French of the present and the past: the representation of the Parisian vernacular in Maurice Chevalier's songs [*]

Michaël Abecassis (Oxford)

1 Introduction

The influence of the mass media on language has led to a large number of studies (Chambers 1998, Bernet 2000, Colin 2000), but hardly any linguist has worked on a corpus of songs with the exception of a few references to popular French songs (Bernet 1995, Carton 1995). Whether the media plays a major influence on language is highly controvertial, leading to generalisations such as that the media invents words and disseminate them, sometimes even arguing that this entails sound and grammatical changes. The sociologist S. Lieberson (2000), looking at the correlation between names familiar from the media and the way people name their children, found that people tend to overestimate the influence of the media. Instead of popularising names, he showed that the media are not responsible for an active change and only reflected language changes that were already entrenched. Labov (1984) showed that in the inner-city of Philadelphia, daily passive exposure to standard English language on television and education has little effect on the dialect of young African American residents. Milroy /Milroy argue on the contrary that the media, such as television do have an influence in spreading idioms and popularising innovations, but viewers would not necessarily adopt them in the long run (Bauer 1994: 7-8). The analysis here will investigate what Chevalier's songs, although highly stereotypical and artificial, can tell us about the Parisian vernacular, and whether they had some kind of influence, keeping the vocabulary popularised at the time by Chevalier alive today.

For this study, I compiled a corpus of Maurice Chevalier's best known songs, recorded over almost three decades between the 1920s and 1950s, in order to analyse his use of non-standard vocabulary. Maurice Chevalier is the archetype of popular Parisian singers and his songs have become world-wide classics. In the first part of this analysis, after giving a brief outline of the phonetic features of Chevalier's songs, we will concentrate on the proportion of non-standard items in his songs and follow the evolution of his vocabulary diachronically. The main purpose of this research is to gauge the evolution of Chevalier's lexicon by taking into account the attitudes of lexicographers of Le Petit Larousse (2004) towards non-standard items and the perception of native French speakers towards their usage.

Maurice Chevalier was born in 1888 in Ménilmontant, a working-class district of Paris. His father, a house painter, was Parisian by adoption and his mother was of Belgian origin. After holding various jobs in Paris as a child, such as a circus acrobat and an impersonator of the French Comique-troupier Dranem, he became a "comic" singer in Paris café concerts and later in the Folies Bergères. An admirer of Fréhel and Mistinguett, he moved from a popular repertoire to that of a dandy, clad in a suit and wearing his famous boater. During the années folles, the French singer started a career as a Hollywood actor where he became an ambassador for French culture.

The popular songs of the music-hall made their way onto French radio as early as the 1920s with Aristide Bruant, Monthéus and Fragson. While the female singers (Damia, Fréhel and Marie Dubas) characteristically indulged in melodramatic songs, singing of the misfortunes of prostitutes, drug-addicts and abandoned lovers, the male singers (Mayol, Georgius, Milton) sang farcical songs, with an often purposefully idiotic vocabulary. Since they were masters of parody, manipulators of words, and virtuosos of phraseology, their songs were often mere exercises in elocution which consisted of pronouncing as many words as possible before taking breath. Ostentatiously rejecting grammatical rules, Comique-troupiers, like Chevalier deliberately perpetrated in their songs grammatical and pronunciation errors, the French pronunciation of English words ("Liverpoole", "Novyorke") for example, and excelled in the art of innuendo. Recordings of their songs remain today as some of the oldest archives of Parisian vernacular:

Apart from the 'cris de la rue' of street vendors and conversations recorded on sapphire disc available at the national sound archive of Paris (Antoine/Martin 1985: 298, note 8), recordings which provide information about this period are extremely rare. Therefore, 1930s songs, even if they offer us a pastiche of Parisian speech, are valuable evidence of vernacular speech.

