4K
.41:

In J. .I..‘..JZ
In! it: 2......

a "a. .harns J. x
. . . ..
a}? £313.: {haw tr
{$3.152 £35. .5
It "I 30.11!!!
_F»n...\§fi
z,!:i:cis All

1.... .. .. i...
.5139“ “as... ‘
4.1 .
.r ‘
‘4 .‘u

.e . 3.. .2
53......

2.3.... .3.

3...

mun"

In... an;

[li‘lanr-i
5&9. r.
:

 

 

 

...L... a .m.‘ .

an r
63

m

l4:

 

 

 

 

“my LlBRARY l
. - .-i,._ l
,7 ‘1 ., Michigan c. Late
,5. 1) i Universuty
This is to certify that the

dissertation entitled

A CURRENT STUDY OF BANLIEUE LANGUAGE IN
THE PARISIAN SUBURBS

presented by

Teresa L. Kent

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in French Language and
Literature

 

 

l‘le
W

ignature

EbCSEQCff (25 9g; 3

Date

MSU is an Affinnative Action/Equal Opportunity Employer

 

 

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K lProj/Acc8Pres/CIRCIDateDue. indd

 

 

 

in'

A CURRENT STUDY OF BANLIEUE LANGUAGE
IN THE PARISIAN SUBURBS
By

Teresa L. Kent

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
French, Language and Literature

2009

Farr;
fin- -_
*w- lei:
.AA
"' WCW

ate

‘M
”I.
U

U

Q!

Bub v u-
3"

menu
6

‘- 1v-
U

VII!

:9
II
C

”'3‘

c u r Qx ad ..i.\ C
n. 3 la Rue MW.» n2» Wu 8
we E an M. :1 3w .Mw mm
.. V
.r. 0,. w . .u w... 3.. a» we.
... .2. my... . f. .w. a. .. w... mu.
. .. n I u u, n I .- l ti .1-

ABSTRACT

A CURRENT STUDY OF BANLIEUE LANGUAGE
IN THE PARISIAN SUBURBS

By

Teresa L. Kent

This study discusses the banlieue language in an innovative and multi-
fronted manner. It seeks to challenge many theories in research on banlieue
language, especially stereotypes regarding its speaker base. By expanding
current research may be expanded to include more sociolinguistic factors than
ever before, we can begin to see that banlieue language is changing not just in
form but also in number of speakers.

More specifically, this study examines current stereotypes of gender, age,
education level, and ethnicity of speakers of the variety as well and also
proposes new categories of sociolinguistic factors that may affect use of
banlieue language. The main ideas include these hypotheses:

1. Banlieue language is spoken by people older than any other previous

research has indicated.

2. Gender is not as male-oriented as previous studies may have suggested.

3. Regardless of stereotypes and research that relegate the deepest vernaculars
to those speakers of the lowest levels of education, that banlieue language,
because of the function of identity, may also be spoken by those with much

higher levels of education than previously expected.

7’5. 533‘) 3‘53
92': 3t“ ““9"“ ‘

'da:{*. ‘a’IIE‘ :
Btu 36 I “T

This study also addresses two never-before-investigated areas: the
speakers' own views on who uses banlieue language and the effect of religion

on speaking banlieue language.

xgmwmmnu
t’tehancb
in; ever. from
':'.‘r Anna Norr
smtomeen

Tcmylnen

33% Elise. Pie

N -
ie in France

ACKNOWLEDGMENTS

This dissertation is in memory of ‘Stoon,’ Sebastien de la Place, a friend
whose life tragically ended at far too early an age. A core member of the group
of friends who make up much of the older group in my study, Stoon was one of
the many inspirations for this dissertation, causing me to question the
stereotypes of banlieue language. Though truly a product of the banlieue, he
received an education and became a police officer, but he still spoke banlieue
language in its deepest vernaculars, despite his age and employment.

I would also like to thank some members of the MSU community. First of
all, utmost appreciation goes to my advisor, Dr. Anne Violin-Wigent, for her
guidance and understanding throughout the doctoral experience. Without her
endless patience and support, negotiating the statistical rapids of the thesis
would have been impossible. To Dr. Dennis Preston, I am eternally grateful for
his weekly FOS (Friends of Sociolinguistics) sessions in which each week the
discussion and instruction he brought us, in every facet of the discipline, taught
me more than classes ever could have. His adoption of anyone with interest in
socio, even from other departments, exhibits a collegiality that is hard to equal.
To Dr. Anna Norris, I am thankful for the open door to her office, a place to just
slump into the empty chair in the corner for a therapeutic chat.

To my friends in France, Jenny, Gaél, Véro, Denis, Melanie, David,

Jérome, Elise, Piq, and others, thank you for adopting me as one of your own.

My time in France was wonderful because of you. Jenny, you are truly part of my

at “Du 35555-3":
. \

EFIE 43:2 To J

533.37 3": ‘Dl 1‘

. i
1'35 31: 5:371:

um ! "t“ "1:
1%:' U“ 3 5‘ 'U
1

rr ‘3'3'59‘1', for ‘

nhv our V' '
u

g:- DOA” ‘34 ‘f
M :uu‘b'd s6 'U ' A

I
“,W‘fl'g ,._r A an
N a {DUI VC‘QF

’* :H'fmn
"UUJ :a an

00: ’55” A.‘ .e' T?

" "VF-f . u'l'
1" h AA
. was you f:

T'c-M
“Va-iv; I.

finally. I wg

I "l "trr. for fit

family. Your assistance in so many areas of life in France and with the thesis
was invaluable. To Vincent, thank you for your help in many facets of this
dissertation and for introducing me to your language, culture and friends. To
Andrée and Jean-Louis, my French parents, thank you for running me around
endlessly during the data gathering phase, as well as for all your help and
encouragement for the many years I have known you. To Chantal, thank you for
your support and friendship, not to mention the miracles you performed to get
me into some of the venues. To those at the Mission Locale, thank you so very
much for your cooperation. Yours is a truly wonderful organization.

To Donna and Stacy, thanks for being there through it all. To Richard,
your support over the last year and a half made finishing this possible. To Jenn
McNeil, bless you for your patience and hard work editing. You saved me months
of frustration.

Finally, I would like to thank my parents, who have always been there for

me; my mom for fighting her cancer so valiantly and my dad for taking care of us

all.

TABLE OF CONTENTS

TABLE OF CONTENTS ....................................................................................... vi
LIST OF TABLES ................................................................................................ xii
LIST OF FIGURES ............................................................................................ xiv
CHAPTER 1: INTRODUCTION ............................................................................ 1
1.0 Introduction ............................................................................................... 1
1.1. Historical Background to French language variation ............................... 2
1.2. Breaking the norm ................................................................................... 6
1.3 Definitions ................................................................................................. 8
1.3.1 Argot ............................................................................................... 9
1.3.2 Banlieue .......................................................................................... 9
1.3.3 Banlieue language ........................................................................ 10
1.3.4 Dialect ........................................................................................... 10
1.3.5 Jargon ........................................................................................... 11
1.3.6 Variety ........................................................................................... 11
1.3.7 Verlan ............................................................................................ 12
1.4 Goals ...................................................................................................... 12
1.4.1 Age ................................................................................................ 12
1.4.2 Education ...................................................................................... 13
1.4.3 Gender .......................................................................................... 14
1.4.4 Ethnicity ......................................................................................... 15
1.4.6 Religion ......................................................................................... 16
1.4.7 Languages .................................................................................... 16
1.5 Hypotheses ............................................................................................. 17
1.5.1 Age ................................................................................................ 17
1. 5.2 Education ...................................................................................... 17
1. 5.3 Gender .......................................................................................... 17
1.5.4 Ethnicity ......................................................................................... 18
1. 5. 5 Religion ......................................................................................... 18
1.5.6 Languages .................................................................................... 18
1.6 Dissertation Overview ............................................................................. 18
CHAPTER 2: PREVIOUS RESEARCH .............................................................. 20
2.1 Urban languages .................................................................................... 20
2.1.1 Language contact .......................................................................... 20
2.1.2 City melting pot erodes other languages ....................................... 21
2.1.3 A new shift of function ......... . ......................................................... 22
2.1.4 General emergence of urban languages ....................................... 23
2.1.4.1 Examples of emerging language contact varieties ............... 24
2.1.4.1.1 The case of Sheng and Engsh .................................. 24

vi

21 4 1.:

2.1 4 i .
23.13399 a'ld K
22? L363”???
222 :n-grouQ—’
233 foertfcaf
224163.792"-

nn' - “Ah‘-fl
'0‘ b—
ou|Pavkvrs a UVU'

.3-
as

‘ .

h-
c-) (g: .‘A'
(1) (A, R) -—~

(1)

'< (J )
é) (.9
-:~,. ('1

' L

T (b

'a.’
N V'\ ‘l

in

'3

(_
u)
‘LJ
(0
m

‘

o

337166;:an . .
Zer'eécpmeni l
22.ex:cai terns
:1 1%.,

. :rfl
“Iv uu‘V'lI'al'C

iiiDemog'
332 W360 I15
illuies .........
3.4.1L'unrre'
34.2LaMss
3.4.3Pri.ale
3'5 Ifmanl has
35.1 Gende'
3.5.2. Nation
353Emnmn
3.549eiigro
ISSDepart
3.5.6Age
3.5.7Educa'
3.5.8 Langu.
3.5.8.1l
3.5.8.2:
3.5.8.3
. 3.5.9Profe:
II SIaIlSIlcaI a
3.6.1 Systa
3.62 Gold
3.6.21
3.62.2
3.62 '
3.6.2.
3.6.2

3'61th RE

 

2.1.4.1.2 The case of Camfranglais/Francamglais .................... 25

2.1.4.1.3 The case of Dutch straattaal (‘street language’) ......... 27

2.2 Language and identity ............................................................................ 27

2.2.1 Language change as it relates to identity ...................................... 27

2.2.2 ln-group/ out—group behavior ......................................................... 28

2. 2. 3 Identification with the environment ................................................ 29

2. 2.4 Maghrebi connection ..................................................................... 31

2.3 Factors affecting the use of banlieue language ...................................... 33

2.3.1 Age ................................................................................................ 33

2.3.2 Gender .......................................................................................... 35

2.3.3 Youth language ............................................................................. 36

2.4 Language attitudes and perceptions of speakers ................................... 37

CHAPTER 3: METHODOLOGY .......................................................................... 40

3.0 Introduction ............................................................................................. 40

3.1 Development of the questionnaire .......................................................... 40

3.2 Lexical items ........................................................................................... 41

3.3 Questionnaire format .............................................................................. 42

3.3.1 Demographic information .............................................................. 42

3. 3. 2 Word list ........................................................................................ 44

3.4 Venues ................................................................................................... 47

3.4.1 L’université de Saint Denis ............................................................ 47

3.4.2 La Mission Locale ......................................................................... 48

3.4.3 Private holiday gatherings ............................................................. 49

3.5 lnforrnant base ........................................................................................ 50

3.5.1 Gender .......................................................................................... 50

3.5.2. Nationality .................................................................................... 50

3.5.3 Ethnicity ......................................................................................... 51

3. 5. 4 Religion ......................................................................................... 51

3. 5. 5 Department of Residency .............................................................. 51

3. 5. 6 Age ................................................................................................ 52

3. 5. 7 Education ...................................................................................... 52

3.5.8 Languages .................................................................................... 53

3.5.8.1 First language ...................................................................... 54

3. 5. 8. 2 Second language ................................................................. 54

3. 5. 8.3 Third language ..................................................................... 54

3. 5. 9 Professions ................................................................................... 55

3.6 Statistical analysis methodology ............................................................. 55

3.6.1 Systat ............................................................................................ 55

3. 6.2 Goldvarb ........................................................................................ 56

3.6.2.1 Age ...................................................................................... 57

3.6.2.2 Religion ................................................................................ 58

3.6.2.3 Ethnicity ............................................................................... 58

3.6.2.4 Education ............................................................................. 59

3.6.2.5 Languages ........................................................................... 59

CHAPTER 4: RESULTS ..................................................................................... 61
vii

" “auction of res
1.13155. of indz‘ric
4‘. 1 Fiona
11.111452
4.1.1.11
41112
4.1115
4.1.1.2010:
4.1.1.2‘
4.1.1.2.}
4.1.1.2..
1.1.1.3891
4.11.3
411.3
4113
411.433,
41.14
4.114
4114
411.533.
4.115
4115
4.1.1:
4.1.1.558
4.1.1.1
4.1.1.1
411
4.1.1.75.
4.1.1
4.1.1
4.1.1
4.1.1.85
4.1 ‘
4.1
4.1
4.119
4.1.2 Clear
41.21
4.1
4'4.

 

 

 

 
   

4.0 Introduction of results ............................................................................. 61

4.1 Analysis of individual lexical items .......................................................... 61
4.1.1 Roma ............................................................................................ 62
4.1.1.1 Marave ................................................................................. 62
4.1.1.1.1 Gender ....................................................................... 63
4.1.1.1.2 Education ................................................................... 63
4.1.1. 1. 3Age ............................................................................. 64
4.1.1.2 Chourave ............................................................................. 65
4.1.1.2.1 Gender ....................................................................... 66
4. 1.1.2.2 Education ................................................................... 66
4.1.1. 2. 3 Age ............................................................................. 66
4.1.1.3 Se (la) natchave ................................................................... 67
4.1. 1 3.1 Gender ....................................................................... 68
4.1. 1. 3. 2 Education ................................................................... 68
4.1. 1. 3. 3 Age ............................................................................. 68
4. 1.1.4 Schmidts .............................................................................. 69
4.1.1.4.1 Gender ....................................................................... 69
4. 1. 1. 4. 2 Education ................................................................... 7O
4.1.1.4.3 Age ............................................................................. 70
4.1.1.5 Gadjo ................................................................................... 71
4.1.1.5.1 Gender ....................................................................... 71
4.1.1 5. 2 Education ................................................................... 71
4.1 1.5.3 Age ............................................................................. 71
4 1.1.6 Gadgi ................................................................................... 72
4.1.1.6.1 Gender ....................................................................... 72
4.1. 1. 6. 2 Education ................................................................... 73
4.1 1 6.3 Age ............................................................................. 73
4.1.1.7 Bédave ................................................................................ 74
4.1.1. 7.1 Gender ....................................................................... 74
4. 1. 1. 7. 2 Education ................................................................... 74
4.1 1 7.3 Age ............................................................................. 75
4.1.1.8 Bédo .................................................................................... 76
4.1.1.8.1 Gender ....................................................................... 76
4.1. 1. 8. 2 Education ................................................................... 76
4. 1.1.8.3 Age ............................................................................. 77
4.1.1.9 Roma analysis ..................................................................... 78
4.1.2 Creole ........................................................................................... 81
4.1.2.1 Goumer ................................................................................ 81
4.1.2.1. 1 Gender ....................................................................... 82
4.1.2.1. 2 Education ................................................................... 82
4.1.2.1.3 Age ............................................................................. 82
4.1.2.2 Sa ou fé ............................................................................... 83
4.1.2.2.1 Gender ....................................................................... 83
4.1.2.2.2 Education ................................................................... 83
4.1.2.2.3 Age ............................................................................. 84
4.1.2.3 Sa ka maché ......................... . .............................................. 85
viii

41.23
41:23
41.25
4.124.
412.1
41.2
412
4125 C1
”Arabic
4.13.1K
413
4.13
413
41.32C
41.1
41.:
4.11
4133:
41..
4,1,
41.
4.1.34
4.1411643:
4.1.41

4.1.2.3.1 Gender ....................................................................... 85

4.1.2.3.2 Education ................................................................... 85
4.1.2.3.3 Age ............................................................................. 85
4.1.2.4 Makoumé ............................................................................. 86
4.1.2.4.1 Gender ....................................................................... 86

4. 1.2. 4.2 Education ................................................................... 87
4.1.2.4.3 Age ............................................................................. 87
4.1.2.5 Creole analysis .................................................................... 87
4. 1.3 Arabic ............................................................................................ 89
4.1.3.1 Kiffer .................................................................................... 89
4.1.3.1.1 Gender ....................................................................... 9O
4.1.3.1.2 Education ................................................................... 90
4.1.3.1.3 Age ............................................................................. 90
4.1.3.2 Choucrane ........................................................................... 91
4.1.3.2.1 Gender ....................................................................... 91
4.1.3.2.2 Education ................................................................... 91
4.1.3.2.3 Age ............................................................................. 92
4.1.3.3 Ouaiche ............................................................................... 93
4.1.3.3.1 Gender ....................................................................... 93
4.1.3.3.2 Education ................................................................... 93
4.1.3.3.3 Age ............................................................................. 94
4.1.3.4 Arabic analysis ..................................................................... 94
4.1.4 Verian ............................................................................................ 96
4.1.4.1 Keum ................................................................................... 96
4.1.4.1.1 Gender ....................................................................... 97
4.1.4.1.2 Education ................................................................... 97
4.1.4.1.3 Age ............................................................................. 97
4.1.4.2 Sky ....................................................................................... 98
4.1.4.2.1 Gender ....................................................................... 99
4.1.4.2.2 Education ................................................................... 99
4.1.4.2.3 Age ............................................................................. 99
4.1.4.3 Verlan analysis .................................................................. 100
4.1.5 Argot ........................................................................................... 101
4.1.5.1 Binouse .............................................................................. 101
4.1.5.1.1 Gender ..................................................................... 102
4.1.5.1.2 Education ................................................................. 102
4.1.5.1.3 Age ........................................................................... 102
4.1.5.2 Daron ................................................................................. 103
4.1.5.2.1 Gender ..................................................................... 103

4. 1.5.2.2 Education ................................................................. 103
4.1.5.2.3 Age ........................................................................... 104
4.1.5.3 Daronne ............................................................................. 105
4.1.5.3.1 Gender ..................................................................... 105
4.1.5.3.2 Education ................................................................. 105
4.1.5.3.3 Age ........................................................................... 106
4.1.5.4 Savater .............................................................................. 107

4.1.5 4
4.1.54

41 5 4
4,161.911081'
1.2 Etta? Resdts
421 Gender
4.21 1 PE
4212 A;
421 3 Ct
4225316553:

422.1136

4.2 3 Age .,
4.2 3.1 Pi
4.23.3 C

4.2 4 Langua

425 Adm'fte

42.6 Venue

4.2.7 Overal:

‘3 3318133111,! 5

 

4.1.5.4.1 Gender ..................................................................... 107

4.1.5.4.2 Education ................................................................. 108
4.1.5.4.3 Age ........................................................................... 108
4.1.6 Lexical items ............................................................................... 111
4.2 Global Results ...................................................................................... 114
4.2.1 Gender ........................................................................................ 116
4.2.1.1 Perceived gender usage by gender of informant ............... 1 16
4.2.1.2 Admitted usage by gender of informant ............................. 1 19
4.2.1.3 Comparative analysis ........................................................ 120
4.2.2 Education .................................................................................... 123
4.2.2.1 Perceived education-level usage by education of informant
....................................................................................................... 1 25
4.2.2.2 Admitted usage by education of informant ......................... 129
4.2.2.3 Comparative analysis ........................................................ 130
4.2.3 Age .............................................................................................. 132
4.2.3.1 Perceived usage by age of informant ................................ 133
4.2.3.3 Comparative Analysis ........................................................ 140
4.2.4 Language of the lexical item ........................................................ 143
4.2.5 Admitted usage by the original language of lexical item .............. 143
4.2.6 Venue .......................................................................................... 148
4.2.7 Overall statistics analysis ............................................................ 151
4.3 Statistically significant factors influencing knowledge of lexical items ..154
4.3.1 Factors tested ............................................................................. 154
4.3.2 Factors eliminated ....................................................................... 156
4.3.2.1 Gender ............................................................................... 156
4.3.2.2 Age .................................................................................... 156
4.3.2.3 Languages spoken ............................................................ 157
4.3.2.4 First ethnicity ...................................................................... 157
4.3.3 Statistically significant factors kept for knowledge ....................... 157
4.3.3.1 Education ........................................................................... 158
4. 3. 3. 2 Residential history ............................................................. 159
4. 3. 3. 3 Birthplace ........................................................................... 161
4. 3. 3. 4 Nationality .......................................................................... 161
4.3.3.5 Second ethnicity ................................................................ 162
4. 3. 3. 6 Religion .............................................................................. 164
4.3.3.7 Word origin ........................................................................ 164
4.3.4 Statistically significant factors for admitted usage ....................... 165
4.3.4.1 Factors eliminated ............................................................. 166
4.3.4.2 Factors retained ................................................................. 166
4. 3.4. 2.1 Gender ..................................................................... 166
4. 3.4. 2.2 Age ........................................................................... 167
4. 3.4.2.3 Education ................................................................. 169
4.3.4. 2.4 Department of residence .......................................... 171
4. 3.4.2.5 Ethnicity .................................................................... 172
4. 3.4.2.6 Religion .................................................................... 174
4.3.4.2. 7 Languages of informants .......................................... 175

43426
lit-535:5 01 5335
21733.5 CONCLUS
E‘ 32:: 5‘1‘3135‘5
E; see-'31 00006
53:3:er extia

531 Ward or;
5 3 2 Sosa." ‘a:
532.1 69'
3322435

53 2.3 E 3;
5325 Ext
532 6 Rs.
53.2.? La:
ElCrdcs-ro-ns

:: *4 {A
In: us-Iltawunl .....

