Grammar in English learners' dictionaries 9783111340074, 9783484309166

Citation preview

EJTOMMIFMgA

S

LEXICOGRAPHICA Series Maior Supplementary Volumes to the International Annual for Lexicography Suppléments à la Revue Internationale de Lexicographie Supplementbände zum Internationalen Jahrbuch für Lexikographie

Edited by Sture Allén, Pierre Corbin, Reinhard R. K. Hartmann, Franz Josef Hausmann, Hans-Peder Kromann, Oskar Reichmann, Ladislav Zgusta

16

Published in cooperation with the Dictionary Society of North America (DSNA) and the European Association for Lexicography (EURALEX)

Marcel Lemmens / Herman Wekker

Grammar in English Learners' Dictionaries

Max Niemeyer Verlag Tübingen 1986

CIP-Kurztitelaufnahme der Deutschen Bibliothek Lemmens, Marcel: Grammar in English learners' dictionaries / Marcel Lemmens ; Herman Wekker. - Tübingen : Niemeyer, 1986. (Lexicographica : Series m a i o r ; 16) NE: Wekker, Herman:; Lexicographica / Series maior ISBN 3-484-30916-4

ISSN 0175-9264

© Max Niemeyer Verlag Tübingen 1986 Alle Rechte vorbehalten. Ohne Genehmigung des Verlages ist es nicht gestattet, dieses Buch oder Teile daraus photomechanisch zu vervielfältigen. Printed in Germany. Druck: Weihert-Druck GmbH, Darmstadt.

CONTENTS

Preface Chapter I. Introduction 1.1

The background

1.2

The five learners' dictionaries

1.3

2.2

2.3

Preliminaries to a new coding system

13 15

2.5

17

The Oxford Advanced Learner's Dictionary of Current English (ALP)

17

2.1.1

The coding system: introduction

17

2.1.2

Nouns

17

2.1.3

Adjectives

18

2.1.4

Verbs

19

The Longman Dictionary of Contemporary English (LDOCE)

21

2.2.1

The coding system: introduction

21

2.2.2

The letters

23

2.2.2.1

Nouns

24

2.2.2.2

Adjectives

25

2.2.2.3

Adverbs

26

2.2.2.4

Verbs

27

2.2.2.5

Pronouns

29

2.2.3

The numbers

29

The Oxford Dictionary of Current Idiomatic English Volume 1 (ODCIE 1)

2.4

9 12

Notes to Chapter I Chapter II. Describing the five coding systems 2.1

9

30

The Oxford Dictionary of Current Idiomatic English Volume 2 (ODCIE 2)

39

The Longman Dictionary of Phrasal Verbs (LDOPV)

45

Notes to Chapter II

51

Chapter III. A critical review of the coding systems

53

3.1

53

3.2

3.3

3.4

3.5

ALD 3.1.1

The Introduction

53

3.1.2

The coding system used in the entries

61

3.1.3

Conclusion

LDOCE

62 63

3.2.1

The Introduction

64

3-2.2

The coding system used in the entries

69

3.2.3

Conclusion

ODCIE 1

73 73

3.3.1

The Introduction

74

3.3.2

The coding system used in the entries

78

3.3.3

Conclusion

80

ODCIE 2

81

3.4.1

The Introduction

81

3.4.2

The coding system used in the entries

85

3.4.3

Conclusion

88

LDOPV

89

3.5.1

The Introduction

89

3.5.2

The coding system used in the entries

93

3.5.3

Conclusion

98

Chapter IV. Towards a more adequate coding system

99

4.1

Verbs

99

4.2

Nouns

110

4.3

Adjectives

111

Notes to Chapter IV

114

References

115

PREFACE

This book, is intended as a contribution to the current debate on the coding of grammatical Information

in dictionaries

for foreign

language

learners.

It is

primarily based on an unpublished thesis by Marcel L e m m e n s (1985), written under the supervision of Herman Wekker. Parts of it were adapted and presented in papers which w e read at the EURALEX Seminar on "The Dictionary and the Language Learner" at Leeds in April 1985, and to the English Departments of Lund University in May 1985 and Birmingham University in December 1985. We are very grateful to our audiences for their reactions to our ideas and to several friends and colleagues for c o m m e n t i n g on an earlier version of this text. In particular, w e wish to thank Tony Cowie, Patrick Hanks, Reinhard

Hartmann,

Stig Johansson and Pieter Loonen for reading the text and saving us from quite a few errors. However, in no w a y do we imply that they agree with any of our c r i t i c i s m s or ideas. W e accept full responsibility for the views expressed here. Thanks also go to Ninette Nelis, Nico de Milliano and Diane Crook for helping us produce this text in the appropriate form. In

this

book

we

deal

with

monolingual EFL dictionaries:

the

grammatical

coding

systems

in

five

the Oxford Advanced Learner's Dictionary of

Current English (ALD, 3rd edition,

revised and reset, 1980), the Longman

Dictionary of Contemporary English (LDOCE, 1978), the two volumes of the Oxford Dictionary of Current Idiomatic English (ODCIE 1, 1975 and ODCIE 2, 1983), and the Longman Dictionary of Phrasal Verbs (LDOPV, 1983). The structure of the book is as follows: Chapter I (Introduction) briefly discusses

the general problem of including grammar

attention to what

little is now known

in dictionaries,

and pays

from recent surveys about the language

needs and reference skills of foreign dictionary users. Chapter II (Describing the five coding systems) provides an introductory description of each of the g r a m m a t i c a l coding systems examined. Chapter III (A critical review of the coding systems) deals in some more detail with the strengths and weaknesses of each of the systems. Finally, in Chapter IV (Towards a more adequate coding system) we submit to the reader an outline of an alternative proposal which, we

8

feel, overcomes some of the shortcomings of the other coding systems, and has the advantage of being consistent, self-explanatory and reasonably easy to use. Like many colleagues, we are critical of the lexicographical work that has been done in the area of grammatical coding (which, in our view, should include fixed collocations and usage). However, we would like to emphasize that our proposal leans heavily not only on what we have found in the five dictionaries under

review,

but

also on

the criticisms

that

the compilers

of

these

dictionaries have provoked from fellow-lexicographers and fellow-linguists. It is perhaps also worth mentioning here that our own background is in English syntax and linguistics, not specifically in lexicography. We are aware that this may cause a certain bias in the views we express. Still, we have been persuaded by colleagues to believe that a "syntactic" contribution to the current debate on grammatical coding would be appropriate and helpful.

Nijmegen, Spring 1986

M.L. & H.W.

CHAPTER X INTRODUCTION

1.1 The background Lexicographers and foreign language teachers would probably all agree that an Ideal learner's dictionary should provide the user not only with a full description of the meanings and usage of language forms, but also (especially for encoding^)) with the necessary grammatical guidance (cf. Carstensen 1969:16). As Bejoint (1981:210) puts it, "the best dictionary for encoding is one that provides the most detailed guidance on syntax and collocation, including

perhaps

advice

on

pitfalls

to avoid." Every

good

learner's

dictionary should thus to some extent also be a good pedagogical grammar, giving Information with examples on how to form sentences and phrases in the foreign language by analogy. The compilers of the currently available

learners' dictionaries of

E n g l i s h ^ ) have recognized this need for grammatical information, and have attempted to capture the intricacies of English grammar at least partly In coded forms. However, it is well known that there are problems with the specification of grammar in EFL dictionaries. The difficulties are of two main kinds: firstly, there is the problem of arrangement (i.e. unlike the entries in the dictionary, the grammatical information in the dictionary does not lend itself to alphabetical ordering, and is thus far less easily accessible), and secondly, there is the related problem of grammatical labelling (i.e. devising a system of coding which is both reliable, clear and usable requires a great deal of ingenuity). To these two difficulties we would add the point that, in order to avoid confusion, the grammatical Information supplied should not be in conflict with the information generally found in the mainstream grammars of modern English. The users of learners' dictionaries are often also users of the well-known survey grammars of English (in particular those by Randolph Quirk and his associates).

Our knowledge

of English

structures has

increased

enormously over the past 10 to 15 years, and this should somehow be reflected in EFL dictionaries. Moreover, in recent years we have gained a far greater

10

Insight not only into the structural differences between languages, but also into processes of foreign language learning. What we need are dictionaries which take into account the relevant findings of (applied) linguistics and language pedagogy, paying particular attention to the areas of English grammar and usage which are known to cause problems to foreign learners from different linguistic backgrounds. There

appears

to be a considerable

demand

among

foreign users of

dictionaries for grammatical information of some kind. Tomaszczyk (1979:112) reports that 70% of his 449 informants (foreign students at American colleges and Polish universities,

and Polish foreign language students,

language

instructors and translators) answered that they used their dictionaries to solve grammatical comparable

problems.

to Tomaszczyk's

B/joint's because

(1981:215) results are not

his

questionnaire

quite

and his group of

informants (122 second, third and fourth year students of English at the University of Lyon) were rather different from Tomaszczyk's; moreover, Bejoint was exclusively concerned with the use of monolingual general dictionaries of English. He reports that 53% of his students said they looked for grammatical information in their dictionaries. Hartmann (1982:82) found that 61.6% of his informants (185 English learners and teachers of

German In south-west England)

said

points

that

they

often

looked

up

grammatical

in

their

bilingual

dictionaries (German-English). Although again there are certain differences between Hartmann's survey and the others, similar^)

and

these results are

strikingly

a xi point to a real need for grammatical guidance.^^

We also know, of course, that foreign students do not seem to make full use of the grammatical information supplied by their dictionaries. Students appear to use their EFL dictionaries

mainly

for

decoding

activities

(principally reading), and often neglect the encoding information (for example on syntax). Bejoint (1981:219) concludes from this rather pessimistically that "monolingual

dictionaries

are not used

as fully as

they

should:

their

introductions are not commonly referred to, and neither are the coding systems for syntactic patterns. Certainly many students are not even aware of the riches that their monolingual dictionaries contain." He also notes that "it is for

encoding

that

students

need

the

most

information;

information which is the most difficult to supply;

it is

encoding

and yet this is the

information which students use the least. Lexicographers, consequently, should be wary of embarking upon innovations to help students with their encoding activities. [. . .] Students need to be taught how to use the monolingual dictionaries which they already possess so as to get the most use out of them."

11

Cowie (1983:107) confirms this view, noting that there is "considerable evidence that foreign learners use their EFL dictionaries for interpretive rather than productive purposes, and may in addition be disinclined to master the systems used to codify grammatical patterning". He recommends that students should be taught to use such information to fuller advantage. In addition, it has been observed that examples play a role which is at least as important as that of the grammatical codes. According to Bejoint (1981:218), the quotations and examples in the dictionaries are studied far more frequently than the codes. ^ ^ While agreeing that students should be taught and encouraged to use the grammatical coding systems more fully, we also wonder what could be done by publishers and dictionary-makers themselves to improve the quality of the current coding systems.^^ Compilers of pedagogical dictionaries seem to be faced with a paradox and a dilemma. The paradox is that dictionary users indicate, on the one hand, that they highly appreciate guidance on grammatical matters, while, on the other hand, they do not appear to make full use of all the grammatical information which is already supplied: very few learners read the introductory sections in their dictionaries where the coding system is explained, and the tables of codes are hardly studied. Bejoint (1981:216) reports that 55% of his students (and note that they constitute the most advanced category of learner for which the dictionaries are intended) only read the introductory pages cursorily, others (34%) did not study them at all. The lexicographer's dilemma, obviously, is whether to abandon the grammatical coding scheme altogether, or to improve it in such a way that it will be used more fully, or to leave it just as it is. B/joint's advice to lexicographers not to embark upon innovations to help students with their encoding activities reflects a fairly general feeling to leave things more or less as they are. In support of B/joint's recommendation, Cowie (1981a:205) takes the view that it is the teachers' responsibility to make users "more aware of the riches which their monolingual dictionaries already contain." Perhaps it would be wise to restrict oneself first to dictionaries of limited coverage, and then to attempt to extend the developed system to general EFL dictionaries. However, it would seem to us that on many points the current coding systems in general dictionaries can be greatly improved without too much difficulty. One

of

the

things

that

one

can

learn

from

the

above-mentioned

investigations into users' language needs and reference skills is that it cannot be taken for granted

that learners are prepared to study all the

12

relevant introductory sections and to refer to the tables of codes. Our solution would be to make the grammatical codes as transparent and selfexplanatory as possible, so that frequent reference to introductions or tables will become

unnecessary.

We

assume

that

the grammatical

codes

in

the

dictionaries are mainly intended for intermediate and advanced learners, some of whom will be foreign students of English, who will also attend introductory classes on English syntax. If that is the case, the terminology and the system presented in the dictionary should on the whole be consistent with

the

mainstream approaches adopted in syntax classes. What students should be taught is how to apply their knowledge and insight to dictionary items. In the following chapters an attempt will be made to show how this could be done, and to show that dictionary users are not exclusively to blame for not fully using the wealth of information hidden in their dictionaries. It will be argued that part of the problem lies in the Inadequacy of the current coding systems themselves. The time may be ripe for discussing the various

principles

underlying systems of grammatical coding, and for considering a possible alternative.

1.2

The five learners' dictionaries

As noted in our Preface, we have selected five monolingual EFL dictionaries whose grammatical coding systems will be examined in some detail.

The

dictionaries are: ALD

: Oxford Advanced Learner's Dictionary of Current English. General editor: A.S. Hornby with A.P. Cowie. Third edition. Oxford: Oxford University Press, 1974 (11th revised and updated impression, 1980). First published 1948. Second edition 1963.

LDOCE

: Longman Dictionary of Contemporary English. Editor-in-Chief: P. Procter. Harlow, Essex: Longman, 1978.

ODCIE 1:

Oxford Dictionary of Current Idiomatic English. Vol. I: Verbs with Prepositions and Particles. Editors: A.P. Cowie and R. Mackin. Oxford: Oxford University Press, 1975.

13

ODCIE 2:

Oxford Dictionary of Current Idiomatic English. Vol. II: Phrase, Clause & Sentence Idioms. Editors: A.P. Cowie, R. Mackin and I.R. McCaig. Oxford: Oxford University Press, 1983.

LDOPV

: Longman Dictionary of Phrasal Verbs. Editor: Rosemary Courtney. Harlow, Essex: Longman, 1983.

In Chapter II we shall try to provide a brief and an objective as possible description of the grammatical coding systems of these dictionaries. ALD and LDOCE are general pedagogical dictionaries, while the two volumes of ODCIE and LDOPV are more specialized. In Chapter III each of the systems will be looked at critically, in an attempt to discover the strengths and weaknesses of each dictionary. We shall point out, in particular, where we think the systems could be improved by simplifying them or by making them more consistent and easier to use for

the learner.

exclusively

to

the

In our coding

discussion of

verbs

we (i.e.

shall

pay

lexical

attention verbs

almost

and

their

complementation) in English, this being, not surprisingly, the one category that has been coded in all the five dictionaries. In Chapter IV, finally, we submit our own proposal for the coding of verbal structures, with

some

illustrations of its possible expansion to grammatical patterns with nouns and adjectives.

1.3

Preliminaries to a new coding system

We have noted above that in B^joint's (1981) survey a majority of his students (55%) stated

that they do not use the grammatical information in their

dictionaries at all, probably because they find the codes too hard to master. As a result, Be'joint (1981: 217) observes, students "are unable - or unwilling - to master the codes used in many dictionaries." Other complaints that have been heard from students are that very often they fail to find the specific information they are looking for, that examples illustrating a grammatical point are sometimes ambiguous, that the codes are awkward and hard to remember, that more often than not grammatical information found in one dictionary is contradicted by that in another (even in the case of companion volumes), that different dictionaries do not always define grammatical terms in the same way and do not recognize the same grammatical categories, and that even internally

14

dictionaries are frequently inconsistent. Any new coding system should, in our view, meet certain minimal conditions. We list the following (see also Chapter IV): a)

The grammatical codes should be unambiguous and easy to use. We propose that the codes should provide information about the grammatical function (do,

10, etc.)

and the grammatical category (NP, AdjP, etc.) of the

elements which may follow the verb as its complements. b)

Every verb entry (and, where appropriate, every sub-entry) will include one or more codes, all of which should be self-explanatory. It will remain necessary, of course, to have a Guide in the introductory pages to explain elementary function and category labels such as do, 10, NP and AdjP to the absolute beginner. The terminology should be that of the standard modern grammars of English.

c)

The dictionary user should have all the information he needs at hand, and should not be required to consult the Guide or Table of Codes in order to interpret the codes.

d)

All the grammatical patterns indicated in each verb-entry or sub-entry should be illustrated by example sentences arranged in the same order as the patterns.

e)

Transformational possibilities are another useful feature of learners' dictionaries. If passivization, indirect object movement, extraposition and the like are allowed, these transformations should also be indicated in the entry or sub-entry.

f)

Codes must be as explicit and complete as possible, and should not be collapsed; no part of the code should be put in brackets.

The scheme we shall propose in Chapter IV is intended to take account of these criteria, in an attempt at devising a pedagoglcally acceptable alternative to the currently used grammatical coding systems. Our system is centred round a classification of verb complementation into nine basic structures, but closely related alternatives for nouns and adjectives are also provided.

15

NOTES TO CHAPTER I

Following a now common practice, we adopt a division of language skills into "decoding" activities (reading, listening, translation from L2 to LI) and "encoding" activities (writing, speaking, translation from LI to L2). It is with the encoding information supplied by monolingual

English

dictionaries (particularly the grammar) that we shall be mainly concerned here (cf. Be'joint (1981:210), Cowie (1981:225)). Although our discussion is mainly restricted to EFL dictionaries, it is hoped that colleagues concerned with dictionaries for learners of languages other than English will also find elements here which may be relevant to them. As Bergenholtz and Mugdan (1985:10ff.) observe, the reliability of some of the results of these investigations is somewhat diminished by the fact that we do not know how truthfully informants have answered the questionnaires: they may have wanted to indicate how often they thought they ought to use their dictionaries, rather than how often they actually used them. There is a strong subjective element in many of the questions. The questions asked were sometimes also ambiguous and vague. Informants may indeed have been tempted to say how often they thought they looked up particular Information rather then how often they actually did so. The alternative research technique involving users' protocols under strictly controlled conditions of dictionary use, as advocated by Hatherall (1984), Wiegand (1985) and others, has clear advantages over the questionnaire, as long as steps are taken to increase the "naturalness" of the test situation.

W

Bergenholtz and Mugdan (1985:10-11): "Jedenfalls scheint der Bedarf nach grammatischer Information im W"6rterbuch erheblich zu sein."

O

In a series of papers Bergenholtz and Mugdan have shown that the treatment of grammar in monolingual and bilingual German dictionaries leaves a great deal to be desired (cf. Bergenholtz and Mugdan (1982); Mugdan 1984a);

Bergenholtz

(1984a,

1984b)). The reader

(1983,

is also referred

to

Bergenholtz and Mugdan (eds.) (1985) for critical discussions of the grammatical information found in German dictionaries, and some suggestions

16

for improving the coding systems. The Bergenholtz and Mugdan plan to produce a "grammatical dictionary" of German is worth mentioning here. Mugdan (1984b:5) describes the aims of their four-year project (1985-1989) as follows: 1)

to develop and experiment with new methods of dictionary-making, using a computer corpus of approximately 5 million words;

2)

to compile a German grammar suitable for purposes of lexicography, on which all descriptive devices used in individual dictionary entries are to be based;

3)

to determine the morphological and syntactic properties of some 2000 common German words, including all function words and some highfrequency nouns, verbs and adjectives;

4)

to present the results in the form of a dictionary which can serve as a source of information for linguists, authors of textbooks, grammars and dictionaries, teachers etc., and from which learners' dictionaries and the like could be derived.

Most of these points and some others were raised at a seminar at Essen in June 1984, organized by Henning Bergenholtz and Joachim Mugdan ("Essener (Colloquium zur Grammatik im W&rterbuch"). The proceedings of that meeting have been published by Niemeyer Verlag as Volume 3 of "Lexicographica Series Maior" (H. Bergenholtz and J. Mugdan (eds.) (1985)).