There are two types of songs by Chevalier: some are written in highly literary style with hardly any colloquialisms, while others use colloquial items to create a humorous effect. As the heir of Comique-troupiers such as Ouvrard and Dranem, Chevalier indulged in the 1930s in rather coarse and light-hearted songs ("Le chapeau de zozo", "Prosper" "Valentine") which made him famous as "le p'tit gars de Ménilmuche", then he gradually shifted to a more serious repertoire ("Ça sent si bon la France", "Heure exquise", "Notre espoir") asserting the values of "famille, travail, patrie". Les années folles were a time of extravagance and the unbridled pleasure of intellectual emancipation and linguistic freedom. We find a type of verbal emancipation in les années folles, but "l'état de 'guerre froide' qui instaure sournoisement de nouvelles restrictions, voire de vraies censures par rapport à l'expression verbale de 1944-1955" (Colin 2000: 151) was very quick to establish itself. Like many Parisians at the end of the 19th century (Petitpas 2003), Chevalier favours the abbreviated and resuffixed forms "Pantruche" (1835) for Paris, "Boul' Mich'" (1878) for Boulevard St Michel, "Ménilmuche" (1881) for Ménilmontant and "Lap" for la Rue de la Paix.

2 The evidential value of the data

It goes without saying that songs are not a direct reflection of spontaneous language and are conditioned to a great extent by conventional expectations and stereotyping [1]. Here we make no strong claims about the representativity of this data, since they are a product of a stabilised tradition and a collective creation partly of the lyric writer and partly of the singer. However, though this recorded material is not spontaneous speech, it remains a document of the 1930s-1950s period and is anchored in the linguistic realities of the day. Chevalier's songs give us a subjective view of the objective linguistic reality of that period.

The linguistic analysis of this type of song poses many methodological problems and one has to make some ad hoc choices when compiling a corpus. It is often difficult for instance to establish which singer was the first interpreter of a song. Chevalier commonly sang Trénet's "Y'a d'la joie" and although the lyrics are exactly the same in both versions, Chevalier's interpretation (his tone of voice, the stress and intonation) is very different from Trénet's. We have included a version of Chevalier's "Y'a d'la joie" in the corpus, although the lyrics were not originally written for him. Besides, it can be argued that each interpretation of a song offers a different version of the lyrics. In the version we used of "Valentine", the famous line "elle avait de jolis petits tétons…" becomes "elle avait un si joli piton…". This corpus of texts can consequently be said to be, like the performance of a play, a one-off, stylised representation of the Parisian vernacular at a particular moment in time.

3 The corpus

Our corpus of Chevalier's songs comprises the following 41 songs presented here in alphabetical rather than chronological order, as some dates are unknown. The dates refer to when the songs were recorded, rather than to when they were composed. The total number of words for each song is indicated in parentheses.

4 Phonetic features

In this section, a brief picture will be given of firstly the suprasegmental features recurrent in the corpus and secondly of the segmental features. Maurice Chevalier's songs show a reinforcement and a lengthening of the penultimate syllable. Carton notices that in music-hall songs by Mistinguett or Maurice Chevalier "l'avant dernière syllabe de groupe est longue, intense et monte souvent 'en creux' " (Antoine & Martin 1995: 55). "Monter en creux" means a melodic concave ascent and is the opposite of "en bosse". In French, rising intonations on the accentuated syllable are generally concave. Concavity gives an impression of dawdling (Carton, personal communication, 2000). The example of Chevalier's "ma pômme/ c'est moi" is, as Carton asserts, stereotypical Parisian vernacular with the stress on the penultimate long open vowel [ma'pɔm]. The /mwɑ/ is lengthened to make the voice tremor (ibid.).

The trilled uvular /r/ is another salient feature of Chevalier's songs (Tranel 1987: 141). As Carton has indicated, Maurice Chevalier's /r/ was not a lower-class feature of Ménilmontant, but belonged to the singing tradition and can still be found today in opera singing. Chevalier imitated Georgius "qui avait des /r/ systématiquement 'd'avant' " (Carton 2003, personal communication). The following table presents the various pronunciations of /r/ in a sample of Maurice Chevalier's songs recorded in 1955.