 

 

 

4.3.4.2.8 Language of the lexical item. .................................... 176

4.4 Analysis of statistically significant factors ............................................. 177
CHAPTER 5: CONCLUSION ............................................................................ 181
5.1 Study synopsis ..................................................................................... 181
5.2 General commentary ............................................................................ 182
5.3 Factors explaining knowledge and usage ............................................. 184
5.3.1 Word origin .................................................................................. 184

5.3.2 Social factors ............................................................................... 185
5.3.2.1 Gender ............................................................................... 185

5.3.2.2 Age .................................................................................... 185

5.3.2.3 Education ........................................................................... 186

5.3.2.5 Birthplace, nationality, and ethnicity ................................... 188

5. 3. 2. 6 Religion .............................................................................. 189

5. 3. 2. 7 Languages ......................................................................... 190

5.4 Conclusions .......................................................................................... 191
5.5 Limitations ............................................................................................ 191
5.6 Areas for further study .......................................................................... 192
Appendix A ........................................................................................................ 195
REFERENCES ................................................................................................. 206

xi

 

 

 

, . a . I '4
It: .939...ch

$1»!
I.
c 1
3

t
p
1

)

IIIII

'145 3521911901 “5

I g I I '
It 4me°” “35

:53 from “A4

43.51. CS
‘11:} “1'1th of ac
'te‘t 113911101 l
'tt‘1 Wag): of.
'It‘i Weight of
36.3 Wei-96301
.1614 Weight of
be ‘6 Weight 01
.413 ‘6 Weight 0:
.4417; Weight 0

3613; Weight 0

41519; WGI'gtlt r

41520: Weight ,

 

LIST OF TABLES
Table 1: Perceived gender usage by gender of informant ................................ 1 17
Table 2: Admitted usage by gender of informant .............................................. 119
Table 3: Perceived education-level usage by education of informant ............... 125
Table 4: Admitted usage by education of informant .......................................... 129
Table 5: Perceived usage by age of informant .................................................. 133
Table 6: Admitted usage by age of informant .................................................... 137
Table 7: Admitted usage by original language of lexical item ........................... 143
Table 8: Admitted usage by venue .................................................................... 148
Table 9: Weight of education ............................................................................ 158
Table 10: Weight of residential history .............................................................. 159
Table 11: Weight of Birthplace .......................................................................... 161
Table 12: Weight of nationality .......................................................................... 161
Table 13: Weight of second ethnicity ................................................................ 162
Table 14: Weight of religion .............................................................................. 164
Table 15: Weight of word origin ........................................................................ 164
Table 16: Weight of gender ............................................................................... 166
Table 17: Weight of age .................................................................................... 167
Table 18: Weight of education ............................................................. . ............ 169
TEllble 19: Weight of department of residence ................................................... 171
Table 20: Weight of ethnicity ............................................................................. 172
Table 21: Weight of religious affiliation ............................................................. 174
Table 22: Weight of infonnants' language ......................................................... 175

xii

 

 

it

Q.

.-

‘II-IHIH"

vi
6:

Table 23: Weight of word origin ........................................................................ 176

 

 

xiii

 

LIST OF FIGURES
Figure 1: Perceived vs. admitted usage by gender ........................................... 121
Figure 2: Perceived vs. admitted usage by age groups .................................... 141

 

 

xiv

 

tail was '

W. Ania F'l
|:v|: ahb‘ a b'

.3.-Dan.” Iran;

a 1th

P‘II 1.. Inna
7:0- :1 w: 35:3

{3'13111‘ 'N‘I"

. "“ LU hive“
zi‘x‘a'hlqa
~~ «3:18.11

53366131

CHAPTER 1: INTRODUCTION

1.0 Introduction
This study wishes to discuss banlieue language in an innovative manner, and
therefore approaches the problem at hand—that of banlieue language—in a
multi-fronted manner. Many theories in research on banlieue language,
regarding its speaker base, have not been challenged, and this study wishes to
investigate several existing stereotypes. I believe that current research may be
expanded to include more sociolinguistic factors than ever before, and that we,
as researchers, must be open to the idea that banlieue language is changing not
just in form but also in number of speakers.

Therefore, this study wishes to investigate some of the current stereotypes
of gender, age, education level and ethnicity of speakers of the variety as well as
to propose new categories of sociolinguistic factors that may affect use of
banlieue language. Specifically, I hypothesize:

1. Banlieue language is spoken by people older than any other previous

research has indicated.

2. Gender is not as male-oriented as previous studies may have suggested.

3- Regardless of stereotypes and research that relegate the deepest vernaculars
to those speakers of the lowest levels of education, banlieue language, because
0f the function of identity, may also be spoken by those with much higher levels

0f education than previously expected.

 

 

 

I .1 F
State’s own V"

2': 2.3:: 5'3 55 m
mie'atihn 5
:3“ s :3th. I
m was m0”
:3 tame? C56
1:559? troops
:3" Battle an:
17:2: in what 15
‘31: ”my; ge I

From its au'
“3 he many c
mmmMi

use; their pref

In addition, I would like to investigate two never-before-investigated areas:
the speakers' own views on who uses banlieue language and the effect of

religion on speaking banlieue language.

1.1. Historical Background to French language variation

In order to understand why a phenomenon that started out as a language
game could be so important today, France's long linguistic history must be taken
into consideration. Starting as early as 842, Ies Serrnents de Strasbourg, 'the
Strasbourg oaths,’ become the first text ‘in French,’ though the French language
at the time was more pseudo-Latin than French (Walter 1988). The oaths were
taken between Charles the Bald and his brother Louis the German, along with
both of their troops, the two promising to defend each other against their brother
Lothair (Battye and Hintze 1994). The Sennents are important not just as the
first text in what is recognized as French, but for the choice of the fledgling
French language over Latin for such an important purpose.

From its auspicious beginning, the French language gained in popularity;
and of the many dialects, several fought for recognition. A few centuries later,
the poets of the twelfth century were already lamenting their lack of prowess (or
lauding their proficiency, depending on whether or not the poet came from the
r e9i0n) in Francien, the precursor to standard French spoken in the L'I‘Ie de

France region (Battye and Hintze 1992). Already a variety of French had started

to garner favor.

 

 

 

 

Four 38"th
mm certb'y

l -‘ l
. 732.0%: L

’Ca iocume'

'esg'ale was ;

Four centuries of linguistic evolution later, the trend had solidified. The
sixteenth century brought Francois I, who represented a more formal move away
from traditional Latin and towards not just French but the variety spoken in L'ile
de France. Under his rule, the College de France was built, which taught in
French instead of the traditional Latin (Battye and Hintze 1992). His Edit de
Viliers-Cottérets, in 1539, proclaimed that French should be used in all legal or
official documents (Battye and Hintze 1992). The first grammar book by John
Palsgrave was printed in 1530 (Battye and Hintze 1992) and Du Bellay
published his Deffense et illustration de la langue francoyse in 1549 (Ayres-
Bennett 1996). This century was marked by a perceived need among the
scholars to reform French. Indeed ‘concerted efforts are made not only to codify
but also to “enrich” the language, in particular by attempting to give it the lexical
resources necessary to compete with its classical forebearer’ (Ayres-Bennett
1996: 140).

The 17th century ushered in the Académie Francaise, which fought for
linguistic unanimity and a standardization of the only acceptable French: Ie bon
usage (Ayres-Bennet 1996, Walter 1988). Malherbe and many other
grammarians worked with the court, championing and upholding the linguistic
variety of the elite, that is to say the French of I'ile de France, condemning that
of anywhere outside of the Parisian region (Battye and Hintze 1992).

The 18th century was a period of more reforms. The quest for knowledge
was both linguistic and cultural. France was setting out to define itself physically

and Politically as well as linguistically. In the late 18th century, before the

liq-5'

{LES-7,. Cal «1:
their? and ‘

llea"ll"i?e

tests shell
'.| '. M:
"at" We; 1

: .Aa;. .5l_
1:” '06 .:

l
fit:- JC 70: l
Shake-3'

'9 “li‘n‘ ,4
fl ‘IU: H u'

wens

“We in a'

revolution, cartographers defined France using geometry and rulers to find both
the Meridian and to attempt to map their country (Robb 2007).

Meanwhile, in 1790, Abbey Gregoire conducted what may be considered
one of the first linguistic studies (Walter 1988). Wanting to arrive at an idea of
the number of French speakers in the country, he sent surveys to priests. The
results showed a large number of people speaking regional languages. in fact,
the results showed that a mere three million French people were fluent in French
(Walter 1988). To put this in perspective, twelve million people could not speak
French fluently or take part in a simple conversation. Of those, over six million
people did not know French at all.

Shocked at the number of regional languages spoken, and the
consequent dearth of French speakers outside the Paris area, he helped
promote the crusade that led to the subsequent squelching of regional tongues
in France in an effort to promote linguistic unity as a tangible symbol of equality
and political harmony of the people (Gadet 2003). The very name of his report
illustrates both his goals and the extent to which France has a history of
linguistic assimilation. The report was entitled Rapport sur la nécésité et les
moyens d'anéantir les patois et d'universaliser I'usage de la langue francaise
(Walter 1988).1

The Ecoles Norrnales, or places of instruction for those who wished to
teach, were created as a direct result of Abbey Grégoire's study, since the

original goal of having a primary school taught in French each community was

 

 

i .
'3990“ .0" the necessity and the means to annihilate dialects and to
Unlversallze usage of the French language.‘

4

 

 

l
.L

 

91"." .

5
I: 5' ”NS 3.

.l‘ R‘ ”an
a

E! By: M P‘ .
13"} var-9‘13:

-l‘: t H .fiH
2...: .. -~

art's“ “'

to. "W‘L "
V

rc' “3" 3'
- 0

.a-a-V

its? 'EEBL !

cut-n fl ”V

.::I. lit. . .t
#5385 F
E”: by his
resatzis. it
2": :‘Frang
‘8' 2"6- ital
Theatre

fctt'a‘e? bl

too ambitious at first due to the lack of instructors (Walter 1988). Indeed, the

very idea of primary school education stemmed from a desire to propagate
French as not only an official national language of administration and laws, but
the only language of the people as well (Walter 1988). From that time on,
especially during the period after 1789 — 1815 when regional languages were
first targeted, they experienced a steady decline in popularity, and it is not until
the latter part of the 20th century that a re-burgeoning interest has developed
(Walter 1988). Finally, it is in the 18th century as well where the Encyclopédie is
created with the goal of purity for the French language (Walter 1988).

The 19th century brings schools—free, secular, and obligatory—with the
laws of Jules Ferry in 1880. The patois still survived until World War l, when they
were lost by historical coincidence rather than the false imposition of French
over patois. The men who were sent to war were placed with soldiers from all
parts of France, with French as the only lingua franca (Walter 1988). When the
men came home, they kept speaking French out of habit, which made it easier
for the children also to speak French in the home, and the patois were dealt an
almost fatal blow (Walter 1988).

Throughout the ages, then, from the time that the Roman soldiers help lay
a linguistic superstrate on Galois, the French have striven to define themselves
and their language. The concepts of ‘nation’ and ‘language’ have been intricately
intertwined over the centuries, with authors and political figures regulating the
F r ench language in order to define the French nation, and conversely using the

concept of a French nation to justify the unification and standardization of the

 

 

i.-

 

Steal...“

gaze! (flat

“1 when

“:- IIA~ A
#4: ”Saw:

MIA“ AI
c: ck"

‘I"‘
rear

Cy.

‘t its for
Wis Va."

hit this

French language. Thus, since the 12th century poets, French history has

intrinsically linked ‘being French’ with ‘speaking good French.’

1.2. Breaking the norm
Speaking a non-standard language or dialect in a country that equates
standard speech with citizenship is a form of rejection of the language and the
culture. However, among the speakers of the minority dialect or of the

marginalized form, there is a certain prestige that comes precisely from being a

 

member of the group on the margins and rejecting mainstream language and l‘ 1
ideology. Pierre Bourdieu (1982) calls this ‘contre-Iégitimité Iinguistique.’ Others I

call this ‘covert prestige’ (Labov 1972, Trudgill 1984). Regardless of the moniker
Used, the covert prestige of defying the linguistic norm through the use of forms
such as Verlan and other non-standard linguistic elements effectively distances
the speaker from mainstream French culture and all established precepts,
whether governmental or familial.

Language games and argots in France, however, have not been lacking,
despite (or perhaps because of) this linguistic domination. The functions of these
idioms vary; some secret languages are simply games, others serve as a secret
Code. All serve the function of identity marker to some extent for the speakers
(Goudailler 2002). These argots and secret languages have typically been

Created by segments of society in order to separate themselves from the

mainstream.

 

4S“

' .
iv '

 

Historically, there have been many such phenomena. Largonji, from
‘jargon,’ a language game of the ‘Louchébems,’ (among the bouchers or
‘butchers’), allowed them to talk amongst themselves in a code of syllable
manipulation (Walter 1988). In the early stages, argot thrived as a secret
language of the underground, revealed by its speakers only under torture
(Goudailler 2002, Merle 1997). In the 20th century, the most popular, and
the most studied, is banlieue language, which has as one of its key
elements the syllable reversal of Verlan. The origin of Verlan itself, and its
extension of banlieue language, is much debated.

Verlan, originally a jeu de langue of syllable reversal, has evolved. Its
definition has broadened, and what was once a simple game is now part of a
9reater linguistic struggle. The ensemble may now be better termed ‘banlieue
l«‘=Inguage’ since it now includes so many contact languages as well as argot and
the original syllable-reversal of Verlan. Among those contact languages
specifically cited are Arabic, Roma or Tzigane, and Creole, in addition to some
Al’rican languages (Goudailler 2002, Mela 1997).

Therefore, in today’s political climate, Verlan and banlieue language
should no longer be viewed as a jeu de société 'a parlor/board game' among
stOCially marginalized youth but rather a linguistic symbol of a potential enjeu de
Société 'societal stake' if the message intended by its speakers is ignored. Those
Speakers are making a conscious choice to use a nonstandard variety in a
Country where the standard is so intrinsically related to nationhood and

Citizenship. Doran (2002) and Meliani (2000) have both. tied Verlan to the

 

 

 

 

.nE-nv n “1"
,\
It. 9‘ ‘

. 3.1:!“ as

‘vw I

C 1..

fats ar

'*‘_' ”l" n
"u 'i '\

‘tvb

p
l'

:13”:

l 0i ll

concept of identity in youth of Maghrebi heritage in France and cite their
informants of said origin as claiming the language. Yasir Suleiman (2006)
underscored the role of ‘language as proxy to tell you about (political)
undercurrents. Periods of stress, conflict, and crisis evoke issues of identity.’
The riots in the banlieue in 2005 by the suburban youth emphasize their
discontent, as interest in Verlan and banlieue language increases.

This dissertation examines banlieue language using empirical data. The
informants answered several perceptual questions in a questionnaire about who,
in their opinion, would use those words and phrases. The target demographic
was people from ages 18 to ages 36. I cannot claim that the group of informants
in this study was representative, since the selection was not random. The

Venues were chosen specifically to target people most likely to be familiar with
banlieue language, yet who were older than 18, in order to see if those who had
grown up in this area admitted to speaking the variety. This means that while the
informants did contribute empirical data, I cannot make any sweeping

generalizations; I can only hypothesize by examining the data provided to me by

this group of informants.

1.3 Definitions
Many of the terms in this dissertation are used in popular culture in one manner
at‘lt‘l in another in linguistics. In addition, there is even some variation in terms

among linguists. It is therefore necessary to specify what I mean by each term in

Order to avoid confusion.

 

 

 

pr—

 

’. u l
'vS’w

..
BIA AP
- f'

.vv

.9-
I-

‘Q'

I
"au- 4
k v.

n? '
i1; '

ll-
fi~
I. ,‘v.

1.3.1 Angot
According to the Concise Oxford Dictionary of Linguistics (online), argot is
a “special vocabulary used eg. by criminals which is designedly unintelligible to
outsiders. Thus one form of antilanguage’ (Mathews 2008).
In this dissertation, the word argot is used in a more general sense to
mean the vocabulary in non-standard or non-reference French. It has its origins
in the criminal element of the 18th and 19th centuries (Goudailler 2002, Merle
1997), but I am not speaking of the specific variety that enabled thieves to speak
among themselves without being understood by the general populace. My

informants often used the word argot in reference to everything in non-standard

French that was not specifically Verlan syllable reversal.

