Dictionaries and spoken language: : a corpus-based review of French dictionaries.

DICTIONARIES AND SPOKENLANGUAGE: A CORPUS-BASEDREVIEW OF FRENCH DICTIONARIES

Dirk Siepmann: School of Languages, Literatures and Cultures, University of Osnabrˇ[email protected]

Abstract

Starting from the observation that traditional lexicography has tended to rely on

corpora of written text, the present article argues that this might be to the detriment

of covering the commonest colloquial lexical units which carry the main burden

of everyday conversation. Using a new reference corpus of French (Corpus de

reference du francais contemporain or CRFC), it presents a number of case studies

of highly common informal words and expressions, each of which sets out with a

corpus-based dictionary entry and then goes on to compare this with the treatment

accorded the entry word or phrase in ten major monolingual and bilingual diction-

aries. The general findings are that colloquial words, far from being stylistically ‘in-

ferior’ substitutes of more formal words, are imbued with their own specific shades of

meaning, phraseology, and pragmatics, and that medium-sized spoken corpora like

the CRFC shed light on lexical patterns and collocations about which the dictionaries

under survey and large written corpora are largely uninformative. This leads to the

conclusion that there may well be a second corpus revolution ahead which will apply

Sinclair’s famous dictum that ‘the language looks rather different when you look at a

lot of it at once’ to the investigation and documentation of intimate and colloquial

language use.

1. Introduction

Lexicography, and more especially French lexicography, has traditionally

relied on models of ‘good usage’ (le bon usage), and there seems to have

been a tacit understanding that, as Hanks (2012a: 416) puts it, ‘a corpus of

carefully written, edited, and published texts’ is a better source of data than a

corpus containing spoken data, which may show ‘traces of speakers struggling

towards successful encoding of their thoughts and interactions, relying on the

conventions of their language, but it does not provide very good evidence for

International Journal of Lexicographydoi:10.1093/ijl/ecv006 1

# 2015 Oxford University Press. All rights reserved. For permissions,please email: [email protected]

International Journal of Lexicography Advance Access published March 29, 2015 at U

niversitÃ¤tsbibliothek O

snabrÃ¼

ck on April 19, 2015

http://ijl.oxfordjournals.org/D

ownloaded from

http://ijl.oxfordjournals.org/

the conventions themselves’. The present article takes issue with such claims,

arguing that lexicography should provide a faithful record of all uses of a

language. It will be demonstrated that large spoken corpora offer an entirely

new perspective on the common uses of common words, a perspective which is

still largely absent from dictionaries although it is of vital importance to vari-

ous groups of dictionary users, such as translators involved in dubbing films

and non-native speakers learning a language.

There are several reasons why lexicography must take account of colloquial

speech. Firstly, though perhaps least importantly, the lexical aspect of the

struggle for words that Hanks mentions may be interesting in itself, especially

for foreign users of a language; a simple example is French speakers’ use of

the phrase ‘je vais y arriver’ when they have difficulty pronouncing a word, a

use that may profitably be taken over into the non-native speaker’s repertoire.

Secondly, it would be surprising to find that colloquial language, as a variety

in its own right, employs the same word combinations with the same frequen-

cies as other varieties, and that spoken usage can be ascertained from written

texts. Just as the investigation of academic text tells us nothing or very little

about newspaper language, so ‘carefully written, edited and published texts’

in general are unlikely to contain much evidence of the real nature or fre-

quencies of particular conversational uses of language. While fiction, and

fictional dialogue in particular, has often been used as a source attesting

the existence of particular colloquialisms and regionalisms, it cannot be

relied on for evidence of their natural lexical environment. Thirdly, corpus

linguistics has taught us that lexicographers’ linguistic intuitions may be de-

ficient or downright untrustworthy, and it is difficult to see why this should

not apply to spontaneous spoken language. There may indeed be a second

corpus revolution ahead, which will alert us to the distinctive lexical proper-

ties of spoken language.

The present article uses the French of France as a test case for the above

argument. In France, the establishment in the 17th century of classical French

as the norm governing formal language use has since led to a diglossic situ-

ation, with spoken French and written French, or rather the high and low

forms of French, developing into starkly divergent varieties (cf. Koch and

Osterreicher 2011: 145–153). One of the most striking characteristics of ordi-

nary French conversation in all walks of life is the frequent use of low-register

lexis which dictionaries tend to label as familier, populaire, vulgaire or argo-

tique. Strictly speaking, the latter term applies to the lexis used by a particular

social or occupational group, but such lexis has often spilled over into general

usage. Similarly, lexical items which originated in the French suburbs are now

often adopted by young people throughout France (e.g. kiffer, pecho). This

raises the question of the extent to which such items are adequately represented

in dictionaries.

2 of 30 Dirk Siepmann

at UniversitÃ

¤tsbibliothek OsnabrÃ

¼ck on A

pril 19, 2015http://ijl.oxfordjournals.org/

Dow

nloaded from


Authoritative reference works of French claim to portray the high and low

varieties with the same degree of fidelity. As Alain Rey writes of the Petit

Robert in 2006,

Il balaie un spectre tres large d’usages du francais allant de la pensee

abstraite et des techniques contemporaines a l’expression spontanee des

usages langagiers de cette France qu’on dit « d’en bas », alors qu’elle est de

partout et de tous. (PR, xxiv).

In a similar vein, dictionaries for learners of French, such as DDF, make

claims like the following:

Il arrive souvent qu’un synonyme familier soit beaucoup plus employe que

le mot neutre (P.-V. est plus courant que contravention, flic plus courant

que agent ou policier), et on en a tenu compte. (DDF, ix)

However, as the present article will demonstrate in some detail, the mere

inclusion of words typical of informal speech does not necessarily mean that

these receive adequate treatment. It is assumed here that there are three main

reasons for this. One is that the corpora used in compiling the standard French

monolingual dictionaries are heavily biased towards the language of ‘highbrow’

literature, as in the case of the TLF (‘deux siecles de production litteraire

francaise’ [‘two centuries of French literary production’]; Pierrel, Dendien

and Bernard 2004: 166). A second reason is that some dictionaries, like the

Petit Robert, still rely on manual analysis and intuition rather than electronic

corpora:

A la documentation, nous lisons beaucoup : romans, essais, poesie,

autobiographies, bandes dessinees. . . Nous y relevons des mots, nouveaux

ou pas, des expressions modernes sous la plume d’auteurs recents. (‘In the

documentation department, we read a lot : novels, essays, poetry,

autobiographies, comic books and comic strips. . . From these we glean

words, both new and old, as well as modern expressions from the pen

of recent authors’; http://www.lepetitrobert.fr/le-petit-robert/comment-les-

mots-entrent-ils-dans-le-dictionnaire)

Thirdly, even where corpora are used, it is a reasonable assumption that the

quality of dictionary entries is significantly affected by the size of these corpora

and the diversity of genres found in them. While the publishers of the bilingual

dictionaries examined in this study claim to use electronic corpora in the com-

pilation process, there is no clear information on the make-up of such corpora.

The present article suggests that the use of genre-diverse corpora containing

a large proportion of spoken language leads to considerable improvements in

Dictionaries and spoken language 3 of 30

at UniversitÃ


¼ck on A


Dow

nloaded from

http://www.lepetitrobert.fr/le-petit-robert/comment-les-mots-entrent-ils-dans-le-dictionnaire

http://www.lepetitrobert.fr/le-petit-robert/comment-les-mots-entrent-ils-dans-le-dictionnaire


the treatment of everyday language use. Evidence from the new Corpus de

reference du francais contemporain (hitherto abbreviated as ‘CRFC’) will be

marshalled to show that heavy-duty low-register words such as mec and

peter are not mere substitutes of more formal lexis, although their treatment

in most dictionaries tends to suggest just that. The items in question have

developed their own specific senses, complementation patterns, collocations

and phraseologies but, as will be seen, there is an almost complete absence

of relevant information in most of the dictionaries under investigation.

Particular attention will be given to the extent to which dictionaries meet the

decoding and encoding needs of foreign learners of French.

Section 2 discusses the composition of the CRFC and outlines the method-

ology which informs this study. Section 3 presents a number of case studies,

each of which sets out with a corpus-driven description of the lexical ecology of

a word or phrase and then goes on to compare this with the treatment accorded

the item in ten major dictionaries. Section 4 summarizes and discusses the

results of this investigation.

2. Corpus andmethodology

Research into spoken French has so far had to rely on modest-sized corpora,

such as those compiled by the GARS (Groupe Aixois de Recherche en Syntaxe)

and DELIC teams at the University of Provence, which currently go under the

name of Corp-Aix-2. Corp-Aix-2 comprises around 1.7 million words of tran-

scribed interviews, of which only 300,000 have been made available to other

researchers on DVD (Cresti and Moneglia 2005). The combined size of all

spoken corpora of French (see Cappeau and Seijido 2005 for a detailed inven-

tory) can be estimated at a maximum of 10 million words. Thus, theoretically

at least, there is enough data to build a spoken corpus that would at least be

equal in size to the spoken section of the British National Corpus (BNC), but

the heterogeneous nature of the data collected and their restricted accessibility

has prevented a pooling of resources and research efforts (Debaisieux 2010).

Analysis of the existing corpora has yielded valuable insights into the syntax

and phonology of spoken French (see, for example, Blanche-Benveniste 1991;

1997; 2010 and the journal Recherches sur le francais parle). However, their

small size has meant the inability to undertake work on lexical or lexico-gram-

matical features of present-day spoken French. It is well established that small

corpora such as ESLO and Corp-Aix-2 will contain a large proportion of

hapaxes and are therefore unsuitable for research into medium- or low-fre-

quency words and multi-word units (see, for example, Carroll, Davies and

Richman 1971).