In this corpus of Chevalier's songs, other salient features are the elision of the schwa in "je" as well as the use of the reduced form "y'a" for "il y a", but we will focus here on non-standard lexical items. In numerous films from the 1930s to 1960s (Fric-frac 1939, Touchez pas au grisbi 1954, Du Rififi chez les hommes 1955, Les Tontons flingueurs 1963) which play upon the attractive and mysterious language of criminals in the collective imagination; it is the lexical characteristics which most clearly give speech a slang 'colour' [2]. The same is true of the popular songs of the period.

5 Use of non-standard items in Maurice Chevalier's French songs

By "non-standard", we refer to lexical items which have been labelled fam., pop., arg. etc. by Le Petit Larousse (2004) or which have been excluded from it. The latter words would appear in the following statistics as absent.

In these statistics, 'items' include phrases as well as as individual words. The use of colloquial proper nouns ("Momo", "Mimi" and "Mimile") is very frequent in Chevalier's songs and such examples have been included. We will make no distinction between words and idioms. An idiom like "s'en mettre plein les trous de nez" will count as one item. Further, we have left out onomatopoeic expressions such as "tra la la la", "dzim pa poum pa la" or "ton ton tontaine" which are one-off onomatopoeic creations.

We use Le Petit Larousse's latest edition (2004) as a yardstick, to evaluate the percentage of non-standard items in the corpus. The purpose of this analysis is to investigate lexicographers' perception of these items and their evolution from 1920 to today.

We count tokens rather than types, taking into account several occurrences of a lexical item. Lexical items which appear in a chorus appear in the quantification as many times as they occur, as the lyrics often vary from one chorus to the other and new vocabulary is introduced.

The total of non-standard words in the Chevalier corpus, when one averages the percentage for each individual song, amounts to only 3%. Lexical variables are highly self-conscious and highly salient. Few are needed to create the 'slang effect'. The percentages of non-standard words in each individual song shown below (Figure 1) indicate that the proportion rarely exceeds 20%. The song "Appelez ça comme vous voulez", where Chevalier accumulates instances of colloquial words and expressions, obtains the highest score with 21%, which is a large proportion of the 3% overall figure quoted earlier. Taking into account that there is a predominant percentage of function words (prepositions, conjunctions, articles and interjections for instance), this percentage of non-standard lexical items represents quite a high ratio.

In the following statistics, we have grouped percentages for individual songs in a chronological order and intend to show the progression of slang as Chevalier moves from pre-war to Occupation and to Collaboration. We have classified our sample of Chevalier's songs into three chronological periods: 1920-9, 1930-9, 1940-9. Some of these songs Chevalier sang throughout his career, but by looking at different time periods, we hope to bring out a trend in Chevalier's career.

The results show a slight decline in the use of non-standard items in Chevalier's songs from the 1920s to the 1940s, as he moves from a popular register to a more literary and poetic repertoire.

6 Variety of lexical items

In this section, we examine at the Chevalier song "App'lez ça comme vous voulez" which has the highest number of non-standard words. We have divided the song into its different word-classes (verbs, nouns, adjectives and auxiliaries), grouping together articles, particles, prepositions, conjunctions, clitics and interjections as 'tool-words'. In this quantification, we will consider lexical items as tokens rather than types. If the adjective "petit" occurs twice, this will count as two occurrences.

By means of comparison, we present here results from Müller's 1985 study and show the categories of vocabulary items, with their respective proportions, that can be found in three different corpora:

According to Müller's findings, nouns and verbs are the most frequent word-classes in the three corpora investigated. The percentage of adjectives is high, but only relatively high in Le français fondamental. The number of tool-words only accounts for a small percentage except in statistics for the oral corpus of Le français fondamental where they amount to 17.1%.