1.3.2 Banlieue
This term designates any network of suburbs surrounding a large city in
F r ance, but in this context, I am referring specifically to the Northern suburbs of
P"eil’is. While originally there were two terms for suburbs in French, faubourgs
and banlieue, the former, later modified to banlieue aisée, has come to refer to
the more affluent suburbs. Banlieue by itself is becoming the default term to
designate those suburbs of more modest means. In many cases, the term has

cleveloped to signal not just modest but impoverished areas rife with

u“employment and violence.

 

 

 

 

«uh
A. v
I‘-

"O

‘QE l‘l
«wars
533W WEE
in ‘

't all“:

ed

l
f

”l "3‘53“

‘-

it

ii
'

:rz‘l'!

In. a
Hui
I

a .

1.3.3 Banlieue language

I use the term banlieue language to mean the more global sense of the
variety some researchers are calling Verlan. While Doran's (2002) informants
used this term to define not only the simple process of syllable reversal but also
to capture lexical items brought in from other languages and the grammatical
aspects that are referred to in the media, my informants were loathe to do so.
While talking to several informants, they suggested that it would be better
termed langage de banlieue, which I have just simply translated to ‘banlieue
language.’ Merle (1997: 23) refers to the language of youth in the suburbs as ‘la
langue des banlieues,’ which also translates to ‘banlieue language’ in English.
My infonnants' reasoning was that while a person who spoke the deeper
vernaculars would most undoubtedly use Verlan (and said syllable reversal
linguistic element would be a large component of that person's vocabulary), not

everyone who uses Verlan would also fully utilize the variety. They also flatly

 

r ejected lexical items from contact languages as being part of Verlan. To them,
Verlan was a separate entity, which involves the reversal of syllables, simply
Stated.
1.3.4 Dialect
According to Crystal (2003: 136), dialect refers to ‘a regionally or socially
c"Stinctive variety of language, identified by a particular set of words and
9"ammatical structures. Spoken dialects are also usually associated with a

distinctive pronunciation, or accent.’ Usually dialects of the same language can

10

 

o .7 IA
E...'3W"" “To

1
roflflfl a
.r': . a"

1’
(I‘D
r o-

(I)

be mutually understood by speakers of other dialects, but the case of Italian is a
longstanding example to the contrary (Mathews 2008).
1.3.5 Jargon
By jargon I mean the set of vocabulary specific to any one profession or

peer group, such as the ‘Largonji des Louchébems.’ I do not mean any

grammatical variation, which is what sets this word apart from dialect and variety.

The Concise Oxford Dictionary of Linguistics (online), defines jargon ‘in the

ordinary sense of technical or pseudo-technical vocabulary (Mathews 2008).

1.3.6 Variety
This term is very similar to dialect in that it encompasses the various

accent, vocabulary, and grammatical features of an area. The Concise Oxford

Dictionary of Linguistics (online), states that a variety is ‘any form of a language

Seen as systematically distinct from others: thus the dialect of a specific region

(6.9. Cornwall), any more general form distinguished as a whole by speakers

(e-g. American English or British English), a social dialect, one of the forms
distinguished in diglossia, a dialect used in a specific genre of literature, and so
0'" (Mathews 2008). Crystal (2003: 489) describes it thus: ‘a term used in
sOciolinguistics . . . to refer to any system of linguistic expression whose use is
Qovemed by situational variables . . . For some sociolinguists “variety" is given a
more restrictive definition, as one kind of situational distinctive language—a

slbecialized type of language used within a dialect.’

11

 

PQ'
.,.. .
1
All! A
. 4 r
have b
.
0.!“
, _..,.,-:

PP, uni.)
l a
:4 flat

It. as: ‘
:94 . ‘
'1

‘U

ltll

MM :1-

'U*' n. 1.

1.3.7 Verlan
This game of syllable reversal may have stemmed from the cryptic
language of the undenlvorld in the 19th century, which was used for
communication among thieves (Merle 1997). The term has expanded and for
some researchers means not just the syllable reversal oft studied, but the
greater set of vernacular variables which include but are not limited to argot,

grammatical aspects, and accent (Doran 2002, Melliani 2000). However, when l

Speak of Verlan, I mean strictly syllable reversal.

1.4 Goals
This dissertation was derived from a need to question some of the
accepted stereotypes of banlieue language and compare those stereotypes to
both the perceptions the inhabitants of the area have about banlieue language

and its speakers and how they perceive themselves with regards to banlieue

'anguage.

1.4.1 Age
Ohe of the main goals of this dissertation is to examine the extent to which
banlieue language is retained among ‘older’ speakers, that is to say those older
them 25. Most recent studies (Doran 2002, Lepoutre 1997, Melliani 2000) deal
primarily with adolescent minority youth, some stating that Verlan, their term for
banlieue language, is only used after the age of 16 by ‘délinquants’ (Lepoutre

1 997). Mid-twenties seems to be the longest estimated point among those who

12

 

 

 

indicate age as a factor of retention, even among those who admit to prolonged
peer usage (Bachmann and Basier 1984, Boyer 2001, Calvet 2006, George
1993, Mela 2000, Zerling 1999). Furthermore, I would like to investigate the
hypothesis that the speakers are retaining banlieue language, though it may not
be used in public among strangers.
1.4.2 Education

Another goal is to question the established stereotypes pertaining to the
education level of those who speak banlieue language. The media presents
‘typical’ speakers as being those who have a very low level of education. This
conception has been reinforced by certain researchers such as Lepoutre (1997)
and Doran (2002).

In addition, while some studies have been done with no real regard for
age limits (Bachmann and Basier 1984), most investigations were done with
bv'anlieue language speakers having a low level of education, either those who
have dropped out or those who are in grade school. In France, a country in
which having the baccalauréat is so very important socially, education and class

are closely linked. In fact, students are routed early (classes is the term used by
banlieue authors such as Amara [2003], Leila [2005], though it may also be
referred to as orientés) towards a technical degree if their teachers do not feel
they have the educational acumen to continue along the route to the Dec and
higher education. Many students born to immigrants complain of being placed in
the technical track despite wanting to earn a bac, (Amara 2003, Leila 2005), and

INSEE (2005) recounts a large disparity between the number of children

13

 

 

:N '1

‘Vls’

Ween.

sac. .

‘ ALIA A.‘
a

filo-‘9 s

t s
Cl)
(I,

{x ?
J .
(T)

immigrants in the technical tracks and the number of those born to French
citizens on both sides. INSEE (2005) also attributes a large number of technical

degrees to those areas in economic decline, such as Val d'Oise and Seine-

Saint-Denis.

For the purposes of this study, a low level of education will be defined as a
participant not having a baccalauréat (the bac). The bac is a degree roughly
equivalent to high school, but with a grueling final exam that has to be passed in
order to receive the diploma and the degree. Simply having a technical degree
such as the BEP, which is equivalent to the bac only on time spent in school, is
still considered a low level of education. It is an alternative to the traditional
education route of the bac in high school, a type of trade or technical school.
Having the bac in France is more prestigious than having a simple high school
education in the US, though it is an entry-level degree for many jobs.

We will therefore see if speakers of banlieue language are not all from a
low level of education. In that case, class distinctions pertaining to speaking
banlieue language may be blurring.

1.4.3 Gender
Investigation of both perceived usage among the sexes and admitted
Usage explained in terms of gender is another goal of this study. Usage of
banlieue language has been attested among women (Doran 2002, Lefkowitz
1 989, Lepoutre 1997, Meliani 2000), but has never been the direct focus of a

Study. In addition, though the above studies do mention girls and young women

14

 

. .np. A
1.! 3...

'IN' V

. A I
2's: 3-

&2» ~ a:

N v'v- .‘d

J'-
are, 3*

FA...,

. I-
""V'-\1\.
‘9 “kc
fl 2.“:

{Etta

IN

using the variety, no study has included empirical data concentrating on self-

admission of usage among females.

Sociolinguistic research has shown that, in Western cultures, women lean
towards a tendency to use a more ‘standard’ variety (Labov 1972, Trudgill 1972,
for exceptions see Milroy 1980) and would therefore tend to eschew usage of
banlieue language while men would tend to embrace it. We will see if women
are admitting to using banlieue language vocabulary and will also study their

Parceptions of whether or not the vocabulary in banlieue language is used

equally among the sexes.
1.4.4 Ethnicity

Another major goal of the dissertation includes a study of the different
ethnicities of the informants. Many studies done on banlieue language
concentrate primarily on speakers of Maghrebi heritage. Some researchers refer
t0 it as a product of a mixed heritage of the speakers of banlieue language and
c<>ntinue by saying that the speakers master neither the language of their
parents nor the language of France (Hargreaves 1997). It is important to see if
this is indeed the case that those of Maghrebi descent are more likely to use
banlieue language than those of other heritages, or if banlieue language is, as

Tejedor de Felipe (2004) suggests, a French phenomenon at its base.

1.4.5 Place of residence
In theory, use of banlieue language should be most concentrated in those areas
Where poverty is the most extreme and education is the lowest. However, place

of residence is a factor that has not been investigated empirically. Two areas

15

 

 

 

with a very high concentration of poverty and immigration as well as a low level
of education are Garges-les-Gonesse, in the département (‘department’ or state-
like government entity) Val d'Oise and Sarcelles in the département Seine-Saint-

Denis (INSEE 2005). For this reason, it must be noted that all research was

 

conducted in these départements.
1.4.6 Religion
Since religion is not typically a question presented to informants in

sociolinguistic research, some justification may be necessary. No one has

questioned religion as it relates to banlieue language prior to this study. In fact,

 

the questions were added due to sociological research in France that indicated
that practicing a religion had more impact on voting preferences than the social

class of the person (Lijphart 1971). Brulé (1966) discovered that religious

devotion was an even stronger influential sociolinguistic factor than typical
Sociolinguistic factors such as gender, age, and socio-economic status.
Therefore some investigation into religion and practicing a religion should be
investigated, though both these studies are dated.

1.4.7 Languages
No empirical data has ever been collected on any potential

Correspondence between languages spoken by banlieue language speakers and
promotion or demotion of banlieue language. Not only should the languages in
banlieue language be taken into account but also the languages the informants

Speak. By this I hope to question the precept that banlieue language is primarily

a product of those who are truly second generation and non-integrated, and

   

I ‘U
«to

a?»
.

v.«.

E

hint

propose that it might be more of a French based phenomenon as Tejedor de
Felipe (2004) suggests.

1.5 Hypotheses
Based on the goals described above, I have formulated the following
hypotheses:
1.5.1 Age
The vernacular extends into the early to mid 305 among friends, whether
or not it is used publicly. This opens the possibility for discussion of banlieue
language being an in-group vernacular that ages with the group in question.
1.5.2 Education
Formal education will have some bearing on whether or not someone from
the banlieue still uses the variety. However, those with higher levels of formal
education will still admit to using banlieue language vocabulary. Few associate
banlieue language lexical items with those of extremely low levels of education.
1.5.3 Gender
While women do use banlieue language slightly less than men due to a
Variety of reasons, including sociolinguistic norms (Labov 1972, Trudgill 1972), it
is not as male-oriented as the media has led us to believe. For this reason, the
Vocabulary is perceived as being used equally among the sexes. In addition,
empirical data on admitted usage will allow us to compare what the speakers

perceive as usage among the sexes and what each sex admits to using
D6rsonally.

17

 

 

1.5.4 Ethnicity
Banlieue language is not as specific to speakers of Maghrebi heritage as

current studies claim. It is more a matter of cultural identity with the banlieue

than actual ethnicity of the speaker.

1.5.5 Religion
l hypothesize that religion will have some effect on whether or not a
person uses banlieue language.
1.5.6 Languages
The languages spoken by a person will be an important factor in whether
or not a person will use banlieue language. The languages a person speaks will

help identify him or her more strongly with either the periphery, and thus the

banlieue, or mainstream French culture.

1.6 Dissertation Overview
Chapter two provides an overview of previous research on this topic as
Well as of related research. In particular, several key studies will provide
bv'-'3ickground for the analysis of the results. Calvet’s (1994) discusses the urban
Situation as it relates to language change. Together with Labov’s work in
Martha’s Vineyard (1972), and Milroy’s social network (1982), Bourdieu’s (1982,

1 991) concepts of linguistic marketplace and linguistic capital provide the

$0(:iolinguistic framework.

Chapter three outlines the methodologies used in the dissertation. This

includes the creation of the data collection instruments as well as an in-depth

18

 

 

'elflk

14

\‘i

‘1)

"I

discussion of where the research was conducted. Finally, data compilation
techniques are outlined as well as the statistical operations used in the data
analysis.

Chapter four presents and analyzes the results on the vocabulary used in
banlieue language. Calvet (1994) states that language contact creates a new
variety, which reflects a blending of many influences and thus, a new identity.
With that in mind, this chapter is based on a questionnaire containing lexical
items from Roma, Arabic, Creole, old argot, and purely syllable reversal Verlan
items to test the degree of integration of each component. To discover which
factors were statistically significant, logistic regression tests were run.
Additionally, individual social factors were taken into account, such as age, sex,
origin, family presence in the banlieue, education, and religion.

Chapter five recapitulates the important findings of the research and

analysis and also discusses related studies that may be forthcoming.

19

 

 

 

 

CHAPTER 2: PREVIOUS RESEARCH

2.1 Urban languages
2.1.1 Language contact

Val d'Oise and Seine-Saint-Denis are some of the most geographically
diverse departments in France, with the largest number of immigrants. These
immigrants represent various ethnicities, cultures, and languages from Africa
and Eastern and Western Europe, as well as the Americas (Carpin, Tavan, and
Dugué 2005, Vermes 1988a, Vermes 1988b); therefore a language contact
situation has developed. Because of this, it is important to consider what
happens in a country to which many immigrant communities have flocked, each
bearing their own set of ideologies, as well as cultural and linguistic
backgrounds. A second and related question is what effect is observed in a large
municipal area in which several immigrant communities have come together, or
as in the case of France, have been geographically pushed together?

Urciuoli (1995) takes care to point out that borders, while often thought of
as being national, can also be drawn between territories, so let us begin our
inVestigation into the previous research by looking first at languages in contact,
especially as it relates to large cities.

One of the most instrumental works on that subject, especially as it
applies to France, is Calvet (1994) He underlines the important concept that a
°°Untry in which multiple languages are spoken is not necessarily a multilingual

country; rather, that country is one of monolingual zones that ‘convergent, sur la

20

 

IL

 

 

 

I6

'.a

.2

”J

piste, les marches, les portes et, de facon plus générale, dans la ville’ (Calvet
1994: 11). The result of this convergence can sometimes be what he calls a
‘langue ad hoc, du type pidgin, en fonction de langue véhiculaire’ (Calvet 1994:
11). Calvet (1994) also defines three roles that urban life plays in regards to
language, ‘la ville comme facteur d’unification linguistique, la ville comme lieu de
conflit de langues et la ville comme lieu de coexistence et de métissage
linguistique’ (Calvet 1994: 1 1).
2.1.2 City melting pot erodes other languages

According to Calvet (1994), the first function of language in an urban
setting is unification. Applying this to Paris, Calvet breaks Parisian urbanization
into two separate periods, the first of those an interior migration to Paris during
the early 19th century, when speakers of France’s regional languages—Breton,
Basque, Occitan, etc—came to Paris. These regional languages were quickly
abandoned in this first wave of urbanization, which he calls ‘la période de
croissance maximale’ (Calvet 1994: 64). Those immigrants coming from the
outer regions of France quickly abandoned their regional languages in order to
assimilate.

In addition, as mentioned earlier, the regional languages in the Hexagon
were also deliberately quashed in order to assimilate the population and create a
united France (Gadet 2003, Offord 1996, Vermes 1988a, Vermes 1988b, Walter
1988, Walter 1997). While speakers of each language do exist, very few, if any,
are monolingual in the regional language. The Jules Ferry laws in 1888 for free,

secular, and mandatory education in French constituted one of the final blows

21

.’-,
Q. A

‘U
a"

' '5'... '93

l’.

_ .

"' I"
'uv

O:' ‘

i? ‘.‘e

1"“
'-
t n
.n'.
it v"
h.v

.

I; ':.- I
v.11:
I
'1’“él
V} l.

11"“?
V 5‘

”Mr
Vi iv

iii.

V

a;
9)

var

for regional languages. The vestiges of these languages fight for survival in each
of their respective rural regions. While they did contribute to the French
language (Walter 1988) the borrowings into French took place during the period
when the contact was the greatest, that is to say when the speakers of the
regional languages were mostly monolingual (Walter 1988).

The second wave of urbanization, a ‘période de stabilisation (Calvet 1994:
64), corresponds to a more modern-day France. French is now well established
as the lingua franca; and most of its regional languages, despite recent efforts
for revival, have all but disappeared. Their speakers are at best bilingual, and
French is deeply established.

2. 1.3 A new shift of function

Here French history intervenes to add another layer of superstrate
languages to the mix. After the decimation of F rance's population in the World
War II, there were not enough manual laborers in the country. France looked to
its former colonies for workers, who, over a long history involving several steps,
eventually settled in France and brought their families through the reunification
acts of the seventies (Hargreaves 1997). Many of the wives were not only
monolingual in their respective languages but also illiterate. The children lived in
two worlds, each with its own language. This diglossic situation led to the
children of these first immigrants living between two linguistic and cultural
realities (Doran 2002, Hargreaves 1997, Lepoutre 1997, Melliani 2000).

At this point, exterior urban migration is the norm, and foreigners who are

trying to learn French are able to preserve their language in a diglossic situation

22

h ‘aif ‘.3

“a“; heft-t

«3

<_;f'
‘1

iv
‘1" «7’
by“;

in which French is the ‘official’ language and their native language becomes the
‘low’ language, used only at home. In the larger, poorer, urban areas affected by
immigration, Arabic, Berber, Creole, Roma, Turkish and several African
languages influence French. It is at this point that the language with a foreign
status adopts a function—that of an identity-marker (Calvet 1998).

The second set of languages, those of immigration, are not being
abandoned as quickly due to their new function: identity. One does not have to
look hard to find commentary about the so-called ‘second generation’ and its
being between two cultures, mastering neither the language of the home nor the
language of their adopted country (Amara 2003, Doran 2002, Hargreaves 1997,
Lepoutre 1997, Melliani 2000). The media is rife with stereotypes of this second
generation and its lack of communicative abilities in ‘standard French’.

2.1.4 General emergence of urban languages
Contact among languages and speakers changes the ranking importance of the
languages: that of home and family is less conducive to outside contact, which
leads to a reversal: the lingua franca (Calvet 1998). He calls this a ‘vehicular
Ianguage’ (Calvet 1998: 65), which, he notes, is normally the language of the
dominant ethnicity; but if one is not present, that lingua franca instead takes the
form of ‘une langue neutre’ (Calvet 1998: 65).