This situation has now changed, with the advent of a new genre-diverse

megacorpus of French, the corpus de reference du francais contemporain

(hereafter abbreviated as CRFC; for detailed information on the corpus,


at UniversitÃ


¼ck on A


Dow

nloaded from


see Siepmann and Burgel 2015). The first version of the corpus totals 310

million words of the French of France from 1945 to 2014, with most texts

coming from the last two decades. The ‘spoken informal’ section of the

CRFC was designed to be around six times the size of the largest single

spoken corpus of any language (25m words [Bank of English]; the Corpus of

Contemporary American English, which comprises 90 million words, is made up

mainly of news programmes). Its spoken and ‘pseudo-spoken’ sections com-

prise a total of 155 million words of mainly informal French, or twenty times

the amount of informal data hitherto available. (see Table 1).

Most noteworthy is the large size of the spoken informal section, an

area where previous research had to rely on small, interview-based corpora

(cf. Gadet et al. 2012). 75 per cent of the spoken informal section is made up

of transcripts of unscripted dialogue or monologue from more than 200 dif-

ferent types of television programs broadcast by France 2, France 3 and France

5 in 2013 and 2014, totalling more than 6000 programs and more than 3000

hours of naturally occurring speech. An entire year’s worth of programs was

incorporated to avoid biasing the content of the corpus in favour of certain

days of the week or times of the year (cf. Kennedy 1998: 75). As Meißner (2006:

248–249) has argued in an attempt to answer the question ‘quel francais

enseigner?’ (‘what kind of French should we teach?’), television has long

since set the statistical norm by bringing numerous idiolects and linguistic

Table 1: Composition of the CRFC

Medium Section Size

Spoken formal 30m

informal 30m

pseudo-spoken stage plays and film scripts 30m

film and daily soap subtitles 2,5m

text messages/chat 2,5m

discussion forums 60m

155m

Written academic 30m

non-academic books 30m

prose fiction 30m

newspapers 45m

magazines 10m

diaries and blogs 5m

letters and e-mails 1m

miscellaneous 4m

155m


at UniversitÃ


¼ck on A


Dow

nloaded from


varieties to the eyes and ears of the masses, while at the same time presenting

them in such a way that they can be understood by the vast majority of viewers

and listeners within the language and broadcasting area. As a result, corpus

compilers can use the range of linguistic varieties spoken in the media as a

guide for considerations regarding the composition of spoken corpora and the

development of teaching materials.

This leads to huge differences in coverage between the CRFC and previous

corpora, enabling researchers to explore informal French at a level of delicacy

hitherto unknown. One of the most obvious uses to which such a corpus may

be put is the critical scrutiny of current lexicographic practice concerning

spoken French.

The methodology used for this purpose proceeded in seven steps. The first

step involved generating several frequency lists, viz. a) a frequency list of the

3000 most common lemmas in the spoken portion of the CRFC and b) three

lists of the 3000 most common 3-, 4- and 5-grams; word frequencies and multi-

word strings were identified using the relevant functions of the Sketchengine. In

the second step, lemmas labelled familier, populaire or vulgaire in the Petit

Robert were extracted from frequency list A, and lexical bundles that did not

constitute phrasemes were removed from frequency list B. In the third step, a

random selection was made of four items from the first list and three items

from the second list; the word limit of this journal precludes the treatment of a

larger number of items. In the fourth step, the frequency data obtained for

individual lemmas were compared with native-speaker usage ratings obtained

from LWUF; words were to be included only if assigned the highest or the

second highest usage rating (‘En parlant, j’utilise le mot ou la structure dans la

signification indiquee carrement tous les jours/de temps en temps.’ [LWUF: 6]),

which turned out to be the case for all seven lemmas selected. The resultant lists

were as follows:

Frequency list A (lemmas): mec (‘guy’, ‘mate’), lacher (‘let go of’, ‘drop’),

look (‘look’), peter (‘fart’, ‘snap’, ‘break’)

Frequency list B (multi-word items): autour de (‘around’), n’importe quoi

(‘anything’, ‘rubbish’), c’est bon (‘that’s ok/fine/good/etc.’)

The fifth step involved a detailed analysis of the lexico-grammatical and

pragmatic features of the selected items on the basis of the spoken section of

the CRFC, following the corpus-driven approach to habitual co-occurrences of

words (‘usuelle Wortverbindungen’) developed at the Institut fur Deutsche

Sprache (Steyer and Brunner 2009, Steyer 2013). This approach is based on

three methodological premises which draw inspiration from the British tra-

dition of text analysis established by Firth and Sinclair, viz. a) it derives struc-

ture from the data during the analysis rather than in advance; b) it foregrounds

language as use; and c) it lets the data speak for itself, allowing the observer to


at UniversitÃ


¼ck on A


Dow

nloaded from


form an unbiased picture of authentic language in use. The data are listed in

terms of node words and their primary and secondary collocates (e.g. never

occurred to me in the case of the English node idea) and are then subjected to

thorough scrutiny with a view to determining the internal structure and typical

variation found with node-collocate pairs and establishing the presence or

otherwise of node-collocate pairs with similar characteristics. This fifth step

thus comprises two sub-stages (cf. also Hanks 2013: 92): the first involves

grouping the evidence into recurrent semantic-pragmatic patterns; the second

is the assignment of meaning to each pattern.

There is widespread recognition that such analysis cannot always rely on a

pre-established set of categories but sometimes needs to describe the observed

phenomena by means of lexicographic commentaries. In the case of a noun like

idea, for example, newly created categories might include syntactic types such

as [PAST_OR_PRESENT_PERFECT_TENSES](occur, come to my mind,

spring to mind, etc.): idea came to mind, idea occurred to me] or semantic

types such as [NEGATION_REJECTION: would never have occurred, not

the best idea, etc.].

In the sixth step, the findings of the corpus analysis were cast in the mould of

a dictionary entry in an alphabetical dictionary. The seventh and final step

involved determining to what extent the meanings as well as the collocational

and phraseological patterns thrown up by the analysis were represented in

current dictionaries. Ten major dictionaries were selected: one unabridged

monolingual dictionary (the Tresor de la langue francaise [TLF]), one large

monolingual desk dictionary (the Petit Robert [PR]), two monolingual learn-

ers’ dictionaries (the Dictionnaire du francais [DDF] and the Dictionnaire

d’apprentissage du francais langue etrangere ou seconde [DAFLES]), two un-

abridged bilingual French-English dictionaries (Harraps [HS] and Collins), two

unabridged bilingual French-German dictionaries (PONS Großworterbuch

[PONS] and Langenscheidt Handworterbuch [LH]) and two bilingual slang

dictionaries (Harraps [HS Slang] for the language pair French-English and

Langenscheidts Worterbuch der franzosischen Umgangssprache [LHF] for the

language pair French-German).

3. Results

As explained above, the word limit in this journal precludes an exhaustive

presentation of the results obtained for a large number of lemmas. It was

therefore decided to single out for detailed discussion two verbs (peter,

lacher), two nouns (mec, look) and three multi-word strings (autour de,

n’importe quoi, c’est bon). The discussion of these items will follow a standard

format. The data is presented in the form of an ‘ideal’ dictionary entry based

on the data analysis described in Section 2; entries comprise sense divisions,

including information on collocation and phraseology where applicable.


at UniversitÃ


¼ck on A


Dow

nloaded from


This receives some comment, followed by a tabular presentation of the extent

to which individual dictionaries contain the information thrown up by the

analysis and a discussion thereof.

3.1 PE¤ TER (v)

I. Intr. 1. (= lacher des vents) : qqn pete 7% des peteurs s’amusent a peter

dans l’eau du bain ; � au lit Collocations � bruyamment Phraseologie � de

travers

2. (= eclater ; casser ; ne plus fonctionner) : qqc pete Si vous avez un petit

impact, la vitre peut peter. Collocations+N : obus, grenade; canalisation,

tuyau, ballon, bulle; bouton, braguette, capote, corde, elastique, fil, col-

lant; vitre, soupape, ordi, assiette, paire de souliers, balance; systeme ; +

ADV: net

Pronom. qqc se pete Collocations+N : sac, verre, lampe, appareil ; +

ADV: en deux

(a) (=detruire) faire peter qqc . . . ca m’a inquiete que ca fasse peter un

truc, mais les infirmieres m’ont dit aucun risque ; tout faire peter

Collocations batiment ; voiture ; planete

(b) (=obtenir un maximum, un record) faire peter qqc Collocations

l’audimat, la jauge, le standard, le score, le compteur, un palier, les watts

(c) faire peter Collocations le champagne, l’apero, la bouteille, le bouchon

(d) faire peter (= faire exploser) Collocations bombe, petard

3. (= se disputer) ça pete On est une famille de sanguins, ca pete souvent.

Phraseologie ça ne va peter bien loin

4. (= reussir ; captiver) ça pete Un TF1 avec Ruquier, je suis sur que ca

peterait plus que la formule actuelle.

II. Tr. 1. (=exploser ; casser [plus rare que -> faire peter]) qqn pete qqc

Collocations vase, ballon, porte, lampe, ficelle, . . . ; capote ; jambe, nez,

clavicule ; varice

2. qqn pete qqc a qqn Degage, ou je te pete un bras.