Chevalier's songs are, unlike the literary style of Le Monde, quite poor lexically. Some features of spoken French are present such as short and loose constructions and phrasal repetitions. Adjectives are quite basic ("beau", "joli", "rigolo") and their proportion relatively small, and the high frequency of key-words, the majority of which are articles, interjections and onomatopoeic items, also contribute to the spontaneity and musicality of Chevalier's songs. Non-standard nouns build up, like an exercise in style or an elocution lesson, as in Vincent Scotto's song, sung by Ouvrard:

In this way, Ouvrard, through a succession of nouns and verbs, builds up the anatomical and medical references. We can note that the words generally used in Chevalier's choruses are usually of three syllables or less:

Alliterations using /v/, /p/ and /o/ in Le Chapeau de zozo serve to add to the onomatopoeic and speech-like character of the chorus. The nouns give a basic picture and provide a sketch of the characters involved. In Y'a d'la joie, the lexemes "joie", "toit" and "hirondelles", "soleil", "ruelles", "demoiselles" provide the framework and bring out the spirit of the song whose main character is the singing "moi", "je", or "Maurice". The verbs make the nouns dynamic ("bat", "chavire" and "chancelle").

7 Vitality of non-standard items found in Chevalier's songs in today's French

In this section, we investigate attitudes towards the non-standard items used by Maurice Chevalier implicit in Le Petit Larousse's (2004) treatment of these words (see glossary) and in judgments expressed by a sample of Parisian speakers in 2004. In the study I conducted on non-standard vocabulary in the 1930s film corpus (Abecassis 2000), I pointed out that many social markers have gradually become stylistic markers and have merged into French speakers' linguistic passive repertoires (Posner 1997: 74): even though socio-stylistic variation operates concurrently, the lexis which constitutes the "doublet parasite du standard" (Gadet 2003: 110) has lost part of its sociological dimension. Our findings indicated that non-standard words found in 1930s dialogues were still commonly used, regardless of social status. From these statistics, we evaluate the extent to which items in the Chevalier corpus have disappeared. There seems to be in the language of French youngsters a resurgence of colloquial words through the media and most particularly in songs:

If we agree that television and the media in general have contributed to the increase in formulae and linguistic traits (Bernet 2000: 191), it would not be surprising if songs have been the source of lexical innovations or have spread out-dated popular formulae. It is noticeable that archaic lexical items originating from traditional argot often emerge among the young (Gadet and Conein 1998: 115-116). The singer Renaud, in his 1975 album, reintroduces the terms "Paname" (argot for Paris), "aminche" and "gavroche". Words like "maille", "flouze" and "poteau" are also frequently found in lyrics of rap songs popular in the 21st century. The surveys conducted by Walter between 1987 and 1989 (Walter 1991) show that more than half of the argot words originate from the specialised texts investigated and date back to before 1945 (Colin 2000: 168). As for the non-standard lexical items in the Chevalier corpus, we do not see them as items which have died out and which youth has recently resuscitated, but as items which have always been in the passive linguistic repertoire, absent from public notice and then brought back into fashion by an influential song or film. The verlan 'ripoux', for instance, re-emerged in the public arena when popularised by Claude Zidi's film in the 1980s and its numerous sequels.

7.1 Speaker variables

Informants were asked firstly whether they were familiar with a list of non-standard words extracted from the Chevalier corpus and secondly whether they would use these words themselves. The left-hand column in table 3 shows the percentage of items unknown to the informants, the second presents the percentage actually used and the third the percentage of lexemes the informants are familiar with but do not think that they use. Most of the informants were contacted by e-mail and were unknown to the interviewer, which restricted the sample to those who were computer literate. The informants fall into five broad age categories: a) 15-20, b) 21-30, c) 31-40, d) 41-50 and e) 50+. Considering a sample of twenty educated Parisian speakers in each age bracket is a satisfactory number and should be expected to produce representative results. Moreover, by focusing on informants selected according to the sociolinguistic axes along which language variation operates (age, sex and class), we have tried to establish whether these items have become less stigmatised and become part of their everyday natural speech. Armstrong and Hogg (2000) have conducted a similar study on the use of non-standard lexis which they corroborated by basic statistic tests like ANOVA to see whether the groups are in fact behaving like groups.