Whenever language contact happens, the speakers of all languages
borrow lexicon from the other language or languages (Weinreich 1953). In the
beginning this borrowing is mostly of nouns, and what is borrowed is often

borrowed in a specific geographic location (Myers Scotton and Okeju 1973).

23

a
$ A“
I
. I'u’I

lg. P.‘
.

..~.I

it: '

‘0 !l
a. 3

'f'i

I A

?

The vocabulary borrowed in the case of language contact is not always foreign
to the borrower and may replace existing lexicon in the vehicular language
(Myers Scotton and Okeju 1973). This is especially true in banlieue language,
where many nouns are borrowed for taboo concepts such as theft, sex, and
drugs (Séguin and Teillard 1996, Goudialler 1997). As Guiraud (1963) points
out, ‘argot’, one of the components of banlieue language and debated moniker
for it, has its origins in the undenlvorld as a secret language to conceal thefts and
other crimes.
2.1.4.1 Examples of emerging language contact varieties

It is not just France that has experienced this phenomenon of contact
languages creating a semi-coded language variety among younger generations.
Similar linguistic varieties are being noted in recent times, especially in Africa.
These varieties are badges of identity for their speakers (Githiora 2002, Ginthinji
2008) and serve as a lingua franca in an area where dominant languages and
culture are rejected (Abdulaziz and Osinde 1997). In investigating these
language varieties, we may see that banlieue language has some parallels and
that the latter is not just a code among a segment of society, but may be
becoming a more encompassing variety.

2.1.4.1.1 The case of Sheng and Engsh

One of the more studied mixed-code language varieties in recent years is
Sheng. In Kenya, where Swahili and English fight for dominance among several,
smaller regional languages, the contact has resulted in certain language mixes

(Abdulaziz and Osinde 1997). Sheng, a code-derived moniker for a mixed

24

0"
I!
I 1‘4

“A.

y'-

.u I

III-
I

IQII.
. i

ii Jr

‘0‘.

,‘rgJ

H.”

‘.|.-

language in Kenya, is a combination of Swahili and English, as well as some
other local contact languages (Githiora 2002). This language variety is parallel to
banlieue language in many ways. The primary speakers are ‘primarily
preadolescents through young adults’ (Githiora 2002: 159). In addition, Sheng
seems to be growing, and the use of this variety is based on solidarity and in-
groupness (Githiora 2002, Ginthinji 2008). Githiora (2002) says that over 50
percent of the males questioned in a university setting still admitted to using the
variety, and in fact saw its use as a measure of cultural in-group solidarity. This
becomes particularly important when comparing with banlieue language. Indeed
research on these newer language varieties has been done on older groups in
an educated setting.

Another, similar language variety has evolved in the same area, but with
English as its base (Abdulaziz and Osinde 1997). Engsh, the language game in
question, is spoken by the wealthier youth who use English as their primary
language instead of Swahili (Abdulaziz and Osinde 1997). Both varieties have
become more popular and more fixed as the youth continue its use (Abdulaziz
and Osinde 1997).

2.1.4.1.2 The case of Camfranglais/Francamglais

This language game is also relatively recent, with most studies having
been conducted in the late 1980s It is spoken by adolescents and those ‘older'
speakers in their 303 (Keissling 2005) in Cameroon where two ancient colonial
languages, French and English, vie for power among the citizens (Keissling

2005). In this Cameroon language contact situation, both French and English

25

 

.I-l

-r‘;~
'.
you"

*. A“
d .d‘v

u» .r
I

[31“\ .\

k9. IR

In;-~
D
.51 d

-Ap

'5 I b
any}

I
W rvm

F .
- fare, at

F.

It"

lies
El;

hr

25”.“
N Vi

are official languages. However, due to the colonial situation, most people are
exposed to English as a scholastic language, and French as an official and
scholastic language, in addition to various African mother tongues (Keissling
2005, Kouega 2003).

There are varying theories as to its origin. Koega (2003) attributes its
beginnings to school children. Previous studies by Lobe-Ewan (1989) contribute
the formation of this language game to older students in a university setting. Still
others relegate its creation to the underground, criminal element (T iayon-
Lekobou 1985). Whatever its origins, it still remains a code-based method of
communication with secrecy and identity as its two main functions, using a
French-based syntax and code switching from contributing contact languages.
(Kouega 2003).

As evidenced even by the variability in its name, this language built of
contact languages is rapidly changing. Keissling (2005) cites his informants as
using Francamglals instead of Camfranglais. Its earlier monikers were Camtok
and Camspeak (Kouega 2003).

Interestingly enough, Verlan vocabulary is found in Camfranglais, as
attested by Keissling (2005); mef, rep, rem, ress and refre, are used. Those
lexical items are meuf (femme, ‘woman’ or ‘girlfriend’), reup (from pere ‘father’),
reum (from mere ‘mother’), reus (from soeur ‘sister'), and reuf/ refré (from frere,
‘brother’) respectively (Keissling 2005). These lexical items are tonalized by the

speakers (Keissling 2005).

26

.r_
(I!
"C 1

. ,

.04
WW
n“ “U

u I. a”
I P l

2.1.4.1.3 The case of Dutch straattaal (‘street language’)

As part of the outcropping of language mix varieties not located in Africa,
Dutch straattaal was first was started among the Dutch youth in big cities prone
to immigration. Again, the official language of the country serves as the
language of syntax, and words and phrases from minority languages are

interspersed (Verrneij 2004).

2.2 Language and identity
Here I discuss the research defining the ties between language choice and

identity to an area.

2.2.1 Language change as it relates to identity

As Guiraud (1963: 117) states ‘tout langage est signe.’ Language has long been
viewed as a symbol of identity. We choose to speak in a certain manner in order
to convey our affiliations. Labov’s (1972) Martha’s Vineyard study proved the
relationship between wanting to stay in an area and adopting the local speech
norms because of it. Milroy's (1982) networking study established the usefulness
of determining the extent to which an individual is immersed in the area culture
in order to explain deviation from or adherence to local norms.

Bourdieu (1991) invokes the idea of a ‘linguistic market’, in which each
register or style of speech has a different perceived value in society. In his
theory, the register's perceived value also changes as the group of speakers

changes. For instance, while speaking a very formal register is recognized as

27

 

 

u"

21’!

having a high value in formal circles, its perceived value would be lowered
among informal circles whose members would view the speaker as too formal,
snobby, or self-important. The inverse is also true.

Bourdieu's (1982) idea of ‘contre-légitimité Iinguistique’ in many ways is
very similar to Labov’s (1972) concept of “covert prestige’ in which certain
individuals will consciously or unconsciously go against the linguistic norm. By
doing this, the speakers set themselves apart from mainstream society, which in
turn creates a certain prestige in their own circle. This prestige is ‘covert’, since it
is not shared by the greater population, or by those who share the ‘highest’
values of linguistic currency, meaning the more prestigious circles of society.

One of the many functions of banlieue language, or any code, is its cryptic
nature, which is especially useful when talking about taboo subjects in front of
authority figures (Doran 2002, Gadet 2003, Goudailler 1997, Guiraud 1963, Méla
1997, Tejedor de Felipe 2004, Zerling 1998). Goudailler (1997) explains that a
shift towards identity as the main function of argot has taken place, in what I am
calling banlieue language.

2. 2. 2 ln-group/ out-group behavior

ln-group and out-group behavior applies to the banlieue, not just in what
has been studied, but often on how it is studied. In fact, some of the
researchers, such as Melliani (2000) and Doran (2002) have asked members of
the group to gather information for them, following a kind of Labovian tradition
(1972) when Labov recommended sending and African American in to collect

data among the informants of the same culture.

28

Lepoutre’s (1997) study takes place in a cite‘ (government subsidized
housing projects), near the Cité des quatre mille, his residential choice while
teaching at a La Courneuve college ‘middle school’ in 1990. Though he is
French, he is, by virtue of his profession and age, automatically out-group. While
this would not keep him from observing Verlan, as Gadet (2003) notes, it would
keep him from active participation in a group of older Verlan speakers. His living
in the environment as an ethnographer is highly efficient for observation of his
target group, but by definition automatically restricts him to that particular socio-
economic group. He is viewed as ‘rich’ by his major informant, banlieue speak
for ‘bourge’, an upbringing Lepoutre admits to.

Doran (2002) punctuates the in-group/out-group connection when talking
about the pre-teen group stating that the students were loathe to use banlieue
language with the teachers, who were older and very formal. The students, in
other words, were already well aware that such language was not acceptable in
front of adults.

2. 2.3 Identification with the environment

Identification with the area and culture are also areas that have been
widely studied. Urciuoli (1995: 532) talks about ‘what pulls linguistic elements
into a language’ which focuses on identity as a factor. She talks about
‘deploy[ing] language forms in “acts of identity“ which reflect ‘peOple‘s sense of
community, group and language’ (Urciuoli 1995: 532). Calvet (1994: 62)
proposes that speakers of languages in urban situations undergo two opposing

forces, that of ‘véhicularité’ and ‘grégarité/ldentité.’ In other words, while the

29

‘fl A'
I
awv~

I. ff
, .l
‘ .

.‘9' "Z

l)‘ '2

iI“'

I‘PF

.I-

j I
(11

wt,
v,‘
u 'U

contact of multiple speakers of multiple languages may suppress individuality
(speaking banlieue language instead of another native language), at the same
time, that individuality is highlighted (speaking banlieue language instead of
French).

It is certainly true that banlieue language, by any name, has been tied with ‘
identity and counter establishment by many (Aitsiselmi 1997, Bachmann and
Basier 1984, Calvet 1994, Doran 2002). Melliani's (2000) and Doran's (2002)
informants claim Verlan and banlieue language as their own. Both of them cite
their informants of Maghrebi heritage as vehemently defending 'their’ language
and defending its use from those who are not in-group. Séguin and Teillard's
(1996) students simply stated that if the dictionary project allowed too many
older people to understand them, they would just change the language again to
make it less comprehensible, to keep it theirs.

Gumperz (1983) calls this the ‘we code’, a cryptic language used by
group participants, that both reinforces the group and isolates it from others.
Both Doran (2002) and Tejedor de Felipe (2004) connect the ‘we code’ to
banlieue language (using their own terms for banlieue language). Tejedor de
Felipe (2004) studies it as it was illustrated in France by Séguin and Teillard
(1996) two teachers/researchers at a local middle school (college) in Pantin les
Courtilliers during the school year 1994 - 1995. They observed the students over
a long period, and in an effort to both study the language and improve the

‘francais standard of the students, a dictionary project was created in which not

30

or"

fit...

r3 '

I.
In!-
\t-

u’ 'v

o u
”Sw

,p--

"O"

:Fflzl

I .v.

r?-

‘vi ll

 

 

only did the teachers journal the experience but the students collected and
defined many banlieue language lexical items.

Doran’s (2002) section on identity in which her informants oppose
themselves to les Francais de souche underlines her attention to the Maghrebi
youth, more importantly, she highlights the function of Verlan as ‘Affirrning
Solidarity and Shared Values: ‘We’re not bourge’" (Doran 2002: 203).

Calvet (1998) discusses linguistic history in France as well as in other
countries. Diglossia is an early focal point of the work and I would like to expand
this to the future of banlieue language among well-educated speakers. He
emphasizes the importance of family in language choice as well as a normal
tendency towards using the vernacular.

2. 2. 4 Maghrebi connection

As referred to above, identification with the local area promotes use of the
language. Thus, many scholars have concentrated on what is assumed to be the
most productive subset of speakers, and certainly those who have claimed the
language as their own, those of Maghrebi heritage (Doran 2002, Meliani 2000).
Doran (2002) emphasizes the dual role of adolescence and minority status of
Verlan speakers. Indeed, her study focus is Maghrebi youth to the almost total
exclusion of other ethnicities. She recounts the story of a young Maghrebi girl
harassed by some obviously privileged Parisian youths, who have adopted a
stereotypical Call/era (‘gangster’) persona. The girl refuses to relate to them on a
banlieue level because they are not in-group. Doran’s (2002) analysis is that

Verlan is not a slang belonging to all youth groups equally. Rather, she says that

31

‘92!"

3‘4

it is a ‘sociolect’ generated and controlled by those who associate themselves
with the culture of the working class.

Melliani (2000: 16) focuses on l’identité mixte ‘mixed identity’ of the
speakers. In fact, she concentrates on second generation Maghrebi culture and
first generation ‘Maghrebi-Franco’ (Melliani 2000: 166) culture in Haute-
Nonnandie. This group claims Verlan as theirs, as in Doran’s (2002) study. The
city is racially diverse and includes the same main components as the northern
banlieues of Paris, though of slightly different proportions.

An interesting counter example to the dual ethnicity idea has been taken
up by Bouzini (2000). He refutes the fact that the younger generation is still
between two languages. In his thesis, he discovered that in fact, the younger
generation was learning the two languages simultaneously as native languages
instead of mastering neither of them. Carpin, Tavan, and Dugué (2005: 91)
support this claim as well, saying that 98 percent of children of immigrants either
born in the country or brought before the age of three admit to a maitn'se of the
language (speaking it well). Indeed, Tejedor de Felipe (2004) discovered in the
dictionary corpus that most innovations were French in nature, rather than more
linguistically Maghrebi. Billiez (1990) surnamed banlieue language ‘le parler
interethnique véhiculaire’, which may give us reason to explore the extent of

diffusion of this variety despite ethinicity.

32

2.3 Factors affecting the use of banlieue language

The following are social factors that may indeed affect whether or not a

person will use banlieue language.
2.3.1 Age

When talking about banlieue language, there are many major
generalizations about those who speak it. One of the most prevalent of these is
age. In the past, research has indicated age as a major factor, and is often
reflected in the very name for the linguistic qualities that pertain to banlieue
language. Therefore, many scholars used monikers such as ‘youth Ianguage’
(Doran 2002) or ‘young people's French’ (Boyer 1997), ‘I‘argot desjeunes’
(T ejedor de Felipe 2004), ‘the language of French adolescents’ (George 1986),
‘the lexis of the very young’ (Walter 1997) or the like. For others, it is not the
name that they have given banlieue language, but rather the scope of their
study, which limits the age of their informants to those who are adolescents and
very young adults (Doran 2002, Lepoutre 1997, Meliani 2000). Even among
those researchers who do not call it youth language, or one of the derivatives,
many of them have still mentioned a cut off that they believe to happen in the
late adolescent years.

Lepoutre (1997: 333) specifically indicates a cutoff, at age 16 to 25, at
which time said youth stops embracing the street culture and adheres more
closely to mainstream society. His assessment of those who continue to use the
variety into later adolescence and adulthood is defined as ‘des sous-cultures

déviantes et délinquantes’ (Lepoutre 1997: 334).

33

Doran (2002) agrees with Lepoutre, accepting the established view that
speakers of Verlan are quick to abandon it in their mid to late teens. She does,
however, admit that this may vary as the target population ages and that it may
eventually be carried on to the university. She analyses Verlan use according to
various age groups, starting at the pre-teen group where she identifies them as
looking to the older youth in the community. The second group is ‘early
adolescence’ (Doran 2002: 241), which she says was the most prolific. High
school is her cutoff. The age of 16, when students are legally allowed to quit
school, defines a turning point for Doran. She says that those who choose leave
school tend to also increase Verlan use. She says the early emphasis on career-
path choice in France leads to students evaluating the impact of language use
and its effect on a speaker’s future and that those who with to continue their
education will eschew less standard language in order to improve their linguistic
abilities. The only older speakers were the disruptive older siblings who would
occasionally interrupt the sessions, loudly speaking in Verlan for attention
purposes.

Not all scholars have completely relegated the use of Verlan or banlieue
langue to those under the age of 16. Boyer (2001) is based on questionnaires
given to students in the humanities and sciences at the DEUG level. His
informants, even as early as 1995 to 1997 when the surveys took place, rejected
the notion of banlieue language, in his case specifically Verlan, as being a youth

language. Zerling (1999) and Calvet (2006) also have both indicated that use of

34

this sociolect is not relegated uniquely to youth, though no empirical data has
been studied to prove this.

In recent years there seems to be a move away from ‘youth’ as it pertains
to banlieue language. Scholars have been more prone to name the variety in
question something that pertains to its identity factors, such as ‘language of the
inner and outer suburbs’ (Valdman 2000), ‘slang of the inner city’ (Nicolas 2005),
‘urban verlan’ (Zerling 1999), or finally, ‘contemporary urban French’ (Goudailler
2002)

2.3.2 Gender
Gender is one of the many social factors that can affect the ‘social monetary
value’ of the language (Bourdieu 1991). It is a well-established concept in
sociolinguistics that because of covert and overt prestige, women are usually
more inclined to defend the language of overt prestige (Labov 1972, Trudgill
1974). Especially during the child-rearing experience, women tend not only to
use the standard, or their approximation of the standard, but also to hypercorrect
in order to pass on ‘correct’ language to their children (Labov 1972). Along those
same lines, Trudgill's study (1972) shows that women tend to over-report their
use of the standard and men tend to under-report the same. Labov (1972)
speculates that women, because of a lack of real power in many societies,
depend on symbolic power (Bourdieu 1991) derived from language. In addition,
traditional sociolinguistic studies show that women are less likely to use a non-
prestige dialect and that men will normally be leaders in its change (Labov 1972,

Trudgill 1972). Doran's (2002) studies of the dialect support this as it applies to

35

French.

Applying these theories to France, Lefebvre (1991) found that French
women do tend to uphold the sociolinguistic gender norms. Pooley's (2003)
study upheld covert-overt rules among regional languages in France, finding that
French had become the language of women in the areas where Occitan was still
spoken, and the fact that men had preserved the regional language. Since a
regional language would have less prestige in a country in which the ‘standard’
is so important, it seems reasonable that women would use French, the
language with the most linguistic capital, and that men would tend to use the
regional language. Houdebine-Gravaud (2003) further proves the sociolinguistic
truths by underlining the fact that even in this day and age, vulgar vocabulary
often is still 'forbidden' to women and girls. Melliani's (2000) study did include
some younger female informants who used the variety and Doran (2002) did
discuss females using it, though again, less than the males. No one, however,
has investigated this quantitatively.

2. 3.3 Youth language
Eckert ( 1999) described language choices among students of different social
cliques in a local high school environment. In this study, she found direct ties
between the language variety a student chose to use and his or her social
position in the high school. Youth is a time of experimentation and construction
of identity, and language has been proven to be an expression of a person’s
identity (Aitsiselmi 1997, Doran 2002, George 1986, Labov 1972, Mela 1997). It

is therefore not surprising that during an era of uncertainty and rebellion that the

36

language choices made by those youth would also follow this trend. Youth tend
to be more innovative in their language choice. Banlieue language was first
marked as a language created by youth and its use has been linked to

identification with the urban environment (Boyer 1997, Boyer 2001, Doran 2002).