Phraseologie peter les plombs/un plomb/un cable/un fusible/une durite/un

boulon/les boulons; peter plus haut que son cul/derriere; peter la gueule/

tronche a qqn; (se/s’en) faire peter la panse/le bide/la sous-ventriere; peter le

feu/la forme/la sante/des flammes; peter les couilles (=1. enerver 2. suivre de

trop pres), peter le litron ; peter la dalle/la peter; etre pete de thunes; peter dans

la soie; se la peter (avec qqc {une grosse bagnole}); [surtout Quebec] se peter les

bretelles (avec qqc) ; peter a la figure/gueule/au nez (a qqn); envoyer qqc/qqn

peter; se faire peter (=descendre); **peter le cul/la rondelle (etc.) a qqn

(=sodomiser); pouvoir entendre une mouche peter ; n’avoir rien a peter de

qqc; (faire) peter la baraque ; ne plus se sentir peter ; se peter la gueule (=1.

tomber ; 2. s’enivrer) ; se faire peter le caisson


at UniversitÃ


¼ck on A


Dow

nloaded from


Pronomin. qqn se pete qqc Je me suis pete le coccyx en faisant du surf.

Collocations cheville, poignet, [col du] femur, rein, dos, tympan,. . . ; cœur

pete (= casse; ivre; sous l’influence de drogues) Collocations completement �,

definitivement �

A thorough examination of peter shows that the verb is rarely used in the

sense that most native speakers and most dictionaries (e.g. PR) would identify

as its most common or ‘literal’ sense (‘lacher des vents’). In fact, this use ac-

counts for less than 5 per cent of all occurrences in the CRFC. Although

learners’ dictionaries have recently prioritized frequency as a measure of the

importance to learners of particular words, there is still an equally strong case

for listing first those senses which are acquired by native speakers in the earliest

stages of language acquisition. In the case of peter, this seems to be the sense

‘lacher des vents’, as suggested by the entry in Duhamel and Balaz (1993: 502),

which shows child respondents giving this sense first:

C’est un bruit qui fait “prout”.

On pete de temps en temps quand on fait caca : ca sent mauvais.

On peut aussi peter un ballon ou une lumiere. (italics in the original)

Thus, even though the word is not used frequently in this sense by most

adults (except in child-directed speech), it is probably this meaning that is the

most ‘deeply anchored’ in their brains and hence the most readily ‘available’

(cf. the distinction made in studies on core vocabulary between frequency and

availability [‘disponibilite’]).

Around 50 per cent of the occurrences of peter in CRFC (spoken) are within

a comparatively small number of fixed expressions, many of which are syn-

onymous (e.g. peter la forme/le feu/la sante). Six of these, which have been

found to be particularly frequent, have been shaded grey. Two expressions

can take prepositional complements (se la peter/se peter les bretelles avec

qqc), where the prepositional complement indicates the cause of sb’s pride.

Peter appears in five common patterns of use, four of which are intransitive

and one is transitive. In the cognitively most central intransitive use already

discussed, the subject is human or animal. In the second pattern, the verb is

also intransitive and has two major sub-senses which are not usually distin-

guished in dictionaries. These may be glossed as ‘eclater’ or ‘casser’ (with sub-

jects such as ‘obus’ or ‘capote’) and ‘ne plus fonctionner/etre casse’ (with

subjects such as ‘ordi(nateur)’ or ‘balance’), respectively. A similar meaning,

albeit typically with different subject nouns (e.g. verre, sac), is encoded by the

reflexive variant se peter. Most commonly, however, the intransitive verb is

transitivized by the addition of delexical faire, giving rise to four distinct sub-

senses.


at UniversitÃ


¼ck on A


Dow

nloaded from


Impersonal constructions involving ca are a particular source of difficulty for

the foreign learner (cf. van Peteghem 1994) and, despite the continuing linguis-

tic debate on how to classify them, should therefore be listed separately, espe-

cially if, as here, they encode specific meanings. Thus, there is clearly no

personal construction that corresponds to impersonal constructions illustrated

by the following sentence:

Un TF1 avec Ruquier, je suis sur que ca peterait plus que la formule

actuelle. (?Un TF1 avec Ruquier peterait plus que la formule actuelle)

Even sense 3 only bears a vague metonymous relationship to senses 1 and 2

in that situations where the impersonal construction is used will sometimes

involve the breaking of objects.

As can be seen from Table 2, there is a clear difference between bilingual and

monolingual dictionaries in that monolingual dictionaries tend to give undue

priority to phrasemes based around peter, while bilingual dictionaries provide

more useful collocational information. This poses a serious problem for users

wishing to use peter productively, since they will turn to the monolingual dic-

tionary rather than the French side of the bilingual dictionary. With the ex-

ception of the two Harraps dictionaries, the bilingual dictionaries under

scrutiny appear to have borrowed information on phraseology from the

large monolingual dictionaries or from other works on French argot, since

none of them record any phrasemes that are not also found in the latter.

This is particularly evident in the case of LWUF, which lists rare expressions

such as faire peter le conometre or il faut que ca pete ou que ca dise pourquoi,

which is now usually abbreviated to il faut que ca pete.

Table 2: peter in ten French and French-English dictionaries

I.1 I.2 I.2Pron I.2a I.2b I.2c I.2d I.3 I.4 II.1 II.2 II.2

Pron

ADV+V Phr

(out

of 23)

PR + + - - - - - - - + - - - 9

TLF + + - - - - - - - + - -

DDF + + - - - - - - - - - + 7

DAFLES - - - - - - - - - - - - - -

HS Slang + + + - - - - - - + - - se peter

en deux

12

LWUF - - - - - - - - - + - - - 7

Collins + + - - - - - - - + - 10

HS + + + + - - + - - + - 16

LHF + + - - - - - - - + - 4

PONS + + - - - - - - - + - peter net 1


at UniversitÃ


¼ck on A


Dow

nloaded from


Most of the dictionaries, among them TLF and PR, miss out the reflexive

verb se peter. Given the absence of transitivized faire peter from TLR and PR

and their choice of examples, both wrongly suggest that the transitive use is

more common. The majority of examples in TLR and PR are outdated, al-

though some convey the illusion of living speech:

L’affaire va vous peter dans la main. (PR, s.v. peter ; italics in the original)

This expression, which means ‘come to nothing’ is now on its way out, while

peter a la figure/au nez (etc.) are far more common.

Hardly any provision is made for verb-adverb collocations, with two minor

exceptions (see Table 1).

Of the two learner’s dictionaries, only DDF has an entry for peter.

Remarkably, apart from se la peter, it records those idioms which were found

to be most frequent in the above corpus analysis. It also goes a good deal further

than the two unabridged native-speaker dictionaries in providing examples fea-

turing collocations that the foreign learner may confidently extrapolate from

(e.g. il s’est pete la jambe; note, however, that it would be preferable to draw

sharper lines between collocations and examples, cf. Siepmann 2007). On the flip

side, these collocations are often embedded in made-up examples that lack some

naturalness feature or other. Consider the following sentence:

Il est dans un etat de surexcitation, il va peter les plombs ! (DDF, s.v. peter)

There is an obvious clash here between colloquial peter and the rather formal

dans un etat de surexcitation. Here is a natural-sounding example from the

CRFC:

En fait, on peut dire que j’ai pete les plombs mais je n’ai pas insulte

l’arbitre, loin de la.

The two Harraps dictionaries provide the most extensive treatment of peter,

including reflexive uses and regionalisms. The unabridged Harraps also records

two sub-senses of transitivized faire peter and a wide range of common phrase-

ology. This is an area in which the large French-German dictionaries fail abys-

mally to meet the decoding needs of their users, offering only four

phraseologisms (LH) and one phraseologism (PONS), respectively.

Overall, the dictionaries in question neglect to record impersonal and ditransi-

tive patterns and fail to record, or give due prominence to, the various patterns

featuring transitivized faire peter. The reason may be that lexicographers find it

easier to retrieve collocations from memory than valency patterns.


at UniversitÃ


¼ck on A


Dow

nloaded from


3.2 LA“ CHER (v)

I. tr. 1. (=delaisser, laisser, abandonner qqc) qqn lache qqc a) (definitive-

ment) ils ne veulent pas lacher leur taf ; depuis qu’elle a lache les films X,

elle fait de la publicite Collocations etudes, metier, poste, projet b) (pendant

un moment) Ils reussissent a lacher leur smartphone ? c) (+ nom abstrait)

Les gendarmes ont lache cette piste. Collocations+N occasion+ADV pas

facilement, jamais Phraseologie ne rien � Il peut aller chercher une

medaille. On sait qu’il ne lachera rien. tout � Un pere medecin qui lache

tout pour vous coacher. ne pas lacher le rythme/la pression/. . . Allez, ne

lache pas le rythme. (=ne change pas de rythme) ; ne pas lacher l’affaire/le

morceau ; lacher prise ; lacher la main (a qqn)

2. (=quitter, abandonner qqn) qqn/qqc lache qqn a) quand on est dans la

difficulte, ce n’est pas le moment de lacher les copains ; la droite l’a lache

; son editeur l’a lache ; au debut, on etait laches un peu comme ca dans

l’internat, sans savoir quoi faire. ; Alain Duclos est lache par sa moto.