Table 5 shows differentiation between the male and female speakers' use of these non-standard items. Only an average 53% of traditional slang found in the Chevalier corpus is recognised by the informants to be part of their general vocabulary. The remaining 47% is either unknown or unused by them. Table 6 and Figure 2 show considerable variation in the perceived use of the vocabulary according to age and gender parameters.

When we compare the type of slang used by Maurice Chevalier with that found in 1930s gangster films (Abecassis 2000), fewer lexical items in Chevalier's songs have survived in common usage. Column 2 of Table 5 shows that percentages of unknown items vary between 5% to 42%.

Column 3 shows clear sex differences in all age groups where the percentage of use of non-standard forms by male speakers is relatively higher than that of the female informants. Female informants still achieve quite high scores. These results comply with the "sociolinguistic gender pattern" illustrated by Milroy (1987) and Armstrong/Hogg (2000). The results in the 15-19 age-group indicate quite a high percentage of use of non-standard items among the youth. The questionnaire results also testify that the percentage of use of non-standard lexis increases with the age of the informants reaching its peak among the 50+ cohort. However, over the whole range, the 40-49 cohort was born between the mid-50s and 60s, at a time when popular French music experienced some major changes and was largely influenced by America.

Column 4 shows that on the whole, the female informants present through these results a more conservative image, as the figures demonstrate that they know a higher number of the investigated items but would not use them. This indicates a high degree of awareness on the stylistic value attached to this sample of Chevalier's lexis.

The fact that older informants rather than the younger generations were exposed to some of the cryptic slang of Chevalier ("calter", "un bibi", "un zozo", "un piton") is rather unsurprising. Chevalier's heyday, both as singer and actor, was between the 1930s and the 1960s. In 1958, he made the film Gigi and that year received an Oscar. In the 1970s, in spite of singing the title song of Walt Disney's Aristocats, he became less well-known among adolescents and fell out of fashion in favour of other chansonniers (Brel, Brassens, and Bécaud among others). Another related factor should be taken into account: the disappearance of some lexical items popularised by Chevalier. Young people's culture, concerns and topics of conversation have shifted, and traditional slang associated with drinking, love and the milieu of gangsters are no longer representative of the social persona of the youngsters of the time. The class structures of the 1930s have been superseded by new ones. Nonetheless, the percentage of use by teenagers is still relatively high, showing a resurgence of some of these words. Quite interestingly, the percentage of lexical items they claim to know but do not use is extremely high. Is it mere boastfulness from the younger informants to argue so or is this vocabulary really spreading in the language of young people? The sample presented is not sufficiently representative to confirm the exact tendencies of youth behaviour in this regard. From 50+ speakers to 15-19 informants there is clearly, as shown in table 6, a progressive decline in the use of slang.

The high frequency of non-standard items in female speech in the data can be interpreted in various ways. Chevalier's songs have always been particularly popular among women. His accent and charisma made him the epitome of the French lover in France as in America. It could also be that the males questioned use a different colloquial vocabulary, drawing on other sources such as verlan, thereby appearing to diverge strongly from Chevalier's findings. This would mean that, contrary to the results, the languages of females may be a lot more conservative than that of their male counterparts. However, as Holmes (1997) remarks "no satisfactory explanation has emerged of why women should orient more readily than men to a prestige norm" (Quoted in Milroy and Gordon, p. 101) and we suggest that we might be dealing here with a process of "social re-valuation" (Ibid.: 103). Female speakers favour stigmatised varieties and subvert traditional female roles by means of a familiar usage, enabling a more cohesive mentality, marked by solidarity and group identity, in a society which is increasingly liberal. The real innovation in youth language is the emancipation of female language both among educated speakers, such as those in our survey, and in less privileged groups. Popular features and cheeky humour have often been stereotypically associated with young men who frequent public places like bistros (Gadet 2003: 206), whereas women are considered more likely to remain within the home, and, under parental authority, conform linguistically to prestige varieties. In the 21^st century, the use of a non-standard lexis is no longer prompted by the desire to appear "male", as connections between men and women become more and more subtle, owing to equality between sexes and shared vocabulary.

