times, some schools have had pupil-free days, when the teachers

8.2.31 (14) - less   257 can get on with tasks other than teaching; but this use of -free surely indicates a strange attitude towards pupils! The suffix is full y productive and may be added to a multi-word combination, as in The resort is [    fat American tourist]-free, This village is [cat and dog]-free. 8.2.30 Contrasting -some and -free Suffix -some has a limited distribution, but -free may be used with virtually any noun that has an appropriate meaning. Just a few forms may take either suffix. For example, with trouble, they create adjectives with almost opposite meanings:

• •

trouble-some (for example: back injury, or teenagers) ‘likely to cause trouble’ trouble-free (for example: carnival) ‘lacking trouble (which would be undesirable)’

Other examples include worri-some and worry-free, toil-some and toil-free.

8.2.31 (14) -less OE had free form le¯as ‘lacking, devoid of’ and a related adjective-creating suffix -le¯as, as in care-le¯as ‘lacking care, anxiety’ and wı¯f-le¯as ‘without a wife’. Only the suffix, -less, /-ləs/ or /-lis/, has come through into the modern language. (The present-day index of comparison less has a quite different origin, being a grammaticalization of læssa, the original irregular comparative form of OE adjective lȳtel ‘little, small’.) Unlike -free, suffix -less has a reduced vowel and does not take secondary stress. Whereas -free indicates lack of something considered undesirable, -less generally relates to the lack of something which one would expect (and, usually, hope) to have. Like -free, the suffix -less applies to nouns of virtually every type—hair-less youth, mother-less child, fish-less pond, horse-less cavalryman, flower-less room, water-less plain, moon-less night, thought-less decision, joy-less laughter, taste-less food, end-less sermon. A motion-less person refers to someone who would be expected to be in motion; and a rest-less child is someone whose moving about is a nuisance and who should ideally be at rest.

258    8 making new adjectives It might appear that not all nouns taking -less necessarily describe things one would expect to have. What about fear-less? In 5.5b we noted that it is generally a sensible course to exercise a modicum of fear in a dangerous situation, in order to survive. Someone who is fear-less may be admired for their courage but they may also be putting themselves at risk. One should feel shame after doing wrong; hence shame-less describes those who do not do so. Power-less refers to someone who lacks power when they should have it and ought to exert it, as in The general was power-less to stop his troops looting and rampaging. Effort-less describes something which would require considerable effort for most people—The champion lifted the heavy weight with effort-less ease. The suffix -less is also added to a fair variety of verbs; see 8.4.3. It is fully productive, with both nouns and verbs. And it may be added to a multi-word combination, as in She’s [mother and father]-less, and Sadly, we’re a [wealthy sponsor]-less club.

8.2.32 Contrasting -free and -less The referent of a noun can be regarded as what one should have, in one circumstance, but as something which is undesirable, in another; it will take -less in the first instance and -free in the second. Compare:

• •

parent-less child—an orphan, with no one to look after them parent-free evening—when teenage children have the house all to themselves, for a party

A vehicle may be noise-less, when a little noise would be welcome so that one can hear it approach. But a noise-free environment is a fine place to study. A grazier will look with disdain on a grass-less plain, while a gardening-phobic householder will welcome a grass-free backyard. Child has two meanings. It is in semantic type B of 8.2 (and Table 8.2) as an age-group label, ‘young person’. It is also in type C as a kin term, ‘son or daughter’. Suffixes -free and -less, added to this noun, relate to the different senses:

• •

child-free environment—type B sense, no young people present (who might provide annoyance) child-less woman—type C sense, someone who has not given birth to a daughter or son (as most women do)

8.2.33 contrasting -ful and - less   259 A nightmare is a bad dream. A dream which is not categorized as a nightmare is likely not to be too unpleasant. Hence it is usual to employ -less with dream and -free with nightmare, as in She fell into a deep dream-less slumber and He has been suffering from nightmares a lot recently, but last night he experienced a nightmare-free repose.

8.2.33 Contrasting -ful and -less One can read in a large and magisterial English grammar book that -less ‘is the negative counterpart of -ful’ (Bauer and Huddleston 2002: 1711). But the facts speak differently. Affix -ful is almost restricted to abstract nouns and is no longer productive, whereas -less may be added to a noun of almost any type and is fully productive. There are some nouns which derive adjectives by adding either -ful and -less, others which only take -less and a small number which are confined to -ful. Examples are in Table 8.3 (see also 5.1e). Both suffixes can be added to verbs, and we again find all three possibilities: thank-ful and thank-less, count-less but not *count-ful, forget-ful but not *forget-less (there are more verbs in each set). See 8.4.3–5. In 8.2.22b there is discussion of roots which take either negative prefix un- plus suffix -ful, or else suffix -less. Table 8.3  Contrasting the use of -ful and -less pain-ful

pain-less

in type A

thought-ful power-ful grace-ful fear-ful

thought-less power-less grace-less fear-less

and some others in types L–N

—

brain-less

and many others in types A–C

— — —

horse-less hat-less name-less

and many others in types D–J

— —

luck-less age-less

and many others in types K–Q

delight-ful scorn-ful

— —

and a few others in types L–N

260    8 making new adjectives 8.2.34 Contrasting -ful, -free and -less Care (which can be a noun or a verb) is of interest since it may form an adjective with all three suffixes:

• • •

care-ful (e.g. person, or approach, or driving) ‘characterized by appropriate care’ care-less (e.g. person, or approach, or driving) ‘lacking care, when this should be observed’ care-free (e.g. vacation, or summer) ‘situation in which everyday cares are absent’

8.2.35 Contrasting -less and other suffixes Since -less describes lack of something, it can contrast with quite a few of the adjective-deriving suffixes which relate to some positive characteristic. Some examples are in Table 8.4. Table 8.4  Suffixes which contrast with -less (2)-ly (5) -y

(7) -ed (8) -ous (10) -al (11) -able

shape-ly brain-y noise-y luck-y spirit-ed humour-ous function-al comfort-able

shape-less brain-less noise-less luck-less spirit-less humour-less function-less comfort-less

and hair, breath, bone, nerve, flower, etc. and milk, sun, dust, draught, etc. and fault, taste, price, worth, etc and limit, money, shade, spot, etc. and ambitious/ambition-less, etc. and count, profit, value, etc.

8.2.36 (15) -able, -ible This suffix relates to Latin -a¯bilis or -ibilis, giving rise to alternative spellings -able and -ible, both realized as /-əbl/. There was a torrent of loans from French into ME. Some—such as durable, available, and compatible—were not analysable in English. But there were many pairs of words with and without the suffix. Most were verbs, such as change and change-able; these are discussed in 8.4.7. But there were also a good few nouns, including comfort and comfort-able, profit and profit-able. The adjective able was also borrowed from French; this has no genetic connection with the derivational suffix, but there is a coincidental similarity

8.2.39 (16) - ary   261 of meaning which probably assisted in the recognition of -able as a productive suffix. It was then added, within English, to many verbs and nouns of Romance and also of Germanic origin (always with the form -able) For example fashion-able, pleasur-able, siz-able (Romance), tam-able, ford-able, bridge-able (Germanic). Adjective-creating -able is today perhaps the most thoroughly naturalized of all suffixes of Romance origin. Suffix -able derives adjectives from nouns with the meaning ‘suitable for’ (for example, marriage-able, memor-able, tax-able), or ‘characterized by, showing’ (as knowledg-able, reason-able), or ‘providing’ (comfort-able, profit-able). It occurs predominantly with non-concrete nouns, as illustrated in Table 8.2. The suffix is fully productive with nouns and especially with verbs, occurring also with phrasal verbs; see 8.4.7.

8.2.37 Contrasting -free and -able It is interesting to compare the use of -free and -able with the same noun. Tax is something which has to be paid on certain types of income and profits—these are tax-able ‘suitable for tax’. But everyone welcomes something which is declared to be tax-free, meaning that tax does not have to be paid on it.

8.2.38 Contrasting -ical and -able| -ible Adjective sens-ible and noun sense were borrowed from French into ME in the fourteenth century (the i reflecting Latin form sensibilis, French sensible). Sensible means ‘characterized by sense, showing sense’; we can talk of a sens-ible person, a sens-ible decision, or a sens-ible idea. Early in the seventeenth century, non-sense was formed, by adding non- to sense. Within a few years an adjective was founded on nonsense, but in a different way from that inherited by sense—adding Romance suffix -ical, rather than -able| -ible. A decision, idea, or argument can be described as nonsens-ical, meaning that it appears to have no sense to it. Unlike sens-ible, nonsens-ical is not used to describe a person.

8.2.39 (16) -ary The Romance suffix -ary /-(ə)ri/ reflects -arius| -arium in Latin. Loans from French into ME included contrary and arbitrary (which are not analysable) and also such pairs as pigment and pigment-ary, station and station-ary, tribute

262    8 making new adjectives and tribut-ary. In many of these loans -ary is preceded by -((a)t)ion or -ment and it was later appended to further words ending in these nominalizing suffixes—for example, discretion-ary, evolution-ary, document-ary, rudiment-ary. When added to a noun of three or more syllables ending in -ment, primary stress may shift to the syllable immediately preceding -ary; for example, fragment /'fragmənt/ and fragment-ary, pronounced either as /'fragməntəri/ or as /frag'mentəri/; and also for parliament-ary, complement-ary, complimentary, among others. Adjectives derived with -ary are unusual in that they are almost always used to modify abstract nouns, rather than nouns referring to people or other concrete objects (revolution-ary and vision-ary are among the few exceptions, being used to describe ideas and also people). The meaning of -ary is most often just ‘is a’. A document-ary film is a film which is a document; a compliment-ary remark is a remark which is a compliment. Or it can mean ‘is a part of’, or ‘is like a’—diet-ary fibre is fibre which is part of a diet, and a revolution-ary idea is an idea which is like a revolution in the way of thinking about something. As can be seen, the suffix can be added to words referring to activities (revolution, evolution, inflation), to society (custom, discretion, diet), to some human qualities (caution), and to types (example, complement), among others. The suffix -ary may be productive on further nouns ending in -(a)(t)ion or -ment. It is also much used in scientific work. The OED quotes from a mathematical text of 1940 the definition: ‘An m-ary n-ic . . . is a homogenous polynomial, with arbitrary constant coefficients, of degree n in m independent variables’. There is further discussion of -ary in 9.6.1 and 9.3.12b.

8.2.40 (17) -proof ME borrowed verb prove from Old French and, on the basis of this, created proof, functioning as noun and as adjective. In the sixteenth century, proof developed from being a compounding element to become a derivational suffix -proof /-ˌpruf/, always bearing secondary stress. It basically means ‘keeps out’ and may be added to nouns referring to types of people (burglar-proof, child-proof, fool-proof  ), to types of explosive devices (bullet-proof, bomb-proof  ), to natural phenomena (water-proof, fireproof, sound-proof  ), and also to a selection of verbs (see 8.4.18).

8.2.43 (20) - ward(s)   263 For example, a locking device may be described as child-proof if it is so designed that a child could not open it. However, the suffix takes on a meaning shift in fool-proof and idiot-proof, which describe something which is so simple even a fool or an idiot could understand or use it. Suffix -proof is productive, and can be used with multi-word combinations; for example It’s both [water and draught]-proof and This lock is both [child and thief ]-proof.

8.2.41 (18) -ern OE had suffix -erne, which attached just to names of compass points, and this is retained today as -ern /-ən/. It is only used with the four compass points and combinations of them, forming adjectives, as in east-ern boundary ‘boundary to the east’, southwest-ern side ‘the side that is to the south-west’. There is consonant voicing between north /nɔ:θ/ and north-ern /'nɔ:ðən/, and both vowel and consonant change between south /sauθ/ and south-ern /'sʌðən/.

8.2.42 (19) -most Modern suffix -most, /-moust/ ~ /-ˌmoust/, can optionally take secondary stress. It is not related to free form most, but reflects suffix -mest in OE meaning ‘most extreme’, as in u¯t-mest ‘furthest out’ from u¯t ‘out’, and æfte-mest ‘last’ from æfter ‘behind, after’. Today the suffix indicates ‘furthest in a particular direction’ and is added to compass points terms either directly or after -ern; for example, south-most and south-ern-most appear to be interchangeable. The suffix is also added to terms of orientation—such as top, bottom, front, fore, left, right—and to a number of prepositions/adverbs—including in, out, up, down.

8.2.43 (20) -ward(s) The forms derived by -ward(s) /-wəd(z)/, ‘in the direction of’ function mainly as adverbs and the major discussion of this suffix is in 10.6.1. However, many of the derivations also do double duty as adjectives.

264    8 making new adjectives The suffix applies to compass points—north-ward extension, south-ward excursion. And it can be added to many nouns referring to a place or to a thing in a fixed location, as in home-ward journey; and sky-ward, gardenward, bed-ward, heaven-ward, Paris-ward, and so on. It is also used on some prepositions/adverbs, including on, out, in, down, back, and after. The shorter form, -ward, is generally used when it is employed adjectivally, to modify a noun—in a sea-ward direction, a back-ward move—and -wards is preferred when it is used adverbially—fly up-wards, drive homewards. However, there is a considerable degree of interchangeability. The question of whether or not to include the final s is essentially a matter of whim, and perhaps considerations of euphony.

8.2.44 (21) -ate ME borrowed from Romance a number of adjectives ending in -ate (which reflects Latin participle -atus). For example, desolate and separate (which are not analysable in English), but also fortun-ate alongside fortune, and passion-ate together with passion, consider-ate plus consider. This should have been enough to establish -ate /-ət/ ‘full of’ as an adjective-deriving suffix in English. Indeed, affection-ate was created in the late fifteenth century, based on noun affection. But that seems to have been all. (Other -ate suffixes are discussed in 7.3.5 and 9.5.6.)

8.2.45 (22) -ine, -ene Suffix -ine or -ene, pronounced variously as /-ain/ or /-i:n/ or /-i:ən/, emanates from Latin -inus (which is cognate with Germanic -en). This ending is recognizable in many Latinate adjectives for which there is no underlying noun in English; for example masculine, marine, bovine (and many further terms used in Chemistry). There are a handful of loan pairs where the adjective describes the property associated with the noun—for example, crystal and crystall-ine. There have been many coinages in Chemistry (such as benzene), and a few which are in general use—for example, alkal-ine /'alkəl-ain/ came in during the seventeenth century, and labyrinth-ine /'labərinθ-i:ən/ during the eighteenth. It is scarcely productive today. The suffix is also used with some proper names; for example Florent-ine /'flɔrənˌt-ain/ ‘associated with Florence’ (this is also an adaptation from Latin), and Cair-ene /'kaiər-i:n/ ‘associated with Cairo’.

8.3 adjectives derived from proper names   265 8.2.46 (23) aAlmost without exception, the adjectives dealt with so far in this chapter may function both as modifier to a noun within a noun phrase—for example, my envi-ous neighbour—and as the complement within a copula clause (what is sometimes called ‘nominate predicate’)—as in: My neighbour is envi-ous. Prefix a- /ə-/ derives adjectives from verbs (see 8.4.20) and from some nouns, but these may only function as copula complement, not as modifier. One can say My son, the sailor, is a-shore for a couple of weeks, but not *my a-shore son (only my son who is a-shore). This constraint is explained by the etymology. OE preposition on ~ an ‘on’ was first compounded with locational terms, and then developed into a prefix. With nouns it retains the original meaning ‘on’—if a boat runs a-ground its keel is resting on the ground, and if someone is a-bed they are lying in or on the bed. Others include a-stern, a-midships, a-side. The point is that an a-derived adjective betrays its origins in still behaving like a preposition followed by a noun phrase. One can say The duchess is in (the) bed or The duchess is a-bed. However, *the in-bed duchess is not acceptable, and neither is *the a-bed duchess. The prefix is still used productively, but in a jocular sense and mainly with verbs; see 8.4.20.

8.3 Adjectives derived from proper names There is generally some suffix available for deriving an adjective from the name of a nation or place and we first focus on these. Some also apply to names of people but the possibilities here are more limited; these are summarized and assessed in 8.3.11. The seven main suffixes used to form adjectives from the proper names of nations and places are exemplified in Swed-ish, Iceland-ic, Canaan-ite, Bengal-i, Chin-ese, Alaska-n, and Egypt-ian. Of these, only -ish is native Germanic. Most of the suffixes are used productively; they bring with them some characteristics of occurrence from source languages but have been so adopted into English that there are a number of rough-and-ready rules for which suffix a given proper name should take. (It should be noted that there are a few exceptions to the principles of suffix assignment suggested here.)

266    8 making new adjectives There are basically three forms associated with (a) the name of each nation. These are (b) the name for an inhabitant; (c) an adjective describing the culture and language; and (d) a generic term covering all the inhabitants. Names of nations can be divided into Type A and Type B, with respect to the derivations which apply to them. These are discussed in turn.

8.3.1 Type A derivations The forms for Type A are displayed in Table 8.5. In sets Ia and Ib, the basic term is in column b. For some of the nation names, in column a, -land is added to the basic term (for example, Fin-land); others have a distinct nation name (such as Denmark). The sole term in set Ib, Jew, lacks any nation name. The derived adjective involves addition of suffix -ish to the term in column b (Finn-ish) and the generic label is the b-s (the Finns). The same principles originally applied for most of the terms in set IIa; there were column b terms from which the adjectives in column c were derived by adding -ish—Walh-ish > Welsh, Iras-ish > Irish, Angle-ish > English,

Table 8.5  Type A of nation names with derived forms (a) NATION

(b) AN INDIVIDUAL

Denmark Sweden Po-land Fin-land Scot-land Turkey

Dane Swede Pole Finn Scot Turk

Ib

—

Jew

IIa

Wales Ire-land Eng-land France Holland/The Netherlands

Ia

c plus -man

IIb

Spain

Spaniard

IIIa

Ice-land Green-land

a plus -er

IIIb

New Zealand

(c) ADJECTIVE

(d) THE POPULATION

b plus -ish

the b-s

Wel-sh Ir-ish Engl-ish French Dutch

the c

Span-ish a plus -ic b-s =a

8.3.1 type a derivations   267 Frank-ish > French. However, the forms Walh, Iras, Angle, and Frank fell out of use. A new column b term was created by adding -man—or -woman—to the adjective in column c (Irish-man); the generic term, in column d then was simply the c (the Irish). Adjective Dutch has a different origin, being a German term with wide reference to Germanic people, later contracting its meaning in English to relate just to Holland (a.k.a. the Netherlands). In IIb, Span-ish is from Spain plus -ish with vowel reduction. The inhabitant term Spaniard is a loan from French (and see 9.3.12c). There are minor variations on the pattern shown in Table 8.5. For example, Scott-ish has reduced to Scotch, with both terms having adjectival function (together with Scots). An inhabitant can be called a Scot or a Scots-man (or Scots-woman). For set IIIa the name of the country (column a) is taken as basic, with -er added for column b (Iceland-er) and -ic for column c (Iceland-ic). The generic term, in column d, is b-s (Iceland-er-s). There was an older form Iceland-ish, but it has now pretty well been replaced by Icelandic. There is no special adjective for New Zealand, in IIIb; just the name of the country is used (for example, New Zealand cheese). Derivations relating to Britain show a fascinating development. In OE times there was a Ia system (normalizing this to present-day spellings): (a) Britain

(b) Brit

(c) Brit-ish

(d) the Brit-s

Then the term Brit fell out of use. (Briton is generally used for referring just to the Celtic inhabitants of Britain.) We then had a IIa system with a gap in column b: (a) Britain

(b) —

(c) Brit-ish

(d) the Britt-ish

How did one refer to an individual? Simply as ‘an inhabitant of Britain’, or by specifying which part of Britain they came from: English-man, Scots-man (or Scot), Welsh-man, Irish-man (or, of course, -woman). In the nineteenth century, Americans coined British-er to fill the gap (like a set III term). Finally, in the twentieth century, the term for an inhabitant was back-formed from British—this was Brit (reproducing the noun which had been lost a millennium before). We now have (a) Britain

(b) Brit

(c) Brit-ish

(d) The Brit-ish (IIa) or The Brit-s (Ia)

268    8 making new adjectives We thus have, for type A, two of the adjective-deriving suffixes discussed above for their occurrence with common nouns, -ish (see 8.2.4) and -ic (8.2.16). Their use with proper names can now be summarized.

8.3.2 (3) -ish with proper names This is the only Germanic suffix used with proper names. Most of the forms in column c for sets I and II in Table 8.5 go back to Old English. The suffix was extended to apply to Jew in the sixteenth century, to Finn in the eighteenth, and to Turk in the nineteenth. We also find -ish with the name of an English country, as in a Kent-ish custom. Plus Pict-ish, Gaul-ish, and a number more. It was noted in 8.2.4 that the great majority of common nouns taking -ish are monosyllabic; so also for the names of nations and the like. Interestingly, unlike -ic and -ian (see 8.3.3, 8.3.8–11), we do not find -ish added to the proper names of people. Note that, with names of nations and other places, -ish does not carry any of the negative overtones it often does when used with common nouns. Although still productive with common nouns, this suffix is probably not productive with names. 8.3.3 (9) -ic with proper names As in its use with common nouns (8.2.16), suffix -ic /-ik/ generally requires primary stress to go on the syllable immediately preceding -ic, with some consequential vowel changes. Whereas common nouns may take -ic or -ic-al (or both, see 8.2.20), proper names only occur with -ic, never -ic-al. Nation names plus -ic generally reflect Latin forms, as in Celt-ic, Gall-ic, and German-ic (today used, together with German, as an adjective relating to Germany). It was extended to apply to Iceland in the seventeenth century, and to Greenland in the nineteenth. The suffix -ic is often used in scientific nomenclature, including the names of language families and subgroups (here alternating with -ian and -an). For example, the Turk-ish language belongs to the Turk-ic family; the German language belongs to the German-ic subgroup of the Indo-Europ-ean family. A few adjectives are derived from proper names of people by adding -ic. For example, Homer-ic, Plato-n-ic, Socrat-ic, and Satan-ic; all these came from Greek (some via Latin) between the sixteenth and eighteenth centuries. The suffix has also been used to derive adjectives from post-classical names, as Milton-ic and Byron-ic; see 8.3.11. It is no longer fully productive with proper names, as it is not with common nouns.

8.3.4 type b derivations   269 8.3.4 Type B derivations We can now examine Type B, which has many more members than Type A. Just a sample of members is provided in Table 8.6, which has vowel-final names on the left (sets Ia–g) and consonant-final sets on the right (IIa–f). For all Type B names, there is one derived form covering all of functions (b), (c) and (d) (plural -s may be added for the noun in column d). For example: (a) C  hina Africa

(b) a Chin-ese (c) Chin-ese culture an Africa-n Africa-n culture

(d) the Chin-ese the Africa-n-s

We can now examine each box in Table 8.6, positing rough-and-ready rules for which suffix a given name will take. Table 8.6  Type B of names of nations and other places plus derived forms a

b=c=d

China Burma Malta Guyana

Chin-ese Burm-ese Malt-ese Guyan-ese

Ib Africa America Corsica Jamaica

Africa-n America-n Corsica-n Jamaica-n

Ic

India-n Syria-n Russia-n Samoa-n

Ia

India Syria Russia Samoa

Id Australia Bohemia Galicia Utopia

Australia-n Bohemia-n Galicia-n Utopia-n

Ie

Fiji Fiji-an Malawi Malawi-an Zimbabwe Zimbabwe-an

If

Morocco Mexico

Morocc-an Mexic-an

Panama Peru

Panama-n-ian Peru-v-ian

Ig

a

b=c=d

IIa Japan Vietnam Nepal Surinam Senegal

Japan-ese Vietnam-ese Nepal-ese Surinam-ese Senegal-ese

IIb Lebanon Portugal

Leban-ese Portugu-ese

IIc

Egypt-ian Lao-t-ian Jordan-ian Brazil-ian Zaire-ian

Egypt Laos Jordan Brazil Zaire

IId Texas Caucasus Barbados Mauritius

Tex-an Caucas-ian Barbad-ian Mauriti-an

IIe

Israel Canaan Brooklyn Sydney

Israel-ite Canaan-ite Brooklyn-ite Sydney-ite

IIf

Punjab Bangladesh Yemen Israel

Punjab-i Bangladesh-i Yemen-i Israel-i

270    8 making new adjectives Set Ia. Two syllables, ending in -a, which is unstressed /ə/. Replace final /ə/ by -ese, /-'i:z/, which takes primary stress. Thus Malta /'mɔ:ltə/, Maltese /mɔl'ti:z/. Also for China, Burma, Vienna. From Guyana /'gi:ənə/ is derived Guyan-ese /gaiə'ni:z/, suggesting that for this purpose the material before final /nə/ counts as one syllable. Tonga /'tɔŋə/, the name of a Pacific Island nation, satisfies these conditions and one would expect the derived adjective to be Tong-ese. Indeed it was, in early mentions such as the narrative of the US Exploring Expedition of 1838–42 (Wilkes 1845: 380). But this was soon replaced by Tonga-n, perhaps by analogy with adjectives derived from the names of other Pacific Island groups: Samoa-n, Fiji-an, Hawai 'i-an, Tahiti-an. In similar fashion, one would expect the adjective derived from nation name Cuba to be *Cub-ese. In fact it is Cuba-n, quite possibly a loan from Spanish cuba-no. Set Ib. More than two syllables, ending in -a, which is unstressed /ə/. Add -n, /-n/, with no change in stress. There are many examples here; besides those in Table 8.6 we can mention Formosa-n, Alaska-n, Angola-n, Sri Lanka-n, Uganda-n, and Venezuela-n. Sets Ic and Id. Words of any length, ending in ia, which is unstressed /iə/. Add -n, /-n/, with no change in stress. Disyllabic names are given in Ic and longer ones in Id of Table 8.6. There are many further examples, including Austria-n, Gambian, Zambi-an; Albania-n, Algeria-n, Bolivia-n, Colombia-n, Ethiopia-n, Tunisia-n. Words of any length ending in oa or ua, which is unstressed /ouə/ or /uə/, behave in the same way, adding -n /-n/ with no change in stress; for example Samoa-n, Papua-n, Nicaragua-n. Interestingly, disyllabic words ending in /jə/ pattern with those showing final /iə/. From Kenya /'kenyə/ is formed Kenya-n /'kenyə-n/; similarly for Libya-n. Canada is an exception. It should fall in set Ib with derived adjective *Canada-n; the occurring form Canadian is probably a loan from French Canadien(ne). Set Ie. Any length, ending in an unstressed high vowel. Add -an, /-ən/, with no shift in stress. Names ending in /i/ include Fiji-an, Malawi-an. Ending in /e/ there is Zimbabwe-an. Chile can be pronounced with the final vowel either /i/ or /e/; in either case, /-ən/ is added.

8.3.4 type b derivations   271 Set If. With three or more syllables, ending in o, unstressed /ou/. Replace final o by -an /-ən/, with no shift in stress. This applies to Morocco /mə'rɔkou/, giving Morocc-an /mə'rɔk-ən/; similarly for Mexico, Mexic-an and for San Francisco, San Francisc-an. From Chicago we would expect to get *Chicag-an; in fact the derived term is Chicago-an. There are other trisyllabic names ending in /ou/ which add n and then -ian; for example, Buffalon-ian, Toronto-n-ian. I know of one disyllabic name ending in /ou/, Faeroe (or Faroe) /'fɛərou/; this forms an adjective in -ese, Faro-ese /ˌfɛərou-'i:z/. Set Ig. Ending in a stressed vowel. Peru and Panama belong in this set. Each forms a derivation with -ian /-iən/, which must be added after a consonant. The consonant comes about in different ways for the two names. From Peru /pe'ru:/ is derived Peruv-ian /pə'ru:v-iən/, probably from French péruvien (said to be based on modern Latin name Peruva). From Panama /panə'ma:/ there is Panaman-ian /ˌpanə'mein-iən/; the /n/ included here may be a copy of the preceding consonant in the word. Note that -ian requires the syllable preceding it to bear primary stress, with some consequential shift in vowel values. We can now turn attention to Type B names ending in a consonant, illustrated on the right-hand side of Table 8.6. Starting near the bottom, derivations from proper names of nations and peoples which end in -ite (set IIe) are generally reflexes of Greek and Latin forms. Although -ite nowadays has a certain productivity—with names of people and in scientific terminology—it is unlikely to be added to any new nation name. Suffix -i (set IIf) occurs in loans from the languages concerned. There appear to be no phonological rules for which nouns may take -ite or -i; the two suffixes are discussed a little later. There does, however, appear to be a phonological principle at work for names in sets IIa-d. Set IIa. Ending in -n, -m or -l, preceded by low vowel with primary stress. Add -ese, /-'i:z/, which now takes primary stress. For example, Japan /dʒə'pan/, Japan-ese /ˌdʒapə'n-i:z/. Besides the names in Table 8.6, this applies to Taiwan, Sudan, Canton, Siam. We also have Pekinese, based on Peking (an older form of the name Beijing) which is unusual since the final syllable is a stressed high vowel plus n. The common form derived from Java is, surprisingly, Javanese. It appears that initially the adjective corresponding to Java was Java-n (attested in the

272    8 making new adjectives OED from 1606), with Java-n-ese (attested from 1704) presumably being based on this. Set IIb. Shows some expected derivations in -ese for names which do not have a stressed final syllable. Both Portugal and Portugu-ese are loans from the Portuguese language. Leban-ese is probably a loan from French libanais, based on Liban, the name for Lebanon in French. The adjective derived from Congo might be expected to be *Cong-an or *Congo-an; the occurring form Congol-ese is probably also a loan from French. Set IIc. Essentially covers nation and place names ending in a consonant which are not in set IIa (or in IIb). They add -ian, /-iən/, which attracts primary stress to the syllable immediately preceding the suffix. Egypt and Laos end in consonants other than n, m, and l. Jordan ends in /-ən/ (or syllabic /n̩/). Brazil has final l but preceded by a stressed high vowel. Although Zaire and Singapore end in a vowel in British English, there is an underlying /r/ which selects -ian (and surfaces before it). Laos changes final s to t before -ian; this is again probably French influence. In Egypt-ian the ti becomes /ʃ/, /i'gipʃən/. In Laot-ian, t becomes /ʃ/ before i with the i retained, /'lauʃiən/. Set IId. A number of names ending in /əs/ (spelled as or us or os) replace this by -an /-(ə)n/ or -ian /-iən/ to form an adjective. Corresponding to Texas /'teksəs/ there is Tex-an /'teks-ən/ or /'teks-n̩/. Also Kansas/ Kans-an, Caucasus/Caucas-ian, Barbados/Barbad-ian, Mauritius/Mauriti-an, Honduras/­Hondur-an. There are a few more interesting exceptions to the rough-and-ready rules given above. The adjective based on Ghana is not *Ghan-ese (or Ghana-n) but Ghana-ian; this is likely to be based on the adjective ghanéen in French. There may be a similar reason for why the adjective based on Iran (which does end in a stressed low vowel plus n) is not *Iran-ese but Iran-ian; French has iranien. Then there is Tibet-an rather than the expected Tibet-ian— again, possibly influence from French. Why do we encounter Argentin-ian rather than *Argentina-n? There are in fact two adjectives relating to the nation name Argentina—Argentine and Argentin-ian. It is possible that the latter is simply founded on the former (as a set IIc term, ending in a stressed high vowel plus n).

8.3.6 (25) -i   273 The derivation Norweg-ian goes back to Norweg, the name in OE for what we today know as Norway. Glaswegian from Glasgow was created by analogy with Norwegian and so, more recently, was Taswegian (from Tasmania) as a rather far-out analogy to the other two. There are other irregular forms, such as Belgium| Belgian, Cyprus| Cypriot and Greece| Greek, some going back to forms in classical languages. And no doubt more besides. It is now time to consider, in turn, the five suffixes which occur with Type B proper names.

8.3.5 (24) -ite Derivations with -ite /-ait/ can function as noun or as adjective, referring to a person belonging to a certain place (a suburban-ite) or an associated characteristic (a suburban-ite attitude). The suffix comes from Greek -ites via Latin -ita and French -ite. The first loans into ME related to ancient peoples and places, such as Israel-ite, Canaan-ite, Sodom-ite. Its productive use today is for association with a certain place—New Jersey-ite, Durham-ite—or to a type of location— city-ite, beach-ite—or to a type of activity—social-ite, rugger-ite. The suffix can be added to the name of a memorable person—as Trotsky-ite, Jacobite—indicating adherence to a doctrine or cause associated with the person. The suffix is also used in scientific terminology, for names of substances— anthracite, dynamite, vulcanite, and further quite recondite materials.

8.3.6 (25) -i The derivational suffix -i /-i/ mirrors a form in the language which it describes. It so happens that an -i suffix is found both in languages of the Semitic branch of the Afro-asiatic family, and in languages from the IndoIranian branch of the Indo-European family. Like -ite, it can refer to a person or a quality associated with a certain place (an Iraq-i, and Iraq-i attitudes towards women’s education). Rather rarely, this suffix engenders stress shift in the base to which it is attached; for example Israel /'izreiəl/, Israel-i /i'reil-i/. Indo-Iranian nation/state names taking -i include Pakistan, Bangladesh, Bengal, Punjab, Rajasthan, Gujarat, and the old Hindustan. In the Semitic sphere we find Iraq, Kuwait, Oman, Yemen, and Zanzibar. The derivation

274    8 making new adjectives from Israel had been Israel-ite. In the OED there is a quotation from the New York Times of 5 December 1948: ‘Shortly after the proclamation of the new State of Israel . . . Moshe Shertok announced that its citizens would be called Israelis.’ Thus it came to be that Israel-ite was used in connection with the ancient land of Israel and Israel-i with the modern one. 8.3.7 (26) -ese 8.3.7a Suffix -ese /-'i:z/ takes primary stress. A suffix of similar form is found in a number of Romance languages, but the primary source for -ese in English is probably Italian. Loans commenced in the late fifteenth century, with Milan-ese. The meaning is similar to those of -ite and -i ‘person or quality associated with a certain place’. As pointed out in 8.3.4, it is primarily used on nation names which are disyllabic and end in unstressed /ə/ (set Ia in Table 8.6), and on Faeroe which is disyllabic and ends in unstressed /ou/ (discussion of set If ). And on those which end in a stressed low vowel plus n, m, or l (set IIa). The suffix is also used with some city names: Vienn-ese and (as would be expected) a number in Italy—Geno-ese (or Genov-ese, based on it name in Italian, Genova), Veron-ese, Bologn-ese among them. 8.3.7b There is another suffix -ese /-'i:z/—or else a quite different sense of the one suffix—which derives nouns from the names of places, people and professions, describing the style of language associated with them; for example, New York-ese, Kipling-ese, journal-ese, telegraph-ese. Within the last fifty years, the style of ‘baby talk’ used for communication with a small child has come to be called mother-ese. These derivations cannot be used adjectivally. The endings -n, -an, -ian and -ean are often taken to be forms of a single suffix. However, on both formal and functional grounds, it seems most appropriate to regard them as two suffixes, one with orthographic forms -an and -n, the other with -ian and -ean. (But it will be seen that there is a little blurring of the boundary between them.) 8.3.8 (27) -(a)n This suffix has form -n /-n/ after a vowel and -an /-ən/ after a consonant, and generally does not affect stress placement. It reflects adjectival suffix -anus in Latin.

8.3.10 contrasting - an and -ian/-ean   275 This suffix is added to names in sets Ib–f of Table 8.6, and is used on some in set IId. It can also be added to names of people which end in an unstressed vowel; for example Chomski-an /'tʃomsk-iən/ (an alternative description is Chomsky-ite). Just a handful of common nouns form adjectives with -an, such as republic-an. 8.3.9 (28) -ian, -ean 8.3.9a The suffix written as -ian or -ean, pronounced /-iən/, reflects Latin -anus added to a stem ending in i. Primary stress is generally moved to the syllable immediately preceding the suffix, sometimes with consequential vowel change; for example Devon /'devn  ̩/ and Devon-ian /de'von-iən/, Iran /i'ra:n/ and Iran-ian /i'rein-iən/. (A rare exception to this rule is Europe /'yuərəp/, Europ-ean /yuərə'piən/ where stress falls on the suffix itself.) The initial /i/ of the suffix may fuse with a root-final /t/, producing /ʃ/, as in Egypt /'i:dʒipt/and Egypt-ian /i'dʒipʃ-ən/. Adjective-deriving suffix -ese is used with the names of places but seldom with names of people. The same applies—but to a lesser extent—with -(a)n. Instead, -ian is employed, as in Aristotel-ian, Darwin-ian, Dickens-ian, Freud-ian, Christ-ian and very many more. A number of adjectives not related to names end in -ian but there is no underlying noun root in present-day English; for example, plebian, equestrian, crustacean, barbarian. We do get guard and guard-ian, both borrowed from Romance. 8.3.9b It is not always easy to distinguish between suffixes -(a)n and -ian. Should historian be analysed as histori-an (with suffix -an) or as historian (involving -ian)? The fact that there is stress shift, from /'histəri/ to / his'tɔ:riən/, could be cited in favour of the -ian analysis. The derivation historian only functions as a noun. But we also find stress shift in some adjectival forms derived from nation names, suggesting analyses Ital-ian and Hungarian (rather than Itali-an and Hungari-an). Suffixes -ian and -an are much used—together with -ic—in scientific nomenclature, including the names of language families; for example, Dravidian, Algonkian (or Algonquian), Athapaskan. 8.3.10 Contrasting -an and -ian| -ean Suffixes -an and -ian| -ean have similar origin and identical meaning. They are essentially in competition. We can enquire what motivates which suffix is used with which noun?

276    8 making new adjectives Generally speaking, -ian is used after the names of people (or common nouns) which end in a consonant. There are a few exceptions which do add -an—notably Elizabeth-an, Luther-an, Tibet-an, suburb-an, and Mohammedan. In fact, for the first four of these, both -an and -ian derivations have been in use, with the same meaning. Elizabeth-ian is recorded from 1807, but was soon replaced by Elizabeth-an. The OED has quotations for Luther-ian from 1530 to 1589 but it was then eclipsed by Luther-an (also first noted in 1530). Tibet-ian is attested from 1747, and there is a quotation from 1973, but it has now largely been over-shadowed by Tibet-an (from 1822). Suburb-an (from about 1400) and suburb-ian (from 1606) are both still in use, although suburb-an is more than a hundred times the more frequent. The two suffixes differ in terms of stress. Whereas -(a)n generally does not affect stress placement on the root (which is typically on its first syllable), -ian| -ean transfers stress to the syllable immediately preceding the suffix. However, we do find exceptions, with penultimate stress together with the -an suffix, in Elizabeth-an /ilizə'bi:θ-ən/ from Elizabeth /i'lizəbəθ/ and in suburb-an /su'bə:b-ən/ from suburb /'sʌbə:b/ These adjectives have had both -an and -ian forms. It is the -an form which has prevailed, but with stress placement which would be appropriate for the -ian alternative. There is a degree of variation in the use of -(a)n and -ian. Indeed, it can be a matter of whim, and of altercation. Mencken (1948) reports that ‘in Atlanta, Ga., for example, a battle rages between Atlantan and Atlantian, the former being favoured by the Atlanta Constitution and the latter by the Atlanta Journal.’ 8.3.11 Derivations from names of people 8.3.11a It is the custom in Britain to refer to particular styles of architecture, furniture, customs, and the like, through the name of the monarch reigning at that time. In a short story ‘The three Garridebs’, Sherlock Holmes examines a house and says: ‘I was wondering if this was Queen Anne or Georgian.’ The interesting point is that he uses an adjective, Georg-ian, based on the name of King George, but Queen Anne has to take on adjectival role all by itself, un-affixed. Why is there no Ann-ian? When one runs through the names of rulers of England since 1500 we find Mari-an, Elizabeth-an, Georg-ian, Edward-ian, Victoria-n. The Latinate Jacob-ean refers to James I (James-ian is in use, but only relating to the style of—or a follower of—philosopher William James or novelist Henry James). However, there are no adjectives based on Henry, Charles, William, or Anne.

8.3.11 derivations from names of people   277 8.3.11b The two common ways of creating an adjective from the name of a person are by adding -ic or -ian| ean (or -an). Affixes used with the names of ancient personages reflect the original forms in the classical languages— Aristotel-ian from Greek Aristotéleios, Latin Aristolele¯us, Socrat-ic from Greek So¯kraticós, Latin So¯craticus, and similarly for Euclid-ian, Sophoclean, Homer-ic, Platon-ic, and others. From the eighteenth century, the suffixes were added to the names of contemporary or recently dead people. For example, the first instances of both Milton-ic and Milton-ian are from 1708. Each of these is still in use, with roughly equal frequency. An important difference is that an -ian derivation can be equally used as an adjective (Milton-ian verse) or as a noun (a Miltonian is an admirer of the poet), whereas the -ic derivation is predominantly an adjective (the Milton-ic epic), only rather rarely a noun. We also find -ite added to just a few people names; this generally functions as a noun (the Trotsky-ites), but can be used adjectivally (a Trotsky-ite idea). The questions then to ask are: which people have had their names adjectivalized, and which suffix (or both) has been used. The answer to the first question is that only particularly notable people were accorded this derivation. 8.3.11c We can deal first with monarchs and the like. Queen Mary I (1553– 8), a Catholic, was known as ‘Bloody Mary’ for her cruel execution of the Protestant followers of her late father, Henry VIII. The adjective Mari-an was first recorded in 1579, in such collocations as Mari-an persecution and Marian martyrs. The adjective was also used, but only from 1868, in relation to Mary, Queen of Scots (1542–87), as in the Mari-an conspiracy, and also, from 1701, for the Virgin Mary, as in Mari-an theology. Bloody Mary’s sister, Queen Elizabeth I (1558–1603) did not merit an adjective until 1807 (Elizabeth-an age| style). Derivations contemporary with a monarch’s rule commenced with Georg-ian (1745), for styles associated with Kings George I (1714–27) and George II (1727–60), then extended to later Georges. Victoria-n came in 1839, two years after the Queen’s accession. The business of making adjectives from people’s names then embraced some long-dead kings. From 1762 there is Alfred-ian, relating to the AngloSaxon King Alfred the Great (871–99) and, from 1861, Edward-ian, harking back to the first three King Edwards (1272–1377). But nothing for any of the Williams or Henrys. It seems that kings and queens uniformly merited -ian or -an. As already mentioned there are both Milton-ic and Milton-ian. Byron-ic (from 1823) is

278    8 making new adjectives today far more common than Byron-ian (1822). Nelson-ian (first instance 1809) and Nelson-ic (1846) are both in use today. These are unusual. Most people’s names occur with just one adjectivalizing suffix. For example, Shakspear-ean, Darwin-ian and Napoleon-ic (not *Shakespear-ic, *Darwin-ic, *Napolean-ian). In some instances this may be just the way things happened (and it could have turned out the other way round). In others there may be a phonological reason. 8.3.11d It appears that there is a preference that suffix /-ən/ or /-iən/ should not be added to a noun ending in /ən/ or /əm/, to avoid two unstressed syllables of similar form. For that reason *Napolean-(i)an /'na'poulyən-(i)ən/ and *William-(i)an /'wilyəm-(i)ən/ are not acceptable, Napoleon-ic is used in the first instance, and nothing at all in the second. (If the suffix were -an /-ən/ stress would be unchanged; if it were -ian /-iən/, primary stress should be on the syllable before the suffix, which would have to strengthen.) However, this can only be a preference, not a rule—from Johnson /'dʒɔnsən/ is derived Johnson-ian /dʒɔn'soun-iən/. The lack of any adjective relating to good Queen Anne or to bad King John may also, at least in part, be due the un-euphonious character of the putative derivations *Ann-(i)an /'an-(i)ən/ and *John-ian /'dʒɔn-(i)ən/. However, the OED reports that John-ian has been used with a quite different meaning, relating to St John’s College at Cambridge (but not, apparently, to the college of the same name at Oxford). All this shows that the making of adjectives from the names of people is not upon any really principled basis. Chapman (1934) includes useful information on adjectives derived from proper names, particularly with respect to their Greek and Latin origins. Note that there are a number of rather esoteric Latinate forms which have limited meaning and use, such as Caroline, Carolingian (relating to both Charlemagne and Charles I and II of England), Henrician (relating to Henry VIII and to various European kings), Jacobite (an adherent of James II after his abdication) and the opposing Williamite (a supporter of William of Orange).