(contexte : rallye) ; on redevient humain quand on est lache par les medias

? Collocations copain, amant b) (= laisser qqn en paix) Putain, lache-moi,

je veux dormir. ; Ca va/c’est bon, lache-moi Collocations+ADV un peu,

jamais Phraseologie Lache-moi les baskets. (= laisse-moi tranquille) c)

(= laisser partir ; mettre en liberte) qqn lache qqn/un animal Avant de

lacher Patrick, qui doit rejoindre Pierre Dubois. . . ; le prof nous a laches

plus tot ; Avant d’etre lache dans la reserve, Atlas a passe plusieurs mois

dans cette cage. Phraseologie lacher qqn dans la nature tous les deux laches

dans la nature dans Paris, c’etait pas triste ; Il y a parfois une inquietude de

la part des patients, qui ont la crainte notamment d’etre laches dans la

nature juste apres l’intervention. (ne pas�=se cramponner a) En voyant

leurs petits-enfants, les grands-parents ne les lachent plus. Phraseologie ne

pas� des yeux/du regard ; ne pas� d’une semelle ; lacher la proie (pour

l’ombre)

3. (= captiver) qqc (idee, film, livre) ne lache pas/plus qqn Je sais, pour l’avoir

vecu, que lorsqu’un livre vous attrape, il vous lache plus. ; C’est un film qui

vous prend a la gorge des le depart et ca ne vous lache plus.

4. (= faire sortir) a) qqc lache qqc (cuisine) Ca lache plus de jus. ; Au moment

du changement de temperature le matin, la fleur lache le pollen. Collocations

*pet/perle, rafale (de mitraillette) ; vapeur b) qqn lache qqc (sur qqn)

Phraseologie *en lacher une (=peter) ; ** � la puree/la foutre (sur)

(= ejaculer)

5. (= cesser de tenir ; laisser tomber ; perdre le controle de qqc) qqn lache qqc

(a qqn) a) Vous pouvez me� le bras ? ; J’ai lache le portable. Il est tombe

au fond de la voiture. ; il a lache le savon dans l’eau Collocations ballon,

volant, guidon ; les amarres, du lest ; pedale Phraseologie lacher du lest, un

peu de lest b) (= rendre moins tendu ou moins serre) Collocations corde,


at UniversitÃ


¼ck on A


Dow

nloaded from


ceinture Collocations+ADV doucement, brusquement Phraseologie la

taille d’un vetement ; du fil ; � la ligne ; � la bride a un cheval ; a qqn

(= le laisser plus libre) ; a qqc (emotions, imagination) ; � la bonde a qqc

c) (= lancer) qqn lache un animal (sur qqn/qqc); une bombe Il a lache son

Rottweiler sur moi. Collocations bombe, bordee ; chien, meute, pigeon(s) d)

(=donner, distribuer [de l’argent]) C’est vrai que j’ai bien gagne la-dessus.

Je peux bien en lacher un peu. Phraseologie ne pas les lacher (facilement) ;

les lacher avec un elastique/des elastiques ; a toi de les lacher (= c’est toi

qui paies)

6. (= laisser echapper; emettre brusquement) a) Elle lache des mots d’anglais

pour se donner un genre. Collocations+N soupir, cri, gemissement ; betise,

plaisanterie, remarque, juron b) (dire enfin) : Il n’a rien lache sur le gaz de

schiste ? Le mot est lache. (= enfin quelqu’un l’a dit) Collocations+N

mot, phrase, infos, detail+ADV au compte-goutte Phraseologie lacher

le morceau/le paquet (= tout avouer) c) (= laisser libre cours a) Lacher

des emotions si profondement enfouies, c’est douloureux.

7. qqc lache qqn (=ne plus fonctionner)

8. (= sport : distancer) qqn lache qqn Il lache le peloton.

9. (=ordinateur) lacher un icone

II. intr. 1. a) (=abandonner) Je ne lache pas facilement. Collocations+ADV

mentalement (= s’accrocher, s’affirmer) Les Irlandais prennent le score et

ne lacheront plus malgre cet essai des la reprise sur le poteau de coin. ;

Depuis des annees, je fais respecter l’ordre. Il ne faut pas lacher maintenant.

b) (= inciter qqn a abandonner ou a demander moins d’efforts aux

autres) Parfois, tu as envie de lui dire “Lache un peu.” Il est tellement

perfectionniste que la frustration qu’on a dans l’equipe, c’est qu’on

n’arrive jamais a le satisfaire. ; C’est bon, lache. Il est remis a sa place,

c’est parfait comme ca.

2. (ne plus fonctionner ; se rompre, se detacher). . . un moteur qui lache. . . ; Et

c’est le noeud simple qui lache le premier. Collocations+N : corde ; frein,

embrayage, soudure ; digue (a. fig.) Phraseologie les nerfs lachent

Pourquoi, je ne sais pas, les nerfs ont lache.

3. (peter) Qui a lache ?

III. pron. 1. se lacher (avec qqc) a) (= se laisser aller ; laisser libre cours a ses

emotions) Vous vous lachez pendant les fetes ?; Tout le monde s’est lache

avec cette musique. ; Pour une plante de salon, on peut se lacher. On peut

se lacher dans les formes, les couleurs. . . ; Ca te ferait du bien de te lacher,

tu as le droit de pleurer. b) (s’exprimer librement) Vas-y, lache-toi, ma

cherie ! Collocations + ADV completement, totalement, vraiment 2.

(= lacher II. 1a) Encore 300m de course. Elles ne vont pas se lacher.

It may be intuitively surprising that lacher is among the 2000 most common

words in spoken French. One reason for this is the bewildering variety of


at UniversitÃ


¼ck on A


Dow

nloaded from


identifiable senses, ranging from the abandonment of people through the expen-

diture of money to the breakdown of machinery, each of which contributes to

the verb’s overall commonness. However, as Zipf’s law would predict, the com-

monest senses are far more common than the less common senses, raising the

age-old question of the order in which senses should be arranged. In the case of

peter above, it was argued that the ontogenetically earliest sense should be listed

first. With lacher, this sense is more difficult to determine, but there are good

reasons for believing that the sense which is prevalent in spoken language overall

may also be the sense that is first acquired. In this sense, lacher has become the

preferred equivalent of abandonner, taking both persons and things as its direct

object. Sense I.6 combines with objects designating language.

There are four other senses in evidence which may be broadly grouped under

sense I. Sense I.3 is a figurative extension of senses I.1 and I.2 which takes

works of art as subjects and humans as objects. Sense I.4 is broader in scope,

referring, as it does, to the release of any kind of substance, including highly

informal and even vulgar uses (lacher la puree). Sense I.5 is the literal meaning

of ‘no longer holding on to sth’ or ‘slacken’, which has two more specific

variants meaning ‘letting (animals) loose’ and ‘giving away money’.

Most of the (sub-)senses identified have highly specific collocations and

phraseology, with the exception of the most common senses I.1 and I.2, which

interestingly have a wide range of unspecific collocates. Thus, lacher I.4 b) com-

bines with animal objects and bombs, and II.1 collocates strongly with the

adverb mentalement, especially in the context of sport. It is also worth noting

that most of the senses of transitive lacher hardly ever occur in the passive voice.

Two major intransitive uses are discernible, one referring to people giving up,

the other to devices or manipulations that fail to work. The reflexive se lacher, with

its basic meaning of relaxation, covers a broad spectrum of uses ranging from

clubbing through architecture and decoration to the release of strong emotions.

Lastly, attention should be drawn to the dense use of cliches containing

lacher in colloquial French, the most common of which are lacher prise,

lacher l’affaire, lacher le morceau and lacher le rythme.

A fairly clear pattern emerges from the dictionary checks (see Table 3), show-

ing that, in the case of the complex verb lacher, the bilingual dictionaries make

the same sense divisions as their monolingual counterparts, and rely on these for

information on collocability. One example among many is the use of creancier as

a collocate in LHF, which may be traced back to its appearance in the following

example sentence in PR: les creanciers ne le lacheront pas. Equally evident is the

almost complete absence of the most colloquial uses of lacher. HS and PR are

the only dictionaries to include the reflexive use. However, somewhat oddly, HS

wrongly labels this highly common use ‘literary’. HS is also alone in illustrating

sense I.3 with an example sentence (cette idee ne m’a pas lache), but, by placing it

under the sense ‘cesser de tenir’ (e.g. lacher la pedale du frein), it neglects the

specific valency pattern associated with this use (qqc ne lache pas/plus qqn).


at UniversitÃ


¼ck on A


Dow

nloaded from


Table

3:lacher

intenFrench

andFrench-English

dictionaries

la-cher

I.1a)

I.1b)

I.1c)

I.2a)

I.2b)

I.2c)

I.3

I.4

I.5a)

I.5b)

I.5c)

I.5d)

I.6a)

I.6b)

I.6c)

I.7

I.8

I.9

II1a

II1b

II2

II3

III.a

III.b

PR

++

-+

++

++

++

-+

++

++

-

TLF

++

++

+-

++

++

++

++

--

DDF

++

++

++

DAFLES

++

++

HSSlang

++

-+

LWUF

++

Collins

++

++

++

++

++

HS

++

++

++

++

++

++

++

+

LHF

++

++

++

++

++

PONS

++

++

++

++

+


at UniversitÃ


¼ck on A


Dow

nloaded from


(Almost) all the dictionaries under investigation miss out other important

senses, notably I.1.b) and c); I.6.b) and c); I.9. DAFLES especially fails dis-

mally to include even some of the most basic senses; the reason for this may be

that it is based on a newspaper corpus and on folk definitions which rely

heavily on collocations. However – and this is a general lesson to be derived

from the analysis of lacher - lacher has a very limited range of typical collocates

in most of its senses, and very often these are adverbs rather than nouns. Hence

the difficulty experienced by the authors of DAFLES to identify more than a

handful of senses; it would be interesting to investigate whether this points to a

general weakness in DAFLES.

The great semantic complexity surrounding lacher means that even in HS

many senses have been illustrated with only one suggestive example and that no

general valency patterns are provided in any of the dictionaries.