CHAPTER II DESCRIBING THE FIVE CODING SYSTEMS

2.1

The Oxford Advanced Learner's Dictionary of Current English (ALP)

2.1.1

The coding system: Introduction

ALD 1s claimed to be a dictionary "for the learner of English who has mastered the rules of English grammar and pronunciation" (Introduction, p.xiii).^ As a dictionary intended for decoding as well as encoding, It includes features enabling "the learner to use words correctly in sentences of his own" (p.xiv). The three types of information considered to be of greatest importance for the correct active use of headwords are :

(i) spelling and pronunciation, (ii)

grammatical patterns and forms, and (ill) examples of the contexts In which words are generally used. Of chief interest to us will be especially (ii), but also (ill). The grammatical information provided in the introductory pages (xxiv-xxv and xxviii-xxxix) concentrates mainly on nouns, adjectives and verbs (verbs being treated most extensively). The

grammatical information in ALD is

supplied in the form of codes, abbreviations and explicit grammatical forms.

2.1.2

Nouns

The nouns in ALD are followed by the abbreviation n, to which may be added the label JJI^ in cases where the noun occurs in the plural form only, e.g. scissors. In other cases additional information is given in brackets, e.g. police ri

18

(collective ri, always sing in form, used with j>1 v). If a noun has an irregular plural

form,

this

is

indicated

in brackets

after

the abbreviation ri,

e.g.hippopotamus...n (pi t\Jes /-siz/ or -mi /mai/). The only information in coded form which ALD gives on nouns concerns their countability. It uses the symbols [C], [U] and [C,U], but as a general rule countable nouns

are not

coded (only the less obvious ones). A noun is coded

[C,U] if it can be used either as a countable or an uncountable noun, e.g. drink. Noun entries may also contain information on are

typically used

for complementation.

the preposition(s) that

A distinction

is made

between

obligatory and optional prepositional phrases (optional prepositions being placed in brackets). For example, on/about, and after difference we find

at the entry for information we find t\J (between) in bold print.

Information on the use of pronouns, etc. is explicitly included in the entries, e.g.: that^...(pl those...) ad j, pron (contrasted with this, these. [T], [V], [X] and [WvJ. [D] is used for verbs followed by two sentence elements of which at least one is obligatory. The two e l e m e n t s "come in positions III and IV, often as INDIRECT OBJECT (the boy) + DIRECT OBJECT (a book). These nouns,

PRONOUNS,

etc., a l w a y s represent (REFER TO) something else, not each other" (p. xxix). These sentence patterns are exemplified

on page xxix of

the Guide to the

Dictionary, as follows: Give the boy a book.

(both sentence elements are obligatory)

Buy him a_ book.

(him is optional)

He aade her a_ good husband. Spray paint on the wall. Indirect object m o v e m e n t is indicated by means of prepositions (to or for) added to the grammatical codes, e.g.: buy

... 1 [D1 (for);

sell

...1

send

...1 [D1 (to);

..] ...

[D1 (_to); ..] ... ..] ...

In the case of sell, we find the following example sentences to illustrate the transformation: ...1^ sold my brother my car for 100 pounds/ brother

sold my car to my

for 100 pounds.

In certain cases information is given on whether passivization is allowed: e.g. the senses of give^ ... 15

[D1 n £ pass.] ... 17 [V3 often

pass.] ... 18

[X9 usu. pass.] ... . [I] is used for "verbs, most of which are INTRANSITIVEs, that need not be followed by anything in position III or IV, and need have no ADVERBIAL" (p. xxx).

[0] is added

Examples are: We paused.

to the

[I] if the verb is followed

by nothing at all-

28

He spoke with his mouth full of food. [L] stands for linking verbs followed by a complement in position III or an adverbial referring to the location of the subject. For example: She became president. He lives here. [T] is used for "verbs, many of which are TRANSITIVEs, that are followed by a noun or nounlike expression in position III as a DIRECT OBJECT which does not represent (REFER TO) the subject, unless REFLEXIVE" (p. xxxi). For example: She kicked the boy. [V] is used for "verbs that are followed, in position III, by a 2-part DIRECT OBJECT" (p. xxxi). This is, in fact, a sub-category of [T],

since it also

applies to transitive verbs followed by a direct object, but in the case of [V] the direct object is realized by a non-finite clause with a. subject. For example: _I saw the man leave. _I saw the man leaving. Transformation into the passive is only demonstrated for one sentence in the description of [V], i.e. for the type with the infinitive without to^, which gets an infinitive with tj> in the transformed sentence:

saw the man leave

becomes The man was seen to leave. [X] is used for verbs that are followed by a direct object in position III and a complement or an adverbial referring to the direct object. The two possible sentence patterns are: They considered him a fool. Put i_t In the box. There are 6 [Wv] codes for verbs: [Wvl] refers to a table of inflected forms of the verb to_ t>e (p. xxxviil). [ Wv2 ] refers to a table of inflected forms of auxiliaries (p. xxxviii). [Wv3] is a code for verbs in which the schwa is not pronounced when -lng or -er is added. [Wv4] is used for verbs that are often used In the -lng form as adjectives. [Wv5] Is used for verbs that are often used in the -ed form as adjectives. [Wv6] is a code for stative verbs. If a verb can occur with a particular preposition (e.g. on) or adverb (e.g. UP), this preposition or adverb is put in brackets immediately after the grammatical code within the square brackets. If a verb idiomatically occurs with a particular preposition and/or adverb, this combination gets a separate entry in LDOCE. For example: congratulate ... v. [T1 (on)] ... •uddle 2 ...v [T1 (UP)]...

29

get a c r o s s v prep ... keep out _v adv ... hold out on ^v adv prep ...

2.2.2.5

Pronouns

There are only two pronoun codes:

[Wpl] and [Wp2].

[Wpl] r e f e r s to a table on

page x x x v i i , which gives the i n f l e c t i o n s of the personal pronouns. [Wp2] r e f e r s to a table of r e l a t i v e pronouns on page x x x v i i i .

2.2.3

The numbers

Apart from l e t t e r s ,

the coding system a l s o uses numbers. "The numbers g i v e

information about the way the rest of a phrase or CLAUSE i s made up in r e l a t i o n to the word being described" (p.

xxxiii).

The numbers d e s c r i b e d i f f e r e n t r e a l i z a t i o n s of sentence e l e m e n t s and complementations in phrases. Each number r e f e r s c o n s i s t e n t l y

to the same

constituent class (except f o r the numbers in the [W] codes). The numbers must always be accompanied by a l e t t e r ,

whereas l e t t e r s can

occur without numbers. [0]

or

"nothing".

This

is

a "number"

for

intransitive

verbs

only,

indicating that they need not be f o l l o w e d by any elements, e.g. Vte paused [10]. [ 1 ] "means that a v e r b i s f o l l o w e d by a n o u n l i k e e x p r e s s i o n i n p o s i t i o n III

... or 2 nounlike e x p r e s s i o n s

in p o s i t i o n s

III

and IV ..." ( p .

xxxiii),

e.g. She became President [ L I ] , Give the boy a book [ D l ] , They considered him a fool

[ X I ] . The r e a l i z a t i o n s

of

the sentence e l e m e n t s

f u n c t i o n i n g as v e r b

complementation must be noun phrases, [ l a ] "means that the adverb i n a v adv marked

[Tla,

xxxiv) [lb]

...]

...

comes j u s t

after

the v e r b (even b e f o r e

a noun)"

(p.

indicates that the two elements of the v adv can be separated by a

noun or pronoun in

[Tib].

[ 2 ] means that the verb i s f o l l o w e d by an i n f i n i t i v e w i t h o u t Jto, e.g. can f l y

[12].

[3] means that the word i s f o l l o w e d by an i n f i n i t i v e with to, e.g. _I want to go [T3], John i s eager _to please

[F3].

30

[4] means that the verb is followed by an -lng form, e.g. _I enjoyed singing [T4]. [4a] means that adverbs or prepositions that are part of the v_ adv or _v prep cannot be separated frbm the verb, e.g. He^ left off (v adv) working [T4a], We looked at (v prep) him jumping [V4a]. [4b] means that the prepositions in the _v prep is separated from the verb, e.g. They inhibited her from doing it. [V4b]. [5] means that "a word is followed by that + a CLAUSE" (p. xxxiv). [5a] means that that can be left out, e.g. I know (that) he'll come [T5a]. [5b] means that a verb can be used with

or sometimes with not, e.g.

believe

so/not [T5b]. [5c] means that the that-clause can contain putative should as an alternative to a subjunctive, e.g. a desire that she (should) go/not go. [6] "means that a word is followed in position III by a wh-word, how, or as if, + either a CLAUSE ... or a phrase..." (p. xxxiv), e.g. H£ decided who should go [T6a], the reason why he came [C6a] and I'm not sure where to go [F6b]. The code [6a] refers to words that are followed by a wh-word + a clause. In [6b] the wh-clause or phrase functioning as verb complementation or postmodifier includes an infinitive with to. [7] means that the verb is followed by an adjective, either immediately or not, e.g. She became famous [L7], They considered him foolish [X7]. [8] means that the verb is followed by a past participle, e.g. He got trapped [L8] and to have

house built [V8].

[9] "means that a verb needs an ADVERBIAL, and that a noun or adjective needs a descriptive word or phrase" (p. xxxiv). The adverbial is obligatory, e.g. She lives in Amsterdam [L9]. The "descriptive word or phrase" either premodlfles or postmodifies the headword and can be of many types. For example: politically aware [B9] a film buff [C9] his posterity [GU9] powers £f memory [P9]

2.3

The Oxford Dictionary of Current Idiomatic English, Volume 1 (ODCIE 1)

ODCIE 1 contains verbal expressions made up of a verb + particle and/or preposition, e.g.: go about^

(verb + particle)

keep froa^

(verb + preposition)

31

boil down to

(verb + particle + preposition)

In most cases, the particle and/or preposition can be regarded as an integral part of the combination. The dictionary also records expressions which are not idiomatic, but in which the chances of co-occurrence of particle and/or preposition with the verb are very high. The optional part is enclosed by parentheses: consent (to) coop up (in) convert (from) (to) A great number of entries also include other parts of speech as fixed elements. They are mainly nouns (cf.the index of nouns used in headphrases on pages 371393), but adjectives, adverbs and present participles also occur: lose sight of1

(noun)

leave for dead

(adjective)

live over again

(adverb)

be beyond caring

(present participle)

These parts of speech can also be extended to longer phrases as, for example, the NP in go through (the) proper channels. As the title of the dictionary suggests, one would expect 0DC1E 1 to contain idioms only. However, it is not easy to distinguish between idioms and non-idioms. The compilers of the dictionary describe three semantic criteria to establish idiomaticity, starting from

the assumption that an idiom functions

as a unit of meaning: 1) One-word-substitution test. The idiom can be replaced by one word without change of meaning: Bill

took off Winston Churchill to perfection, mimicked imitated

2) Particle deletion test. The impossibility of leaving out the particle or preposition without changing the meaning: Bill took Winston Churchill to perfection. 3) Verb replacement test. The verb cannot be replaced by a related verb without a change of meaning: Bill took off Winston Churchill to perfection. * grabbed *snatched However, these criteria can only be fully applied to "true" idioms. Semi-idioms do not fulfil all three criteria, e.g. muck in Buck up can be replaced by mess,

32

and the particle

could be deleted without a real change of meaning.

Nevertheless it is included in ODCIE 1 on the basis of the following test: The weather really mucked up our weekend. spoiled ruined The compilers conclude that "the boundary between highly idiomatic items and the rest is not sharply drawn but hazy and imprecise. We shall do better to think in terms of a scale of idiomaticity, ..." (p. x). The policy adopted in this dictionary has been to include a great many semi-idioms. Five criteria for the Inclusion of semi-idioms are mentioned on page xi: 1) The preposltlon(s) or particle(s) is (are) obligatory, e.g. seize on/upon. 2) Semi-idiomatic expressions with one of the major intransitive verbs like come and go are also included when there is a close relationship with a corresponding transitive semi-idiom. Compare coae down and bring down. Homonyms of Idioms are Included because they are related in sense to idioms. For example, put aside^ (a book, one's knitting) is recorded because it is related 9 1 to pat aside (money, cash) and to put aside (a grievance, one's differences) . 3) Predictable co-occurrences of verb + particle

and/or preposition, even

though the particle/preposition is not obligatory, are also included, e.g. abstain (from), develop (froa), develop (fro«) (into). 4)

Combinations of verbs of motion + particle and/or preposition are included

when they have a specialized meaning, e.g. walk, on (»'have a small part, appear briefly, in a play'). 5)

Verbs such as puff, steam and zoom are also Included when they can be

combined with a particle of direction and thus be interpreted as verbs of motion (e.g. steam across • 'move across under steam power' etc.). The particle cannot be removed without changing the sense of the verb. Compare: The train steamed out noisily. The train steamed noisily. It is partly because of semi-idioms of this kind that ODCIE 1 does not use the conventional three-way division into: transitive phrasal verbs (e.g. make up (one's face)), intransitive phrasal verbs (e.g. (of a witness) come forward)) and prepositional verbs (e.g. run Into (difficulties)). Another reason for not adopting this division is that it does not account for complex Idioms such as put ideas into one's head, which consist of a direct object and a prepositional complement as integral parts of the idiom. The system of categories adopted in ODCIE 1 could have been extended to

33

provide for both the semi-idioms and the complex idioms. However, the compilers felt that this would have led to "a system that was both cumbersome to operate and difficult to interpret" (p. xii). 0DC1E 1 is not only a work of reference, but its design should also "promote active language use ..." (p. xxlv). To enable the user to construct correct

sentences,

the dictionary

provides a grammatical coding system

supported by collocations, example sentences and cross-references. Some entries also contain notes, which provide semantic and grammatical details. For example: pull up*

... O a

pull-up is a caf^ where drivers (usu of

commercial

vehicles) can stop for refreshments, not look at ... • usu after will/would. aake out*' ... • usu in direct and indirect questions after how. The square brackets that are used in the coding system contain two types of grammatical information. First, there is a letter/number combination, and second, there is abbreviated information on possible transformations. The letter/number system refers to six active sentence patterns, which are represented schematically as follows (cf. p. xxix): intransitive

transitive

Particle

[AL]

[BL]

Preposition

[A2]

[B2]

Particle + Preposition

[A3]

[B3]

The following examples are used to illustrate the patterns: [Al] The electricity supply vent off. [A2] We were banking on a change of heart. [A3] The committee fell back on an earlier plan. [Bl] Fred tipped the police off. [B2] Peter foists all his problems on his unfortunate friends. [B3] YOU can put the shortage down to bad planning. According to ODCXE 1 (page viii), the difference between a particle and a

34

preposition in the patterns [ B l ] and [A2] depends on two c r i t e r i a : a) The order of the d i r e c t o b j e c t NP and the p a r t i c l e can be reversed, which i s impossible in the case of verb + preposition + NP: Pop music turns on many young people. Pop music turns many young people on. The machine turns on £ central

pivot.

*The machine turns a central pivot on. b) Pronominal objects must come before the p a r t i c l e but a f t e r the preposition: Pop music turns them on. The machine turns on i t . I d i o m a t i c e x p r e s s i o n s i n ODCIE 1 are t r a n s i t i v e i f

their constituents

are

[ B l i ] and [ B l i i i ]

(see

separated by a d i r e c t object (O), with the exception of below). ODCIE 1 gives the f o l l o w i n g d e f i n i t i o n of

transitivity:

"according

to

whether or not they c o n t a i n a D i r e c t O b j e c t " ( p . x x v i i i ) . A change of h e a r t i n [A2], an e a r l i e r plan in [A3], his unfortunate f r i e n d s in [B2] and bad planning in [B3] are c o n s i d e r e d , not as d i r e c t o b j e c t s , but as p r e p o s i t i o n a l

objects

(o). They are complements of the prepositions. [ B l ] i s subdivided i n t o t h r e e c a t e g o r i e s :

[ B l i ] "a noun or noun phrase

O b j e c t can be placed on e i t h e r s i d e of the p a r t i c l e " , [ B l i i ] "such an O b j e c t must appear between the (main) verb and the

p a r t i c l e " , and [ B l i i i ] " i t must be

placed a f t e r the p a r t i c l e " (p. x l i v ) . In a l l these cases a pronominal

object

precedes the p a r t i c l e . Grammatical information cannot only be obtained from the codes, but also from the c o l l o c a t i o n s . as r e a l i z a t i o n s transitive,

sentence e l e m e n t s or p a r t s of

entry

contains

a number

of

phrases. I f

collocating

an idiom

is

direct

objects

introduced by a c a p i t a l 0 i n b o l d t y p e . I f the s e t of c o l l o c a t e s i s

severely

restricted,

the

Collocates are words or expressions that are t y p i c a l l y used of

ODCIE 1 gives the danger sign A .

The f o u r main g r a m m a t i c a l

functions indicated

i n the e n t r i e s a r e : S

subject, 0: d i r e c t o b j e c t , o: prepositional o b j e c t and A: adverbial. Other l e s s frequently occurring c o l l o c a t i o n a l l i s t s in ODCIE 1 are: Inf:

infinitive

(either

standing

alone

or

introducing

a

non-finite

c l a u s e ) , as in b r i n g up^. adj: a d j e c t i v e (usu modifying the noun in an expression of which the noun I s a f i x e d p a r t ) , as in play a p a r t / r o l e ( i n ) .

35

n: adverbial modifier of a particle, as in be o u t 1 2 (e.g. ... was some way out.) .

While

the

codes describe

the g r a m m a t i c a l

the entry,

use of

the sets of

collocates give possible realizations of sentence elements or parts of phrases. The example sentences illustrate the use of the entries in authentic or made~up pieces of text. For example: deceive into [B2 pass] ... S: enemy, rival; salesman, agent, confidence trickster. 0: firm, purchaser, shopper, o: supposing, believing, thinking; buying,

selling.Q The Prussians deceived Napoleon

withdrawn away from the British. Q tha t wa s

into thinking they had

We_ were deceived into buying a_ house

scheduled for demolition,

- i n g form of a v as object of

into.

The following substitution table could be made for deceive into:

S

0

o

The en emy

our firm

supposing...

Our rivals

the purchaser

believing...

The salesman

DECEIVE

the shopper

INTO

thinking...

The agent

me

buying...

The confidence trickster

the students

selling...

With a proper knowledge of English tenses, etc. and under the guidance of the collocates,

the

example

sentences

(the

passive

transformation

is

also

illustrated) and the extra information supplied in the footnote, the dictionary user should be able to form sentences like: The salesman deceived me into buying the vacuum cleaner. Our rivals deceived our firm into thinking that they wanted £ merger.

Transformations in ODCIE 1 are indicated by means of abbreviations after,

or

before,

the

letter/number

code

within

the

square

immediately brackets.

36

"Transformation ... is a matter of structural change, and the changes may be of various kinds. Sentences which are transformationally related are often closely related in meaning also" (p. xxx and xxxi). A transformation of a particular type is possible only if this is indicated in the grammatical code.

Five

transforms are made mention of: a) [pass]: passivization of active sentence patterns. In this transform the 0 or o of the active sentence becomes the S of the passive. Rules for forming passive sentences are given (and illustrated) in the Introduction. [A2] and [B2] have a subcategory [pass(o)]. In the case of [A2] this means that the second prepositional object becomes the S in the passive transform, because the first is too closely connected with the idiom: got to grips vith the problem. The problem was got to grips with. In [B2 pass(o)] it is again not the 0 (direct object) which becomes S in the passive sentence, but the o, because of the close link between 0 and the rest of the

idiom:

The governor made an exaaple o£ these prisoners. These prisoners were made an example of. In most cases the transform is mentioned after the letter/number code, e.g. [A2 pass], indicating that the passive is derived from the active

sentence.

However, there are codes in which the transform is mentioned first because there is no corresponding active form. The pattern from which that passive form may have been derived is given in brackets: infatuated with [pass (B2)] cut out for

[pass (B3)]

In the transitive patterns the 0 of the active sentence becomes an extraposed S in the passive if it is a clause. Anticipatory 'it' fills the subject position then: We worked out that we should need 100 pounds. [Bli] It was worked out that we should need 100 pounds. You should bear In aind that he wasn't present. [B2] It should be borne in aind that he wasn't present. b) [nom]: "... in general terms, a nominalized form is a noun derived from a verb + particle combination ... or from a combination of verb + preposition ..." (p. xxxi). Nominalized forms can be divided into three groups: 1) The verb precedes the particle or preposition. The noun may be hyphenated or

37

not, e.g. crack-up, hold-up and category

have

[nom] added

to

look-over. Idiomatic their

expressions

code. The nominalized

form

in

this

is often

illustrated in the example sentence. 2) The verb follows the particle. In this case the verb is either a bare infinitive

or an -ing

form.