7.2 Stylistic variables

In a second stage of this survey, we asked our informants to rate the non-standard items of the Chevalier corpus to establish whether their views differ from those of lexicographers. The following table shows the proportions of items in the song corpus rated as fam., arg. and abs. by our reference dictionary (Le Petit Larousse 2004).

The majority of the items investigated are labelled by Le Petit Larousse as fam. There is a small number of argot words, but none is rated as populaire. This suggests that formerly pop. words have gradually become fam. in the eyes of lexicographers. 27% of these items are exluded from the dictionary, which either implies that they have not survived in modern French or that they are still heavily stigmatised.

To obtain the following statistics, we asked our informants to rate each lexical item using the labels fam., standard or archaic. We are aware that this is a simplification of the stylistic continuum used by lexicographers and other labels could be attributed to the items of our corpus such as pop., arg. or both fam. and archaic. However, for the purpose of this quantification, this enables us to gauge whether the people questioned thought that these items have merged into the average French person's linguistic repertoire and have become standard or whether they are considered to be obsolete.

As one can see, the informants rate a large percentage (47%) of these items as fam., which is very close to the attitude of lexicographers. However, their view is less stigmatising as they consider that 24% of these items have become standard and 20% are now obsolete.

Linguists have attempted to identify the reasons why argot words should remain current in French, and the growing number of dictionaries of argot (Bernet 2000: 174) shows the French interest in these curiosities which have almost reached the status of national cultural heritage. We only have to think of clichés and other quirky lexical features which have been collected together. There is indeed an oral tradition which passes from generation to generation and which sometimes goes underground for a generation, to become current again much later, as in Renaud's songs. Slang is an identity marker and the more it is generalized, the less useful it is, but if in today's French it has a tendency to spread through all levels of society, this is as Colin (2000: 160) has shown, because of a number of factors. Our language is becoming increasingly technical and the intrusion of information technology and types of jargons associated with technical specialized professions has a tendency to crop up in everyday language (e.g. "CD ROM", "digitaliser", "startup"). There is not simply the ludic pleasure of playing with words demonstrated by "bons mots", irony and self-mockery, puns and other rhetorical devices used by many French humorists (Pierre Desproges, Karl Zéro, Vincent Roca, Guy Carlier), heirs to cabaret singers, who find the mark of a group identity in the use of argot, and in this way confine themselves to the outskirts of society, because it is primarily the desire to upset, provoke or shock which keeps argot active. Roca's witticism for his satirical TV programme 'Sur le fil dérisoire' (2004) is only one example of this continual seach for a play on words combining spirituality and double-entendre. Some radio stations (NRJ, Fun Radio) and the cable television channel Canal +, in seeking to convey a young, lively image, use a subversive vocabulary. This desire to veil the message is a form of rejection of all forms of authority, both the symbolic rejection of the constraints of parental authority and good linguistic manners imposed by the norm, and the refusal to conform to rules. The overuse of the adjective "cool" by the youth today pinpoints a state of mind and the need to project an image of protest against the propriety imposed by society. According to Calvet's formulation, argot "est une façon de se situer […] une façon de revendiquer son appartenance à un groupe social, à un lieu ou à une classe d'âge" (1994: 115). In the 1930s, the effect sought by Maurice Chevalier was the same. He frequently used popular expressions for humorous effect and increased his notoriety with risqué and politically incorrect words.