3.3 MEC (n)

(I) (=homme, par opposition a femme) Ca, c’est les mecs. Tous pareils.;

Les mecs aiment le foot.; Je me sens un peu seul, le seul mec!

(II) (= individu masculin quelconque): Le mec parle super bien l’anglais.;

T’as vu les trois mecs la-bas?; Salut, les mecs.; Mec, t’as pas une clope?

Collocations+ADJ: beau mec, jeune mec, mec bien/cool/formidable/

genial/super/sympa; pauvre mec, petit mec; mec louche; mec

bourre+DET: ce mec, le mec en question

(III) (=petit ami, compagnon, y compris dans les couples gay) Je suis

amoureuse de ce mec.. . . la liste de tous les mecs avec qui elle est sortie. . .

Les filles quittent un mec pour un autre. Collocations+ADJ un autre

mec (=un rival), le premier mec, nouveau mec, mec parfait; mec

chiant, mec jaloux+V : trouver/rencontrer/avoir/sortir avec/coucher

avec/se taper/rester avec/draguer un mec+DET : mon/ton/son mec

(IV) (=homme viril, partenaire sexuel) Les seuls mecs qui m’ont plu sont des

mecs beaucoup plus mecs que moi. Collocations+ADJ beau mec, vrai

mec, mec mignon, *mec bien membre, mec viril, mec muscle, mec

mature+V: **sucer un mec, **chevaucher un mec

At first blush, it may seem that mec is a fairly unremarkable word which

substitutes for homme in colloquial French, and this is indeed one of its major

senses (I.). In this sense mec enters into a number of cliched expressions which

ascribe negative qualities to men (les mecs sont tous des [salauds, cretins, . . .]).

On closer inspection, however, it is possible to identify at least three other

senses that are distinct from the meaning ‘homme’ and from other vague

words such as type and gus. The first sense serves to categorize male individuals

and to address them in the vocative. The adjectival collocations that may be

grouped under this sense are usually imbued with positive semantic prosodies


at UniversitÃ


¼ck on A


Dow

nloaded from


(mec bien, mec super, etc.), with few exceptions (mec louche, mec bourre; pauvre

mec). It is worth noting that both the vocative use of the singular noun mec and

the collocation beau mec, when employed with copular verbs, take the zero

article (t’es beau mec, je veux dire), while the plural vocative is preceded by the

definite article (eh, les mecs). Sense III, referring to men as (heterosexual or

homosexual) boyfriends, is usually associated with singular personal pronouns

(mon/ton/son mec) and a wide range of verbs to do with dating and love rela-

tionships (rencontrer un mec, rester avec un mec, etc.). Closely related to this is

sense IV, which refers to men as objects of sexual desire (mec viril, mec muscle,

mec bien membre). A comparison of the collocational preferences of mec and

synonyms such as homme and type highlights that these are by no means inter-

changeable. Thus, the semantic feature [+ sexually desirable] associated with

sense IV precludes collocations of the type ?mec impuissant, while homme im-

puissant is notably common. By contrast, explicitly sexual collocations abound,

such as the vulgar mec a grosse bite, where type a grosse bite seems less likely.

Similarly, Sense II does not usually combine with adjectives such as mechant or

dangereux, while the synonym type does.

Table 4 shows that TLF and PR make the same binary sense distinction

(1. Homme energique, viril; 2. Homme, individu quelconque), with PR listing

‘individu masculin’ and ‘compagnon d’une femme’ (which overlooks the homo-

sexual use) as sub-senses of sense 2. TR makes no mention of the ‘boyfriend’

sense, instead giving too much weight to the dated sense ‘client (of a prosti-

tute)’. While PR captures the entire meaning spectrum encoded by mec, it

places the emphases wrongly, since the frequency of the ‘boyfriend’ sense

and the abundance of collocations associated with it clearly require a separate

sense division to be set aside. A measure of the general datedness of TLF is its

inclusion of the following comment:

Dans ce sens, mec s’emploie beaucoup (. . .) avec des adjectifs pejoratifs

(pauvre mec, petit mec) (. . .)

This observation may have been correct in the 1970s, but evidence from the

CRFC clearly shows the predominance of neutral and positive associations in

present-day French.

DDF’s and LWUF’s treatment of mec is rudimentary and fails to clarify

important sense distinctions for the foreign learner. This is to some extent

mitigated by the choice of pertinent examples sentences containing useful col-

locations and, in the case of LWUF, workable translation proposals, although

it is doubtful whether foreign users will be able to infer the different shades of

meaning conveyed by pauvre mec from a mere quotation in DDF (Pauvre mec!)

or the translation ‘ein armer Typ’ in LWUF.

HS Slang has a very short entry which gives due weight to senses I and IV

but provides no information on collocability. The unabridged HS and PONS


at UniversitÃ


¼ck on A


Dow

nloaded from


have the same sense division, with HS also offering more examples; it is the

only dictionary to illustrate the use of mec with the zero article: t’en fais pas,

mec!

Collins’ and LHF’s handling of mec is disappointingly poor, confining itself

to the enumeration of a few general translation equivalents.

There is a complete absence of verbal collocates from all the dictionaries

examined and, probably under the influence of TLF, a sparse coverage of

positively connoted adjectival collocates. No mention can be found of the

common use of mec to refer to homosexual partners.

3.4 LOOK (n)

(I) (personne ; = style vestimentaire, coiffure, chaussures, maquillage,

accessoires, etc. ; souvent volontairement etudie) T’as un look excentri-

que. ; Je porte le look gothique. ; La base du look punk, c’est le cuir. ;

Le look n’est pas une question d’age, mais de gouts!; Le look homosexuel

Table 4: mec in ten French and French-English dictionaries

I. II. III. IV. Collocates

PR (listed as

a sub-sense)

+ + + mec a la

redresse ;

pauvre mec

;

petit mec

TLF - + - +

DDF - + - (sense mentioned

in an example)

- drole de mec,

beau mec,

pauvre mec

DAFLES - - - -

LWUF - - - - pauvre mec,

le mec des

mecs

HS Slang + - + - -

HS + - + - drole de mec,

pauvre mec,

petit mec,

vrai mec

Collins - - - - -

LHF - - - -

PONS - (sense mentioned

in an example)

+ + - pauvre mec


at UniversitÃ


¼ck on A


Dow

nloaded from


permet de rentrer dans certaines discotheques de Paris. ; On a quatre

looks. Qu’est-ce que vous preferez?; Les differents looks qu’elle avait,

gothique, punk, tous ces styles melanges. . . ;. . . les looks des people. . . ;

Elle cultive un look un peu decale. ; Que pensez-vous du look de votre fille

aujourd’hui?; J’ai toujours bien vecu mon look, mais depuis l’arrivee de

mon fils je me sens moche et ridicule dans ce look. ; avec son look de

baroudeur Collocations+ADJ androgyne, chic, decontracte, gothique,

impeccable, intemporel, ravageur, retro, sexy, soigne, vestimentaire; le

total look+V adopter, adorer, ameliorer, avoir, completer, creer, cul-

tiver, donner, personnaliser, porter; changer de+PREP dans, avec

Phraseologie� du tonnerre, d’enfer ; t’as le look, coco (chanson)

(II) (chose : appartement, batiment ; magazine=design ; image donnee par

qqc) un mobilier avec un look fifties ; un episode de Star Trek avec son

look seventies; Perso, j’aime bien le look tole ondulee, et les enceintes

chaudes bouillant avec 4000 fans dechaines. . .+ADJ avant-gardiste, chic,

contemporain, elegant, futuriste, minimaliste, moderne, retro

Look is a comparatively long-established Anglicism which originated among

the users of francais branche, i.e. the language of the metropolitan in-groups

(cf. Merle 1989), but has now passed into general use. Although its first sense in

particular has developed its own patterning (e.g. look de baroudeur) and con-

notations, which set it apart from French near-synonyms such as style and

allure, it is probably the word among those examined which could most

easily be regarded as a simple substitute for the less fashionable style, sharing,

as it does, many of its collocations with this near-synonym (cf. for example

personnaliser son style/look, un style/look decontracte/glamour). This is even

more true of its second sense, which is almost interchangeable with its closest

synonym (‘design’). Look is particularly common in the worlds of fashion and

show business, as another comparison with its near-synonyms style and allure

shows: one speaks of ‘le look des chanteurs/des people (i.e. celebrities)’ rather

than ‘l’allure/le style des chanteurs/des people’. Since it describes a fairly per-

manent defining feature of people’s personalities, look has a strong tendency to

combine with the definite article (le look homosexuel/gothique, etc.). Like items

of clothing, it also collocates readily with the prepositions dans and avec. It is

also worth noting that the phraseology of look suggests that branche words

tend to team up with each other to form branche phrases (un look d’enfer/du

tonnerre).

When dealing with colloquial words which mainly act as substitutes for core

words of the language, dictionary makers should explore to what extent col-

locational features are shared by the words in question. Only the more specific

collocations should be listed at the entry for the colloquial word.

Although six out of the ten dictionaries under survey correctly capture the

two comparatively distinct senses of look, the extreme commonness of the word


at UniversitÃ


¼ck on A


Dow

nloaded from


would lead one to expect more than the extremely rudimentary treatment ac-

corded to its lexical environment in all the dictionaries (see Table 5). Harraps

provides the most detail, including four collocational items. None of the dic-

tionaries would allow the foreign-born user to produce sentences such as elle

cultive un look un peu decale.