8.4 Adjectives derived from verbs 8.4.1 Syntactic orientation A transitive verb has two core arguments—A (transitive subject) and O (transitive object)—while an intransitive verb has a single core argument,

8.4.1 syntactic orientation   279 S (intransitive subject). An adjective derived from a verb generally relates to one of the verb’s core arguments. Consider derivational suffix -some in the following examples. (1) transitive verb tire [Filling in the form]A tired JohnO

adjective with A-orientation, tire-some [Filling in the form]CS was [a tiresome task]CC for John

(2) transitive verb loathe MaryA loathes leechesO adjective with O-orientation, loath-some MaryA considers leechesO to be [loathsome creatures] (3) intransitive verb meddle JohnS is always meddling adjective with S-orientation, meddle-some JohnCS is [a meddlesome character]CC Some verbs derive adjectives with all three orientations. Others are restricted to just A and S, or just S and O, or just O, or just S. This is exemplified in Table 8.7 for the main affixes which derive adjectives from verbs (together with their reference numbers used throughout this chapter). Table 8.7  Syntactic orientation of adjectives derived from verbs ARGUMENT

A

S

O

(14) -less

harm-less

rest-less

count-less

(11) -ful

watch-ful

boast-ful

use-ful

(12) -some

tire-some

meddle-some

loath-some

(5)

-y

choose-y

sleep-y

bend-y

(29) -(at)ive

abus-ive

explos-ive

—

(30) -(at)ory

condemn-atory

migrat-ory

—

(31) -ant, -ent

defi-ant

exist-ent

—

(15) -able

—

change-able

accept-able

(3)

—

snapp-ish

tickle-ish

(17) -proof

—

shatter-proof

tamper-proof

(32) -worthy

—

—

trust-worthy

(23) a-

—

a-blaze

—

-ish

280    8 making new adjectives 8.4.2 Adjectives derived from syntactic types of verbs 8.4.2a The major semantic types of nouns in English were set out in 8.2, followed in Table 8.2 by examples of derivations from each type, for a number of derivational suffixes. A detailed list of semantic types of verbs in English is in the Appendix. Table 8.8 shows, for the four suffixes which derive the most adjectives from verbs, how these are used with verbs from the various semantic types. Table 8.8  Occurrence of some adjective-deriving suffixes with semantic types of verbs Semantic type

(14) -less

(11) -ful

(29) -(at)ive

motion rest

(15) -able, -ible climb-able transport-able

rest-less

rest-ful

install-able catch-able

affect

explos-ive extens-ive

hitt-able break-able

giving

contribut-ative possess-ive

market-able bribe-able

corporeal

harm-less

harm-ful mourn-ful

drink-able kiss-able

competition

compet-itive

winn-able beat-able

social contract

dismiss-ive

appoint-able

manipulat-ive

us-able wear-able

perform-ative

process-able

using

use-less

use-ful

obeying attention

watch-ful

percep-tive investigat-ive

observ-able recogniz-able

thinking

forget-ful doubt-ful

imagin-ative speculat-ive

forget-able imagin-able

decis-ive select-ive

un-decid-able

talk-ative descrip-tive persuas-ive

in-describ-able forgiv-able persuad-able

deciding speaking

thank-less

thank-ful boast-ful

liking

fear-less relent-less

fear-ful dread-ful

like-able enjoy-able

8.4.2 adjectives derived from syntactic types of verbs   281 Table 8.8 (continued) Semantic type

(14) -less

(11) -ful

(29) -(at)ive

(15) -able, -ible

annoying

tire-less

delight-ful

offens-ive impress-ive distract-ive

excit-able distract-able

imitat-ative

imit-able

acting happening

change-less

comparing

count-less

change-able vari-able compar-ative

count-able compar-able

relating

indic-ative

relat-able

beginning

continu-ative

complete-able

trying wanting

attempt-able hope-less need-less

hope-ful

expect-ative

postponing

un-avoid-able

making helping

desir-able

help-less

help-ful

permiss-ive

allow-able permiss-ible

cooperat-ive

in-support-able help-able

Some forms do double duty as verb and as noun. In a number of instances, it is debatable what the primary word class membership is (for example, harm and fear). Nevertheless, it is instructive to consider the syntactic character of the derived adjective with respect to that of the underlying root as verb. The twelve affixes creating adjectives from verbs, shown in Table 8.7, include eight of those listed in 8.2, which also derive adjectives from nouns and adjectives. Seven of these are Germanic: -ish, -less, -ful, -some, -y, -proof and a-; the other is the now fully-assimilated Romance-origin suffix -able. The remainder are used exclusively (or almost exclusively) with verbs. They comprise three Romance suffixes -(at)ive, -(at)ory, -ent| -ant, and Germanic -worthy. 8.4.2b There are other suffixes which play a small role in deriving adjectives from verbs. We find horrif-ic, relating to horr-ify. There is also beautif-ic, with a meaning rather different from beautify, and terrific, which has moved

282    8 making new adjectives right away from the meaning of terrify. The suffix -ity was used a good deal to derive adjectives in days gone by, and is retained in itch-ity. 8.4.2c Regular verbal participles may be used as modifying adjectives— those in -ing have A-orientation and those in -ed (or in -en) O-orientation. Examples are: [An interest-ing problem]Ainterests [an interest-ed scholar]O, [A trust-ing person]Atrusts [their trust-ed friend]O, [A bor-ing meeting]Abores a [bor-ed executive]O. Of course, felicity of style normally precludes a verb and two participial forms derived from it being included in the same sentence, but these do illustrate the semantic and grammatical relationships involved. 8.4.2d We first discuss the four suffixes of major use, illustrated in Table 8.8: -less, -ful, -(at)ive, and -able, also noting contrastive use of these suffixes with the same verb root. The eight more minor affixes are dealt with last.

8.4.3 (14) -less with verbs When added to nouns (8.2.29, 8.2.31), -free indicates lack of something considered undesirable (for example, trouble-free) and -less the lack of something which one would expect—and hope—to have (  joy-less). Only -less is used with verb roots and here covers both senses, just meaning ‘not having a potentiality to do or feel something’. For example, harm-less, which is a good thing since one doesn’t want to be harmed, and help-less which describes wanting to help and not being able to, something that is not a good thing. Adjectives derived with -less include a number with each of the three syntactic orientations. For example: A-type O-type S-type

harm-less, fear-less, help-less count-less, tire-less, use-less, need-less rest-less, change-less, relent-less

That is, someone who is fear-less is unlikely to fear anything; someone is tire-less if scarcely anything tires them; something is change-less if it seldom changes. There are not a large number of adjectives derived from verbs with -less; but, as can be seen from Table 8.8, they span a fair number of semantic types. The suffix appears to be fairly productive with verbs, as with nouns.

8.4.5 contrasting - ful and - less with verbs   283 8.4.4 (11) -ful with verbs Like -less, this yields adjectives of all three syntactic types. They include: A-type O-type S-type

harm-ful, watch-ful, forget-ful, doubt-ful, resent-ful, delight-ful, hope-ful, help-ful use-ful, dread-ful mourn-ful, boast-ful

The meanings of the A- and S-type derivations is ‘likely to do’; for example ‘likely to forget’, ‘likely to boast’. For the O-type it is ‘likely to be verb-ed’. Something which is use-ful will be of use; anything dread-ful (in the original meaning of this word) is likely to be dreaded. Rest-ful is a little different, relating not to the core argument of verb rest (as rest-less does) but to the referent of a peripheral argument. An ambience is rest-ful if it is conducive to properly resting; for example We had a really restful time at the coast last weekend. Plag (1999: 79) suggests that -ful has only been added to monosyllabic verbs, or to disyllabics with final stress. The suffix appears not to be productive with verbs, as it is not with nouns (8.2.22).

8.4.5 Contrasting -ful and -less with verbs As can be seen from Table 8.8, there is some overlap in the semantic types of verbs that -less and -ful are used with. There are some instances where the two derivations have the same semantic orientation. For example:

both A-type

harm-ful harm-less

‘liable to harm’ ‘not capable of harming’

help-ful help-less

‘able and willing to help’ ‘unable to help (but would like to)’

(A further sense of help-less is ‘unable to help oneself, weak’.) It is instructive to compare help-less with: A-type

un-help-ful ‘won’t help (but could)’

284    8 making new adjectives We also find: both O-type

use-ful use-less

‘can be used a lot’ ‘cannot be used’

For others the orientation differs. A person can be thank-ful (A-type), and a task undertaken can be thank-less (O-type). For example: John was thank-ful that Mary had painted the wall (that is, he thanked her) but Mary found painting the wall a thank-less task (nobody thanked her for it). Un-thank-ful means ‘doesn’t thank people when they should’ whereas thank-less is ‘something one doesn’t get thanked for doing when one should be’. Hope-ful and hopeless pattern similarly. And a person may be rest-less ‘not at rest when they should be’ whereas, as explained just above, a place or a time may be rest-ful.

8.4.6 (29) -ive, -tive, -ative, -itive 8.4.6a This suffix can be written as -ive, -tive, -ative, or -itive and pronounced as /-iv/, /-tiv/, /-ətiv/, or /-itiv/. It goes back to the Latin suffix -ı¯vus, which was added to participial stems, such as demonstra¯t-ı¯vus ‘demonstrative’. Loans into ME from French included furtive and pensive, which are not analysable in English, and also a goodly number of verb/adjective pairs. Reflecting the original Latin forms, there are some consonantal alternations; for example, s replacing d or t, as in extend| extens-ive, permit| permiss-ive. The suffix may have form -ive, or -tive (as in describe| descrip-tive, where the original b assimilates in voicing to the t of -tive), or -ative (imagin-ative) or -itive (compet-itive). There may be vowel change, as in compete /kəm'pi:t/, compet-itive /kəm'pet-itiv/. Verbs which form an adjective by adding -(at)ive typically also form a noun with -((a)t)ion (see 9.4.7 for the phonology and meaning of suffix -((a)t)ion); there are similar consonant and vowel changes associated with the two suffixes. However, some verbs show only one of these derivations. For example: verb extend perceive

derived noun extens-ion percep-tion

derived adjective extens-ive percep-tive and very many more

organize

organiz-ation

—

and quite a few more

offend

—

offens-ive

and a few more

8.4.7 (15) - able, -ible with verbs   285 Some verbs which derive adjectives with -(at)ive form nouns in other ways; for example, dismiss| dismiss-al| dismiss-ive and perform| perform-ance| perform-ative. 8.4.6b The meaning of -(at)ive is ‘liable to do, can do’; for example offensive ‘liable to offend’. Almost all adjectives derived with this suffix have ­A-orientation. As shown in Table 8.8, the suffix is found in words from most semantic types, the main exceptions being motion, rest, corporeal, and liking. Adjectives formed with -(at)ive can describe either an agent, or something they do or have; for example, impress-ive person| teacher or ­impress-ive lecture| idea. There are just a few intransitive verbs which accept the suffix. The resulting adjectives have S-orientation; for example extens-ive ‘extends a far way’ and explos-ive ‘liable to explode’. A number of -(at)ive derivations, predominantly based on Speaking verbs, are most often used to describe a speech act, rather than the person uttering it; for example, descrip-tive title, declar-ative statement, suggest-ive remark. 8.4.6c This suffix is used with many more verbs than are -less and -ful; only a sample are included in Table 8.8. Most are of Romance origin (generally, having two or more syllables), but there are some of Germanic origin, such as talk-ative. There are just a few nouns forming adjectives with -(at)ive, including mass-ive and argument-ative. (Historically, secret-ive is formed on verb secrete, rather than on noun secret.) The suffix is scarcely productive. However, we do find stick-to-it-ive in colloquial American English (from about 1870). Unlike (15) -able, discussed just below, the derivational suffix -(at)ive is generally not used with phrasal verbs.

8.4.7 (15) -able, -ible with verbs 8.4.7a The basic form of this suffix is -able, with variant -ible, both pronounced as /-əbl/. As described in 8.2.36, it relates to Latin -a¯bilis. There were many loans from French into ME, and the suffix was soon added, within English, to further Romance roots and also to many Germanic ones. It is fairly productive with nouns and by far the most productive suffix deriving adjectives from verbs. As shown in Table 8.8, it can be added to verbs from every semantic type.

286    8 making new adjectives The semantic orientation of -able derivations is almost exclusively towards the O of an underlying transitive verb, with the meaning ‘can be verb-ed’, ‘suitable to be verb-ed’, or ‘likely to be verb-ed’, as in climb-able, wear-able, and forget-able, respectively. There are just a few derivations with S-orientation. Vary is an ambitransitive verb of type S = O—for example [The athlete]Avaried [his training schedule]O and [The wind speed]S varies— but the derived adjective vari-able only has S-orientation—as in The wind speed is vari-able. For other S = O ambitransitive verbs, the derived adjective can have either O- or S-orientation: relating to The boss is likely to change [the date of the meeting]O there is The date of the meeting is change-able, and corresponding to [The weather]S may change we can get The weather is change-able. (Similarly for break.) 8.4.7b This suffix very often co-occurs with negative prefix un- (or in-). In fact, some adjectival derivations are more used with un- than without it (for example, in-describ-able and in-support-able were included in Table 8.8). In a number of cases the form with un- came into the language first. The OED gives 1430 for un-think-able but 1854 for think-able; similarly for derivations on speak, know, and bear (see 5.2b and Marchand 1969: 230). There are some forms which go back to Latin, often with the suffix being -ible rather than -able, and without the non-derived verb root persisting into NE. For many of them, there are now derivations on Germanic verbs which are taking over. Compare combustible and burn-able, legible and read-able, intelligible and understand-able (for others see Jespersen 1942: 403). In 5.4c we contrasted in-ed-ible and un-eat-able, in-evit-able and un-avoid-able. 8.4.7c The productivity of -able is shown by the fact that it can be added to a phrasal verb, which may consist of simple verb and one or more prepositions (see 9.3.1f, and also the discussion in Dixon 2005a: 293–7). Marchand (1969: 230–1) and Jespersen (1942: 400–1) show that, since the sixteenth century, it has been permissible to add -able to the end of the whole complex, as in That salary is very live-off-able. More recently, it has become acceptable to add -able after each component, as in That salary is very live-able off-able. Other examples include:

The meeting is This government is

older version put-off-able bring-down-able

newer alternative put-able off-able bring-able down-able

8.4.9 contrasting - less and - able with verbs   287 However, when negative prefix un-2 is added, it appears that the preference is for -able to be appended to the simple verb component, as in This meeting is un-put-able off (certainly, one wouldn’t say *un-put-able off-able). If there are two prepositions in a phrasal verb, -able is added after the first, and optionally also after the simple verb component: She’s very fall-out-able with/fall-able out-able with Your inheritance is fall-back-able on/fall-able back-able on He’s entirely get-on-able with/get-able on-able with 8.4.8 Contrasting -(at)ive and -able with verbs When added to transitive verbs, -(at)ive derives adjectives with ­A-orientation and -able adjectives with O-orientation. Since -able applies to the great majority of transitive verbs, and -(at)ive to a considerable number (mostly of Romance origin), there are many instances of the two suffixes applying to the same verb, with a difference of syntactic orientation (and sometimes also variation in sense of the verb). If we hear John persuaded Mary then it is likely that John is persuas-ive or Mary is persuad-able, or both. Similarly for The noise distracted Fred—the noise may be distract-ive and/or Fred may be easily distract-ible. Among very many other examples are: decis-ive| decid-able, offens-ive| offend-able, permiss-ive| permiss-ible. A more subtle difference occurs in: The garden is extens-ive—that is, it extends a long way (S-orientation) The garden is extend-able—that is, it can be extended (O-orientation) These derivations relate to the intransitive and transitive senses, respectively, of the S = O type ambitransitive verb extend.

8.4.9 Contrasting -less and -able with verbs Pairs of adjectives derived with these suffixes include: O-type

count-less ‘so many it is impossible to count’ ­ count-able ‘such a number that they can be counted’

S-type

change-less ‘scarcely changes’

S-type or O-type

change-able ‘liable to change’ or ‘liable to be changed’

288    8 making new adjectives Note that un-count-able has a very similar meaning to count-less, as does un-change-able to change-less.

8.4.10 Contrasting -ful and -able, and -less, with verbs The following derivations have different syntactic orientations: A-type

forget-ful ‘liable to forget’

O-type

forget-able ‘liable to be forgotten’

As mentioned above, help occurs with -less and -ful; and also with -able: A-type

help-ful ‘able and willing to help’

A-type

help-less ‘unable to help (but would like to)’

O-type

help-able ‘can be helped’

We can now examine the eight suffixes which also derive adjectives, but from a smaller number of verbs. 8.4.11 (12) -some with verbs Germanic suffix /-səm/, /-sm̩/ (see 8.2.25) is no longer productive and occurs with rather few verbs. However, the verbs that do take it, and their syntactic orientations, are of especial interest. The meaning is ‘apt to (cause)’. The liking and annoying semantic types of verbs involve the same semantic roles, but mapped onto syntactic functions in different ways. For the liking type the Experiencer is in A function and Stimulus in O, with this being reversed for annoying verbs. For example: JohnEXPERIENCER:A likes [winning prizes]STIMULUS:O [Winning prizes]STIMULUS:A pleases JohnEXPERIENCER:O The suffix -some occurs with a couple of liking verbs and with half-a-dozen from the annoying type. All of the derived adjectives carry a negative meaning, and all are oriented towards the Stimulus role. That is, we get: O-type with liking verbs: loathe and fear A-type with annoying verbs: weary, tire, trouble, irk, bother, worry

8.4.13 (5) - y with verbs   289 That is, if John fears X, then X is likely to be fear-some. And if Y irks Fred, then Y could be described as irk-some. The suffix occurs with just a few other verbs, from different semantic types. We find quarrel-some, meddle-some (both S-type) and venture-some (A-type); they mean ‘likely to do’. Rather out on its own, there is the O-type derivation (with a most definitely positive sense) cuddle-some ‘suitable and nice to be cuddled’. As mentioned in 8.2.25, a number of the roots to which -some is added may function as noun or as verb; these include fear, trouble, and venture (plus cuddle). The adjective could be derived from either.

8.4.12 Contrasting -less, -ful and -some with verbs The root fear takes all three suffixes:

A-type

fear-ful

A-type O-type

fear-less fear-some

‘likely to fear’ (may also mean ‘likely to inspire fear’) ‘unlikely to fear’ ‘likely to be feared’

One can say: The fear-ful person ran away from the fear-some monster, or The fear-less person attacked the fear-some monster. Tire takes just -less and -some. Note that the syntactic orientations are the opposite of those for fear: O-type A-type

tire-less tire-some

‘does not get tired easily’ ‘behaves in such a way that it tires others’

8.4.13 (5) -y with verbs Germanic suffix -y /-i/ is productive with nouns (8.2.8), and also with verbs of certain semantic types. All the verbs to which it is added end in a consonant and almost all are monosyllabic (exceptions include fidget-y and shiver-y). They are predominantly intransitive, yielding S-orientation adjectives with the meaning ‘liable to’. Most of the adjectives have unpleasant overtones (tast-y is one of the few derivations with a positive sense). The semantic types involved are:

290    8 making new adjectives —motion: runn-y (as of a nose), fidget-y, wriggl-y, shak-y —affect (stretch subtype): bend-y, curl-y, twist-y, stretch-y (break subtype): crumbl-y, squash-y —corporeal: sleep-y, drows-y, dream-y, weep-y, sulk-y, wheez-y, shiver-y, smell-y, tast-y —speaking: squeak-y, creak-y, croak-y Some of the verbs are S = O type ambitransitives; bend-y could be ‘bends easily’ (S-orientation) or ‘can be bent easily’ (O-orientation). A few are essentially transitive verbs, and their adjectival derivations thus have ­O-orientation; these include shak-y, squash-y, and tast-y. The verb choose, from the deciding type, forms choos-y; this is of ­A-orientation and does have negative overtones, describing someone being more particular than is generally considered appropriate about what they choose. 8.4.14 (30) -(at)ory 8.4.14a Suffix -ory, -tory or -atory, /-əri/, /-təri/, /-ətəri/, appears almost exclusively within Romance loans. There is often (but not always) a corresponding verb, and a nominalization in -((a)t)ion. It is interesting to note the first dates of attestation in English for verb, adjective, and noun in each row of the following paradigm (which is just a sample of the whole corpus):

(a) (b) (c) (d) (e)

verb introduce (1475) contribute (1530) congratulate (1548) condemn (c1340) migrate (1623)

adjective introduc-tory (1400) contribut-ory (1467) congratulat-ory (1523) condemn-atory (1570) migrat-ory (1708)

noun introduc-tion (c1386) contribut-ion (1387) congratulat-ion (1591) condemn-ation (1382) migrat-ion (1527)

It can be seen that the words in each row came into the language in varying order, eventually completing the paradigm, which mirrors forms in Latin and French. The order of acceptance was Noun-Adjective-Verb for

8.4.14 (30) -(at)ory   291 rows (a–b), Adjective-Verb-Noun for (c), Verb-Noun-Adjective for (d), and Noun-Verb-Adjective for (e). Note that for some adjectives there is only a partial paradigm. For example, corresponding to noun illus-ion and adjective illus-ory, there is no corresponding verb. And there are a fair number of adjectives ending in -ory for which no verb (with the appropriate meaning) without the suffix exists; for example, auditory and perfunctory. Within a synchronic description of the modern language, adjectives in the middle column can be regarded as derived from forms in the verb column. In some instances the stressed vowel changes value, as in introduce /ˌintrə'dyus/, introduc-tory /ˌintrə'dʌk-təri/ and inflame /in'fleim/, inflamm-atory /ˌin'flam-ətəri/. This may also be stress shift, as in migrate /mai'greit/, m ­ igrat-ory /'meigrət-əri/ (an alternative is the straightforward /mai'greit-əri/).

8.4.14b The suffix -(at)ory is used mainly with verbs from a small number of semantic types. All derived adjectives are of A- or S-orientation, with the meaning ‘does’, or ‘tending to do’: —motion (all S-type): migrat-ory, transit-ory —speaking (A-type): i ntroduc-tory, congratulat-ory, condemn-atory, confirm-atory, accus-atory, contradict-ory, (S-type): ejaculat-ory, expostulat-ory —giving (A-type): contribut-ory, compensat-ory —annoying (A-type): satisfact-ory —happening (S-type): circulat-ory Plus compuls-ory from the making type and inflamm-atory from corporeal. The suffix is not nowadays used to create new adjectives. As described in 9.6.1, suffix -(at)ory also derives location nouns from verbs; for example deposit-ory ‘place where one deposits things’. This occurs in Latinate forms for which there is no underlying root in English; for example, dormitory.

292    8 making new adjectives In British English, suffix (30) -ory (from Latin -orius)—which can be regarded as added to verbs—is homonymous with (16) -ary (from Latin -arius)—which is generally added to nouns (see 8.2.39); both are pronounced /-əri/. However, in American English the two endings may have different pronunciation in certain phonological environments— see Marchand (1969: 338).

8.4.15 A-orientation adjectives derived from annoying verbs Verbs from the annoying type derive adjectives with A-orientation in a number of different ways: with -ful: with -ive: with -some: with -y: with -(at)ory:

delight-ful offens-ive, impress-ive, distract-ive weari-some, tire-some, trouble-some, irk-some, bother-some, worri-some scar-y, trick-y satisfac-tory

The remainder use participle -ing. For example: The terrify-ing monster terrified John, The surpris-ing solution surprised Mary, The upsett-ing events upset Tom, The pleas-ing result pleased Jim. There is no strong correlation between historical origin of verb and of suffix. Delight is a Romance root taking a Germanic suffix, -ful. All of offend, impress, and distract are Romance, adding the Romance suffix -ive. Most verbs taking Germanic suffix -some (all with negative overtones) are Germanic, but trouble is of French origin. Of those taking -y, scare is Germanic but trick is Romance. In fact, the majority of verbs in the annoying semantic type are of Romance origin, including most of those which take the -ing suffix (terrify, surprise, please and many more); however, upset is Germanic.

8.4.16 (31) -ent, -ant 8.4.16a The suffix which is written as -ent or -ant, and pronounced /-ənt/, serves to derive both adjectives and nouns from verbs. It comes from Latin -ant(em) and -ent(em), through French -ant. The derived nouns are discussed in 9.3.4. Derived adjectives tend to use -ent and derived nouns -ant, but there is a good deal of mixing, as shown by:

8.4.17 (3) - ish with verbs   293

-ant

-ent

adjectives resist-ant observ-ant expect-ant emerg-ent indulg-ent diverg-ent

nouns applic-ant claim-ant inhabit-ant oppon-ent stud-ent preced-ent

Verbs depend and descend each have an -ent and an -ant derivation. In his 1755 dictionary, Samuel Johnson identified descend-ant as a noun (‘the offspring of an ancestor’) and descend-ent as an adjective (‘falling, sinking, coming down, descending’), adding: ‘It seems to be established that the substantive [noun] should derive the termination from the French, and the adjective from the Latin.’ However, this distinction does not prevail today— the noun is generally written as descend-ant but sometimes as descend-ent, and the other way round for the adjective. Similarly for depend-ent and ­depend-ant, each as both adjective and noun. This is, of course, simply a matter of spelling, both variants being pronounced the same, as /-ənt/. Some borrowed adjectives in -ent| -ant did not bring with them a corresponding verb and are not analysable in English; for example violent and elegant. However, there are many verb/adjective pairs. The suffix is used only with Romance forms and is not now productive. 8.4.16b Adjectives derived with -end| -ant indicate ‘having the quality of verb-ing’. Some are of A-orientation—defi-ant, triumph-ant—but most have S-orientation—repent-ent, diverg-ent, recurr-ent. A number show phonological differences from the underlying verb, reflecting their Romance origins; for example, abound /ə'baund/ and abund-ant /ə'bʌnd-ənt/, prevail /prə'veil/ and preval-ent /'prevəl-ənt/.

8.4.17 (3) -ish with verbs The Germanic suffix -ish /-iʃ/ is used with many common nouns (8.2.4) and with some proper nouns (8.3.2). It also occurs with a small number of Germanic verbs. The O-orientation derivation tickl-ish is used of someone who has an extreme reaction to being tickled; it is also used to describe a tricky situation or problem (and here the verb tickle could not be employed). We

294    8 making new adjectives also encounter the S-orientation derived adjectives snapp-ish ‘liable to snap’ and peck-ish ‘feeling slightly hungry’ (with a rather tenuous semantic connection to the verb peck at (one’s food)—as a bird does—referring to nibbling at bits of a meal).

8.4.18 (17) -proof with verbs As discussed in 8.2.40, the Germanic suffix -proof /-ˌpruf/ (which always bears secondary stress) is used with a number of nouns referring to unwelcome things and then has the meaning ‘keep out’. It is also used with a small set of verbs referring to unwanted happenings and means ‘to stop [something] happening’. Some of the verbs are S = O type ambitransitives, and the derived adjective may have either S- or O-orientation. Something may be shatter-proof if it is either ‘unlikely to shatter’ (S-orientation) or ‘such that it cannot be shattered’ (O-orientation) or both. Similarly for tear-proof. We can recognize a transitive prepositional verb tamper with, and adjective tamper-proof is an O-orientation derivation ‘cannot be tampered with’.

8.4.19 (32) -worthy The Germanic suffix -worthy /-wə:ði/ always takes secondary stress. It originated as the second element of compounds in ME times, and thence into a suffix. It is added to a number of verbs from the speaking type (praise, blame, note, quote), and to trust from the liking type. The suffix is historically related to noun worth and has the meaning ‘be appropriate to be verb-ed’. For example, someone is trust-worthy if it is appropriate to trust them (that is, if they are worth trusting). The few adjectives formed from verbs with -worth all have O-orientation. The suffix also derives adjectives describing the appropriateness of a vehicle for travel in a certain medium, from a noun describing that medium: air-worthy, sea-worthy, road-worthy, and so on. Note-worthy and quote-worthy describe something which it is appropriate to note or quote. However, note and quote also function as nouns. Newsworthy ‘appropriate to be an item of news’ was coined (in the 1930s), undoubtedly by analogy with note-worthy and quote-worthy, even though news is only a noun (never a verb).

8.4.20 (23) a - with verbs   295 As Marchand (1969: 357) points out, the fact that -worthy is a bona fide suffix (rather than the second element in a series of compounds) is shown by its co-occurrence with negative prefix un-, as in un-trust-worthy, un-seaworthy, and un-news-worthy. (As mentioned in 3.3b, compounds only rather seldom accept un-.) The suffix does have a limited productivity with verbs and nouns of a certain type. Indeed, it may be added to a conjunction, such as He’s both [trust and praise]-worthy, or to a complex noun phrase, as in This new four-wheeldrive vehicle is really [rough dirt track]-worthy. Credit-worthy is an old formation (the OED provides quotes from 1562 and 1639) meaning ‘worthy of belief or trust’; for example, Reported by credit-worthy writers. A quite new derivation came into use in the middle of the twentieth century, credit-worthy meaning ‘financially responsible person to whom it is appropriate to extend credit (the favour of allowing to owe money)’. And one might hear She’s both [trust and credit]-worthy, adding the suffix to a coordination of trust (which can be a noun or a verb) and credit (here only a noun).

8.4.20 (23) a- with verbs As described in 8.2.46, prefix a- /ə-/ goes back to preposition on ~ an ‘on’ in OE, and modern adjectival derivations in a- (whether based on noun or on verb) have the same functional role as a preposition followed by a noun phrase— that is, as copula complement. One can say The house is a-blaze, but not *The a-blaze house (only The blaz-ing house, employing a verbal participle). A limited number of intransitive verbs form an adjective with a-, showing the meaning ‘being in a state of’. For example, a-glow, a-boil, a-sleep, a-float, a-quiver (as in Every nerve in my body was a-quiver). The prefix is productive in an ad hoc manner, for jocular effect. One might hear: He was a-singing and a-dancing and a-jumping up and down. Or even, employing an old grammatical device to describe new-fangled activities: He’s a-facebook-ing and a-twitter-ing. We now turn, in the next chapter, to the heady topic of nominalization— creating nouns from verbs, from adjectives, and from other nouns.

9 Eager-ness, bidd-ing, pay-ment, owner-ship, satisf-action: Making new nouns

9.1 Introduction 296 9.2 Descriptions of states: nouns from adjectives  298 9.3 People, agentive and otherwise: nouns from verbs and nouns  306

9.5 Abstract nouns from concrete nouns 347 9.6 More on location  360 9.7 Bits and bobs: -(i)ana, -ade, -ometer, -in 362

9.4 States, units, activities, results, objects, and locations: nouns from verbs  329

9.1 Introduction 9.1a This chapter deals with nominalizations—suffixes which create nouns from verbs (for instance, observ-er from observe, donat-ion from donate) and from adjectives (eager-ness from eager, wis-dom from wise). And also, suffixes which may be added to a noun and produce a new noun of quite different type, such as vicar-age, description of a residence, from vicar, label for a person’s occupation, and curv-ature, description of a property, from curve, name for a geometrical shape. Chapter 6 ‘Keeping the same word class’ was concerned with prefixes which may be added to a verb, adjective, or noun, and derive a verb, adjective, or noun of the same type. For example, re-build is a transitive verb

9.1 introduction   297 of the affect type, just like build; extra-wide is a dimension adjective, just like wide; and arch-villain is a description of a type of person, just like villain. With very few exceptions, derivational processes which form verb from verb or adjective from adjective do not affect syntactic status or type of meaning. This also applies to the ‘noun from noun’ prefixes of Chapter 6. In contrast, the suffixes discussed in this chapter, when they apply to nouns, always create a new noun of a different subclass from the base form. For example:

• • •

Art-ist and cell-ist are practitioners of the discipline of art and of the cello, a musical instrument, respectively. Darwin-ism describes the evolutionary doctrine proposed by Charles Darwin. Mother-hood describes the state associated with the kinship designation mother.

9.1b Nominalizing suffixes can be roughly divided into four sets (with a certain amount of overlap between them). Section 9.2 describes suffixes which derive a noun referring to a state from an adjective describing the corresponding property; for example, stupid-ity from stupid. Section 9.3 concentrates on suffixes which create descriptions of people and of things, both from verbs (bak-er, lubric-ant) and from nouns (mountain-eer). ­Section 9.4 surveys a variety of suffixes which apply largely to verbs and form nouns describing states, activities, results, objects, and locations (bewilder-ment, assist-­ance, declar-ation, discover-y, enclos-ure). Section 9.5 focuses on suffixes which create abstract nouns out of concrete nouns ( friend-ship, myth-ology, tonsil-itis). Section 9.6 deals with location, while four minor nominalizing suffixes make up 9.7. As mentioned in 2.8g, there are limited possibilities for adjectives undertaking double duty as nouns, the main instance being potential (and see 9.2.1a). Most apparent examples involve omission of a head noun which will be understood from the context; for example, human (being), medical (examination), social (function). In contrast, very many verbs have secondary function as a noun, including cheat, slice, envy, bite, visit. Table 2.4, of 2.8h–j, contrasted double duty with explicit derivations—show as both verb and noun as against demonstrate/demonstrat-ion, and so on. These double duty items are included in the discussions of 9.3 and 9.4.

298    9 making new nouns 9.1c Of the suffixes described in this chapter, ten take primary stress: -eer (9.3.2), -ese (9.3.7), -aire (9.3.12a), -ee (9.3.14), -ology (9.5.8), -itis (9.5.10b), -orium and -arium (9.6.1), -ade (9.7b) and -ometer (9.7c). Secondary stress goes on -monger (9.3.13a), -maker (9.3.13d), and -in (as in sit-in, 9.7d). There are also a number of suffixes which may affect stress placement in the base: -ity and -ability (9.2.1), -ian (9.3.7), -((a)t)ion (9.4.7), -icide (9.5.10a), -(at)ory (9.6.1), and -(i)ana (9.7a). Other suffixes neither customarily take stress, nor affect stress placement in the base.

9.2 Descriptions of states: nouns from adjectives 9.2a An adjective refers to a property of the referent of a noun; for example, happy (as in the happy dog). A noun may be derived from an adjective and it describes the state of having that property; for example, happi-ness. If someone says I knew happi-ness only once in my life, this is equivalent to I was in a state of being happy for only one period in my life. Nominalizations of adjectives, describing states, are one variety of what are called ‘abstract nouns’ since they describe abstract (that is, non-­concrete) states. 9.2b For most pairings of state noun with property adjective, it is the adjective which is in root form, and the noun is derived from it by one of the suffixes soon to be discussed. There are a handful of exceptions, where the noun is the basis, with an adjective derived from it by one of the suffixes discussed in Chapter 8. Thus, nouns anger and luck form adjectives with suffix -y, grace with -ous, and beauty with -ful. Some verbs from the liking and annoying types have similar semantic import to some adjectives from the human propensity type (details of types are in the Appendix). For instance, verb like (from liking), verb please (from annoying) and adjective fond (of) (from human propensity). Compare: (1) (a) (b) (2)

John likes [Mary’s red dress] John has a lik-ing for [Mary’s red dress]

(a) [Mary’s red dress] pleases John (b) John gets pleas-ure from [Mary’s red dress]

9.2.1  -ity, -ety, -ty, -city   299

(3) (a) (b)

John is fond of [Mary’s red dress] John has a fond-ness for [Mary’s red dress]

In summary, we have: verb expressing emotion like please

adjective describing property

fond (of   )

nominalization referring to state lik-ing (for) pleas-ure (from) fond-ness (for)

Other verb and adjective correspondences are: grief| grieve together with sad| sad-ness, and regret (doing double duty as verb and noun) with sorry| sorrow. A variety of nominalizing suffixes form state nouns from liking and annoying verbs, illustrated in admir-ation, satisf-action, enjoy-ment, excitement, annoy-ance, lov-ing. Each of these suffixes has a range of meanings, which are discussed together in 9.4. Among the several senses of suffix -(e)ry is one where it is added to an adjective and derives a state noun (bravery, gallant-ry, prud-ery); see (5) in 9.5.9c. 9.2c We now discuss, in turn, those affixes whose primary role is to derive state nominalizations. First, suffixes of Romance origin which are pretty well restricted to employment with Romance adjectives: -ity, -y, and -(a)cy (with their variants). Then the Germanic suffix -th, found in a handful of Germanic derivations. And then -ness, of Germanic origin but today used with every sort of adjective, and also with phrases. There has been competition between these suffixes, and this is reviewed in 9.2.6–7.

9.2.1 -ity, -ety, -ty, -city 9.2.1a This suffix has the form -ity or -ety, generally pronounced /-əti/, sometimes /-iti/. It can be just -ty /-ti/ after a consonant, as in safe-ty /'seif-ti/. Another alternative form is -city /-səti/. Stress is always on the third syllable from the end. Compare modern /'mɔdən/ with modern-ity /mɔ'də:n-iti/, and specific /'spi'sifik/ with specif-icity /spesi'f-isiti/. There is consonant alternation in brief|  brevity, reflecting the time when v (used medially) and f (finally) were variants of a single phoneme.

300    9 making new nouns This suffix emanates from Latin ending -tatem (with or without a preceding i or e). Many loans came into ME and early Modern English, either directly from Latin or through French. A few were unanalysable in English, including alacrity and felicity. There were also many adjective/noun pairs such as pure and pur-ity, cruel and cruel-ty, certain and certain-ty, infirm and infirm-ity. These were sufficient to establish -(i)ty as a nominalizing suffix in English. However, it has been applied only sparingly, and then almost exclusively to adjectives of Romance origin. It was mentioned in 2.8g that potential is one of the few adjectives doing double duty as adjective and as abstract noun. There is also potential-ity, with very similar meaning, as in This new block-buster has the potential| potentiality to make a vast profit for us. 9.2.1b Some abstract nouns in -(i)ty have a corresponding adjective ending in -ous, both being taken over from the Romance source. For example: adjective loquac-ious vorac-ious dextr-ous

abstract noun loquac-ity vorac-ity dexter-ity

9.2.1c Adjectives derived with -able or -ible (8.2.36) typically go on to create an abstract noun by adding -ity, the -able| -ible suffix then becoming -abil /-ibəl/ to accommodate the antepenultimate stress (note that this alternation goes back to Latin). For example read-abil-ity /ˌri:d-ə'bil-əti/, desir-abil-ity, revers-ibil-ity. In the discussion of the adjective-creating suffix -able, in 8.2.36, we mentioned that there is no genetic connection between suffix -able and adjective able (also a loan from French), but the coincidental similarity of form and meaning has probably re-inforced the productivity of -able as a derivational process. Suffixes -able and -ity, in sequence, have effectively fused, and the resulting -ability could in some instances now be regarded as a suffix in its own right. The coincidence of form and partial similarity of meaning with noun ability (to which it is not genetically related) has assisted this recognition. There is further discussion of this, and comparison with -able-ness, in 9.2.7.

9.2.3  -cy, -acy   301 Nominalizing -ity is also added to adjectives ending in able or ible which are not themselves analysable within English, such as plausibil-ity, culpabil-ity, possibil-ity (note that -ble is again adjusted to -bil-). However, these could not be re-analysed in terms of a suffix -ability.

9.2.2 -y There are several grammatically distinct suffixes with basic form -y /-i/. Diminutive/endearing -y (as in Ruth-y, dogg-y) was discussed in 6.11.3. ­Adjective-deriving -y (greed-y, green-y, dream-y) was the subject of 8.2.8 and 8.4.13. We are here examining the suffix -y which is added to adjectives and creates a noun describing a state. This goes back, through French -ie, to Latin -ia, and Greek suffixes -ia and -eia. ME borrowed jealous,  jealous-y, honest, and difficult, with honest-y and difficult-y following in the early sixteenth century. We can certainly recognize -y as a suffix which derives abstract nouns from adjectives in English, but it is by no means productive today. What is essentially the same suffix derives abstract nouns from verbs— enquir-y, discover-y, and so on; see 9.4.11.

9.2.3  -cy, -acy Suffix -(a)cy /-(ə)si/ has a similar origin to -y. It comes from Latin ending -ia onto a preceding c or t. Most English abstract nouns in -(a)cy relate to a Latinate adjective. First there are a number where the adjective ends in -ate and the nominalization in -acy: adjective

delic-ate intric-ate

state noun delic-acy intric-acy

Also, among others, legitimacy, intimacy, obstinacy, literacy, accuracy. There are also pairs where the adjective ends in -ant or -ent and the abstract noun in -ancy or -ency: adjective

pregn-ant consist-ent

state noun pregn-ancy consist-ency

Also buoyancy, urgency, permanency, coherency, leniency, and many more.

302    9 making new nouns The adoption of -(a)cy as a productive affix in English was a slow business. In the early sixteenth century, suprem-acy was coined on the basis of supreme (probably by analogy with delicacy, intricacy, and the like). There is an example of -(a)cy being added to a free adjective root in secret and (with a minor phonological change) secre-cy, both Romance loans into ME. The productive use of -(a)cy to create abstract nouns from adjectives came in the nineteenth century (it is said to be predominantly in the American dialect) with such forms as normal-cy. (And see 9.2.6d.) A minor feature of -(a)cy is that it is added to a few concrete nouns, deriving abstract nouns, as in chaplain-cy ‘the position or work of a chaplain’ and pap-acy ‘the office of Pope, and the administrative infrastructure associated with it’. There are also a number of abstract nouns in -acy relating to concrete nouns ending in -at, such as diplom-acy and diplomat; see (5) in 2.5k. Suffix -acy is also added to the verb conspire, deriving a result noun conspir-acy.

9.2.4 -th, -t Suffix -th /-θ/ goes back to OE nominalizing suffix -ð(o| u), which sometimes involved umlaut of the root vowel. From strang ‘strong’ was derived streng-ðu ‘streng-th’. This has given rise to modern vowel alternations as in long| leng-th, deep| dep-th. We also have, from OE, true| tru-th. Warm-th was created in ME times, and then cool-th, which has unfortunately not survived. Within early Modern English, wid-th and bread-th were created, from wide and broad. There were also some -th nominals based on verbs, such as steal-th from steal, and (in the sixteenth century) grow-th from grow. Plus others which are not today analysable, including birth, health, and wrath. A most interesting derivation was OE noun fȳl-ð ‘filth, impurity’ from adjective fu¯l ‘foul, putrid, unclean’. These developed slightly different meanings. From fu¯l came modern foul, which in ME times took on a new nominalization foul-ness. Noun filth, in turn, formed its own adjective with -y (8.2.8), filth-y. If you enter an apartment which has not been cleaned for ages, with balls of dust in the corners, you may describe it as a filth-y apartment. Foul is prototypically used of a smell, such as that of rotten eggs or a particularly

9.2.6 contrasting -ity , - y and - (a)cy with - ness   303 egregious fart. Metaphorically, people speak of foul language and filth-y language, with much the same import. We can link -th derivations with the few employing -t, such as heigh-t, which also goes back to a -ð(o| u) derivation in OE, hieh-ðo. A quite different -th suffix is used to form ordinal numbers from cardinals—seven-th, ten-th, and so on. Also see 9.4.14. 9.2.5 -ness Alongside Romance -ity, -y, and -(a)cy, there are two Germanic suffixes forming abstract nouns from adjectives. Suffix -th had limited use in OE and survives in just a handful of words today; it is no longer productive. In contrast (as mentioned in 2.4a), suffix -nes(s) was highly productive in OE and has remained so through all following periods. Germanic adjectives (other than the few with -th or -t) uniformly take -ness /-nəs/. From ME times it has been added to a selection of Romance adjectives, such as strange, round, usual, common, willing, appropriate, and correct. The applicability of -ness includes verbal past participles—such as ashamed-ness—and a wide variety of other words. For example, same-ness, otherness, belonging-ness, much-ness, and me-ness. From the sixteenth century, -ness has been added to complex forms, as in kind-heart-ed-ness and self-­ conscious-ness. More recently, we find matter-of-fact-ness, up-to-date-ness, man-of-the-world-ness, and many similar coinages. Indeed, the possibilities of use for -ness seem to be as free as one’s ability-to-dream-up-newthings-ness. 9.2.6  Contrasting -ity, -y and -(a)cy with -ness 9.2.6a It is sometimes said that -ness may be added to any adjective (whatever its genetic origin), forming a noun which describes a state. There is, however, a difference between ‘can happen’ and ‘does happen’. It is instructive to look carefully at what has occurred with Romance adjectives. For almost every Romance adjective which—in ME times or later— adopted a nominalization in -ness, that is all there is. We have correct-ness but no *correct-ity, *correct-y, or *correct-cy, common-ness but no *­ common-ity etc. (And see 2.6h.) There was mention in 2.6a of how I imagined that there should be a form appropriacy (related to appropriate in the way that accuracy is to accurate, and so on). But no such word exists in standard English; one just says appropriate-ness.