The

noun

is not

hyphenated,

e.g.

downpour,

outbreak, outpouring and upbringing. In the case of outpouring there is only an illustration and in the case of upbringing the nominalized form is mentioned in a note, but not illustrated. 3) ODCIE 1 includes a group of nouns that have no corresponding verb + particle expression. They have the same structure as the nouns mentioned under 1) above, but

their

codes

abbreviation

differ.

The

[nom] first, and

codes

of

nouns

in

this

then the letter/number

category

give

combination,

the

which

Indicates which pattern the nominalized form is related to: come-down

[nom (Al)]

going-over

[nom (A2)]

hook-up

[nom (Bl)]

All nominalized forms are listed in the Index on pages 394 - 396. The functions that nominalized syntactic

forms may fulfil are not coded,

uses of particular

nouns are displayed

but the

characteristic

in the illustrations (cf. p.

xxxiil). In the notes to the tables in the introductory pages there is some information on the functions of the nouns and the accompanying verbs.

c) [adj]: This abbreviation is enclosed in the square brackets if the idiomatic expression can be "adjectivized" and used attributively in an NP: the quickly glanced-through article ...[A2] his very much looked-forward-to retirement ... [A3] The "adjective" always occurs in the past participle form. In the notes to the tables for [A2 adj], [A3 adj] and [Bli adj] (p. xxxix and following), reference is made to the grammatical link between the basic active sentence, the passive sentence and the participial adjective phrase. For example: [A2] He glanced through the article quickly. [A2 pass] The article was quickly glanced through. [A2 adj] The quickly glanced-through article.

d) [emph]: This transformation places the prepositional phrase in Initial position, e.g.: For his family he has provided well [A2 emph] In the case of [Al emph], where there is no preposition, the particle is placed

38

at the beginning of the transformed sentence: Down caae the prices

[A1 emph].

This transform is seldom illustrated, so that the dictionary user has only the tables and notes in the introduction to rely on. The notes give details about the order of the elements in the transformed sentences and the use of the transform. Cf. [A1 emph] notes (a), (b) and (c), pages xxxvi and xxxvii, [A2 emph] note (c), page xl.

e) [rel]: This abbreviation is meant to indicate that basic patterns like We were hanking on on

change of heart can be changed into NPs like The change of

heart on which we were banking. Other examples of [rel] are: The article through which he glancedThe sandbank on which the coaster went aground. Remarks of which I_ have taken careful note. Since

this

relative

transform

is

rather

formal

alternatives with the preposition placed after the verb

in style,

colloquial

are given in the notes

to tables [A2 rel], [A3 rel], [B2 rel] and [B3 rel]. For example: The change of heart which we were banking on.

Finally, it is worth pointing out that ODCIE 1 contains two types of crossreference. One type is marked (qv) and is used for reference to synonyms. The second type of reference is expressed by an arrow. The purpose of this crossreference is "to indicate special grammatical connections between entries, or to point out important relationships of meaning (other, that is, than sameness of meaning)" (p. lxx). For example; decide against 2 ... < = 0 decide on...

decide on

decide against 2

put on^® ...^^^be on^, bring on^, come one® The cross-references often connect transitive and intransitive equivalents, e.g.

coae

back^

[Al] ... c = £ b r i n g

back 2 .

The

expression

bring

back 2

is

transitive and has the code [Bli pass]. Cross-references are also useful for the active use of the entries, in that they give additional collocations to the ones already,mentioned. For example: coae a t 2 ...o: facts; truth; causes. ... i = ^ g e t at'' •3 get at

...o: truth, facts; real wishes, intentions;

trouble; the cause of the disturbances ... I

^come

at 2

the root of

the

39

2.4

The Oxford Dictionary of Current Idiomatic English, Volume 2 (ODCIE 2)

This dictionary

of phrase,

clause and sentence

idioms is also designed

primarily for the advanced foreign student. The problem of the inclusion of (semi-)idioms arises here too. Again we find that there is no clear-cut division between idioms and non-idioms, and that these are no more than endpoints on a scale of idiomaticity. ODCIE 2 also includes many semi-idioms. On the other hand, the compilers have excluded open collocations. The dictionary, for example, does not include an entry for on one's return, since on in the sense of 'at the time of' can be combined with a number of other nouns: arrival, departure, demise, death, dismissal, etc. In doubtful cases the compilers have tended "to be accommodating, so that some combinations in which both or all the constituents are used in a straightforward sense have been included." (p.xiii). Semi-idioms such as the cold war are not considered as nouns but as NPs, because they do not constitute the same unity of meaning as full or pure idioms. As a result, all idiomatic expressions of this type are classified as NPs. The same principle governs the grammatical categorization of clause and sentence patterns. A pure idiom such as kick the backet, which as a whole could be regarded as an intransitive verb (cf. die, its one-word substitute), is coded [V + 0], i.e. transitive verb + direct object, parallel to semi-idioms of the same

syntactic

structure,

e.g. jog one's/sb's memory

[V + 0 pass].

ODCIE 2, like ODCIE 1, contains a good deal of syntactic information. The student is helped to "determine what kind of pattern an idiom corresponds to, how it can be broken down into elements smaller than itself, and what other elements precede, follow or interrupt it" (p. xxvii). The grammatical coding system is divided into four main types: 1) Clauses containing objects, complements and/or adjuncts e.g. catch/take sb unawares [V + 0 + Comp pass]. 2) Possessive clauses, e.g. have etc occasion to do sth [possess]. 3) Phrases with a noun, adjective, preposition or adverb as the headword, e.g. a howling success [Comp (NP)]. 4) Phrases in which a particular class of word is repeated, e.g. hither and thither [adv + adv non-rev].

40

Not a l l e n t r i e s in ODCIE 2 are coded. Most of the sentence i d i o m s ( e . g . tbe bird has flown) do not have codes, because they "allow very l i t t l e

lexical

v a r i a t i o n " and " t h e s t r u c t u r e of such i d i o m s i s n o r m a l l y f i x e d : t h e r e i s no need f o r a mechanism to explain t r a n s f o r m a t i o n a l p o s s i b i l i t i e s " ( p .

xxviii).

Only "the commonest clause and phrase patterns and/or functions are i d e n t i f i e d by means of a grammatical code". Grammatical v a r i e t y in expressions with l e s s common patterns i s i l l u s t r a t e d in the example sentences. 0DC1G 2 distinguishes f i v e clause patterns, as f o l l o w s : [V + Comp]

verb + complement pattern

[V + 0]

verb + d i r e c t object pattern

[V + 0 + Comp]

verb + d i r e c t object + complement pattern

[V + 10 + 0]

verb + i n d i r e c t o b j e c t + d i r e c t o b j e c t pattern

[V + 0 + A]

verb + d i r e c t object + adjunct pattern

The [V + Comp] p a t t e r n ( e . g . be thick w i t h sth and be three sheets In the wind) c o n t a i n s as r e a l i z a t i o n s of the complement an AdjP and an NP, but t h i s information i s not e s s e n t i a l ,

since they are f i x e d elements in the

idiomatic

c l a u s e . The l i n k between t h i s p a t t e r n and [V + 0 + Comp] i s made c l e a r by the c o l l o c a t i o n s , e.g. go etc crackers etc [V + Comp] ... V: g o . ^ b e ; send sb

drive

sb,

The f o l l o w i n g sentence can be constructed: H£ drove ne crackers.

Expressions of the [V + 0] type e.g. f l o g a dead horse and f l u t t e r the dovecotes

may or

may not

be p a s s i v i z e d .

If

they

can be p a s s i v i z e d

the

abbreviation pass i s added to the code, e.g f l u t t e r the dovecotes [V + 0 pass]. Similarly,

in the case of

the

[V + 0 + Comp] p a t t e r n ,

right/wrong and make matters/things worse, i f

e.g.

get

the p a s s i v e t r a n s f o r m a t i o n

sth is

possible, the abbreviation pass i s added to the code. The r e a l i z a t i o n of the 0 i s e i t h e r a f i x e d part of the i d i o m or i t i s n o t . I f

the r e a l i z a t i o n of

the

d i r e c t o b j e c t i s v a r i a b l e and r e l a t i v e l y l o n g , i t may be t r a n s p o s e d , e . g . The Chancellor made plain the p r i n c i p l e s on which the Budget had been designed.

In

some entries with this pattern the d i r e c t object i s optional. In that case i t i s p l a c e d in b r a c k e t s i n the headphrase, e.g. make etc ( s b / s t h ) a laughing stock [V + 0 + Comp p a s s ] . Many e x p r e s s i o n s of the [V + 10 + 0 ] p a t t e r n ( e . g . s e l l sb a pup and s e t (sb) a (good etc) example) have an optional i n d i r e c t o b j e c t ,

which i s

enclosed

in parentheses in the headphrase. P a s s i v i z e d sentences of t h i s type have the i n d i r e c t o b j e c t of the a c t i v e sentence as S ( e x c e p t when the o t h e r form i s i l l u s t r a t e d ) . There may be an a l t e r n a t i v e

form (transform) of

this pattern in

which the 10 becomes a p r e p o s i t i o n a l complement to to^ or towards, e . g . Mary sets an example to us a l l . This a l t e r n a t i v e form, however, i s not coded.

41

In the [V + 0 + A] pattern (e.g. take sb/sth as he/it coses and Cake stb hard/lightly [V + 0 + A pass] pass Is added if the Idiomatic expression can be passivized. If the 0 is long, it can be transposed, e.g. A doctor has to put first his poorer patients and their families. The five clause patterns contain sentence elements, most of which are obligatory (cf. the optional 10 in some sentences of the [V + 10 + 0] pattern). Generally the verbs in these patterns are obligatory and unique, i.e. they cannot be substituted by other verbs. Yet even if the verb can be replaced, either by an open or a restricted set of verbs, it remains an Integral part of the idiomatic expression (e.g. hear/see the last of sb/sth and scare etc (sb) stiff).

Possessive clause patterns, marked

[possess] (e.g. The leaders got a taste of

their own medicine) contain verbs expressing possession, get, glve and have being the commonest. The structure of a possessive clause idiom can be [V + 0], e.g. get etc a bad oaae, or [V + 10 + 0], e.g. give sb/sth the slip. In the first structure get is the most frequently occurring verb, in the second give. The two structures are said to be connected because the S of the first can become the 10 of the second pattern. Compare: The leaders got a taste of their owl Medicine. The voters gave the leaders a taste of their own medicine. The codes of phrases carry two types of information, the function(s) the phrase can have (sentence element) and the form of the phrase (NP, AdjP, PrepP or jComp/A (NP)]. Only the commonest functions AdvP): (as) a matter of principle are identified by means of a grammatical code (see p. xxviii) Whereas in the clause patterns the verb is always part of the code, in phrase patterns it is never coded. The verb is excluded from the phrase code because it is not part of the idiom. However, the related clause pattern can be deduced from Comp in the code, as well as from the collocates. Consider: easy on the eye [Comp (AdjP)] ... V: !

be, find sb. ...

The collocations play an important part in ODCIE 2. The information given by them is twofold. Because of the way they are presented, the user knows where to insert the given realizations of phrase constituents. As far as the grammatical functions are concerned the following abbreviations are used: S (subject), V (main

verb/verb

phrase), ^0

(complement) and A (adjunct).

(indirect

object), 0 (direct

object),

Comp

42

Collocations are Included In entries if realizations of elements in clause patterns can be replaced and if the user needs guidance on these substitutions. Sometimes

realizations

of optional clause elements are added

(e.g. the

adverbials iji the soil and iji their minds). get etc a standing ovation [possess] ... S^: speaker, leader. W_: get, ^

have, receive, give sb. ...

take/strike root [V + 0] ... ¡3: seeds, the idea, his words. A: in the soil, in their minds. ... Especially in the case of phrases, which constitute only one element in a clause, the collocations can be of great help for the active use of the entries, since they indicate the possible clause patterns (e.g. easy on the eye). The four phrase types in ODCIE 2 are: [NP]

noun phrase

e.g. •oney veil spent,

[AdJP]

adjective phrase

e.g. soft/wrong in the head,

[PrepP]

prepositional phrase

e.g. in sb's honour,

[AdvP]

adverbial phrase

e.g. far afield.

The constituents of the phrases as listed on p. xxxv + xxxvi are:

NP determiner a

adjective

noun

blue

fit

a

prep phrase/clause

feather in one's cap

Ad jP adv modifier

adjective

fast

asleep

not

fit

prep phrase/inflnitive/clause to wash his feet

43

PrepP preposition

object of preposition determiner

adjective

at

one's

own

noun discretion

in

the

ordinary

way

AdvP adv modifier

adverb

all too

briefly

prep phrase/clause

as

far

as his legs could carry him

It is important to know the order of the phrase constituents, so that one knows where to insert or replace collocates in the headphrase. The collocates in the entries are always preceded by abbreviations, indicating which parts of speech or sentence elements we are dealing with: det:

determiner

adj:

adjective

n:

noun

cl:

(non-)finite clause

Inf:

infinitive or infinitive phrase

pron:

pronoun (usu replacing a whole NP)

prep:

preposition

o:

object of preposition

adv mod:

adverbial (pre-)modifier

adv:

adverb

The following entry serves as an example: a slip of a boy etc [Comp (NP)] ... det: just a, no more than a. adj: mere, tiny, slender, n: boy, ^ ^ girl, thing, creature. ...

A table of this NP would be: NP determiner

adjective

noun

PrepP prep.

object of prep.

44

a

mere

just £

tiny

no more than

slender

...

determiner

noun

a

girl

boy slip

of

thing

...

creature

The abbreviations of phrase constituents can also occur in entries of clause patterns (i) and uncoded idiomatic expressions (ii): (I) carry conviction [V + 0] ... det: some, more, less, adj: total, absolute. ... (II) go their different etc ways ... pron: their, our, your, adj: different, ^

separate; several, various. ... (2)

0DC1G 2 contains a special group of phrases with repeated elements: [n + n non-rev]

e.g. double or quits

[adj + adj non-rev]

e.g. fast and furious

[v + v non-rev]

e.g. hire and fire

[det + det non-rev]

e.g. each and every

[adv + adv non-rev]

e.g. first and foremost

non-rev means that the constituents cannot be reversed; rev is included if the nouns, etc can change order. The functions of these phrases are not usually added. Exceptions are: day and night

[A (n + n non-rev)]

bright and early

[A (adj + adj non-rev)]

Some of the other functions are illustrated in the example sentences. In the notes on the functions of [PrepP], page xxxvi, Disjuncts and Conjuncts are mentioned. This function is not only realized by Prepositional Phrases, although they are the commonest. Other possibilities include: either way [Disj (NP)] and far from it [Disj (AdjP)]. Many codes of Disjuncts and Conjuncts in ODCIE 2, however, do not include information on the realization in terms of phrases, e.g. more's the pity [Disj] and even as [Conj]. Only the passive transformation is expressed in the codes. The possibility of passivlzatlon of active sentences, a syntactic change, is indicated by means of the abbreviation pass, e.g. dig a pit (for sb) [V + 0 pass]. A second transformation is not explicitly pointed out, but is inherent in

45

the code [possess]. In this semantic transformation the S of the [V + 0] pattern becomes the indirect object of the [V + 10 + 0] structure, i.e. if the verb changes. Compare: The economy gets a shot in the arm. The Chancellor gives the economy a shot in the arm. A third, non-coded

transformation is sometimes possible in [V + 10 + 0]

patterns: the 10 becomes an object after the prepositions U> or towards. Compare: He bears me no malice. He bears no malice towards me. The last transformation, which again is not coded, is the postponement of the direct object in the patterns [V + 0 + Comp] and [V + 0 + A]. This postponement takes place, or rather can take place, if the 0 is relatively long. Grammatical information is also supplied in the form of footnotes. The footnotes give details about structural variants, tenses, number, etc. Examples are: fight like a tiger ... O also pi fight like tigers. give sth all one's got ... O esp with present simple tense, will, be going to. here and now . . . O n form the here and now. (it is) high/about time (that) ...Q that-cl contains a v in the past tense; ...

2.5

The Longman Dictionary of Phrasal Verbs (LDOPV)

LD0PV (1983)^), compiled by Rosemary Courtney, is a "learner's guide to twoword verbs". These two-word verbs, or phrasal verbs, are said to be idiomatic combinations of a verb and adverb, or a verb and preposition (or a verb with both an adverb and preposition), these being idioms which are known to "cause difficulties for students of English because of their meaning and grammar. However,

the dictionary

does not

only

contain entries for

Idiomatic

combinations such as take in (=» deceive), set about (= start) and pat up with (= bear). It also includes verbs which require a fixed preposition (e.g. depend on and despair of), longer verbal idioms such as let the cat out of the bag

46

(s.v. let out), and certain verb + adjective and verb + noun combinations such as lie low and chuck It. Moreover, if a combination of verb + adverb or preposition has a normal as well as an idiomatic meaning, the normal meaning is also included/^) Verbal combinations of which the meanings are clear from the parts (e.g. sit down) are normally excluded from IDOPV.

The coding system in LDOPV is generally the same as that used in LDOCE for verbs (see section 2.2.2.4). The grammatical codes consist of a letter with mnemonic value (D, I, L, T, V, or X) and a number (0 to 9). However, there is one difference. LDOCE and LDOPV have similar codes for phrasal verbs, but in the case of non-idiomatic combinations LDOCE only codes the verb, whereas LDOPV often includes the adverb and/or preposition in the code. A typical example of an LDOPV entry is the following: hold back v adv 1. to keep (something or someone) in place;

prevent (something or

someone) from coming forward: [X9 + BACK] The men built banks of earth to hold back the rising flood waters. ... *2. to control (something such as a feeling): [Tl] Jlm was able to hold back his anger and avoid a fight. ... *3. to prevent the development of (usu. someone): [Tib] You show promise as ji musician but your lack of practice is holding you back. ... *4. ... etc. Sub-entry 1, the non-idiomatic sense of the combination hold back, is given the code [X9], which means that hold is a transitive verb with a noun as direct object and a complement (back in this case) which refers to the object. The complement has the grammatical function of adverbial; back is repeated in capitals after the letter/number code and preceded by a plus sign.

The

remaining sub-entries are all idiomatic combinations (phrasal verbs marked with an asterisk) which are treated as single verbs and coded [T]: [Tl], for example, means the verb-adverb combination is transitive and takes a noun as direct object (his anger in this case); [Tib], in sub-entry 3, means that hold back is a transitive verb with a noun direct object and that back is always separated from the verb/6) By contrast, LDOCE labels all the above-mentioned sub-entries, including sense 1, as [Tl], that is, as a transitive verb with a noun direct object.

47

LDOPV uses 25 different explained

letter/number

codes,

which are listed

and

in the introduction. The codes seem to apply primarily to the

idiomatic combinations, which is probably why only examples of idioms have been given in the introduction. Codes for non-idiomatic entries or sub-entries, as we have seen, are slightly different, but on the whole they involve the same letter/number combinations. Certain codes (e.g. [D5] and [D6]) do not apply to non-idioms, others (such as [L9] and [X9]) occur frequently with non-idioms. Below we list the grammar codes used in this dictionary. Every code is supplied with a short description and one example here (the Introduction of the dictionary sometimes gives two or three examples). In each case the capital letter shows the type of verb (transitive, intransitive, linking, etc.), the number shows what kind of category follows the verb (noun (phrase), adjective (phrase), to-lnflnitlve, etc.), while the small letters a and b show the position of the adverb or preposition. [Dl]

ditransitive verb with a direct object (NP) and an indirect object (NP). Can _I help you to some more potatoes?

[D5] ditransitive verb with a direct object (NP) and an indirect object (that-clause). The law sets it down that speed limits must be obeyed. [D6]

ditransitive verb with a direct object (NP) and an indirect object (whclause). No one is going to dictate to me where to live!

[10]

intransitive verb. Don't let me interrupt you, just carry on.

[13]

intransitive verb followed by a to-lnflnitlve. I shall really have to push myself to finish this book before the contract date.

[14]

intransitive verb followed by an -lng form. Having missed the last bus, we had to set off walking.