What will become of Maurice Chevalier's argot? It seems, contrary to the suggestion of Colin (2000: 155), that classical argot is not dead. Indeed, it appears that one does not speak of argot anymore but of argots in the plural. Both a small death and a renaissance are going on simultaneously. Verlan for example is a social and geographical marker (commonly heard in the Parisian suburbs, but only very rare in the South of France). Certain frequent words however ("keuf", "meuf" etc.) are integrated into the language and used by various social groups. As the statistics have shown, Chevalier's argot is still in use despite strong competition: it is putting up a strong defence against verlan and forms from all sides, from English to Arabic, and Antillais and Romany languages (Gadet 2003: 2004) which young people pick up and which are disseminated by the media through television. Our survey indicates, fairly unsurprisingly, clear sex differences with male informants using proportionately more non-standard items and shows that 42% of the non-standard terms in the Chevalier's corpus are still used by the informants. The informants' judgements on the stylistic appropriacy of these items demonstrate they are becoming more uniform, although around 20% are considered to be archaic or outdated. The currency of non-standard lexis can ebb and flow in an unpredictable way, probably as a result of the media. For Bernet, "le lexique évolue de manière imprévisible à long terme et rien ne permet de dégager avec certitude des tendances qui pourraient préfigurer l'avenir" (Bernet 2000: 194). Recently, the singer Patrick Bruel revitalised the songs of 1930s with his album "Entre-deux" (2003) and it would not take much for the 1930s vocabulary to come back into fashion.

The following glossary gives a representative sample of non-standard items in the Chevalier corpus with their translation in English and their acceptance in Le Petit Larousse (2004).

Notes

* I am very grateful to Professor Lodge for his numerous comments and Dr W. J. Anderson for her help in proofreading. [back]

1 Stereotypes are highly stigmatised variables recognised by a particular group as mistakes or affected manners of speech (adapted from Baylon 1991: 91-2 and Wardhaugh, 1986: 142). Note that "stereotypes" could be understood in the Labovian sense, as defined above, or in a broader sense as "value judgments about what [the lay-person] think[s] is and is not correct" (Lodge et al. 1997: 3) about a language as well as 'evaluations of speakers' (Wardhaugh 1986: 113). [back]

2 In Fric-Frac, the representation which scriptwriters and actors give of the Parisian vernacular is mostly phonetic and lexical. In films from the 1950s, the emphasis is on lexical features. In Du Rififi chez les hommes, for instance, the speech of gangsters is full with non-standard items. Les Tontons flingueurs is not so much peppered with argot words, as with one-off lexical creations ("Ton Antoine commence à me les briser menu"; "Moi quand on m'en fait trop, j'correctionne plus, je dynamite, je dépense, je ventile"). [back]

References

Abecassis Michaël (2000): "A diachronic analysis of non-standard vocabulary in the dialogues of 1930s French films". Te Reo: Journal of the Linguistic Society of New Zealand 43: 21-37.

Antoine, Gérald/Cerquiglini Bernard (2000): Histoire de la langue française 1945-2000. Paris.

Antoine, Gérald/Martin, Robert (eds.) (1995) Histoire de la langue française (1914-1945). Paris: 191-206.

Armstrong, Nigel/Hogg, Clare (2001): "The Pollyanna principle in French : a study of variable lexis". Journal of Pragmatics 33, 11: 1757-1785.

Armstrong, Nigel/Unsworth, Sharon (1999): "Sociolinguistic Variation in Southern French Schwa". Linguistics 37,1: 127-155.

Bauer, Laurie (1994): Watching English Language: An Introduction to the Study of Linguistic Change in Standard Englishes in the Twentieth Century. New York.

Baylon, Christian (1991): Sociolinguistique: Société, langue et discours. Paris.

Bernet, Charles (1995): "Le français familier et populaire à la radio et au cinéma". In: Antoine, Gérald/Martin, Robert (eds.) (1995): Histoire de la langue française (1914-1945). Paris: 191-206.