304    9 making new nouns 9.2.6b What about those Romance loans which brought with them an -ity or -y or -(a)cy derivation, or had one coined within English? In just about every such case, there was also a -ness form, which was in competition with the -ity, -y or -(a)cy nominalization. A sample, with dates of first attestation, is: jealous (1225) curious (1340) modern (1485) honest (c1300)

jealous-y (1225) curious-ity (1380) modern-ity (1635) honest-y (1509)

jealous-ness (1360) curious-ness (1386) modern-ness (1653) honest-ness (c1420–1556)

The nominalizations in the middle column have a high frequency today. Jealous-ness, curious-ness, and modern-ness have not quite disappeared (as honest-ness appears to have done) but they are all rather rare. It is probably true to say that -ness can be added to any adjective. For Romance adjectives like honest, modern, and the like, -ness did indeed enter the fray; but it has effectively lost out to Romance nominals with -ity or -y or -(a)cy. 9.2.6c Sometimes, two forms with similar meanings continue to co-exist. Noun faith was borrowed from French and, in ME times, soon adorned with Germanic affixes: adjectivalizing -ful, nominalizing -ness, and negator un -1. The OED gives dates of first attestation as: noun adjectives abstract nouns

faith (c1300) faith-ful (c1300) un-faith-ful (1384) faith-ful-ness (1425) un-faith-ful-ness (1388)

Then, in 1509, there came abstract nouns fidelity and infidelity, as direct Romance loans. (These followed noun infidel, in 1425, predominantly used to describe anyone who did not subscribe to Christianity, especially a Moslem.) Some speakers today would use infidelity primarily for religious transgression and un-faith-ful-ness for disregard of marriage vows. However, the Cobuild dictionary (Sinclair 2001: 801, 1705) equates them, saying for each that it describes having sex with someone other than a spouse or a permanent partner. (Perhaps it is slightly posher to indulge in infidelity rather than un-faith-ful-ness?) 9.2.6d The activity of making new words never stops. Normal was borrowed, from Late Latin, about 1500. Around 1850, it appears that the need arose

9.2.7 - ity and - ness following - able/-ible   305 for a noun describing the ‘state of being normal’. Almost simultaneously, normal-ity, normal-ness, and normal-cy came into being. All three are in use today, normal-ness only marginally, with normal-cy said to be preferred in the USA and normal-ity elsewhere. 9.2.7  -ity and -ness following -able| -ible 9.2.7a Suffix -able creates adjectives, and both -ity and -ness form abstract nouns from adjectives. Adjectives in -able differ as to whether they accept -ness, or -ity, or both. One hears comfort-able-ness, not *comfort-abil-ity, reverse-abil-ity not *reverse-able-ness. But for change-able both nominalizations are acceptable, and carry a different semantic nuance. What is the explanation for this? 9.2.7b We can begin with adjectives derived by -able from verbs (see 8.4.7). In terms of the scheme of syntactic orientation set out in 8.4.1 (see Table 8.7), most verbs taking -able are transitive and the syntactic orientation of the derived adjective is almost always towards the O argument of the underlying verb. If an outcome is desir-able, it is desired; if a decision is revers-ible, it can be reversed. It was mentioned, in 9.2.1c, that the suffix sequence -able plus -ity becomes fused into -ability, with an implicit link to adjective ability. An adjective derived by -able from a transitive verb with O-orientation has the ‘ability’ to be done, and thus naturally takes suffix -ability. ‘Ability’ is perhaps not quite the appropriate term, ‘potential’ being more apposite. Thus, only -ability, not *able-ness, is used with desire, reverse, read, accept, and the like. As mentioned in 8.4.7a, change is an ambitransitive verb of type S = O. Adjective change-able can relate to the transitive sense of the verb and then has O-orientation, as in The boss is likely to change the date of the meeting. For this sense, nominalization -ity is appropriate—the OED quotes: The change-ability of phenomena through human agency. When change-able relates to the intransitive sense of the verb, it has S-orientation, as in The weather is change-able. In this instance, nominalizer -ness is most likely to be used—The change-able-ness of the weather. 9.2.7c What about adjectives formed by adding -able to nouns? Choice of nominalizer again depends on ‘potential’. If some venture is profit-able, it

306    9 making new nouns has the potential to show profit, so that we may speak of its profit-abil-ity. Marriage-able is all about ‘potential’, leading to marriage-abil-ity (rather than *marriage-able-ness). In contrast, if a chair is comfort-able, this is just what it is; there is no element of ‘potential’ involved. The nominalization is thus comfort-able-ness. 9.2.7d This semantic principle, of using -abil-ity when there is ‘potential’ and -able-ness otherwise, goes a good way towards explaining the occurrence of the two nominalizers after -able. However there is also a good deal of variation. Indeed, adjective vari-able has given rise to vari-abil-ity and vari-able-ness, with no discernible difference in meaning.

9.3 People, agentive and otherwise: nouns from verbs and nouns There is a range of possibilities for creating labels for people on the basis of what they do, what they are associated with, and (sometimes) what is done to them. A number of verbs undertake double duty, as a designation for the agent of the action they describe as a verb. For example, guide, witness, cook, cheat. In addition there are twenty or so suffixes which create names for people, mainly from verbs but some also from nouns—clean-er, attend-ant, violin-ist, million-aire. A few of these also derive instrumental nouns, such as heat-er, lubric-ant, or descriptions of places, such as sleep-er.

9.3.1 -er, -or, -ar, -ier 9.3.1a Suffix -er /-ə/ is one of the most pervasive derivational affixes in English, having a wide range of meanings. However, it is an etymological jumble. First, there was suffix -ere in OE; from rı¯ pan ‘reap’ was derived rı¯ p-ere ‘reaper’, and from wrı¯ tan ‘engrave, draw, depict, write’ there was wrı¯ t-ere ‘writer’. Many modern -er forms go right back to OE. There are also a considerable number of -er nominalizations which are Romance loans, either from French or directly from Latin; for example, ­defend-er was taken over from French defend-eur. Romance loans often show an -or or -ar ending (both pronounced as /-ə/, just like -er). For example, inspect-or is from Latin inspect-or and French inspect-eur, possess-or

9.3.1  -er, -or, -ar, -ier   307 from Latin possess-or and French possess-eur, schol-ar from Latin schol-a¯ris ‘of a school’. Which form of the suffix is used may not always be a true indicator of origin. Verb beg, which is of Germanic origin, is attested with an agentive nominalizer from the thirteenth century. At first, spellings begg-er and beggar were both in use. Of the OED example sentences between 1250 and 1700, three-quarters employ the -er alternative (as would be expected). But the spelling begg-ar was always hovering, and appears to have been the only one used after 1700. This suffix has a by-form -ier /-iə/, as in glaz-ier /'gleiz-iə/, hotel-ier, ­finance-ier, and so on. 9.3.1b Leaving aside the orthographic form of the suffix, we can examine the types of word to which it may be attached and the meanings produced. The suffix is prototypically added to a verb and then derives an agentive nominalization, describing a person who initiates and/or controls the activity described by the verb; for example, arrang-er, shoot-er, forg-er, employer, investigat-or, preach-er, act-or, tempt-er. Almost all semantic types—listed in the Appendix—include some (or many) verbs forming agentive nouns with -er (or -or or -ar). There are virtually none in the liking and annoying types, since one cannot have an agentive nominalization of a state. We do find worshipp-er, but this implies undertaking something physical as an act of worship (for example, kneeling and praying before a cross or other idol, or making a sacrifice), and not just worshipping in the non-physical sense of ‘showing respect or admiration for’. The conditions for use of -er can be seen by examining verbs from the subtype motion-a. One often hears runn-er, danc-er, div-er, jogg-er, and swimmer since these are all volitional activities that people indulge in habitually. Less common are wav-er and wriggl-er since these actions are generally not done habitually; but if they were to be, nominalizations are available. An agentive nominalization is typically used with a modifying adjective which describes how the agent characteristically performs the activity. For example:

a good tipp-er a fast runn-er

a heavy/light sleep-er a deep think-er a heavy smok-er/drink-er

308    9 making new nouns All of these nominalizations describe a habitual activity. One can say He is the winn-er of the 3.20 race but to describe a horse just as a winn-er will generally imply that it has won a number of races. However, there are a few verbs that describe an action which if performed at all is highly significant. For these, an Agent-nom may be used to describe the action being done just once, not habitually—for example, kill-er, ­murder-er, desert-er (from an army or similar force). Note that -er is not possible with verbs which are unlikely to be agentive, such as fall, spill, upset, capsize (and others from the motion-g subtype) and die. And it is not used with verbs referring to an activity which is unlikely to be performed habitually, such as assume, suspect, consider, imagine, ponder, reward, bequeath, insult. 9.3.1c An animal may control its activity in a way that is significant to speakers of English, as when a horse is described as a good gallop-er, or a fighting cock is lauded as a mighty battl-er. Occasionally, -er can identify a type of animal suitable for being subjected to some processes. In America, a broiler is a young chicken just right for broiling (what the English would call grilling). Some animals have descriptive names involving -er—a woodpeck-er pecks at wood, the type of sporting dog called a retriev-er is trained to retrieve game shot down by its hunter owner. The suffix -er may be added to a verb to describe some inanimate object which facilitates an activity. A transport-er is a kind of large vehicle used to transport goods, a scrap-er is a tool used for scraping; other examples include pok-er, mow-er, mix-er, shredd-er, heat-er, and freez-er. An attachment to an artefact can engender or affect an activity—the revolutions of a propell-er are what propels an aircraft, and the muffling effect of a silenc-er is what reduces the sound of a gun-shot (virtually silences it). Then there is the thrill-er, a work of light fiction which thrills the reader (or which provides thrills for them—the basic word thrill functions as both verb and noun). In today’s English, cook is considered to be primarily a verb, secondarily an agentive noun. However, historically the noun came first. In OE cōc described a person who prepares food for eating, the word taking on double duty as a verb only in late ME times. Since there was already a term for the person who does the cooking, the derivation cook-er was available (in the 1880s) to describe an appliance used for cooking.

9.3.1  -er, -or, -ar, -ier   309 A few words derived by -er from a verb describe locations. A din-er is where people can dine. A sleep-er is a carriage on a train where bunks are provided for sleeping in on an overnight journey. A feed-er is a trough into which food is placed for animals or birds to feed from. There are further possibilities, some verging on the bizarre. To mention just one, in the Australian city of Adelaide a float-er is a dish consisting of a small meat pie floating in a bowl of pea soup. 9.3.1d There are also many kinds of -er nominalization based on nouns. As illustrated in Table 8.5 of 8.3.1, an inhabitant of Iceland is called an Iceland-er. Further designations for people based on where they live include New York-er, northern-er, mid-western-er, highland-er, island-er, villag-er. A back-bench-er is a member of parliament who is not a minister and thus sits on a back bench. Nouns hat and cap go back to OE. ME introduced derivations hatt-er and capp-er, for tradesmen who make and/or sell hats and caps respectively. The OED quotes an Act of Parliament from the reign of King Henry VII (1488– 9): No hatter or capper [shall ] put to sale any hatte . . . above the price of xxd (that is, twenty pence). The nominalization capp-er dropped out of use a couple of hundred years ago, but hatt-er survives. At one time mercurous nitrate was commonly used in the manufacture of felt hats, with the resulting mercury poisoning leading to eccentric behaviour. Hence the expression as mad as a hatter (from 1829) leading to The Mad Hatter in Alice’s Adventures in Wonderland (1865). Similar examples—without any overtones of peculiar behaviour—include pott-er and saddl-er. And a cat which is adept at catching mice can be referred to as a good mous-er. Prison and jail (or gaol ) are alternative names for the same kind of institution. It is interesting that their -er forms have evolved in opposite directions. A prison-er is a convicted criminal who is kept locked up, and a jail-er (or gaoler) is an official who is paid to ensure that the prisoners are kept locked up. Some—but by no means all—names of sports and musical instruments may derive the name of a performer by adding -er. They include golf-er, cricket-er, football-er, netball-er, drumm-er, trumpet-er (ski-er is based on the verb ski). The names of some sciences which end in -y may replace this by -er, creating the name of a practitioner—geograph-er, astronom-er, and so on. Suffix -er has a high prevalence in colloquial and slang registers. We find peep-ers for eyes and opprobrious terms such as rott-er for someone who

310    9 making new nouns behaves in an anti-social manner, doing rotten things. (Note that these appear to be related to verb peep and adjective rotten respectively.) 9.3.1e The suffix -er helps create many compounds. A watch-mak-er is someone who makes watches; note that there is no word watch-make, without the -er. Other compounds which incorporate an object noun include booksell-er, hair-dress-er, brick-lay-er, can-open-er. Then there are well-wish-er ‘someone who wishes you well’ and good-look-er ‘someone (generally a woman) who looks good’. And the first element in a compound can be an underlying peripheral element: movie-go-er ‘someone who habitually goes to the movies’. Indeed, as with -ness (9.2.5), the possibilities for adding -er to a sequence of words are limited only by the imagination (and prosodic tolerance)—he’s always been an every-person-for-themself-er, while she’s a do-as-you-would-be-done-by-er. 9.3.1f One topic which can be investigated systematically is the formation of agentive nominalizations from phrasal verbs. Each of these is a single lexeme made up of a simple verb plus one or two prepositions. There are six varieties, as shown on the left-hand side of Table 9.1— ‘p’ indicates a preposition and ‘N’ a noun phrase. Note that the ‘N’ can move to the right of the ‘p’ in types (iii)—Pick the box up or Pick up the box—and (vi)—Put the accident down to carelessness or Put down the accident to carelessness— but no movement is possible for the other varieties. The right-hand side of the table illustrates nominalizations for (i)–(iii); I have not been able to find any for the other varieties. (There is further discussion, and exemplification, in Dixon 2005a: 144–6.) Table 9.1  Agentive nominalizations of varieties of phrasal verbs PHRASAL VERBS

(i) (ii) (iii) (iv) (v) (vi)

AGENTIVE NOMINALIZATIONS

VARIETY

EXAMPLE

FORM

verb + p verb + pN verb + Np verb + NpN verb + ppN verb + NppN

sleep in verb-er p look after N verb-er p pick N up verb-er p-er see N through N ⎫ fall out with N ⎬ put N down to N ⎭

NUMBER

EXAMPLE

very few very few quite a few

sleep-er in look-er after pick-er upp-er

none

9.3.3  contrasting - er and - eer   311 This account has only been able to summarize some of the more important properties of derivational suffix -er. There are more inclusive accounts in Jespersen (1942: 224–38) and Marchand (1969: 273–81). As Marchand states, and illustrates, ‘this suffix -er can be tacked on to almost any basis: a simple or composite substantive [noun] or adjective, a numeral, all kinds of phrases, on the general semantic basis “he who or that which is connected with or characterized by his or its appurtenance to—”.’ See also Bolinger (1941) and Mencken (1945: 365).

9.3.2 -eer Nominalizations in -eer /-'iə/ came into English beginning in the sixteenth century, reflecting French -ier. This suffix is added to disyllabic nouns that have stress on the first syllable; this is replaced by secondary stress with primary stress now on the final syllable, which includes the suffix. For example profit /'prɔfit/ and profit-eer /ˌprɔfi't-iə/, mountain /'mauntin/ and mountaineer /ˌmaunti'n-iə/. The basic meaning is ‘person concerned with’. A mutin-eer takes part in a mutiny, a sonnet-eer composes sonnets, an engin-eer is a person who makes or repairs or drive engines, a musket-eer is a soldier armed with a musket, a chariot-eer drives a chariot, an auction-eer conducts auctions, a pamphleteer writes or publishes pamphlets. A interesting property is that -eer almost always follows /t/ or /n/. There are rare instances of its following /l/—also an apico-alveolar sound—as in pistol-eer. There are a couple of verbs ending in -eer, both coming from Dutch sources: domin-eer is ‘establish domination over’, and command-eer is ‘seize command of’ (note that here -eer is added after /d/, another apico-alveolar consonant). Many -eer derivations are just descriptive, with no judgemental overtones: a mountain-eer is simply someone who makes a habit of climbing mountains. But others carry a disparaging nuance. For example, a profit-eer is someone who makes an undeserved profit by nefarious means, a racketeer indulges in illegal activity (a racket). The suffix does have a degree of productivity, especially in American English, typically (but not always) with a disdainful overtone.

9.3.3  Contrasting -er and -eer Suffixes -er and -eer have overlapping meanings. It is relevant to enquire which is used when, and why.

312    9 making new nouns If something is disposed of by auction, one will have a sell-er, a buyer (both with -er), and an auction-eer (with -eer). Suffix -eer generally attaches only to nouns. Auction functions as both verb and noun—plus it is disyllabic, ending in /n/—and so can accept -eer. Buy and sell are monosyllabic verbs and are restricted to -er (which may be added to both nouns and verbs). True, we could theoretically get *auction-er, but do not. Auction-eer, with its final stress, places more emphasis on what the person running the auction does, and so is preferred to *auction-er. On the same principle, we have bak-er (based on a monosyllabic verb) and engin-eer (based on a disyllabic noun, ending in /n/). Also dance-er alongside puppet-eer. It is not easy to find a word which is used with both -er and -eer. Market is a noun also doing double duty as a verb. Market-er is based on the verb and means ‘sell some commodity at a marketplace’. Market-eer is prototypically used in the combination black market-eer, based on the noun phrase black market ‘an illicit forum for trading goods’. The use here of -eer is in keeping with the tendency to use it for describing disreputable activity.

9.3.4 -ant, -ent This suffix, which comes from Latin -ant(em) or -ent(em), through French -ant, may be written as -ant or -ent and is pronounced /-ənt/. It is added to verbs, deriving quasi-agentive nouns and also derives adjectives ‘having the quality of verb-ing’, discussed in 8.4.16 (for example, observ-ant and diverg-ent). In deriving nouns, -ant/-ent has a range of meanings. At one end it creates agentive nominalizations, just like -er. A superintend-ent is someone who superintends, just as an observ-er is someone who observes. Some verbs take both suffixes with an institionalized difference of meaning. A ­command-er is a general term for anyone who commands, whereas command-ant is used to describe a particular position, the person placed in command of a certain military unit. In most of its uses, -ant/-ent carries a less strong ‘controlling’ sense than -er. It can be characterized as ‘being involved in an activity (often, as one of a group)’. For example: stud-ent assist-ant serv-ant

discuss-ant contest-ant attend-ant

particip-ant inhabit-ant

resid-ent migr-ant

9.3.5  contrasting - er and - ant/-ent   313 However, the ‘one of a group’ specification does not always apply—president, defend-ant, and claim-ant are likely to be unique individuals. The suffix is found almost entirely on Romance roots, sometimes together with a verb ending in ate, and a noun in -ation, as with participate/ particip-ant/particip-ation, migrate/migr-ant/migr-ation, supplicate/supplicant/supplic-ation. The noun lubric-ant—from lubricate/lubric-ant/lubricaation—refers not to a person but to a fluid. The suffix is scarcely productive today, although we do find cool-ant (from 1930)—in which the suffix is added to a Germanic adjective—by analogy with the Romance form lubric-ant (1822). 9.3.5  Contrasting -er and -ant| -ent Sometimes we find two verbs of similar meaning, one taking -er and the other -ant| -ent. For example: learn-er help-er

stud-ent assist-ant

This is because learn and help are Germanic and may only take -er (not -ant| -ent), whereas study and assist are Romance and were borrowed together with their agentive nominalizations stud-ent and assist-ant. There are a number of pairs with a significant difference in meaning, the -er member being more agentive and the -ant| -ent one verging towards just ‘being involved in’: defend-er—of a castle, of one’s reputation, of political ideals defend-ant—nowadays mostly confined to legal matters oppos-er—of a motion, by analogy with propos-er oppon-ent—in battle, in sporting contests, in argument propos-er—of a motion (in debate) propon-ent—of an idea or course of action inform-er—this has acquired the rather specific meaning of someone who tells the authorities that someone else has done something they should not have done (and may be punished for) inform-ant—used since the late nineteenth century for a person who supplies information to a linguist or anthropologist concerning their language and culture (this has recently been hounded out of use by ‘politically correct’ people who lack sufficient intelligence to perceive the difference in form and meaning between inform-ant and inform-er)

314    9 making new nouns 9.3.6 -ist 9.3.6a Suffix -ist /-ist/ goes back to Greek -istēs, through Latin -ista and French -iste. English has many -ist forms which are unanalysable—sophist is from Greek, baptist and evangelist from Latin. Loans also came in pairs— psalm and psalm-ist, jury and jur-ist—establishing the suffix in English, where it continues to be highly productive. 9.3.6b Most -ist derivations are based on nouns. (1) A few have an agentive sense: a colon-ist helps establish a colony, a theor-ist propounds a theory. (2) Related to these are nominalizations which describe someone subscribing to a system of ideas or beliefs. A ritual-ist is someone who delights in ritual (for example, within a church). Often the name of an important thinker is sufficient to connote the ideas associated with them—Buddh-ist, Marx-ist, Calvin-ist, Darwin -ist, and many more. (3) Suffix -ist may be added to the name of a musical instrument to denote someone who plays it; for example, violin-ist, clarinet-ist, pian-ist. It can refer to someone who is occupied with a vehicle or tool—cycl-ist, motor-ist, balloon-ist, hammer-ist (the term for a field geologist)—or a commodity—tobacco-n-ist. (4) In similar fashion, it may describe the practitioner of some art or science: art-ist, landscap-ist, cartoon-ist, novel-ist, botani-st, abortion-ist, pharmaci-st, geologi-st, meteorologi-st (and with many other nouns ending in -ology). (5) An -ist form can describe someone who takes part in a particular activity. A competitor who gets through to the (semi-)final(s) of some event becomes a (semi-)final-ist. A duell-ist is involved in a duel. 9.3.6c Only a few -ist forms are based on a simple adjective; for example, nud-ist. In contrast, there are many examples of -ist added to an adjective ending in -al—an imperial-ist is in favour of imperial rule (as in the old British empire). Some -al adjectives, including imperial, fatal, and legal, are not analysable in English. Others are: a spirit-ual-ist is concerned with matters spirit-ual; that is, associated with the spirit (as opposed to the physical body) of a person. When the base is an adjective, there may be a doctrinal noun in -ism corresponding to a ‘participant’ noun in -ist (see 9.5.7). A spirit-ual-ist believes

9.3.6 - ist   315 in the doctrine of spirit-ual-ism, that spirits have an independent existence. Similarly for imperial-ist and imperial-ism, nation-al-ist and nation-al-ism, material-ist and material-ism. Similarly for the noun-based examples under (2) of 9.3.6b—a Marx-ist follows Marx-ism, a ritual-ist practices ritual-ism, and so on. 9.3.6d Under (vii) in 7.4.2 we examined verbalization by -ize of adjectives ending in ic. Those for which the -ic is an analysable suffix replace it with -ize whereas when the ic is an inherent part of the root, -ize is added after it. Nominalizer -ist behaves in the same way as verbalizer -ize. For example: adjective romantic dogma-t-ic drama-t-ic

derived verb romantic-ize dogma-t-ize drama-t-ize

derived noun romantic-ist dogma-t-ist drama-t-ist

One could, of course, suggest that dogma-t-ist and drama-t-ist are based on the underlying nouns, dogma and drama, rather than on the adjectives. 9.3.6e Suffix -ist generally does not affect the stress placement or vowel quality of a form to which it is attached. A rare exception is saxophon-ist /sak'sɔfən-ist/ from saxophone /'sakzəfoun/. After a noun ending in /i/, the suffix is simply /-st/, as in strategy /'stratədʒi/, strategi-st /'stratədʒi-st/. There are varied explanations for what happens when -ist is applied to other roots ending in a vowel. In canoe /kə'nu:/ stress falls on the second syllable, so this must be retained in canoe-ist /kə'nu: -ist/. Pian-ist (rather than *piano-ist) is from French and Italian. We get Buddh-ist (rather than *Buddha-ist) on the basis of Buddh-ism. Banjo is a loan believed to be from African sources; its practitioner could be called a *banj-ist, but usage has preferred banjo-ist. However, there are two terms in use for the player of a cello, cell-ist and cello-ist (the former being far commoner). Rather than *tobacco-ist or *tobacc-ist, one says tobacco-n-ist (like Plato-n-ist from Plato), perhaps simply because it has such a euphonious sound. There is the possibility of adding -ist to a multi-word constituent, as in: He’s always been a getting-deeper-and-deeper-into-debt-ist. Jespersen (1942: 334) quotes some examples from literature, but this does seem a rather marginal phenomenon.

316    9 making new nouns 9.3.7 -ese, -i, -ite, -(a)n, -ian, -(i)cian, -arian 9.3.7a The processes for deriving adjectives from the names of nations and places were discussed in 8.3. For Type A, illustrated in Table 8.5, the adjective (for example, Swed-ish) is different from the term for an individual (a Swede) or the population (the Swedes). For Iceland, -ic is added for the adjective and -er (see 9.3.1) for the nouns—Iceland-ic sagas, an Iceland-er, the Iceland-er-s. For names of Type B (illustrated in Table 8.6), a single derivation functions as both adjective and noun—as in Chin-ese calligraphy, a Chin-ese, the Chin-ese, and Afric-an music, an Africa-n, the Africa-n-s. Suffixes in this set are -ese, -(a)n, -ian (Jordan-ian), -ite(Canaan-ite), and -i (Punjab-i). 9.3.7b Suffix -ite may be added to a few common nouns, such as suburban and beach (referring to a kind of place); suburban-ite and beach-ite are each both noun and adjective. A recent coinage is soccer-ite ‘someone who plays and/or likes the game of soccer’. This corresponds to football-er; note that although suffix -er /-ə/ can be added to a noun ending in er /ə(r)/ as in— caterer /'keitərə/—there does appear to be a preference against this. 9.3.7c Suffixes -(a)n and -ian are also used with names of people and with some common nouns. Mohammed-an, Christ-ian, Darwin-ian, and ­republic-an each functions as adjective, describing adherence to a certain notion (Darwin-ian theory), and as noun, for adherents to it (the conference of Darwin-ian-s). 9.3.7d ME borrowed from French a number of names for sciences and suchlike ending in ic or ics, together with agentive nominals in -ian. They include music /'myu:zik/ and music-ian /'myu:ziʃn̩/, magic and magic-ian, mathematics and mathematic-ian, logic and logic-ian. By analogy, suffix -ian came to be added, within English, to similar nouns ending in ic, such as statistic-ian and tactic-ian. In all these examples, the ic is part of the root, and not analysable. There was then a new development, with -(i)cian /-(i)ʃn̩/ being adopted as an agentive nominalizer, as in diet-ician and algebr-ician, from diet and algebra. This is particularly prevalent in the American dialect. Mencken (1936: 178–9, 288) describes—in his typical jovial style—how ‘we have the lovely mortician, and its brothers beauti-cian, cosmetic-ian and boot-ician [‘bootlegger’], to say nothing of whoop-ician, a college cheer leader’ [hyphens added].

9.3.8  contrasting - er with - ist , - (a)n , and - ian   317 9.3.7e Suffix -arian /-eəriən/ (reflecting Latin -arius) indicates ‘association with a place or thing or idea’. It emanates from -ian added to a noun ending in ry—as in librari-an ‘someone who looks after a library’ and seminari-an ‘a student in a seminary’. In a similar manner to -ician, a new suffix evolved, -arian. An authorit-arian exercises authority in an inflexible manner, a sectarian is a narrow-minded adherent to a religious sect, a humanit-arian promotes the welfare of disadvantaged humans, and a veget-arian confines their diet to vegetables and fruit (eschewing meat and fish). At the end of the eighteenth century nothing-arian began to used for someone with no religious or political affiliation (and then the word faded out of use). There are also Latin-based age-group terms such as octogenarian ‘a person aged eighty’, but these are not analysable within English.

9.3.8  Contrasting -er with -ist, -(a)n, and -ian 9.3.8a A label for the agent of some activity can be formed with -er (added to many verbs and some nouns), with -ist (with nouns and some adjectives), or with -(a)n or -ian (with many proper names and some common nouns). Which suffix is used for which sort of agent? This depends on the nature of the activity, on how the name is created (whether from verb or from noun), and on genetic origin. There is also a whiff of randomness. In some instances there are two labels, equally apposite, and both in use at some time, but usage has embraced one and side-lined the other. It is instructive to briefly survey agents for a number of semantic sets. 9.3.8b members of the orchestra (1) A couple of performers are named from a verb, and thus use -er. They are trumpet-er (from French trompet-eur) and drumm-er (based on the Germanic verb drum). (2) Other names for instrumental-ist-s involve -ist, and are taken directly from French or Italian: pian-ist, violin-ist, clarinet-ist, trombon-ist. The -ist suffix was then extended to newly-introduced instruments, as in banjo-ist, and saxophon-ist. (3) There is no agent noun in general use for performers of some instruments. One just has to say french horn player and tuba player. Why so? Perhaps because no need has been felt for coinages here (or perhaps because putative derivations just do not sound quite right).

318    9 making new nouns 9.3.8c players of games and other activities (1) Some agentive labels are based on a verb, and thus take -er. They include ski-er, skat-er, rid-er, batt-er, bowl-er, pitch-er, field-er, danc-er. (2) Those based on a noun which is the name of a game also take -er. For example, golf-er, cricket-er, football-er. (3) Those based on a noun which is the name of some object used in the activity take -ist. For example, cycl-ist, motor-ist, balloon-ist, canoe-ist. Note that this use of -ist is similar to that with the names of musical instruments, (2) in 9.3.8b. (4) There is no term in general use for players of some very popular games. Hockey, tennis, squash, and chess are the names of games and—­ alongside golf-er, cricket-er, and the like—one would expect an agentive nominalization in -er. However, no such terms are in general use. One simply has to say: hockey| tennis| squash| chess player. (It is hard to know why.)

9.3.8d practitioners of trades and professions (1) Some agent nouns are plainly based on verbs, and use suffix -er; for example, clean-er. (2) Others are based on a noun, and also employ -er. A noun describing either what is made—saddl-er, pott-er—or what is practised—law-yer. For a number the base form does double duty as both noun and verb: garden-er, fenc-er, thatch-er. (3) Then there are a number employing suffix -an or -ian. Librari-an, magic-ian, and politic-ian are Romance loans. Electric-ian was probably coined by analogy with other words ending in ic and taking -ian (and see 9.3.7d). (4) Once again, there are some occupations lacking an agent noun. If someone who places tiles on the roof is called a til-er, why shouldn’t there be a name for someone why hangs blinds on windows? We just say post-man (or -woman) for the deliverer of post, probably because there is already a term post-er with a different meaning. There are further possibilities, and also alternations between them. If you move house and hire people to transfer furniture, they may call themselves

9.3.8  contrasting - er with - ist , - (a)n , and - ian   319 remov-al-ist-s (adding -ist to noun remov-al) or else remov-er-s (adding -er to verb remove). 9.3.8e followers of a doctrine We find some agent nouns in -ist (Marx-ist, monarchi-st, imperial-ist, anarch-ist) and some in -ian or -an (Freud-ian, Christ-ian, republic-an). These appear each to reflect the ending used in Latin or French; there seems not to be any semantic principle at work. It is interesting to compare derivations based on the names of two great Greek thinkers, Plato and Aristotle: ideas associated with (adjective) Platon-ic

follower of (noun) Platon-ist ⎫

⎬

⎭

Aristotel-ian Aristotel-ian and Platon-ic reflect forms in Greek, coming into English through Classical Latin. Platon-ic was originally, like Aristotel-ian, used as both adjective and noun. Then the noun Platon-ist was taken over from Post-Classical Latin, with Platon-ic being restricted to adjectival function. 9.3.8f scientists and artists (1) Some labels are based on a verb, and take -er; for example, writ-er, paint-er. (2) Nouns ending in -graphy form an agentive nominal with -er. We find, among others, geograph-er, photograph-er, lexicograph-er, biograph-er. (3) A number involve -an or -ian added to the name of a discipline, reflecting the form in the Romance language they were borrowed from. These include music-ian, histori-an, grammar-ian, traged-ian. (4) A considerable number use -ist.

• • •

Nouns ending in ology only take -ist—archaeologi-st, geologi-st, morphologi-st, phonologi-st, and many more. Some, like botan-ist, reflect the original form in Greek. Others are based on a noun referring to an item (similarly to violin-ist and canoe-ist, mentioned above)—novel-ist, cartoon-ist, lyric-ist.

320    9 making new nouns

•

The label for a student of a language or language family uses -ist: Arabist, Semitic-ist, Alqonquian-ist.

There are interesting contrasts, such as comedi-an (from French) for someone whose vocation is being funny—on stage, radio, or TV—and humour-ist (from Italian) for someone who writes or behaves in an amusing manner. 9.3.8g The fluidity of assigning agentive nominalizations can be seen from historical variations. Two labels were originally in use for a practitioner of astronomy, astronom-er and astronom-ien, of which only the former has survived. In the nineteenth and early twentieth centuries, a need arose for labels describing specialists in various areas of linguistics—phonetics, syntax, and semantics. Should one employ -ist or -ian? For each term both possibilities eventuated: phonetic-ian and phonetic-ist, syntactic-ian and syntactic-ist, ­semantic-ian and semantic-ist. All still have some currency but those preferred today are two in -ian (phonetic-ian and syntactic-ian) and one in -ist (semantic-ist). This demonstrates the degree of choice possible, and the arbitrariness of its resolution as the habits of a speech community close in on one alternative and exclude another. 9.3.9 -ess 9.3.9a Feminine suffix -ess—pronounced as /-is/ or /-es/ or /-əs/—goes back to Greek; for example diakōn-issa ‘deacon-ess’. This was taken into Romance languages as Latin -issa and French -isse. ME accepted pairs such as patron and patron-ess, host and host-ess, adulter-er and adult-r-ess, ­torment-or and torment-r-ess. 9.3.9b The suffix is added to a noun referring to a person and indicates that the person is female. The basic noun most often describes an office or role, with the normal expectation being that the person fulfilling it is male. There is a perception that a prophet or mayor or priest is prototypically male. If the holder of such a position should be female, then -ess is added to indicate this: prophet-ess, mayor-ess, priest-ess. Suffix -ess is also added to aristocratic titles to denote the wife of the male holder of the title. The wife of a baron will be a baron-ess, similarly for princ-ess, viscount-ess. However, these titles may also designate a woman

9.3.9 - ess   321 who holds the title in her own right. If someone is referred to as Baron-ess Smith, it is unclear whether she is just the wife of Baron Smith, or is herself the holder of the barony. There are a few further instances of -ess, added to nouns describing people with some notable characteristic. For example, millionair-ess, heir-ess (implying: heir to a considerable fortune), murder-ess, adult-r-ess, tormentr-ess. Adventur-er and adventur-ess are used of people who undertake something particularly dangerous or risky (perhaps employing unscrupulous means). A giant is certainly a notable being, and a giant-ess even more so (but there is no *midget-ess). It appears that just three of the most fierce and dangerous animals take -ess: lion-ess, tigr-ess, leopard-ess (but there is no *jaguar-ess, for instance). 9.3.9c Some -ess nouns retain the fused form they had in the original Romance language: there are abbess alongside abbot, duchess with duke, empress and emperor. In others -ess is added to a simple noun (not including any suffix), as with host-ess, deacon-ess, lion-ess. Importantly, -ess is never added after most of the agentive nominalizers we have just considered: not after -eer, -ant| -ent, -ist, -ese, -i, -ite, -(a)n, or -ian. However, it is used with words derived by -er (or -or). There are a number of different techniques here: (1) Just add -ess after the -er, as in manag-er-ess, brew-er-ess. Suffix -ess is also added after some nouns ending in er or or which are not analysable in English: author-ess, prior-ess, tailor-ess. (2) Suffix -er or -or, /-ə(r)/, loses its schwa /ə/, and is reduced just to /r/, before -ess. These include act-r-ess /'akt-r-is/ from act-or /'akt-ə/, and wait-r-ess, tempt-r-ess, conduct-r-ess, torment-r-ess. (3) Suffix -er is replaced by -ess. Alongside murder-er there is murder-ess, also govern-or and govern-ess, sorcerer and sorcer-ess, adventur-er and adventur-ess, adult-er-er and adult-er-ess. This is to avoid a sequence of /ər/ followed by /ər/, as would occur in *murder-er-ess /'mə:dər-ər-is/. 9.3.9d After -ess became established as a suffix in English, it was added to a fair number of new words, almost all of Romance origin (although, in some instances, there was no corresponding feminine derivation in French). Just a few Germanic nouns have accepted -ess. Noun god (for a supernatural being of pre-Christian times), verb hunt and compound noun shepherd

322    9 making new nouns (from sheep-herd ) are from OE times. ME created agentive noun hunt-er and feminine forms godd-ess, hunt-r-ess, and shepherd-ess. (And there are a couple of instances of -ess added after Germanic suffix -ster; see 9.3.10b.) Suffix -ess is almost limited to Romance forms, but there are many agentive nouns of Romance origin which do not take it. There is no feminine term corresponding to: pupil, scholar, doctor, minister, soldier, writ-er, ­inspect-or, glazi-er, devil, and many more. One simply has to say lady doctor, female glazi-er, she-devil, and the like. At one time, actors and warders were predominantly men, so that a female act-or or ward-er was something to be remarked on, hence we got actr-ess and ward-r-ess. But dancers and nurses were generally female. Unusual gender associated was then brought out by saying male dancer and male nurse. (An English speaker sometimes uses the French nouns danseur and danseuse for male and female dancer respectively.) Agentive derivations based on Germanic verbs and nouns do not—­ because of their genetic character—take -ess. That is, there are no feminine forms of runn-er, follow-er, sweep-er, think-er, among many others. Most English terms for levels of the aristocracy are from French, and brought along their own feminine—prince with princ-ess, and so on. A rare Germanic title is earl, roughly equivalent to the French comte. Being Germanic, earl had no feminine form available and so count-ess was adopted, from French comt-esse, for the wife of an earl. As shown in Table 8.6, the great majority of nationality nouns involve suffix -ese, -i, -ite, -(a)n, or -ian, and these are never followed by -ess. Hence, there can be no *Africa-n-ess or *Chin-ese-ess. We do get Jew-ess alongside Jew (both Romance), but no feminine forms for the Germanic labels Finn, Swede, and Dane (from Table 8.5). Negro and negr-ess were both of Romance origin. 9.3.9e Nowadays, there is an aversion to marking the sex of an office-holder, or member of a profession. A manag-er is a manager; why signal that a particular manager is female by adding -ess? A female film star prefers to be called just an act-or, like her male counterparts. (Although it has proved useful to retain the -ess form for certain purposes, as when Oscars are awarded for best act-or and best act-r-ess.) Note that although most instances of -ess are being dispensed with, the suffix is still in use for referring to members of the aristocracy, and for large and dangerous felines (princ-ess, lion-ess).

9.3.10 - ster and - stress   323 Baron (1986: 120–3) includes useful discussion of ‘feminine’ suffixes -ess, -ine, and -ette.

9.3.10 -ster and -stress 9.3.10a OE had sex-determined suffixes which could be added to a monosyllabic noun and referred to a person who made (or was concerned with) the item referred to by the noun. For example: base noun sang ‘song’ sēam ‘seam’

male performer female performer sang-ere ‘male singer’ sang-estre ‘female singer’ sēam-ere ‘male tailor’ sēam-estre ‘seamstress, female dressmaker’

The original masculine suffix -ere has developed into modern -er, now used primarily (although by no means exclusively) with verbs. Suffix -estre gave rise to present-day -ster /-stə/, which is used almost only with nouns. Beginning from late OE times, -ster lost its feminine reference and was used for creators and performers of both sexes. Seam-ster was employed for both male and female garment-makers, and we soon had dye-ster ‘maker of dyes’, malt-ster ‘maker of malt’, brew-ster ‘maker of brews’, and a number more. These are little used today. A new line of semantic development then opened up, with -ster added to nouns—again monosyllabic—which refer to things non-material (along the lines of song-ster). From the sixteenth century on there were game-ster ‘someone who plays games’, rhyme-ster ‘producer of feeble rhymes’, tip-ster ‘someone who provides tips concerning who will win some sporting event’. The suffix is particularly favoured in the USA with gang-ster and mob-ster ‘member of a criminal gang or mob’, ring-ster ‘member of an illegal ring (e.g. for price-fixing)’, and so on. As can be seen, -ster carries a decidedly negative meaning in such coinages (and see Mencken 1945: 359–60). The reference of a -ster form is not always human. Road-ster was first used (in 1760) of a horse suitable for riding—or pulling a carriage—on a road, and later for an open sports car (plus, it is now the brand name for a type of bicycle). There have been just a few instances of -ster added to an adjective. In The Merry Wives of Windsor (1604, Act 5, Scene 3), Shakespeare has Mistress Page declaim:

324    9 making new nouns Against such lewdsters and such lechery Those that betray them do no treachery Lewd-ster, from lewd, is no longer in use. A few years earlier there was coined young-ster, which does survive today. It implies that the young person referred to is inexperienced. A couple of hundred years later it was joined by the less common old-ster. Among the useful sources on -ster are Lubbers (1965), Adams (1973: 172–5) and Bauer (2006: 179–80).

9.3.10b Once -ster derivations had ceased to denote a female practitioner, the regular feminine suffix -ess (of Romance origin) could follow it. Seamster originally denoted a woman, then someone of either sex, then suffix -ess was added giving seam-str-ess for a female maker of garments (with -sterreducing to -str- before -ess, as -er often does). It is seam-str-ess which continues today, with seam-ster having dropped out of regular use. However, song-ster has survived, alongside its feminine form song-str-ess. On the basis of seam-stress and song-stress, a new suffix -ress came to be recognized during the early days of Modern English. There was hermit-ress, for a female hermit, and poet-ress (coined fifty years before poet-ess). However, these have not survived.

9.3.11  Contrasting -er and -ster Two nominalizations relate to singing. First (from about 1390), there is singer, based on the verb sing, which describes anyone who habitually sings—we can have a jazz singer, a popular singer, an opera singer, and so on. Then there is the continuation from OE of song-ster, based on the noun song. Today this is used of someone who performs songs in a particular genre (and often composes them as well). It is most often a variety of folk or popular music. An opera singer could never be called a songster.

9.3.12  Minor suffixes: -aire, -ary, -ard| -art, -c, -nik, -ette 9.3.12a A handful of words have been borrowed from French (commencing about 1750) with suffix -aire /'-eə/— taking primary stress—added to a noun which was an earlier loan. It carries the meaning ‘someone associated

9.3.12  minor suffixes   325 with’. Thus concession-aire /kənˌseʃə'n-eə/ ‘a person (or company) to whom a concession has been granted’, questionn-aire ‘a form with a list of questions’, legionn-aire ‘a member of some specified legion’, million-aire ‘a person worth a million dollars (or pounds, or whatever)’. The suffix is not productive in English, save in analogic formations such as billion-aire. 9.3.12b Suffix -ary /-(ə)ri/ was discussed in 8.2.39. It serves in the main to derive adjectives, some of which do double duty as nouns—for example, vision-ary, revolution-ary. There are also some -ary alternatives to -aire derivations: legionn-ary and concession-ary (based on Latin) alongside legionnaire and concession-aire (from French). 9.3.12c Suffix -ard /-əd/ or -art /-ət/ comes originally from Germanic -hard or -hart ‘hardy’ (and is reflected in names such as Richard and Bernard). It made its way into French and then back into ME in such words as coward and bastard (not analysable in English). There were just a few early derivations in English. Some are from verbs: a bragg-art brags a lot, a lagg-ard lags behind. And a few from adjectives: a drunk-ard is habitually drunk and a wiz-ard is notably wise. (This ending is also reflected in Spaniard; see Table 8.5 in 8.3.1.) 9.3.12d Suffix -c /-k/, which is of classical origin, may be added to a number of Romance names, ending in -ia, describing maladies. They designate a sufferer from that malady. For example: insomnia-c, aphasia-c, ­hypochondria-c, haemophilia-c, mania-c, kleptomania-c, dipsomania-c. 9.3.12e One of the most unusual suffixes in Modern English is -nik /-nik/ ‘person associated with’. This originated in Russian, was borrowed into the Germanic language Yiddish, and from there into the Semitic language Modern Hebrew. It came to be used among the Jewish community in the USA, who also knew Yiddish. The OED records real-estate-nik from 1909 and all-rightnik (‘an upstart’) from 1919. American English took from Yiddish the verb noodge ‘pester, nag at’, and then nud-nik ‘a pesterer, a bore’. From Hebrew there came kibbutz (1931) and then kibbutz-nik (1949). The first space satellite, in 1957, was called Sputnik, a Russian word which is not analysable in English. But the nik ending set off a flurry of

326    9 making new nouns new coinages. San Francisco journalist Herb Caen relates how, in early 1958 ‘beat-nik slipped out of my typewriter one day when I was writing about one or another of the Beat types—Kerouac, Ginsberg et al’ (Rex 1975: 329). This word caught on, and spawned a multitude of ad hoc formations such as hold-up-nik, protest-nik, computer-nik. The suffix has humorous and often disparaging overtones. Around 1970, followers of Chomsky at MIT were called mit-nik-s by linguists with a wider purview (and the mitniks didn’t like it at all). 9.3.12f Suffix -ette /-et/, which comes from French, may indicate ‘small version of’, as in kitchen-ette, novel-ette. (It has similar effect to the more productive suffix -let, discussed in 6.11.1.) From the early twentieth century, it developed female reference—undergradu-ette for female undergraduate, usher-ette in a cinema, major-ette for a girl in uniform marching at the head of a band. The suffix has had greatest currency in the USA—see Mencken (1936: 178, 1945: 362). As with -ess, the feminine sense of -ette has now fallen out of favour.