2.4 Language attitudes and perceptions of speakers

Baker (1992: 9) states, ‘In the life of a language, attitudes to that language
appear to be important in language restoration, preservation, decay or death. If a
community is grossly unfavourable to bilingual education or the imposition of a
“common” national language is attempted, language policy implementation is
unlikely to be successful.’ In other words, the way that people view a language,
and their identification with it will affect whether or not the speakers choose to
change. Their attitudes toward a given language (or variety) may affect its
longevity. While the bulk of his study concerned Welsh versus English in the UK,
the same can apply towards banlieue language, since the ties between identity
and banlieue language have already been established.

Preston (2002: 40) has done extensive work on linguistic attitudes and folk
linguistics in the United Statese ‘It is perhaps the least surprising thing
imaginable to find that attitudes towards languages and their varieties seem to
be tied to attitudes towards groups of people’ Language attitudes are inherent in
our society, and as Preston (2002) highlights, a speaker of a given variety is
often viewed in the same light as his language. Assumptions are made about his

or her intelligence level, amicability, work ethic, etc., simply based on the variety

37

of language a speaker chooses to use. Therefore, an understanding of this
folk correlation between group stereotypes and linguistic facts appears to be
particularly important to the study of the social identities we maintain and
respond to’ (Preston 2002: 41).

Githingi (2008) applies this area of study to Sheng, which we have
mentioned has marked similarities with banlieue language and is itself an urban
language. He explores the mixed attitudes towards Sheng in conjunction with
the perceptions of its speakers. Other studies on attitudes towards Sheng have
been done by Githiora (2002) and Fink (2005).

French, however, is the subject of very few attitudinal studies. Pecheur
(1988) studied attitudes of the 1000 French people toward their language,
querying literature and ‘degrading influences’ on French in recent history.
However, with that one exception, the few attitudinal studies concerning the
French language relate to areas outside the hexagon, where French is not the
only official language.

There exists a marked dearth of scholarly articles concerning banlieue
language is the speakers' own perceptions. Many scholars have reported that
their informants claim ‘Verlan’ specifically, and banlieue language by extension,
is a language they, themselves, created and that they ‘own’ (Doran 2002,
Melliani 2000, Séguin and Teillard 1996) even though Verlan itself dates much
further back than this century (Goudailler 1997, Goudaillier 2002, Merle 1997).
The French media is rife with stereotypes concerning the speakers’ level of

education and their linguistic prowess in French, or lack thereof. And even

38

among the scholars and educators, many (Calvet 2006, Doran 2002, Lepoutre
1997) have reported that those who speak the deepest vernaculars are often
those at the furthest margins of society. However, no one has asked speakers

what they think about the speakers‘ typical education level, age, or gender.

39

mg:
hmmw
3:513; M:

imwm

L l

CHAPTER 3: METHODOLOGY

3.0 Introduction

The goal of this study was not simply to test typical Verlan words but also
the extent to which the many languages in contact in the banlieue influence
banlieue vocabulary. As with any contact language situation, lexical items tend
to be borrowed or calqued from one language to another (Weinreich 1953), and
this study wishes to examine that phenomenon in French to a fuller extent than
previous studies have done (Boyer 2001, Calvet 1981,1987, Doran 2002, etc).
To that end, a questionnaire with a list of lexical items was developed with the

aid of a speaker in the banlieue area.

3.1 Development of the questionnaire
This speaker, age 30, was chosen because of his ties to the banlieue. He
identifies strongly with the area. His personal background reflects the diversity
found in the banlieue. His father was born in France to Italian parents who
immigrated to France from Italy after WWII and his mother can be described as
francais de souche, or native French.

The list he created was refined in several stages. It was first distributed
among several French natives from regions other than the banlieues of large,
urban areas. Lexical items this test group knew, such as chouilla, or ‘a little bit’ in
Arabic, were excluded. This was done to limit the questionnaire to specifically

banlieue vocabulary and the number of lexical items in general.

40

(:25 ta‘. ‘ilt

ya we: ~ i,.~‘

.‘6. «NM

‘é‘a‘s that

G‘atulary

 

The second step was to test the list among women from the area. Women
are less likely to be leaders of a change in a non-standard variety, as seen in
2.3.2 (Labov 1972, Trudgill 1974). Therefore the list was distributed to two
female residents of the banlieue, at that time in their mid 208, in order to select
words that were more mainstream banlieue, and also to control for vocabulary
that could be particular of the original speaker's idiolect. Finally, some very
mainstream slang lexical items were inserted in order to represent the francais
de souche who are also a part of this environment. Using more mainstream
French had a secondary function: to see if these words were used as often as
the other lexical items in the list or rejected by the banlieue residents. This also
means that someone outside the banlieue area may well recognize the

vocabulary.

3.2 Lexical items

The lexical items in the questionnaire included both individual words and
phrases that are from some of the many contact languages in the banlieue area,
along with some French slang words. Roma is represented by eight words:
marave ‘to beat up someone,’ bédave ‘to smoke marijuana,’ and its noun bédo,
which means ‘a marijuana cigarette,’ Chourave ‘to steal,’ se (la) natchave ‘to
leave,’ les schmidts, ‘police officers,’ and gadjo and gadgi, which mean a ‘non-
Roma male’ and ‘non-Roma female’ respectively. In the Creole category, there
are four lexical items: goumer ‘to beat up someone,’ makoume, ‘homosexual

male,’ and sa ou fe and sa ka mache, which both mean ‘how are you.’ Sa ou fe

41

is the more formal of the two, the latter being derived from marcher ‘to walk’
(Mondesir and Carrington, 1992). In colloquial French ‘Ca marche?’ comes from
‘ls it walking?’ and has come to mean ‘ls it working for youl how's it going?’
Arabic provides three items: choucrane ‘thank you,’ kiffer, ‘to really like
something,’ and ouaiche ‘how are you?’ Verlan contributes two words: keum
‘guy,’ and sky ‘whiskey.’ Finally, in argot there are four: damn and daronne,
which mean ‘mother’ and ‘father’ respectively, binouse for ‘a beer,’ and savater,

which means ‘to beat up someone.’

3.3 Questionnaire format

The questionnaire is comprised of two separate parts. The first page
concentrates on demographic information of the informant and poses questions
such as age and level of education, while the second part of the questionnaire
asked direct questions about the vocabulary.

3.3.1 Demographic information

This part of the questionnaire asked a subject's age, gender, employment,
languages spoken and other features. Because identity with the banlieue and
rejection of mainstream France have been determined as factors in whether or
not someone will use standard French or banlieue language (Calvet 1994,
Doran 2002, Melliani 2000), the questionnaire asks how long the informant has
lived in Garges and if close family members also live or lived in Garges. When
an informant did not live in Garges, he or she wrote the name of his or her city of

residence and answered the questions accordingly. This was based on Milroy's

42

(1980) social networking model which realized the importance of ties to an area,
such as whether or not informants have family members in the area, if they
socialize mainly with people from the surrounding areas, etc. in determining
what language variety people will be prone to use.2

The category of profession posed unexpected problems since most of the
informants were either actively seeking employment or were currently enrolled in
college. Only the private venue category held full time employment. In the first
two cases, the profession an informant hoped to practice was asked in order to
determine not only current/ future socioeconomic status, but as an indicator of
whether the informant would be more likely to want to stay in the banlieue or find
a way to leave it. As demonstrated by Labov’s Martha’s Vineyard study (1972),
identification with an area is a strong determinant in whether someone will want
to use the dialect of the region. Therefore, if the profession desired is something
that cannot be done in the banlieue, such as marine biologist, one would expect
that the informant in question would be less likely to use banlieue language.

The questionnaire next queries nationality and birthplace, since many
studies indicate that the primary users of banlieue language are of Maghrebi
origin, or the studies investigate that particular ethnic group while excluding
other nationalities (Doran 2002, Lepoutre 1997, Melliani 2000). The current
study wishes to determine if the lexical items do indeed represent banlieue
speech, regardless of their ethnic background, or if they are limited to a

particular ethnic group. Therefore, informants are asked if they were born in

 

2 It should be noted here that very few informants included information on their
family members. It was therefore not used.

43

11:2 as well a
it: a' thorn.
Flatten it
rise tangle
firsts well
13:93 3051
T’ :than so
376‘. Final
”r- 5? an:
3:355!” me
Iii" mtg'a
F I130 mas
iiiii‘i’ie Is
lite this in;
iii invest.
WI Schc

 

 

France as well as what their nationalities and ethnic origins are. Also, I ask
about an informant’s religion and which languages he or she speaks.

Religion was requested in order to see if it had any effect on how much
banlieue language informants know or use. To follow up this question,
informants were also asked whether or not they practiced the religion they had
indicated. Sociological studies have found that for France, religion has a bigger
impact than social class (Brulé 1966, Michelat and Simon 1977), so I decided to
test that. Finally, the language section asks which languages are spoken by the
informant and to what degree, including what his or her maternal language is. As
previously mentioned, media representation of the banlieue language speaker is
of an immigrant in scholastic failure who is typically caught between two cultures
and who masters neither French nor his or her native/heritage language. This
stereotype is also held as within certain scholarly circles (Hargreaves 1997).
While this may be the case for many speakers of banlieue language, the present
study investigates the limits of such an image found in both the media and
current scholarship, In addition to opening the question of whether we should be
examining banlieue language as a much bigger social phenomenon.

3.3.2 Word list

The second part of the questionnaire was comprised of the word list itself.
For each of the 21 lexical items, the informant was asked whether or not he or
she knew the lexical item in question. Each word was presented individually, so
no context was given to the informant. If the participants did not know the word

or phrase, they simply circled ‘I do not know this word’ without answering other

44

questions related to that particular item. They did not answer the following
questions about who, in their opinion, used the word, since by not knowing the
lexical item in question, they would have no basis for judgment.

If the participant claimed to know the word, three possible definitions were
given. One was the definition given by the native speaker, and agreed upon by
the female residents of the banlieue. This was used as the 'expected answer.’
The other two choices I selected by looking in a French dictionary where the
lexical item would have been placed alphabetically, in order to find an entry as
close as possible to the lexical item. For example, for goumer, I used ‘a
childhood illness’ because the word the French word gounne means ‘impetigo.’

Some definitions were clearly misleading, created by using a translation
or something believable for someone not from the area. Here a pertinent
example would be the word sky. In the banlieue, sky means ‘whisky,’ created by
Verlan, skiwhi and then apheresis. Therefore, as a misleading response, the
direct translation, ciel, was given. This is especially important because all
informants were informed that I was an American studying banlieue language.
They were thus informed that their knowledge of banlieue language and
vocabulary was important, not their knowledge of English.

After defining the lexical item, the informant was asked a series of
questions about each particular response. The first of those questions pertained
to self-reported usage. The informants had to report whether they often used the

lexical item in question, rarely used it, or never used it.

45

The remaining questions about each lexical item asked the participant to
describe the type of person that would use the vocabulary item in question.
lnfonnants were told that circling more than one answer was acceptable. The
first of these questions asked about gender. The informants had to decide if the
lexical item was used primarily by males, females, or equally by both genders.
The next question was about age. lnformants were given the choices ‘a younger
generation,’ ‘my generation,’ or ‘an older generation.’ Education level was
queried as well, with ‘not very educated,’ ‘somewhat educated,’ and ‘very
educated’ as possible responses.

The informant first had to circle the definition of the word or phrase for
each of the lexical items. This particular step was pivotal. If the informant
provided an expected answer, all other questions following the lexical item were
counted in the statistics. The opposite was also true: once an unexpected
definition was provided, all following questions pertaining to that particular lexical
item were excluded, since the infonnant's opinions would no longer be relevant.
In the example above, if a participant chose ciel as the meaning of sky, I did not
count that person's answers pertaining to whether or not they used it, who used
it, what age group used it, etc., since the focus of the study was to ascertain who
used banlieue vocabulary. lnformants also had the option of ‘refining’ the
definition on the space provided to the right of the choices, or writing their own

definitions.

46

 

3.4 Venues
l distributed the questionnaire in the Northern banlieues of Paris (93 —
Seine-Saint-Denis and 95 - Val d’Oise), from early December 2005 to early
January 2006, a mere few weeks after the riots that wracked the surrounding
areas. lnforrnants were solicited in three separate venues, including I’Univérsité
de Saint Denis in the 93, la Mission Locale in Garges-Les-Gonesse in the
95,and finally, with a group outside a specific venue; friends and acquaintances
were solicited at private holiday gatherings in Val D‘oise and Seine-Saint-Denis.
At each venue, l circulated, explaining the project and asking for
volunteers, distributing informed consent forms and questionnaires, and then
circled back through to answer any questions and collect the finished
questionnaires. I did not read the questionnaire to anyone or pronounce any of
the words. lnformants were told simply that l was an American student doing my
doctoral dissertation on banlieue vocabulary.
3. 4.1 L’université de Saint Denis
To test the hypothesis of retention, I chose to investigate banlieue language
among older, more educated speakers, especially since the speakers
investigated in many of the Verlan studies are now well into adulthood.
Additionally, it was important to test whether banlieue language really
disappears or simply goes underground in a diglossia. Did this variety grow with
the speakers in question, and do they use it with their peers?

Therefore, the first major venue was I’Université de Saint Denis. Students

participation was solicited during their lunch hour. In all, 35 vocabulary

47

questionnaires came from the university, 25 from female informants and 10 from
male informants. The locale was chosen for several reasons. Though Lepoutre
(1997) and Doran (2002) point out that banlieue language seems to disappear
among speakers after their mid-teens, especially if those speakers wish to
continue their education, no one has concentrated specifically on retention past
the age of 16. Boyer's study (2001 ) involved a questionnaire that was distributed
at colleges in Montpellier, Paris and Lille, but did not focus on age limit of
retention.

Additionally, as a university, the venue assured more diversity. Many of
the informants in this venue came from places other than the banlieue, though
most of them are geographically close, due to a French program called la carte
scolaire (the scholastic map), in which students are required to go to the
university in closest geographical proximity for at least the first year unless the
student places into one of the more prestigious schools. Therefore, some of the
informants did not know where Garges was, while others live there. This venue
allows us to investigate the depths of vocabulary knowledge as well as the
linguistic attitudes, since some speakers may very well recognize the words in
question but may not use the lexicon in order to be more upwardly mobile.

3. 4.2 La Mission Locale

The Mission Locale in Garges was the second venue. This organization is
the unemployment office for people between the ages of 16 and 26, though no
one under the age of 18 was considered for the study. La Mission Locale holds

workshops to help increase the employment among this age group while also

48

 

performing the duties of an unemployment office. While the education level
among this group is lower overall than those at the University, it does provide
another age group and a distinct difference in setting from the other venues.
Though much of the research has involved people of lower socio—economic
status, 18- to 26-year-olds constitute an older group of ‘youth.’

There were only 12 informants here, seven males and five females who
ranged in age from 18 to 25). As a group, their educational level was lower than
the others, though each informant had a degree. Most of them had a BEP
(seven informants or 58%). One person had a bac, one DEUG and one Maitrise
(8% each). Finally, this group also all lived in Garges and has less of a likelihood
of wanting to or being able to get out of the banlieue than the university group. In
fact, the professions they were aspiring to were those that could be done locally,
several at the local international airport, which according to the director of the
Mission Locale, hires quite a few of their youth. The professions they were
seeking were those of sales person, security officer, FedEx personnel,
mechanic, chauffer, business person, health aid, and accountant.

3. 4.3 Private holiday gatherings

Finally, the last venue is more of a group than an actual venue. Twenty
questionnaires were distributed among groups of friends in the area, ten to
females and ten to males. The questionnaires were distributed at various social
gatherings over the holidays. Most of these informants constitute yet an older
group than the two others, and they are informants with whom I am in-group.

This group was instrumental because they are all older, in their late twenties and

49

early thirties, and most of them are quite well educated. The vast majority of this
group also holds steady employment. In theory, the last two factors should lead
to either the most dropping of Verlan/ banlieue language or the most diglossia.

Therefore this group allows me to compare age and employment as factors.

3.5 Informant base
Here I discuss the demographics of the participants in the study.
3. 5.1 Gender

Overall, there were 67 informants counted in the statistics, 40 women
(59.8%) and 27 (40%) men. Of the 35 informants from the university, 25 (71.4%)
were female and 10 (28.5%) were male. There were 12 informants from the
Mission Locale, five (41.7%) females and seven (58.3%) males. The final 20
informants were part of the ‘friends’ venue, which was composed of ten female
and ten male informants.

3.5.2. Nationality

The informants’ nationalities were overwhelmingly French. Only 8 people (12%)
did not report a nationality. Almost 85% (59) of the informants who reported
nationality reported being French, and 84% (58) said that France was their
birthplace. An additional 5% claimed dual nationalities with Portugal, Bulgaria

and Poland, one informant each. Two informants were Algerian and Portuguese.

50

Romanian and the Reunion Island had one informant each. In the latter case,
the informant chose to list the Island, despite its being a DOM. 3
3.5.3 Ethnicity

A study of the informants' ethnicity creates a more accurate picture of the
informant base. Thirty-three informants, or 50%, said that French was their
ethnicity, origine familiale, on the questionnaire. Sixteen informants, or 24%,
said that they were from the Maghreb. Four informants, 6.6%, came from other
European countries, and two informants said they were from Central or South
America. Only 7.5% (five informants) said that they were from the Caribbean
and Mayotte. The rest were from African origins, four informants from West
Africa (6.6%); one from East Africa, and one that just stated African (1.5% each).

3.5.4 Religion

Religion was only counted in the positive; therefore, if an informant
answered the question about religion as ‘atheist,’ it was added to the ‘no’
answers because it is not a formal, organized religion. Therefore, only 36
informants are counted in the question on religion. Of those, 47% are Muslim (17
informants), followed closely by Catholicism at 36% (13 informants). Four
informants, 11%, simply stated that their religion was Christian, and 2.8% (one
informant each) said they were Protestant or Kabyle.

3.5.5 Department of Residency
Departments of residency vary, and six informants did not indicate their

place of residence. Those who did indicate residence came from six different

 

3 Numbers here, and throughout this dissertation, were rounded, in this case to
the nearest tenth. This will account for any discrepancies.

51

departments: 91 Hauts-de-Seine, 92 Essonne, 93 Seine-Saint-Denis, 77 Mame-
et-Seine, 95 Val d'Oise and 75 Paris. The largest group of the informants came
from Val d'Oise, at 51% (31 informants), followed by Seine-Saint-Denis at 24.6%
(15). The remaining 25% were broken up in Paris at 13.1% (eight), Hauts-de-
Seine at just under 4.9% (three), two informants from Val de Marne (3.3%), and
then 1.6%, one informant each, in Essonne and Val de Marne.
3.5.6 Age

Age is calculated in terms of groups laid out for a previous study in the lie
de France region. The logarithm computed the likelihood of buying a house,
based on age (Baccaini 1998). Since owning housing is arguably a sign of
adulthood, I chose this study to define my own age groups: 18 and 19, 20 - 24,
25 — 29, 30 — 36. Three informants were excluded due to their ages of 49, 58
and 59 respectively, because there would be too few results in each of the age
groups and the difference between their ages and the rest of the informants in
the study is quite important“. The largest category was 20 — 24 at 42% (28
informants). 24- to 29-year-olds represent 24% (16) of the informants and 30- to
36—year-olds made up 13.4%, 10 informants.