[15]

intransitive verb followed by a that-clause. Watch out that you don't catch cold, going out in the pouring rain.

[16]

intransitive verb followed by a wh-clause. Mind out where you're walking, the ground is muddy.

[LI ]

linking verb followed by a complement (NP) that refers to the subject. In cricket, a no ball counts as ai run.

[L4]

linking verb followed by a complement (-lng form) that refers

48

to the subject. never dreamed that [L7]

would wind up owning such

l o t of property.

linking verb f o l l o w e d by a complement (AdjP) that r e f e r s to the subject. A f t e r making j» fortune In business, he f i n i s h e d up poor at his death.

[L9]

linking verb f o l l o w e d by a complement (PP, AdvP) that r e f e r s to the subject. Setting out In buslness I s no easy job.

[Tl]

t r a n s i t i v e verb;

the d i r e c t o b j e c t i s r e a l i z e d by an NP.

The children are gathered in the garden to l e t o f f the fireworks. [ T l a ] as [ T l ] , but the adverb of the phrasal v e r b must precede the d i r e c t object. At the end of the row, cast o f f seven s t i t c h e s . [Tib] as [ T l ] , but the d i r e c t o b j e c t i s p l a c e d between the two ( o r three) component parts of the phrasal verb. Only his courage carried Jim through. [T4]

t r a n s i t i v e verb; the d i r e c t object i s r e a l i z e d by an - i n g form. The doctor told Jim to g i v e up smoking and

gave smoking up years ago.

[T4a] as [ T 4 ] , but the adverb cannot be separated from the verb. He didn't leave o f f [T5]

transitive

verb;

talking f o r an hour. the d i r e c t

object

i s r e a l i z e d by a t h a t -

clause. would l i k e to point out that i t i s g e t t i n g very l a t e . [T6]

t r a n s i t i v e verb; the d i r e c t o b j e c t i s r e a l i z e d by a wh-clause. Can you think of why she should do such a thing?

[V3]

t r a n s i t i v e verb with a d i r e c t o b j e c t (NP) and a t o - l n f l n l t i v e . His f a t h e r ' s example drove Jim on to enter the competition.

[V4a] t r a n s i t i v e verb with a d i r e c t o b j e c t (NP) and an - i n g - f o r m . The preposition cannot be separated from the verb. Don't reckon on the weather being f i n e f o r your garden party. [V4b] as

[V4a] ,

but

here

the

preposition

or

adverb

is

always

separated from the verb. The p o l i c e charged the prisoner with s t e a l i n g the j e w e l s . [XI]

transitive

verb

with

a direct

object

(NP)

and an

object

object

(NP)

and an

object

complement (NP). They regard him as their enemy. [X7]

transitive

verb

complement ( A d j P ) .

with

a direct

49

Since she did not answer, [X9]

set her down as fearful and nervous.

transitive verb with a direct object (NP) and an adjunct (AdvP, PP) . Your uncle should be able to fit you up with a_ job in his firm.

The g r a m m a r codes mentioned in the Introduction are s u m m e d up on the inside covers at the back and at the front of the dictionary. Grammatical codes in LDOPV are always placed before the example sentences in the entries. Every code has at least one example. Only one pattern is always coded and exemplified at the same time. On the other hand, no attempt seems to have been made to order the g r a m m a r codes systematically. For e x a m p l e , in a sub-entry of derive from code [Tl] precedes

code [Dl], while it follows

[Dl]

in a sub-entry of form loco. Compare: derive from

v. prep

*2 ... [Tl (no pass.)] The word 'warmth' derives from 'warm'. [Dl (usu. pass.)] The word 'grammar' is derived

form into

*3

_v

from Greek.

prep

... [Dl] Form

the men

into

lines three deep.

[Tl (no pass.)] The

soldiers formed into battle order. As

is

shown

above,

the

(sub-)entries

also

contain

information

on

the

possibility of passivization. If phrasal verbs and verbal combinations can normally undergo passivization, nothing is added to the letter/number code. It is only the constraints (on passive or active voice) that are made mention of. The following abbreviations can be found between the square brackets: [...(pass, rare)], e.g. *act for [...(no pass.)], e.g. *itcb Cor [...(often pass.)], e.g. blow on/upoo *2 [...(usu. pass.)], e.g. invest with *3 The abbreviation

[...(only pass.)] is not used, since combinations of verb +

adverb and/or preposition which occur in the passive only, are printed in bold type as fixed expressions, inflate with

e.g.:

prep

2 be inflated with ...

50

Other grammatical information that can be found between the square brackets includes: [...(simple tenses)], e.g. call for *3 [...(often simple tenses)], e.g. call in 1 [...(usu. simple tenses)], e.g. calculate on/upon *2 [...(simple tenses, usu. present)], e.g. *bleed for 2 [...(continuous tenses)], e.g. brew up *3 [...(often continuous tenses)], e.g. bugger about/around with [...(usu. perfect tenses)], e.g. go for 15 [...(often inf.)], e.g. cut short 5 [...(imper.)], e.g. cut off *9 [...(usu. imper.)], e.g. *button up 2 [...(often neg.)], e.g. blink at [...(usu. neg.)], e.g. *breathe of The prepositions of optional prepositional phrases are also frequently given in the code, and placed between round brackets. Consider: go forward v adv

*4 ... [ ICl(wlth)) ] The council gave us permission to go forward with our building plans. Bracketed grammatical information is exemplified where possible. No examples, however, have been provided for the nominallzed and adjectivized forms printed in bold type at the end of entries or sub-entries. Contexts for these forms must be deduced from the definition and examples of the related verb. For example: cut off _v adv *3

to

prevent

or

stop

(some thing):[ T1 ] "_I w i l l

give

them

an

everlasting name, that shall not be cut off." (The Bible) "He that cuts off 20 years of life, cuts off so m a n y years of fearing death." (Shakespeare, Julius Caesar). - cut off n, adj

51

NOTES TO CHAPTER II

One might very well question this first editorial statement in ALD. If the learner "has mastered the rules of English grammar and pronunciation," what exactly does he need to learn?

(2)

The title of the dictionary is especially misleading to students who are familiar with the "grammatical" definition of phrasal verbs as verbparticle/adverb combinations which display certain phonological and syntactic characteristics (cf. Quirk et al. (1972:815)).

(3)

We are quoting here from the introduction of LDOPV. As the pages of the introduction are not numbered, we are unable to provide page references.

(4)

The (sub-)entries regarded as Idiomatic are marked with an asterisk, e.g.:

carry out v adv 1. to lift (something or someone) and take it out: ... *2. to fulfil or perform (something): ... In such cases the dictionary always lists the normal, literal sense first, then the "phrasal" sense, and finally possible fixed expressions. The use of the asterisk for this purpose is somewhat unfortunate in the light of its different function in grammars and linguistic publications. We shall return to the problem of defining the notion of idiomaticity in section 3.5.1. (5)

Prepositions are also repeated after the letter/number code between the square brackets. However, they are printed in italics, not in capitals. For example: cut through v prep 1. to cut

(something) with

sharp

metal

through

(something):

[X9+through] They had to cut their way through the forest with axes. *2. to shorten (something); work faster with (something): [Tl] Let's cut through the usual committee stage and get the plan approved faster.

Consider the two sub-entries of fry up: fry up T/ adv 1. to dry or destroy (something) with heat: [X9 + UP] The unusually hot sun has fried up the crops. *2. to reheat (food) by cooking it in hot fat: [Tl] We can fry up some of these pieces of cold meat for our lunch. ... The first of the two senses of fry up is regarded as non-idiomatic in LDOFV, and the second as idiomatic. One wonders why. Both senses are clearly non-idiomatic, which may be the reason why neither of them is listed in LDOCE.

CHAPTER III

3.1

A CRITICAL REVIEW OF THE CODING SYSTEMS

ALP

A.S. Hornby was the first lexicographer to include grammar in a learner's dictionary. However, the compressed form in which his grammatical information is presented has had its effect on the usefulness of the codes. In this section an attempt will be made to deal with the inadequacies of the coding system as it is described in the Introduction and used in the dictionary.

3.1.1

The Introduction

Nouns are coded for countabllity with the letters [C], [U] or [C,U], as pointed out on page xxiv of the Introduction. Unfortunately, one of the examples of an uncountable

noun ([U]) given

there

is noise,

whereas

the entry

in

the

dictionary codes it [C,U]. In most cases the use of these codes is very clear. For example, in the entry for pain the codes are clearly differentiated and supported by example sentences. However, there is a problem with the use and non-use of [C]. Given the principle that "when no Information is given in a noun entry, it is an obvious countable noun" (p. xxiv), one would expect to find this [C] code only with entries that have other codes as well. If the noun is countable in all its senses, it should not be given the code [C]. For example, delight1 is coded as follows: ...n 1 [UJ...2 [C]

, while pack 1 has no code since it is countable

54

in a l l i t s senses. However, s e v e r a l nouns are coded [ C ] , even i f

they occur

only as countable nouns, e.g. decline^, d e c l i v i t y , decree, d e f e c t ^ , package, panegyric,

paradox and paragraph.

One reason f o r adding [C] to some noun e n t r i e s may have been that these nouns are less common than the non-coded countable nouns. But why then do the f o l l o w i n g u n f a m i l i a r nouns r e c e i v e no code: delphinium, pachyderm, paean, pailasse, palanquin, panacea and panjandrum? Degree has also been coded inconsistently. are coded i f

Normally a l l

senses of a noun

there i s at least one sense that i s not countable,

as in the case

of d e f i c i e n c y . In the e n t r y of degree o n l y senses 3 and 4 are coded [C,U] and [U],

respectively. A u s e f u l type of g r a m m a t i c a l i n f o r m a t i o n on nouns would have been an

i n d i c a t i o n of t h e i r complementation by p r e p o s i t i o n a l phrases. However, the p r e p o s i t i o n i s u s u a l l y not coded. In most cases the user has to r e l y on the example sentences and/or on references to a verb that may be

morphologically

r e l a t e d to a p a r t i c u l a r noun, e.g. declamation and deduction. In the case of deliverance from, the preposition i s given in bold type. Idioms l i k e take d e l i g h t in and on a par ( w i t h ) are a l s o p r i n t e d in bold type. Optional prepositions can only be found in example sentences or in verb entries. Nominal complements other than prepositional phrases are not indicated and can only be found in example sentences,

e.g.

decision...His decision to

retire

surprised a l l of us. Herbst (1984) makes the o b s e r v a t i o n that both ALD and LDOCE l a c k b a s i c and s y s t e m a t i c i n f o r m a t i o n as f a r as a d j e c t i v e s are concerned. His paper, which deals with a d j e c t i v e complementation within a valency approach,

distinguishes

between complements (which may be o b l i g a t o r y or o p t i o n a l ) and elements.

Herbst argues

that both ALD and LDOCE confuse optional

peripheral complements

and peripheral elements in particular. He i s r i g h t in demanding that

learners'

d i c t i o n a r i e s should aim at comprehensiveness, since foreigners who use them must be able to f i n d out whether constructions are grammatically correct. As an example he gives the f o l l o w i n g sentence:*Valency theory i s useful to dictionary w r i t i n g . ALD does not supply i n f o r m a t i o n on whether t h i s type of sentence i s c o r r e c t or i n c o r r e c t . No mention i s made in the s u b - e n t r y f o r u s e f u l of the p r e p o s i t i o n that i s normal here ( t o ,

for,

e t c . ) . Nor do the examples

given

provide the information that the user needs. Sometimes we f i n d i n f o r m a t i o n on the complementation in the example

55

sentences, but this is not done very consistently. For example: pleased... He looked A J

with himself. I'm very (much) with what he has done. It Is odd

that with does not appear in bold type and in brackets immediately after pleased, indicating that it is

an optional complement as, for example, the

preposition _tc> in accessible (to). For the use of adjectives it is important to know whether they can occur predicatively and/or attributively. ALD does not use codes to indicate this, but adds the sub-class labels pred adj and attrlb adj. Sometimes, however, the information Is formulated differently. For example: loath, loth... ad j (pred only)... •aln1...ad j (attrib

only;...)...

If in the case of predicative adjectives not all senses of an adjective can be used predicatively, the dictionary gives the appropriate label in the right place, cf. e.g. certain, conscious, faint and fond. A great number of attributive adjectives, however, are not characterized as such, e.g. sere, only, saae, sheer and utter. Neither is there consistent use of the abbreviation attrib if an adjective can be used attributively in only one or more of its senses. Examples of how it could be done properly are certain and aged, but late^ in sense 3 does not inform us about its restriction to attributive use. Adjectives that do not form their comparatives and superlatives with more and most have their inflections immediately after the headword. The Introduction states (p. xxiv) that "the comparative and superlative forms of all adjectives of two or more syllables are made by using more and most before the adjectives", but there are a number of disyllabic adjectives that do not follow

this rule and get ~(e)r/-(e)st endings. Quirk £t al.

(1972:292) lists some of these adjectives that can be Inflected. Of those mentioned there, the following are not given inflections in ALD: hollow, shallow, feeble, noble, aature, obscure, coaion, handsome, polite and wicked. As an illustration of succesfully integrated (non-coded) grammatical information on adjectives in ALD the following example may be quoted in full: afraid/.../ pred adj 1 f\i(of), frightened (of): There's nothing to be r\>of. Are you

£f snakes? 2 C O of ...gerund; ( M that, doubtful or

anxious about consequences:

was

hurting his feelings/that

might

hurt his feelings. She was ^>("Don't hesitate) to ask for

56

my h e l p . 4 r \ j ( t h a t ) ,

( t h a t usu o m i t t e d ) (a p o l i t e formula used w i t h a

statement that may be unwelcome): I ' m c v j ( t h a t ) we shall be l a t e . We missed the l a s t t r a i n , I'm f\). I'm e is misleading. Another example

of bad

grouping is [VP2C], where sentences like Please come in, I'll soon catch up with you and It looks like rain are all regarded as instances of the pattern: intransitive V + adverbial adjunct. Again, it is wrong to suggest that J^n, up with you, like rain etc. have the same g r a m m a t i c a l function; some of these strings are not even constituents. The examples given also show that ALD often mixes categorlal and functional information, instead of providing both types of Information

separately and consistently. Categorlal information seems as

essential as information about the grammatical

function of a constituent. For

the learner it is important to know, for example, that a verb like last may be

61

followed by an NP (two hours) or a prepositional phrase (PP) with for (for two hours). But he must also know that these NPs or PPs are adverbial adjuncts (cf. [ VP2B ], p. xxix).

3.1.2

The coding system used in the entries

The grammatical coding of .the entries and sub-entries in the dictionary leaves a great deal to be desired. There are, for example, many verbs which have been given an incomplete coding or no VP-code at all (e.g. decease, decelerate, deepen, deject, package,

pal, panel, panic). In some cases, verbs are followed

not only by their syntactic code but also by elements in bold type which constitute the verbal complement, as already indicated by the VP. In these cases the VPs are more or less redundant. Examples: deign ...vl [VP 4A]

to do something, ...

pander ... vi [VP 3A] O i

_to, •••

Example sentences can be of great help to the dictionary user as illustrations of VPs, since he need not consult the tables in the Introduction every time. However, ALD does not always present its example sentences, if there are any, in the same order as the VPs in the entries (degenerate, delay). For example: degenerate ...[VP2A.3A] ... Thrift is desirable, but do not let it n O into avarice. He denied that the young men of today were degenerating. ...

delay ...[VP6A,2A,B] ... Don't

. The train was r\J ed (for) two hours.

_I was r\> ed b^ the traffic. Moreover, the dictionary does not always illustrate all VPs, whereas others are sometimes illustrated more than once. For example: decrease

... [VP6A,2A]

... Your hunger

s. £ £ you eat. The

population of the village has A J ed b£ 1500 to 500. The VPs are often given in what seems to be a random order. Cf. decide ... vt, vl 1 [VP6A, 14,2A,3A] ... pair ... vt,

[VP6A, 15B,2A,C] ...

The reason for this seems to be that transitive and intransitive VPs are grouped

together in the order indicated by the abbreviations v ^ and vi^

immediately after the headword. There is no reason why vl could not always precede vt, so that the VPs could always be ordered numerically (and the order of examples adapted to this). Again there are exceptions to the rule:

62

parachute ...vt, vl [VP2A,C,6A,15A,B] ... The codes of verbs which are only transitive or intransitive are not always ordered numerically either: deduce ...vt [VP6A,14,9] ... expect ...vt [VP6A, 17 ,7,9,14] seem ...vi [VP4D,E,2A] ... Frequency of occurrence may be the reason for ordering the VPs in this way, but if it is, this is not made an explicit criterion in ALD. ALD provides no code for forms that occur in the passive only (although in the following case the example sentences supply sufficient information): delight^ ... 2 (passive) be ( \ ) ed, be greatly pleased: I was f Q ed to hear the news of your success/ CNJ ed jit the news

.../ fVJ ^d that you were

successful. ...

3.1.3

Conclusion

In many cases, ALD does not make consistent use of abbreviations and codes. As far as nouns and adjectives are concerned, information on complementation should be extended, both by means of words in bold type, e.g. ^^(for), ^ J to do, f\i sb to do and by means of well-chosen example sentences. The system of Verb Patterns has been made unnecessarily complicated. Principles governing

the choice and ordering of VPs should be

strictly

syntactic. The choice of example sentences could be improved. Furthermore, codes and examples should be ordered in such a way that the dictionary user is less dependent on the tables. It would be an enormous improvement if each VP was systematically illustrated by at least one example. All the relevant syntactic

information

should be given at each entry, with

sufficiently

transparent and self-explanatory codes. The VP-system does not distinguish properly between obligatory and nonobligatory sentences elements. Sometimes adjuncts are said to be "possible but not essential" (e.g. [VP2A]). Obligatory adjuncts play an important role in [VP15A] (e.g. Put these papers on that desk), but in [ VP16A] anon-obligatory adjunct (of purpose, etc.) is taken to be a distinctive characteristic of the VP (e.g. He opened the door to let the cat out). Cases where the dO/NP is followed by an (obligatory) complement NP or AdjP (cf. [VP22] and [VP23]) could be grouped together.

63

The c l a s s i f i c a t i o n

of

all

patterns

consisting

p a r t i c u l a r l y p r o b l e m a t i c . For p e d a g o g i c a l purposes, state that, apart from an NP, the verb leave,

of

V + NP + V i s

i t might be u s e f u l t o

f o r example, may

take AdvP (e.g.

They l e f t me o u t s i d e ) , a PP ( e . g . They l e f t me i n the l u r c h ) , or the - l n g form of

the v e r b ( e . g . They l e f t me w a i t i n g . . . ) ,

etc.

However,

we f e e l

that

the

grammatical function of the AdvP, the PP and the V-lng i s equally relevant.

If

pressed, we would be prepared to c a l l a l l these patterns SVOC or SVOA. T r a n s f o r m a t i o n a l r e l a t i o n s between sentence p a t t e r n s are brought out insufficiently

by t h e ALD s y s t e m

of

VPs;

indirect

movement,

object

p a s s i v i z a t i o n and e x t r a p o s i t i o n are among the structural changes that should somehow be r e f l e c t e d in the scheme. In some c a s e s , verbs in ALD are f o l l o w e d not o n l y by t h e i r codes,

but

also

by

strings

in

bold

type

which

indicate

grammatical the

complementation as a l r e a d y e x p r e s s e d by the VP. In these c a s e s , grammatical information i s given t w i c e ,

verb's

the same

which i s pedagogically perhaps not a

bad i d e a , but i t seems to us t h a t , i n p r a c t i c e , the VPs are made redundant in t h i s way. I t i s c l e a r t h a t the s t r i n g s i n bold type are e a s i e r t o

interpret

than the VPs. U n f o r t u n a t e l y , t h i s idea i s not e x p l o i t e d c o n s i s t e n t l y in ALD. For example, i t would have been p o s s i b l e t o add t o , say, [VP6A] and [VP6B] rvjsb/sth in bold type, to [VP6C], fs^sb sth and to [VP13A]

[VP6D] and [ VP6E]

that

to [VP12C]

f s j sth t o sb.