9.3.13  Occupational suffixes: -monger, -wright, -smith, -maker, -man, -woman In 3.3e–g there was consideration of a number of forms—primarily referring to occupations—which are often said to be just the final elements in compounds. Reasons were given why they should be accorded the status of affixes (perhaps called ‘semi-affixes’). They can now briefly be reviewed. Note that -wright and -smith are unlikely to be used in new coinages, but the others have varying degrees of productivity. 9.3.13a OE had agentive noun mangere ‘trader’ (from mangian ‘trade’) and within ME this developed into suffix -monger /-ˌmʌŋgə/ (with secondary stress) ‘dealer in a line of goods’: cheese-monger, fish-monger, iron-monger. From the sixteenth century the suffix took a different semantic turn, now describing disreputable reportage, as in gossip-monger, scandal-monger, scare-monger. When in the 1930s, Winston Churchill warned of the Hitlerian threat, naive citizens of England dubbed him a war-monger (accusing him of fostering war, when he was in fact urging for preparation against the inevitable).

9.3.13  occupational suffixes   327 Interestingly, -y can be added to just one of these -monger forms; ironmonger-y is what an iron-monger sells (these include—but are by no means confined to—artefacts made of iron). 9.3.13b OE had a noun ‘maker of constructions’, written as wyrhta or wryhta (related to weorc ‘work’). This continued into ME and beyond as wright, but is scarcely used today as a free noun, surviving mostly as suffix -wright /-rait/, which attaches to a noun indicating what is being constructed: shipwright, mill-wright, wheel-wright. To these was added, in the seventeenth century, play-wright. (The pronunciation /'plei-rait/ has sometimes been reinterpreted as play-write, taking cognizance of the activity involved.) 9.3.13c In OE smith could be used for a worker with metals or with wood, but the meaning soon narrowed to refer just to an artificer of metals. It also developed as a suffix, added to the names of metals. OE had gold-smith and silver-smith with these being followed by copper-smith, iron-smith, and tinsmith. Black-smith is an alternative to iron-smith (and generally preferred to it nowadays) for someone working with a black metal, while white-smith preceded tin-smith for someone dealing with tin, a white metal. We then had lock-smith and gun-smith, for makers of small metal objects. The 1870s saw an innovative use of the suffix with word-smith ‘an expert in the use of words’. This has led to further derivations; I recently heard: You can’t do any word-smith-ing with the banner (meaning, you are not allowed to alter its wording). 9.3.13d The Germanic word mak-er functions as a free noun, as in She’s a maker of furry toys and He’s a maker of all sorts of trinkets. There has also, since ME times, been a derivational suffix -maker /-ˌmeikə/—with secondary stress—‘person (or machine) that produces or engenders something’. It is added to words of all genetic types. The suffix typically applies to material things. Cabinet-maker, watchmaker, tile-maker, (scientific-)instrument-maker, and dress-maker all refer to people; a coffee-maker is most likely to be a machine; hay-maker and icecream-maker are used to describe both machines and people. It is also added to nouns with incorporeal reference, such as law-maker, verse-maker, fun-maker, mischief-maker. A taste-maker is used of an influential person who sets the style in some artistic field (that is, determines people’s tastes).

328    9 making new nouns Bed-maker relates not to the plain verb make but rather to phrasal verb make up, ‘someone who makes up—tidily rearranges—a bed after someone else has slept in it’. As mentioned in 3.3e -maker can be attached to a coordination, as in He’s a [cabinet and desk]-maker. 9.3.13e The conundrum of how to deal with -man was fairly fully aired in 3.3f. There are a score or so words with second element -man which is pronounced /man/, the same as the free-form noun, and bears secondary stress. For example, snow-man /'snow-ˌman/. This is typical of a compound. Then there are around a hundred words in which -man has its vowel reduced and bears no stress, giving -man /-mən/, as in mad-man /'mad-mən/. This is typical behaviour for an affix. How about -woman /-wumən/. This never reduces vowels and has the same form in snow-woman /'snou-ˌwumən/ and mad-woman /'mad-wumən/. There is just the inclusion of secondary stress in the former which can be taken as evidence for its being a compound. In contrast the latter, with no secondary stress, can be regarded as involving a derivational suffix.

9.3.14 -ee Suffix -ee /-'i:/—with primary stress—is from French participial ending -é(e). Legal terms borrowed by ME includes pairs lessor| lessee and donor| donee. English soon developed -ee as a suffix added to verbs (the majority of them being of Romance origin). (1) Most instances involve transitive verbs, with the -ee derivation relating to the reference of the NP in O (transitive object function). The verbs mainly refer to social actions. We get employ-ee, appoint-ee, nomin-ee, grant-ee, pay-ee, train-ee, address-ee, and invit-ee, among others. (2) Some -ee forms relate to the indirect object of a transitive verb. If you mortgage your house to a bank, the bank is the mortgag-ee. (3) The suffix also applies to S (intransitive subject) function for a few intransitive verbs. This appears to have begun in the sixteenth century with absent-ee, which was based on the now archaic intransitive sense of verb absent (as in He absented from the meeting). It was joined, three hundred years later, by escap-ee and then retir-ee. Note that these all

9.4 nouns from verbs   329 refer to a person being out of something which they had previously been in (or ought now to be in). O-derivations such as divorc-ee and deport-ee have similar import. A rather different meaning attaches to stand-ee, for someone who stands, generally when there are no seats available (as in a theatre or at a sporting event). The suffix is highly productive today, particularly in the USA. Mencken (1945: 364–5) mentions ‘fantastic forms’, such as hold-up-ee, crack-up-ee, and quiz-ee. Note the quite different suffix -ee, which can be regarded as a by-form of the diminutive suffix -y (6.11.3). For example boot-ee ‘infant’s woollen boot’.

9.4 States, units, activities, results, objects, and locations: nouns from verbs We now examine nouns, other than those describing people, which are derived from verbs. These basically relate to six semantic categories (except for location, they can all be called ‘abstract nouns’):

• • • • • •

a state such as enjoy-ment, admir-ation, annoy-ance, depend-ence a unit of activity, such as a jump, a chat, a depart-ure extended activity, such as assist-ance, recover-y, cry-ing the result of some activity, such as a wound, a solution, a loss something which can be the syntactic object of the verb from which it is derived, such as an assump-tion, a pay-ment, a discov-ery a location, such as a settle-ment, a resid-ence, an enclos-ure

There are seven key derivational suffixes involved—one is Germanic, -ing, and the remainder Romance: -((a)t)ion, -ment, -ance| -ence| -ancy| -ency (­abbreviated to -a| ence| y), -y, -ure, and -al. Each of these relates to several (four, five, or all six) of the semantic categories. There are also many verbs which do double duty, covering all six categories. And other verbs which create nominals by internal change, such as lose|  loss and grieve| grief. We first discuss each of the six semantic categories, in 9.4.1–5, and then deal with the derivational suffixes and so on, one at a time, in 9.4.6–13. (Note that a couple of the suffixes are also added to a few nouns and adjectives.) The minor suffix -t, which creates some object and result nominalizations (such as complain-t), is the topic of 9.4.14. Suffix -age, discussed in 9.4.15,

330    9 making new nouns has a variety of uses, one of which is to derive result nouns from verbs (for example, wreck -age). See Dixon (2005a: 348–52) for a summary of occurrences of these suffixes across the various semantic types of verbs (listed in the Appendix).

9.4.1 States 9.4.1a An adjective describes a property. Section 9.2 dealt with suffixes which can be added to an adjective and derive a noun referring to a state (for example, state noun happi-ness from adjective happy). As mentioned in 9.2b, verbs from the liking and annoying semantic types (see Appendix) describe ‘being in a state’ and they have similar semantic import to human propensity adjectives. We compared fond-ness (from adjective fond ) with lik-ing (from liking type verb like) and pleas-ure (from annoying type verb please). Verbs in these two types share the same semantic roles, but for liking verbs the Experiencer is the A and the Stimulus is the O argument, with this being reversed for annoying verbs. A state nominalization may be possessed by the noun phrase which is in the Experiencer role—in A function for liking and in O function for annoying verbs. For example: liking verbs JohnA loathes clergymenO MaryA admires athletesO

[John’s loath-ing] for clergymen [Mary’s admir-ation] for athletes

annoying verbs [The result]A satisfied MaryO [The joke]A amused BillO

[Mary’s satisf-action] at [the result] [Bill’s amuse-ment] at [the joke]

9.4.1b Six of the seven key suffixes (all but for -y) are used to derive state nouns from liking and annoying verbs. Examples include: suffix -ing -((a)t)ion -ment -a/ence/y -ure -al

from liking verb lik-ing, loath-ing admir-ation enjoy-ment, resent-ment prefer-ence, abhorr-ence — approv-al

from annoying verb — confus-ion, attract-ion disappoint-ment, bewilder-ment annoy-ance pleas-ure —

9.4.2  units of activity and extended activities   331 There are also a few irregular correspondences—verbs hate and grieve match state nouns hatred and grief. In addition, a few liking and many annoying verbs also do double duty as state nouns. These include love, dislike, scare, shock, surprise, delight, envy, and concern. For example, [John’s dislike] of clergymen and [Mary’s surprise] at the result. There is also state noun depend-ence from verb depend (in the relating semantic type). Compare: JohnA depends on [his mother]O [[John’s depend-ence] on his mother] is worrying 9.4.2  Units of activity and extended activities 9.4.2a Relating to a number of kinds of verbs, there can be (a) a noun describing a unit of activity, and (b) a noun describing a period of extended activity. For example, corresponding to verbs laugh and throw we get: (a)  John gave [a loud laugh] (b)  [[John’s loud laugh-ing] throughout the meeting] annoyed the president (a)  John made [an accurate throw] (b)  [[John’s accurate throw-ing] throughout the season] pleased the captain 9.4.2b A unit activity noun is often a verb undertaking double duty, with an extended activity noun being derived by suffix -ing, as here. But there are other possibilities. These include: suffix -ing -((a)t)ion -ment -a/ence/y -y -ure -al

unit of activity happen-ing, beginn-ing assassinat-ion, promot-ion commence-ment, announce-ment occurr-ence, perform-ance apolog-y, deliver-y fail-ure, depart-ure arriv-al, recit-al

extended activity runn-ing, understand-ing — — assist-ance, guid-ance recover-y — —

To these can be added a few irregular correspondences, such as verbs think, believe, and speak with unit of activity nouns thought, belief, and speech.

332    9 making new nouns 9.4.2c A unit of activity can be counted, as in John emitted loud laugh-s, There were four assassinat-ion-s in the Middle East last year and Mary offered two apolog-ie-s, one to mother and one to father. In contrast, an extended activity noun is not normally countable. One can say Mary’s contrite apologis-ing went on all weekend (not *Mary’s apologis-ing-s). It is interesting that there can be unit nouns relating to liking verbs. In John’s hatred of clergymen, the noun hatred describes a state which is naturally extended in time. But in John has three pet hate-s: dogs, clergymen, and children the verb hate is doing double duty as a unit (and countable) noun. With love, the base form covers both functions: Mary’s love of music (extended activity noun) and Mary’s three dearest love-s: poetry, gardening, and bebop (unit noun). (Like behaves in the same way as love.)

9.4.3  Results of activities 9.4.3a A large number of nominalizations describe the result of an activity. For example: verb X imitated Van Gogh X arranged some flowers X declared war X lost a diamond ring X solved the problem X wounded John X injured Mary

result nominalization imitat-ion of Van Gogh arrange-ment of flowers declar-ation of war loss of a diamond ring solu-tion to/of the problem wound on John injur-y to Mary

9.4.3b All seven of the key derivational suffixes are utilized in creating result nouns. For example: -ing -((a)t)ion -ment -a/ence/y -y -ure -al

bett-ing inform-ation, solu-tion, decis-ion, imitat-ion, declar-ation arrange-ment, judge-ment, punish-ment, imprison-ment differ-ence, resembl-ance, assist-ance, depend-ancy injur-y, enquir-y, assembl-y expos-ure, eras-ure, seiz-ure revers-al, acquitt-al

9.4.4  referents of syntactic objects   333 There are also a fair number of verbs doing double duty as result nouns; for example wound, bite, scratch, break, and plan. And some irregular correspondences, such as verb lose and result noun loss.

9.4.4  Referents of syntactic objects 9.4.4a A nominalization can refer to the target of the activity described by the verb it is derived from. For example: verb John discovered a cure for cancer Mary paid $50 for the book They propose to abolish all fees Tom inherits the business

object nominalization John’s discover-y was a cure for cancer Mary’s pay-ment for the book was $50 Their propos-al is to abolish all fees The business is Tom’s inherit-ance

Noun discover-y identifies the noun phrase in O (transitive object) function for verb discover. And similarly for the other illustrations. 9.4.4b Object nominalizations involve all seven of the key suffixes. Examples include: -ing -((a)t)ion -ment -a/ence/y -y -ure -al

weav-ing, knitt-ing, build-ing assump-tion, suppos-ition, select-ion, possess-ion pay-ment, announce-ment utter-ance, inherit-ance discover-y, deliver-y sign-ature, expend-iture propos-al

Build-ing can function both as an object noun (They intend to build the tallest build-ing in the world ) and as an extended activity noun (The build-ing of the cathedral has now been going on for seventy years). Announce-ment and deliver-y can both be unit of activity noun and object noun. There are also a fair number of irregular correspondences—what one knows is knowledge, a choice is chosen, an object lent is a loan, one bears a burden, and thinks thoughts. If John loses $100 at the races, then his loss is $100. And many verbs undertake double duty as their own object nouns, including plant, taste, smell, supply, pay, suspect (and see 2.8h).

334    9 making new nouns 9.4.4c It is unusual to have an object nominalization functioning alone as the O noun phrase of its verb. A semi-synonym of the verb is likely to be ­preferred—They will erect a build-ing there sounds better than They will build a build-ing there, and John is the one who put forward that propos-al is more likely to be heard than John is the one who proposed that propos-al. An object nominalization is only likely to function as head of an object noun phrase for its verb if it is sufficiently embellished. For example: I was just thinking the most beautiful thoughts They supplied us with ample supplies for a long voyage She selected an amazingly wide select-ion of fruits He just signed the most exquisite sign-ature I’ve ever seen Note that suffix -ee (9.3.14) derives object nouns with human reference, such as employ -ee and address-ee.

9.4.5 Locations 9.4.5a There are a number of verbal nominalizations which refer to a location. These show two basic possibilities: (a) The nominalization of a verb of rest may refer to a place of rest. John resides in the vicar-age implies John’s resid-ence is the vicar-age. And They located the corpse means They found the locat-ion of the corpse. Nominalizations place-ment and plac-ing have overlapping meanings relating to location. One can say The curator carefully placed the sculptures in the exhibition, or The curator saw to the careful plac-ing/place-ment of the sculptures in the exhibition. A variant on this pattern is shown by the verb enclose. If we hear The gardener enclosed the fruit trees, then we know The fruit trees are in an enclos -ure. (b) The nominalization of a verb of motion may refer to a place with respect to which the motion takes place. Instead of The plane is approaching the runway one could say The plane is on the approach to the runway. And alongside The duchess is entering the ballroom there is The duchess is at the entr-ance to the ballroom.

9.4.6 - ing   335 9.4.5b A few verbs do double duty as location nouns, including approach and exit, both with sense (b). And there are examples with six of the seven key suffixes (all save -al ): -ing -((a)t)ion -ment -a/ence/y -y -ure

dwell-ing (a), plac-ing (a) locat-ion (b) settle-ment (a), encamp-ment (a), place-ment (a) resid-ence (a), entr-ance (b) entr-y (b) enclos-ure (a)

Entr-y and entr-ance have overlapping meanings. They each function as a unit of activity noun (Mary’s entr-y| entr-ance onto the stage was breath-­ taking) and as a location noun (The entr-y| entr-ance to the building is difficult to find ). Note that exit does triple duty, as verb and as noun with both these senses. See also the discussion of such place nouns as brewery under (6) in 9.5.9c. We can now examine each of the seven key suffixes. 9.4.6 -ing 9.4.6a OE had a suffix with the form -ung or -ing which applied to verbs and derived nouns, generally referring to an extended activity. For example hēof-ung ‘lamentation, mourning, grieving’ from verb hēof-an ‘lament’. In ME times the form was rationalized to -ing /-iŋ/ and its scope extended. The suffix is used primarily with Germanic roots. The majority of -ing derivations describe an extended activity (throw-ing) or a state (enjoy-ing) although, as just illustrated, they can be used for a unit of activity. Further examples of this are meet-ing and sitt-ing (for example, of a parliament). And, as just seen, there are a handful of -ing forms referring to a state, a result, a syntactic object, or a location. There has been an extension of use for -ing whereby it may be added to certain nouns. Balloon-ing is going up in a balloon, blackberry-ing describes picking blackberries, and if someone says they are off soldier-ing it just means that they will have a spell as a soldier. That these derivations are nouns is shown by sentences such as Balloon-ing always scares me, ­Blackberry-ing often involves getting nasty scratches, and Soldier-ing is a dangerous occupation.

336    9 making new nouns Nominalizer -ing is fully productive. It can be added to a wide variety of verbs (and to some nouns), especially in the extended activity sense. 9.4.6b There is a quite different suffix -ing, which must be carefully distinguished from the nominalizer. This is used to mark the verb in a type of complement clause construction. The simple sentence JohnA throws [the ball]O accurately can be made into a complement clause through adding ’s to the subject NP and -ing to the verb, as in (1) MaryA admires [John’sA throw-ing [the ball]O accurately]COMPL.CLAUSE:O We can compare this with the use of nominalization throw-ing as head of an NP, modified by possessor John’s, adjective accurate and post-head element of the ball: (2) MaryA admires [John’s accurate throw-ing of the ball]NOUN.PHRASE:O There are a number of important grammatical differences between throw -ing as the verb of a complement clause, in (1), and throw-ing as a nominalization, in (2). These include (a fuller list is in Dixon 2010b: 376): (a) The fact that throw-ing is a noun in (2) is shown by its modification by adjective accurate, and the fact that throw-ing is functioning as a verb (within the complement clause) in (1) by its modification by adverb accurate-ly. (b) As a verb, in (1), throw-ing is directly followed by the object NP the ball. In (2), throw-ing is a noun and requires an of before the ball. (c) In a noun phrase a possessive element may be replaced by the. This applies in (2)—we can say Mary admires the accurate throwing of the ball. In (1) John’s cannot be replaced by the—one cannot say *Mary admires the throwing the ball accurately—showing that John’s throwing the ball accurately in (1) is not a noun phrase. Note that both (1) and (2) could be abbreviated to Mary admires John’s throwing, which is ambiguous between complement clause and nominalization interpretations. It is only when accurately or accurate, and the ball or of the ball are added that things become clear.

9.4.7  -ation, -ion, -ication, -tion, -ition   337 This possibility of ambiguity only arises for those verbs which use -ing as a nominalizer. Romance verb avoid employs -ance for its extended activity nominalization. In a complement clause it takes -ing, like every other verb. There is then a difference of form between the complement clause construction in (3) and the nominalization in (4). (3) WeA noted [John’sA carefully avoid-ing [the issue]O]COMPL.CLAUSE:O (4) WeA noted [John’s careful avoid-ance of the issue]NOUN.PHRASE:O 9.4.6c There is a further morphological use for suffix -ing. It can be added to a noun to indicate a collection of things, as in cloth-ing, floor-ing (materials to construct a floor), fenc-ing (materials for making a fence), and bedd-ing (sheets, blankets and suchlike associated with a bed). And -ing has several further syntactic uses, as in Hav-ing made his will, John jumped off the cliff and The buses were on strike, forc-ing us to walk. There is a summary of these in Dixon (2005a: 54–6). 9.4.7  -ation, -ion, -ication, -tion, -ition 9.4.7a This multi-form suffix relates to Latin -atio and learned French -ation. It always ends in /-ʃən/ and stress goes on the penultimate syllable. That is, it is on the first syllable of the suffix in its long form -ation /-'eiʃən/, as in i­ nstall-ation /ˌinstə'leiʃən/, and on the syllable preceding the suffix in its short form -ion /-ʃən/, as in exempt-ion /ig'zemp-ʃən/. There were many loans from French into ME. Some, like duration, are not analysable in English but many were accompanied by the underlying verb; for example, exempt and exempt-ion, imagine and imagin-ation, assume and assump-tion. And there were a few adjective/noun pairs—perfect and perfect-ion, absent and absent-ion. Typically, a set of corresponding forms crossing several word classes were taken over from Romance. These include: (1) Verbs ending in ify /ifai/: replace final /ai/ with -ication /-i'keiʃən/ as in justify /'dʒʌstifai/ and just-if-ication /ˌdʒʌstif-i'keiʃən/, pur-ify and purif-ication. (Note that justify is not analysable in English whereas pur-ify is; see Chapter 7.) On the pattern of these borrowed pairs, English then formed further nominalizations on other loans ending in -ify, including beauti-f-ication, and rectif-ication.

338    9 making new nouns (2) Verbs ending in -ize: simply add /'eiʃən/ as in organize /'ɔ:gənaiz/ and organiz-ation /ˌɔ:gənai'z-eiʃən/. Many new nominalizations were created on this pattern, including colon-iz-ation and civiliz-ation. (3) There were pairs of loans involving a verb ending in -ate and a noun in -ation (effectively replacing final /t/ on the verb with /ʃən/), as in nominate /'nominait/ and nominat-ion /ˌnɔmi'ne-ʃən/. Alongside the result noun ending in -ation there is often a concrete noun ending in -ant, describing some person or thing which is involved in the activity (see 9.3.4), or else an adjective ending in -ant (8.4.16). Repeating the forms from 2.5j: verb stimul-ate particip-ate lubric-ate emigr-ate

concrete noun stimul-ant particip-ant lubric-ant emigr-ant

result noun stimul-ation particip-ation lubric-ation emigr-ation

radi-ate domin-ate

adjective radi-ant domin-ant

radi-ation domin-ation

Note that, although there are no free forms stimul, particip, lubric, and so on, each of the endings does occur with free nouns—for example alienate (see 7.3.5) and inhabit-ant—and so can be recognized as a bona fide suffix. Looking now at verbs other than those ending in -ify,- ize or -ate, we find: (4) Ending in /t/, /d, or /s/. There are two possibilities here.

•

either (a) /-ʃən/ replaces final /t/, /d/ or /s/. For example, extend /eks'tend/ and extens-ion /eks'ten-ʃən/. This is shown in various ways in the spelling. Final t is sometimes retained, as in desert| desert-ion, and sometimes replaced by ss or s, as in permit| permiss-ion, convert| conversion. Final d is sometimes changed to t, as in intend| intent-ion, but more often replaced by s, as in extend| extens-ion. Final ss or s (before e) is generally retained, as in discuss| discuss-ion, confuse| confus-ion. or (b) add /-'iʃən/ or /-'eiʃən/, with vowel change and the normal stress shift. For example: suppose /sə'pouz/ and suppos-ition /ˌsʌpə'ziʃən/, converse /'kɔnvə:s/ and convers-ation /ˌkɔnvə's-eiʃən/. Also compete| compet-ition, oppose| oppos -ition, recite| reci-tation, accuse| accus-ation.

•

9.4.8  -ment   339 Note that process (a) applies for convert| convers-ion and (b) for converse| convers-ation. If the same process applied, the two nominalizations might coincide. (5) Ending in /m/, /b/, or /v/. Add /p-ʃən/ with vowel change, retaining /m/ but replacing /b/ or /v/. For example assume /ə'syu:m/ and assum-ption /ə'sʌmp-ʃən/. Also describe| descrip-tion, perceive| percep-tion. But note that observe| observ-ation is an exception, being of type (6). Some verbs ending in /v/ replace this by /'u:ʃən/, with vowel change and stress shift. For example, solve /sɔlv/ and solution /sə'lu:-ʃən/. Also resolve| resolu-tion (which comes from the same Romance root). (6) Ending in /l/ or /n/. Generally add /-eiʃən/, with vowel change and stress shift, as in install /in'stɔ:l/ and install-ation, /instə'l-eiʃən/. Also imagine| imagin-ation, resign| resign-ation, explain| explan-ation, console| consol-ation. (7) Verbs which used to end in /r/, but have now lost this in most dialects, include it before /-'eiʃən/ (as before other suffixes commencing with a vowel). For example declare, /dik'leə/ and declar-ation /ˌdeklə'reiʃən/. Also consider| consider-ation, admire| admi-ration, prepare| prepar-ation (note the vowel shift in each of these). 9.4.7b As seen above, this suffix derives state, unit of activity, result, object, and location nouns. It is almost restricted to Romance forms, being added (from the eighteenth century) to just a few Germanic verbs—flirt-ation, starv-ation and the rather cute bother-ation. Productivity is probably limited to replicating familiar patterns, such as applying the nominalizer to new verbs created with -ify or -ize (see 7.7). For example, from Trotski-fy and Reagan-ize could be derived Trotski-f-ication and Reagan-iz-ation. 9.4.8 -ment Suffix -ment is, like -((a)t)ion, of Romance origin—going back to Latin -mentum and French -ment—but the two suffixes differ markedly in character. We have seen that -((a)t)ion has several alternative forms and typically fuses with the base to which it is attached, engendering vowel change and stress shift. In contrast, -ment /-mənt/ (or sometimes just /-mn̩t/, with a syllabic /n̩/) is simply added to the base, generally with no effect upon vowel

340    9 making new nouns quality or stress (British pronunciation of advertise-ment is one of the few exceptions). And whereas -((a)t)ion is found almost exclusively with Romance words, -ment has been well extended to Germanic forms. ME borrowed from French pairs of words such as agree and agree-ment, commence and commence-ment, judge and judge-ment (and see 2.4e). It took no time at all for -ment to become a productive suffix in English, being added to further Romance verbs (such as enhance-ment) and to a good number of Germanic ones; for example, amaze-ment, settle-ment and bereave-ment. However, it is scarcely productive today. One feature of -ment is its use on verbs derived from adjectives and nouns by means of prefix en- or em- (see 7.3.3 and 11.4a), such as en-large-ment, encamp-ment, en-lighten-ment, and em -power-ment. Verbalizing prefix en-| emindicates ‘make be’ or ‘make have’ and then applying -ment produces a result noun. For example, en-large is ‘make large’ and en-large-ment is ‘the result of making large’. Em -power is ‘make have power’ and then em-power-ment is ‘the result of making have power’, in other words ‘having been accorded power’. As has been illustrated, -ment nominals may refer to state, unit of activity, result, object, and location (but probably not to extended activity). The expansive character of -ment is shown by its direct addition to a small number of nouns and adjectives. For example, devil-ment ‘activity befitting a devil’ and merri-ment ‘the state of being merry’ (although in fact this was originally based on the now obsolete verb merry, meaning ‘to be merry’). 9.4.9  -ance, -ence, -ancy, -ency 9.4.9a There are two nominalizers, one of which is -ance or -ence, pronounced /-əns/ or /-ns/, the other -ancy or -ency, pronounced /-ənsi/. Both evolved from Latin -āntia/-entia; some of the words in English were taken directly from Latin, others coming via French. The two suffixes have similar form and overlapping meanings. But they do differ a little, and can contrast. ME and early Modern English accepted pairs of loans which included these affixes. There were verb/noun pairs such as resist| resist-ance, assist| assist-ance, accept| accept-ance, and depend| depend-ence. And also pairs linking an adjective which ends in ent or ant with a derived noun, including efficient| efficiency, piquant| piquancy, and buoyant| buoyancy. 9.4.9b There are some pairs of nominalizations on the same base, one in -ance| -ence and the other in -ancy| -ency. The difference between them can roughly be characterized as follows:

9.4.9 - ance, -ence, -ancy, -ency   341 (I) -ance| -ence forms tend to refer to something extended in time, and they relate to a verb. (II) -ancy| -ency forms tend to refer to a result, and they relate to an adjective ending in -ant| -ent. This is illustrated in Table 9.2. A depend-ence is a continuing quality, as in The city’s depend-ence on tourism (it depends on tourism). In contrast, a depend-ency is a state or result; for example, if someone has a depend-ency on drugs (that is, they are depend-ent on drugs) this is a state which results from taking a lot of drugs. One might talk of John’s continual insist-ence, all through the meeting, that working hours should be increased (he kept insisting on it) or, speaking generally, of John’s habitual insist-ency (he is always being insist-ent). One can modify an -ance| -ence form (but not an -ancy| -ency one) with something like a degree of or quite a bit of, as in The city has a high degree of depend-ence on tourism and He exhibited quite a bit of insist-ence concerning that proposal. In contrast, one would be less likely to hear *a high degree of depend-ency or *quite a bit of insist-ency. (For discusssion of further pairs, see Jespersen 1942: 371.) There is no verb relating to perman-ence but the same difference applies. For example, This product kills weeds with a fair degree of perman-ence, and He is happy to have achieved perman-ency (he has been appointed to a ­perman-ent position). 9.4.9c Principles (I) and (II) help to explain some of the occurrences of -ance| -ence and -ancy| -ency, but there is also a good deal of arbitrariness involved. For example, the OED records instances of ignor-ance and

Table 9.2  Comparing -ance| -ence and -ancy| -ency nominalizations -ance|-ence nominalization ‘extended in time’

relating to verb

-ancy|-ency nominalization ‘result’

relating to adjective

depend-ence

depend

depend-ency

depend-ent

insist-ence

insist

insist-ency

insist-ent

consist-ence

consist

consist-ency

consist-ent

transcend-ence

transcend

transcend-ency

transcend-ent

inher-ence

inhere

inher-ency

inher-ent

perman-ence

perman-ency

perman-ent

342    9 making new nouns ignor-ancy, import-ance and import-ancy, benevol-ence and benevol-ency, relev-ance and relev-ancy, effici-ency and effic-ience, redund-ancy and ­redund-ance, poign -ancy and poign-ance. The item first listed for each pair is in favour today, with the other having fallen into disuse. And then there are disturb-ance, remembr-ance, utter-ance, presid-ency, and so on. Why do we say abs-ence, corresponding to adjective abs-ent, rather than *abs-ency? It appears that this is just the way the language developed. 9.4.9d Generally, the addition of -ance| -ence or -ancy| -ency involves no shift in stress or vowel change. There are a few exceptions, such as residence /'rezid-əns/ from reside /ri'zaid/. A number of derivations from a basic form ending in er, pronounced as /ə/, replace this by /r-əns/; for example, remember /ri'membə/ and remembr-ance /ri'membr -anz/. Then there are some where the base form ends in er, pronounced as /ə:/, and this is replaced with /(ə)rənzs/ with vowel change and stress shift, as in refer /ri'fə:/ and refer-ence /'refr-əns/. Note also prefer-ence, differ-ence, infer-ence, and entr-ance. As shown in 9.4.1–5, -ance| -ence covers all six of the semantic categories; the predominant use of -ancy| -ency is for result. Almost all instances of these nominalizations involve Romance bases. There are just a few with Germanic roots, including hindr-ance, ridd-ance and forbear-ance. Suffix -ance has been used for terminology in electrical engineering—imped-ance, induct-ance, react-ance. Neither suffix has any significant productivity today.

9.4.10  Contrasting -((a)t)ion and -ance| -ancy 9.4.10a In 9.4.7a there was illustration of result nominalizations in -((a)t)ion corresponding to verbs ending in -ate. For some (not all) of these there are also nominalizations in -ance| -ancy. We can repeat—and expand on—part of the tabulation under (3) of 9.4.7a: verb domin-ate radi-ate particip-ate

result noun domin-ation [domin-ancy] radi-ation [radi-ancy] particip-ation [particip-ancy]

extended activity noun domin-ance radi-ance particip-ance

9.4.12  -ure, -iture, -ature   343 Looking back at Table 9.2, none of the forms there have an -((a)t)ion derivation, and so use -ency for a result and -ence for a state which is extended in time. However, corresponding to each of domin-ate, radi-ate, and participate there is an -((a)t)ion form which refers to a result. Both -ance and -ancy forms are attested for the three verbs, but -ancy ones (in square brackets above) are little used, -((a)t)ion being preferred. However, the -ance nominalizations, referring to a state extended in time, have a fair degree of use. 9.4.10b I have tried to indicate some of the principles underlying the use of the -ance| -ence and -ancy| -ency nominalizers. However, these are at best rough-and-ready indicators, there being no shortage of exceptions to them. To mention just one example of this, corresponding to verb expect there is result nominalization expect-ation (for example, Our high expectations were not fulfilled ) and extended activity nominalization expect-ancy (There was an excited air of expectancy all through the counting of votes). From what has been said, we would predict that the latter form should be expect-ance. Both expect-ance and expect-ancy have been in use (from about 1600) but it is the -ancy form which has found greatest favour. Why? Because that is just the way the language community has chosen to go. 9.4.11 -y As discussed in 9.2.2, the Romance nominalizer -y /-i/ derives some state nouns from adjectives (for example, jealous-y). What is essentially the same suffix derives abstract nouns (of all of the semantic categories discussed in 9.4.1–5 except for state) from verbs—enquir-y, discover-y, and so on. Most of the verbs it attaches to end in re or er, /ə/. This is replaced by /-ri/ or /-əri/. Thus, from injure /'indʒə/ is derived injur-y /'indʒ(ə)r-i/, and from enter /'entə/ we get entr-y /'entr-i/. We also get victor-y from agent noun victor. Note that this is a quite different suffix from the Germanic -y which derives adjectives from nouns and verbs (8.2.8, 8.4.13), and also from the -y which has been used in Modern English to form diminutive and endearing forms (6.11.3). 9.4.12  -ure, -iture, -ature 9.4.12a Suffix -ure relates to Latin -ura and French -ure. There were a number of loans from French into late ME and early Modern English, often

344    9 making new nouns following the loan of the underlying verb; for example, press-ure after press, and clos-ure after close. On this basis, -ure could be recognized as a derivational suffix in English, and was indeed added to some existing Romance loan verbs, such as eras-ure from erase. But -ure never really became established as a productive suffix in English, and this is surely responsible for its range of phonological forms, and of meanings, which typically reflect those in the Romance source. 9.4.12b Although the orthographic form is -ure (or -iture or -ature), there is a range of pronunciations:

•

Most often, -ure follows a root ending in /z/, with this being replaced by /-ʒə/, as in expose /ik'spɔuz/ and expos-ure /ik'spou-ʒə/. Also pressure, enclos-ure, compos-ure, among others. With pleas-ure /'ple-ʒə/ from please /pli:z/ there is also vowel change. A few base forms end in /t/ and this is replaced by /tʃə/. Thus depart /di'pa:t/ and depart-ure /di'pa:-tʃə/. Also forfeit-ure, legislat-ure. The nominalization from fail /feil/ is fail-ure /'feil-yə/, here adding /-yə/. A longer form of the suffix, -iture /-itʃə/ occurs in expend-iture and invest-iture. And we find -ature in curv-ature and sign-ature (note the vowel change here).

• • •

9.4.12c Most -ure derivations refer to a result; for example, seiz-ure, clos-ure, mixt-ure. But, as illustrated in 9.4.1–5, -ure derivations occur in the semantic categories of state, unit of activity, object, and location. And there is more besides. For example, curva-ture is the property of having the geometrical shape curve, and legislat-ure is an assembly of people with the power to legislate. Although the predominant use of -ure is with verbs, there are a few derivations with nouns and with adjectives. For example, architect-ure, the profession of an architect (and the style and results of this), portrait-ure, the art of making portraits, and moist-ure. At first sight, mixt-ure might be thought to relate to verb mix. But then whence the -t-? The actual story is as follows. Late ME took over the French adjective mixte ‘mixed’, and mixt-ure was based on this. The adjective was then re-interpreted as past form of a verb, mix-ed, and verb mix created by back-formation (this was after mixt-ure was already in use).

9.4.15  -age   345 9.4.13 -al From Latin -alis have developed two suffixes -al. One derives adjectives from nouns (for example nation-al ) and was dealt with in 8.2.19. Here we discuss the -al suffix which derives nouns from verbs. It is pronounced either /-əl/, or just as /-l̩/, with a syllabic l; dismiss-al can be /'dismis-əl/ or /'dismisl̩/. The addition of this -al does not affect stress or vowel quality. ME had pairs of loans such as arrive and arriv-al, rehearse and rehearsal, leading to the establishment of -al as a nominalizing suffix in English. It was then added to further Romance loans (deni-al, recit-al, dispos-al, among many others) and to some Germanic verbs (for example, betroth-al, bestowal, and withdraw-al ). Most nouns derived with -al refer to results; for example reviv-al, surviv -al, and apprais-al. A number could be taken as either unit of activity or result; these include arriv-al, revers-al, acquitt-al. Then there is approv-al, a state noun, and propos-al, an object noun. This suffix is not nowadays productive.

9.4.14 -t One could recognize a suffix -t which serves to derive nouns from verbs. First, there are a few Germanic forms. Gift /gif-t/ and weight /wei-t/, both object nominalizations, come from give /giv/ (note the voicing assimilation of /v/ to /f/ before /t/) and weigh /wei/. Fligh-t /flai-t/ serves as a result nominalization for both fly /flai/ and flee /fli:/. Also thrive| thrif-t and think| though-t (here with a major vowel change). There are also a few Romance loans where the final consonant of a verb is replaced by /t/ to form an object or result noun—receive /ri'si:v/ and receip-t /ri'si:-t/, descend and descen-t, ascend and ascen-t, deceive and decei-t. The /t/ is simply added to complain, giving complain-t. There are more complex changes between bequeath and beques-t, and conquer and conques-t. Neither variety of this pseudo-suffix has ever been productive. Note that it is quite different from Germanic suffix -th (with by-form -t) which derives state nouns from adjectives (dep-th, heigh-t, and so on), and was discussed in 9.2.4.

9.4.15 -age Suffix -age /-idʒ/ comes from French -age. It is added to both verbs and nouns (as it was in an early stage of French). Loans into ME included hermit

346    9 making new nouns and hermit-age, baron and baron-age, use and us-age. These established -age as a suffix in English and it was soon added to many bases, mostly Romance but also some Germanic (for example, wharf-age). It has a number of senses. (1) Added to verbs, a result noun is derived. A break-age is what results from breaking something (All break-ages must be paid for!). Leak-age describes what escapes from something which is leaking. Similarly with spillage and wreck-age, among others. (2) Added to some nouns, -age derives a noun of quantity, often indicating a charge or tax. Some restaurants allow you to bring in your own wine, but charge a cork-age fee for each bottle (that is, for each cork). Wharf-age is the fee for using a wharf; post-age is how much you pay for having a letter or parcel conveyed through the post. Tonn-age has two senses both relating to quantity: the weight of a ship in tons, or a tax levied on each ton of goods. The literal sense of mile-age is a measure of the number of miles travelled (as shown by a mile-ometer, see 9.7c). It also has an extended sense, as in I’ve had a lot of mile-age out of this pair of jeans (I’ve worn them many times) and You ought to get considerable mile-age from that idea (you can present papers on it at many conferences). Suffix -age can also describe spatial extent, as in acre-age and pasture-age ‘portion of land available for pasturing livestock’. (3) A number of verbs combine the result and quantity senses. Us-age can refer to the extent to which something was used (His electricity us-age was very high). If something shrinks it produces a degree of shrink-age (Allow for ten per cent shrink-age on washing the garment for the first time). Adjective short can derive a noun with similar sense; if the amount available of some commodity is less than that needed, then there is a short-age of it (an alternative description is with the compound short-fall ). (4) By analogy with hermit-age ‘the dwelling of a hermit’, we have such derivations as orphan-age, vicar-age, parson-age, and also anchor-age ‘a place where a vessel may anchor’. (5) The suffix is added to a noun which portrays a person in a certain role and derives an abstract noun describing the role. Someone who is a vassal is in the condition of vassal-age. A baron is a member of the baron-age. A rather different sense attaches to parent-age; if someone enquires concerning your parent-age they want to know what sort of people your parents are, in terms of social class, ethnic origin, wealth, and education.

9.5.1  -hood   347 There are other meanings as well: a pilgrim may go on a pilgrim-age to some holy place. When a number of people assemble together they form an ­assembl-age. A person-age is a person of significant social importance. And more besides. Nevertheless, suffix -age is scarcely productive today. For a fuller account of this multi-faceted suffix, see Jespersen (1942: 436–8), Marchand (1969: 234–6), and Gadde (1910: 50–69).

9.5 Abstract nouns from concrete nouns We can now examine a set of suffixes whose main function is to be added to concrete nouns (sometimes to a few other items as well), deriving an abstract noun.

9.5.1 -hood Germanic suffix -hood /-hud/ (which does not affect stress placement) emanates from OE noun hā d ‘rank condition, character, nature’. The noun has dropped out of use, but the suffix to which it gave rise, within OE, continues. From the earliest times we have such derivations as child-hood and priest-hood, with many more being added in ME and modern times, based on nouns of all genetic complexions. It has a degree of productivity today. The meaning is ‘role of’ or ‘status of’. Basically, -hood is found on: (a) Many kin terms. Father-hood describes the state of being a father (that is, of having children) as when one hears of the joys of father-hood. Similarly for mother-hood, daughter-hood, and so on. Sister-hood and brother-hood have a secondary sense, referring to a group of likeminded women or men. (b) Terms describing marital status, such a spinster-hood, bachelor-hood, and widow-hood. (c) Terms describing stage of life, as in He was remembering a story learnt in child-hood. Also baby-hood, adult-hood, boy-hood. Man-hood may describe the state of being a man, or manly characteristics, or virility (or it may be used to describe the male sexual organs). (d) Terms describing social or political position, such as knight-hood, saint-hood, and waitress-hood. When a colonial territory obtains

348    9 making new nouns independence from its erstwhile ‘mother country’, it has achieved state-hood (or nation-hood ). (e) Derivations under (d) may have a collective sense, as in John belongs to the priest-hood. Relating to this, there is neighbour-hood, describing a certain settled area and all the people in it. There are a handful of derivations from adjectives, all rather irregular. False-hood does not refer to some state of being false, but rather to a particular false statement or idea. Likeli-hood is a noun with essentially the same meaning as adjective likely—one could say either He is very likely to come today, or There is a strong likeli-hood that he will come today. Both of these go back to ME, whereas hardi-hood was coined a little later. 9.5.2 -ship In OE, scipe functioned as a noun ‘dignity, office’ and gave rise to suffix -scipe, as in hla¯ ford-scipe ‘lord-ship’, tu¯n-scipe ‘the inhabitants of a tu¯n “manor, village”’, and fre¯ond-scipe ‘friend-ship’. Only the suffix survives today, as -ship /-ʃip/, which does not affect stress placement. Whereas in OE it could be added to nouns and adjectives (hard-ship is a relic from this) in Modern English it is used almost exclusively with nouns. There are two basic meanings, both still used productively: (a) A reciprocal relationship. If two people are friends, they are in a relationship of friend-ship (for example, Nothing could ever mar our friend-ship). Similarly with comrade-ship, acquaintance-ship, fellowship, companion-ship, partner-ship, kin-ship (as in Kin-ship ties are the strongest of all). (b) An office, and the responsibilities associated with it. For example, I never realized that the duties of the principal-ship would be so onerous. The derivation king-ship refers to the job of being a king and all its obligations. Also trustee-ship, guardian-ship, apprentice-ship, owner-ship, librarian-ship, citizen-ship, dictator-ship, and many more. 9.5.3  -manship A further meaning of -ship is ‘skill associated with’, as in musician-ship ‘the skill of being a musician’ and scholar-ship. The great majority of i­nstances

9.5.4  -dom    349 of this sense follow a noun derived with -man (as discussed in 3.3f and 9.3.13). For example sea-man-ship, horse-man-ship, crafts-man-ship, swordsman-ship, sales-man-ship, and states-man-ship. Note that these all relate to set (3) in 3.3f, in which the -man has a reduced vowel /-mən/, not to those in set (2) where it is /-ˌman/. (That is, we do not get *super-man-ship.) Around 1950, -manship /-mənʃip/ emerged as a suffix in its own right, rather than just as a sequence of -man plus -ship. Someone who is good at getting grants for research projects could be said to display excellent grants-manship. English humorist Stephen Potter coined games-manship, life-­manship, and one-up-manship, all of which have achieved wide currency. (The point is that there are no intermediate derivatives, *grants-man, *one-up-man, and so on.) New formations abound. For example, brink-manship is the skill of going up to—but not quite over—the brink, in some dangerous situation. In 3.3f it was noted that -woman can substitute for -man in words of set (3). Alongside horse-man and swords-man there are horse-woman and swords-woman. However the addition of -ship to the -woman derivations produces forms which sound clumsy and are scarcely acceptable.