3. 5. 7 Education
Sixty-five informants answered the question of education. The informants'

education levels ranged from no formal degree to the doctorate level

 

4 In looking over the questionnaires, the 49-year-old knew almost all the words
due to his deep attachment to the cité, and the 58-year-old woman used many of
the words as a sign of linguistic capital (Bourdieu 1982) with her sons between
the ages of 27 and 32. The 59-year-old man knew very few of the words, with the
exception of the Creole vocabulary.

52

(DEA/DESS). The question asked them the highest level of education that they
had completed, not what they were doing at the present time. Only one
informant did not answer the question. Of those who did, only one informant had
no formal education, and he was excluded due to his age. In order from lowest
education to highest: 18.5% (12) of informants had completed the bac (short for
baccalaunéat, an equivalent to high school but with a grueling exam at the end),
21.5% (14 informants) had a BEP (an alternative, technical school degree),
3.1% (two) had a BTS (a degree that is two years after the high school or
technical school experience, geared toward learning in a technical environment.
Some engineers take this route.)—15.4% (10) had done the DEUG (a degree
that no longer exists in France, equivalent to an associate's degree in the US),
23% (15) had done the Licence (a three-year degree equivalent to the US.
bachelor’s degree), 13.9% (nine) had a Maitrise (equivalent to a Master's
degree), 1.5% (one) had completed a DEA, and 3% (two) had completed a
DESS (both terminal degrees that can be done right before a doctorate). Overall,
it was a fairly well educated group.
3. 5.8 Languages

Languages were recorded as follows: if an informant wrote down the
fluency of the language in question, the languages were put in order of
knowledge. For instance, if an informant wrote: ‘English: beginner, French:
native, German: fluent,’ French was recorded as L1, German as L2, and English

as L3. If no level was reported, languages were recorded in the order they were

53

written on the questionnaire, based on the assumption that more solid
languages would come to mind first.
3.5.8.1 First language

Sixty-five people answered this question. A high percentage of 78.5% (51)
of the informants reported French as their native language. Arabic, Kabyle and
Spanish were next, though a long way behind, with 4.6% each (three speakers).
Creole, Polish, Portuguese and Soninke, a language of Senegal, had one
speaker each (1 .5%).
3. 5. 8. 2 Second language

Fifty-seven informants responded to the question of second language. Of
those, only 19.4% said that French was their second language (10 informants).
The effects of language instruction in school are also shown in the L2 category,
since 52.6% (30) of the informants listed English as a second language. Five of
the informants (8.8%) reported Arabic as an L2. Spanish and Italian were at 5%
(three) each, followed be German at 3.5%, (two informants) and Berber, Creole,
Portuguese at one informant each (1.8%).
3.5.8.3 Third language

Forty-two informants had a third language on their questionnaire. English
was the biggest group here as well, at 42.9% (18). Spanish followed closely with
38.1% (16 informants), and Portuguese had 7.1% (three). German was reported
by almost 5% (two) of the informants. Arabic, French and Italian were reported

by one informant each, or 1.8%.

54

‘Nq.’ .

"'33;

3. 5.9 Professions

It is important to remember here that well over half of the informants were
recruited in environments where they were either actively seeking a job or were
at the university in order to complete a degree to obtain a certain job. Therefore,
for the case of this study we do not distinguish between a profession an
informant is currently practicing and a profession an informant would like to
practice.

The largest group in this category was those who were or wanted to be
teachers (10.2%). Business was next with 8.47%, closely followed by
chauffeur/driver at 6.8%. The professions of day-care professional (animateur)
and engineer were represented at 5% each. The next most popular professions
were computer specialist and accountant at 3.4% each. The rest of the
professions were chosen by one informant each: Fedex worker, mechanic,
health aid, trainer, human resources, supervisor, manager, coordinator of
production, film distribution, cinematographer, artist, documents specialist, and

ethnographer.

3.6 Statistical analysis methodology
Two statistical programs were used for this dissertation. The following
sections describe both programs, their functions, and why they were used.
3. 6.1 Systat
One statistical program used to compile results was Systat. It was used to

compute all simple, mathematical procedures, such as the raw percentages for

55

. *i ,

£3103 Th

9.73895 '
azepfior to
2’. 21. etc.)
East-aim I!
MAI-4'

bate 0P. 3

3:: wcuéd

 

the descriptive statistics. The program was able to count tokens among the
various factors and show them side-by-side to facilitate comparisons. For
example, with this program I was able to find out how many people in any given
age group said that a particular word was used by people with a low level of
education. This program was not used to select statistically significant factors,
but to simply compile percentages.

In most cases, factors in this section were treated individually, meaning
that those informants who came from Paris and those who came from Seine-
Saint-Denis were not grouped together, i.e. the factor of residency. The notable
exception to this rule is in the section of age, where l divided the ages (18, 19,
20, 21, etc.) according to the age groups given in the study of lle de France
(Baccaini 1998), as mentioned earlier. Again, this study created age groups
based on a logarithm that determined how likely it was that a person of a given
age would own a home in France. This is a study on a variety that was once
discussed as a ‘youth’ language. Therefore the likelihood of home ownership—-
arguably a mark of adulthood—is a good parallel.

3.6.2 Goldvarb

Goldvarb is a binomial logistic regression program that is used especially
when independent variables are sufficiently complex to include cells of unequal
number. In my study, in order to be as thorough as possible, initially, all social
factors named above were included. Because of this, it is virtually impossible
that in a sample of only 70 original informants, there would be an equal number

of males and females in the same age group from Val d'Oise with a licence and

56

the same ethnicity. This makes most other statistical programs invalid for
statistical procedures of this nature.

The study of lexical items was twofold. I decided for the purposes of this
study to test actual knowledge of the vocabulary items in question among the
demographics first and then test the self-admitted usage of the items, i.e.
whether or not informants admitted to these lexical items being part of their
vocabulary. Therefore, the first dependant variable was knowledge, and the
second variable was self-admitted usage.

Since Golvarb only recognizes binary categories, the information had to
be adapted. To compute knowledge, therefore, any non-answer or unexpected
answer was counted as not knowing the lexical item, and any expected answer
was counted as knowledge of the word. To compute usage, those who did not
know the lexical item and those who said they never used it were grouped
together. Those informants who admitted to using the vocabulary at either level,
sometimes or rarely, were also grouped together.

The original independent factors were gender, age, education level last
completed by the informant, which French department they live in, country of
origin/birthplace, nationalities, ethnicities, religion, languages they speak, and
the origin of the word. A discussion follows on each of the factors.
3.6.2.1 Age

As a reminder, the age groups were broken up into the following groups:
18 and 19, 20 - 24, 25 - 29 and 30 - 36 based on Baccaini‘s (1998) logarithm.

No age groups could be collapsed.

57

 

3. 6. 2. 2 Religion

For statistical purposes, religion was divided into two categories——
practicing versus non-practicing—since Goldvarb discarded religious affiliation.
3. 6.2.3 Ethnicity

Initially ethnicity was divided into two categories: first ethnicity and second
ethnicity. After initial testing, the only significant ethnicity group in the first run
was the second ethnicity. In the initial run, it was determined, by both similarity
of weights and category, that much of the information in the second ethnicity
could be combined. The original categories in this section were Italian, Polish,
Irish, Portuguese, Algerian, Tunisian, Morocan, West African, Spanish, German,
Bulgarian, Rumanian, Eastern European, Caribbean, Commoran, Mayotte and
South/Central American. Italian was eliminated as a ‘knock-out,’ which means
that the one informant of Italian descent knew all of the vocabulary items and
therefore his information had to be eliminated in order for the statistical program
to run.

The weights for each of the individual countries that make up the Maghreb
were close enough to justify their being collapsed, and since they also share
similar geography and culture, Algeria, Morocco, and Tunisia could be collapsed
as the Maghreb. African countries as well provided similar weights, regardless of
location within the continent. Finally, when another ethnicity was combined with
French, the weights were similar. Therefore, the final categories for second
ethnicity were African, Polish, Spanish, Portuguese, Maghrebi, German, and

South/Central American.

58

3. 6.2.4 Education

There were informants in each educational level from no official degree to
BEP, the bac, BTS, DEUG, License, Maitrise, and DEA, DESS. In this group,
because of weights on the first few runs, very little could be collapsed, which
resulted in only a merger of DEA and DESS because their weights were nearly
identical and they are of the same level of education (5 years after the
baccalauréat).

3. 6. 2. 5 Languages

Languages, originally classified as L1, L2 and L3, contained German,
English, Spanish, Kabyle, Arabic, Soninke, Berber, Bulgarian, Rumanian, Italian,
Creole, Portuguese, Polish, and Irish. In the French system, students usually
learn two languages during their school years, often including Spanish, English
and German. With the first few runs, it became apparent that these categories
could be further simplified, again because of weights, into ‘scholastic language’
and ‘heritage language.’

To determine which category applied to each informant, l cross-referenced
his or her ethnicity, nationality, and country of origin. For example, in almost
every case, English, Spanish and German are scholastic languages. In the rare
event when an informant listed Spanish as an ethnicity, nationality, or origin, the
answer was grouped with heritage language. If not, the language was listed as
scholastic. Portuguese, Arabic, Kabyle, and less common African languages

such as Soninke are not often taught in French schools, but again, the

59

 

ethnicities, nationalities, and origins were cross-referenced. No categories were

collapsed.

60

CHAPTER 4: RESULTS

4.0 Introduction of results

This chapter will present and discuss the results from three different
angles. First we will look at the individual lexical items, organized by language.
For each word, I will discuss what perceptions the informants have toward these
words. Their results are compiled according to what they believe to be the
gender, educational level, and age of someone who would use each word. I will
then provide some analysis about what these results mean.

After that will come a section that addresses the discrepancy between the
perception of who uses such words and the actual usage claimed by the
informants for themselves. This section will first focus on gender, then age and
education. In addition to this, self-reported usage by the informants is analyzed
in terms of venue and the items’ original language.

Finally, factors that are significant in both the perception and admitted/self-
reported usage will be investigated. These factors include the social factors
discussed in section 4.3, including ethnicity, nationality, religion, birthplace, and

place of residence.

4.1 Analysis of individual lexical items

In this section, I explore each of the contributing languages separately and

the lexical items that fall into those languages.

61

4.1.1 Roma
This language was included because of the recent increase of Eastern
immigration and the nomadic tendencies of the Roma people. As they are not as
integrated in the society as other groups, it will be interesting to see the extent of
the influence of this language in banlieue language. Tejedor de Felipe (2004)
notes their lack of contribution in the Séguin and Teillard (1996) corpus.

At the time of the fieldwork, there were several camps of Roma people in
the Garges area. The government sets up areas for these camps in order to give
nomadic peOple access to clean water and the children access to schooling.
However, from what my informants said, most French people tend to avoid
Contact with these areas. There are also many people from the east who have
settled permanently in the area and enrolled their children in school.
4.1.1.1 Marave5

This verb, meaning ‘to fightl beat up,’ belongs to the category of banlieue
language that does not show its flexion by changing forms while tensed. This
well-documented phenomenon (Goudailler 1997, Séguin and Teillard 1996) has
allowed some researchers to compare banlieue language to a Creole (Jablonka
2003). Whether it is used in passe compose, future, or present, there is no
audible, or overt, flexion to the word itself. Je l’ai marave. Je vais le marave. Je
le marave ‘I beat him up. I’m going to beat him up. I’m beating him up.’ As se

battle or massacrer ‘to fight,’ it is used in the physical, literal sense only. It

 

5 The reader should note that these spellings were given to me by my informant,
but as Séguin and Teillard (1996) note, since banlieue language is oral in nature,
spellings vary. In fact, in their study, there were as many as 11 different spellings
for some lexical items.

62

cannot be used figuratively to fight for something. It is, therefore, a word with
inherently violent connotations. 6

4.1.1.1.1 Gender

Out of a total of 52 expected answers, only one informant (2%) said that
primarily women used the word marave, while 22 informants (42%) claimed that
its users were masculine, and 29 informants (56%) attributed its use to both
sexes.

4.1.1.1.2 Education

Forty-eight informants responded to this question. It is interesting that
when l tabulate the lowest three responses, ‘not very,’ ‘somewhat,’ and the
combination of those two, 68.8% (33 informants) believed that marave was in
those lower categories for education. Of those 33 informants, nine chose ‘not
very’ (18.8%) and 20 chose ‘somewhat’ educated (41.7%). An additional four
informants (8.3%) chose the combination of said lower categories. One could
postulate that the violent nature of the word would cause it to be used mainly by
people who would physically fight, which is often something left to lesser-
educated people. Perhaps most surprising is the fact that five peOpIe circled
‘very’ educated (10.4%) and another 10 said that people of all educations use it,
which is 20.8% of the informants. I should at this point like to propose that
indicating ‘very educated’ and indicating all three categories constitutes the
same thing. Since we are looking at a non-prestige variety, the covert prestige in

use comes from the lower classes and less educated. A non-prestige variety

 

6 Again, numbers in the following statistics may not add up to 100 due to
rounding to the nearest tenth.

63

does not usually start out as change from above. If a non-standard lexical item is
used also by the very educated, one would expect that the word is used by
education levels below as well, given sociolinguistic tendencies toward change
from below (Labov 1972).

4.1.1.1.3 Age

Fifty-one informants chose to answer this question. I will discuss age by
age groups, starting with the youngest group and ending with a discussion of the
oldest age group. Nine members of the youngest group answered this question.
Of those, the largest group selected ‘a younger generation,’ at 55.6%, or five
informants. The second largest category for them was ‘my generation’ with three
informants, or 33.3%. A combination of his or her own generation and a younger
generation was chosen by one informant, or 11.1%. This is what we might
expect as being part of the stereotypical ‘youth language.’ However, the results
do not follow the expected pattern as well as one might expect as the age
groups of the informants get older.

The next group, those from age 20 to 25, had 21 informants participate. Of
those, the largest group chose their own generation (10 informants, at 47.7%).
Seven informants chose a younger generation (33.3%) and two informants
chose a combination of the latter two categories (9.5%). However, for the first
time we see and older generation in the statistics. One informant (4.8%) chose
an older generation, and one informant chose a combination of his own

generation with that of an older generation.

64

For the 25— to 29-year-olds, there were 13 responses. Of those, the largest
category was a younger generation in combination with their own generation—
six informants, or 46.15%. Five informants (38.5) chose simply their own
generation; one informant (7.7%) chose a younger generation and one chose all
generations (77%).

Finally, let us look at the oldest group. There were eight responses. Of
those, the largest group, three informants (37.5%) chose a younger generation
as the primary users of this word. Two informants chose their own generation
(25%) and one informant chose a combination of the two (12.5%). Two (25%)
informants chose all three categories of age.

In analyzing this data, it looks as if the word belongs mostly to people
under 30. After 30, the ‘younger’ category increases, and the category of those
who claim specifically their own generation decreases. If we consider that the
word means ‘to fight,’ it may be that fighting is an act more associated with youth
and less associated with those who are older and who have fulltime jobs.
4.1.1.2 Chourave

This verb means ‘to steal,’ and just like marave, does not show overt
flexion. While some people knew the word outside of the banlieue area, many
sources include this word in examples of banlieue vocabulary (Goudailler 1997,

Séguin and Teillard 1996).

65

4.1.1.2.1 Gender

Out of 52 informants who responded to the question, 17 (32.7%) chose
males as the primary users of the word. Only one person (2%) chose females,
and 34 respondents (65.3%) stated that both sexes use the word.

4.1.1.2.2 Education

Eleven respondents chose all three categories of education, which
corresponds to 23% out of a total of 48 responses on this section. Another three
people (6%) chose ‘very educated,’ and one person chose both ‘not very
educated’ and ‘very educated,’ skipping the middle category. I will read both the
latter responses as ‘all,’ which, when condensed, amount to 30% of
respondents. The category chosen most often was “somewhat educated’ at 15
informants (31%). ‘Not very educated’ was chosen by 10 informants (20.8%),
and an additional seven informants (14.5%) chose a combination of the lower
educational levels.

4.1.1.2.3 Age

The youngest generation had eight respondents. Of those, three chose
their own generation (37.6%), two chose their own generation in combination
with an older generation (25%), and two chose their own generation in
combination with a younger generation (25%). Only one informant in the
youngest age group chose an older generation.

Of the second group, 20 informants answered the question on age. Of
those, the largest category chose their own generation with nine informants

(45%). ‘A younger generation’ at eight informants (40%) was the next most

66

 

 

 

popular choice among people from 20 — 24 years of age. Two informants (10%)
chose their generation in combination with a younger generation, and one
informant (5%) chose an older generation.

In the third age group, there were 14 respondents. Of those, the largest
group chose their generation in combination with that of a younger generation,
six informants, (42.9%). Their own generation by itself was chosen by the next
largest group of informants, five, at 35.7%. Two informants said that the word is
used by all generations (14.3%), and one informant relegated it to an older
generation (7.1 %).

Finally, in the oldest group eight people responded. Of those, 50%, or four
informants, chose their own generation in combination with a younger
generation. One informant each chose a younger generation, an older
generation, their generation in combination with an older generation, and their
own generation.

Interestingly enough, almost all of the respondents who chose ‘a younger
generation’ under the age of 24, are female (11 females, one male). This may be
due to the fact that they are relegating banlieue language to a younger age
group in order to distance themselves from the banlieue. Most of the informants
in this case who chose the younger generation were in the university setting or
had their own job.
4.1.1.3 Se (la) natchave

This lexical item is a verb, which means ‘to leave.’ The original informant

who helped to compile the list omitted the la but the two females who helped

67

revise and modify the list stated that they had only heard it with the particle. This
lexical item was not as well known among my informants.

4.1.1.3.1 Gender

Only 16 people answered this question, and of those, seven (43.7%) said
that men were the primary users and the rest attributed its youth equally to both
sexes.

4.1.1.3.2 Education

Out of a total of 15 answers on this section, six people chose ‘somewhat
educated,’ which, at 40%, constitutes the largest group. Four others (26.6%)
chose ‘not very educated’ and three people (20%) chose a combination of those
two factors. Only two people chose all levels of education, corresponding to
13.3%.

4.1.1.3.3 Age

Only 16 informants in total replied to this question. When we look at these
by age groups, we see that the answers are mixed. No one in the youngest
group even replied to this question. In the second group, eight respondents
replied. Of those, five (62.5%) claimed that it belongs to their own generation.
Two informants (25%) claimed it belongs to a younger generation, and one
claimed that it belongs to an older generation ( 12.5%).