ALD does not g i v e example sentences f o r e v e r y code, and some VPs are i l l u s t r a t e d more than once. Moreover, ALD does not always present i t s examples in the same order as that of the VPs. Example sentences play an important part in providing grammatical information. As i l l u s t r a t i o n s

of codes,

they

vitally

support the learning process. A well-organized system of examples i s at

least

as important as the codes themselves. I f examples are always ordered according to the codes and i f each code g e t s at l e a s t one example, the d i c t i o n a r y user can r e l y l e s s on the explanations given in Introductions,

3.2

Guides or Tables.

LDOCE

Whereas ALD has a separate system to indicate the grammatical use of each part of speech, LDOCE has a s i n g l e

comprehensive system f o r a l l word c a t e g o r i e s .

This i s summarized in the table at the back of the dictionary. Although the presentation of grammatical information in LDOCE seems more

64

systematic, more consistent and easier to use than ALD's system, it still shows some deficiencies both in the Guide to the Dictionary and in the application of the coding system to the entries.

3.2.1

The Introduction

The coding system, of which the editor-in-chief claims that it "requires no knowledge of grammatical theory to be fully understood" (p. viii), is mainly described

in section

concerned with the

17 of the Guide to the Dictionary. Other

sections

coding system are 7 (Inflections) and 16 (Usage notes). The

Usage notes are very successful in conveying both semantic and grammatical details. Here we shall concentrate on section 17. He shall point out the strengths and comment on the weaknesses of the descriptions and examples. As we saw in 2.2.1 above, the dictionary uses the notion of "positions" in sentences and phrases, in an attempt obviously

to explain to users what

functional elements are. Position III, for example, gives three possibilities: She was He gave the He kicked the

happy. boy

a boat.

ball.

What we have here, in fact, are three different sentence elements, which are strangely enough classified under one "position". The table also seems to suggest that boy and ball are sentence constituents, which they are not. This explanatory device, we are afraid, is not a very good one. Instead of introducing "positions", the compilers would perhaps have done better to stick to rigorous distributional criteria, using labels that are more informative and precise than "positions". Consider, for example, the definition of position IV: it is said to be the position "that a second OBJECT or COMPLEMENT has after a first OBJECT or COMPLEMENT" (p. xxix). The inaccuracy lies in the fact that one might read into this that a complement can be followed by another complement. The "positions" in the phrase seem to be more or less redundant, since they are restricted to and also described in only a few codes, i.e. in [A], [B] and [E].

65

In the description of [D] the term "nounlike expression" is used for the first time. It also occurs, for example in [L], [T], [V] and [X]. However, LDOCE remains very vague about what exactly is meant by "nounlike expression". Examples used in the Guide are: what to do President fool One wonders whether the ordinary dictionary user will understand what the nounlike characteristics are that these examples share. [D] has a rather unsatisfactory description: "verbs that are followed by 2 nouns (boy, book; wall, paint), PRONOUN^ (him, them), or nounlike expressions (what to do), which come in positions III and IV, often as INDIRECT OBJECT (the boy) + DIRECT OBJECT (a book). These nouns, PRONOUN£, etc., always represent (REFER TO) something else, not each other" (p. xxix). Although there are sentences to exemplify what is meant, this description does not exclude ungrammatical sentences like *He gave him them or *He gave them him. There is no information on the restriction on combinations of noun/noun, pronoun/noun, pronoun/nounlike expression and noun/nounlike expression. And what is an inexperienced user to do

with the information "often as INDIRECT OBJECT ... +

DIRECT OBJECT"? The other possible construction is SVOiCs as in the example He made her a good husband. We think LDOCE should have been more explicit about this. Where ALD uses separate patterns for Give the boy a book and Give a book to the boy ([VP12A] and [VP13A]) and Buy him a book and Buy a book for him ([VP12B] and [VP13B]), LDOCE codes these sentences [D1 (to)] and [D1 (for)]. The advantage of this is that the dictionary user immediately sees that it is possible to transform the sentence into one in which the original indirect object has become a complement in a prepositional phrase with ttj or for. It is worth noting

that there is an interesting conflict between the

coding of the verb spray in the Guide and In the entry Itself. In the Guide examples of spray, as in spray paint on the wall and spray the wall with paint, are incorrectly coded [D1 + with/on]. This is incorrect since the V is not followed by two NPs. The mistake is corrected in the entry of spray^, which has different codes: 1 [X9; L9: esp. on] ... 2 [T1 (with); id] ...

Spray with is

apparently not fixed enough to be given a separate entry and should therefore not be coded [Dl].

The wrong coding in the Guide can be ascribed to what

seems to be the difficulty of defining "transitivity", especially in relation to multi-word verbs. Verbs are transitive, [T] and [V], if they take only one

66

object. This also holds good for verb + preposition, verb + adverb and verb + adverb + preposition combinations. Verbs are ditransitive, [D], if they require two objects.

Problems

arise

with

the combination

verb + adverb

and/or

preposition. If the combination of verb + adverb and/or preposition is fixed (and gets a separate entry), then the same rules of transitivity apply as for "normal" one-word verbs, cf. defer [Tl] and defer to [T1 ]. Both LDOCE and ALD code defer as a transitive verb, but the combination defer to is coded as intransitive by ALD and as transitive by LDOCE. ODCIE 1 codes defer to as intransitive rather than transitive. The problem is inherent in the decision as to what combinations of verb + adverb and/or preposition count as idiomatic (cf. ODCIE 1). Coding is again seen to depend at least partly on semantic considerations. The difficulty of determining what an idiom is may be illustrated by looking at two arbitrary examples: angle for and fish for, both of which are mentioned as examples of semi-idioms on page xi of ODCIE 1. They are coded as intransitives in ALD and ODCIE 1. LDOCE has a separate entry for angle for and codes it [Tl], but the third sense of fish 2 is coded [10; (for)]. The intransitive code is the only one that is given, since optional prepositions are not coded. There are also problems with the coding of transitivity in the case of [V], e.g. We looked at the man jumping [V4a] and They inhibited her from doing it [V4b]. [V] stands for transitives followed by a 2-part direct object. [V] requires a subcategorization into [4a] and [4b] to distinguish between the different word orders in the two sentences. It is doubtful whether [V4b] is a pattern of V followed by a 2-part dO.and not for example [X9], i.e. V + one object + something else. Consider, for example, the entry of inhibit from, [V4b; (Dl)]. [Dl] seems to imply that there are two objects in a sentence like Fear inhibits me from evil Impulses, and [V4b] suggests that Fear inhibits me from talking has a completely different structure with a 2-part direct object. See also, for example, the entry for the verb sour on. We suggest that these patterns are all [X9].

It seems very difficult to determine whether a verb should be coded [X] or not, especially [X9]. Put it in the box is coded [X9], because it conforms to the description of [X] given on page xxxiii. The third sense of spread 1 is coded [Dl + with/on] indicating that either with or on are obligatory complements, giving sentences like spread butter on bread, spread a_ piece of bread with butter (these are the examples used in LDOCE). It is strange that on bread and with butter should be objects as indicated by the code. They are adverbials

67

like ¿n the box in Put ^t iji the box. This is, in fact, the same problem as with spray above. There are also entries with [V] that should be coded [X] as in the case of delegate^. Its second sense is coded [Tl; V3] and gives the example

have

delegated Captain Roberts (to serve in my place). However, the [V3] sentence with ^o serve in my place does not have a 2-part direct object, its pattern is SVdOA, fitting in with [X]. The sixth sense of put 1 :

(to ask (officially) for (something) to be

considered) is coded [X9; (Tl)] and one of the examples is The laywer put several questions to the witness. We would argue that in an attempt to remain consistent, LDOCE has coded this [X9]. In doing so, it gives the Impression that

the witness is either a complement or an obligatory adverbial, but this

is incorrect. As a rule, optional and peripheral elements are not coded in LDOCE, for example deck out v adv [Tl (in) often pass.] ... The street was decked out in flags. In this sentence i^n flags is optional. In the active sentence, there is only one obligatory object, i.e. the street. In the sentence with put we mentioned above, the to-complement is considered obligatory (which it is not). Therefore it had to be coded, but the compilers failed to notice that it is a transformation of a [D] pattern.

The labelling of accused as a case of [GU] is not correct. Whereas Admiralty and left are obvious group nouns, which can be used with either singular or plural forms, accused may refer to a group or an individual. Therefore it should have been coded [Wn3] like, for example, deceased . In the Guide to the Dictionary dramatics is used as an example of [P] (i.e. nouns that are only used with plural verbs). It must be very confusing for students to find dramatics coded with [U] in the dictionary, which means that it is to be used with a singular verb. It is not always easy to establish the distinction between [Wnl] and [Wn2]. This is probably the reason why pheasant is given as an example of [Wnl], whereas it is coded [Wn2] in the dictionary. In most cases, noun entries coded [Wn2] give both singular and plural forms. The noun hookworm is mentioned as an example of [Wn2], but in the dictionary it is given no [Wn] code at all. In their review article on LDOCE, Wekker and Hyams (1979) pointed out the inconsistent use of the [Wv6] code. They found that out of 66 stative verbs mentioned in Quirk/Greenbaum (3.35) only 18 are coded as such in LDOCE. This may lead to ungrammatical sentences like

He i^s detesting that.

68

There are three comments we wish to make on the number coding on pages xxxiii and xxxiv. As an example of [2], LDOCE gives the following sentence:

helped clean

the windows [T2], but the entry of help does not have the code [T2]. As an example of [3 ], it gives a^ yen J^o be alone [ S3 ], which is a correct phrase. However, the code for yen 2 is [S (for,3)], which may not seem very straightforward to many users, because both the Guide and the example sentence in the entry (He has & yen to be alone in a boat) seem to suggest that the infinitive is a more or less obligatory complement. The description of [6] on page xxxiv again shows that the term "position" is not very helpful. According to this description, He decided who should go and the reason why he came are of the same type. Who should go and why he came are said to be in the same "position", but surely in the first case we are dealing with an object realized by a clause and in the second case there is a clause functioning as a postmodifier in a Noun Phrase. An additional problem is that LDOCE calls a non-finite clause a phrase. The little letters, whose values are explained on pages xxxiv and xxxv, could have been integrated much better into the coding system. We think they are only added in this way to keep the number of letters as low as possible. The difficulty, however, is that in each combination they stand for something else, e.g. [a] can be interpreted in three different ways. Another problem is that the little letters are not incorporated into the grammar table. Finally, we would like to comment briefly on the grammar table on the endpapers of LDOCE. The categories are well-arranged and the short descriptions are good, with the exception of [X], "v with one object + something else", which is not specific enough. Four things might be changed. [Wpl] and [Wp2] refer to information which cannot be found in the Table of Codes, but in separate tables on pages xxxvii and xxxvlli. As an example of [S9] the table contains a fine intelligence. This code, however, is not mentioned in the entry of intelligence. The same holds good for the example of [T2]. We wonder if [18] and [L8] really give examples of a different type, cf. Smoking is not permitted [18] and He got trapped [L8 ]. Both sentences are of

69

the same construction, which is [18]. The table should also contain information on the little letters, which are completely left out. The

greatest merit of the coding system in LDOCE is that its descriptions of

letters and numbers is very concise and clear, so that it does not take much time to study the system or to refresh one's memory. Each category also gives enough examples to illustrate what is meant. We think that after some time a student will remember the meanings of most letters and numbers because they are mnemonic, e.g. [D] stands for ditransitive and [S] for nouns used in the singular only. The numbers always have the same meaning, e.g. [5] means "followed by a that-clause".

3.2.2

The coding system used in the entries

LDOCE provides no code for nouns which are [C] in all senses, for adjectives which are [B] in all senses and adverbs which are formed regularly. All other entries are coded, and this is done very economically. Compare, for example, afraid in LDOCE with the corresponding entry in ALD. Where ALD uses rather a lot of space to inform the student of all possible forms of complementation, LDOCE uses no more than 3 codes: ...1 [F (£f, 3,5a)] ... 2 [F (£f,3,5a)] ... 3 [F5a,b] ... . The advantage of the ALD system is that the user does not need to refer to a guide as the information is very explicit, although ALD is not very consistent in this respect. LDOCE is more efficient and economical. The advantage of the LDOCE system is that the coded forms are very short. The system of letters and numbers also seems to be more logical, because identical forms of complementation of, for example, verbs or nouns are always indicated by means of the same number, e.g. deduce [Tl,5] and deduction [C,C5], which means that they can both be followed by a that-clause. However, closer examination of the codes in the entries shows that LDOCE does not fully exploit the potential of its system.

70

Nouns

Although the coding system can be used to indicate optional complementation of the noun

in a Noun Phrase, LDOCE very often

fails to supply

this kind

of

information. For e x a m p l e , decision is coded [C] in its second sense, but the fact that it can be postmodified indicated

by a to-infinitive

construction,

is not

(e.g. 'Who made the decision to go there?'). This could have been

expressed in a code like this: [C(3)] or [C,C3]. Delusion ...2 should have had [C5] to indicate the possibility of using a postmodifying that-clause. Another case in point is pace^. The first sense is coded

[S], but [S9] would have been

better. The examples are helpful here, in that they indicate that the noun "needs a descriptive word or phrase" (p. xxxiv) e.g.: The old man can walk only at jj very slow pace and The plans are being prepared at quite a_ good pace. This shows that there is too much still that is taken for granted. Information

on optional prepositions

in LDOCE is not consistent enough.

The entries for panorama, paragon, paraphrase and parcel add of as an optional complement, whereas others which can also take the of-complement do not mention that possibility, e.g. pack (only in sense 3), package and packet.

We stated earlier that LDOCE is more careful to code all entries than ALD. Good examples are those of adjectives + suffix and of nouns derived from verbs: deftness [U] and decapitation

[U;C]. H o w e v e r , we have also observed a few

problems in the coding. For example declination has no code for its second sense, although normally either all or none of the senses are coded. The first sense

of

deed

is coded

[C; in+U].

The

example

of

the

second

code

is

an

honourable man in word and in deed. The odd thing here is that word is not given this code. Moreover, in word and deed is a fixed expression and should deserve to be printed separately. Although LDOCE always gives irregular plural forms, it does not add the irregular plural of paper knife. The only example in the entry of parachute is a parachute jump. This shows that the noun can also "come in position I, before a noun" (p. xxix). However, it is not coded

[A].

Adjectives

LDOCE

is

more

explicit

than

ALD

in

its

information

on

adjective

complementation. As far as prepositional c o m p l e m e n t a t i o n of adjectives is concerned, both dictionaries s o m e t i m e s fail to supply essential information. Examples that Herbst (1984) mentions are available and useful. Neither ALD nor

71

LDOCE mentions or illustrates optional to-compiementation. LDOCE, is more explicit than ALD about restrictions on syntactic structures cf. aware and ready. Where ALD often only gives examples, LDOCE codes the structures and exemplifies them. A good example of this is adjective complementation with infinitive clauses. LDOCE, for example, codes hard [B3], difficult [B3], easy [B3], willing

[F3] and keen [F3] and gives at least one example of that

construction. Of all the adjectives mentioned in section 12.13 of Quirk et al.'s University Grammar of English it is and

convenient

that have

no

only splendid, hesitant, impossible

information

on

possible

to-lnfinltlve

constructions. ALD only speaks of ^-complementation in the entries for eager, reluctant and ready. Hard, difficult, easy, willing and keen give example sentences of to-complementation. LDOCE normally encloses optional and obligatory complements within square brackets (with additional parentheses for the optional ones), together with the letters and numbers. However, there are also quite a few cases of inconsistent coding of adjective + obligatory preposition constructions. The eighteenth sense of open does not have the code [F + to], but mentions open to separately in bold print (also compare alive to and alive with). Another case is that of bent on, which gets its own entry, coded [F, FA]. It is strange that the grammar table on the end-papers does not include the possibility of [FA]. Although this entry is an exception, it is an improvement on the other entries, because this code extends information, i.e. [FA] indicates that bent on is followed by a gerund. Keen on, for example, does not have a separate entry and the dictionary user can only deduce from the examples that it can be followed by a gerund. This system would work just as well, of course, if examples consistently supplied complete information on complementation. Yet this is not always done consistently. LDOCE should have exploited this system more fully. It should have printed adjectives (and nouns) + obligatory preposition in full in the entries and then added codes in the following way: keen^ ... 5 keen on [Fl,4]. One can compare this with a verb entry like decide on [Tl,4]. LDOCE

is

consistent

in

its

coding

of

attributive

and

predicative

adjectives. If words are restricted to particular uses in different senses, the dictionary is often very effective in showing these restrictions. Sick may serve as an example of this. Those adjectives which are not coded sufficiently in ALD all have the correct codes in LDOCE. LDOCE has also been far more careful than ALD in coding adjectives for their comparative and superlative forms. Of the dl-syllablc adjectives that can

72

be inflected only obscure does not have [Wa2]. Quiet and wicked are coded [Wal] Instead of [Wa2], Indicating that these adjectives can only be inflected and do not occur with more and most.

Verbs

We have already pointed out some of the problems we had with the LDOCE system of transitivity. Comparison with ALD shows that the main difference between the two dictionaries, as far as verb codes are concerned, is that ALD codes all verbs except the combinations of verb + optional adverb and/or preposition. For example, decide ...4 is coded [T1;V3]. The example in this sub-entry is Your words have decided me (to help you). Both possibilities are coded. The greatest disadvantage of the system of coding in LDOCE is that it does not inform the user of related constructions. With a few exceptions (cf. [D1 (to)]), each related construction is given a separate code. For example, deliver ... 5 is [X9] and is exemplified by She delivered a^ hard blow to his jaw. Sense 6 is coded [D1 (to); Tl] and has the following example: They'll deliver me all their support. Although the structures of these examples are very similar, this cannot be deduced from the codes. LDOCE codes all verbs. The codes are normally ordered alphabetically at each verb entry; the intransitive and linking verb codes, ([I] and [L]), are mentioned after the transitive codes. We have observed a number of problems. For example, the first sense of define is coded [T1;I0],

but there is only one example: Some words are hard to

define because they have many different uses. Must we assume that this is the transitive or the intransitive use of define ? The fourth sense of define is coded [Tl (as) ] and the last example is What defines us as human?. This is an example of a basic SVOC construction. The code, however, makes sis human appear as optional. In our view, this is due to the way in which transitivity is handled in LDOCE. The same kind of inadequate coding can be found in e.g. electJ, Interpret and recognize . On the other hand, regard as., and describe as are given separate entries. In the code of degenerate^, the compilers have failed to put brackets around the prepositions: [10 from and/or into]. Park^ is coded [T1;I0] for its first sense. The examples are Don't park the car in this street and I'm parked over there. We think that these sentences are examples of [X9] and [L9] respectively.

73

LDOCE is more consistent than ALD in illustrating every code with one example, especially if an entry has two or more codes. There are very few entries in which one code is illustrated more than once, while others are not used in examples at all. However, like ALD, LDOCE does not consistently give the examples in the same order as the codes. For example declare*, decrease* and deem. Rearrangement and extension of examples would greatly Improve the quality of LDOCE.

3.2.3

Conclusion

On the whole LDOCE is more consistent than ALD. Although LDOCE still does not exemplify all codes, it is more successful in supplying examples for most of them. The grammar table is easy to use. It is a pity that it does not include information on the little letters, which seem to be an unnecessary complication to the system. The LDOCE system of transitivity is bound to create problems for the users of that dictionary. Especially those who have been taught grammar from A University Grammar of English or A Grammar of Contemporary English (which LDOCE mentions as a work of reference on page xxxiii) may find the dictionary confusing in this respect. The coding system is potentially very good, but it has not been exploited to the full. Minor changes and additions, but also better and more authentic English examples,could make LDOCE a reliable dictionary for productive use.

3.3

ODCIE 1

Since ODCIE 1 has restricted its range of coded information by only accounting for transitivity and numbering the word class of the second and third parts of the verb + particle and/or preposition combinations the grammatical coding system could be kept fairly simple. The greatest strength of ODCIE 1 is that the codes and transforms are easy

74

to learn and more or less s e l f - e x p l a n a t o r y . The user does not need to r e f e r to a table of codes to f i n d out what they mean, once he has f a m i l i a r i z e d himself with the system. This i s not only due to the straightforward coding system but a l s o to the c o l l o c a t i o n s at each e n t r y , which g i v e c o n c r e t e r e a l i z a t i o n s of phrase and sentence elements that can be inserted in the headphrase or added to it.