9.5.4  -dom OE had noun dō m ‘judgment, authority’ and suffix -dō m. This has developed into modern suffix -dom /-dəm/, which does not affect stress placement. It is used with just a few adjectives and then indicates ‘state of being’; free-dom and wis(e)-dom go back to OE. Bore-dom, based on a verb, was coined during the nineteenth century. The major use of this suffix is with concrete nouns: (a) It may be added to specific titles and then refers to the power and authority associated with the title, plus the related administrative infrastructure and the territory ruled over. For example, king-dom, dukedom, sheikh-dom. (b) With a general term, the suffix has a collective sense: official-dom denotes the entire galaxy of officials, and Christen-dom (the irregularity of form goes back to OE) describes the world-wide community of Christians. Similarly, serf-dom, bachelor-dom, and spinster-dom. This suffix is used a great deal in many new words, often rather jocular—for example, loafer-dom and topsy-turvy-dom. Many of these have a short life.

350    9 making new nouns Recent coinages that have become fairly established include gangster-dom ‘the intimate world of gangsters’ and star-dom ‘the life-style of being a recognized star’. Bauer (2001: 162–72) has an instructive history of -dom across Germanic languages. The multitude of nonce formations are illustrated in Wentworth (1941) and Mencken (1936: 178, 1945: 358–9).

9.5.5  Contrasting -hood and -ship with -dom Just a few nouns may accept either -hood or -dom. For example:

• •

bachelor-hood—the state a man is in when not married bachelor-dom—the collection of all unmarried men, and the sort of life-style they follow

Spinster-hood and spinster-dom contrast in similar fashion. In the OED we find an evocative quotation from the Illustrated London News of 1899: The winning ways of ‘catdom’ and ‘kittenhood’. Kitten refers to a stage of life, and so takes -hood, while cat simply denotes a type of animal and bears the ‘state of being’ suffix -dom. Looking now at -ship and -dom, Shakespeare’s Richard III (Act 5, Scene 7) famously exclaimed (after his horse had been killed in battle): A horse! A horse! My king-dom for a horse He was offering to exchange all the power and authority and wealth of his king-dom for a horse. It would not have been felicitous to substitute kingship here, since this has a quite different meaning—the duties and responsibilities of being a king.

9.5.6 -ate English has four suffixes written as -ate. Verbalizing -ate pronounced /-eit/ (for example, liquid-ate) was discussed in 7.3.5. The rather minor adjectivizing suffix -ate, pronounced /-ət/ (affection-ate) was mentioned in 8.2.44. A suffix -ate /-eit/ is much used in Chemistry (as in acetate, carbon-ate). ME borrowed from French a wide range of nouns ending in -ate (reflecting Latin -ātus, -ātum), most of which are not analysable within English. For instance, magistrate, curate, prelate. Also pairs such as consul and consul -ate,

9.5.7  -ism   351 which established -ate, pronounced /-ət/, as a nominalizing suffix in English (it does not affect stress placement). One of its senses is to indicate a qualification or position—Our most recent doctor obtained his doctor-ate for a thesis on English morphology and His father was director of three companies, and John has finally been appointed to his first director-ate. Director-ate can also have a collective meaning, referring to a board of directors. The derivation consul-ate can refer to the position of consul, and also the premises where the consul and their staff work. A protector-ate is the territory administered by a protector. Elector-ate refers to all the people registered as electors for a particular purpose, and also the region they live in. The noun professor /prə'fesə/ was borrowed from French at the end of the fourteenth century. In the 1800s, two abstract nouns were formed from it, and today they have contrastive meanings:

•

a professor-ate /prə'fessər-ət/ describes the position of professor—Mary Jones has just been appointed to a professor-ate (an alternative term is professor-ship) professor-iate /ˌprɔfi'sɔ:r-iət/ describes the body of professors in a ­university—The professor-iate meets each Friday afternoon

•

An interesting question to ask is: how did the -i- get into professor-i-ate? Just possibly, the nominalization was formed on adjective professor-ial /ˌprəfe'sɔ:riəl/ (and this would also explain the stress shift in professor-iate). There are a limited number of nominalizations ending in -ate. It may be significant to note that in almost all instances the root has two or more syllables and the -ate is preceded by an -r or -l. We do find alderman-ate, after an -n. But there appear to be no productive derivations in English onto a root ending in a stop. This may explain why emir-ate has been coined for the office of an emir, and the territory governed, but there is no corresponding *sheikh-ate (because sheikh is monosyllabic and ends in a stop). We have instead to use sheikh-dom.

9.5.7 -ism 9.5.7a Like agentive suffix -ist (9.3.6) and verbalizer -ize (Chapter 7), suffix -ism goes back to Greek. Indeed triplets such as baptism, baptist, and

352    9 making new nouns baptize directly reflect the Greek originals. Ending -ismō s in Greek gave rise to Latin -ismus, French -isme, and thence English -ism /-izm/ or /-izəm/. Romance loans such as pagan and pagan-ism, catholic and catholic-ism, established the suffix in English and it was soon added to many Romance forms, and also Germanic ones (for example, heathen-ism and tru-ism). This derivation generally does not affect stress or vowel quality; it does cause a preceding /k/ to assimilate to /s/, as in sceptic /'skeptik/, sceptic-ism /'skeptis -izəm/. 9.5.7b Abstract nouns derived with -ism have a variety of meanings. They can describe a doctrine or ideology (spirit-ual-ism, Darwin-ism, anarch-ism), a practice (critic-ism), a habit (nud-ism), a quality (hero-ism), and more besides; see the examples below. The suffix has a wide functional range. It can be added to nouns of various types, and to simple and derived adjectives. It is useful to systematize types of -ism derivations in conjunction with related -ist and -ize forms. In the discussion below, the -ism form is always coded as (c), with a related adjective as (b), a noun describing a type of person as (d), a verb in -ize as (e), and other nouns as (a). (1) -ism derivation based on a derived adjective (b). We begin with: (a) Concrete noun: industry, a type of manufacturing enterprise. From this is derived: (b) Adjective: industri-al, describing anything related to industry (for example machinery, chemicals, nations). Three derivations are based on this adjective: (c) Abstract noun: industri-al-ism, an endeavour on industrial principles. (d) Noun describing a type of person: industri-al-ist, someone involved in organizing an industrial enterprise. (e) Transitive verb: industri-al-ize, make an enterprise industrial.

9.5.7  -ism   353 A further example of the same type begins with: (a) Concrete noun: tradition, a custom or belief from times past. From this is derived: (b) Adjective: tradition-al, describing any aspect of tradition (for example, songs, dress, teaching methods). Three derivations are based on this adjective: (c) Abstract noun: tradition-al-ism, an adherence to traditional behaviour. (d) Noun describing a type of person: tradition-al-ist, someone who follows traditional principles. (e) Transitive verb: tradition-al-ize, make something more traditional. Others include idea-l-ism, romant-ic-ism, nation-al-ism, commerc-ial-ism, spirit-ual-ism. Also many based on the names of countries or areas, as in Africa-n-ism. In (1) we had (c), (d), and (e) based on adjective (b). An alternative is to have (d) as the basis, with (c) and (e) derived from it. (2) -ism derivation based on a noun which describes a type of person, (d). An example is: (d) Noun describing a type of person: critic, a person who judges and evaluates. Two derivations are based on this noun: (c) Abstract noun: critic-ism, a judgement or evaluation. (e) Transitive verb: critic-ize, make a critical assessment. Others include vandal-ism, mystic-ism, cynic-ism, cannibal-ism. And there are a number which lack (e), an -ize verb: agnostic-ism, pagan-ism, lesbian-ism, absentee-ism, fanatic-ism, despot-ism, pauper-ism, hero-ism, Nazi-ism.

354    9 making new nouns We also get derivations from (d) athlete, to (b) athlet-ic and then to (c) athlet-ic-ism. This is a sort of blend of types (1) and (2). A third type has an -ism noun, (c), based directly on a noun, (a), without any intervening adjective stage, as there was in (1). (3) -ism derivations based on other types of noun (a). An example is: (a) Noun: anarchy, a society in which little attention is paid to codified rules and laws. Two derivations are based on this noun: (c) Abstract noun: anarch-ism, a doctrine that rules and laws should be abolished. (d) Noun describing a type of person: anarch-ist, a person who espouses anarchism. Others include cub-ism, impression-ism, tour-ism, journal-ism, and also many based on the names of important political and religions thinkers: Platonism, Buddh-ism, Marx-ism, and so on. A variant on this pattern also has an -ize derivation, (e). For example: (a) terror, (c) terror-ism, (d) terror-ist, (e) terror-ize. A fourth type has an -ism noun, (c), based, not on a derived adjective as in (1), but on a simple adjective, (b). (4) -ism derivations based on an unanalysable adjective, (b). An example is: (b) Adjective: nude, not wearing any clothes. Two derivations are based on this noun: (c) Abstract noun: nud-ism, the habit of not wearing clothes. (d) Noun describing a type of person: nud-ist, someone who practises nudism.

9.5.7  -ism   355 Similar derivations include femin-ism, archa-ism (in which the final -ic is omitted from archaic). Some simple adjectives add -ism without there being a corresponding -ist derivation. For example, tru-ism. Others also have an -ize derivation, (e). For example: (b) secular, (c) secular-ism, (d), secular-ist, (e) secular-ize; also modern-ism, popular-ism. In summary, an abstract noun in -ism is derived from a noun describing a type of person in (2), from other types of noun in (3), from a simple adjective in (4), and in (1) from an adjective which is itself derived from a noun. There are also -ism forms where the base is not a word in English. A number have -ist and sometimes also -ize correspondents:

•

hypnotism, (d) hypnotist, (e) hypnotize. Also exorcism and plagiarism (although there was originally a noun plagiary, nowadays replaced by plagiarism). pessimism, (d) pessimist; also optimism, hedonism, Methodism, communism.

•

This account does not cover every nuance of -ism derivation. But it should serve as an outline of the main features. 9.5.7c In some instances, the meaning of an -ism noun is slightly different from what might be expected from the meaning of the base. For example, journal-ism (a direct loan from French) describes contributing to a newspaper (   journal in French) or other popular media outlet, rather than to keeping a daily journal, or diary, or publishing in an academic journal. Most differences in meaning are ad hoc, and are explained in good dictionaries. However there are some recurrent patterns. One concerns a recent variant of pattern (2) where the base noun is race, sex, or age (or height or size or weight, etc.). The -ism derivation has a pejorative meaning, stating belief in the superiority of and/or discrimination in favour of one type of race, sex, or age bracket, and so on. In another variant of pattern (2), -ism is added to the name of a substance and indicates addiction to it, as with alcohol-ism and cocaine-ism (here there is no corresponding -ist form). A subtype of pattern (1) involves -ism derivations based on adjectives, which are in turn based on names of places, such as America-n-ism. These can refer to a custom or habit typical of the place or else to some speech

356    9 making new nouns expression. For example, I just love it when you tell me to ‘ring somebody’, it’s such a cute English-ism. Similarly, colloquial-ism generally refers to some word or phrase from colloquial speech. 9.5.7d The -ism suffix is today highly productive, being used in all sorts of ad hoc derivations. And it is freely added to multi-word sequences, such as bigshot-ism, know-nothing-ism, red-tape-ism, and devil-may-care-ism. For several centuries, a noun ism has been used as a cover term for an array of doctrines all of whose names end in -ism, The OED quotes, from 1688: He was the great Hieroglyphick of Jesuitism, Puritanism, Quaquerism, and of all other Isms from Schism. And from 1811: He is nothing—no ‘-ist’, professes no ‘-ism’ but superbism and irrationalism. Also from the OED (1841), there is: His name deserves to be ism-ated (i.e. his ideas are so important that an abstract noun should be based upon his name, similar to Darwin-ism). At the end of 3.2k, we noted the existence of a number of books which contain extensive listings of ‘isms’ and ‘ologies’. 9.5.8 -ology Suffix -ology /-'ɔlədʒi/ attracts primary stress to its first syllable, as in climate /'klaimət/ and climat-ology /ˌklaimə't-ɔlədʒi/. It reflects Greek -ologia¯ . English words come either directly from Greek, or through Latin -ologia or French -ologie. The meaning is straightforward: ‘the study of, or the science of’. It is often followed by -ist, indicating a practitioner of that study or science, as in climat-olog-ist. Early loans into ME were unanalysable—theology, philology, mythology (the simple noun myth did not enter English until the 1800s). Later came such as chronology, geology, ecology. Demon, a loan into ME, was followed, in the late sixteenth century by ­demon-ology ‘the study of demons’, establishing -ology as a suffix in English. From the late 1700s there were new derivations, mostly on Greek roots— ­insect-ology, method-ology, climat-ology and (around 1900) music-ology. The suffix was extended to Romance roots in crimin-ology (1950s) and sexology (1950s) and to a Germanic base in gemm -ology (1960s). Suffix -ology is today a favourite for ad hoc formations; for example, folklore-ology, clock-ology, and hut-ology ‘a study of the scientific importance of huts’. The Chase Bag Company of Chicago named their house journal

9.5.9  -ery, -ry   357 bag-ology (Mencken 1945: 371), presumably ‘the art of making bags’. As with ism, the noun ology has been adopted as a cover term for abstract nouns ending in -ology (see 3.2k). 9.5.9  -ery, -ry 9.5.9a The suffix -ery ~ -ry has heterogeneous origins, conditioned realizations, and disparate meanings. It relates to two endings in Old French. The first was -erie, as in ­juel-erie ‘jewel(le)ry’, corresponding to juel ‘jewel’. The second involved addition of -ie to an agent noun ending in -ier; thus boucher-ie ‘butcher-y’ from bouchier ‘butcher’. ME accepted loan pairs like this, and -(e)ry became established as a derivational suffix in English, applying to many nouns—of all genetic hues—and to just a few adjectives. 9.5.9b There are two spellings, and corresponding pronunciations, of the suffix, depending upon the nature of the base to which it is attached. (Note that it generally does not affect vowel quality or stress placement.) (i) After a closed monosyllable, ending in any consonant, the suffix is written -ery and pronounced /-əri/. For instance, pigg-ery /'pig-əri/, and cann-ery /'kan-əri/. Also snobb-ery, bind-ery, rock-ery, droll-ery, creamery, nunn-ery, nurs-ery, fish-ery, pott-ery. (ii) After a form of two (or perhaps more) syllables, with initial stress and ending in /l/, /n/, /t/, or /d/, the suffix is generally written -ry and pronounced /-ri/. For example, yeoman-ry /'youmən-ri/, bandit-ry /'bandit-ri/, wizard-ry /'wizəd-ri/. We also get /-ri/ after a word written with final r (pronounced when word-final in only some dialects), as in soldie(r)-ry /'sɔldʒə-ri/. After a root ending in an unstressed vowel plus /l/, the unstressed vowel may be dropped; for example, from devil /'devil/ we get devil-ry /'devl-ri/. The -(e)ry form based on jewel is interesting. Noun jewel can be pronounced either (a) as a single syllable with a diphthong, /dʒu:əl/, or (b) as a disyllabic form with vowels separated by semi-vowel /w/, /'dʒuwəl/. There are two forms of the -(e)ry derivation. Jewel-ry /'dʒuwəl-ri/—prevalent in America—adds /-ri/ to the disyllabic form, (b), while jewell-ery /'dʒu:əl-əri/—­encountered most in Britain—adds /-əri/ to the monosyllabic form, (a).

358    9 making new nouns There are a few exceptional instances. Generally, polysyllabic words ending in something other than /n/, /l/, /t/, or /d/ do not have -(e)ry derivations. Savage /'savidʒ/ does have one, and here the suffix is -ery: savag-ery /'savidʒəri/. Jew /dʒu:/ is a rare example of an open monosyllable taking -(e)ry. Although written jew-ry, the form of the suffix is /-əri/ (as with closed monosyllables), plus shortening of the root vowel: /'dʒu-əri/. 9.5.9c As previewed in 2.3d, -(e)ry has a range of somewhat disparate meanings. (1) Added to a noun referring to a person exhibiting a certain character, it derives a noun describing the associated quality or activity. For example, bigot-ry, exhibiting intolerance of opinions other than one’s own, as a bigot does, and wizard-ry, a clever achievement by means not understood, such as a wizard might accomplish. Also snobb-ery, devil-ry, knave-ry, drudg-ery, pedant-ry, savag-ery, and more. (2) Describing the occupation or condition of a person described by the root noun—dentist-ry is the profession of being a dentist, and slav-ery is the condition of being a slave. Also cook-ery, artist-ry, outlaw-ry. (3) Added to a noun describing a type of job, -(e)ry derives a noun describing a collection of people, all of whom have that job. They include soldie-ry ‘a body of soldiers’ and also yeoman-ry, peasant-ry, citizenry, and constabul-ary (a derivation which does involve vowel and stress changes). Jew-ry, describing all those with a certain ethnic affiliation, also belongs here. (4) Similar to (3), but referring to a collection of things—weapon-ry, gadget-ry, jewel(l-e)ry, toilet-ries, ballad-ry. Also shrubb-ery, rock-ery, and (based on an adjective) green-ery, ‘an array of living plants, which are green in colour’. (5) There are a few -(e)ry derivations based on an adjective, then describing the state relating to the adjective—brav-ery, gallant-ry, prud-ery. Pleasant-ry has a rather different meaning: ‘a remark which is of a pleasant nature’. (6) A common use of -(e)ry—and perhaps the only sense which is now productive—is to derive a noun describing a place in which referents of the base noun live or are kept. Thus, nunn-ery, pigg-ery, catt-ery, swann-ery.

9.5.10  death and disease: - icide and - itis   359 There are forms which end in ery and describe where a particular thing is made; they include bindery, brewery, bakery, printery. There are two ways of analysing these. Either by saying that -ery is added to a verb (and these would be the only examples of this)—bind-ery, brew-ery, etc. Or by adding -y (see 9.4.5) to the agentive noun—bind-er-y, brew-er-y, etc. Deciding between these alternatives is basically a matter of personal preference. (7) We also find forgery, a result nominalization, which could be analysed as either forg-ery or forg-er-y. Similarly for robbery and mockery (and perhaps also bribery and embroidery). 9.5.9d Jespersen (1942: 346–7) comments: ‘Though -(e)ry cannot compete with -ment, -age and -ness with regard to popularity, it is still used rather freely in nonce formations, even from phrases.’ He quotes: She kept a [tamecatt]-ery of adoring young men. The productive use of this suffix centres on the locational sense, (6). An Englishman might frequent the local booz-er, while an American could repair to a grogg-ery. Or to an eat-ery, cak-ery, doughnut-ery, or breakfast-ry, or to a car -wash-ery, or even to a stomp-ery ‘dancing school’ (Mencken 1945: 348–9). A popular place of eating in New Zealand is a ­[fish-and-chipp]-ery. And so it goes on. 9.5.10  Death and disease: -icide and -itis 9.5.10a Suffix -icide /-isaid/ means ‘killer of’ or, more often, ‘the act of killing of’. It relates to Latin -icı¯ da and French -icide. Loans into late ME and early Modern English were unanalysable: homicide, fratricide, matricide, suicide, regicide (‘killing a monarch’), and so on. We then had infant-icide /in'fant-isaid/ alongside infant /'infənt/. Productive derivation began in the late nineteenth century, then just meaning ‘substance which kills’. Thus, fungi-icide ‘a chemical which kills fungi’, herb-icide ‘something which kills certain plants, particularly weeds’, germ-icide, insect-icide, pest-icide, and sperm-icide (a contraceptive cream which kills sperms). (Unlike with infant-icide, these modern uses of -icide do not affect stress or vowel quality.) 9.5.10b Suffix -itis /-'aitis/ relates to Greek -itis and refers to a disease, particularly of an inflammatory nature. Terms were borrowed into—or coined

360    9 making new nouns within—the medical Latin used by English doctors. Many are not analysable within English, including arthritis, peritonitis, meningitis. However some are: alongside tonsil /'tɔnsl̩/ and sinus /'sainəs/ there are tonsil-itis /ˌtɔnsi'l-aitis/ and sinus-itis /ˌsaini's -aitis/. Note that the suffix attracts primary stress to its first syllable, with secondary replacing primary stress in the root form. In recent years, -itis has come into popular use as a disparaging comment on some addiction considered by the speaker to be unhealthy—for example, golf-itis, crossword-puzzle-itis, let-George-do-it-itis, idle-itis (it is interesting that idle-itis sounds more euphonious, and is more used, than lazy-itis). In the preface to a recent monograph (Dixon 2010a: xiii), we read: ‘There is today a fashion in linguistics—and no doubt in other disciplines as well— of what can be called “quotationitis”. That is, attempting to cite every single thing published on or around a topic, irrespective of its quality or direct relevance.’ Instructive examples of nonce uses of -itis are in Mencken (1936: 179, 1945: 364) and in the OED entry for -itis.

9.6 More on location Nominalizations referring to location and place are produced by a variety of derivational suffixes. Recapitulating, we have had:

• •

One sense of -er (9.3.1c) derives names of places, including din-er and sleep-er. Six of the seven ‘key suffixes’ of 9.4 include among their many senses some locations (9.4.5):



° -ing, as in dwell-ing, plac-ing ° -((a)t)ion, as in locat-ion ° -ment, as in settle-ment, encamp-ment, place-ment ° -a| ence, as in resid-ence, entr-ance ° -y, as in entr-y ° -ure, as in enclos-ure.

• •

Suffix -age (9.4.15) forms some place nouns, such as hermit-age, ­orphan-age, and anchor-age. One sense of -ate (9.5.6) can derive a noun referring to a place, as with consul-ate, protector-ate, and elector-ate.

9.6.1  -(at)ory, -orium, -ary and -arium   361

•

The multi-faceted suffix -(e)ry (9.5.9) derives nouns referring to places, such as nunn-ery and pigg-ery.

The unprincipled distribution of these locational derivations is exemplified by the nomenclature for residences of a Church of England minister:

• • •

If he is a parson or vicar, he lives in a parson-age or vicar-age If he is a rector, he lives in a rector-y If he is a dean, he lives in a dean-ery

Rector-y and parson-age were borrowed entire from French or Latin. The other two were created within English, vicar-age by analogy with parsonage, and dean-ery through adding the productive suffix -(e)ry to dean. And there are four further suffixes involved in the making of designations of places. 9.6.1 -(at)ory, -orium, -ary and -arium In 8.4.14 there was discussion of Romance suffix -(at)ory /-(ət)əri/, which derives adjectives from verbs. It is also used in some names of places. The French loan dormitory is not analysable in English, but we do find suffix -(at)ory in reform-atory, deposit-ory, and observ-atory (note that these are derived from verbs). There are related loans ending in -orium (reflecting Latin o¯rium); for example auditorium and emporium. In the late nineteenth century a term was needed for a place where cremations are performed. Both cremat-ory /'kremət-əri/ and cremat-orium /ˌkreme't-ɔ:riəm/ came into use, more-or-less simultaneously, the latter employing -orium /-'ɔ:riəm/ as a suffix. (Note the difference in stress and vowel quality from verb cremate /kri'meit.) Suffix -ary /-(ə)ri/, of similar origin, is typically added to nouns, deriving many adjectives (for example, document-ary, see 8.2.39) and some nouns referring to people (concession-ary, 9.3.12b). There are also nouns ending in ary which describe places—unanalysable loans, such as seminary, and also dispens-ary (based on verb dispense) for a place where medicines are dispensed, and gran-ary /'gran-əri/ for a place where grain /grein/ is stored (note the difference in vowel quality here). Parallel to -(at)ory and -orium, there is—alongside -ary—a longer form -arium /-'eəriəm/ (from Latin -ā rium). In the 1700s there were coined

362    9 making new nouns h­ erb-arium /hə:'b-eəriəm/, for a place with a collection of dried herbs and other plants, and planet-arium /plani't-eəriəm/, for a building which hosts models of planets and stars. Two hundred years later came ocean-arium for a salt-water aquarium. It will be seen that both -orium /-'ɔ:riəm/ and -arium /-'eəriəm/ have primary stress on their first syllable. These suffixes are continually being used anew, especially in the USA. Mencken (1936: 179, 1945: 354–5) mentions establishments called nicknack-atory, beaut-orium, motor-torium, abort-arium (a hospital specializing in abortions), and many more.

9.7 Bits and bobs: -(i)ana, -ade, -ometer, -in We can round off the chapter with a handful of rather minor noun-deriving suffixes. 9.7a Suffixes -(a)n and -ian, which derive adjectives from proper names, are discussed in 8.3.8–9—giving America-n, Shakespear-ian, and the like. These can have the further addition of -a, then producing a noun which can be used of various things associated with the referent of the proper name. Victoria-na may refer to furniture, or just a style of fashion, associated with that queen’s reign. America-na might be used of a collection of maps and documents relating to the USA, Shakespear-iana may refer to anecdotes or literary studies about the playwright, Churchill-iana could describe a collection of the statesman’s pithy sayings. The suffix -(i)ana /-(i)a:nə/ can be traced back to Latin -ana. Note that whereas -(a)n and -ian do not affect stress, adding an -a causes stress to move to the second last syllable; compare America(n) /ə'merika(n)/ with Americana /əˌmeri'ka:-nə/. There is mention in 8.3.7 of a special sense of suffix -ese /-'iz/ (or perhaps it should be regarded as a separate suffix) which can be added to the names of places, people and professions, and describes a style of language associated with them. Thus Glasgow-ese, Johnson-ese, and official-ese. 9.7b There are a number of words in English ending with ade (for example, tirade) which relate to Latin ending a¯ ta or its descendants in modern

9.7 bits and bobs: - (i)ana, -ade, -ometer, -in   363 Romance languages. Just a few appear to be analysable, but with varying meanings. Block does double duty as noun and as verb and so does blockade, describing the complete blocking of, say, a road or a port. Cannonade describes a period of heavy cannon fire. And escapade, ‘a daring adventure’, does not really relate to escape. The lack of a recurrent semantic element here militates against the recognition of ade as a suffix. There is a suffix -ade /-eid/ which is productive and means ‘drink made from the juice of a fruit’. Lemon-ade /'lemən-eid//was borrowed (from Spanish or French) in the seventeenth century. It was followed a hundred years later by orange-ade. The suffix is now available for use as required: we get ginger-ade, pineapple-ade, and more. (As well as brand names of ‘energy drinks’ such as Power-ade and Gator-ade.) 9.7c Greek noun mētron ‘measure’ was taken into Latin and then French, where it developed into a suffix meaning ‘quantity of’. From the seventeenth century on, English has had a fair number of scientific terms ending in meter (some loans, others created on a classical model), including thermometer, barometer, chronometer, pedometer. None of these is analysable within English. However, we also have speed-ometer /spi'd-ɔmitə/, which measures the speed /spi:d/, at which one is travelling, and mil-ometer /mai'l -ɔmitə/, which measures the number of miles /mailz/, travelled. On the basis of these we can recognize a suffix -ometer /-'ɔmitə/, which attracts stress to its first syllable. Suppose that there were a story about some strange place where the unit of measure is a glaut /glaut/; one would naturally expect an instrument which measures the number of glauts traversed to be called a glaut-ometer /glau't-ɔmitə/. When, in the early 1970s, measures in Australia were decimalized, one wondered what to call the gauge which had replaced the milometer on the dashboard, and now measured the number of kilometres travelled. ­*Kilometr-ometer would be long and ugly. In fact, the name adopted was odometer, a loan in the 1790s from French (based on Greek hodos ‘way’), another term which is not analysable within English. A household will measure its consumption of utilities by means of a gasmeter /'gas-'mi:tə/, an electricity-meter, and a water-meter. These are compounds, in which each component retains the stress and vowel quality it would have as a free form. That is, -meter, as the second element in such a compound, is quite different from the suffix -ometer, as tentatively identified in the previous paragraphs.

364    9 making new nouns 9.7d During protests by African-Americans against racial segregation, in the 1960s, groups would sit together in places they were banned from. Following such sit-ins, there were kneel-ins at churches reserved for whites, drive-ins at segregated ice cream stands, and bury-ins, to integrate cemeteries. Thus did -in /-ˌin/ evolve as a suffix which could be added to a verb and derived a noun describing a form of protest (it generally takes secondary stress). As apartheid was dismantled, protests began to be against other things, such as wars (or an eat-in, against the poor standard of catering in a college). Then verb-in nominalizations were extended to any social or cultural activity (nothing to do with a protest), which involves a fair number of people, such as sing-ins and cook-ins. Having, in the last three chapters, discussed ways of making new verbs, adjectives, and nouns, we can now turn our attention to ways of creating members of the smaller—but still important—class of adverbs. Parts of this chapter are based on sections of Chapter 10 of Dixon (2005a).

10 Care-ful-ly, side-ways, home-wards, crab-fashion: Making new adverbs

10.1 Introduction 365 10.2 -ly 366 10.3 Functions of adverbs  368

10.4 Adjectival semantic types and their adverbs  371 10.5 Emerging double duty  373 10.6 Further suffixes  376

10.1 Introduction Adverbs in English roughly divide into the following sets: I Those which are derived from a noun or adjective or preposition by adding one of the rather minor suffixes -ward(s), -way(s), -wise, -fold, -style, -fashion. These are discussed in 10.6. II Those which are derived from an adjective by adding the powerful derivational suffix -ly, and which maintain the meaning (or a major part of the meaning) of the adjective; for example, quick-ly, rude-ly, narrow-ly, humb-ly, simp-ly, probab-ly, sensib-ly. Most of this chapter is devoted to investigating which types of adjectives may take -ly, and examining the functional properties of the adverbs thereby derived. III The rest, which further divide into two sub-kinds:

366    10 making new adverbs IIIa Those which end in -ly, but have a rather different meaning from the form obtained by omitting the -ly. They include hardly, fairly, really, terribly. IIIb Those which do not end in -ly. Generally, these are not morphologically analysable; for example, quite, just, even, also, plus time adverbs such as still, now, soon, then, and spatial adverbs such as here and out. Others have an historical analysis but are perceived as unanalysable forms in the modern language. They include almost, somewhat, and time adverbs such as always, today, tomorrow, and yesterday. In fact, the division between II and IIIa is not clear-cut. Some adverbs ending in -ly have a degree of semantic similarity to the corresponding adjective, but not so much as prototypical members of II; for example, truly, scarcely, surely. Before looking at the functional categories of adverbs, and their possibilities for positioning within the sentence, we can discuss the origin of adverbderiving suffix -ly. Dixon (2005a: 375–431) provides a full account of the syntax of adverbs (including phrasal and clausal adverbs). Parts of this chapter (including Tables 10.1 and 10.2) are closely based on that account, sometimes quoting from it verbatim.

10.2 -ly 10.2a There are two suffixes -ly /-li/ in present-day English. In 8.2.2 there was discussion of the -ly which derives adjectives from nouns—for example, man-ly, world-ly, cost-ly—and from a few adjectives—good-ly, dead-ly. The other -ly suffix is basically added to an adjective and derives an adverb, as in bad-ly, clever-ly. The origin of the two suffixes is fascinating. In essence, OE formed an adjective by adding suffix -lic to a noun, and then formed an adverb from this by adding a further suffix -e after the -lic. For example:

noun eorÞ ‘earth’ frēond ‘friend’

adjective eorÞ-lic ‘earthly’ freˉond-lic ‘friendly’

adverb eorÞ-lic-e ‘in an earthly manner’ frēond-lic-e ‘in a friendly manner’

10.2 - ly   367 By ME times, two distinct suffixes had emerged. First, -lic, forming adjectives from nouns, reduced to -ly. In OE, adverb-forming -lice had begun to be recognized as a single suffix, being added directly to some adjectives; for example beald-lice ‘boldly’ from beald ‘bold’, and sweˉt-lice ‘sweetly, pleasantly’ from sweˉt ‘sweet’. This suffix also reduced to -ly, forming adverbs from adjectives. Until the seventeenth century, adverb-forming -ly could be added after adjective-forming -ly, as in kind-li-ly and god-li-ly. However, today it is not regarded as acceptable to add adverbalizer -ly to a base form ending in ly. That is, in place of *coward-li-ly one must say something like in a coward-ly manner. (One does sometimes hear friend-li-ly—as in He didn’t behave very friend-li-ly—but only in informal, spoken contexts.) One important difference between the two suffixes concerns what they can follow. Adjective-deriving -ly cannot be added to a form ending in /i/ or /l/; one may say friend-ly but not *enemi-ly, god-ly but not *devil-ly. In contrast, adverb-deriving -ly may follow these segments; examples include angri-ly and careful-ly. 10.2b Adverb-deriving suffix -ly is used with simple adjectives and also with many kinds of adjectives derived from common nouns and from verbs (see Chapter 8). For example, fool-ish-ly, greed-i-ly, wood-en-ly, glori-ous-ly, pain-ful-ly, fashion-ab(l)-ly, fragment-ari-ly, affection-ately, harm-less-ly, tire-some-ly, abus-ive-ly, defi-ant-ly. As mentioned in 8.2.20, -al is generally added to an adjective ending in (-)ic before the -ly; adverbs corresponding to adjectives authentic and poet-ic are authentical-ly and poet-ic-al-ly. (Note that adverbs are not generally formed from adjectives ending in -ed, -free, -proof, -like, and other minor adjectivalizing suffixes.) Just a few adverbs are derived directly from nouns by adding -ly. Part-ly and purpose-ly were coined around 1500. Interestingly, adjective purposeful came into use around 1850, together with adverb purpose-ful-ly. There is a semantic difference, purpose-ly meaning ‘on purpose’ and purpose-fully indicating ‘with purpose’. One may say They keep a dog purpose-ly (not *purpose-ful-ly) to deter intruders, and He strode into the room purpose-ful-ly (scarcely *purpose-ly). Nouns indicating a period of time may take -ly, the resulting forms functioning both as adjective (his week-ly tasks) and as adverb (he is paid weekly). This covers: hour, day, night, week, fortnight, month, quarter, and year.

368    10 making new adverbs 10.3 Functions of adverbs Whereas an adjective can modify a noun, or occur as copula complement, adverbs extend over a wide range of functions. These are, in outline: (a) Modifying a complete clause or sentence—sentential function. (b) Modifying a verb (plus object, if it has one)—manner function. (c) Modifying an adjective—adjectival modifying function. (d) Modifying a complete noun phrase—NP modifying function. (e) Modifying another adverb—adverbial modifying function. These can now be illustrated, one at a time, indicating for each type of adverb its possible positions in sentence structure. (a) Sentential function. For example, obvious-ly as in (adverbs are underlined within example sentences or phrases): (1)  John has obvious-ly been selling secrets to the Ruritanians (2)  Obvious-ly, John has been selling secrets to the Ruritanians (3)  John has been selling secrets to the Ruritanians, obvious-ly These three sentences have essentially the same meaning: ‘it is obvious that John has been selling secrets to the Ruritanians’. That is, obvious-ly modifies the meaning of the complete sentence; A sentential adverb has the potential to occur at any of three positions in sentence structure:

•

After the first word of the auxiliary (coded as ‘A’), as in (1). If there is no auxiliary, the sentential adverb immediately precedes the verb (John obvious-ly sells . . ). Either initially (coded ‘I’), as in (2), or finally (‘F’), as in (3).

•

A sentential adverb in position I or F is frequently—but not invariably— set off by contrastive intonation (shown by a comma in writing). This is also possible for position A: John has, obvious-ly, been selling secrets to the Ruritanians. Note that not all sentential adverbs are used in all three positions. For example, common-ly is more-or-less restricted to position A, certain-ly to A and I, and linguistical-ly to F and I. However, most are acceptable in all three positions.

10.3 functions of adverbs   369 (b) Manner function. For example perfect-ly, as in: (4) Mary could have [perfect-ly summarized the report] in one paragraph (5) Mary could have [summarized the report perfect-ly] in one paragraph These two sentences have essentially the same meaning. The adverb modifies verb, plus object (if there is one), and indicates the manner in which the action was performed. The constituents in square brackets are equivalent to a perfect summary of the report. A manner adverb has the potential to appear in either of two positions:

• •

Immediately before the verb (coded as ‘V’), as in (4) Immediately after the verb, plus object if there is one (coded ‘O’), as in (5)

Most manner adverbs have both positions available, but a few are more restricted. For example, bad-ly and beautiful-ly are more-or-less restricted to position O, while the monomorphemic adverb quite is only found in V. Obvious-ly only has sentential and perfect-ly only has manner function. However, some adverbs span both functions. For instance, stupid-ly. Consider the basic sentence: (6)  The soldier might have answered the officer’s question Suppose that the question was rhetorical and did not expect an answer. It would have been stupid for the soldier to answer it. This can be expressed using stupid-ly in sentential function, in any of the three positions (commas are optional in each sentence): (7) The soldier might, stupidly (A), have answered the officer’s question (8)  Stupidly (I), the soldier might have answered the officer’s question (9) The soldier might have answered the officer’s question, stupid-ly (F) Now suppose a different scenario. The question is not rhetorical, and requires an intelligent answer. But, knowing this soldier, we are aware that he may provide a stupid answer. To express this, stupid-ly can be used in manner function, in either of the two positions:

370    10 making new adverbs (10) The soldier might have [stupid-ly (V) answered the officer’s question] (11) The soldier might have [answered the officer’s question stupidly (O)] Sentences (7–9) describe a potentially stupid soldier, while (10–11) relate to a potentially stupid answer. Note the contrast between sentential adverb stupid-ly in position A, after the first word of the auxiliary, in (7), and manner adverb stupid-ly in position V, immediately before the verb, in (10). Stupid-ly occurs sentence-finally in both (9) and (11), which could lead to ambiguity (although this might be avoided by contrastive intonation in (9)). There is detailed discussion in Dixon (2005a: 385–421) of the positional possibilities for individual adverbs in sentential and manner functions.

(c) Adjectival modifying function. An adverb may immediately precede an adjective within a noun phrase and indicate the degree or status of the property described by the adjective. This function extends to nonderived adverbs such as rather and slightly, and also derived ones such as appropriate­-ly, obvious-ly, and phonetic-al-ly, as in: (12)  He provided [a [phonet-ic-al-ly accurate] transcription] (d) Noun phrase modifying function. An adverb may come right at the beginning of an NP, before the article, as in They promote [just the people who work hard]. Most of the adverbs with this function are simple forms, such as only, even (which may either precede or follow the NP) and really. And there are just a few adverbs derived with -ly, such as exact-ly and precise-ly, as in: (13) He brought [precise-ly the sum of money demanded by the ransomers] (e) Adverbial modifying function. An adverb may be modified by another adverb, indicating its degree. For example, almost exact-ly, quite stupidly, really bad-ly, just perfect-ly. It is rare for an adverb derived with -ly to have this function, although we do find it with some adverbs derived from value adjectives, and with similar-ly and deliberate-ly, as in: (14)  She behaved [deliberate-ly disgusting-ly]

10.4 adjectival semantic types and their adverbs   371 10.4 Adjectival semantic types and their adverbs 10.4a Adjectives do not all behave alike with respect to the derivation of adverbs by suffix -ly. There is a principled basis, which is determined by the semantic type of the adjective (as listed in the Appendix). Table 10.1 provides an outline summary of the possibilities for each adjectival semantic Table 10.1  Functions of adverbs derived from adjectival semantic types (a) sentential function

(b) manner function

(c) adjectival modifying function

dimension

no

some—relating to the meanings of adjectives in semantically diverse ways

no

physical property

no

generally relate only to secondary sense of the adjective

yes

age

no (except for newly as a time adverb)

no

no

colour

no

some

some

speed

yes (similar meaning to manner function)

yes

no

value

no

yes

yes (except for badly, well)

difficulty

no

yes

yes

volition

no

yes

yes

qualification definite possible usual likely sure correct

yes yes yes yes yes yes

no no no no no some

yes yes yes no no yes

human propensity

yes

yes

limited

similarity

yes

yes

yes

372    10 making new adverbs type forming adverbs in (a) sentential function, (b) manner function, and (e) adjectival modifying function. Adverbs derived from the human propensity, speed, and similarity types score ‘yes’ for the manner and sentential columns in Table 10.1. The contrasting semantic effects of the functions are illustrated and explained through examples (7)–(11). All qualification items have sentential function (and many also adjectival function) but most lack manner function. The remaining semantic types occur in manner but generally not in sentential function (some also in adjectival function). Almost all members of the physical property type, and some from colour, have derived adverbs, but the adverb most often corresponds to a secondary, metaphorical sense of the adjective. Thus, hot-ly refers not to temperature (as in hot tea) but to temperament (as in hot temper) and bitter-ly not to taste (as in bitter chocolate) but to attitude (as in bitter dispute). Many of these adverbs are mainly used with speaking and thinking verbs—hot-ly deny, cold-ly assert, dry-ly remark, warm-ly invite, cool-ly consider, sweet-ly request, sour-ly reject, bitter-ly complain, and—with a colour item—dark-ly frown. A number of dimension adjectives—such as big, small, little—do not form adverbs. For some others, the adverb relates to the result of an action: He sliced the bread thin-ly indicates that he produced thin slices. Adverbs narrow-ly and broad-ly relate to a metaphorical sense of the adjective, like physical property items, as in narrow-ly conceived and broad-ly intended. Deep-ly has both a concrete sense (dive deep-ly) and a metaphorical one (think deep-ly). Adjectives short and long each has a secondary sense relating to time (short meeting, long interval); adverbs short-ly and leng-th-i-ly (this being based on the adjective leng-th-y, from noun leng-th, from adjective long) are time adverbs. In essence, an adjective (or the sense of an adjective) which relates to an objective or (semi-)permanent property is unlikely to form an adverb. An adjective (or the sense of an adjective) referring to people’s skill and ability, and to types and degrees of success, is a good candidate for adverbialization. Time adverbs derived from nouns have sentential function and are restricted to F position, as in The magazine is published month-ly. In an article entitled ‘Why great : greatly but not big : *big-ly?’, Kjellmer (1984: 18) concludes that while big is only used as a dimension adjective, great also has ‘a (transferred) value sense to use Dixon’s term (great poet)’, and it is in this sense that it forms an adverb.