In the third group, five people answered. Two (40%) said that it belonged
to their own generation and two others (40%) said that it belonged to their own
generation in combination with a younger generation. A final informant attributed

this word to his generation in combination with an older generation (20%).

68

In the third group, there were three people who answered this question,
and each chose a different answer. One chose his own generation, and the
other two each combined their own generation with an older and a younger
respectively. Ironically, while five informants between the ages of twenty-one
and thirty claimed that the principle users of the word are either younger or
younger in combination with their own generation, no informant under 21
responded to the question. This seems to support that they may have heard it
but did not use it, and assuming it was new banlieue language, attributed it to
younger speakers. It also may be that this lexical item has an aging user base,
which is supported by Calvet (2006).
4.1.1.4 Schmidts

This plural noun means refers to an officer of the law. Its more literal
equivalent would be ‘cops,’ or another moniker with more derision. Goudailler
(1997) explains that the lexical item is descendent from the Roma word for
metal, and by extension handcuffs, which by metonymy came to mean police
officer. Many informants gave an alternative meaning, ‘Germans,’ and some
even reported using the word as such, but only expected answers were counted
in the statistics since I had not been able to verify this meaning elsewhere and
rule out the possibility of a guess on the part of an informant.

4.1.1.4.1 Gender

Only 30 informants responded to this question with an expected answer.
Of those 30 expected answers, 18 informants (60%) said both sexes use the

word equally, and the remaining 12 (40%) stated that men mainly used the word.

69

4.1.1.4.2 Education

When we look at expected answers, there are 24 for education level of
those using the word schmidts. Of those, only two informants (8.3%) chose a
low level of education. Four others split their choice between the two lower
levels of education (16.7%). From there the majority speaks with 'somewhat
educated' at 33.3%, or eight informants; and the largest category, ‘all levels of
education,’ at 10 informants, or 41.7%.

4.1.1.4.3 Age

Twenty-seven informants gave information of age groups for users of the
word schmidts, using the expected answer. Breaking this information down into
groups, we see that the youngest group had only four expected responses. Of
those, two selected an older generation (50%), and one informant each selected
their generation and their generation and all generations. The second age group
had 12 respondents. The largest category was ‘my generation’ at five informants
(41.7%), followed by ‘an older generation’ at three informants (25%). Two chose
a younger generation (16.7%), and one informant each chose all generations
and a younger generation in combination with their own (8.3%). The third oldest
age group had six respondents. Of those, three chose an older generation in
combination with their own (50%). The other three each chose different answers.
one for all generations, one for a younger generation in combination with their
own, and one an older generation by itself, (16.7% each). Finally, the oldest
group had five informants reply to this question with an expected answer. Of

those, the largest category was ‘all,’ with three votes (60%). One person chose

70

their own generation exclusively and one chose their own generation in
combination with that of an older generation at 20% each. This seems to support
an aging user base.
4.1.1.5 Gadjo

This masculine noun refers to a male person, though some people specify
a non-gypsy male, which is the original Roma definition (Goudailler 1997).

4.1.1.51 Gender

Forty-five informants answered this section. Of those, 35.5% (16
informants) said that gadjo is used primarily by males. Only 8.8% (four
informants) said that it is used primarily by females. A majority of 55.5% (25
informants) said that it is used equally by both sexes.

4.1.1.5.2 Education

Overall, 17% of 41 informants answering this question indicated that the
word is used by all education groups. Fourteen people (34%) stated that the
word is used by the ‘somewhat educated.’ Eight people (19.5%) said that it is
used by those with little education, and an equal number of informants selected
a combination of ‘not very educated’ and ‘somewhat educated.’

4.1.1.5.3 Age

In global terms, 44 informants answered this question. The youngest
9roup had six respondents. Of those, three, or 50%, chose their own generation,
one (16.7%) chose an older generation and the remaining two informants, one
each (16.7% each) chose a combination of their own generation with a younger

and an older generation respectively.

71

The second age group had 19 respondents, but even these were split as
to which generation the lexical item belonged to. The largest group was ‘my
generation’ at six informants (31.6%), four people chose a younger generation
(2 1 .1%), and another four respondents chose an older generation (21.1%). Two
informants (10.5%) chose ‘all,’ and two others chose their generation in

combination with a younger generation. A final participant chose his own

generation in combination with an older generation (5.3%).

The 24- to 29-year-olds had 11 respondents. Their largest group was their

Own generation, at four participants, or 36.4%. Three others chose an older
generation (27.3%). Two chose all generations (18.2%). One chose his own

generation and an older generation, and another chose his own generation in

Combination with a younger generation (9.1% each).
The oldest age group had six informants. Two of them (33.3%) picked an

Older generation, and two picked ‘all’ (33.3%). One participant (16.7%) chose his

Own generation, and another (16.7%) chose the his own in combination with a

Younger generation.

4- 1.1.6 Gadgi
This lexical item is a feminine noun, meaning ‘a woman,’ or woman who is

n0t a gypsy (Goudailler 1997). Fewer people knew this word, in spite of the fact

t"lat it closely resembles the previous entry.

4.1.1.6.1 Gender
Only 30 peOple responded to this question. Of those informants, slightly

o\Ier half (16 informants, or 53%) chose ‘both sexes’ as the primary users of the

72

word. What is extremely interesting is that the remaining 14 chose equally
between men and women, both at seven informants, or 23.3%. Gadji was also
one of the few in which women were highly chosen as primary speakers of the
lexical item by a relatively large percentage of the informants. It is possible that
in this case the informants connect the word's meaning with who would use it.
4.1.1.6.2 Education
Twenty-eight people responded to this question. The largest category was
'somewhat educated’ at nine informants, or 32%, followed closely by ‘not very
educated’ at eight people, or 28.5%. Another four people (14%) chose a
Combination of the latter two categories. Five people chose ‘all,’ amounting to

1 8% of those who responded, and a mere two informants (7%) chose ‘very

e<incated.’
4.1.1.6.3 Age

Only 26 informants answered the question on age for gadgi. The youngest
QFOUp had three respondents, two (66.7%) of whom chose their own generation
and one (33.3%) who chose an older generation.

The 20- to 24-year-olds had 11 respondents. The largest group, five

people (45.5%), chose their own generation. One of them (9.1%) chose all
generations. Three chose a younger generation (27.3%), and two chose a older

generation (18.2%).

The 25- to 29-year-old group had only eight respondents. Of those, three
Chose their own generation (37.5%), followed by two who chose all generations,

(25%). One person chose a younger generation (12.5%), one informant chose

73

at older generattr
mbhatlon with
The oldest
heir responses.
generation and t
With these
Scaused by the
a"fitters attribu
an aging user b.
lelletatjons, an
llliBédave
Meaning
liestionnalre (1
4.1.1.7. l
Founeen
“New, and .
bomsexes.
41.1.7.2

EdUCafiC

rhoanswered

htereslingly er

 

an older generation (12.5%), and a final informant chose his own generation in
combination with that of an older generation.

The oldest set had five respondents, but they were more consistent in
their responses. Two (40%) chose all generations, one (20%) chose his own
generation and two others (40%) chose an older generation.

With these results, a possible interpretation is that the variety in responses

is caused by the fact that the informants hear the word but do not use it, and
therefore attribute it either to an older or a younger generation. It seems to have

an aging user base, since the oldest group attributes it to their own, all, or older
generations, and never to a younger generation.

4- 1. 1.7Bédave

Meaning ‘to smoke a joint,’ this verb, in accordance with the others in the
questionnaire that end in -ave, does not carry overt flexion.

4. 1. 1. 7.1 Gender

Fourteen respondents (26%) claimed that men were the primary users of

this word, and 40 informants, an overwhelming 74%, said the word was used by

both sexes.

4. 1 . 1. 7.2 Education

Education level proved difficult for this question. Out of 46 respondents
Who answered this question, 34% ( 16 informants) circled all three categories.
lf‘lterestingly enough, only one person chose ‘very well educated.’ ‘Not very
educated’ was only chosen by 17.5% of the respondents, while ‘moderately

educated’ was the second largest category with 28% (13 informants). Four

74

respondents (8.6%) chose a combination of ‘not very educated’ and ‘somewhat

educated.’ Also allowing us to infer that education may not be as important as
previous research would indicate, one informant chose a combination of ‘not
very educated’ and ‘very educated.’ Since it would be unusual for a word to be

used only by the highly educated and the poorly educated, skipping those in the
middle completely, I counted this answer under ‘all.’

4.1.1.7.3 Age

The youngest group of informants had eight respondents. Of those, six, or 75%,
Chose their own generation. One person (12.5%) chose all generations, and a
fi"a! person (12.5%) chose an older generation.

The second youngest group had 19 informants. Eleven of those, or 57.9%
Chose their own generation. Four chose the their own generation in combination
With a younger generation (21.1%), and two relegated the word to a younger
generation (10.5%). One person said ‘all’ (5.3%), and another relegated the
Word to an older generation (5.3%).

The group of 25- to 29-year-olds has 12 informants. Of those, the largest
group was their own generation in combination with a younger generation, with
fiVe informants, or 41.7%. Four others chose simply their own generation, at
33.3%. Three people chose all generations, at 33.3%.

The oldest group had seven informants respond to this question. Of those,
most (four, or 57.1%) chose their own generation in combination with a younger

generation. Two (28.6%) chose all generations, and a final informant (14.3%)

chose his own generation.

75

What this tells us about this word is that it is well ingrained into the
vocabulary. Very few people, two in total over all the groups, attribute it to an

older generation, and the same number of people attribute it to a younger

generation.

4. 1.1.8Bédo

This noun means ‘a cigarette containing marijuana.’

4.1. 1.8.1 Gender

The ovenrvhelming majority of the informants (50 out of 59, or 84.7%)

Chose both sexes as using bédo equally. Only one person, or 1.7%, chose

fethales as the primary users. The remaining eight respondents (13.5%) chose

males as the primary users.

4.1.1.8.2 Education

Bédo had the highest level of informants choosing all three categories for
the educational levels of all of the lexical items on the questionnaire. Out of a
tOtal of 46 responses, a majority of 24 (52%) chose all levels of education. Two
additional informants (4.3%) chose ‘very educated,’ and two others (4.3%) chose
'Very educated” in combination with ‘somewhat educated.’ On the other end of
the spectrum, five informants (10.9%) chose ‘not very educated,’ and another

6.5% (three informants) chose ‘not very educated’ in combination with

'somewhat educated.’ Eleven respondents (23.9%) chose ‘somewhat educated’

alone.

76

4.1.1.8.3 Age
When we break down the results by age group, we see that the youngest
group had nine informants that replied to this question. Of those, five (55.6%)
said that the word belonged to their own generation. Two informants (22.2%)
claimed that it belongs to all generations. One person chose a younger
generation (1.1%), and another chose his own generation in combination with
that of a younger generation (1.1%).

The second group had 23 respondents. Of this group, 12 (52.2%) chose
their own generation. Five (21.7%), the next largest group, chose the latter in
cOmbination with a younger generation. Three (13%) chose ‘all.’ Two informants,
or 8.7%, chose an older generation. One final informant chose a younger
generation (4.4%).

The third group had 14 informants. Of those, only one chose his own
generation by itself (7.1%), but six chose their own generation in combination

With a younger generation (42.9%). Four others (50%) chose all generations.

For the oldest group, we see that there were eight informants who
reSponded to this question. Of those, 50%, or four informants, chose all
generations. One informant (12.5%) chose his own generation, and three

(37.5%) chose the latter in combination with a younger generation.
It is clear from this information that bédo is fully integrated into banlieue
Ianguage. The most p0pular response, for all age groups, included the

respondant’s own generation as users of the word.

77

4. 1.1.9 Roma analysis

As one of the languages best represented by this study in terms of sheer
number of lexical items, the informants were familiar with much of the
vocabulary. The vocabulary in this section ranges from se (la) natchave or gadji,
with relatively few informants knowing the lexical item, to bédave and bédo,
Where almost everyone knew the word. This would be in direct contrast to
Tejedor de Felipe's (2004) analysis, which suggested that Roma was not as

large a contributor to banlieue language as other languages.

In most cases throughout the Roma lexical items used in this study,7 we
See that previous literature partially predicts this study's gender outcome.
Women by themselves are not considered the primary speakers of this variety
by any majority of informants. This follows Doran (2002), specifically relating to
banlieue language; and in the broader focus of France as a country, the
l’eeearch done by Lefebvre (1991), in which French women upheld

Sociolinguistic gender norms, and Pooley (2003), in which women spoke the

prestige dialect of French instead of the regional variety.

Men were viewed as the primary speakers of the variety at around 40% for

most of the lexical items in Roma. Very few (under 4% for most lexical items in

Roma) suggested that women are primary users. When we compare the two, we

\

A reminder to the reader may be necessary: the vocabulary in this (and all
1“ellowing language sections) was chosen by the original informant from the area
and heavily invested in its culture. While some of the vocabulary is found in
Séguin and Teillard's (1996) study as well as Goudailler's (1997) study, which at
least confirms that it was used in the mid to late 90s, I cannot state that this

Vocabulary is representative, or that they are the most frequent Roma words in
banlieue language.

78

find that previous literature concerning both the concepts of overt/covert prestige

(Labov 1972) and linguistic market (Bourdieu 1991) seem to predict the

outcomes of the research.
However, when we include the percentage of informants who feel that the

items in question are spoken primarily by both genders, the previous literature
On gender is not as accurate a predictor. In fact, throughout the category of
Roma, around 60% of the informants say that these lexical items are used
Primarily by both genders. This may indicate that, instead of a simple

"Wale/female division, other factors transcend gender.
Linguistic capital (Bourdieu 1991) and solidarity among speakers of the

V3 l‘iety generated by using the vocabulary may be more important than gender.

TI"le results for bédo (84% for equal usage among the sexes) and bédave (74%)
Seem to support this. These lexical items refer to an illicit activity (smoking

marijuana). By the very nature of its proscription by hegemonic society, which
created and upholds the very laws prohibiting this activity, those who choose to
partake separate themselves from the society that forbids it. In theory, women,

prone to use a standard more than men, should avoid usage of this vocabulary.
These words, however, seem to be a large part of banlieue culture. It is possible

that the linguistic capital of the lexical items is more important than gender

norms.
With regards to education, scholars have indicated that the variety is

dropped by those who go on to a formal educational setting (Doran 2002,

Lepoutre 1997). Though Boyer's (2001) focus was not to prove or disprove the

79

stereotypes regarding the education level of someone who uses the variety, he
did conduct research at a university setting on the assumption that the students
would speak it. In general, since this is a non-prestige dialect, the hypothesis is
that it would not be attributed to very high levels of education.

Lexical items for Roma were partially viewed by the informants as we
might expect. A range of 25% to 46% of the informants for each of the lexical
items choose either low education by itself or low in combination with a mid-level
of education. Few informants marked the vocabulary as specifically belonging to
the highest levels of education. There are exceptions, though. For bédo and
bédave, the perceived educational level was quite high, with many informants
choosing all three categories of education levels. Moreover, informants did
choose all levels between 18% and 42% on each of the lexical items. This
means that the vocabulary is not viewed as exclusively the variety of poorly
educated people, contrary to the stereotypes and research (Lepoutre 1997).

Age is another matter that must be discussed. As I have mentioned,
scholars (Doran 2002, George 1986, Lepoutre 1997, Walter 1997) originally
overwhelmingly attributed the variety to youth. Boyer (2001) conducted research
at a university in which he would, by default, find students older than 16; and
Zerling ( 1999) underscores the fact that banlieue language is not limited to
youth. However, as | indicated earlier, no one has specifically studied age of
retention with empirical data.

What we find with Roma is that the majority of the lexical items were

attributed to an informant's own generation alone or in combination, including

80

the oldest group. That group, however, also frequently chose all generations.
Very few items were attributed to younger generations, regardless of the
informant’s age group. This is not what we would have expected if this were a
youth language, and it allows us to suggest that Zerling (1999) may be correct
that age does not matter. The notable exception is marave, where it was
ironically only the youngest age group that used it with any regularity, at 50% for
that lexical item, though the violent nature of the word may have been influential.
4.1.2 Creole

The Creole vocabulary was included since many immigrants in the Garges
area are from the Antilles. In addition, many students come from the ‘DOMTOMs
to complete their education in the Paris area. Having come from a French
département, French has been the language of instruction in their schools, and
they are French citizens. Thus, though people who speak Créole may come to
the same area as immigrants, they do not need Creole in order to communicate
with others in the area. They are already partially assimilated by virtue of their
nationality and academic system. The question remains: is there enough Créole
influence in the banlieue for banlieue language to reflect it?
4.1.2.1 Goumer

This lexical item, a verb, means ‘to fight,’ the same as savater and

marave. It does overtly follow French rules for conjugation and flexion.

81

4.1.2.1.1 Gender

Almost half the informants knew the lexical item, and 35 informants
responded to the question on gender. The majority, 21 informants (60%), chose
males as the primary users of this verb. One participant (2.9%) said that females
were the primary users, and the remaining 13 (37.1%) chose both males and
females as equal users of the word. In comparison to marave and savater,
goumer is the only word where ‘men’ is chosen at a higher rate than ‘both.’ This
reinforces the idea that a lexical item that connotes violence seems to be viewed
as one used more frequently by males than females or both sexes.

412.12 Education

Thirty-three participants responded to this question. Of those, 21 %
(seven) said goumer was used by people of all educational levels. Another 6%
(two respondents) chose ‘very educated.’ 24% (eight respondents) chose ‘not
very educated,’ while slightly more, 27% (nine) chose ‘somewhat educated.’ The
remaining 21% (seven) chose a combination of ‘somewhat educated’ and ‘not
very educated.’ Again, the connotations of violence associated with the word
could explain the higher number of people choosing lower levels of education.

4.1.2.1.3 Age

The youngest group had only four respondents for this word. Two, or 50%,
chose their own generation, one chose the latter in combination with an older
generation (25%), and the final informant chose an older generation (25%).

The second group had 15 informants. Of those, the largest group chose a

younger generation, with seven informants (46.8%). Four chose their own

82

generation (26.7%), and three people chose the latter in combination with a
younger generation (20%). Finally, one informant chose an older generation
(6.7%).

The third group had eight informants answer the question. Of those,
37.7%, or three informants, chose their generation and a younger generation in
combination. Two informants chose solely their own generation (25%). Two
more chose all generations (25%). Only one informant chose an older
generation (12.5%).

The oldest group had five responses. Of those, three, or 60%, chose their
generation in combination with a younger generation, and the remaining two
chose all generations (40%).

4.1.2.2 Sa ou fe

This is a phrase meaning ‘how are you?’ This particular phrase was very
rarely identified, since between 12 and 15 informants answered each of the
perception questions.