3.5 AUTOUR DE

(I) (=dans l’environnement de ; a proximite de) Elle supporte mon bordel

quotidien autour de l’ordinateur, mes disques qui s’eparpillent et les boıtiers

qui disparaissent. J’enroulais ma serviette autour de la taille. (+ pronom

personnel) 72 cameras autour de nous (+ partie du corps) Il s’est mis du

lait autour de la bouche. ; une serviette autour de la taille

Collocations+V: regarder autour de soi, avoir/mettre/enrouler/nouer

qqc autour de qqc, s’asseoir autour de (la table), graviter/orbiter

(autour d’une planete/d’une etoile)/voyager autour du monde+N :

autour de la table/de la maison/de la Mediterranee/d’un feu de bois/de

la bouche/du cou/de la taille Phraseologie autour d’un cafe/dejeuner/

repas/verre (etc.) (=en prenant un cafe, etc.) Ca vous dirait de parler

de tout ca autour d’un the?; tourner autour de qqn (lui tourner autour) ;

faire le vide autour de soi ; autour d’une valeur La cote argus n’est pas un

prix mais une valeur de reference autour de laquelle se fabrique un prix.

(II) (=dans l’entourage de qqn ; en rapport avec qqn) Autour de moi, tout

le monde a des amis. ; Tout change autour de nous ; Vous avez forcement

des gens autour de vous qui. . . ; il y a beaucoup de pretendants autour de

la petite ; il faut coordonner les interventions autour de la personne agee ;

Table 5: look in ten French and French-English dictionaries

I. II. Collocates

PR + + (‘image’)

TLF - -

DDF + - changer de look

DAFLES - -

LWUF + +

HS Slang + + avoir un look d’enfer

HS + + T’as le look, coco!; changer de look, soigner

son look, avoir un look d’enfer

Collins + +

LHF + -

PONS + +


at UniversitÃ


¼ck on A


Dow

nloaded from


Il avait une grosse equipe autour de lui. ; Meme si la couveuse n’est pas

la, elle est dans notre tete et on fabrique autour de ce bebe comme une

bulle de protection. ; Mais a 10 ans, j’ai decouvert que j’avais une voix

particuliere et que je provoquais des choses autour de moi.

(III) (=environ) un cadeau autour de 100 euros ; J’avais commence a prendre

du poids autour de la quarantaine. ; Vous pouvez voir ici qu’on est tou-

jours au-dessous, autour de 0,006. ; Les voix des filles changent un peu

autour de 20 ans. Phraseologie On (en) est autour de (+ quantite) Dans

le commerce, a qualite egale, on sera plutot autour de 15 a 20 euros

minimum. ; On en est autour de 100000 ventes.

(IV) (=avec qqn ou qqc au centre) se retrouver autour d’une passion com-

mune ; Ta vie est trop centree autour de lui. ;. . . une famille qui a ete

importante dans notre pays et qui se reconstitue autour de deux hommes

qui ont decide de mettre leurs egos de cote ; En creant l’UDI autour de

Jean-Louis Borloo, nous voulons incarner autre chose. ; La famille se fait

autour de l’enfant.

Collocations+V (s’)articuler/construire/(re)grouper/organiser/reunir/federer/

(se) rassembler/structurer/(se) mobiliser autour de (axes/idees/poles/projets/

valeurs/themes/. . .)

(V) (= en ce qui concerne ; sur) J’entends surtout ceux qui ont essaye de faire

beaucoup de tapage autour de ce dossier, qui ne disent pas la verite. ; Le

pharmacien pourrait avoir un role accru autour de l’adaptation de cette

posologie a la personne, en fonction du conditionnement.

Collocations+V bavarder/debattre/echanger/tourner/graviter/. . . autour

de+N animation/enjeu/mystere/polemique/. . . autour de

Table 6: autour de in ten French dictionaries

I. II. III. IV. V.

PR + +- (sub-sense of I.) +- (sub-sense of I.) - -

TLF + +- + +- +-

DDF + - - - -

DAFLES - - - - -

HS Slang (unrecorded)

LWUF (unrecorded)

Collins + - + - -

HS + - + - -

LHF + - + - -

PONS + - + - -


at UniversitÃ


¼ck on A


Dow

nloaded from


Surprisingly, all the dictionaries under investigation are seriously flawed in

their handling of the complex preposition autour de (see Table 6):. TLF is the

only dictionary to record definitions or examples that go at least some way

towards explaining the various uses of autour de, although it, too, contains a

number of misjudgements and leaves out important senses (such as autour d’un

verre). Thus, it posits a figurative sense in which autour de is followed by an

inanimate abstract noun, citing examples like the following:

22. Nous avons groupe ces remarques autour de la vie de grande ville. La

petite ville, le bourg, presentent autant de variantes entre l’influence urbaine

et l’influence paysanne, et aussi autant de traits originaux.

MOUNIER, Traite du caractere, 1946, p. 87. (TLF, s.v. autour de; italics in

the original)

However, the CRFC data show that autour de should here be analysed as

being dependent on the verb grouper rather than the noun remarque (cf. sense

IV. above). Autour de is part of the complementation pattern of verbs such as

construire, articuler and grouper, and this use is by no means restricted to

abstract nouns (cf. (se) grouper autour de lui). The fact remains that autour

demay be part of a nominal complementation pattern (cf. sense V), but the two

uses are not clearly distinguished in TLF.

All the other dictionaries merely record the locative and approximation

senses, with HS including the sub-sense ‘autour d’un verre’. Nowhere do we

find any attestation of fixed expressions such as on (en) est autour de or autour

de sa personne or of the sub-sense ‘autour d’une valeur’ (une valeur de reference

autour de laquelle se fait un prix), which needs to be distinguished from the sub-

sense ‘environ’. The treatment in the two learners’ dictionaries is particularly

rudimentary.

3.6 N’IMPORTEQUOI

(I) (= tout, sans discernement) Les jeunes aujourd’hui mangent n’importe

quoi. ; L’avantage de ce sport, c’est qu’on peut utiliser tout et n’importe

quoi. ; On laisse faire des decorateurs qui font n’importe quoi, ca faisait

beau. Phraseologie faire n’importe quoi (pour/pour que) J’aurais fait

n’importe quoi pour rester. dire/raconter n’importe quoi (=dire des

mensonges) Elle ne peut pas dire/raconter n’importe quoi sur moi.;

C’est pire depuis que cette fille est arrivee. -Tu dis n’importe quoi.;

(= tu dis des inepties) Il se dit n’importe quoi; (= toutes sortes de

choses)

(II) (= incroyable, impossible) Shakira reprend Cabrel. Ca devient n’importe

quoi. ; Il y a eu du n’importe quoi. Tout le monde le sait. (=des choses

incroyables) ; C’est n’importe quoi. (=des inepties)


at UniversitÃ


¼ck on A


Dow

nloaded from


(III) Marqueur du discours

(a) (= en se plaignant du comportement de qqn) (p.ex. mere qui ramasse

une sucette que son enfant vient de jeter par terre) N’importe quoi ! ;

b) (=pour marquer son desaccord) - Je l’ai apercu, il est sexy. -

Pff. . . N’importe quoi! ; A chaque appel, tu me parles de Lola. -

N’importe quoi! (= c’est pas vrai) ; (suivi de « ah si », etc.)

N’importe quoi. – Ah si./Ben si./Ben oui.

(IV) pas n’importe quoi

(a) (negation du sens I) Je vais vous demander de chanter. Mais pas

n’importe quoi. Vous allez chanter du rap. ; Il y a des regles strictes.

On ne fait pas n’importe quoi et cela protege le consommateur.

(b) (negation du sens II) Tu sais ce que je t’ai dit hier sur ses pouvoirs? -

Oui. -C’etait pas n’importe quoi. (= serieux)

(c) (=grave, « du lourd ») C’est un pilier de la societe democratique qui

est atteint. Ce n’est pas n’importe quoi d’abattre quelqu’un dans un

journal, de se rendre dans un media avec une arme comme cela.

Collocations+ADV un peu n’importe quoi ; + V prendre/acheter/vendre

n’importe quoi

All the dictionaries under discussion assign low priority to the use of

n’importe quoi as a discourse marker, if mentioning it at all: (see Table 7).

TLF has an entry for n’importe + interrogative adverb which fails to include

n’importe quoi. PR treats n’importe qui and n’importe quoi under the same sub-

entry, but fails to distinguish between the literal and the discoursal use of the

sequence; the discoursal sense cannot be inferred from the only example given:

c’est vraiment n’importe quoi! The same goes for DDF.

The handling of n’importe quoi in the bilingual dictionaries is more adequate,

even though no individual dictionary captures any of the niceties of its

Table 7: N’importe quoi in ten French dictionaries

I. II. III. IV.

PR + - - -

TLF - - - -

DDF + - - -

DAFLES (unrecorded)

HS Slang (unrecorded)

LWUF (unrecorded)

Collins + + + -

HS + + + +

LHF - - - -

PONS + + + -


at UniversitÃ


¼ck on A


Dow

nloaded from


communicative use. Collins, for example, provides the following translation

equivalents:

n’importe quoi anything; il fait/dit n’importe quoi he has no idea of what

he’s doing/saying!; n’importe quoi! rubbish!, nonsense!

HS, which treats n’importe as an adverb with entry status, goes a little further

by illustrating the two major senses of pas n’importe quoi:

3000 euros, ce n’est pas n’importe quoi! 3000 euros is not to be sneezed at!;

une table Louis XIII d’epoque, ce n’est pas n’importe quoi a genuine Louis

XIII table is no ordinary table

PONS provides an excellent illustration of the sense which has the speaker

rejecting what seems to him an outrageous suggestion:

je reve d’un coupe decapotable (sic!) – n’importe quoi! Et pourquoi pas d’un

jet prive ? ich traume von einem Cabrio – sonst noch was! warum nicht

gleich von einem Privatjet?