10.5 emerging double duty   373 Table 10.2  Interaction of adverb and nominal derivation SEMANTIC TYPE

ADJECTIVE PLUS DERIVED NOUN

DERIVED MANNER ADVERB PLUS VERB

dimension

deep embedd-ing

deep-ly embed

physical property

warm recommend-ation

warm-ly recommend

speed

quick consider-ation

quick-ly consider

colour

clear enunciat-ion

clear-ly enunciate

value

bad organiz-ation

organize bad-ly

difficulty

easy ascen-t

easi-ly ascend

qualification

definite decis-ion

definite-ly decide

possible resign-ation

possib-ly resign

usual respon-se

usual-ly respond

correct judge-ment

correct-ly judge

loyal greet-ing

loyal-ly greet

clever solu-tion

clever-ly solve

similar behav-iour

behave similar-ly

human propensity similarity

10.4b A fascinating property of English is that a manner adverb derived from an adjective is used with an underived verb, but the underlying adjective is used with a nominal derived from the verb. A small sample of examples is in Table 10.2. In the middle column, the adjective must precede the noun. For the right-hand column, the adverb has generally been placed before the verb but it could, equally acceptably, follow—deep-ly embed or embed deep-ly, warm-ly recommend or recommend warm-ly. Only for the main value terms, and for similarity, do we find that the adverb must follow the verb. It will be seen that either the modifier is derived and the head non-derived (in the right-hand column of Table 10.2) or the modifier is underived and the head derived (in the middle column). All of the adverbs are derived by means of suffix -ly. The derived nouns involve a variety of suffixes, -((a)t)ion, -ment, -ing, and -(i)our, besides changes ascend to ascent and respond to response. There are, in addition, some items which do double duty as noun and as verb; for example, a generous laugh| laugh generous-ly, a rough copy| rough-ly copy. 10.5 Emerging double duty 10.5a In standard English, there are just a couple of adjectives which do double duty as adverbs, fast and hard. For example:

374    10 making new adverbs use as adjective He has [a fast car] She has [a hard job]

use as adverb He drives fast She works hard

In some colloquial varieties (perhaps especially in the USA), there has arisen the convention of employing further adjectives in a double duty role, as adverbs. Instead of the standard locutions (15)  Mary [pays the rent regular-ly] (16)  John [did well] in the test one may hear (17)  Mary [pays the rent regular] (18)  John [did good] in the test The substitution of suppletive adjective form good for adverb well, in (18), shows that it is not just a case of omitting -ly from an adverb, but rather of requiring an adjective to undertake double duty as adverb. This is confirmed by the fact that people say He did it great but one cannot say *He did it great-ly, showing that great here is not a shortening of great-ly. Sentences such as (17) and (18) are decried by self-consciously well-­ educated users of the language. However, a linguist’s task is not to prescribe, but rather to describe the language as it is used. This emerging convention in colloquial varieties may in time spread throughout the language, with many adjectives doing double duty as adverbs, just as today many nouns also function as verbs and vice versa (see 2.8). This phenomenon is at present in an incipient stage. It is instructive to briefly consider which types of adjectives are involved, and in what function(s). 10.5b In essence, this emerging double duty applies to:

• •

manner function in O position (after the verb and its object, if there is one)

Well only occurs in O position, whereas regular-ly can be in either O position, as in (15), or in V position: (19)  Mary [regular-ly pays the rent]

10.5 emerging double duty   375 In (19) regular-ly may not be replaced by regular. Adjectives which are taking on double duty as adverbs are all fairly highfrequency items. They include value adjectives such as good, bad, and perfect: (20)  The system [works [just perfect]] Also speed adjectives such as slow and quick: (21)  She [solved the problem quick] Plus human propensity adjectives such as stupid, clever, careful, generous, and honest. Also, among others, deliberate, accidental, correct, similar, different, and easy: (22)  You can [take it to pieces easy] 10.5c Really is perhaps the most versatile adverb in English, occurring in all five functions. For example: (a) sentential function (23)  Mary has really (A) been hiding the Easter eggs [We thought she wasn’t going to do it, but she has been] (b) manner function (24)  Mary has been [really (V) hiding the Easter eggs]  [She has hidden them so well that they will be very hard to find] (c) adjectival modifying function (25)  John is [a [really nice] fellow] (d) noun phrase modifying function (26)  Mary is [really the cleverest student in the class] (e) adverbial modifying function (27)  She [solved the problem [really quick-ly] The adverb really, as in these examples, cannot—within the grammar of present-day English—be regarded as derived from adjective real since the meanings are quite distinct. Sentence (1), John has obvious-ly been selling secrets to the Ruritanians, with sentential adverb obvious-ly, can be restated as It is obvious that John has been selling secrets to the Ruritanians. Manner adverb plus verb, perfect-ly summarized, in (4) is equivalent to a perfect summary. In contrast, (24) can not be re-stated as It is real that Mary has been hiding the Easter eggs. And so on. The interesting point is that, alongside the innovation of double-duty adjectives—saying played bad as an alternative to played bad-ly—we get really

376    10 making new adverbs replaced by real—saying played real bad as an alternative to played really bad-ly. This applies in just two of the five functions. First, (e) adverbial modifying, as in: (28)  She [solved the problem [real quick]] And also (c) adjectival modifying function: (29)  John is [a [real nice] fellow] What appears to be happening is that a new lexeme real has been created (quite distinct from the established adjective real) and it is used as an adverb in functions (e) and (c), in concert with the new double-duty adjectives. (And, completing the circle, it could be said the adverb real-ly is derived from this new adjective by adding -ly.) 10.6 Further suffixes We now turn to a number of suffixes each of which has a rather limited role in forming adverbs. All the derived forms are confined to sentential function. 10.6.1 -ward(s) Suffix -ward(s) /-wəd(z)/ ‘in the direction of’ is a direct descendant of -weard(es) in OE, where we find, among others, uˉfe-weard ‘further up, upper part of’, norÞ-weard ‘northward’ and haˉm-weard(es) ‘on the way home, homewards’. Derivations with -ward(s) function as sentential adverbs and, as noted in 8.2.43, many also do double duty as adjectives. Forms with and without final s are generally interchangeable, which one is used being to a large extent at the whim of the speaker. However, there is a tendency for an adjective to omit the final s and an adverb to include it. For example The up-ward climb was tiring as against We climbed up-wards, and The home-ward journey took a long time alongside We are now travelling home-wards. There are basically three domains of usage for -ward(s): (1) In, out, up, down, back, and after each functions as both preposition and adverb. In the latter instance, it may be expanded by -ward(s). One can say The feather floated down or The feather floated down-wards.

10.6.3 - wise   377 Interestingly, once the -ward(s) form has been derived as an adverb, it may also function as an adjective, as in The feather floated in a downward(s) direction. Note that the meaning of on-ward(s)—as in We must move on-wards and The on-ward march of progress—relates to the sense of adverb on found in Walk on! (2) Simple and complex compass point names may take -ward(s). An example in sentential adverb function is The tropical cyclone is moving south-east-wards at a steady rate, and one in adjective function is The north-ward journey was tedious. (3) Alongside home-ward(s), which goes way back, there is the productive possibility of adding -ward(s) to any expression designating a place, with the meaning ‘in that direction’. Heaven-ward(s), land-ward(s), and wind-ward(s) are well-established. One could equally felicitously say She’s travelling Delhi-wards or Why don’t we direct our steps my-pied-àterre-wards? (Note, though, that this nonce use is much more likely in sentential adverb than in adjective function.) 10.6.2 -way(s) The Germanic suffix -way(s) /-wei(z)/ relates to noun way. It is historically an element in always and leastways, and also in anyway(s) and someway(s), meaning ‘in any way’ and ‘in some way’ respectively. The suffix can be added to some dimension adjectives, either in plain or nominalized form. One may hear either Cut the piece of timber long-ways! or Cut the piece of timber length-ways!, both meaning that it should be cut along the longest dimension. Similarly, side-ways means ‘to the side’, as in look side-ways and walk side-ways. If you put something into a box edgeways this means that the edge goes in first. This use is extended in I couldn’t get a word in edge-ways, meaning that I was not able to intrude even the edge of my proposed utterance into the conversation. The number of -way(s) derivations is severely limited. They can be used as sentential adverbs, in final (F) position. And sometimes also as adjectives, as in His side-ways stagger indicates that he is drunk.

10.6.3 -wise In ME times, -wise /-waiz/ evolved as a derivational suffix, a development from the OE noun wı¯s(e) ‘manner’. There is a rather minor spatial

378    10 making new adverbs sense—long-wise and length-wise have very similar meanings to long-ways and length-ways (it would be interesting to investigate what the difference is, if any). If, say, a stick is slightly longer than the box it is intended for, one may have to put it in slant-wise; that is, slant-ways. The central use is with nouns and it then means ‘in relation to’. Thus clock-wise ‘in the direction that the hands of a clock move’. Also, She’s rather knowledgeable weather-wise (she’s good at predicting it), and He’s been behaving foolishly health-wise (smoking, and eating all the wrong foods). Despite its antiquity, -wise had rather limited use until it took off (predominantly in the USA) from about the 1950s. Houghton (1968) quotes a fine spread of examples, demonstrating the vibrant productivity of this suffix. They include: I think I can say I am leaving Northwestern a little better off football-wise than when I arrived, Business-wise , we believe we are on the right basis, and (Mick Jagger talking about the Rolling Stones) Sound-wise we were different. As has been illustrated, -wise formations may occur in final (F) or initial (I) position. Or, like many other sentential adverbs, some of them may follow the first word of the auxiliary (but do then require contrastive intonation), as in She has, career-wise, been taking after her father and grandfather. They may also function as adjectives, as in a clock-wise direction. These -wise-derived adverbs should be clearly distinguished from compound adjectives whose second element is -wise, such as worldly-wise and street-wise.

10.6.4 Minor suffixes: -fold, -style, -fashion 10.6.4a Old English suffix -feald has become modern -fold /-fould/, retaining the same applicability; it is added just to cardinal numbers, and to many. The original meaning was ‘folded over that many times’ or ‘having that many parts’ (as in a three-fold blessing). In modern usage it generally indicates ‘that number of times’. A -fold derivation can function as a sentential adverb in final position, or as an adjective. Thus, The population increased fifteen-fold and There was a fifteen-fold increase in population both mean that there are now fifteen times as many people as there were before. 10.6.4b During recent times, two Romance nouns have developed into productive suffixes, creating final-position sentential adverbs. These are -style /-stail/ and -fashion /-faʃən/. Thus, He dresses cowboy-style (in the style that

10.6.4 minor suffixes: - fold , - style , - fashion   379 a cowboy adopts) and He licked the lollipop schoolboy-fashion (in a manner reminiscent of a schoolboy). These derivations may also have adjectival function, as in a holiday-style atmosphere. Mainly in their sentential adverb function, both suffixes may be added to long constituents; for example, He talked in circles member-of-parliament-style, and She yodelled Swiss-mountaineer-fashion. The last six chapters have looked at derivational affixes largely on an individual basis. We will now, in the next chapter, take a brief look at how they may be combined.

11 Un-relent-less-ly de-west-ern-iz-ing non-mean-ing-ful infra-structures: Combining affixes

11.1 Introduction 380

11.3 Combining prefixes  392

11.2 Combinations of word-class-

11.4 Everything together  394

changing suffixes  382

11.1 Introduction We have thus far looked at derivational affixes more-or-less one at a time. But they can be combined, in ways that are interesting and complex. 11.1a Consider noun joy, which refers to a quality. Based on it, there are two nouns referring to a state, with rather different meanings: joy-ful-ness and en-joy-ment. This is how they come about. (1)

• • •

Start from quality noun joy. Add suffix -ful, deriving property adjective joy-ful ‘having joy’. Add suffix -ness, deriving noun joy-ful-ness ‘the state of having joy’, as in He appears to be brimming over with joy-ful-ness today.

11.1 introduction   381 (2)

• • 



• 

Start from quality noun joy. Add prefix en-, deriving transitive verb en-joy, which requires an O (object) noun phrase, as in She en-joy-ed the play. The verb can be explained in terms of the underlying noun as ‘have joy about’. Add suffix -ment, deriving noun en-joy-ment ‘the state of having joy about something’. Note that this noun must be accompanied by specification of what the enjoyment was about (corresponding to the O of the underlying verb enjoy). For example, I noticed her absolute en-joy-ment of the play.

Until recently, enjoy was exclusively a transitive verb. An object had to be stated, as in Enjoy it! Over the last few years the habit has arisen of a waiter in a restaurant saying just Enjoy! with no object stated. This new intransitive usage is basically restricted to imperatives.

11.1b Affixes can be combined in many ways. Indeed, extremely long words may be built up. Let’s have a go at this.

• • • • • •

Begin with concrete noun: industry ‘a type of manufacturing enterprise’. Add suffix -al, forming adjective: industri-al ‘related to industry’. Add suffix -ize, forming transitive verb: industri-al-ize ‘make industrial’. Add suffix -ation, forming abstract noun industri-al-iz-ation ‘the process of making industrial’. Add suffix -ist, forming a noun describing a type of person, an industrial-iz-ation-ist ‘a proponent of the process of making industrial’. We can also add prefixes. Suppose that the person described is only pretending to be a proponent of the proposal. The appropriate description is pseudo-industri-al-iz-ation-ist. But maybe this is all in the past. The person now has quite different ideas. So we can describe them as an ex-pseudo-industri-al-iz-ation-ist.

•

11.1c This by no means exhausts the possibilities. But—with four suffixes and two prefixes—we have created a word so long that it would be unlikely to be occur spontaneously. It seems too ugly, and too complex. (Maybe it would be too difficult for our brains to readily process it. This is a question which begs investigation.)

382    11 combining affixes However there are many words in general use which include two or three or even four derivational affixes. We shall first, in 11.2, survey the combinations of suffixes which change word class, or sub-class for nouns. Then, in 11.3, there is consideration of prefixes, which by-and-large do not change word class. Finally, in 11.4 prefixes and suffixes are linked together. It must be stressed that this is a preliminary—and pioneering—survey. There will surely be further possibilities beyond those listed. If the account here can serve as a foundation, to be amended and extended by others, then it will have served its purpose. There have been a number of recent publications on affix combining. They include Fabb (1988), Plag (1999), Hay and Plag (2004) and Plag and Baayen (2009) on suffixes and Zirkel (2010) on prefixes. Each considers only a selection of the relevant affixes. All include useful information. Bauer, Lieber, and Plag (2013: 591–610) features an extensive list of affix combinations.

11.2 Combinations of word-class-changing suffixes 11.2a Standard abbreviations will be used: N for noun, A for adjective, V for verb, Adv for adverb. Arrow ‘→’ indicates the direction of a derivation. The sign ‘√’ comes before a root (the starting point for a derivation). For example: √concept (N) → concept-ual (A) → concept-ual-ize (V) Or, indicating just the word classes involved: √N → A → V. 11.2b We will first consider nouns, adjectives, and verbs, extending the discussion to adverbs in 11.2.5. There are a number of general exclusions; that is, derivations which just do not occur: I No verb is derived from a verb. That is, there is no √V → V, or → V → V. II We do get a few adjectives derived from adjective roots √A → A; for example, yellow-ish, sick-ly, tame-able, right-eous, drunk-en. However, no adjective is derived from a derived adjective. That is, there is no → A → A. III No verb is derived from an adjective which is itself derived from an adjective or verb root. That is, there is no √A → A → V or √V → A → V.

11.2.1 two derivational suffixes onto a noun ro ot   383 Leaving these aside, every possibility of √X → Y → Z is attested, where each of X, Y, and Z can be N, A, or V. We can now survey the seventeen occurring combinations of two derivational suffixes—first those built on a noun, then on an adjective, then on a verb root.

11.2.1 Two derivational suffixes onto a noun root 11.2.1a Of the nine possible sequence-type combinations based on a noun root, two are not attested: √N → V → V by exclusion principle I, and √N → A → A by II. The seven which do occur are exemplified in Table 11.1. We can now comment on the cells in this table. 11.2.1b In (1a), from the concrete noun myth is derived abstract noun myth-ology ‘the study of myths’ and then myth-olog-ist ‘person who studies myths’. In (1b) a person noun is first derived from concrete noun library by adding -an, librari-an, with -ship then being added, giving abstract noun librari-an-ship ‘the professional skills and responsibilities of running a library’. Whereas there are a fair number of examples of (1a) and of (1b)— including crimin-olog-ist, sex-olog-ist, guard-ian-ship, music-ian-ship—for type (1c), iron-monger-y is the only one I know. It describes objects sold by

Table 11.1  Two derivational suffixes added to a noun base (1)

√N→N→N

(a) myth-olog-ist

(b) librari-an-ship

(2)

√N →N→A

(a) victor-i-ous

(b) senat-or-ial

(3)

√N→N→V

(a) myth-olog-ize

(4)

√N→A→N

(a) world-li-ness snobb-ish-ness mind-ful-ness

(b) spirit-ual-ity sens-ibil-ity Christ-ian-ity

(5)

√N→A→V

(a) metr-ic-ize person-al-ize west-ern-ize

(b) document-ari-fy Chin-ese-ify Israel-i-fy

(6)

√N→V→N

(a) subsid-iz-ation cod-if-ication orchestr-at-ion

(b) agon-iz-ing liqui-fy-ing

(7)

√N→V→A

(a) honor-if-ic

(b) liquid-iz-able class-ifi-able

(c) iron-monger-y

(c) Celt-ic-ism/t trib-al-ism/t sect-arian-ism/t

(c) colon-iz-er cod-ifi-er orchestr-at-er

384    11 combining affixes an iron-monger, someone who deals mainly, though not exclusively, with things made of iron. 11.2.1c I only know of two examples for (2) √N → N → A. Nominalizing suffix -y (9.2.2) is generally used with an adjective, as in jealous-y, but is added to the person noun victor thus deriving victor-y ‘the success a victor has achieved’, and thence adjective victor-i-ous ‘quality of being a victor’. And from the noun senate ‘legislative body’ is derived the person noun senat-or ‘member of the senate’ and thence adjective senat-or-ial ‘characteristic of being a member of the senate’. The sequence √N → N → V, in row (3), is also rather rare. We do have myth-olog-ize ‘work on mythology; assign mythic status to’. 11.2.1d In contrast, there are many instances of √N → A → N in row (4), only a sample being provided in Table 11.1. As described in 9.2, a state noun can be formed from most adjectives, both simple and derived; the two main nominalizers are Romance suffix -ity and the highly productive Germanic suffix -ness. We find that adjectives derived by Germanic suffixes—if they form a state noun—only take -ness. For example coward-li-ness, fool-ish-ness, mood-i-ness, wood-en-ness, one-legg-ed-ness, grace-ful-ness, friend-less-ness, burglar-proof-ness, sea-worthy-ness. We find that -ness is also added after some suffixes of Romance origin, such as in pictur-esque-ness, nerv-ous-ness, passion-ate-ness, Chin-ese-ness. Romance nominalizer -ity is only found after Romance suffixes. One says alkali-n-ity in preference to *alkali-ne-ness. However, a Romance adjectivalizer may be followed by -ity in some instances and -ness in others. We find, for example, product-iv-ity and secret-ive-ness, complement-ar-ity, and fragment-ari-ness. After -(u)al, the nominalizer is generally -ity, as in form-al-ity, univers-ality and spirit-ual-ity, but we also get natur-al-ness. Both -ity and -ness are found after Romance suffix -able| -ible. In 9.2.7 it was suggested that -ity tends to be used when there is ‘potential’ involved (for example, profit-ability) and -ness otherwise (as in comfort-able-ness). Romance suffixes -ism and -ist tend to pattern together. They may follow any of a smallish number of adjectivalizing suffixes, as illustrated in (4c). A tradition-al-ist follows the doctrine of tradition-al-ism, adherence to t­ radition-al habits and values. Both -ism and -ist may follow a few

11.2.1 two derivational suffixes onto a noun ro ot   385 adjectives in -ic (such as histor-ic-ism| -ist), quite a number in -al (sensational-ism| -ist, univers-al-ism| -ist, nation-al-ism| -ist), many in -(a)n (American-ism| -ist) and a couple in -arian (authorit-arian-ism| ist). 11.2.1e We can now consider row (5) in Table 11.1, √N → A → V. This involves productive verbalizers -ize and -ify. As explained in 7.4–7, their use is phonologically conditioned. For instance adjectivalizer -ic requires primary stress to be on the syllable immediately preceding the -ic, and this selects a following -ize. An example is noun German /'dʒə:mən/, adjective German-ic /dʒə:'man-ik/ and verb German-ic-ize /dʒə:'man-is-aiz; also poetic-ize, metr-ic-ize, and more. The conditions for using -ize also apply after -(u)al (giving such derivations as concept-ual-ize, person-al-ize, industrial-ize), -(a)n (America-n-ize), -al| -ar (pol-ar-ize), and -ern (west-ern-ize). The phonological conditions for using -ify yield Chin-ese-ify, Israel-i-fy, ­Swed-ish-ify, document-ari-fy, element-ari-fy. 11.2.1f Row (6), √N → V → N, deals with nominalizations of verbs derived from a noun by -ize or -ify or -ate. There are dozens involving suffix -((a)t)ion added after -ize—including standard-iz-ation, neutral-iz-ation, local-iz-ation—plus many added after -ify—pur-if-ication, glor-if-ication, vers-if-ication—and perhaps only one after -ate—orchestr-at-ion. Abstract nouns are also made from derived verbs by adding -ing; examples include agon-iz-ing (as in His continual agon-iz-ing over the decision exasperated us) and liqui-fy-ing (for example, The liqui-fy-ing of the gold took all day). Cell (6c) in Table 11.1 deals with nominalizer -er, in its varied senses, being added to a derived verb. A colon-iz-er is a person who colon-izes some place, a class-ifi-er is a grammatical element which class-ifies nouns (or verbs), and an orchestr-at-er is someone who prepares a score for all the members of an orchestra. There are just a few adjectives based on verbs derived from nouns, shown in row (7). We find -ic after -ify, as in honor-if-ic and beauti-fi-ic, and -able after both -ize and -ify, as in liquid-iz-able, class-ifi-able, and not-ifi-able. 11.2.1g It is interesting to enquire concerning the genetic origins of combining affixes. Most often, a Romance affix attaches to another Romance affix, and Germanic to Germanic. For example, in Table 11.1, cell (4a) involves two Germanic suffixes, and cells (4b) and (4c) two Romance ones. However,

386    11 combining affixes there is a degree of mixing: libari-an-ship has a Romance suffix followed by a Germanic one, while the reverse holds for west-ern-ize.

11.2.2 Two derivational suffixes onto an adjective root 11.2.2a Of the nine possible sequence-type combinations based on an adjective root, three are not attested: √A → V → V by exclusion principle I, √A → A → A by II, and √A → A → V by III. The six which do occur are exemplified in Table 11.2. We can now comment on the cells in this table. 11.2.2b All of adjectives perfect and abstract, verbs perfect and abstract, and nouns perfect-ion and abstract-ion were borrowed into ME from French. In present-day English, perfect and abstract are regarded as primarily adjectives, also doing double duty as verbs. On this basis, perfect-ion and abstraction are √A → N, and adding -ism or -ist creates rare examples of √A → N → N, in row (8). (But an alternative analysis would be as √V → N → N, in row (14) of Table 11.3.) 11.2.2c Suffix -ist is generally added to a noun but there are also a few instances onto an adjective. Many -ist derivations may be extended by -ic, giving √A → N → A. For example legal-ist-ic, fatal-ist-ic, imperial-ist-ic in (9a). Nominalizer -((a)t)ion can be added to adjective discrete, and followed by adjectivalizer -ary, in discret-ion-ary, cell (9b). And (9c) also has a single instance of its type, tru-th-ful.

Table 11.2  Two derivational suffixes added to an adjective base (8)

√A→N→N

(a) perfect-ion-ism/t

(9)

√A→N→A

(a) legal-ist-ic

(10)

√A→N→V

(a) moist-ure-ize

(11)

√A→A→N

(a) clean-li-ness right-eous-ness hard-i-ness

(b) tame-abil-ity

(12)

√A→V→N

(a) standard-iz-ation simpl-if-ication

(b) popular-iz-er pur-ifi-er hard-en-er

(13)

√A→V→A

(a) equal-iz-able

(b) discret-ion-ary

(c) tru-th-ful

11.2.3 two derivational suffixes onto a verb ro ot   387 Instances of → N → V are rare. I know of only one for row (10), √A → N → V, moist-ure-ize. 11.2.2d Looking now at row (11), it was mentioned before that there are only a handful of adjectives derived from root adjectives. Just about all of these can be nominalized, giving √A → A → N. As before, we find -ness used after Germanic adjectivalizers (sick-li-ness, lone-li-ness, hard-i-ness, drunk-en-ness, yellow-ish-ness), and after Romance suffix -ous (right-ous-ness), and Romance nominalizer -ity after Romance suffix -able (practic-abil-ity, perfect-abil-ity). 11.2.2e Verbalizations on an adjective base, with both -ize and -ify, can typically be nominalized by adding -((a)t)ion, in (12a); they include modern-ization, steril-iz-ation, solid-if-ication. Cell (12b) has further instances of √A → V → N, involving agentive nominalizer -er; for example urban-iz-er, ­beauti-fi-er. The minor verbalizing suffix -en may also be followed by -er: a soft-en-er is a substance you put into water to remove minerals and make it soft. For row (13), √A → V → A, there are just a couple of examples of -able added to -ize—equal-ize-able and perhaps real-ize-able.

11.2.3 Two derivational suffixes onto a verb root 11.2.3a Of the nine possible sequence-type combinations based on a verb root, five are not attested: √V → V → N/A/V by exclusion principle I, √V → A → A by II, and √V → A → V by III. The four which do occur are exemplified in Table 11.3. We can now comment on the cells in this table. Table 11.3  Two derivational suffixes added to a verb base (14) √V→N→N (a)

exhibit-ion-ism/t escap-olog-ist

(b) own-er-ship (c) manag-er-ess stud-ent-ship guard-ian-ship

(15) √V→N→A

inspir-ation-al manag-er-ial

(b) inflat-ion-ary

(a)

(d) mean-ing-ful

(e) begg-ar-ly

(16) √V→N→V

(a)

comput-er-ize

(17) √V→A→N

(a)

resent-ful-ness (b) read-abil-ity squash-y-ness sens-itiv-ity compet-itive-ness

(c) own-er-less weigh-t-less

388    11 combining affixes 11.2.3b Row (14) deals with √V → N → N. Nominalizer -((a)t)ion typically follows a verb and may often be followed by either -ism or -ist. The many examples of this include protect-ion-ism| -ist, obstruct-ion-ism| -ist and prohibit-ion-ism| -ist. Nominalizer -ology may only be followed by -ist (never by -ism), as in crimin-olog-ist and climat-olog-ist. Cell (14b) deals with nominalizer -ship, describing the responsibilities associated with a role; it can be added after suffixes which describe types of people, as in lead-er-ship, receiv-er-ship, superintend-ent-ship, guard-ian-ship. Cell (14c) records the fact that agentive suffix -er| -or can be followed by feminine -ess, as in act-r-ess. 11.2.3c Row (15), √V → N → A, features several combinations of suffixes. Adjectivalizer -al can be added to abstract nouns in -((a)t)ion and to agentive nouns in -er| -or. Among the many examples there are educat-ion-al, convers-ation-al, instruct-ion-al, profess-or-ial, dictat-or-ial. We also find -((a)t)ion followed by -ary, in react-ion-ary, inflat-ion-ary. Cell (15c) illustrates agentive -er followed by -less, as in rid-er-less, lend-er-less, and also -less after state noun weigh-t. Minor combinations are shown in (15d–e): -ful following abstract noun mean-ing, and adjective-forming -ly after person noun begg-ar. Examples of (16), √V → N → V, are hard to find. One recent coinage is comput-er-ize ‘use computers to perform a certain task’. 11.2.3d Finally, (17) √V → A → N deals with state nouns derived from adjectives which are based on verbs. As in rows (4) and (11), we find -ness after Germanic adjectivalizers (tickl-ish-ness, crumble-y-ness, forget-ful-ness¸ need-less-ness, irk-some-ness, trust-worthy-ness) and also after Romance suffixes -ous, -ate, and -atory (covet-ous-ness, consider-ate-ness, inflam-atoriness). Romance adjective-forming suffix -able typically takes Romance nominalizer -ity; the many examples of this include desir-abil-ity, predictabil-ity, adapt-abil-ity. But we also get some instances of -able-ness, as in change-able-ness (this is discussed in 9.2.7). And we find examples of both -ness and -ity after Romance suffix -(it)ive: permiss-ive-ness and sens-itiv-ity. 11.2.4 Three (and more) derivational suffixes in sequence Having established the basic patterns of combination, these can now be extended. Some common suffixes may have nothing added after them; they

11.2.4 three (and more) derivational suffixes   389 include -ness, -ity, -ship, and -ism. But others are open to the application of further derivational suffixes, so long as the result is semantically plausible. 11.2.4a First, some possible sequences of three suffixes can be suggested. (A) Verbalizers -ize and -ify may be extended by nominalizers -er or -((a)t)ion or -ing, or adjective-creating -able. Considering those rows of Tables 11.1–3 which end in →V, plausible three-suffix sequences include: √N→N→V→N (3a) √N→A→V→N (5a/b)

√N→A→V→A √A→N→V→N (10a) √V→N→V→N (16a) √V→N→V→A

myth-olog-iz-er institution-al-iz-ation person-al-iz-ing Chin-es-ify-ing west-ern-iz-able moist-ur-iz-ation comput-er-iz-ation comput-er-iz-able

myth-olog-iz-ing element-ari-fi-cation America-n-iz-ation

moist-ur-iz-er

(B) A state noun may be formed by adding -ness or -ity after some ­adjective-deriving suffixes. Looking now at those rows in the tables which end with →A: √N→V→A→N √A→N→A→N √A→V→A→N √V→N→A→N

(7b) (9c) (13a) (15c) (15d) (15e)

liquid-iz-abil-ity tru-th-ful-ness equal-iz-abil-ity own-er-less-ness mean-ing-ful-ness begg-ar-li-ness

weigh-t-less-ness

(C) An agentive noun may sometimes be formed by adding -ist after adjectivalizer -al: √V→N→A→N

(15a)

conserv-ation-al-ist

(D) An adjective may be formed from an agentive noun in -er by adding -less: √N→V→N→A

(6c)

cod-ifi-er-less

(E) A person noun in -er may be made into an abstract noun by adding -ship: √N→V→N→N

(6c)

orchestr-at-er-ship

390    11 combining affixes (F) An adjective may be formed from a person noun in -ist by adding -ic: √N→A→N→A (4c) √A→N→N→A (8a) √V→N→N→A (14a)

form-al-ist-ic perfect-ion-ist-ic obstruct-ion-ist-ic

nation-al-ist-ic isolat-ion-ist-ic

(G) A person noun may be formed from an abstract noun in -((a)t)ion by adding -ist, or a further abstract noun by adding -ism: √N→V→N→N (6a) √A→V→N→N (12a)

subsid-iz-ation-ist/m vers-ifi-cation-ist/m pur-if-ication-ist/m divers-if-ication-ist/m

(H) An adjective can sometimes be formed from an abstract noun by adding -al after -((a)t)ion: √N→V→N→A (6a)

cod-if-ication-al

√A→V→N→A (12a)

simpl-if-ication-al

subsid-iz-ation-al

(J) Verbalizing suffix -ize may be added after adjectival -al: √V→N→A→V (15a)

convers-ation-al-ize

Note that this is just a sample of the possibilities. With suitable context-ualiz-ation, an ingenious speaker could manufacture many more. 11.2.4b Some of these processes may be combined, producing sequences of four derivational suffixes. For example: (A) plus (B) (A) plus (D) (A) plus (G)

√N→A→V→A→N √N→N→V→N→A √N→A→V→N→N

west-ern-iz-abil-ity myth-olog-iz-er-less industri-al-iz-ation-ist/m metr-ic-iz-ation-ist/m comput-er-iz-ation-ist/m conserv-ation-al-ist-ic pur-ifi-cation-ist-ic subsid-iz-ation-al-ist convers-ation-al-iz-er

√V→N→V→N→N √V→N→A→N→A √A→V→N→N→A √N→V→N→A→N √V→N→A→V→N Some of these multi-derivations may be hard to appreciate when considered alone. However, within a suitable context, they will seem normal and natural. (C) plus (F) (G) plus (F) (H) plus (C) (J) plus (A)

11.2.5 bringing in adverbs   391 Suppose that a drama is being written about Plato and that it includes the story of how his philosophical treatises came to be written in dialogue form. It is supposed that they were at first produced as straight prose. Then we hear: athenian: Why not make your treatises as if you were conversing with someone? plato: Why should I do that? athenian: They would be more understandable if written in the form of a convers-ation. plato: Really? athenian: A convers-ation-al style would attract attention. plato: You may be right. But I have not the inclination to conversation-al-ize them myself. athenian: Then let me offer to be your convers-ation-al-iz-er. plato: Very well. So long as you allow me to check the convers-ation-aliz-ation when it is completed. 11.2.4c Why stop at a sequence of four derivational suffixes? There are theoretical possibilities for further extension, albeit that these are seldom likely to be utilized in the ordinary use of language. How about: (H) plus (C) plus (F)

√N→V→N→A→N→A subsid-iz-ation-al-ist-ic ‘characteristics of a person in favour of things relating to the process of providing subsidies’.

If so inclined, the reader may have fun constructing long words for themself, always taking care to provide a context for their use (science fiction stories are permitted).

11.2.5 Bringing in adverbs 11.2.5a The productive adverbializing suffix -ly may be added to a few nouns, to many simple adjectives, and to a fair few adjectives which are derived from common nouns and from verbs (but not from proper names). This can be illustrated with the various adjective-forming suffixes:

392    11 combining affixes

√N→A→Adv

fool-ish-ly wood-en-ly energet-ical-ly life-less-ly affection-ate-ly

statu-esque-ly spirit-ed-ly habit-ual-ly fashion-ab(l)-ly

nois-i-ly murder-ous-ly peace-ful-ly custom-ari-ly

√V→A→Adv

relent-less-ly vari-able-ly satisfact-ori-ly

mourn-ful-ly tire-some-ly expect-ant-ly

dismiss-ive-ly shake-i-ly

11.2.5b Adverbializer -ly can also be added after some sequences of two derivational suffixes. Here are a few samples: √N→N→A→Adv √A→N→A→Adv √V→N→A→Adv

(2a) (9a/c) (15a)

victor-i-ous-ly legal-ist-ical-ly tru-th-ful-ly inspir-ation-al-ly

11.2.5c And, indeed it can be added after the three-suffix sequences in (F): √N→A→N→A→Adv (4c)

nation-al-ist-ical-ly

Note that nothing can be added to an adverb to derive a member of another word class. 11.3 Combining prefixes The majority of prefixes do not change word class or subclass but just add a semantic feature. They can be combined in many ways. 11.3a Indeed, sometimes two prefixes may occur in either order, always with a difference in meaning. For example: pro-[anti-fluoridation] ‘in favour of the plan not to fluoridate water’ anti-[pro-fluoridation] ‘opposed to the plan to fluoridate water’ We can contrast a pseudo-[anti-Marxist] ‘someone who pretends to be opposed to Marxism’ and an anti-[ pseudo-Marxist] ‘someone who is against people who pretend to support Marxism’.

11.3 combining prefixes   393 11.3b It is possible to have a prefix repeated. A revolution may be followed by a counter-revolution, and then that by a counter-counter-revolution. I might ask Tom whether he is planning to join in with the anti-nuclear demonstration and receive the reply: ‘No, I’m anti-anti-nuclear.’ ‘Oh, does that mean you’re pro-nuclear?’ ‘Not at all, I have an open mind on the issue and can appreciate arguments on both sides. What I mean is that I’m out of sympathy with how these anti-nuclear activists behave.’ What he is saying was that he is anti-[anti-nuclear activists]. It is important to mark constituency (with brackets) and indicate the scope of application of each prefix. (See also the discussion of anti-anti- in 5.24b, and of counter-counter- in 5.25c.) Suppose a new boss takes over in the office and re-assigns duties. A month passes, he re-assesses things and re-re-assigns duties. One could say this, to emphasize the repetition of re-assign-ment, but it would be more felicitous to rephrase it as: ‘after a month he re-assigned duties all over again’. Prefix repetition is stylistically marked and generally employed for rhetorical effect. For example, ‘I thought you said these boxes had been unpacked—they look distinctly un-un-packed to me’. And ‘Where are you going? I thought the pre-season training didn’t start till next week!’ ‘I’m off to a sort-of pre-pre-season workout’. Commercial enterprises employ doubling for hype: a super-super-bargainprice or a mega-mega-holiday. But it can also fulfil a more serious purpose. An archaeologist who has been investigating the sub-soil, and found little, may dig a little deeper into the sub-sub-soil. 11.3c Many prefixes are constrained in their positioning, and can only appear immediately before the base. They include mis-, mal-, mono-, poly-, infra-, peri-, arch-. At the other end of the scale there are a dozen or more with considerable fluidity of placement. A sample of some of the more productive is:

• • • • • •

non-sub-titled film, non-mis-guided advice, non-uni-lateral decision un-pre-cooked meals, un-re-integrated group, un-inter-related topics ex-vice-president, ex-under-secretary, ex-para-medic anti-extra-marital sex, anti-de-segregation, anti-under-cover activities pro-dis-engagement, pro-self-examination, Pro-Vice-Chancellor pre-inter-galactic travel, pre-proto-Indo-European, pre-co-educational

394    11 combining affixes

• • • • • • •

semi-un-employed, semi-mis-pronounced word, semi-self-serviced apartment self-un-packing dish washer, self-re-appointed dictator auto-de-frosting refrigerator, auto-re-winding tape re-dis-interred corpse, re-en-slaved boy, re-sub-let apartment pseudo-extra-special event, pseudo-re-education, pseudo-pre-Inca relic quasi-ultra-conservative, quasi-non-conformist, quasi-ultra-sharp knife half-trans-planted garden, half-inter-woven stories

11.3d Of course we can go further. There are three prefixes in half-dis-entangled string and in ex-Pro-Vice-Chancellor. The ‘not quite right’ prefixes, pseudo- and quasi-, may readily be added to a complex derivation. A cheap quasi-auto-de-frosting refrigerator only partly does the job. And John now realizes that he purchased what is in fact a pseudo-self-un-packing dishwasher (the con-man door-to-door salesman assured him it did do this and he was taken in by the line of jive).

11.4 Everything together Prefixes can of course be combined with suffixes. Of the words with two prefixes quoted in 11.3c, half-a-dozen end in -((a)t)ion or -ment or -(a)n or -(at)ive or -ist. And pre-co-educa-ation-al has two prefixes and two suffixes. 11.4a There are a number of common combinations of prefix plus suffix. First, verbalizer en-| em- attracts suffix -ment. Prefix en-| em- means ‘make have’ or ‘make be’ or ‘make be in’, and adding -ment gives ‘the result of making have’ or ‘of making be’ or ‘of making be in’. These can apply to many nouns: em-power-ment, en-slave-ment, em-prison-ment, and also with courage, force, place, list, noble, and throne. Also to adjectives in en-largement and em-bitter-ment, and to a verb in en-tangle-ment. (See 7.3.4 on en-light-en-ment.) Negative prefixes dis- and mis- also combine with -ment. With nouns we get dis-place-ment and dis-courage-ment, and with verbs dis-arrange-ment, dis-engage-ment, mis-state-ment, and mis-treat-ment.

11.4 everything together   395 Interestingly, the dis- correspondent of en-tangle is dis-en-tangle, with dis- added before en-. Something must be en-tangled before it can be dis-entangled. In contrast, the correspondent of en-courage is dis-courage, with dis- replacing en- (rather then being added before it). One can either bolster someone’s courage or mute it; it does not have to be bolstered before being muted. The use of un- and in- with derived adverbs was illustrated in 5.2b and 5.4e. 11.4b Going beyond these conventionalized combinations, it is more or less a case of ‘anything goes’. Derivational affixes may, potentially, be combined in any manner provided that:

• • •

The grammar is right. Typically, each derivational affix may only be added to a form belonging to one (or one of several) word classes. The semantics is right. Each combination of derivations must add up to a plausible whole. There is an appropriate context. Both short and long words only have significance as they are used in discourse, not quoted in isolation.

At the end of 11.1b, we built up ex-pseudo-industri-al-iz-ation-ist. This is one example of the huge number of possibilities, available for use as required.

12 Envoi: The way of words

Words are mysterious and also magnificent creatures. They can fix laws, instruct, beseech, praise, and anger. There is a proverb: Sticks and stones may break my bones but words will never harm me. Nothing could be less true: a wound will heal, a broken leg will mend, but an unkind word—delivered with spite and malice—may echo in one’s mind for ever. The word word has a range of meanings. On the one hand, it can describe anything said or written. The Bible is referred to as ‘the Word of God’ (a rendering of logos in Greek). One can ‘have a quick word’ with someone, ‘put in a good word’ for a friend, news may travel ‘by word of mouth’, a truthful person ‘keeps to their word’. There are special kinds of words, mystical words. Abracadabra is used by a conjuror to produce magical effect, or by a wizard to ward off misfortune. At one time, co-conspirators would recognize one another by each knowing ‘the word’, a secret signal. Nowadays this is called ‘a password’, and is increasingly needed for every kind of electronic transaction. Except that it may be a phrase, or if it is a single unit this may be required to include both letters and numerals, something like ‘xGvV3yt7’. Funny sort of word! Leaving aside such disembodied items, we can focus on words as units of speaking and writing. Language is undoubtedly humankind’s most precious possession, enabling people to live together socially and to pursue aesthetic endeavours such as story and song. People typically think of a language in terms of its inventory of words. But we cannot communicate by just emitting a string of words. They must be linked together—to form a statement, a question, a command, or whatever—and the principles of such linking are called grammar.

12 envoi: the way of words   397 One basic unit of grammar is the clause, with a predicate which is generally filled by a verb. To complete its clause, a verb needs a number of core arguments, each of which is filled by a pronoun or noun (which may be modified by adjective, demonstrative, and so on). For example, Accidents happen has an intransitive verb (happen) as predicate and noun accidents as intransitive subject argument (abbreviated as S; see 1.2). And Cats devour liver has a transitive verb (devour) as predicate, plus two arguments: cats as transitive subject (A) and liver as transitive object (O). A convention of English grammar places the A argument before the predicate and the O argument after it. Grammar provides a framework with slots that are filled by lexemes (nouns, adjectives, verbs, adverbs), thus producing utterances which communicate meaning. One way in which languages differ is in how one can tell which argument is which. In OE, this was achieved by a system of case inflections. A noun in subject (A or S) function took nominative ending, and one in object function (O) took accusative ending. Within a clause, subject noun, object noun, and verb could occur in any order; one could tell who was doing what to whom from the nominative and accusative endings on the nouns. During the development to Modern English, case inflections have been lost (save on pronouns). We know who is doing what to who by the order of constituents—the argument in A function comes before the verb, and that in O function follows it. Another parameter of difference between languages is whether a lexeme relates to just one word class, or whether it undertakes ‘double duty’ (2.8), relating to several classes. In some languages the various word classes—noun, adjective, intransitive verb, transitive verb—are mutually exclusive; each lexeme belongs to just one class. To use a noun in verbal function, it is necessary to apply a morphological process of verbalization, as in English class-ify or hospital-ize. The alternative is to have many lexemes which can be used—without change or form—in two (or more) word classes, as in English we can say take the goods to the market (noun) or market (verb) the goods, and make a list (noun) of the items, or list (verb) the items. In OE there was very little double duty. This increased markedly during the ME period and has continued to proliferate ever since. (The augmentation of double duty happened at about the same time as the loss of case endings. Opinions differ as to the degree of connection between these two developments—see Jespersen 1942: 152–61, Marchand 1969: 362–5.)

398    12 envoi: the way of words Almost every day, one sees or hears what had been just a noun now used as a verb, and vice versa. This can lead to the diminishing use (and eventual extinction) of some derivational processes. There is discussion in 10.5 of the increasing employment of adjectives in adverbial function (as in He pays the rent regular), usurping an adverb derived with suffix -ly (regular-ly). Double duty can lead to ambiguity. Consider a newspaper (or e-bulletin) headline: Royal witnesses cheat. This has two quite different interpretations: (1) [[Royal]ADJECTIVE [witness-es]NOUN]S [cheat]VERB (2) [Royal]NOUN:A [witness-es]VERB [cheat]NOUN:O We have here three double duty lexemes: royal is primarily an adjective, as in (1), but can also be used as a noun meaning ‘a member of a royal family’, as in (2); witness and cheat both function as verb and as agentive noun. Interestingly, there are here two quite different suffixes -es; that in (1) indicates plural of witness used as a noun, while that in (2) marks present tense with a 3rd person singular subject for witness used as a verb. If the case inflections of OE—to mark argument function—had been retained, there would be no ambiguity— cheat would take accusative case just in (2) and witnesses would take nominative case just in (1). And note that if witness and cheat were replaced by lexemes with similar meaning, which do not do double duty as verb and as agentive noun, ambiguity would again be avoided. For example, using observe and swindle we get: (1a) [[Royal]ADJECTIVE [observ-er-s]NOUN]S [swindle]VERB (2a) [Royal]NOUN:A [observe-s]VERB [swindl-er]NOUN:O These involve agentive nouns derived by adding -er to the verb: observ-er in (1a) and swindl-er in (2a), rather than the verb acting in bare form as an agentive noun. It can be seen that the effectiveness of communication is diminished through the loss of grammatical markings such as case inflections, which mark syntactic function within a clause, and through the loss of derivational affixes, which indicate word class membership. The order of words in a clause plays an important role but, as we saw with Royal witnesses cheat, the emergence of double duty does lead to ambiguity. It may be considered improper to say such a thing, but there is no doubt that, as English (or any other languages) simplifies its morphology, it is likely to become less efficient as a vehicle for the communication of meaning. On the positive side, of the two hundred or so derivational affixes discussed in this book, well over half are still fully productive, providing a basis

12 envoi: the way of words   399 for the fashioning of new words. This creation can take many forms. Most often it is quite spontaneous, as when journalist Herb Caen tells how ‘beatnik slipped out of my typewriter one day’ (9.3.12e). At the opposite extreme, a new word may be deliberately crafted. It was mentioned in 5.8b that when, in 1869, Thomas H. Huxley needed a noun to mean ‘someone who believes that nothing can be known outside human experience (including: it cannot be known whether or not there is a god)’, he coined a-gnostic, employing negative prefix a-, which is of Greek origin (5.8), onto Greek gno¯sto¯tos ‘relating to knowledge, especially of a spiritual nature’. Words are the life-blood of a language, and the creation of new words helps to keep it healthy and active. The continued generation of new words enriches the descriptive depth available. All the major derivational affixes of English have been described in this volume. Here they are, available to be used anew.