In ODCIE 1 t h i s kind of information i s included in the entries

themselves

in the form of c o l l o c a t i o n a l l i s t s of phrase and sentence elements,

introduced

by abbreviations l i k e adj, S, 0 and o. Charge ( f o r ) may serve as an example: charge ( f o r ) [A2 pass B2 pass] ask in payment ( f o r ) 0: (a h e a v y ) p r i c e , fee;

ten pounds;

service;

ten per cent commission, o: admission,

seat;

treatment,

c l e a n i n g , r e p a i r l n g . Q " H o w much do they charge f o r washing a

car?"QThe conductor t r i e d to charge me fivepence f o r a fourpence journey. Q o f t e n with an i n d i r e c t o b j e c t as in the second example. Both the code and 0 show that charge ( f o r ) i s t r a n s i t i v e . Examples of the types of d i r e c t o b j e c t and prepositional o b j e c t which can occur in sentences with the headphrase, are l i s t e d a f t e r 0 and o. B e s i d e s the codes and c o l l o c a t i o n s , usage notes supply more g r a m m a t i c a l d e t a i l s ( a s in charge ( f o r ) ) .

3.3.1

The Introduction

On pages v i i

to x i i i ,

Cowie and Mackin describe

which verb + p a r t i c l e and/or

p r e p o s i t i o n combinations have been i n c l u d e d and the reason why they have decided on the coding system consisting of six basic patterns. On page x i types of seml-ldiom are l i s t e d . The f i r s t category i s described i n terms of d e l e t l o n / n o n - d e l e t i o n

of

the p a r t i c l e and/or p r e p o s i t i o n . The

second category i s described as f o l l o w s : The weighting given to this c r i t e r i o n also in part explains the inclusion of many s e m i - i d i o m a t i c e x p r e s s i o n s c o n t a i n i n g one of the come, go, pat,

major verbs

take e t c . Because of the r e l a t i o n s h i p of " i n t r a n s i t i v e " to

" t r a n s i t i v e " which e x i s t s between come down ( f o r example) In certain of i t s senses and b r i n g down, these e x p r e s s i o n s are o f t e n not f u l l I d i o m s . The f i x i t y of the p a r t i c l e ensures t h e i r inclusion. In many such cases, we have also been guided by the need to Include items which, while not Idioms t h e m s e l v e s , may throw l i g h t on the meanings of i t e m s which a r e . So put aside (a book,

one's k n i t t i n g ) i s included because i t i s r e l a t e d in sense

75

to

put

aside

(money,

cash)

and

put

aside

(a

grievance,

one's

differences). This category should, in fact, have been divided into two, one group consisting of pairs of related semi-idioms with major verbs (often with a causative relationship), the other consisting of homonyms. Although both groups contain major

verbs,

in

the

first

group

we

find

verb

+

particle/preposition

combinations with different elements, e.g. coae down^ and go down^ , coae o n ^ and

be

on^,

in

the

second

group

the

same

combination

of

verb

+

particle/preposition has more than one entry to differentiate between all its meanings. Cowie and Mackin's reasons for not adopting the familiar sub-division into phrasal verbs and prepositional verbs, as described on page xii, are clear. Their first argument refers to semi-idioms: To call such expressions 'phrasal verbs' or 'prepositional verbs' would imply that they are unities, which is inconsistent with the facts. A second practical objection is that the dictionary contains many complex items - push the boat out and bring the house down for example - which the scheme does not provide for. While both these expressions contain a verb and a particle it would not make sense to speak of them as "phrasal verbs".

Finally, if we were to adapt and enlarge the scheme so that it

took account both of the scale of idiomaticity and of the difference between simple and complex expressions we should end up with a system that was both cumbersome to operate and difficult to interpret. The system of only indicating the grammatical structure of expressions (e.g. verb + particle, verb + preposition) indeed ... has the descriptive

advantage of enabling us to speak of make up

(one's face) etc as a unit of meaning while at the same time leaving us free to speak of the grammatical separability, or mobility, of its parts (as in make up one's face/aake one's face up. (p. xii)

On pages xxi and xxxii three types of nominalized form are discussed. Examples of the first two types are break-in

and upkeep.

The

third

category

is

represented by only two nouns in 0DC1E 1, i.e. outpouring and upbringing. No other nouns of this pattern can be found in the Index of nominalized forms (p.394 -396). Although this is a category of minimal size, the nouns are not dealt with consistently in the entries. Outpouring is only used in an example and

upbringing is mentioned separately as the nominalized form of bring up 1

(no example).

76

Oil pages xxviii ff. the grammatical coding system is described in detail. The classification of the six basic patterns depends on whether the verb is transitive or intransitive, and on whether it combines with a particle, a preposition or particle + preposition (cf. xxix). Transitivity

is essential to the classification. The absence or presence

of the direct object determines the choice of the letters A and B in the code. For that reason, the definition of transitivity could have been given more prominence. It has only been Included in brackets (p.xxviii): These

"basic" patterns

can

be divided

transitivity (i.e. according

to whether

into

two

groups according

to

or not they contain a Direct

Object). F u r t h e r m o r e , the difference between direct object and prepositional object could have been dealt with more extensively to avoid confusion. The six example sentences already referred

to in 2.3.1 above, can be represented in a table as

follows:

subject

verb

direct

particle

preposition + prep, obj

object

[Al] The electricity

went

off

supply [A2] We

were banking

on a change of heart

[A3] The committee

fell

[Bl] Fred

tipped

the police

[B2] Peter

foists

all his

on his unfortunate

problems

friends

[B3] You

put

back

the

on an earlier plan

off

down

to bad planning

shortage

The advantage of such a table is that it clearly illustrates the differences between transitives and intransitives, particles and prepositions and direct objects and prepositional objects. The comparisons between [A1]/[B1], [A2]/[B2]

77

and [A3]/[B3], mentioned on page xxix, make the information complete. From page xxxv onuards all patterns and related transforms are extensively illustrated

in tables.

Under

these

tables

information

on all kinds of

grammatical, syntactic, semantic and stylistic details is given in the form of notes. Sometimes, however, these notes also contain important grammatical information, for example on the place of (optional) adverbs in the headphrases, which can only be found there. Neither from the example sentences nor from the collocations can it be deduced whether adverbs or adverbial phrases may be inserted between parts of the headphrase. Consider notes (c) and (d) to [A3] on page xlii: (c) In some cases, an adverb or adverbial phrase may be inserted between the particle and the prepositional phrase. As the following examples show, it is possible to separate particle and preposition in this way even when the latter cannot be removed: 5 They had done away reluctantly with this piece of legislation. 7 We put up cheerfully with these interruptions. (d) The verb may sometimes be divided from the particle, though generally only by the adverbs "right" or "straight": 5 They had done right away with this piece of legislation. 9 The family came straight up against fresh problems. (Occasionally, we find such adverbs of degree as "completely", "totally", "entirely" in the position of "right" in no 5.). In no way have these details been included in the entries of do away with, put up with and come up against. Other examples of notes with information that cannot be found in entries are: -[Al nom] note (c) (except break-In), which points out the possibility of forming sentences with There is/was + nominalized form. -[Al emph] note (b), which reads as follows: (b) In some sentences in which the subject is a noun, or noun phrase, [the particle] may precede the verb: 10 Down the snow came. 12 In the sun went."(p. xxxvi) -[A2 emph] note (c): In some examples of this transform, a contrast is implied between the noun in the prepositional phrase and another unspecified noun. This noun may be made explicit, as follows: 5 For his family he has provided well (but not for his employees).

78

-[Bli. nom] notes (b), (c) and (d), which point to related sentences of the basic pattern with respectively the verbs do/make, give and get. The sentences Bill took the premier (him) off _to perfection [Bli] and Fred tipped the police (them) off [Bli] can be transformed into: Bill did £ perfect take-off off the premier (note (b)) Fred gave the police the tip-off (note (c)) The police got the tip-off from Fred (note (d)) -[Bli pass] notes (b) and (c): (b) When the Object of the active sentence is a clause, it remains in the final position in the transform and "it" is introduced initially. Compare: We worked out that we should need 100 pounds. It was worked out that we should need 100 pounds. (c) There may be two types of passive sentence when the active pattern contains an Indirect (as well as a Direct) Object.

Compare

these

acceptable transforms: was sent down a_ list of missing stock. A list of missing stock was sent down £o me. with the unacceptable £ A list of missing stock was sent down me.

With the exception of break-in in [A1 nom] note (c), the entries of headphrases used to exemplify the information in the notes do not contain that information. Although these notes are not of primary importance to the construction of grammatically correct sentences, they make the user more dependent on the 70page introduction.

3.3.2

The coding system used in the entries

The strengths of ODCIE 1 are that all its entries have codes. Both obligatory and optional particles and prepositions that are shown in the headphrases are coded. Furthermore, many entries supply the user with collocations for productive use, each headphrase is illustrated by at least one example and usage notes often give additional information. The advantage of collocations is that not every code and/or transform needs to be illustrated, because the user can make his own sentences with the

79

help of the realizations of sentence elements as printed in the form of collocational lists. It is a pity, however, that some entries give several examples of one code, whereas others are not illustrated at all. Claim back (froa) [Bli pass B3 pass rel], for example, has three [Bli] examples (in active sentences), whereas the other possibilities are not illustrated. Another example is carry about (with one) [Blii B3]. This entry only gives two examples of [B3], [Blii] has not been illustrated. The codes are always ordered in such a way that the Intransitive uses come first, followed by the transitive ones, if any (numerically). The transforms are printed in the following order: nom, pass, pass (o), adj, emph, rel. Compare:

call (In) (at)

[Al A2 rel A3 rel] ...

catch up (with)

[Al A3 pass rei Blii pass] ...

cast off^

[Bli nom pass adj] ...

cast up (on)

[Bli pass adj B3 pass emph rel] ...

Like OALD and LDOCE, ODCIE 1 does not always order example sentences in the same way as the codes. For example, call (In) (at) orders its three examples [A3], [A2], [Al] and cancel (each other) out [A1 Blii] reverses the order of illustrations. Other entries with this kind of inconsistency are: -call up^ [Bli nom pass adj], which orders the example sentences [Bli], [Bli adj], [Bli nom]. -carry on (with) [Al A3 rel], which reverses the order of its two examples, -castaway/cast away [nom pass (Bl)], where the example with the passive is mentioned before the sentence with the nominalized form, -cast up (on) [Bli pass adj B3 pass emph rel], which includes two examples in the following order: The shipwrecked sailors were cast up on the shore of a desert island (= [B3 pass]), The sea casts up the detritus of civilization on rocks and beaches all over Europe (= [Bli]), and -chalk out [Bli pass adj], which first gives an illustration of a passive sentence ([Bli pass]) and then of an active one

([Bli]).

The exemplification of codes could probably be improved. A good example of how an entry could be organized

is go across (to) [Al A2 A3], where the

individual examples are preceded by the relevant code:

[A2]

so that the villagers could go across the marshy area.

Planks were laid

[A3] I'm just going

across to the pub for half an hour. Call away is coded [Bli pass]. The usage note says "very often passive, with no agent mentioned." As S-collocates are given: doctor, midwife, nurse; police constable; priest. This implies that these are the subjects of the

80

active sentence, which is not true. They are the subjects of the passive pattern, but this is not indicated (as in carry on^). Carry out 1 has a code which is not dealt with in the introduction, i.e. [Bliii adj]. This code should not have been included in the square brackets, because carried out is not used as a participial adjective. The nominalized form of cast off^>3 (cast-off) is not included in the Index of nominalized forms on pages 394-396. In the second example sentence of cast up (on) the preposition on should be printed in bold-face, although it could be argued that the prepositional phrase in the first example, The shipwrecked sailors were

cast up on the shore of the desert island, is "less optional"

than the one

in the second

sentence,

The sea casts up the detritus of

civilization on rocks and beaches all over Europe. n Catch up'' is coded [Bll pass B3 pass]. Either a preposition should have been added to the headphrase or [B3 pass] should have been left out. Now we either have an incomplete entry or the wrong coding. Chicken out/out of is coded [A1 A3], but out of is not a combination of particle + preposition, it is a preposition. The code should be [A1 A2 ] as in the cases of check out/out of and ooze out/oat of.

3.3.3

Conclusion

On the whole ODCIE 1 has a reliable coding system. The Introductory pages give an excellent description of the system adopted. The collocations which create a context for the headphrases are especially valuable. ODCIE 1 is a specialized dictionary, working with one small set of related categories. It would be difficult to adopt a similar coding system for general dictionaries like OALD and LDOCE, since the number of letter/number combinations would have to be enormously

extended.

As a result,

the system

would

become

less

self-

explanatory. Furthermore, the space in a one-volume dictionary would simply not be available. Each meaning has a separate entry in ODCIE 1, so that clearly differentiated sets of collocates can be given. If this were to be done in OALD and LDOCE the dictionaries would be many times the size that they are now. ODCIE's coding system lives up to the demands that can be made on such a system: it is complete, consistent, clear and easy to use, although some users will have to get used to the somewhat unusual classification into six basic patterns.

81

3.4

ODCIE 2

The grammatical coding system of ODCIE 2, which appeared eight years after ODCIE 1, is completely different from that in its predecessor. This is mainly due to the differences between the types of entries recorded in the two dictionaries. Volume 1 contains a rather coherent set of entries, whereas ODCIE 2 gives Idioms of various kinds. The sub-title of the book is: Phrase, Clause Sentence Idioms. In this section we concentrate on the coding system as described in the introductory pages (3.4.1) and as applied to the entries (3.4.2).

3.4.1

The Introduction

A striking

feature of ODCIE 2 is that is does not have a coding system

consisting of letter/number combinations as In the other three dictionaries. Instead, it uses abbreviations for sentence elements and phrases. Advanced foreign

students,

the

main

target

group

of

ODCIE

2,

will

find

these

abbreviations very convenient, so that no extra guidance for understanding the codes is needed. Descriptions and examples of sentence and phrase elements are given on pages xxix to xxxvii. Of the usual seven basic clause patterns (cf. Quirk et al. (1972)), only five are coded in ODCIE 2, i.e. SV and SVA are missing. The reasons for the restriction to 5 clause patterns may be the low frequency of occurrence of the SV and SVA patterns and the fixity and uniqueness of the clause elements, so that adding codes would become redundant. The clause patterns distinguished are the following: Verb + Complement

go berserk

Verb + Direct Object

ease sb's conscience/mind

Verb + Direct Object + Complement

paint the town red

Verb + Indirect Object + Direct Object

do sb credit

Verb + Direct Object + Adjunct

take sth aaiss

The examples and notes to the clause patterns on p.xxix ff. are clear. There are, however, three comments we wish to make.

82

a) The fourth example of [V + 0] is The minister can't blink the fact any longer. The way this sentence is Included 1 n the table seems to suggest that the direct object of the sentence is the fact any longer, but obviously the pattern of this sentence is SVOA, with any longer as an optional adverbial. b) Note (a) of [V + 10 + 0] states that the 10 in blow sb a kiss is optional, but this is in no way indicated in the entry. c) The first part of note (b) to [V + 0 + A] reads as follows (p.xxxiii): "When the direct object is relatively long and the adjunct relatively short, they can sometimes be transposed: A^ doctor has to put first his poorer patients and their families." Although this is an interesting note, the dictionary user cannot find this information in the entry of put sb/sth first Itself. There are a few other notes which contain information that cannot be found in the entries themselves, e.g. note (b) to [V + 0] and note (d)

to [V + 10 + 0].

If a headphrase has a code, all its elements are generally expressed by means of an abbreviation in that code, e.g. make bricks without straw, which is coded [V + 0 + A]. Optional elements in headphrases are not always coded. In change/swap horses (in mldstreaa) [V + 0 pass], clear the decks (for action) [V + 0 pass], go etc blue (In the face) [V + Comp], Bake sth clear/plain (to sb) [V + 0 + Comp pass] the adjuncts, although part of the headphrase, are not coded. Optional elements other than adjuncts, however, do have a code, e.g. •ake (sth) fast [V + 0 + Comp] and pay (sb) a compliment [V + 10 + 0 pass]. If an entry of a clause idiom in 0DCIE 2 is coded, the code usually represents the structure of the headphrase only (with the exception of optional adjuncts). Go etc berserk, for example, is coded [V + Comp]. The grammatical code

informs

us

only

about

the

structure

intransitive verb go, but the etc indicates

of

the headphrase

that

with

the

other verbs are also

possible. The collocational list of verbs in the entry mentions drive sb and send sb as transitive alternatives for go. Transitivity is here only expressed by means of the abbreviation sb; the code does not reflect this fact. Make etc (sb/sth) a laughing stock is coded [V + 0 + Comp pass]. The code only applies to sentences like John has made his colleagues a laughing stock. Collocating verbs in this entry are be and become. Since these verbs can only occur in the [V + Comp] pattern, sb/sth in the headphrase is in brackets. This, however, is misleading, for it suggests that sb/sth is also optional in sentences with the verb make. The only way to prevent the dictionary user from

83

drawing such wrong conclusions is to add another code to the entry and to place etc after sb/sth: make sb/sth etc a laughing stock [V + 0 + Comp pass] make a person, institution or object appear ridiculous. V: make sb/sth; [V + Comp] be, become. ... In this way, the codes explicitly inform the dictionary user that there are two possible patterns. The coding system of ODCIE 2 is more informative as far as phrases are concerned. If a phrase can occur as the realization of more than one sentence element, this is usually indicated in the code. For example, a clean sheet^ is coded [0/o(NP)], which means that the NP can function both as a direct object and as a prepositional object. The collocating verbs in the entry are ordered in the same way as the abbreviations in the code, e.g.: V: have, give sb; start with, begin with. The coding system for clause idioms in ODCIE 2 differs from that in the other dictionaries. The codes in ALD, LDOCE and ODCIE 1 are Intended to deal with all potential constructions and/or inflections of headphrases. An entry in those three dictionaries may have several codes. For example, catch up (with) in ODCIE 1 is coded [A1 A3 pass rel Blii pass], representing all possible constructions in which this idiom can occur. However, if entries of clause idioms

in ODCIE 2 are coded,

they have only one code.

Information on

constructions other than the one represented by the code, must be deduced from collocations, examples and/or usage notes. Besides five basic patterns, ODCIE 2 also has a so-called "possessive clause pattern", labelled [possess] (see pp. xxxiii f.). The purpose of this grammatical code is to identify "possessive" idioms (principally clauses with get, give and have) and to show how these patterns are related. The code [possess] applies to [V + 0] and [V + 10 + 0] patterns. For example: The leaders got a taste of their own medicine as opposed to The voters gave the leaders a taste of their own medicine. It is noted that the verb get in such possessive idioms can often be replaced by have and that there is also frequently a corresponding prepositional construction with with or without. Compare: The customer was getting cold feet/The customer had cold feet/The customer with cold feet ... etc. The obvious advantage of this is that the code attempts

to identify an important grouping of clause

idioms.

However, the code will not always be clear to dictionary users. It does not always have the same meaning as in get etc cold feet. This headphrase is of the [V + 0] pattern, whereas give sb, which is mentioned as an alternative for

84

get in the list of verbs in the entry, has a [V + 10 + 0] pattern. Both patterns are possible here. However, the same code [possess] is used if only one of the two patterns is allowed. For example, get etc a good/bad press, which has a limited set of verb collocates, (have, enjoy, and suffer), can only occur in the [V + 0] pattern. Give sth all one's got has the same code, but is restricted to the [V + 10 + 0] pattern. Furthermore, there are headphrases coded [possess] which have three patterns rather than two. Get etc a shot in the ara, coded [possess], has the following list of verb collocates: V: get, have (got), give sb; be ... (for), provide ... (for); bring ... (to). Besides sentences of the [V + 0] pattern and the [V + 10 + 0] pattern, there is a third pattern, [V + Comp] , with be ... for. Another example is give sb/sth etc the thumbs down, which is also coded [possess]. The verb collocates are give sb/sth, get and be. Again the three patterns are [V + 10 + 0], [V + 0] and [V + Comp], respectively. The code [possess] is an exceptional element in the coding system. It is meant to cover two related patterns. The choice between [V + 10 + 0] and [V + 0] depends on the verb. But there are many similar relations which have not been coded in 0DCIE 2. Make etc (sb/sth) a laughing stock, as we have seen, has no code representing both the

[V + 0 + Comp] pattern, as in They

made

themselves a laughing stock, and the related [ V + Comp] pattern, as in He and his attempts to revive the old customs are just a local laughing stock. The idea of grouping patterns by means of a word-code is certainly not a bad one, but the possessive code in 0DC1E 2 lacks generality and precision. Moreover, it does not form part of an overall system for indicating structural relations. It seems to be an arbitrary choice. The tables and examples of the functions of phrases are short and clear. There is, however, one wrong example in the table of [AdvP]. Number 4, full in the face, is not an AdvP, but a PP, in which full is a modifier of the PP. The heading of the fourth category of codes is: Noun phrase pattern with repeated element (p. xxxvii). This is wrong, since it is not only NPs that occur there, but also AdjP, VPs, etc. The second example of the table of [n + n] is (the) deaf and dumb. This example is potentially misleading since this pattern can also be [adj + adj]. On the whole the description of the grammatical coding system in the introductory pages is excellent. It is also important that it has been kept fairly brief. With the exception of the possessive clause pattern, the codes are clear and self-explanatory.