3.7 C’ESTBON

(I) (appreciation positive sur ce qu’on mange ; interrogation possible) -

C’est comment les œufs a la coque, Antoine ? - C’est bon. ; Ce que

c’est bon ! ; Les quenelles, c’est bon. Collocations C’est bon, hein ?

(II) (=confirmation qu’on n’a pas besoin d’aide ou que quelque chose

fonctionne; interrogation possible) Tu peux la monter la c’est bon?

C’est bon. ; -Ca va? C’est bon? Tout est boucle? Pas besoin d’un coup

de main? -Non, ca va aller. ; Tu veux que je t’aide? -Non, c’est bon.

Collocations Non, c’est bon.

(III) (=desir d’etre laisse en paix, mecontentement, refus ; interrogation

impossible) - Je vais prendre ma douche. -N’y reste pas 2 heures. -

C’est bon! ; Va te faire foutre! Trois fois, mec! - C’est bon, degage!

Collocations C’est bon, j’ai compris.

(IV) (= jugement que quelque chose est suffisant ou adequat ; interrogation

possible) a) (par rapport a une quantite) On est a 120 kg. – C’est bon.

Arrete, Florian. ; 50 euros, c’est bon ? (en payant une note=gardez

tout) C’est bon. (par rapport a la qualite) - Je ne voudrais pas qu’il y ait

trop de blanc la-bas. C’est bon comme ca? – Je pense, oui. ; (a un enfant)

Plus qu’une cuillere et c’est bon. ; (= en faisant l’amour) Humm. Leche-

moi. C’est bon. b) (resolution satisfaisante par rapport a une situation

anterieure) J’ai eu des frayeurs la premiere fois. Mais quand tu l’as

fait une fois, apres, c’est bon. ; Tu remues comme une petite omelette.


at UniversitÃ


¼ck on A


Dow

nloaded from


Des que le jaune est destructure, c’est bon. ; -Marc a trouve un appart. -

Tu as trouve? - Oui, c’est bon. ; C’est bon. Elles m’ont convaincue. ; - On

a repere l’individu, il est directement sur sa droite. - OK, c’est bon, on le

tient. Collocations C’est bon, il y arrive. ; C’est bon. Impeccable. ; Allez,

c’est bon. c) (=pour marquer la fin) Voila c’est bon j’ai fini. d)

(= t’inquiete pas ; il n’y a aucun probleme) ah ben je sais pas il doit

y avoir du jus d’ orange ou quelque chose. . . ouais ben c’est bon je vais

me debrouiller va! ; - Tu risques plus rien? - J’ai pris sur moi pour pas lui

peter la gueule, mais c’est bon. - Il t’a donne une garantie? -T’inquiete, ca

va aller. ; - Je me suis pas echauffee. -C’est bon, echauffe-toi. ; On fait

une pause ou c’est bon ? ; Il faut que je m’approche ou c’est bon la ?

Collocations C’est bon la (./?) Mais c’est bon. C’est tout bon. e) (= en

conduisant : la voie est libre) La, va tout droit. C’est bon. f) (=pour

dire que quelque chose est pret ou pour s’assurer que quelqu’un est

pret) -Vous etes prets a virer? -Oui. -C’est bon? -Allez, virez de bord. ;

C’est bon. Vous pouvez prendre le bain. Collocations C’est bon. On y

va./On peut y aller.

C’est bon is a notably common and highly polysemous discourse marker

whose senses diverge markedly from a hypothetical ‘compositional’ meaning

based on its constituent parts. There is a remarkably wide spectrum of senses,

ranging from ‘nothing coming’ (navigator/passenger to driver) through ‘keep

the change’ to ‘I’ve had enough of this’. Close scrutiny of the lexical environ-

ment of c’est bon also reveals the need to redefine the notion of collocation with

regard to discourse markers in spoken language. We do not find the well-

known types of collocations with lemmas belonging to particular parts of

speech, but rather combinations with specific conjugated verb forms (c’est

bon+present perfect tense of particular verbs: j’ai compris/il a compris/j’ai

fini; c’est bon+present tense of arriver: j’y arrive/il y arrive), connectors

(mais c’est bon), long-distance collocations (des que. . . c’est bon) and other

phrases (C’est bon. On y va.). Some of these collocations are in their turn

polysemous, such as C’est bon, j’ai compris, which may express either annoy-

ance or a genuine epiphany on the part of the speaker. It is also evident that

‘dialogic’ exemplification is needed to illustrate many of the uses of c’est bon.

While the discourse marker in question suffers total neglect in most of the

dictionaries, there is some attempt in HS and LH to cover basic uses. However,

as can be seen from the following reproduction of the entry in LH, these do

no more than scratch the surface of the complexities underlying the use of

c’est bon.

int (c’est) bon! a) accord, satisfaction (es ist) gut!; b) conclusion,

deduction (na, also) gut !; c) mecontentement, refus schon gut !, lassen

wir’s gut sein !


at UniversitÃ


¼ck on A


Dow

nloaded from


4. Discussion

The present study started from the observation that traditional lexicography

has tended to rely on corpora of written text. It was hypothesized that this

might be to the detriment of covering the commonest colloquial lexical units

which carry the main burden of everyday conversation, are usually acquired

early on in life and are therefore more deeply anchored in speakers’ brains than

words first encountered in the course of education.

The results of the analysis summarized in the previous section have provided

some evidence of the viability of this hypothesis. The examination of the treat-

ment of four words and three multi-word strings in ten different French dic-

tionaries has shown that current lexicographic descriptions of spoken French

are often patchy and inadequate with respect to various lexico-grammatical

features.

Most notably, there was found to be an almost complete absence of infor-

mation on the collocational range of colloquial items in the dictionaries under

investigation, although, at least theoretically, native-speaker lexicographers

could have retrieved some collocations from memory. There is on the whole

a considerable uniformity in the content of both monolingual and bilingual

dictionaries, with monolingual learners’ dictionaries and bilingual dictionaries

tending to adopt collocations and sense divisions from PR and, to a lesser

extent, TLF. The only serious exception is Harraps, which in many cases,

such as peter, offers its own sense divisions and goes further than the other

dictionaries in meeting the user’s encoding needs.

Since the extraction of collocations is an essential prerequisite for the deter-

mination of meanings, it was hardly surprising to find that the marking of sub-

senses may not be sufficiently clear for the encoding needs of speakers of

French as a second or foreign language. Here the clearest examples were the

highly polysemous verb lacher and the complex preposition autour de. The only

way to enable learners to gain an overview of, and ultimately to make pro-

ductive use of, a verb like lacher is to opt for splitting up the various senses

derived from its literal meaning ‘ne plus tenir’ rather than lumping them all

together, as PR does. It would be unrealistic to expect learners to derive such

specific uses as the following from the general sense, especially since this use

features a non-human subject:

C’est un film qui vous prend a la gorge des le depart et ca ne vous

lache plus.

Another main finding concerns the rudimentary treatment of common multi-

word items with a clear discoursal function. Thus, nine of out of ten diction-

aries fail to record the sense in which the common preposition autour de is used

to indicate that someone or something is at the centre of a particular


at UniversitÃ


¼ck on A


Dow

nloaded from


endeavour, and all the dictionaries give low priority to the use of n’importe quoi

as a discourse marker. Most seriously perhaps, there is no indication of the

typical contextual embedding or common lexical collocations of discourse

markers (e.g. – N’importe quoi. – Ah si.)

A discussion of the dictionaries’ policy on exemplification would require a

separate article. Suffice it to say here that most of the examples found in the

dictionaries under investigation illustrate written usage, and of those that illus-

trate spoken usage many lack some naturalness feature or other (cf. the dis-

cussion of peter). There is some evidence that different types of exemplification

may be needed for different words (cf. Hausmann 2005). lacher is one example

found in this article of a low-collocability word which cannot be illustrated by

means of typical cooccurrences. With such words, users will need a large

number of (at least) sentence-length examples to grasp the various meanings

of the word. The case is different with look, a clear example of a high-

collocability word which requires little exemplification beyond information

on collocation.

One minor flaw that does not just concern the lexicography of spoken

French is the poor coverage of delexical causative faire, which was found to

be almost completely absent from dictionary entries for peter, although it is a

notably common feature of its use.

Two dictionaries deserve special comment: DAFLES and Harraps.

Although it has several compensating strengths which fall outside the scope

of this study (cf. Verlinde, Binon and Selva 2006), DAFLES has almost no

entries for colloquial words, a fact which may be due to its corpus base.

However, as illustrated by the five big monolingual dictionaries for learners

of English, no serious learner’s dictionary can currently afford to neglect the

spoken language.

By contrast, Harraps achieves a remarkable harmonization of descriptive

and pedagogic needs. A measure of the overall quality of this monument to

bilingual lexicography is the inclusion of a large number of colloquial senses of

both lacher and peter and the provision of illustrative sentences from which

students may confidently extrapolate personal choices. Like all current diction-

aries, however, even Harraps is still weak on discoursal items.

There are two important theoretical lessons to be drawn thence. The first,

which concerns corpus linguistics, is that medium-sized spoken corpora like the

CRFC will shed light on lexical patterns and collocations about which even

very large mega-corpora of written language are completely uninformative.

This means that there may well be a second corpus revolution ahead which

will apply Sinclair’s famous dictum that ‘the language looks rather different

when you look at a lot of it at once’ to the investigation of intimate and

colloquial language use.