Appendix 1: List of adjective and verb semantic types, with sample members These lists include some of the more important words in the semantic types, including those which have been quoted for affix attachment in this volume. (A full list would run to many hundred pages.) Many verbs and adjectives have several senses. Generally only one (sometimes more) of the more central senses is included here. Criteria for membership of the semantic types, and discussion, will be found in Dixon (2005a: 84–206; 1991: 73–204).

Adjective types dimension type, e.g. big, large, little, small, slim, great, long, tall, short, deep, shallow, broad, wide, narrow, straight, steep, thick, thin, round, square, flat location subtype, e.g. high, low, external, internal physical property type, e.g. hard, soft, strong, weak, dirty, clean, clear, hot, warm, cool, cold, wet, damp, moist, dry, heavy, light, sweet, sour, bitter, foul, fresh, ripe, smooth, rough, coarse, sharp, blunt, dull, tight, loose, solid, stable, stiff, dense, intricate, rich, slack, stale, delicate, cheap, expensive, quiet, noisy, pure, free, safe corporeal subtype, e.g. fat, well, fit, sick, feeble, deaf, blind, infirm, ill, pregnant, alive, dead, awake, conscious, nude, active, absent, beautiful, handsome, ugly, male speed type—quick (at), fast (at), slow (at), rapid, sudden, frequent age type—new, old, ancient, young, modern colour type, e.g. white, black, red, green, yellow, blue, brown, crimson, violet, spotted, mottled, golden, dark, light, fair

adjective types   401 value type, e.g. (a) good, bad, lovely, pleasant, atrocious, lewd, poor, perfect, real, fake, ordinary, supreme; (b) odd, strange, curious, necessary, crucial, important, amazing, lucky, fortunate, popular difficulty type, e.g. easy, difficult, tough, hard, simple, complex volition type, e.g. deliberate, accidental, purposeful qualification type definite subtype, e.g. definite, probable, true, false, obvious possible subtype, e.g. possible, impossible, plausible, incredible, potential usual subtype, e.g. usual, normal, unusual, extraordinary, common, natural, regular, permanent, special, particular likely subtype, e.g. likely, certain sure subtype, e.g. sure correct subtype, e.g. correct, accurate, precise, legal, standard, proper, right, wrong, rational, coherent, appropriate, apt, pertinent, sensible, significant human propensity type fond subtype, e.g. fond (of   ), familiar (with) angry subtype, e.g. angry, indignant (with/at/about), aghast (at), jealous (of    ), bold (in/about), arrogant (in/about), modest (in/about), mad (about), violent (about), sad (about) happy subtype, e.g. anxious, keen, happy, thankful, shy, sorry, glad, serious (all taking about), proud, ashamed, afraid (all taking of   ) unsure subtype, e.g. certain, sure, unsure (all taking of or about), careful (about), curious (about), aware (of   ) eager subtype, e.g. eager, ready, prepared (all taking for), reluctant (to), willing clever subtype, e.g. clever, intelligent, smart, wise, adept, stupid, lucky, kind, cruel, brave, generous (all taking at/in), benevolent, lenient, loyal, polite, rude (all taking to), strict (to/in), severe (in/on), prudent (in), famous (for) honest subtype, e.g. honest, fair, noble (about/in/at), frank, just (in/about) busy subtype, e.g. busy (at/with), occupied (with), preoccupied (with), lazy (over), capable (of   ), alert (to), patient (in/with/over) similarity type, e.g. like, unlike, similar to, different from, equal to/with, identical to, analogous to, separate from, independent of, consistent with

402    appendix 1: list of adjective and verb semantic types Primary-A verb types motion type (Roles: Moving, Locus) motion-a, the run subtype, e.g. run, walk, crawl, slide, spin, roll, turn, wriggle, quiver, fidget, swing, wave, rock, shake, climb, dive, stroll, trot, gallop, jog, charge, dance, march, jump, bounce, swim, fly, play, ride, drive, ascend, descend motion-b, the arrive subtype, e.g. (i) arrive, return, go, come; (ii) enter, exit, emerge, diverge (from), cross, embark, depart, travel, pass, escape, flee, come in, go out, emit, migrate, emigrate; (iii) reach, approach, visit motion-c, the take subtype, e.g. (i) take, bring, fetch, retrieve; (ii) send; (iii) move, divert, raise, lift, steal motion-d, the follow subtype, e.g. (i) follow, pursue, track, lead, guide, precede, accompany; (ii) meet motion-e, the carry subtype, e.g. carry, bear, transport, cart, remove motion-f, the throw subtype, e.g. throw, chuck, fling, propel, ejaculate, pour, spray, water, push, shove, press, pull, jerk, drag, tug, draw motion-g, the drop subtype, e.g. fall, drop, spill, tip (over), upset, overturn, capsize, trip, slip rest type (Roles: Resting, Locus) rest-a, the sit subtype, e.g. sit (down), stand (up), lie (down), kneel, crouch, squat, lean, hang (down), float rest-b, the stay subtype, e.g. stay, settle (down), live, dwell, inhabit, inhere, stop, rest, remain, reside, attend rest-c, the put subtype, e.g. (i) put, place, lay, locate, set, arrange, assemble, install, align, put NP on, sow, plant, inter, fill, load, pack, hide, beach, land, shelve, store, dump; (ii) leave, desert, abandon, deposit, ground, take NP off rest-d, the contain subtype, e.g. contain, enclose, encircle, adjoin, surround, integrate rest-e, the hold subtype, e.g. (i) hold, handle; (ii) grab, grasp, seize, clutch, clench, catch, gather, pick up, capture, trap rest-f, the open subtype, e.g. open, expose, close, shut, lock affect type (Roles: Agent, Target, Manip) affect-a, the touch subtype, e.g. touch, stroke affect-b, the hit subtype, e.g. hit, strike, punch, poke, bump, kick, knock, tap, bash, slap, spank, whip, belt, stone, cane, hammer, iron, shoot, tread (on)

primary-a verb types   403 affect-c, the stab subtype, e.g. pierce, prick, stab, dig, sting, knife, spear, cut, prune, mow, saw, slice, chop, hack affect-d, the rub subtype, e.g. rub, wipe, scrape, erase, scratch, mark, sweep, brush, shave, rake, polish, lubricate, lick, wash affect-e, the wrap subtype, e.g. wrap, bind, cover, butter, roof, veil, clothe, dress, grease, plaster, paint, coat, surround, frame, put NP on, unwrap, uncover, unroof, undress, take NP off, peel, shell affect-f, the stretch subtype, e.g. stretch, extend, compress, bend, curl, fold, coil, twist, pinch, squeeze, vaporize, liquefy, solidify, melt, thaw, dissolve, freeze, cool (down), warm (up), heat (up), burn, singe, blaze, radiate, glow affect-g, the build subtype (Product role), e.g. build, knit, tie, make, produce, weave, sew, shape, form, connect, link, unify, divide, stir, mix, knead, fry, bake, cook, boil, inflate, mend, repair, draw, invent, compose, write, sign, forge affect-h, the break subtype (Breaking role), e.g. break, crush, squash, shatter, crumble, tangle, tamper (with), meddle (with), destroy, damage, wreck, collapse, tear, split, shred, chip, crack, snap, smash, crash, burst, explode, blow NP up, blow NP down, let NP off, erupt giving type (Roles: Donor, Gift, Recipient), e.g. give, hand (over), lend, sell, rent, hire, pay, invest, charge, owe, acquit, bequeath, bestow, award, compensate, serve, feed, supply, present, donate, contribute, encumber, deliver, let, tip, reward, bribe, market, exchange, trade, borrow, buy, purchase, bet, accept, reject, receive own subtype (Roles: Owner, Possession), e.g. have, lack, get, inherit, keep, obtain, redeem, come by, gain, own, possess, belong to, lose, forfeit corporeal type (Roles: Human, Substance), e.g. eat, consume, dine (on), chew, digest, suck, drink, smoke, bite, nibble, peck (at), sip, smell, feel, taste, sniff, swallow, breathe, starve, smile, fart, burp, cough, spit, shit, pee, urinate, vomit, yawn, sneeze, laugh, leer, wink, blink, mourn, sob, sigh, sleep, sleep in, drowse, dream (of/about), think (of/about/over), die, weep, cry, shiver, faint, pass out, wheeze, sweat, sulk, rest, ache, suffer, live, survive, come to, recover, be born, wake, waken, grow, swell, inflame, hurt, bleed, heal, drown, bring NP to, comfort, console, cure, soothe, ease, nurse, doctor, tickle, kill, murder, assassinate, harm, beat up, injure, wound, poison, give birth to, kiss, embrace, hug, cuddle, fuck, rape weather type (no roles), e.g. rain, snow, hail, thunder

404    appendix 1: list of adjective and verb semantic types competition type (Competitor role), e.g. conquer, beat, defeat, triumph (over), dominate, overcome, bring down, race (against), resist, fight, play, win, lose, attack, guard, shield, surrender, defend, compete (with), struggle (against) social contract type, e.g. appoint, employ, dismiss, sack, fire, promote, nominate, convert, swindle, arrest, prosecute, impeach, punish, ordain, govern, rule, civilize, missionize, participate (in), join, manage, preside (over), appeal, supplicate, apply for, qualify for, resign from, withdraw (from), work (at), marry, betroth using type, e.g. use, operate, manipulate, work, employ, wear, waste, fiddle with, indulge (in), spend, expend obeying type, e.g. obey, defy, execute, process, deal with, grant, refuse, perform

Primary-B verb types attention type (Roles: Perceiver, Impression) attention-a, the see subtype, e.g. see, hear, smell, taste, feel, observe, notice, perceive attention-b, the show subtype, e.g. show, demonstrate, exhibit attention-c, the recognize subtype, e.g. recognize, spot attention-d, the discover subtype, e.g. discover, find, see NP through attention-e, the witness subtype—witness attention-f, the look subtype, e.g. look (at), listen (to), stare (at), glare (at), peep (at), peer (at), squint (at), eavesdrop (on), search (for), look (for), hunt (for), inspect, study, investigate, scan, scrutinize, examine, check, view, explore, survey, look after, superintend, visit attention-g, the watch subtype, e.g. watch, listen (to) Also: (i) ignore, disregard, overlook, pass over; (ii) appear, disappear; (iii) look, sound, smell, taste, feel thinking type (Roles: Cogitator, Thought) thinking-a, the think subtype, e.g. think (of/about/over), consider, imagine

primary-b verb types   405 thinking-b, the assume subtype, e.g. assume, suppose thinking-c, the ponder subtype, e.g. ponder (on/over), meditate (on/ about), brood (on/over), speculate (on/about), wonder (at/about), reflect (on/about), dream (of/about), contemplate thinking-d, the remember subtype, e.g. remember, forget thinking-e, the know subtype, e.g. know, sense, feel, realize, learn, understand, teach, educate thinking-f, the conclude subtype, e.g. conclude, infer, reason, argue, prove, demonstrate, show, guess thinking-g, the solve subtype, e.g. solve, work NP out, devise, make NP up, analyse thinking-h, the believe subtype, e.g. believe, suspect, doubt deciding type (Roles: Decision-maker, Course) deciding-a, the resolve subtype, e.g. decide (on), determine (on), resolve, plan, settle deciding-b, the choose subtype, e.g. choose, select, pick NP (out), appoint, elect, vote (for/on), judge, navigate speaking type (Roles: Speaker, Addressee(s), Medium, Message—with components Message-Label and Message-Content) speaking-a, the talk subtype, e.g. speak, talk, expostulate, chat, gossip, converse, communicate, quarrel, argue, joke, write, compose speaking-b, the discuss subtype, e.g. discuss, refer to, describe, talk about speaking-c, the shout subtype, e.g. shout, call, cry, ejaculate, roar, swear, pray, preach, narrate, interpret, quote, recite, intone, read, sing, whistle, warble, croak, creak, squeak, translate, pronounce, mispronounce, utter, name speaking-d, the report subtype set (i), e.g. say, declare, assert, deny, observe, joke, put NP about, give NP out, let NP out, put NP across, let on about set (ii), e.g. state, avow, affirm, confirm, rumour set (iii), e.g. announce, proclaim, mention, note, report, regret set (iv), e.g. remark (on), comment (on), explain, justify set (v), e.g. boast (about/of   ), brag (about/of   ), complain (about/of   ), grumble (about) set (vi), e.g. suggest, hint, claim, acknowledge, admit, confess (to), repute set (vii), e.g. undertake, offer, bid, propose, agree (on/with) set (viii), e.g. promise, threaten

406    appendix 1: list of adjective and verb semantic types speaking-e, the inform subtype, e.g. inform, advise, lecture, agree (on/ with), remind speaking-f, the tell subtype, e.g. tell, ask, question, request, beg, supplicate, enquire, demand, answer, reply (to) speaking-g, the order subtype, e.g. (i) order, command, insist (on), urge, instruct, encourage, warn, caution, persuade, invite, recommend (to), tell, remind, ask, request, beg; (ii) forbid, discourage, dissuade, prohibit speaking-h, the forgive subtype, e.g. (i) insult, slander, curse, scold, criticize, blame, condemn, rebuke, contradict, abuse, forgive, pardon, praise, thank, congratulate, compliment, tell NP off, pick on; (ii) accuse, excuse; (iii) greet, welcome, introduce; (iv) cheer, applaud, apologize liking type (Roles: Experiencer, Stimulus—with components StimulusLabel and Stimulus-Content), e.g. (i) like, love, hate, prefer, fear, dread, trust; (ii) dislike, resent, loathe, abhor, admire, value, honour, respect, regret, grudge, rejoice (in/at), (don’t) mind (about), (don’t) care (about); (iii) enjoy, favour, object to, approve of; (iv) worship, fall for; also: envy, pity, relent, fall back on, fall out with, rely on annoying type (same roles as liking), e.g. frighten, terrify, scare, shock, upset, surprise, astonish, offend, delight, please, satisfy, entertain, amuse, thrill, excite, stimulate, inspire, impress, concern, trouble, worry, grieve, dismay, depress, sadden, madden, infuriate, annoy, anger, irk, disappoint, confuse, bewilder, deceive, trick, perplex, puzzle, interest, distract, bore, attract, embarrass, disgust, tire, exhaust, weary, bother, influence acting type, e.g. act, behave, flirt, adopt, copy, imitate, rehearse, mimic, mime, reproduce happening type, e.g. happen, take place, prevail, abound, exist, establish, organize, arrange, bring NP about, circulate, commit, attend to, neglect, put NP on, take NP on, get on with, do, tie NP in with, increase, decrease, change, devise, vary, repeat, reverse, experience, undergo, transpire, transcend comparing type, e.g. resemble, differ (from), take after, distinguish (between), compare, class, group, cost, grade, match, balance, calculate, measure, weigh, time, count, fit, suit, equal, include, comprise, consist in/of, correspond (to), be made up of

secondary-c verb types   407 relating type, e.g. depend (on), result (from), indicate, relate (to), conform (to), imply, be due (to), put NP down to, show, demonstrate, suggest Secondary-A verb types (no independent roles) modal type, e.g. will, would, shall, should, ought to, must, can, could, be to, may, might, used to, had better, need, dare semi-modal type, e.g. be going to, have (got) to, be able to, be about to, get to, be bound to beginning type, e.g. (i) begin, start, commence; (ii) continue (with), keep (on (with)), go on (with); (iii) finish, cease, stop, complete, discontinue, interrupt, punctuate; also: set in, break out trying type, e.g. (i) try, attempt, try out; (ii) succeed (in/at), manage; (iii) miss, fail; (iv) practise, repeat hurrying type, e.g. hurry (over/with), hasten (over/with), dawdle (over), hesitate (over/with) daring type, e.g. dare, venture Secondary-B verb types wanting type (Role: Principal), e.g. (i) want, wish (for), desire, crave, long (for), pine (for); (ii) hope (for); (iii) demand; (iv) need, require, deserve; (v) expect, wait (for), dread; (vi) intend, plan (for), aim (for), mean, prepare (for); (vii) pretend postponing type (Role: Timer), e.g. postpone, defer, put off, delay, avoid Secondary-C verb types making type (Role: Causer), e.g. (i) make, force, compel, cause, drive, get, have, tempt; (ii) let, permit, allow; (iii) prevent, stop, spare, save, check (oneself), rescue, release; (iv) ensure, control

408    appendix 1: list of adjective and verb semantic types helping type (Role: Helper), e.g. help, aid, assist, oblige, cooperate (with), collaborate (with), hinder, support, oppose

Secondary-D verb types (Role: Arbiter) seem type, e.g. seem, appear, look, sound, feel, happen, come about matter type, e.g. matter, count

Appendix 2: Alphabetical directory of adjectives and verbs in the list of semantic types abandon: rest-c/put abhor: liking abound: happening absent: physical property/ corporeal abuse: speaking-h/forgive accept: giving accidental: volition accompany: motion-d/follow accurate: qualification/correct accuse: speaking-h/forgive ache: corporeal acknowledge: speaking-d/report acquit: giving act: acting active: physical property/ corporeal adept (at/in): human propensity/ clever adjoin: rest-d/contain admire: liking admit: speaking-d/report adopt: acting advise: speaking-e/inform affirm: speaking-d/report afraid (of   ): human propensity/ happy

aghast (at): human propensity/ angry agree (on/with): speaking-d/report and -e/inform aid: helping aim (for): wanting alert (to): human propensity/busy align: rest-c/put alive: physical property/ corporeal allow: making amazing: value amuse: annoying analogous to: similarity analyse: thinking-g/solve ancient: age anger: annoying angry (with/at/about): human propensity/angry announce: speaking-d/report annoy: annoying answer: speaking-f/tell anxious (about): human propensity/happy apologize: speaking-h/forgive appeal: social contract appear: attention and seem

410    appendix 2: alphabetical directory applaud: speaking-h/forgive apply for: social contract appoint: social contract and deciding-b/choose approach: motion-b/arrive appropriate: qualification/ correct approve of: liking apt: qualification/correct argue: thinking-f/conclude and speaking-a/talk arrange: rest-c/put and happening arrest: social contract arrive: motion-b/arrive arrogant (in/about): human propensity/angry ascend: motion-a/run ashamed (of   ): human propensity/ happy ask: speaking-f/tell and -g/order asleep: physical property/ corporeal assassinate: corporeal assemble: rest-c/put assert: speaking-d/report assist: helping assume: thinking-b/assume astonish: annoying atrocious: value attack: competition attempt: trying attend: rest-b/stay attend to: happening attract: annoying avoid: postponing avow: speaking-d/report awake: physical property/ corporeal

award: giving aware (of   ): human propensity/ unsure bad: value bake: affect-g/build balance: comparing bash: affect-b/hit be able to: semi-modal be about to: semi-modal be born: corporeal be bound to: semi-modal be due to: relating be going to: semi-modal be made up of: comparing be to: modal beach: rest-c/put bear: motion-e/carry beat: competition beat up: corporeal beautiful: physical property/ corporeal beg: speaking-f/tell and -g/order begin: beginning behave: acting believe: thinking-h/believe belong to: giving/own belt: affect-b/hit bend: affect-f/stretch benevolent (to): human propensity/ clever bequeath: giving bestow: giving bet: giving betroth: social contract bewilder: annoying bid: speaking-d/report big: dimension

appendix 2: alphabetical directory   411 bind: affect-e/wrap bite: corporeal bitter: physical property black: colour blame: speaking-h/forgive blaze: affect-f/stretch bleed: corporeal blind: physical property/ corporeal blink: corporeal blow NP down: affect-h/break blow NP up: affect-h/break blue: colour blunt: physical property boast (about/of   ): speaking-d/ report boil: affect-g/build bold (in/about): human propensity/ angry bore: annoying borrow: giving bother: annoying bounce: motion-a/run brag (about/of   ): speaking-d/report brave (at/in): human propensity/ clever break: affect-h/break break out: beginning breathe: corporeal bribe: giving bring: motion-c/take bring (to): corporeal bring down: competition bring NP about: happening broad: dimension brood (on/over): thinking-c/ ponder brown: colour

brush: affect-d/rub build: affect-g/build bump: affect-b/hit burn: affect-f/stretch burp: corporeal burst: affect-h/break busy (at/with): human propensity/ busy butter: affect-e/wrap buy: giving calculate: comparing call: speaking-c/shout can: modal cane: affect-b/hit capable (of   ): human propensity/ busy capsize: motion-g/drop capture: rest-e/hold care (about): liking careful (about): human propensity/unsure carry: motion-e/carry cart: motion-e/carry catch: rest-e/hold cause: making caution: speaking-g/order cease: beginning certain: qualification/likely certain (of/about): human propensity/unsure change: happening charge: giving and motion-a/run chat: speaking-a/talk cheap: physical property check: attention-f/look check (oneself   ): making cheer: speaking-h/forgive

412    appendix 2: alphabetical directory chew: corporeal chip: affect-h/break choose: deciding-b/choose chop: affect-c/stab chuck: motion-f/throw circulate: happening civilize: social contract claim: speaking-d/report class: comparing clean: physical property clear: physical property clench: rest-e/hold clever (at/in): human propensity/ clever climb: motion-a/run close: rest-f/open clothe: affect-e/wrap clutch: rest-e/hold coarse: physical property coat: affect-e/wrap coherent: qualification/correct coil: affect-f/stretch cold: physical property collaborate (with): helping collapse: affect-h/breaking come: motion-b/arrive come about: seem come by: giving/own come in: motion-b/arrive come to: corporeal comfort: corporeal command: speaking-g/order commence: beginning comment on: speaking-d/report commit: happening common: qualification/usual communicate: speaking-a/talk compare: comparing

compel: making compensate: giving compete (with): competition complain (about/of   ): speaking-d/report complete: beginning complex: difficulty compliment: speaking-h/forgive compose: affect-g/build and speaking-a/talk compress: affect-f/stretch comprise: comparing concern: annoying conclude: thinking-f/conclude condemn: speaking-h/forgive confess (to): speaking-d/report confirm: speaking-d/report conform (to): comparing confuse: annoying congratulate: speaking-h/forgive connect: affect-g/build conquer: competition conscious: physical property/ corporeal consider: thinking-a/think consist of/in: comparing consistent with: similarity console: corporeal consume: corporeal contain: rest-d/contain contemplate: thinking-c/ponder continue (with): beginning contradict: speaking-h/forgive contribute: giving control: making converse: speaking-a/talk convert: social contract cook: affect-g/build

appendix 2: alphabetical directory   413 cool (down): physical property and affect-f/stretch cooperate (with): helping copy: acting correct: qualification/correct correspond (to): comparing cost: comparing cough: corporeal could: modal count: comparing and matter cover: affect-e/wrap crack: affect-h/break crash: affect-h/break crave: wanting crawl: motion-a/run creak: speaking-c/shout crimson: colour criticize: speaking-h/forgive croak: speaking-c/shout cross: motion-b/arrive crouch: rest-a/sit crucial: value cruel (at/in): human propensity/ clever crumble: affect-h/break crush: affect-h/break cry: corporeal and speaking-c/shout cuddle: corporeal cure: corporeal curious (about): value and human propensity/unsure curl: affect-f/stretch curse: speaking-h/forgive cut: affect-c/stab damage: affect-h/break damp: physical property

dance: motion-a/run dare: modal and daring dark: colour dawdle (over): hurrying dead: physical property/ corporeal deaf: physical property/ corporeal deal with: obeying deceive: annoying decide (on): deciding-a/resolve declare: speaking-d/report decrease: happening deep: dimension defeat: competition defend: competition defer: postponing definite: qualification/definite defy: obeying delay: postponing deliberate: volition delicate: physical property delight: annoying deliver: giving demand: speaking-f/tell and wanting demonstrate: attention-b/show, thinking-f/conclude, and relating dense: physical property deny: speaking-d/report depart: motion-b/arrive depend (on): relating deposit: rest-c/put depress: annoying descend: motion-a/run describe: speaking-b/discuss desert: rest-c/put

414    appendix 2: alphabetical directory deserve: wanting desire: wanting destroy: affect-h/break determine (on): deciding-a/resolve devise: thinking-g/solve and happening die: corporeal differ (from): comparing different from: similarity difficult: difficulty dig: affect-c/stab digest: corporeal dine (on): corporeal dirty: physical property disappear: attention disappoint: annoying discontinue: beginning discourage: speaking-g/order discover: attention-d/discover discuss: speaking-b/discuss disgust: annoying dislike: liking dismay: annoying dismiss: social contract disregard: attention dissolve: affect-f/stretch dissuade: speaking-g/order distinguish (between): comparing distract: annoying dive: motion-a/run diverge (from): motion-b/arrive divert: motion-c/take divide: affect-g/build do: happening doctor: corporeal dominate: competition donate: giving don’t care about: liking

don’t mind about: liking doubt: thinking-h/believing drag: motion-f/throw draw: motion-f/throw and affect-g/build dread: liking and wanting dream (of/about): corporeal and thinking-c/ponder dress: affect-e/wrap drink: corporeal drive: making and motion-a/ run drop: motion-g/drop drown: corporeal drowse: corporeal dry: physical property dull: physical property dump: rest-c/put dwell: rest-b/stay eager (for): human propensity/ eager ease: corporeal easy: difficulty eat: corporeal eavesdrop (on): attention-f/look educate: thinking-e/know ejaculate: motion-f/throw and speaking-c/shout elect: deciding-b/choose embark: motion-b/arrive embarrass: annoying embrace: corporeal emerge: motion-b/arrive emigrate: motion-b/arrive emit: motion-b/arrive employ: social contract and using

appendix 2: alphabetical directory   415 encircle: rest-d/contain enclose: rest-d/contain encourage: speaking-g/order encumber: giving enjoy: liking enquire: speaking-f/tell ensure: making enter: motion-b/arrive entertain: annoying envy: liking equal: comparing equal to/with: similarity erase: affect-d/rub erupt: affect-h/break escape: motion-b/arrive establish: happening examine: attention-f/look exchange: giving excite: annoying excuse: speaking-h/forgive execute: obeying exhaust: annoying exhibit: attention-b/show exist: happening exit: motion-b/arrive expect: wanting expend: using expensive: physical property experience: happening explain: speaking-d/report explode: affect-h/break explore: attention-f/look expose: rest-f/open expostulate: speaking-a/talk extend: affect-f/stretch external: dimension/location extraordinary: qualification/ usual

fail: trying faint: corporeal fair (about/in/at): human propensity/honest fair: colour fake: value fall: motion-g/drop fall back on: liking fall for: liking fall out with: liking false: qualification/definite familiar (with): human propensity/ fond famous (for): human propensity/ clever fart: corporeal fast (at): speed fat: physical property/corporeal favour: liking fear: liking feeble: physical property/ corporeal feed: giving feel: corporeal, attention-a/see, thinking-e/know, and seem fetch: motion-c/take fiddle with: using fidget: motion-a/run fight: competition fill: rest-c /put find: attention-d/discover finish: beginning fire: social contract fit: comparing fit: physical property/corporeal flat: dimension flee: motion-b/arrive fling: motion-f/throw

416    appendix 2: alphabetical directory flirt: acting float: rest-a/sit fly: motion-a/run fold: affect-f/stretch follow: motion-d/follow fond (of   ): human propensity/ fond forbid: speaking-g/order force: making forfeit: giving/own forge: affect-g/build forget: thinking-d/remember forgive: speaking-h/forgive form: affect-g/build fortunate: value foul: physical property frame: affect-e/wrap frank (in/about): human propensity/honest free: physical property freeze: affect-f/stretch frequent: speed fresh: physical property frighten: annoying fry: affect-g/build fuck: corporeal gain: giving/own gallop: motion-a/run gather: rest-e/hold generous (at/in): human propensity/clever get: giving/own and making get on with: happening get to: semi-modal give: giving give birth to: corporeal give NP out: speaking-d/report

glad (about): human propensity/ happy glare at: attention-f/look glow: affect-f/stretch go: motion-b/arrive go on (with): beginning go out: motion-b/arrive golden: colour good: value gossip: speaking-a/talk govern: social contract grab: rest-e/hold grade: comparing grant: obeying grasp: rest-e/hold grease: affect-e/wrap great: dimension green: colour greet: speaking-h/forgive grieve: annoying ground: rest-c/put group: comparing grow: corporeal grudge: liking grumble (about): speaking-d/report guard: competition guess: thinking-f/conclude guide: motion-d/follow hack: affect-c/stab had better: modal hail: weather hammer: affect-b/hit hand (over): giving handle: rest-e/hold handsome: physical property/ corporeal

appendix 2: alphabetical directory   417 hang (down): rest-a/sit happen: happening and seem happy (about): human propensity/ happy hard: physical property and difficulty harm: corporeal hasten (over/with): hurrying hate: liking have: giving/own and making have (got) to: semi-modal heal: corporeal hear: attention-a/see heat (up): affect-f/stretch heavy: physical property help: helping hesitate (over/with): hurrying hide: rest-c/put high: dimension/location hinder: helping hint: speaking-d/report hire: giving hit: affect-b/hit hold: rest-e/hold hollow: physical property honest (about/in/at): human propensity/honest honour: liking hope (for): wanting hot: physical property hug: corporeal hunt (for): attention-f/look hurry (over/with): hurrying hurt: corporeal identical to: similarity ignore: attention ill: physical property/corporeal imagine: thinking-a/think

imitate: acting impeach: social contract imply: relating important: value impossible: qualification/ possible impress: annoying include: comparing increase: happening incredible: qualification/possible independent of: similarity indicate: relating indignant (with/at/about): human propensity/angry indulge (in): using infer: thinking-f/conclude infirm: physical property/ corporeal inflame: corporeal inflate: affect-f/stretch influence: annoying inform: speaking-e/inform infuriate: annoying inhabit: rest-b/stay inhere: rest-b/stay inherit: giving/own injure: corporeal insist (on): speaking-g/order inspect: attention-f/look inspire: annoying install: rest-c/put instruct: speaking-g/order insult: speaking-h/forgive integrate: rest-d/contain intelligent (at/in) human propensity/clever intend: wanting inter: rest-c/put interest: annoying

418    appendix 2: alphabetical directory internal: dimension/location interpret: speaking-c/shout interrupt: beginning intone: speaking-c/shout intricate: physical property introduce: speaking-h/forgive invent: affect-g/build invest: giving investigate: attention-f/look invite: speaking-g/order irk: annoying iron: affect-b/hit jealous (of   ): human propensity/ angry jerk: motion-f/throw jog: motion-a/run join: social contract joke: speaking-a/talk and -d/report judge: deciding-b/choose jump: motion-a/run just (in/about): human propensity/ honest justify: speaking-d/report keen (about): human propensity/ happy keep: giving/own keep on (with): beginning kick: affect-b/hit kill: corporeal kind (at/in): human propensity/ clever kiss: corporeal knead: affect-g/build kneel: rest-a/sit knife: affect-c/stab knit: affect-g/build knock: affect-b/hit

know: thinking-e/know lack: giving/own land: rest-c/put large: dimension laugh: corporeal lay: rest-c/put lazy (over): human propensity/ busy lead: motion-d/follow lean: rest-a/sit learn: thinking-e/know leave: rest-c/put lecture: speaking-e/inform leer: corporeal legal: qualification/correct lend: giving lenient (to): human propensity/ clever let: giving and making let on about: speaking-d/report let NP (off   ): affect-h/break let NP out: speaking-d/report lewd: value lick: affect-d/rub lie (down): rest-a/sit lift: motion-c/take light: colour and physical property like: liking and similarity likely: qualification/likely link: affect-g/build liquefy: affect-f/stretch listen (to): attention-f/look and -g/watch little: dimension live: rest-b/stay and corporeal load: rest-c/put loathe: liking

appendix 2: alphabetical directory   419 locate: rest-c/put lock: rest-f/open long: dimension long (for): wanting look: seem look (at/for): attention-f/look look after: attention-f/look loose: physical property lose: giving/own and competition love: liking lovely: value low: dimension/location loyal (to): human propensity/ clever lubricate: affect-d/rub lucky (at/in): value and human propensity/clever mad (about): human propensity/ angry madden: annoying make: affect-g/build, and making make NP up: thinking-g/solve male: physical property/ corporeal manage: social contract and trying manipulate: using march: motion-a/run mark: affect-d/rub market: giving marry: social contract match: comparing matter: matter may: modal mean: wanting measure: comparing meddle (with): affect-h/break

meditate (on/about): thinking-c/ponder meet: motion-d/follow melt: affect-f/stretch mend: affect-g/build mention: speaking-d/report might: modal migrate: motion-b/arrive mime: acting mimic: acting mind (about): liking mispronounce: speaking-c/shout miss: trying missionize: social contract mix: affect-g/build modern: age modest (in/about): human propensity/angry moist: physical property mottled: colour mourn: corporeal move: motion-c/take mow: affect-c/stab murder: corporeal must: modal name: speaking-c/shout narrate: speaking-c/shout narrow: dimension natural: qualification/usual navigate: deciding-b/choose necessary: value need: modal and wanting neglect: happening new: age nibble: corporeal noble: human propensity/honest noisy: physical property

420    appendix 2: alphabetical directory nominate: social contract normal: qualification/usual note: speaking-d/report notice: attention-a/see nude: physical property/ corporeal nurse: corporeal obey: obeying object to: liking oblige: helping observe: attention-a/see and speaking-d/report obtain: giving/own obvious: qualification/definite occupied (with): human propensity/busy odd: value offend: annoying offer: speaking-d/report old: age open: rest-f/open operate: using oppose: helping ordain: social contract order: speaking-g/order ordinary: value organize: happening ought to: modal overcome: competition overlook: attention overturn: motion-g/drop owe: giving own: giving/own pack: rest-c/put paint: affect-e/wrap pardon: speaking-h/forgive

participate (in): social contract particular: qualification/usual pass: motion-b/arrive pass out: corporeal pass over: attention patient (in/with/over): human propensity/busy pay: giving peck (at): corporeal pee: corporeal peel: affect-e/wrap peep at: attention-f/look peer at: attention-f/look perceive: attention-a/see perfect: value perform: obeying permanent: qualification/usual permit: making perplex: annoying persuade: speaking-g/order pertinent: qualification/correct pick on: speaking-h/forgive pick NP (out): deciding-b/choose pick NP up: rest-e/hold pierce: affect-c/stab pinch: affect-f/stretch pine (for): wanting pity: liking place: rest-c/put plan (for): deciding-a/resolve and wanting plant: rest-c/put plaster: affect-e/wrap plausible: qualification/possible play: motion-a/run and competition pleasant: value please: annoying

appendix 2: alphabetical directory   421 poison: corporeal poke: affect-b/hit polish: affect-d/rub polite (to): human propensity/ clever ponder (on/over): thinking-c/ponder poor: value popular: value possess: giving/own possible: qualification/possible postpone: postponing potential: qualification/possible pour: motion-f/throw practise: trying praise: speaking-h/forgive pray: speaking-c/shout preach: speaking-c/shout precede: motion-d/follow precise: qualification/correct prefer: liking pregnant: physical property/ corporeal preoccupied (with): human propensity/busy prepare (for): wanting prepared (for): human propensity/ eager present: giving preside (over): social contract press: motion-f/throw pretend: wanting prevail: happening prevent: making prick: affect-c/stab probable: qualification/definite process: obeying proclaim: speaking-d/report

produce: affect-g/build prohibit: speaking-g/order promise: speaking-d/report promote: social contract pronounce: speaking-c/shout propel: motion-f/throw proper: qualification/correct propose: speaking-d/report prosecute: social contract proud (of  ): human propensity/happy prove: thinking-f/conclude prudent (in): human propensity/ clever prune: affect-c/stab pull: motion-f/throw punch: affect-b/hit punctuate: beginning punish: social contract purchase: giving pure: physical property purposeful: volition pursue: motion-d/follow push: motion-f/throw put: rest-c/put put (off   ): postponing put NP about: speaking-d/ report put NP across: speaking-d/report put NP down to: relating put NP on: rest-c/put, affect-e/ wrap, and happening puzzle: annoying qualify (for): social contract quarrel: speaking-a/talk question: speaking-f/tell quick (at): speed

422    appendix 2: alphabetical directory quiet: physical property quiver: motion-a/run quote: speaking-c/shout race (against): competition radiate: affect-f/stretch rain: weather raise: motion-c/take rake: affect-d/rub rape: corporeal rapid: speed rational: qualification/correct reach: motion-b/arrive read: speaking-c/shout ready (for): human propensity/ eager real: value realize: thinking-e/know reason: thinking-f/conclude rebuke: speaking-h/forgive receive: giving recite: speaking-c/shout recognize: attention-c/recognize recommend (to): speaking-g/ order recover: corporeal red: colour redeem: giving/own refer to: speaking-b/discuss reflect (on/about): thinking-c/ponder refuse: obeying regret: speaking-d/report and liking regular: qualification/usual rehearse: acting reject: giving rejoice (in/at): liking

relate (to): relating release: making relent: liking reluctant (to): human propensity/ eager rely on: liking remain: rest-b/stay remark (on): speaking-d/report remember: thinking-d/remember remind: speaking-e/inform and -g/order remove: motion-e/carry rent: giving repair: affect-g/build repeat: happening and trying reply (to): speaking-f/tell report: speaking-d/report reproduce: acting repute: speaking-d/report request: speaking-f/tell and -g/order require: wanting rescue: making resemble: comparing resent: liking reside: rest-b/stay resign (from): social contract resist: competition resolve: deciding-a/resolve respect: liking rest: rest-b/stay and corporeal result (from): relating retrieve: motion-c/take return: motion-b/arrive reverse: happening reward: giving rich: physical property ride: motion-a/run

appendix 2: alphabetical directory   423 right: qualification/correct ripe: physical property roar: speaking-c/shout rock: motion-a/run roll: motion-a/run roof: affect-e/wrap rough: physical property round: dimension rub: affect-d/rub rude (to): human propensity/ clever rule: social contract rumour: speaking-d/report run: motion-a/run sack: social contract sad (about): human propensity/ angry sadden: annoying safe: physical property satisfy: annoying save: making saw: affect-c/stab say: speaking-d/report scan: attention-f/look scare: annoying scold: speaking-h/forgive scrape: affect-d/rub scratch: affect-d/rub scrutinize: attention-f/look search (for): attention-f/look see: attention-a/see see NP through: attention-d/ discover seem: seem seize: rest-e/hold select: deciding-b/choose sell: giving

send: motion-c/take and giving sense: thinking-e/know sensible: qualification/ correct separate from: similarity serious (about): human propensity/ happy serve: giving set: rest-c/put set in: beginning settle: deciding-a/resolve settle (down): rest-b/stay severe (in/on): human propensity/ clever sew: affect-g/build shake: motion-a/run shall: modal shallow: dimension shape: affect-g/build sharp: physical property shatter: affect-h/break shave: affect-d/rub shell: affect-e/wrap shelve: rest-c/put shield: competition shit: corporeal shiver: corporeal shock: annoying shoot: affect-b/hit short: dimension should: modal shout: speaking-c/shout shove: motion-f/throw show: attention-b/show, thinking-f/conclude, and relating shred: affect-h/break shut: rest-f/open

424    appendix 2: alphabetical directory shy (about): human propensity/ happy sick: physical property/ corporeal sigh: corporeal sign: affect-g/build significant: qualification/ correct similar (to): similarity simple: difficulty sing: speaking-c/shout singe: affect-f/stretch sip: corporeal sit (down): rest-a/sit slack: physical property slander: speaking-h/forgive slap: affect-b/hit sleep: corporeal sleep in: corporeal slice: affect-c/stab slide: motion-a/run slim: dimension slip: motion-g/drop slow (at): speed small: dimension smart (at/in): human propensity/ clever smash: affect-h/break smell: attention-a/see and corporeal smile: corporeal smoke: corporeal smooth: physical property snap: affect-h/break sneeze: corporeal sniff: corporeal snow: weather sob: corporeal

soft: physical property solid: physical property solidify: affect-f/stretch solve: thinking-g/solve soothe: corporeal sorry (about): human propensity/ happy sound: attention and seem sour: physical property sow: rest-c/put spank: affect-b/hit spare: making speak: speaking-a/talk spear: affect-c/stab special: qualification/usual speculate (on/about): thinking-c/ponder spend: using spill: motion-g/drop spin: motion-a/run spit: corporeal split: affect-h/break spot: attention-c/recognize spotted: colour spray: motion-f/throw square: dimension squash: affect-h/break squat: rest-a/sit squeak: speaking-c/shout squeeze: affect-f/stretch squint (at): attention-f/look stab: affect-c/stab stable: physical property stale: physical property stand (up): rest-a/sit standard: qualification/correct stare (at): attention-f/look start: beginning

appendix 2: alphabetical directory   425 starve: corporeal state: speaking-d/report stay: rest-b/stay steal: motion-c/take steep: dimension stiff: physical property stimulate: annoying sting: affect-c/stab stir: affect-g/build stone: affect-b/hit stop: beginning, making and rest-b/stay store: rest-c/put straight: dimension strange: value stretch: affect-f/stretch strict (to/in): human propensity/ clever strike: affect-b/hit stroke: affect-a/touch stroll: motion-a/run strong: physical property struggle (against): competition study: attention-f/look stupid (at/in): human propensity/ clever succeed (in/at): trying suck: corporeal sudden: speed suffer: corporeal suggest: relating and speaking-d/report suit: comparing sulk: corporeal superintend: attention-f/look supplicate: social contract and speaking-f/tell supply: giving

support: helping suppose: thinking-b/assume supreme: value sure: qualification/sure sure (of/about): human propensity/ unsure surprise: annoying surrender: competition surround: rest-d/contain and affect-e/wrap survey: attention-f/look survive: corporeal suspect: thinking-h/believe swallow: corporeal swear: speaking-c/shout sweat: corporeal sweep: affect-d/rub sweet: physical property swell: corporeal swim: motion-a/run swindle: social contract swing: motion-a/run take: motion-c/take take after: comparing take NP off: rest-c/put and affect-e/wrap take NP on: happening take place: happening talk: speaking-a/talk talk about: speaking-b/ discuss tall: dimension tamper (with): affect-h/break tangle: affect-h/break tap: affect-b/hit taste: corporeal and attention-a/see

426    appendix 2: alphabetical directory teach: thinking-e/know tear: affect-h/break tell: speaking-f/tell and -g/order tell off: speaking-h/forgive tempt: making terrify: annoying thank: speaking-h/forgive thankful (about): human propensity/happy thaw: affect-f/stretch thick: dimension thin: dimension think (of/about/over): thinking-a/ think and corporeal threaten: speaking-d/report thrill: annoying throw: motion-f/throw thunder: weather tickle: corporeal tie: affect-g/build tie NP in with: happening tight: physical property time: comparing tip: giving tip (over): motion-g/drop tire: annoying touch: affect-a/touch tough: difficulty track: motion-d/follow trade: giving transcend: happening translate: speaking-c/shout transpire: happening transport: motion-e/carry trap: rest-e/hold travel: motion-b/arrive tread (on): affect-b/hit

trick: annoying trip: motion-g/drop triumph (over): competition trot: motion-a/run trouble: annoying true: qualification/definite trust: liking try: trying try out: trying tug: motion-f/throw turn: motion-a/turn twist: affect-f/stretch ugly: physical property/corporeal uncover: affect-e/wrap undergo: happening understand: thinking-e/know undertake: speaking-d/report undress: affect-e/wrap unify: affect-g/build unlike: similarity unroof: affect-e/wrap unsure (of/about): human propensity/unsure unusual: qualification/usual unwrap: affect-e/wrap upset: annoying and motion-g/drop urge: speaking-g/order urinate: corporeal use: using used to: modal usual: qualification/usual utter: speaking-c/shout value: liking vaporize: affect-f/stretch

appendix 2: alphabetical directory   427 vary: happening veil: affect-e/wrap venture: daring view: attention-f/look violent (about): human propensity/ angry violet: colour visit: motion-b/arrive and attention-f/look vomit: corporeal vote (for/on): deciding-b/choose wait (for): wanting wake(n): corporeal walk: motion-a/run want: wanting warble: speaking-c/shout warm: physical property warm NP up: affect-f/stretch warn: speaking-g/order wash: affect-d/rub waste: using watch: attention-g/watch water: motion-f/throw wave: motion-a/run weak: physical property wear: using weary: annoying weave: affect-g/build weep: corporeal weigh: comparing welcome: speaking-h/forgive well: physical property/ corporeal wet: physical property

wheeze: corporeal whip: affect-b/hit whistle: speaking-c/shout white: colour wicked: value wide: dimension will: modal willing: human propensity/eager win: competition wink: corporeal wipe: affect-d/rub wise (at/in): human propensity/ clever wish (for): wanting withdraw (from): social contract witness: attention-e/witness wonder (at/about): thinking-c/ ponder work: social contract and using work NP out: thinking-g/solve worry: annoying worship: liking would: modal wound: corporeal wrap: affect-e/wrap wreck: affect-h/break wriggle: motion-a/run write: affect-g/build and speaking-a/talk wrong: qualification/correct yawn: corporeal yellow: colour young: age

Glossary A: subject of a transitive verb, a syntactic core argument. ablative case inflection: basically marking movement away from some location (‘from’); may also be the case on a noun phrase which follows a certain preposition with a certain meaning. accent, see stress. accusative: case inflection: the ending on the words in a noun phrase which is in transitive object (O) function, as in Latin and Old English; may also be the case on a noun phrase which follows a certain preposition with a certain meaning. adaptation: a new word made by analogic adaptation from a model word, substituting a new initial element of similar phonological form; for example, chocaholic and workaholic modelled on alcoholic. adjectivalization: morphological derivation which forms an adjective stem from a noun or verb root or stem. adjective: class of words which typically refer to properties and have two main roles: (a) makes a statement that something has a certain property through functioning in copula complement slot; and (b) helps to specify the referent of the head noun in an NP by functioning as ­modifier to it. adverb: class of words which can modify a clause, a verb (plus object if it has one), an adjective, a noun phrase, or another adverb. affix: a bound form added to a root or stem. affixation: morphological process which involves adding an affix to a root or stem. ambitransitive: verb which can function in both a transitive and an intransitive clause; of type S = A or S = O. analogy: creating a new pairing of forms which is similar to an existing pairing; for example, in the sixteenth century English had adjective beauty and verb beaut-ify, plus adjective ugly—verb ugl-ify was created by analogy with beaut-ify.

glossary   429 antonym: if X and Y are antonyms, then something which is more X is less Y, and vice versa; for example, big|  little and many other pairs in the dimension, physical property, speed, age, and value semantic types. argument, core: an obligatory argument for a specific verb, which must be either stated or understood from the context. argument, peripheral: non-core argument, which is optional; typically includes instrument, accompaniment, recipient, beneficiary, time, place, manner. back formation: taking a form which looks as if it were derived (such as greedy), assigning it a derived structure (greed-y), and thus extracting a new root (greed ). base: a form (root or stem) to which a derivational affix is added. blend: a part of one word combined with a part of another word; for instance, brunch from breakfast and lunch (not to be confused with adaptation). borrowing: the process by which one language takes a word (a loan), or a grammatical construction, from another language. bound form: form which cannot occur alone but must be attached to some other form, e.g. un- in English. clause: the description of some activity, state, or property; consists of an obligatory predicate which requires certain core arguments and may also have peripheral arguments. clitic: a surface element part-way between a word and an affix in its properties; it is typically a separate grammatical word which is attached to a contiguous phonological word. cognates: two forms which are historically related; that is, go back to a single original form. complement clause: clause which fills a (normally core) argument slot in a higher clause. complementary terms: two terms are complementary if the assertion of one implies denial of the other (for example, single/married, male/female); colours constitute a larger set of complementary terms. compounding: morphological process which joins two roots to form one stem. constituent: anything which fills a slot in a syntactic structure. construction: type of clause (or, sometimes, phrase) with specified properties. converses: if A is X to B implies that B is Y to A, then X and Y are converses; for example, husband| wife, buy| sell. conversion: alternative description of double duty.