85

3.4.2

The coding system used in the entries

In contrast with Volume 1, ODCIE 2 has not coded all its entries. It is only "the

commonest

clause

and

phrase

patterns

and/or

functions

identified by means of a g r a m m a t i c a l code" (p. xxviii). A m o n g

[that]

are

the non-coded

entries we have found not only sayings and catchphrases, but also clause and phrase idioms (for example, catch fire and a Batter of fact). It is obvious why a saying like a cat aay look at a king or a catchphrase like iao bites dog is not coded: both have fixed e l e m e n t s and do not allow variations.

One might indeed argue that clause and phrase patterns which are

invariable need no code (although phrase patterns should have indications of functions). This, however, is not done consistently in ODCIE 2: catch (a) cold is coded

[V + 0], but catch fire is not coded, even though neither of them

allows variations or transforms. We can compare this with catch sb's attention [V + 0 pass], which does not give V or 0 collocates. This entry is coded because a passive transformation is possible. Other examples of coded headphrases without collocations or transforms are: call a spade a spade [V + 0 + Comp], carry/hold one's liqnor [V + 0] , carry the can [V + 0], carry/win the day [V + 0], catch one's breath [V + 0], change colour [V + 0], change gear [V + 0], chew the fat/rag [V + 0], sake/pull a face [V + 0], make it [V + 0], m a k e old bones [V + 0], make one's presence felt [V + 0 + Comp], aarry m o n e y [V + 0], mean business [ V + 0], aeet sb halfw a y [V + 0 + A ] and meet one's aaker [V + 0]. E x a m p l e s of non-coded entries which do allow collocational collocates),

lists

are:

can(not) call

a soul

variations and etc one's o w n

include (with

0

can't/couldn't do sth for the life of one etc (with V and

A

collocates), (not) care/give a daan (with 0 and A collocates), make sb free of sth (with O and o collocates), a a k e good s t h 1 + ^ (with 0 collocates), a a k e sth stick (with O collocates) and Bean it etc (with O collocates). As the compilers say (p. xxviii), the criterion ODCIE 2 is whether

for coding entries in

they are c o m m o n or not. This can be compared with the

criterion for including entries in the dictionary. In both Volume 1 and 2 it is stressed that a large number of semi-idioms has been included in order to be as informative as possible and because the distinction between Idioms and nonidioms is not easy to establish. For the same reasons one would have expected more entries to be provided with a code (commonness being equally hard to

86

determine). It must be added that, as in Volume 1, the collocations in ODCIE 2 are of great help. Mean it etc, for example, is not coded, but the collocational list of direct objects supplies the dictionary user with enough information to see that the headphrase is an instance of a [V + 0] pattern. Still, the entry should have been coded. The collocations are the best part of the grammatical information. They give exact and concrete realizations of clause and phrase elements. The abbreviations preceding the collocational lists inform the dictionary user about where to insert alternatives in the headphrase. As we have noted above, an important difference between the coding system of ODCIE 2 and the systems of ODCIE 1, LDOCE and ALD is that ODCIE 2 does not code potential constructions, but only the headphrase. Of course, there is an enormous difference between the entries in ODCIE 2 and those in ALD, LDOCE and ODCIE 1. The clause idioms in ODCIE 2 are 'semifinished' products, their structures and realizations are fixed in most cases. In ALD, LDOCE and ODCIE 1 the headphrases are raw materials. Their codes should be guides to all patterns in which these headphrases can occur. Phrases are not coded if they are not restricted to a particular use, e.g. card-carrying aeaber etc has a wide range of functions. It can be S, Comp, do, iO and prepositional object (o) and is therefore not coded. If phrases are restricted to one or two functions, this is indicated by means of a code: cap in hand [A (NF)], a change for the better/worse [Comp/0 (NP)]. The coding of phrases is not always complete. A man etc after one's own heart is coded [Comp (NP)], indicating that the NP can be used as a complement in a clause like John is £ man after my own heart. However, the collocational list of verbs also includes flnd, a transitive verb. In sentences with this verb, as illustrated in the example sentence Near Huelgoat, in the middle of Flnistere, we found an oppidum after our own heart, the headphrase functions as direct object. In another example with meet it also functions as direct object. The code fails to provide this information. There are two reasons why codes of clause idioms in ODCIE 2 never contain o: (1) Idioms (headphrases) with verb + preposition are included in ODCIE 1, (2) ODCIE 2 only codes headphrases. If verbs in headphrases can be replaced by collocating VPs consisting of verb + preposition, this alternative pattern is not coded. For example, Bake etc a decision is coded [V + 0]. This code only represents the analysis of that headphrase. Verb collocates in the entry are: make; reach, arrive at, come to. In terms of ODCIE 1, arrive at and come to are

87

intransitive, i.e. [A2]. Not only does this show that the coding systems of ODCIE 1 and ODCIE 2 are based on different starting points, but also that the systems of labelling clause and phrase elements in ODCIE 2 differ. The coding system of ODCIE 1 consists of two parts, the letter/number combinations and the abbreviated information on transforms. Of the latter, only [pass] is'found in ODCIE 2. Whether entries in ODCIE 2 can be nominalized or used attributively is not indicated by means of an abbreviation in the square brackets. This kind of information, with the exception of [pass], is always mentioned in the notes (except in chips with everything). Examples: cash and carry ... • attrib use a cash-and-carry store, business. chill sb's/the spine etc ... • adj compound spine-chilling; n compound spine-chiller. a chinless wonder ... • ...; attrib use a chlnless wonder type make etc mischief ...'• n compounds a mischief-maker, mischief-making In Volume

1 this information

is only given in the notes if there is a

structural change in the headphrase. Similarly, Volume 2 gives nominalized forms in the notes, because they always involve these changes, cf. make etc mischief. However, there are many headphrases which can be 'adjectivized' without changing order or form, but which are still not given the abbreviation [adj], e.g. a candid camera, cash and carry, a chinless wonder, make or break etc, a man/woman of the world. ODCIE 2 always includes two or more example sentences in every entry. Together with the collocations they give the dictionary user enough contexts for understanding and using the headphrase. If noun or adjective compounds can be made of the headphrase,

they are explicitly

mentioned

in notes

and

illustrated in sentences (e.g. call sb names, chocolate box, a mad dog, make do and mend). On the other hand, chips with everything gives only an illustration and change/turn one's coat gives only the nominalized form turncoat in a note. Unfortunately, the only coded transform, i.e. [pass], is not

always

illustrated, even if there are more than enough active example sentences, e.g. cast one's net wide etc, catch sb's attention, catch sb's eye^+^, make matters/ things worse, make etc a speech. Finally, we would like to point out some of the inconsistencies and errors that we have found in the entries. Can have sth for the asking and can have sth for the catching etc are both coded [V + 0 pass]. The 0 in the codes apparently stands for sth for the asking and sth for the catching etc. The O-collocates

88

are confusing since in can have sth for the asking the list of O-collocates is: it, money, help, admittance, fame, love. This suggests that only sth in the headphrase is ()• In can have sth for the catching etc there are even two lists of O-collocates: can have sth for the catching etc [V + 0 pass] ... 0: fish, berries, profits, opportunities. 0: picking, grabbing, having, taking (away). ... This implies that there are two direct objects. The problem could be resolved by coding the headphrases [V + 0 + A pass]. A further advantage of this is that passivization of both headphrases becomes more straightforward, because only the direct object (i.e. sth) becomes subject of the passive sentence. Besides changing the code, the second collocational list of can have sth for the catching etc should be changed into A: for the catching, picking, grabbing, having, taking (away). Although optional elements

in the headphrase are coded in ODCIE 2,

change/swap horses (in midstream) does not code the adverbial. Phrases with a repeated element are not always coded consistently. Chapter and verse, for example, is coded [0 (NP)] and adds as a note that it is "nonrev". One would expect the entry to have been coded [0 (n + n non-rev)]. Make or break etc is not coded at all, although a code would have been useful here. The VP is one with a repeated element, [v + v non-rev], and is transitive. For that reason, etc should have been substituted by sb/sth. Furthermore, the [V + 0] pattern we have here can be passivized. A complete code fitting in with the system of ODCIE 2 might be: [V (v + v non-rev) + 0 pass]. Man and boy is coded [n + n non-rev]. A note adds: ... • functions as an A (of duration). This was not necessary in the light of what is stated on page xxxvii: "Some phrases have a more limited function, and this is shown as follows: A ( n + n non-rev)] 1 She devotes herself to George body and soul. 3 He plays that damned trumpet day and night." The note could have been left out and the information incorporated into the code: [A (n + n non-rev)].

3.4.3

Conclusion

The main conclusion we can draw after studying the coding system in ODCIE 2 is that the codes as such are less informative than the ones in the other three

89

dictionaries.

Moreover,

many e n t r i e s are not coded.

The codes m e r e l y supply us w i t h the a n a l y s i s of the headphrase, and the "dynamic" dimension i s added by the c o l l o c a t i o n s ,

examples and notes.

The codes play a less important r o l e than the combination of c o l l o c a t i o n s , examples and notes. However, the basic information necessary f o r the a c t i v e use of entries is there, even though i t i s not f u l l y Incorporated into the codes. The advantage of

ODCIE 2 i s

that

grammar i s

g i v e n such a prominent

p o s i t i o n and that the user need not c o n s u l t the i n t r o d u c t i o n e v e r y time he wants to use the dictionary. A comparison between ODCIE 1 and ODCIE 2 shows that i t i s hard to develop a comprehensive coding system f o r both types of ( i d i o m a t i c ) items. The e n t r i e s i n both d i c t i o n a r i e s account of

are so d i f f e r e n t

too many requirements.

that one system would have to take

Such a comprehensive coding system becomes

either very complicated or inadequate. Cowie and Mackin have d e c i d e d dictionary,

to d e v e l o p

s e p a r a t e systems f o r

with sub-divisions in ODCIE 2. Both d i c t i o n a r i e s

each

g i v e an enormous

amount of grammatical information, but ODCIE 1 i s more successful than ODCIE 2 as f a r as coded i n f o r m a t i o n i s concerned. The codes in Volume 1 c o v e r patterns and transforms of

the e n t r i e s .

all

E s p e c i a l l y codes of clause idioms in

ODCIE 2 do not supply the user with a l l the necessary information on changes in the basic

3.5

3.5.1

patterns.

LDOPV

The Introduction

As we have noted in 2.5.1, LDOPV uses the same codes as i t s companion volume LDOCE f o r verbs. The main d i f f e r e n c e between the treatment of phrasal verbs and n o n - i d i o m a t i c verbal combinations

in LDOCE and LDOPV i s that in LDOPV a l l

sequences of verb + adverb and/or preposition,

or verb + a d j e c t i v e or pronoun

are made headphrases. LDOCE only l i s t s such combinations as headphrases i f are considered to be s i n g l e units of meaning, However,

as Cowie and Mackin o b s e r v e ,

i.e.

semantically s i n g l e

idiomaticity

i s hard to

they verbs.

establish:

"...the boundary between h i g h l y i d i o m a t i c i t e m s and the r e s t i s not s h a r p l y drawn but hazy and imprecise" (ODCIE 1, p. x). That Is why they decided to base

90

their coding system on form rather than on meaning, thus avoiding the problem of having to determine whether verbal combinations are idiomatic or nonidiomatic.

It is clear that this boundary is difficult to draw. Yet it is

precisely this problematic division of multi-word verbs into idioms and nonidioms which lies at the heart of the coding system in LDOPV. In LDOPV the difference between phrasal and non-phrasal verbs (or senses) is indicated by means of an asterisk. The phrasal verb is treated as a single unit, and the non-idiomatic combination is treated as a verb with something else (an adverb, a preposition, etc.). The question of the idiomaticity of a multi-word verb (that is, the problem of deciding whether it is a phrasal or a non-phrasal verb) seems to be resolved in LDOPV by an appeal to the following criterion: a multi-word verb is idiomatic (phrasal) if the meaning of the combination cannot be guessed from the verb alone. Thus, it is argued that hold up in the sense of 'to stop (a vehicle) by force in order to rob it', as in The criminals held up the train and stole all the passengers' money, is idiomatic, while hold up in the sense of 'to raise (something)', as in Hold up your right hand..., is non-idiomatic. In the former sense (sub-entry *2 in the dictionary) hold up is said to be a phrasal

verb,

in

the

latter

(sub-entry

1) it

is a non-phrasal

verbal

combination. There are numerous cases where we disagree with the classification made, and thus with the grammatical coding. It is not only multi-word verbs such as hold up in the above-mentioned idiomatic sense *2 that are regarded as phrasal verbs. LDOPV also treats verbpreposition combinations such as depend on or despair of as phrasal verbs (hence they are marked with asterisks in the dictionary). The criterion here does not appear to be a change in meaning of depend or despair (since the meaning does not change and the meaning of the combination can be easily guessed from the verb), but rather that these verbs always require a fixed preposition (on/of, • respectively, not from, with, etc.). In the dictionary the different senses of depend on and despair of are coded [Tl], [V3], [V4a], etc., just like the idiomatic verb-adverb combinations such as hold up 2. The problem is that in the case of verb-adverb combinations the criterion of idiomaticity is applied (a semantic argument), while in the case of verb-preposition combinations

it is the fixity of the preposition

that counts (a formal

argument). It is clear that multi-word verbs such as depend on and despair of are the so-called prepositional verbs, and it would have been better if they had been treated like that.

91

On Che third page of the Introduction the criterion of idiomaticity is repeated in a less extreme form: "Phrasal verbs cause problems for the student not only because of their meaning, which cannot always be guessed from the meaning of the verb, but also because of their grammar". However, a comparison between LDOPV and LDOCG shows that it is often hard to distinguish consistently between idiomatic and non-idiomatic entries or sub-entries. For example, impregnate is coded [Tl] in LDOCE and has [(with)] added to its second sense. This seems a logical choice, because the meaning of the verbal combination is closely linked with the meaning of the verb alone: 'to cause a substance to enter and spread completely through (another substance)'. LDOPV, however, marks senses 2 and 3 of impregnate with with asterisks, although 2 is similar to the above-mentioned sense 2 in LDOCE: impregnate with v prep

*2 to fill (something) completely with (a substance): [D1 (usu. pass.)] This cloth has been Impregnated with a special chemical which protects it against rain. Other examples in LDOCE and LDOPV that point to the same problem include: agitate 3 [10 (for)] /¿agitate for [Tl], chuck 2 [Tl (IN, UP)]/chuck in *2 [Tl], clap 3 [X9, esp. on] /*clap on 2 [Dl]. Some entries are recorded as non-idiomatic in LDOPV and as phrasal verbs in LDOCE. Nip in, clean up and change into are examples of this inconsistency. Compare:

LDOCE: change into v prep 2 [Dl] to cause (someone or something) to become (something different): The scientist tried to change iron into gold. LDOPV: change into v prep 2 to cause (something or someone) to become (something or someone else): [Tl+into]. The prince was changed into

beggar....

In 3.5.2 we shall discuss the effect that the difference in status between phrasal verbs and verbal combinations has on the grammatical coding system.

92

As in LDOCE, the Information given in the introductory pages of LDOPV has been kept to a minimum. It takes the dictionary user only a very short time to read the definitions and to master the grammar codes. In contrast to the introduction of LDOCE, which includes separate descriptions of the letters and numbers, LDOPV immediately informs the user about the meaning of every letter/number combination as a whole. At this point we should like to make a few comments on the LDOPV grammar codes for phrasal verbs. The idiom *lower oneself is given as an example of [13]: it is considered to be an intransitive phrasal verb. Other reflexive verbs have been coded likewise, e.g. *orientate oneself [10 (to)] and ^overreach oneself [1(3]. It is strange that they should be treated in the same way as, for example, *set out and *break down. LDOCE codes them as transitive verbs ([Tl]). Phrasal verbs - at least as conceived here - are units of meaning and the Longman grammatical coding system accounts for that by treating them as single verbs. If this is the assumption, one may wonder why the examples of the various codes in LDOPV so much emphasize the separation

of the parts of the

phrasal verbs? Compare the following: [Dl] *help to v prep Can

help

you

to

v

obj

prep obj

some more potatoes?

[T5] look out v adv Look

out

that you don't catch a cold,

v

adv

that-clause

It would have been sufficient and more consistent to print the phrasal verb in bold italics, and to leave out the abbreviations v, adv and prep below it. These abbreviations are already given after the headphrase. The following representation may be an alternative: [Dl] *help to Can

v prep

help you obj

to some more potatoes? obj

93

We must also point out the confusing use of the term "verb" In the definitions of the codes. Consider for example: [Tla] a transitive verb with a noun direct object. The adverb always comes just after the verb. The first occurrence of "verb" refers to the combination of verb +

adverb as a whole, the second refers to the verbal part of the

combination. One of the objections we have raised

to the LDOCE coding system is that

the letters a, b and c do not always have the same value. This problem has been avoided in LDOFV (a and b always mean the same), and it seems that this was possible because the entries involved constitute a fairly homogeneous category. The small letters in LDOPV refer to the separability, or not, of the verbal part and the adverb and/or preposition. The letters a and b (c does not occur) have been fully integrated into the grammar codes. It is interesting to note that the description of [V], as in [V3], given in LDOCE has been adapted in LDOPV. LDOCE states that the complementation of help in J[ helped

him to clean the windows is a 2-part direct object. In LDOPV,

the verb is said to be followed by a direct object and a to-infinitive, but the grammatical function of the infinitive is not given. Some descriptions of the grammar codes add a note informing the user about the use or non-use of the passive, e.g.: [D5] verbs are not used in the passive [D6] verbs are not common in the passive Unfortunately, this information is not repeated in the individual codes at (sub-)entries.

3.5.2

The coding system used in the entries

A striking feature of LDOPV is the excellent organization and lay-out of its entries. The headphrases with their abbreviations (e.g. v adv) and all subentries are clearly printed on separate lines. The grammatical codes are not lumped together after each definition and each code is separately illustrated by one or more

example

sentences.

This makes

every (sub-)entry

very

systematic. Exemplification of patterns plays an important part. In spite of obvious

94

restrictions of space, the compiler has made a successful attempt to illustrate all patterns, even those that are related, by means of separate

examples.

Brackets and slashes, which are often used in other dictionaries to combine or compose

similar

information

grammatical

added

codes,

have

been

avoided.

in brackets after the letter/number

The

grammatical

code is also

well-

illustrated . LDOPV does not give collocates

for verbs, as ODCIE 1 does. Cowie and

Mackin's main reason for including collocates in their dictionary was to help the student to create as many contexts as possible for productive use. We have already stated that we regard this as a useful feature of ODCIE 1. LDOPV comes rather close to it with its extensive illustrative material, but the idea of collocability is worked out less explicitly than in ODCIE 1. Compare, for example:

ODCIE 1: cheer

on

[Bli pass]

encourage

(to

further

efforts).

0:

competitor,

athlete. The crowd cheered the runners OB as they star ted on the last lap of the race.

LDOPV: cheer on

v adv...

to encourage

(someone, a team, etc.):

[T1 + ON]

Please

come to the sports meeting to cheer our team on.