The second theoretical lesson is that colloquial words, far from being stylis-

tically ‘inferior’ substitutes of more formal words, are imbued with their own


at UniversitÃ


¼ck on A


Dow

nloaded from


specific shades of meaning, phraseology, and pragmatics. It is as if there is a

primary lexis which is even more deeply submerged in the routines of everyday

life and thus even less accessible to native-speaker intuition than the secondary

written lexis, but there can be no doubt that such lexis is communicatively prior

and that its detailed description is of crucial importance to second or foreign

language learners. The present study suggests that much of the primary lexis of

French, and very probably other languages as well, remains almost undescribed

in respect of all but its most basic features, with dire consequences for foreign

learners aspiring to acquire native-like proficiency in the language. Reliance on

the rarer and clumsier words or lexical units may make their French sound

stilted and unnative-like (e.g. il n’abandonne jamais son portable rather than il

ne lache jamais son portable).

This begs the question of whether the lexical units (i.e. the senses) peculiar to,

or most common in, speech should not be given first entry status in diction-

aries. Many dictionaries for foreign learners order word senses by their fre-

quency of occurrence in corpora; the underlying assumption is that the most

important sense will be looked up or should be learned first. Since corpus data

have so far been heavily skewed in favour of the written varieties, current

dictionaries record speech-specific sub-senses rather low down the list, if at

all. If the goal of language teaching is to teach the primary lexis used in every-

day encounters first, or sometimes exclusively, more weight will need to be

given to spoken data and the most common meaning units found therein. A

thought-provoking example is afforded by lacher, a verb which was found to

have a bewildering variety of different senses. All of the dictionaries under

survey give first entry status to the ‘literal’ sense derived from the adjective

‘lache’ (= slack, loose), although other senses are clearly predominant in the

daily output produced by French speakers.

The general feeling which emerges from this study is that the lexicography of

spoken French is still in a nascent, pre-corpus-based stage where lexicographers

can provide ‘little more than a collection of stabs in the dark’ (Hanks 2012b:

63). It is hoped that the sample articles and comments offered here will point

the way forward to a reliable and comprehensive description of the primary

lexis of languages such as French and English on the basis of such corpora as

the CRFC or the new spoken section of the BNC.

ReferencesA. Dictionaries

Atkins, B. T., et al. 2010. Collins/Robert French-English/English-French Dictionary.

(Ninth edition) Glasgow/London: Collins. (Collins).

Rey-Debove, J. (ed.) 1999. Dictionnaire du francais Le Robert & Cle International.

(DDF).

Verlinde, S., Selva, T., Bertels, A. and J. Binon (eds). Dictionnaire d’apprentissage du

francais langue etrangere ou seconde, Leuven, Institut Interfacultaire des Langues


at UniversitÃ


¼ck on A


Dow

nloaded from


vivantes. Accessed on 24 Januar 2015. http://ilt.kuleuven.be/blf/search.php.

(DAFLES).

Nicholson, K. and G. Pilard. 2012. Harrap’s Slang. Dictionnaire d’argot et d’anglais

familier. Paris: Larousse. (HS Slang).

Langenscheidt-Redaktion. 2010. Langenscheidts Handworterbuch Deutsch-Franzosisch/

Franzosisch-Deutsch. Berlin: Langenscheidt. (LHF).

Meißner, F. J. 1992. Langenscheidts Worterbuch der franzosischen Umgangssprache.

Berlin: Langenscheidt. (LWUF).

Stevenson, A. 2007. Harrap’s Unabridged PRO French-English/English-French.

Ottawa: Laurier Books Ltd. (HS).

PONS Großworterbuch Franzosisch. Deutsch-Franzosisch/Franzosisch-Deutsch.

Stuttgart: PONS. (PONS).

Robert, P., J. Rey-Debove and A. Rey (eds) 2008. Nouveau Petit Robert: dictionnaire

alphabetique et analogique de la langue francaise. Paris: Le Robert. (PR).

TLF: Tresor de la langue francaise. Dictionnaire de la langue du XIXe et du XXe siecle

publie sous la direction de Paul Imbs. Paris: Editions du Centre National de la

Recherche Scientifique 1971–1994. Accessed on 24 January 2015. http://atilf.atilf.fr.

(TLF).

B. Other literature

Blanche-Benveniste, C. 1991. Analyses grammaticales dans l’etude de la langue parlee.

In U. Dausendschon-Gay, E. Gulich and U. Krafft (eds), Linguistische

Interaktionsanalysen. Tubingen: Niemeyer, 1–18.

Blanche-Benveniste, C. 1997. Approches de la langue parlee en francais. Gap-Paris:

Ophrys.

Blanche-Benveniste, C. 2010. Approches de la langue parlee en francais. Gap-Paris:

Ophrys.

Cappeau, P. and M. Seijido. 2005. Les corpus oraux en francais (inventaire 2005 v.1.1).

http://www.culture.gouv.fr/culture/dglf/recherche/corpus_parole/Inventaire.pdf.

Cappeau, P. and F. Gadet. 2007. Ou en sont les corpus sur le francais parles? Revue

francaise de linguistique appliquee, 12.1: 129–133.

Carroll, J. B., P. Davies and B. Richman. 1971. The American Heritage Word Frequency

Book. Boston: Houghton Mifflin.

Cresti, E. and M. Moneglia (eds) 2005. C-ORAL-ROM: Integrated Reference Corpora

for Spoken Romance Languages. Amsterdam: Benjamins.

Debaisieux, J.-M. 2010. Corpus Oraux – Problemes methodologiques de recueil et d’ana-

lyse de donnees. Nancy: Presses Universitaires de Nancy.

Duhamel, Claude-Alain and Carole Balaz. 1993. Le gros dico des tout petits: 3000 mots

racontes par les enfants. Paris: Lattes.

Gadet, Francoise., et al. 2012. CIEL_F : choix epistemologiques et realisations empiri-

ques d’un grand corpus de francais parle. Revue Francaise de Linguistique Appliquee,

XVII-1: 39–54.

Hanks, P. 2012a. The Corpus Revolution in Lexicography. International Journal of

Lexicography, 25.4: 398–436.

Hanks, P. 2012b. Corpus Evidence and Electronic Lexicography. In S. Granger and

M. Paquot (eds), Electronic Lexicography. Oxford: Oxford University Press, 57–82.

Hanks, P. 2013. Lexical Analysis: Norms and Exploitations. Cambridge: MIT Press.

Hausmann, F. J. 2005. Isotopie, scenario, collocation et exemple lexicographique.

In M. Heinz (ed.), L’exemple lexicographique dans les dictionnaires francais


at UniversitÃ


¼ck on A


Dow

nloaded from

http://ilt.kuleuven.be/blf/search.php

http://atilf.atilf.fr

http://www.culture.gouv.fr/culture/dglf/recherche/corpus_parole/Inventaire.pdf


contemporains. Actes des "Premieres Journees allemandes des dictionnaires"

(Premieres Journees allemandes des dictionnaires, Klingenberg am Main, 25–27 juin

2004) (Lexicographica Series Maior 128). Tubingen: Niemeyer, 283–292.

Kennedy, Graeme. 1998. An Introduction to Corpus Linguistics. Harlow: Longman.

Koch, P. and W. Oesterreicher. 2011. Gesprochene Sprache in der Romania: Franzosisch,

Italienisch, Spanisch (Romanistische Arbeitshefte 31). Berlin/New York: De Gruyter.

Meißner, F.-J. 2006. Linguistische und didaktische Uberlegungen zur Entwicklung von

Kompetenzaufgaben im Lernbereich Mundlichkeit (Schwerpunkt Horverstehen).

franzosisch heute, 37: 240–282.

Merle, P. 1989. Dictionnaire du francais branche. Paris: Seuil.

Van Peteghem, M. 1994. Il vs ce/ca en construction impersonnelle. In D. Flament-

Boistrancourt (ed.), Theories, donnees et pratiques en francais langue etrangere.

Lille: Presses Universitaires de Lille, 95–110.

Pierrel, J.-M., J. Dendien and P. Bernard. 2004. Le TLFi ou Tresor de la Langue

Francaise informatisee. Actes de EURALEX 2004, Lorient, 165–170.

Siepmann, D. 2007. Collocations and examples: their relationship and treatment in a

new corpus-based learner’s dictionary. Zeitschrift fur Anglistik und Amerikanistik, 3:

235–260.

Siepmann, D. and C. Burgel. 2015. L’elaboration d’une grammaire pedagogique a partir

de corpus: l’exemple du subjonctif. In T. Tinnefeld (ed.), Grammatikographie und

Didaktische Grammatik – gestern, heute, morgen. Saarbrucker Schriften zur

Linguistik und Fremdsprachendidaktik (SSLF), Saarbrucken: htw saar.

Sketchengine. http://www.sketchengine.co.uk/.

Steyer, K. 2013. Usuelle Wortverbindungen. Zentrale Muster des Sprachgebrauchs aus

korpusanalytischer Sicht (Studien zur Deutschen Sprache 65). Tubingen: Narr.

Steyer, K. and A. Brunner. 2009. Das UWV-Analysemodell. Eine korpusgesteuerte

Methode zur linguistischen Systematisierung von Wortverbindungen. Mannheim:

Institut fur Deutsche Sprache.

Verlinde, S., J. Binon and T. Selva. 2006. Corpus, collocations et dictionnaires d’appren-

tissage. Langue francaise, 150: 84–98.


at UniversitÃ


¼ck on A


Dow

nloaded from

http://www.sketchengine.co.uk/


Dictionaries and spoken language: : a corpus-based review of French dictionaries.

Documents