430   glossary copula clause: indicating a relational meaning between CS (copula subject) and CC (copula complement) functions. copula complement (CC): the argument in a copula clause which is shown to be in a specified relation to the copula subject (typically, may be realized as an plain NP, an NP marked with a preposition, a possessive clause, an adjective, or a complement clause). copula subject (CS): that argument in a copula clause which is topic for the discourse in which it occurs (generally realized by an NP or a complement clause). core argument: an obligatory argument for a specific verb, which must be either stated or understood from the context. count noun: a noun referring to a set of individuals, which can be counted; a count noun may take the plural suffix. derivation: optional morphological process which applies to a root or stem and derives a stem; may or may not change word class. derivational affix: an affix which realizes a morphological process of derivation. diachronic description: description of how a language system changes through time. diphthong: vowel phoneme which has two or more phonetic components. double duty: a word, which has primary membership of one word class, having secondary membership of another word class; for example, noun stone also does double duty as a verb (also called ‘conversion’ or ‘zero derivation’). free form: a form which constitutes a grammatical word without any morphological processes having to be applied. fusion: where the combination of two forms involves merging or change of one or both (rather than simple addition); for example, root music /'myu:zik/ plus suffix -ian /-iən/ gives music-ian /'myu:siʃən/, where /ki/ across the root-suffix boundary becomes /ʃ/. genitive: the marker on words of a noun phrase which is in possessive function. grammatical word: a unit on the hierarchy of grammatical units (just below phrase) defined on grammatical criteria; generally (but not necessarily always) coinciding with phonological word. head: obligatory nucleus of a phrase which determines the grammatical profile of the whole phrase (for example, gender of a noun phrase). inflection: morphological process which obligatorily applies to a root or derived stem of a certain word class, producing a grammatical word.

glossary   431 inflectional affix: an affix which realizes a morphological process of inflection. internal change: morphological process which involves changing a vowel (or, less frequently, a consonant) in the middle of a word, for instance, from take /teik/ to took /tuk/ in English. intransitive: clause type with one core argument, in S (intransitive subject) function; verb which occurs in the predicate of such a clause. language: in the technical sense of linguists, a number of forms of speech are said to constitute a single language if they are mutually intelligible. lexeme: a word which belongs to vocabulary, rather than to grammar (as do the, of, to, not, etc.). loan word: a word taken from another language; for example, English café from French. mass: a noun referring to a substance (such as mud ) or a state (such as anger) which cannot be counted; a mass noun may not take the plural suffix. morpheme: the minimum meaningful unit of speech. morphological process: process which applies to a root, forming a stem. morphology: that part of grammar which studies the structure of words. nominative case inflection: the ending on the words in a noun phrase which is in transitive subject (A) or intransitive subject (S) function, as in Latin and Old English. nominalization: morphological derivation which forms a noun stem from a verb or adjective root or stem. nonce word: a word coined for just one occasion (that is, not in general use). noun: word class whose primary function is as head of an NP; many of its members refer to concrete objects. noun phrase (NP): a constituent which can fill an argument slot in clause structure; it has a noun or pronoun or demonstrative, etc. as head. NP, see noun phrase. O: object of a transitive verb, a syntactic core argument. opposites, see antonyms, complementary terms, converses. peripheral argument: non-core argument, which is optional; typically includes instrument, accompaniment, recipient, beneficiary, time, place, manner. phoneme: basic unit of sound distinction (the minimum segmentable unit of phonology). phonetics: articulatory and/or acoustic study of the sounds of speech. phonological word: a unit on the hierarchy of phonological units (just above syllable), defined on phonological criteria; generally (but not necessarily always) coinciding with grammatical word.

432   glossary phonology: description of the phonetic contrasts which are used to distinguish between distinct words in a given language. phrasal verb: a single lexeme which is made up of a simple verb and one or two prepositions/adverbs, where the meaning of the whole cannot be inferred from the meanings of the components; for example, take after, put up with. phrase: a constituent which can fill a slot in clause structure—noun phrase in an argument slot and verb phrase in predicate slot. possessive phrase: a type of NP which is included within a larger NP and indicates the possessor with respect to the head of the larger NP, which is the possessed. pragmatics: the practical consequences of the use of a given portion of language. predicate: the central (and obligatory) structural element of a clause, generally realized by a verb phrase (with verb as head); it determines the number and type of core arguments required in the clause. prefix: an affix which precedes a root or stem. preposition: a marker of a peripheral grammatical relation which is realized as a separate phonological word or as a clitic, not as an affix; a preposition precedes the constituent for which it provides grammatical marking. primary verbs: refer directly to an activity or state. proclitic: clitic which is attached to the beginning of a word. productivity: the extent to which a derivational affix is used to create new words. proto-language: putative single ancestor language for a group of modern languages that are held to be genetically related, each having developed by regular changes from the proto-language. root: unanalysable lexical element. S: subject of an intransitive verb, a syntactic core argument. S = A ambitransitive: the S argument, when the verb is used intransitively, corresponds to the A argument, when it is used transitively (for example, eat, knit). S = O ambitransitive: the S argument, when the verb is used intransitively, corresponds to the O argument, when it is used transitively (for example, trip, melt). secondary concepts: provide modification for a primary verb; may be realized as an affix or as a verb (a secondary verb).

glossary   433 semantic role: the types of participant involved with verbs of a certain semantic type. semantic type: a set of words with similar meanings and grammatical properties. semantics: study of the meaning relations conveyed by the grammatical systems and lexical contrasts of a language. semi-affix: something which functions both as a free form, and also with many of the properties of a derivational affix. stem: the nucleus of a word, to which an inflectional process applies, forming a full word. stress (or accent): a contrastive prosody generally having scope over a word, characterized by some or all of: loudness, vowel quality, pitch, and length. subgroup: set of languages within a language family which descend from a single ancestor language, this being itself a descendant of the proto-­ language for the whole language family. subject, see A, CS, S. suffix: an affix which follows a root or stem. suppletion: when a lexeme has two forms which are not cognate (as go and went in English). syllable: a phonological unit centred on a nucleus (typically a vowel) which may be preceded and/or followed by one or more consonants. synchronic description: description of a language system at one point in time, without taking account of historical changes. syntax: study of the organization and interrelation of the components of a grammar above the level of word. transitive: clause type with two core arguments, in A (transitive subject) and O (transitive object) functions; verb which occurs in the predicate of such a clause. transitivity: whether a verb is intransitive, transitive, ambitransitive of type S = A, or ambitransitive of type S = O. verb: word class whose primary function is as head of a predicate; most of its members refer to actions and states. verb phrase: a constituent which can fill the predicate slot within a clause, typically has a verb as its head. word: the result of applying optional derivational processes to a root, and then any obligatory inflectional process to the resulting stem; subtypes: phonological word, grammatical word (unit at the intersection of morphology and syntax).

434   glossary word class: a class of words which have a common grammatical function, and similar meanings; open word classes are noun, verb, adjective, adverb, closed classes include pronouns, demonstratives, prepositions (word classes are also called ‘parts of speech’). zero derivation: alternative description of double duty.

References Adams, Valerie. 1973. An introduction to Modern English word-formation. London: Longman. Adams, Valerie. 2001. Complex words in English. Harlow: Longman. Aikhenvald, Alexandra Y. 2007. ‘Typological distinctions in word formation’, pp. 1−65 of Language typology and syntactic description, 2nd edition, Vol III: Grammatical categories and the lexicon, edited by Timothy Shopen. Cambridge: Cambridge University Press. Algeo, John. 1971. ‘The voguish uses of non’, American Speech 46: 87−105. Barnhart, Robert K. 1988. Editor of Chambers dictionary of etymology. Edinburgh: Chambers. Baron, Dennis. 1986. Grammar and gender. New Haven: Yale University Press. Bauer, Laurie. 1983. English word formation. Cambridge: Cambridge University Press. Bauer, Laurie. 2001. Morphological productivity. Cambridge: Cambridge University Press. Bauer, Laurie. 2006. ‘Competition in English word formation’, pp. 177−98 of The handbook of the history of English, edited by Ans van Kemenade and Bettelou Los. Malden, MA: Blackwell. Bauer, Laurie and Huddleston, Rodney. 2002. ‘19. Lexical word-formation’, pp. 1621–1721 of The Cambridge grammar of the English language, chief authors Rodney Huddleston and Geoffrey K. Pullum. Cambridge: Cambridge University Press. Bauer, Laurie, Lieber, Rochelle, and Plag, Ingo. 2013. The Oxford reference guide to English morphology. Oxford: Oxford University Press. Beckett, Samuel. 1963. Watt. London: Calder. Berlin, Brent and Kay, Paul. 1969. Basic color terms. Berkeley and Los Angeles: University of California Press. Bloch, Bernard, and Trager, George L. 1942. Outline of linguistic analysis. Baltimore, MD: Waverley Press. Bloomfield, Leonard. 1933. Language. New York: Holt. Bolinger, Dwight L. 1941. ‘Among the new words’, American Speech 16: 144−8. Brinton, Laurel J. 2000. The structure of Modern English: a linguistic introduction. Amsterdam: John Benjamins. Chapman, R. W. 1934. Adjectives from proper names. (SPE Tract No. LII.) Oxford: Clarendon Press. Clark, Eve V. and Clark, Herbert H. 1979. ‘When nouns surface as verbs’, Language 55: 767–811.

436   references Dixon, R. M. W. 1977. ‘Where have all the adjectives gone’, Studies in Language 1: 19−80. [Revision published as pp. 1–62 of Dixon 1982.] Dixon, R. M. W. 1982. Where have all the adjectives gone, and other essays in semantics and syntax. Berlin: Mouton. Dixon, R. M. W. 1991. A new approach to English grammar, on semantic principles. Oxford: Clarendon Press. Dixon, R. M. W. 2005a. A semantic approach to English grammar. Oxford: Oxford University Press. Dixon, R. M. W. 2005b. ‘Comparative constructions in English’, Studia Anglica Posnaniensia 41: 5−27. [Reprinted as pp. 473–93 of Language at large: Essays on syntax and semantics, by Alexandra Y. Aikhenvald and R. M. W. Dixon. Leiden: Brill. 2011.] Dixon, R. M. W. 2010a. Basic linguistic theory, Vol. 1. Methodology. Oxford: Oxford University Press. Dixon, R. M. W. 2010b. Basic linguistic theory, Vol. 2. Grammatical topics. Oxford: Oxford University Press. Fabb, Nigel. 1988. ‘English suffixation is constrained only by selectional restrictions’, Natural language and linguistic theory 6: 527−39. Flexner, Stuart Berg. 1987. Editor in chief of The Random House dictionary of the English language, 2nd edition, unabridged. New York: Random House. Gadde, Fredrik. 1910. On the history and use of the suffixes -ery (-ry), -age and -ment in English. Lund: Berlinoska Boktrvckeriet. Gleason, H. A. 1955. An introduction to descriptive linguistics. New York: Holt, Rinehart and Winston. Goldwag, Arthur. 2008. -Isms and -ologies: All the movements, ideologies and doctrines that have shaped our world. New York:Vintage. Harley, Heidi. 2006. English words: a linguistic introduction. Oxford: Blackwell. Hay, Jennifer and Plag, Ingo. 2004. ‘What constrains possible suffix combinations? On the interaction of grammatical and processing restrictions in derivational morphology’, Natural language and linguistic theory 22: 565−96. Hockett, Charles F. 1958. A course in modern linguistics, New York: Macmillan. Houghton, Donald E. 1968. ‘The suffix -wise’, American Speech 43: 209−15. Huddleston, Rodney and Pullum, Geoffrey K. 2002. Chief authors of The Cambridge grammar of the English language. Cambridge: Cambridge University Press. Jespersen, Otto. 1933. ‘Negation in English and other languages’, pp. 1−151 of Selected writings of Otto Jespersen. London: Allen and Unwin and Tokyo: Senjo. [Originally published in 1917.] Jespersen, Otto. 1942. A modern English grammar, on historical principles, Part VI, Morphology. Copenhagen: Munksgaard and London: Allen and Unwin. Johnson, Samuel. 1755. A dictionary of the English language . . . London: J. and P. Knapton, T. and T. Longman, C. Hitch and L. Hawes, A. Millar, and R. and J. Dodsley. Jones, Daniel. 1956. Everyman’s English pronouncing dictionary, 11th edition. London: J. M. Dent and New York: E. P. Dutton.

references   437 Jones, Daniel. 1977. Everyman’s English pronouncing dictionary, 14th edition, revised by A. C. Gimson and Susan Ramsaran. London: J. M. Dent. Kaunisto, Mark. 1999. ‘Electric/electrical and classic/classical: variation between the suffixes -ic and -ical’, English Studies 80: 343−70. Kaunisto, Mark. 2004. ‘Adjective pairs ending in -ic and -ical: A study of variation and change across centuries’. PhD dissertation. University of Tampere, Finland. Kaunisto, Mark. 2007. Variation and change in the lexicon: A corpus-based analysis of adjectives in English ending in -ic and -ical. Amsterdam: Rodopi. Kellett, Arnold. 1965. Isms and Ologies: A guide to unorthodox and non-Christian beliefs. London; Epworth Press. Kjellmer, Göran. 1984. ‘Why great : greatly but not big : *bigly?’, Studia Linguistica 38: 1–19. Krueger. John B. 1963. ‘The morphemes of “cranberry”’, American Speech 38: 155. Kruisinga, E. 1932. A handbook of present-day English, Part II, English accidence and syntax 3, 5th edition. Groningen: Nordhoff. Lieber, Rochelle. 1998. ‘The suffix -ize in English: implications for morphology’, pp. 12−33 of Morphology and its relation to phonology and syntax, edited by Steven G. Lapointe, Diane K. Brentari, and Patrick M. Farrell. Stanford: ­ CSLI. Lieber, Rochelle. 2004. Morphology and lexical semantics. Cambridge: Cambridge University Press. Lubbers, K. 1965. ‘The development of “-ster” in Modern British and American English’, English Studies 46: 449−70. Malkiel, Yakov. 1977. ‘Why ap-ish but worm-y?’, pp. 341−64 of Studies in descriptive and historical linguistics: Festschrift for Winfred P. Lehmann, edited by Paul J. Hopper. Amsterdam: John Benjamins. Marchand, Hans. 1966. Review of Affixal negation in English and other languages by Karl E. Zimmer, Language 42: 134−42. Marchand, Hans. 1969. The categories and types of present-day English word-­ formation, a synchronic-diachronic approach, 2nd edition, Munich: C. H. Beck. Matthews, P. H. 2007. The concise Oxford dictionary of linguistics, 2nd edition. Oxford: Oxford University Press. Mencken, H. L. 1936. The American language: an inquiry into the development of English in the United States. 4th edition, New York: Knopf. Mencken, H. L. 1945. The American language: an inquiry into the development of English in the United States. Supplement one. New York: Knopf. Mencken, H. L. 1948. ‘What the people of American towns called themselves’, American Speech 23: 161−84. Murray, James A. H. 1888–1933. Chief editor of A new English dictionary on historical principles, founded mainly on materials collected by the Philological Society. Oxford: Clarendon Press. [Later re-titled, by the publisher, the Oxford English dictionary.] Muthmann, Gustav. 1999. Reverse English dictionary, based on phonological and morphological principles. Berlin: Mouton de Gruyter.

438   references Nida, Eugene A. 1946. Morphology: the descriptive analysis of words. Ann Arbor: University of Michigan Press. Oxford English dictionary, see Murray, James A. H. [Most references here are to the on-line version]. Payne, John and Huddleston, Rodney. 2002. ‘5. Nouns and noun phrases’, pp. 323−523 of The Cambridge grammar of the English language, chief authors Rodney Huddleston and Geoffrey K. Pullum. Cambridge: Cambridge U ­ niversity Press. Plag, Ingo. 1999. Morphological productivity: structural constraints in English derivation. Berlin: Mouton de Gruyter. Plag, Ingo and Baayen, Harald. 2009. ‘Suffix ordering and morphological processing’, Language 85: 109−52. Quirk, Randolph, Greenbaum, Sidney, Leach, Geoffrey, and Svartvik, Jan. 1985. A comprehensive grammar of the English language. Harlow: Longman. Reed, Allen Walker. 1933. ‘The scope of the American dictionary’, American speech 8: 10−20. Rex, Richard. 1975. ‘The origin of beat-nik’, American Speech 50: 329−31. Sapir, Edward. 1921. Language: An introduction to the study of speech. New York: Harcourt Brace. Sinclair, John. 2001. Founder editor-in-chief of Collins Cobuild English dictionary for advanced learners. Glasgow: HarperCollins. Sweet, Henry. 1891. A new English grammar, logical and historical. Oxford: Clarendon Press. Watkins, Calvert. 1985. The American Heritage dictionary of Indo-European roots. Boston: Houghton Mifflin. Waugh, Evelyn. 1936. Waugh in Abyssinia. London: Longmans, Green. Wentworth, Harold. 1941. ‘The allegedly dead suffix -dom in Modern English’, PMLA 56: 280−306. Wilkes, Charles. 1845. Narrative of the United States exploring expedition during the years 1838, 1839, 1840, 1841, 1842. Philadelphia: Lea and Blanchard. Zirkel, Linda. 2010. ‘Prefix combinations in English: structural and processing factors’, Morphology 20: 239−66.

Indexes References are to section number and to divisions within sections, not to page numbers. For example, ‘9.3.12’ refers to section 9.3.12 within Chapter 9, and ‘9.3.12b’ to division b within 9.3.12. For each prefix and suffix, its major discussion is given first in bold type. This is followed by further mentions. Generally, discussion of an affix in Chapters 1–4 provides a partial preview of its main entry, in one of Chapters 5–10. The genetic origin of suffixes is abbreviated: Gk, Greek; Gmc, Germanic; Rom, Romance.

Index of derivational prefixes a- (a-shore) Gmc 8.2.46, 8.4.20, 8.4.1 a(n)- (a-political) Gk 5.8, 5.9 after- Gmc 6.7.1, 3.3c all- Gmc 6.2 ante- Rom 6.7.3, 3.2c, 6.7.4 anti- Gk 5.24, 3.2k, 5.26, 6.8.2 arch- Gk 6.8.1, 5.1d auto- Gk 6.8.9 back- Gmc 6.7.5 be- Gmc 7.3.1, 5.1d bi- Rom 6.2, 3.3d by- Gmc 6.8.6 circum- Rom 6.6.3, 3.2j co- Rom 6.8.7 contra- Rom 5.27 counter- Rom 5.25, 2.6c, 5.26 de- Rom 5.10, 5.16, 7.3, 7.3.3b demi- French 6.2 di- Gk 6.2 dis- Rom 5.11, 5.12–13, 5.16, 5.19, 7.3 down- Gmc 6.6.5, 3.3c em-, see enen-/em- Rom 7.3.3, 5.10b, 5.11d, 7.3.4 epi- Gk 6.4.9 ex- Rom 5.22, 3.2h, 5.23 extra- Rom 6.5.3, 3.3b, 6.5.4 fore- Gmc 6.7.2, 6.7.4 half- Gmc 6.2, 3.3d hemi- Gk 6.2 hetero- Gk 6.8.10 hind- Gmc 6.7.5

homo- Gk 6.8.10, 3.2k hyper- Gk 6.3.3, 6.4.4, 6.4.6a hypo- Gk 6.3.3 il-, see inim-, see inin- (in-land) Gmc 6.6.5, 3.3c in/-im-/il-/ir- (in-correct) Rom 5.3, 2.2a, 2.3f, 3.2f, 5.4–5, 5.7 infra- Rom 6.4.2, 6.4.6b inter- Rom 6.5.2 intra- Rom 6.5.1 ir-, see inmal- Rom 5.20, 3.2j, 5.21 many- Gmc 6.2, 3.3d maxi- Rom 6.3.2, 3.2k mega- Gk 2.4g, 6.3.1, 2.5m, 3.2c meta- Gk 6.8.8 micro- Gk 6.3.1 mid- Gmc 6.7.6 mini- Rom 6.3.2 mis- Gmc & Rom 5.17, 2.3f, 5.18–19, 5.21 mono- Gk 6.2, 2.2d, 2.6g, 3.2c, 3.3d multi- Rom 6.2, 3.3d neo- Gk 6.7.7 non- Rom 5.6, 2.6c, 3.2h, 3.3b, 5.7, 5.9, 5.13, 5.23 off- Gmc 6.6.5, 3.3c omni- Rom 6.2 on- Gmc 6.6.5 one- Gmc 6.2, 2.2d, 2.6g, 3.3d out- Gmc 6.6.5, 3.3c, 7.1c over- Gmc 6.4.3, 3.3c, 6.4.4

index of derivational prefixes   441 paleo- Gk 6.7.7 pan- Gk 6.2 para- Gk 6.6.1 peri- Gk 6.6.3 phil(o)- Gk 2.7d poly- Gk 6.2 post- Rom 6.7.1 pre- Rom 6.7.3, 6.7.4 pro- Rom 6.8.2, 3.2k, 6.7, 6.8.4 proto- Gk 6.7.7 pseudo- Gk 6.10 quasi- Rom 6.10 re- Rom 6.9, 2.4d, 2.5d retro- Rom 6.7.5 self- Gmc 6.8.9 semi- Latin 6.2, 3.2k, 3.3d step- Gmc 6.8.5 sub- Rom 6.4.1, 3.3c, 6.4.6a, 6.8.4 super- Rom 6.4.3, 6.4.5, 6.4.6a, 6.5.4

supra- Rom 6.4.7 sur- Rom 6.4.8 tele- Gk 6.6.2 th(o)rough- Gmc 6.6.4 three- Gmc 6.2 trans- Rom 6.6.4 tri-, Gk & Rom 6.2 two- Gmc 3.3d, 6.2 ultra- Rom 6.3.4, 3.2b, 6.4.5, 6.4.6b, 6.5.4 un-1 (un-happy) Gmc 5.2, 2.2a, 2.3f, 2.5b, 3.2f, 3.3b, 5.4–5, 5.7, 5.12, 5.15, 5.18 un-2 (un-tie) Gmc 5.14, 2.3f, 5.15–16, 7.3 under- Gmc 6.4.1, 3.3c, 6.8.4 uni- Rom 6.2, 2.2d, 2.3a, 2.3e, 2.6g, 3.3d up- Gmc 6.6.5, 3.3c very- Rom 6.5.4 vice- Rom 6.8.3, 3.2k, 6.8.4 well- Gmc 6.3.5

Index of derivational suffixes -ability/-ibility Rom 9.2.7, 9.2.1c -able/-ible Rom 8.2.36, 8.4.7, 5.2b, 5.4c, 8.2.35, 8.2.37–8, 8.4.1–2, 8.4.8–10, 9.2.7 -acy, see -cy -ade Rom 9.7b -age Rom 9.4.15 -aire Rom 9.3.12a -al (dismiss-al) Rom 9.4.13, 8.2.19a -al/-ar (nation-al) Rom 8.2.19, 3.2f, 8.2.20, 8.2.35 -an/-n Rom 8.3.8, 9.3.7, 8.3.10–11, 9.3.8 -ana, see -iana -ance/-ence/-ancy/-ency Rom 9.4.9, 2.8j, 3.2j, 9.4.10 -ancy, see -ance -ant/-ent (inhabit-ant) Rom 9.3.4, 2.5j, 9.3.5 -ant (resist-ant), see -ent -ar, see -al -ar, see -er -ard/-art Gmc & Rom 9.3.12c, 3.2e, 8.3.1 -arian Rom 9.3.7e -arium Rom 9.6.1 -art, see -ard -ary Rom 8.2.39, 9.6.1, 9.3.12b -ate (affection-ate) Rom 8.2.44 -ate (consul-ate) Rom 9.5.6 -ate (liquid-ate) Rom 7.3.5, 2.5j, 7.5.1, 7.6.2d, 7.6.2h -atic, see -ic -ation/-ion/-ication/-tion/-ition Rom 9.4.7, 2.5g–j, 2.8j, 9.4.10 -ative, see -ive

-atory/-ory Rom 8.4.14, 9.6.1, 8.4.1, 8.4.15 -ature, see -ure -c Rom 9.3.12d -cian, see -ician -city, see -ity -cy/-acy Rom 9.2.3, 2.5k, 2.6a, 2.6h, 9.2.6 -dom Gmc 9.5.4, 9.5.5 -ean, see -ian -ed Gmc 8.2.12, 8.2.13, 8.2.35 -ee (boot-ee), see -y -ee (employ-ee) Rom 9.3.14, 2.6d, 3.2e -eer Rom 9.3.2, 9.3.3 -en (deep-en) Gmc 7.3.2, 2.3f, 2.6a, 2.6e, 2.8a, 2.8g, 3.2e, 7.3.4, 8.2.10b -en (wood-en) Gmc 8.2.10, 2.2b, 2.3f, 3.2f, 8.2.13, 8.2.17 -ence, see -ance -ency, see -ance -ene, see -ine -ent (resid-ent), see -ant -ent/-ant (emerg-ent) Rom 8.4.16, 3.2j, 8.4.1 -eous, see -ous -er/-ar/-or/-ier Gmc & Rom 9.3.1, 2.3b, 2.3e, 2.6d, 3.3e, 9.3.3, 9.3.5, 9.3.8, 9.3.11 -er(s) recent 6.11.5 -ern Gmc 8.2.41 -ery/-ry Rom 9.5.9, 2.3d–e -ese (Chin-ese) Rom 8.3.7, 9.3.7 -ese (journal-ese) Rom 8.3.7b

index of derivational suffixes   443 -esque Rom 8.2.7, 3.2f -ess Gk 9.3.9 -ette Rom 9.3.12f -ety, see -ity -ey, see -y -fashion Rom 10.6.4b -fold Gmc 10.6.4a, 8.1d -free Gmc 8.2.29, 3.3b, 8.2.30, 8.2.32, 8.2.34, 8.2.37 -ful (cup-ful) Gmc 8.2.22c -ful (grace-ful) Gmc 8.2.22, 8.4.4, 3.2f, 3.3b, 5.1e, 8.2.23–4, 8.2.28, 8.2.33–4, 8.4.1–2, 8.4.4–5, 8.4.10, 8.4.12, 8.4.15 -fy, see -ify

-ish Gmc 8.2.4, 8.3.2, 8.4.17, 2.2c, 2.4b, 2.5c, 3.2b, 3.2e, 6.1b, 8.2.5–6, 8.2.9, 8.4.1 -ism Gk 9.5.7, 2.5j, 3.2k -ist Gk 9.3.6, 2.5j, 9.3.8 -ite Gk & Rom 8.3.5, 9.3.7 -ition, see -ation -itis Gk 9.5.10b -itive, see -ive -iture, see -ure -ity/-ety (itch-ity) Gmc 8.2.8e, 8.4.2b -ity/-ety/-ty/-city (modern-ity) Rom 9.2.1, 2.4a, 2.5j, 2.6j, 9.2.6–7 -ive/-tive/-ative/-itive Rom 8.4.6, 8.4.1–2, 8.4.8, 8.4.15 -ize Gk 7.3.6, 7.4–7, 2.2e, 2.5j, 2.6f

-hood Gmc 9.5.1, 9.5.5

-kin(s), see -ikin(s)

-i Indo-Iranian & Semitic 8.3.6, 9.3.7 -ian/-ean Rom 8.3.9, 9.3.7, 8.3.10–11, 9.3.8 -iana/-ana Rom 9.7a -ibility, see -ability -ible, see -able -ic/-atic/-tic Gk & Rom 8.2.16, 8.3.3, 8.4.2b, 4.2g, 8.2.17–18, 8.2.20, 8.3.11 -ic-al Rom 8.2.20, 8.2.38 -ication, see -ation -ician/-cian Rom 9.3.7d -icide Rom 9.5.10a -ie, see -y -ier, see -er -ify/-fy Rom 7.3.6, 7.4–7, 2.2e, 2.6f -ikin(s)/-kin(s) Gmc 6.11.4 -in Gmc 9.7d -ine/-ene Rom 8.2.45 -ing (meet-ing) Gmc 9.4.6 -ing (bedd-ing) Gmc 9.4.6c -ion, see -ation -ious, see -ous -ise, see -ize

-le Gmc 7.1d -less Gmc 8.2.31, 8.4.3, 3.2f, 3.3b, 5.1e, 5.5b, 8.2.32–5, 8.4.1–2, 8.4.5, 8.4.9–10, 8.4.12 -let Rom 6.11.1 -like (lady-like) Gmc 8.2.1, 3.3b, 8.2.3, 8.2.5 -like (old-like) Gmc 8.2.1b -ling Gmc 6.11.2 -ly (clever-ly) Gmc 10 1–5, 8.2.2b -ly (mother-ly) Gmc 8.2.2, 2.2c, 8.2.3, 8.2.6, 8.2.23, 8.2.26, 8.2.35 -maker Gmc 9.3.13d, 3.3e -man Gmc 3.3f, 9.3.13e, 3.2e, 8.3.1 -manship Gmc 9.5.3 -ment Rom 9.4.8, 2.3c, 2.3e, 2.4e, 2.8j, 3.2g -monger Gmc 9.3.13a, 3.3e -most Gmc 8.2.42 -n, see -an -ness Gmc 9.2.5, 2.5a, 2.6a, 2.6h–j, 3.2b, 3.2d–f, 5.4e, 9.2.6–7 -nik Russian & Yiddish 9.3.12e

444    index of derivational suffixes -ology Gk 9.5.8, 3.2k -ometer Gk 9.7c -or, see -er -orium Rom 9.6.1 -ory, see -atory -ous/-ious/-eous Rom 8.2.14, 2.4f, 2.5e, 2.5h–j, 2.6j, 3.2f–g, 8.2.15, 8.2.18, 8.2.21, 8.2.24, 8.2.27, 8.2.35

-t (heigh-t), see -th -t (gif-t) Gmc & Rom 9.4.14 -th (ten-th) Gmc 9.2.4 -th/-t (warm-th) Gmc 9.2.4 -tic, see -ic -tion, see -ation -tive, see -ive -ty, see -ity

-phile/-philic/-philia Gk 2.7d, 6.8.10 -phobe/-phobic/-phobia Gk 2.7d, 6.8.10 -proof Gmc 8.2.40, 8.4.18, 3.3b, 8.4.1

-ure/-iture/-ature Rom 9.4.12, 2.8j

-ry, see -ery -ship Gmc 9.5.2, 9.5.5 -smith Gmc 9.3.13c, 3.3e -some (four-some) Gmc 8.2.25 -some (trouble-some) Gmc 8.2.25, 8.4.11, 8.2.26–8, 8.2.30, 8.4.1, 8.4.12, 8.4.15 -ster Gmc 9.3.10, 3.3e, 9.3.11 -stress Gmc 9.3.10 -style Rom 10.6.4b, 8.1d -sy (art-sy) Gmc 8.2.8d -sy (dogg-sy), see -y, 6.11.3

-ward(s) Gmc 10.6.1, 8.2.43, 8.1a -way(s) Gmc 10.6.2, 8.1d -wise Gmc 10.6.3, 8.1d, 8.2.14a -woman Gmc 3.3f, 9.3.13e, 3.2e -worthy Gmc 8.4.19, 3.2b, 3.3b, 8.4.1 -wright Gmc 9.3.13b, 3.3e -y (discover-y) Rom 9.4.11 -y/-ey/-ie/-sy/-ee (dogg-y) recent 6.11.3, 8.2.8d -y (greed-y) Gmc 8.2.8, 8.4.13, 8.2.9, 8.2.11, 8.2.13, 8.2.15, 8.2.35, 8.4.1, 8.4.15 -y (iron-monger-y) Gmc 9.3.13a -y (jealous-y) Rom & Gk 9.2.2, 2.8j, 9.2.6

General index A, transitive subject function 1.2a–b, 8.4 ablative case 6.4.1b, 6.7.3a accusative case 6.4.1b, 6.7.5a, chapter 12 acting semantic type 8.4.2a, appendix activity nominalizations 9.4 adaptations 4.2, 6.6.1–2, 6.8.9a Addison, Joseph 6.8.6 adjectivalizations, chapter 8 adverbializations, chapter 10 affect semantic type 5.14, 8.4.2a, 8.4.13, appendix affix, criteria for 1.1c, 2.5k age semantic type 2.9, 3.2b, 5.5b, 8.2, 10.4, appendix agentive nominalizations 9.3 Alice (in Wonderland) 5.2a, 9.3.1d ambitransitive verbs 1.2b, 5.14c, 6.9, 7.1b, 7.1d, 8.4.7, 9.2.7b annoying semantic type 2.9, 5.11b, 8.4.2a, 8.4.11, 8.4.14b, 8.4.15, 9.2b, 9.3.1b, 9.4.1, appendix antonyms 5.5a–b, 6.3.2, 6.3.5d, 6.3.5f, 6.7.7c, 7.3.2b arguments 1.1a, 1.2 attention semantic type 8.4.2a, appendix back formation 1.1h, 5.12e, 5.20b, 6.6.2, 6.7, 7.6.2c–f, 8.3.1, 9.4.12c base 1.1b Beckett, Samuel 2.6a beginning semantic type 8.4.2a, appendix blends 4.2p

Boswell, James 6.11.3a bound root 2.5h–k Carroll, Lewis 5.2a, 9.3.1d Chaucer, Geoffrey 5.17a clause 1.1a clitic 6.7; see also proclitic colour semantic type 2.6c, 2.8g, 3.2b, 5,5b, 8.2, 10.4, appendix combinations of affixes, chapter 11 ‘combining forms’ 2.1b comparing semantic type 8.4.2a, appendix competition semantic type 8.4.2a, appendix complement clause construction 9.4.6b complementary terms 5.2d, 5.5a–b, 5.7a, 5.24b, 6.4.1d, 6.4.3d, 6.6.4c, 6.6.5b, 6.7.5a, 6.7.5d, 6.8.2, 6.8.10 compounds 2.1, 3.1–3 consonants of English 1.3b contrastive stress 1.1f, 7.2b converse 5.10a–b conversion, see double duty copula clause 1.2c copula complement 1.2c, 8.2.26, 8.2.46, 8.4.20. 10.3 copula subject 1.2c core arguments 1.2a corporeal semantic type 8.4.2a, 8.4.13, appendix count noun 1.1b, 1.1e cran- morpheme 1.1c deciding semantic type 8.4.2a, appendix deputy, ways of describing 6.8.4

446   general index derivational affix 1.1b, 1.1d also as free form 3.2k, 3.3b, 6.3.2 criteria for 3.2 in scientific naming terminology 4.1a fallen into disuse 4.1c forms excluded 4.1b–c, 6.11, 8.1c multi-word scope 3.2h, 8.1g difficulty semantic type 5.5b, 10.4, appendix dimension semantic type 2.6c, 2.8h, 2.9, 3.2b, 5.5b, 8.2, 10.4, appendix diminutive derivations 6.11 diphthongs of English 1.3a double duty 2.8, 6.7.1d, 8.1b, 8.2.3, 8.2.11, 8.2.25, 8.4.2a, 9.1b, chapter 12 adjective or verb also as noun 2.8, 5.17c, 8.1b, 9.1b, 9.2–4 adjective or noun also as verb 2.8. 7.1a, 7.2–6, 11.2.2b adverb also as adjective, 8.1d, 8.2.43, 10.4, 10.6 emerging 10.5 noun also as adjective 2.8, 5.6b preposition also as noun or as verb 2.8 Dryden, John 5.4f emerging double duty 10.5 endearing derivations 6.11 free form 1.1c, 2.5h, 3.2–3 fusion 2.5, 3.2g Gilbert, W. S. 7.3.6a giving semantic type 8.4.2a, 8.4.14b, appendix Goldsmith, Oliver 6.11.3a grammatical word 3.2, 5.6a, 6.5 happening semantic type 8.4.2a, 8.4.14b, appendix head of noun phrase 1.1a helping semantic type 5.11b, 8.4.2a, appendix

Holmes, Sherlock 8.3.11a human propensity semantic type 2.6c, 3.2b, 5.5b, 8.2, 9.2b, 10.4, 10.5a, appendix Huxley, Thomas H. 5.8b inflectional affix 1.1b, 1.1e -ing as complementizer 9.4.6b internal change 1.1f, 3.2b, 9.4 intransitive clause 1.2 Johnson, Dr Samuel 5.4f, 5.12e, 6.8.4, 6.10a, 6.11.3a, 8.4.16a kings and queens, adjectives based on names 8.3.11 lexeme 1.1b lexical root 1.1b liking semantic type 2.9, 5.11b, 5.12c, 8.4.2a, 9.2b, 9.3.1b, 9.4.1, appendix locational derivations 6.4–7, 9.4, 9.6 magnitude and degree derivations 6.3 main clause 1.1a making semantic type 5.11b, 8.4.2a, appendix mass noun 1.1e model word for adaptations 4.2 morpheme 1.1a–c criteria for 1.1c, 2.5k morphological process 1.1f, 3.2b motion semantic type 8.4.2a, 8.4.13, 8.4.14b, appendix negative derivations, chapter 5 nominalizations, chapter 9 nominative case, chapter 12 nonce word 2.6, 5.2a, 5.11e, 7.3.6c, 8.1e, 8.2.25, 9.5.4, 9.5.9–10, 10.6.1 noun phrase (NP) 1.1a number derivations 6.2

general index   447 O, transitive object function 1.2a–b, 8.4 obeying semantic type 5.11b, 8.4.2a, appendix opposites, types of, see antonym, complementary terms, converse ordering of derivations 1.1d orthography 1.3 phoneme 1.3 phonological word 3.2, 5.6a, 6.5 phrasal verbs 2.8e, 8.1g, 8.3.2b, 8.4.7c, 9.3.1f, 9.3.13d phrase 1.1a physical property semantic type 2.6c, 2.8g, 3.2b, 5.5b, 8.2, 10.4, appendix postponing semantic type 8.4.2a, appendix pragmatics 3.2.b–c, 6.8.8, 7.5.1, 7.6.2b predicate 1.1a, 1.2 prefix properties, compared with suffixes, 2.7 primary verbs 2.9, appendix proclitic 3.2b, 5.6a, 6.7 productivity 2.6, 3.2f proper names adjectivalizing of 6.3 verbalizing of 7.7 qualification semantic type 5.5b, 10.4, appendix redundant use of un-2 5.14f relating semantic type 8.4.2a, appendix repetition suffix 6.9 rest semantic type 8.4.2a, appendix result nominalizations 9.4 role derivations 6.7 root 1.1b S, intransitive subject function 1.2a–b, 8.4

S = A ambitransitive verbs 1.2b, 7.1b. 7.1d S = O ambitransitive verbs 1.2b, 5.14c, 6.9, 7.1b, 7.1d, 8.4.7a, 8.4.8, 8.4.13, 8.4.18, 9.2.7b scientific affixes 4.1a secondary verbs 2.9, appendix semantic roles 2.9, 8.4.11, 9.4.1a, appendix semantic types 2.9, 2.6b, 3.2b, appendix adjective types 5.5, 7.3.2b, 8.2, 9.2b, 10.4, appendix verb types 2.9, 5.11b, 5.14b, 5.12c, 8.4.2, 8.4.6b, 8.4.11, 8.4.13–15, 9.2b, 9.3.1b, 9.4.1, appendix semi-affix 3.3c, 3.3g Shakespeare, William 6.4.4a, 9.3.10a, 9.5.5 Shaw, G. B. 8.2.2c similarity semantic type 5.5b, 10.4, appendix social contract semantic type 8.4.2a, appendix speaking semantic type 5.11b, 8.4.2a, 8.4.13, 8.4.14b, appendix Spectator, The 6.8.6 speed semantic type 3.2b, 8.2, 10.4, 10.5b, appendix states nominalizations 9.2, 9.4 stem 1.1b stress 1.1f, 1.3d, 2.5d, 2.5m–p, 3.1c, 3.2a–c, 3.3f, 4.2m, 6.1c, 7.2b, 8.1f, 9.1c subordinate clause 1.1a; see also complement clause construction suffix properties, compared with prefixes 2.7 syntactic orientation of derived adjectives 8.4 temporal derivations 6.7 Thackeray, W. M. 4.2a thinking semantic type 5.11b, 8.4.2a, appendix

448   general index time adverbs 10.1–2, 10.4a transitive clause 1.2 transitivity 1.2 trying semantic type 8.4.2a, appendix using semantic type 5.11b, 8.4.2a, appendix value semantic type 3.2b, 5.5b, 8.2, 10.4, 10.5b, appendix verb phrase 1.1a verbalizations, chapter 7

Victoria, Queen 7.7a volition semantic type 10.4, appendix vowels of English 1.3a wanting semantic type 8.4.2a, appendix Waugh, Evelyn 5.2d Wilkes, Charles 8.3.4 word 1.1b, 3.2 Wordsworth, William 6.6.4a zero derivation, see double duty