Bernet, Charles (2000): "Usage et marges du lexique français". In: Antoine Gérald/Cerquiglini Bernard (eds.): Histoire de la langue française (1945-2000). Paris: 152-172.

Carton, Fernand (1995): "La prononciation du français". In Antoine, Gérald/Martin, Robert (eds.): Histoire de la langue française (1914-1945). Paris: 27-52.

Chambers, Jack K. (2003): Sociolinguistic theory: Linguistic Variation and its Social Significance. Oxford,/Cambridge.

Chambers, Jack K. (1998): "Myth 15: TV makes people sound the same" In: Bauer L./ Trudgill P.(eds): Language Myths. Harmondsworth. Penguin books: 123-31.

Colin, Jean-Paul (2000): "Nouvelles pratiques langagières: les argots". In: Antoine, Gérald/Cerquiglini, Bernard (eds): Histoire de la langue française (1945-2000). Paris: 152-172.

Françoise, Densie (1985): " Les argots ". In: Antoine Gérald/Martin, Robert (ed.): Histoire de la langue française (1880-1914). Paris: 295-327.

Françoise, Gadet (1989): Le Français ordinaire. Paris.

Françoise, Gadet (1992): Le Français populaire. Collection Que Sais-Je? 1172. Paris.

Françoise, Gadet (2003): "'Français populaire' : un classificateur déclassant ?". Marges linguistiques 6:103-115.

Françoise, Gadet/Conein, Bernard (1998): " 'Le 'français populaire' de jeunes de la banlieue parisienne entre permanence et innovation". In: Androutsopoulos, Jannis K./Scholz, Arno (eds.): Jugendsprache - langue des jeunes - Youth Language. Frankfurt am Main/New York: 105-123.

Goose, André (2000): "Evolution de la Syntaxe". In: Antoine, Gérald/Cerquiglini, Bernard (eds): Histoire de la langue française (1945-2000). Paris: 105-145.

Goudailler, Jean-Pierre (2000): "Comment le céfran des cités a fait souffler un véritable air de jeunesse sur le français de papa". Le Nouvel Observateur Hors série 41: available at http://www.nouvelobs.com/hs-lesados/instruire/art8.html

Guiraud, Pierre (1965): Le Français populaire. Collection Que Sais-Je? 1172. Paris.

Holmes, Janet (1997): "Setting new standards: sound change and gender in New Zealand English" English World-Wide 18,1: 107-142.

Labov, William (1984): "The transmission of linguistic traits across and within communities". Paper presented at the Symposium on Language Transmission and Change, Center for Advanced Study in the Behavioral Sciences, Stanford University, California.

Lieberson, Stanley (2000): A Matter of Taste: How Names, Fashions and Cultures. Yale.

Lodge, R. Anthony/Armstrong, Nigel/ Ellis, Yvette M.L./Shelton, Jane F. (1997): Exploring the French Language. London.

Milroy, Lesley (1987): Language and Social Networks. Oxford.

Milroy, Lesley/Gordon, Matthew (2003): Sociolinguistics: Methods and Interpretation. Blackwell.

Müller, Bodo (1985): Le Français d'aujourd'hui. Paris.

Petitpas, Thierry (2003): "De Paname à Ripa: histoire d'une rupture" Marges linguistiques 6: 145-155.

Posner, Rebecca (1997): Linguistic Change in French Language. Oxford: Clarendon Press.

Szabó, Dávid (1998): "Methodological Problems of the Sociolinguistic Study of Youth Slang". In: Androutsopoulos, Jannis K./Scholz, Arno (eds.): Jugendsprache - langue des jeunes - Youth Language. Frankfurt am Main/New York: 48-57.

Tranel, Bernard (1987): The Sounds of French. Cambridge.

Walter, Henriette (1991): "Où commencent les innovations lexicales". Langue française: 53-64.

Wardhaugh, Ronald (1986/1992): An Introduction to Sociolinguistics. Oxford.