The coding

system

of ODCIE 1, as we have seen, consists of only 8 codes

(including the three-way division of [Bl]). Types of complement are not coded but exemplified in collocational lists and example sentences. The codes are distinguished on formal criteria only. This system enabled the compilers to apply

the same coding system

to both idiomatic

and non-idiomatic

verbal

combinations. W e have noted that this is different in the case of LDOPV (and LDOCE). Phrasal verbs, marked by asterisks, are coded as single verbs. Entries or subentries which are not considered to be idiomatic are coded differently, in that only the verb is coded, the adverb and/or preposition (optional or obligatory) always being repeated between the square brackets. The elements of the three Idiomatic c o m b i n a t i o n s consisting of verb + adverb, verb + preposition

and

verb + adverb + preposition

are

only

distinguished by the abbreviations v adv, v prep and v adv prep. The codes for

95

these three types of phrasal verb are based on the same principles. This causes a coding difficulty for transitive phrasal verbs, since the code makes no distinction between the required object of a preposition and the direct object of the phrasal verb containing an adverb. However, the codes do not inform the dictionary user about the impossibility of placing the prepositional object between

the verb and the preposition. What is m o r e , the code is available,

[Tla], but for reasons unknown to the learner it is not used. Compare *cater to and *change round, which behave differently, but have the same code:

*cater to v prep 1...[T1 (simple tenses)] Politicians should learn to cater to the man in the street. *change round v adv l...[Tl] J[_f y o u

change

the

words

round,

the

sentence

sounds better.

As w e have noted in 3.5.1, too absolute a distinction has been made in LDOPV between

idiomatic and non-idiomatic

verbal combinations:

the (idiomatic)

phrasal verbs are marked with an asterisk. Again, we draw attention to two cases which we find problematic. They are the entries for call d o w n on and obtrude on/upon.

*call dona on v adv prep ...[Tl (no pass.)] The priests called

down

punishment

on

the people for their evil ways.

This is probably a case of [Dl] rather than [Tl], but note that [Dl] stands for the sequence verb + object + adverb + preposition + object (e.g. I put his bad temper d o w n to recent illness). In the case of call down on the sequence is: verb + adverb + object + preposition + object. Would

it perhaps have

been

useful to distinguish between [Dla] and [Dlb]?

*obtrude on/upon v prep ...[T1]..^I w i s h

he would

stop obtruding hlmseIf

and

his

opinions upon our peaceful gathering. This also s e e m s to us a clear instance of [Dl], with its two objects, rather than of [Tl]. We have already stated that there are headphrases which are marked as

96

idiomatic in LDOCE, but as non-idiomatic in LDOPV, and the other w a y round, which , of course, leads to a difference in coding. We restrict ourselves to three examples.

LDOCE: centre ...1[L9; X9: esp. on, upon, round, around] to (cause to) gather to centre: Our thoughts centred on the young girl about to be married. 2 [L9; X9: esp. on, upon, round, around] to (cause to) have as a main subject: Attention

centred on

the priest....

LDOPV: *centre on/round/upon... v prep 1 to have (something) as its middle or most important point; surround (something or someone):

[T1 (no pass., usu. simple tenses)] The village

centred on its market square. The chlldren a l w a y s centred round the teacher at story time.

LDOCE: c h l a e In v adv...2 [lCt(witb)] to suit; match: Do the

president's ideas

chime in with yours?

LDOPV: *chime in with v adv prep 1

to suit; match

(something):

[ T1 ( ti£ pass., simple

tenses)]

My

opinion chimes in with yours. ...

LDOCE: nail d o w n v adv [Tl] 1 to fasten d o w n , w i t h a nail or nails: Nal 1 down a loose board in the floor.

LDOPV: nail down v adv 1 to fasten (something) down with a nail or nails: [X9 +

DOWN] The lid

was nailed down, so we couldn't get It off.

It is not only entries with a different grammatical status that have been coded differently In LDOPV and LDOCE. There are also verbal combinations of the same type that have been given different codes. Consider, for example:

97

LDOCE

LDOPV

call down 2 [TX]

call down *3 [Tla]

carp

c a r p at

[...10 (ON,at)]

[L9 + a t

(about)]

cavil [10 (at)]

cavil at [L9 + at]

cede [D1 (to);...]

cede to [X9 +

claaour

2

2 [10 (Cor);...]

to...]

claaour for [L9 + for...]

nip at [T1 (AWAY)]

*nlp at [Tl]

occupy 5 [X9]

occupy In [Tl + In...] occupy with [Tl + with...]

Identical

patterns

specifically

are

concerns

also

the

sometimes

codes

coded

differently

[L9]/ [10] and [X9]/[T1]

in LDOPV. in

This

non-idiomatic

(sub-) entries. For example, cadge from/off is coded [Tl + from]: Can I cadge a c i g a r e t te

from

prepositional

yo u?. T h e

entry

has not

been

coded

[X9],

because

the

p h r a s e is o p t i o n a l . If w e c o m p a r e this w i t h c a j o l e Into and

c a j o l e o u t of, the p r e p o s i t i o n a l p h r a s e s of w h i c h are e q u a l l y o p t i o n a l (cf. LDOCE OUT

+

[Tl + I n t o / o u t of]), w e find that they are coded [X9 + into] and [X9 + of],

respectively.

(Curiously

enough,

the

preposition

out

of

is

c o n s i d e r e d to be a c o m b i n a t i o n of a d v e r b and p r e p o s i t i o n throughout LDOPV.) Another example is cheat into, which is coded

[X9 + Into]: The clever

cheated the old lady into giving him all her money. Trick Into,

salesman

cross-referred

to in the cheat into entry, is of the same pattern, with an identical degree of optionality, but it is coded

[Tl + Into]: I was tricked into giving the old man

some money. Also compare the following examples from LDOPV: o b j e c t to v prep

- d e m u r at...

to o p p o s e ; e x p r e s s o p p o s i t i o n to ( s o m e t h i n g ) or s o m e t h i n g being done: [ 10 + to] I_ object to being b l a m e d for s o m e t h i n g All the local farmers objected

that

haven't done.

to the n e w airport...

demur at/to v prep - object to ... fml

to oppose (something) slightly;

express objection

to (an idea):

[X9

+ at/to (usu. simple tenses)] The committee demurred at the suggestion that they should meet again the next day.Even the government's own party demurred to the new election rules. In the first p l a c e , d e m u r is c l e a r l y i n t r a n s i t i v e (cf. L D O C E [10 (at or to)]) and should be coded [L9]. In the second p l a c e , it is odd that there should be

98

different codes for two such closely related headphrases. The Inconsistencies mentioned in this section can be mainly attributed to the rather arbitrary division between the group of so-called phrasal verbs and the mixed group of non-idiomatic verbal combinations. This has resulted in two different ways of coding headphrases and

sub-entries. The first type of coding

is identical to the coding system of LDOCE, while the second is different In that

It also

attempts

to account

for

the presence

of

the adverb

preposition in the non-idiomatic headphrase. As we have seen,

and/or

this leads to a

confusion of obligatory and optional complements and to differences between the codes in LDOCE and LDOPV. The dictionary user is confronted with different codes for

identical

patterns, merely because one (sub-)entry is idiomatic and the other is not (this distinction Is already established by the asterisk). Our main criticism of the LDOPV coding system is that it does not apply to both Idioms and non-idioms. A simple coding system based on formal criteria only (as in ODCIE 1) which could be applied

to all verbs with adverbs and

prepositions, would have been preferable. It seems to us that the LDOPV system for coding non-idiomatic c o m b i n a t i o n s could also have been Imposed on the idioms. This would have resulted in a single coding system for all verb entries in both LDOPV and LDOCE.

3.5.3

Conclusion

The grammar codes in LDOPV are generally consistent and clear, and each code is separately illustrated. A further advantage of the coding system is that it is restricted

to only six (capital) letters and eight numbers. The system

should

be fairly easy to learn, since most of the letters have a m n e m o n i c value. The numbers always have the same meaning. It is unfortunate

that because of the division that Is made in LDOPV

between idiomatic and non-idiomatic LDOCE

could

not

be

verbal combinations the coding system

indiscriminately

applied

to

the

entries

of LDOPV.

comparison of the two companion dictionaries reveals that there are differences,

mainly due

of A

many

to the often arbitrary decisions w i t h regard

to

idiomatlclty. The syntactic similarity between phrasal and non-phrasal verbs is obscured

by

treating

the

former

as g r a m m a t i c a l

combinations consisting of two or three parts.

units and

the

latter

as

CHAPTER IV

TOWARDS A MORE ADQUATE CODING SYSTEM

The provisional scheme which we present In the following pages Is intended to overcome some of the shortcomings of the current grammatical coding systems, but obviously

it leans heavily on them.

In the first instance we have

restricted ourselves to the lexical verb and its (obligatory) complementation, but we have also included examples of how the system might be extended to cover grammatical information about nouns and adjectives. He recognize that there may be practical obstacles to the Implementation of the proposal, and that lexicographical traditions are difficult to change. We do hope, however, that our work will stimulate discussion about the issues involved.

4.1

Verbs

In Chapter I we have briefly listed the six assumptions on which our proposal is based. We regard these points as conditions that must be fulfilled if the inclusion of grammar in learners' dictionaries is to be successful. Let us consider these basic requirements in some more detail again, before we go on to our classification and our coding system of English verbs. Our assumptions are the following: a) The grammatical codes should be unambiguous and easy to use. They should provide the dictionary user with useful and straightforward information about both the grammatical function (e.g. do, 10) and the grammatical category (e.g. NP, AdjP, PP) of the verbal complementation. b) The codes should be self-explanatory and be added to every verb in the dictionary. As for the choice of terminology, we recommend that of the modern comprehensive grammars of English. Most first-year students learning English will be familiar with that terminology after their first few weeks of syntax

100

classes. However, for the absolute beginner and the general user the dictionary will have to provide an introduction to the grammatical coding system. c) A full inventory of grammatical codes should be given at the entries or subentries themselves. The g r a m m a r should not be presented in such a condensed form that frequent reference to explanatory notes elsewhere in the dictionary is required. Of course, printing the codes in the entries themselves will take up

a

great

deal

of

space,

more

than

is now

customary

in

the

learners'

dictionaries. However, if it is felt that grammatical information is a useful feature of such dictionaries, one should also be prepared

to give it the space

it requires. d) Example sentences play a very important part in learners' dictionaries. They do not only function as illustrations of the meaning of the headword, but also as exemplifications of its syntactic behaviour. Therefore

the examples should

consistently illustrate and correspond to each of the codes in the entry or sub-entry. Such an approach will make repetition of codes before each example sentence superfluous. e) Transformationally

related

structures should

also be indicated

as such in

the entries and sub-entries. If verbs allow transformations like passivlzation, indirect object m o v e m e n t and extrapostion, or if there are restrictions on particular transformational possibilities, this should be shown in the codes. f) The coded grammatical information should be as explicit and complete as possible. In order to avoid confusion no parts of the codes should be collapsed or bracketed. It should be noted that the proposed scheme is intended to be a pedagoglca 1 approach to the problem of verb complementation. A purely syntactic analysis or classification

would

look

rather

different.

The

system

is

based

on

the

assumption that it is possible to classify English Verb Phrases (VPs) in terms of the (obligatory) complementation (e.g. He put the money in

that verbs take. For example, the verb put

box) requires an NP as w e l l as a PP or AdvP as its

complementation; grow (as in John grew old) requires an AdjP; live (as in He lives

in Paris) requires a PP

or AdvP,

etc.

These

verbs

are

said

to

be

"subcategorized" as taking, say, NP + PP/AdvP, or AdjP, PP/AdvP, etc. All verbs can, in principle, be classified in this way. We distinguish nine main classes of verbs. Let us c o m m e n t briefly on the scheme presented below. In all basic VPstructures the lexical verb (V) is the Head, followed by different grammatical categories complementing V. Class 1 is the group of verbs, traditionally called intransitive verbs,

101

which take "zero" complementation (no complementation).^^ One might want to distinguish here between John snores and John drinks by noting that in the latter example some NP has been suppressed, whereas in the former there is truly "zero". In Class

we find grouped together the verbs traditionally known as

copulas. Some of these verbs require an NP, an AdjP, a PP, or a CI (for clause). The label CI is used here as a cover term for the various types of finite and non-finite clauses (e.g. to-lnf, inf., -ing part, -ed part, NPtoinf, etc.). Copula be, for example, may take any of these four complements, as is shown by the examples below. Other copulas allow only NP or AdjP (e.g. turn), or

just AdjP (grow,

get). The grammatical

function of all these

categories is that of Predicative complement (Pc)/^ The verbs in Class 3 are intransitive verbs, not copulas; they usually (but not exclusively) denote Location. The GF of the AdvP (Adverb Phrase) of the PP is Loc(ative adverbial complementation). An example of this structure not expressing Loc is The match is tomorrow. Class 4^ is the rather mixed group of English verbs which require an adverbial complement (Ac), also called an 'obligatory adverbial', realized by an NP, an AdvP, a PP or a CI. These verbs are also called intransitive. Class _5 represents verbs that are followed by an NP or an AdjP, which acts as an Ac and a Pc at the same time;

these verbs may be called

complex

transitive. Class 6 covers the so-called monotransitive verbs. All the strings that follow the V in these examples (both NPs and clauses) may be regarded as the direct object (do). The subordinate clause may be finite or non-finite, and if non-finite they may be infinitive clauses, -ing clauses or -ed clauses. Finite clauses always have an explicit subject of their own; non-finite clauses may or may not have an explicit subject. NPto-lnf, for example, means that the object clause contains a to-infinitive and an explicit subject. Class 7 is the group of ditransitive verbs like give, offer, ask, etc, verbs that require NP + NP (iO and do respectively). Class

are the so-called complex (mono)transitive verbs. These verbs are

(mono) trans it ive, and the AdjP, NP or PP following the object NP acts both as an Ac, qualifying V, and as a Pc, qualifying dO/NP. Compare Class 5. Class 9^, finally, consists of verbs that take an object NP and a PP or AdvP which functions as Ac. In a sense, Class 9 corresponds to the Classes 3 and 4 together, which may be another reason for collapsing 3 and 4 (and 2).(3)

102

Basic VP-structures l.V 0 UT

Verb types Intransitive

2.V

Copula

(A)

Examples John snores John drinks a fool John is foolish in a good mood John is to be envied (to-inf.pass.) John grew old The milk turned sour They got rich John becaae an actor

3.V

4.V

/AdvP Loci PP PP I CI Ac/ AdvP

(A)

(A)

NP

Loc Intransitive

Ac Intransitive

,outside i John is \ l n t h e g a r d e n i John lives in Paris

John belongs to a left-wing party. John condescendedit0 bribery. (to help us' He lived to be 90 This car drives smoothly It weighs 10 pounds It lasted (for) three hours

5.V < Ac - Pc, jAdjpj (A)

fl

6.V John likes } s i n ging >

(to-inf) (-ing)

John hopes that you will help him (fin) (leave i John saw Susan heaving* John kept me waiting John believes the dog to be hungry John found the money stolen from the box 7.V

iO/NP + dO/NP (A)

Ditransitive

John gave Susan the book

(NPinf) (NP-ing) (NP-ing) (NPto-inf) (NP-ed)

103

8. V

I AdjPj She painted the door)8 reen dO/NP +{Ac-Pcy NP j(A) Complex (all colours' Monotransitive John saw Susan drunk He flung the door open a fool

)

!

9.V

dO/NP +

H2>

foolish > in a good mood) They appointed him chairman They christened the baby John They beat the prisoner unconscious Ac in a b o x John put the money j t1 1 Monotransitive there John's firm posted him to Singapore John reminds me of his father John worded the letter carefully

Let us now look at some concrete verbs, and consider how the grammatical information about VP-structures, i.e. the verb with its complementation, may be represented

in the dictionary

at each entry or sub-entry.

Consider

following examples which are largely based on the definitions,

the

etc. in

LDOCE:^ 4 ) bring /.../ v brought /.../ 1 [dO/NP; iO/NP + dO/NP objm-to or for; dO/NP + Ac/AdvP,PP] to come with or lead: bring an answer; bring me the book/bring the book to me - The soldier's brave deeds brought him honour and glory John brought Mary some flowers/brought some flowers for Mary; Can you bring your sister back home? - Bring your friend to the party - The prisoner was brought before the judge - The beauty of the music brought tears to her eyes. — s e e USAGE 2 [dO/NP,NPto-inf] to cause or lead to: Spring rains bring summer flowers - What brought you to do It? - He could never bring himself to kill an animal or bird. 3 [dO/NP; iO/NP + dO/NP] to sell or be sold for: This old car will bring about 10 pounds - What/How much would £ new car bring?; The pictures he sells bring him 12,000 pounds £ year. 4 [dO/NP of ten + PP]

(against) law to make officially: The policeman

brought £ charge against the fast driver - The neighbours brought a complaint against the noisy family. 5 [ dO/NP ,NP-lng] to cause to come: One sad letter from his wife brought

104

many offers of help/brought him crying home - A phone call brought him hurrying to Leeds. 6 [dO/NP + Ac/AdvP,PP] to cause to reach a certain state: to bring someone low (In defeat) - bring them ln/out/back/together - Jto bring someone into being/Into action/to an end/to his knees (In defeat).

decelerate /.../ v [0; dO/NP] to (cause to) go slower: Before you decelerate you should check In your rear-view mirror; We decelerated the engine long before we came to a stop —

compare ACCELERATE (1)

decide /.../ v 1 [dO/NP,to-lnf,fIn] to arrive at an answer or way out that ends uncertainty or disagreement about: to decide the question - to decide where to go/where they should go. 2 [dO/to-lnf ,f in] to come to hold or declare a stated belief: She decided to go/(that) she should go. 3 [dO/NP] to bring to a clear or certain end: One blow decided the fight. 4 [dO/NP,NPto-inf] to cause to make a choice: Your words have decided me What decided you to give up your job? 5 [0] to make a choice or judgment: I've been waiting all day for them to decide. 6 [Ac/PP] rvjon,for,against,in favour of to make a choice or judgment: I've decided on

new car/on going there - They had decided

for an

expensive education for their children - We've decided against having £ big family gathering this Christmas - They decided in favour of him and against me. declare /.../ v 1 [dO/NP; dO/NP + Ac-Pc/NP,-ed] to make known publicly or officially, according to rules, custom, etc.: Our government has tonight declared war (on Ruritanla); Jones was declared the winner of the fight declare Alvln B. Schlff elected. 2 [dO/NP,fin,NPto-inf; dO/NP + Ac-Pc/NP,AdjP,PP] to state (or show) with great force that there is no doubt about the meaning: He declared his position - He declared (that) he was right -

He declared himself to be a_

member of their party; His actions declared him an honest man - He declared himself Innocent - They declared themselves for/against the plan. 3 [dO/NP] to make a full statement of (property for which money may be owed to the government): Have you anything to declare?

105

4

[tf; d O / N P + A c - P c / - e d ] (of the c a p t a i n of a c r i c k e t t e a m ) to end the

team's I N N I N G S b e f o r e all its m e m b e r s have been put out: Do you think Brearly will declare before tea?; declare an innings closed. 5 [0; d O / N P ] (in the g a m e of B R I D G E 3 ) to say w h i c h type of card w i l l be played

as

(TRUMPS):

He

declared

after

a

three

minutes'

pause;

John

declared trumps. 6

[Ac/PP] P O against/for to state one's position

to/support for (someone,

something, of doing something): He declared against my proposal.

deem /.../ v [ d o / f i n , N P t o - i n f ; dO/NP + A c / P c / A d j P , N P ] f m l

to c o n s i d e r ; have

the opinion: He d e e m e d that it w a s h i s duty to h e l p - Do you d e e m p l a n to be s e n s i b l e ; _I don't d e e m

this

it s e n s i b l e to leave right n o w - He

would deem it an honour if the minister came to see him.

d e f e n d /.../ v 1 [dO/NP o f t e n + P P ] f\J ( f r o m / a g a i n s t ) to keep safe f r o m h a r m ; p r o t e c t a g a i n s t attack: W h e n the dog a t t a c k e d m e ,

defended myself -

d e f e n d s o m e o n e f r o m h a r m - The fort cannot be d e f e n d e d a g a i n s t an air attack. 2

[0;

dO/NP]

to play at (a p o s i t i o n ) so as

to k e e p

an o p p o n e n t

from

advancing, making points, or winning: There was no point in attacking,

we

just had to defend; They defended their side of the field very well. 3 [dO/NP,-ing] to support, protect, or show the rightness of, by argument: defend £ claim - How can you defend killing someone like that? 4 [dO/NP] to act as a lawyer for (the person who has been charged): defend the accused. 5

[