4.1.2.2.1 Gender

Out of 12 respondents, one chose ‘males’ as the primary users of the
word and the remaining 11 (91.6%) chose both sexes as equal users. Since it is
a greeting, the lexical item is very neutral, which may be why the informants
chose 'both sexes’ in one of the highest percentages of any lexical item.

4.1.2.2.2 Education

Out of 14 respondents, eight (57%) chose all three categories of

education. Five people (36%) chose ‘moderately/somewhat educated’ and one

83

person chose ‘very educated.’ No one chose a low level of education. I suggest

that the low number of informants familiar with the vocabulary in conjunction with

the fact that no one chose a low level of education may indicate that the word is

not well integrated into banlieue language and may be seen more as a foreign

token, since the other lexical items (even those such as bédo and bedave, which

were fairly well viewed) all had informants who chose a low level of education.
4.1.2.2.3 Age

When we look at age by groups, we see that the youngest group only has
three responses, and those are evenly split. One person each (33%) attributed
the lexical item to the following categories: his own generation, a younger
generation, and an older generation.

The second group had four respondents, and again, the answers were
quite varied. Two people (50%) chose their own generation, and one person
each (25%) chose an older and a younger generation.

The third group had three respondents as well, and here, one person
chose ‘all,’ one person chose his own in combination with a younger generation,
and one person chose an older generation, at 33.3% each.

The oldest group had the largest number of informants for this lexical item,
despite being one of the smallest groups in terms of number of informants. Five
people responded, and three of them (60%) chose all generations. One each

(20%) chose his own generation and an older generation.

84

4.1.2.3 Sa ka maché

This is a phrase, a greeting, which means ‘How are you?’ though it is less
formal than 33 cu fé.

4.1.2. 3.1 Gender

Twenty-seven informants responded to this question, and the
overwhelming majority (24 informants or 88.8%) chose both sexes using the
phrase equally. One person chose females using it most (3.7%) and two others
(7.4%) chose males as the primary users of the word.

4.1.2.3.2 Education

Only 14 informants responded to this question. Of those, eight (57.1%)
chose all levels. Five (35.7%) chose ‘somewhat educated,’ and the remaining
informant chose ‘very educated.’ Of striking importance, not a single informant
chose “not very educated.’

4.1.2.3.3 Age

Looking at the results by age group, we see that only four of the youngest
group answered this question. Of those, three (75%) reported that the lexical
item was used by an older generation, and one (25%) reported that it was used
by all generations.

The second group had six respondents, and again, the largest category, at
four informants, or 66.7%, was ‘an older generation.’ One informant attributed it
to his own generation (16.7%) and one (16.7%) combined his own generation

with that of an older generation.

85

The third group only had four informants. Of those, they were evenly split,
(50% versus. 50%) between ‘all’ and their own generation.

Finally, the oldest generation had the most responses: seven. Of those,
three (42.9%) chose all generations, another three (42.9%) chose an older
generation in combination with their own, and one person chose simply his own
generation (14.3%).

These responses make clear the fact that this word has an aging user
base. Most informants among the age groups say that it belongs to an older
generation, and very few people among the younger informants even know the
lexical item. An aging user base would be also be supported by the fact that all
“an older generation’ choices fall between age 18 and 23.
4.1.2.4 Makoumé

This is the only Creole noun on the list. It refers to a homosexual male. It
was relatively unknown by the informants. In fact, out of the eight people who
responded to most of the general questions about its use, five were from the
Antilles. It was also toward the end of the list, and the lack of response may well
indicate participant fatigue.

4.1.2.4.1 Gender

For this category, seven participants responded. Two participants (28.5%)
indicated that men were the primary users of the word, and the remaining five

(71.5%) indicated that it was used equally by both sexes.

86

4.1.2.4.2 Education

Out of the seven responses, participants were split almost evenly between
‘somewhat educated,’ at three informants (43%), and ‘all levels’ at four
informants (57%).

4.1.2.4.3 Age

What is most striking about this lexical item is that only eight people
responded to this question. In addition, those responses are very divided. The
youngest group had only two responses (50%) for ‘all’ and 50% for ‘my
generation.’ The second age group had two responses as well, divided 50/50
between ‘my generation’ and ‘an older generation.’ The third age group had
three responses, at 33.3% each: one for ‘all,’ one who attributed the lexical item
to a younger generation, and one to an older generation. Finally, the oldest
group had only two responses, and those were also divided 50/50 between ‘all’
and ‘an older generation.’
4.1.2.5 Creole analysis

In terms of gender for this category, it is important to treat the lexical items
differently than for other categories. For goumer, with its violent connotations,
and having seen the results for other words of that nature from different
languages, it is not surprising that the lexical item was viewed as being used
primarily by males. Makoumé is in the 70th percentile for both genders, which is
similar to the average among the lexical items in Roma. However, the informants
voted that both genders equally used sa ka maché and sa ou fé at 88.8% and

91.6% respectively. This high percentage of those who chose equal usage

87

among the sexes, along with the high estimation of educational level, seems to
suggest that these tokens are viewed more as Creole than as banlieue. Of
particular reinforcement of this theory is that no one chose ‘not very educated’
as an educational level of someone who would use these lexical items.

In terms of education, goumer is the only item that is treated similarly to
the vocabulary in Roma, meaning a spread that is in the 20th percentile for low.
For the remaining three items, no informant indicated a low level of educational
background for someone who would use the vocabulary. In fact, the informants
are split among choices of moderate levels of education and all levels of
education. This indicates a distinction for Creole vocabulary, though more
research is necessary with a larger amount of tokens.

It becomes apparent when investigating age that the youngest group knew
the least vocabulary from Creole. When they did know a lexical item, they
attributed it fairly often to other generations, equally to older and younger. I
would suggest here that the informants had heard the items enough to be
familiar with them, but the lexical items were rare enough that the informants
only knew they did not belong to their own generation. To support this, we see
that the second age group's answers also varied widely, though there were more
respondents who knew the vocabulary. The two older groups knew the
vocabulary the most and attributed it more to their own generation, an older
generation, or all generations. This may indicate that the lexical items in this

category have an aging user base.

88

It seems as though Creole words are not a large part of banlieue
language. I hypothesize that these phrases are mostly considered to be foreign,
and therefore are not as integrated into the banlieue language. Additionally, I
argue that the oldest group, because of the agreement with the DOMTOM and
state workers, may have had more contact with families that had just
immigrated. Since French is spoken side by side with Creole in the DOMTOM,
and educated speakers are bilingual, it would be easy for a speaker of Creole to
assimilate more quickly and therefore use Creole words less. Based on a limited
questionnaire, it seems as though the contribution Creole has made to the
banlieue is minimal and that those who admit to using most of the lexical items
in this category are either from the Antilles or of a particular group that had more
contact with those who spoke Creole, which was verified by reviewing the
demographic sheets of the informants.

4.1.3 Arabic

Whenever the question of banlieue language arises, Arabic and its
influence are tied to it. Most of the recent research concentrates on the
Maghrebi community in France (Doran 2002, Melliani 2000).
4.1.3.1 Kiffer

This verb, ‘to really like something,’ comes from the Arabic noun kif,
meaning ‘marijuana.’ It has been adopted into the variety and beyond as a verb
and it has overt French morphology: II I’a kiffé. Though now mainstream, it is still
included in several lists for banlieue language (Goudailler 1997, Séguin and

Teillard 1996) and I wanted to test how well the word has been retained.

89

4.1.3.1.1 Gender

Though it was placed toward the end of the list, 58 respondents answered
this question. Of those, the ovemrhelming majority, 54 (93%) chose both sexes
using the word equally. This may have to do with the mainstreaming of the word
in French colloquial language. Two participants each (3.4%) chose males and
females respectively.

4.1.3.1.2 Education

Kiffer had a very high number of informants that chose all three
categories: 20 respondents (40.8%). At the other end of the spectrum, only two
respondents (or 4%) chose ‘not very educated.’ Another 36.7%, or 18
informants, chose ‘somewhat educated.’ Four informants, 8.1% each, chose
‘very educated’ and a combination of ‘not very educated’ and ‘somewhat
educated.’ A single informant (2%) chose a combination of ‘somewhat educated’
and ‘very educated.’

4.1.3.1.3 Age

It is apparent that all generations ‘kiffent’ kiffer. When we look at the
youngest group, we see that five out of a total of nine respondents (55.6%) claim
that the word is part of their generation. An additional two informants (22.2%)
claim that it is their generation in combination with a younger generation. One
person each (11.1%) chose a younger generation and all generations.

As to the second age group, we see that 24 informants answered the
question. Of those, 16 of the informants (66.7%) said that it belonged to their

own generation, and an additional four (16.7%) claimed that it was the latter in

90

combination with a younger generation. Two informants (8.33%) said that it
belonged to all generations, and two informants claimed that it belonged to an
older generation (another 8.3%).

When we look at the third age group, we find that the trend continues.
Four of the informants (28.6%) chose all generations as the primary users of the
lexical item. Six of the informants stated that the word was used by their own
generation in combination with that of a younger generation (42.9%). Three
(21.4%) claimed that it belonged to their own generation alone, and one
informant (7.1%) claimed that it belonged to a younger generation.

The oldest group had seven respondents. Of those, two chose their own
generation (28.6%) and three chose their own in combination with a younger
generation (42.9%). Two more chose all generations (28.6%).
4.1.3.2 Choucrane

This lexical item is a phrase, meaning ‘thank you’ in Arabic. It was chosen
more as a test of whether ordinary words had made their way into banlieue
language.

4.1.3.2.1 Gender

The vast majority of the informants, 30 out of 31 (96.7%), reported equal
usage among the sexes. The remaining informant chose males as the primary
users of the phrase.

4.1.3.2.2 Education

Ten informants (37%), out of 27 responded that all levels of education

used Choucrane equally. Only one person (3.7%), identified the lexical item as

91

being used mostly by people of little education. Another eight people (29.6%)
chose ‘somewhat educated,’ and two informants (7.4%) chose a combination of
the latter two categories. 14.8%, or four informants, chose ‘very educated,’ and
one person chose a combination of ‘very educated’ and ‘not very educated.’

The education level chosen for this lexical item is relatively high, like that
of some of the Creole tokens. In addition, one informant answered ‘des Arabes’
(Arabs) to the question ‘this word is used primarily by_.’ It may be that this
word is viewed as a foreign token. It was also not included in any of the other
sources on banlieue language, but rather serves as a test to view the extent of
absorption of everyday vocabulary into the banlieue.

4.1.3.2.3 Age

When we break this down by age groups, we see that the youngest group
had only five informants. Of those, two chose their own generation (28.6%) three
chose their own in combination with a younger generation (42.9%), and two
others chose all generations, at 28.6%.

The second group had nine informants. Of those, the largest group chose
their own generation, at four informants (44.4%), three (33.3%) chose an older
generation, and two (22.2%) chose all groups.

The third group had eight informants. They were more spread out in their
opinions. The largest group, tour (50%) chose all generations. The rest of the
vote was then split, one informant each (12.5%) for the following categories: ‘my
generation,’ ‘my generation’ with an older generation, ‘my generation’ with a

younger generation, and an older generation.

92

The final group had four informants. Of those, 100% chose ‘all.’
All of this strongly indicates that though this lexical item may be used in the
banlieue, it is viewed as a foreign token.
4.1.3.3 Ouaiche
This word, also spelled wesh, is Arabic meaning ‘How are you doing?’
However, since I had personally always understood it as ‘Hi,’ I put that in as a
choice, along with ‘oui,’ as a distracter. After consulting a native from the area, it
was confirmed that it is sometimes viewed as a simple ‘hello,’ though the original
meaning is ‘how are you?” as it comes from Arabic. I will only be analyzing that
meaning as the expected meaning. The other definitions are not analyzed.
4.1.3.3.1 Gender
When we look at this for the expected meaning, it is divided between ‘both
sexes’ at 17 informants (81%) and ‘men alone’ at four informants (19%). We see
that no one chose females as the primary users of the lexical item. In fact,
excluding the women who spoke Arabic, every other woman had analyzed the
lexical item as ‘hi’ instead of ‘What’s up?/ How are you?’
4.1.3.3.2 Education
Out of 19 expected responses, six informants, or 31.6% chose ‘all levels
of education’ as those who would use the lexical item. Eight informants (42.1%),
the largest group in this category, chose 'somewhat educated' and five others

(26.3%) chose 'not very educated.’

93

4.1.3.3.3 Age

Thirty-two informants responded to this question. In the youngest age
group, there were two informants. Both of them (100%) chose their own
generation as those who used this word the most.

In the second age group, 12 respondents replied. Of those, 66.7%, or
eight informants, chose their own generation. Two informants selected a
younger generation (16.7%) and the remaining two (again at 16.7%) chose all
generations.

In the 24- to 29—year-old group, eight people responded to the question of
age. Their choice was split between only two categories, both of which
contained the younger generation. Five informants, or 63.5% of this group,
selected a younger generation. The remaining three chose the younger
generation in combination with their own.

The 30- to 36—year-old group had only two respondents, one of whom
(50%) chose his own group, and one of whom chose ‘all’ (50%). This was a rare
word in that those in the youngest group claimed the lexical item as their own
and those in the older groups agreed with that assessment.

4.1.3.4 Arabic analysis

First of all, it must be underscored that there are not enough tokens in this
category for a full analysis. As this dissertation was an initial foray into
suggesting a need for further study into these areas, I could not fully study each
contact language. Moreover, because of the varied nature of the vocabulary, I

believe the discussion should reflect that reality, as it did for Creole.

94

One generalization can be made is in terms of gender. All items in this
linguistic category were viewed, at a very high percentage, as belonging to both
sexes equally. Ouaiche was the least equal at 81%, the other two being in the
90th percentile.

Education levels also vary widely, but for choucrane they are very high
and include multiple informants choosing ‘very educated’ by itself, which is rare
among the lexical items as a whole. Only one informant relegated the word to
those with low levels of education (3.7%), compared to two informants for kifl‘er
(4%), and a much larger percentage for ouaiche (five informants, or 26.3%).

Age, too, presents varied results with these three tokens. Kiffer is viewed
overwhelmingly by each age group as belonging to the informants' own
generation. With ouaiche, the tendency was toward youth, which differs with
many of the lexical items in this study and would fit the pattern of previous
research done by Doran (2002) and Lepoutre (1997), among others. Choucrane
responses are diverse, especially in the youngest three age groups, though the
oldest generation chose ’all’ generations at 100%.

I would argue therefore that while much more research is needed with a
larger lexical base from Arabic, of the tokens represented, only choucrane
seems to be viewed as a true foreign token. Interestingly enough, many of the
women who spoke Arabic did not admit to using choucrane. This may be a way
to distance themselves from a native language in order to appear more
assimilated, and not appear to be code-switching or crutching. More research

would be needed on this category for deeper analysis.

95

In general, the vocabulary in this category was viewed as defying previous
research on the subject, with regards to perceived age, gender, and education
level of speakers, as established in the discussion on Roma and Creole
vocabulary. Results do not show an overwhelming choice of males or younger
generations as being the perceived speakers of this variety. Nor is it attributed to
those of low formal education. More complete analysis could be conducted with
a larger vocabulary base for this language category.

4.1.4 Vedan

Though not an actual language, Verlan has become such an integral part
of banlieue language that Doran (2002) refers to the totality of banlieue
language as Verlan, citing her informants; and as previously discussed, she is
not alone. As I stated, my informants were not willing to call the ensemble
‘Verlan,’ but they did admit that speakers of Verlan will use banlieue language
and vice versa. So, while the link between Verlan and banlieue language is clear
and well documented, it still merits its own lexical category.
4.1.4.1 Keum

Keum is the first Verlan word on the list. It means ‘guy,’ and is derived
from mec. It was chosen because it was one of the older and most commonly

cited examples (Calvet 1996, Doran 2002, Mela 1991, etc).

96

4.1.4.1.1 Gender

This was the only lexical item on the list for which the informants
designated more primary users as female than as male. Seven informants
(13.2%) reported females as the primary users, whereas only three (5.66%)
informants chose males. The remainder, 43 (81%), chose both sexes as using
keum equally. Given this information, it would be possible to say that women are
indeed viewed as using this lexical item. Again, this word is neutral, with no
negative connotations, and is well established in the vocabulary; so perhaps that
is affecting the high percentage of perceived female usage.

4.1.4.1.2 Education

Ten people, or 22.2% of the 45 informants who answered this question,
made a point of circling all three categories. An additional informant circled both
‘somewhat educated’ and ‘very educated’ (2.2%). Only six informants chose ‘not
very educated’ by itself (13.3%). The single largest category was ‘somewhat
educated,’ with 16 informants (35.5%). The second largest category was the
combination of ‘not very educated’ and ‘somewhat educated,’ with 12
respondents (26.6%).

4.1.4.1.3 Age

The youngest group had eight respondents. Of those, three (37.3%)
reported that it was a lexical item of their own generation. Two (25%) said that it
was for a younger generation. One informant (12.5%) combined his own

generation with that of an older generation, and one informant (12.5%)

97

combined his own generation with a younger generation. One final informant
attributed the lexical item to an older generation (12.5%).

The second group had 18 informants. Of those, 10 (55.6%) stated that it
was from their own generation. Four combined the later with a younger
generation (22.2%), and two each, at 11.1%, claimed that it was from an older
generation and a younger generation.

The third group had 14 informants. Of those, seven (50%) claimed that it
was their generation with a younger generation. Four (28.6%) claimed it was
from their own generation alone. Two chose all groups (14.3%), and one person
chose his own generation in combination with an older generation at 7.1%.

The oldest group had eight informants. Of those, five, or 62.5%, chose
their own generation in combination with a younger generation. One (12.5%)
chose his own generation, one (12.5%) chose his own generation in combination
with an older generation, and one (12.5%) chose all generations,

Keum is one of the more streamlined of the lexical items. The combination
of the answer ‘my generation’ with ‘a younger generation’ is claimed from age 19
to age 58, when we bring in the outliers as a control group. Our informants
report that it is used virtually equally among the sexes. As to education, very few
people chose ‘not very educated.’

4.1.4.2 Sky
This lexical item is from ‘whisky,’ verlanized to skywhi with apheresis. and

it means ‘whisky.’

98

4.1.4.2.1 Gender
From 49 responses, 61.2%, or 30 informants chose both sexes as using it
equally. The next largest category chose men with 18 informants, or 36,7%.

Women only garnered one vote, or 2%.

4.1.4.2.2 Education

Sky had one of the greatest numbers of informants circling all three
categories of education levels: 19 out of 44, (43.2%). Three additional people
(6.8%) chose ‘very educated.’ At the other end of the spectrum, only one person
chose ‘not very educated,’ (2.27%), and only two people (4.5%) chose ‘not very
educated’ in combination with ‘somewhat educated.’ The remaining 19
informants (43.2%) chose ‘somewhat educated.’

4.1.4.2.3 Age

The youngest group had seven respondents with this lexical item, and all
of them chose their own generation.

The second group was more complicated, with 22 informants. Of those,
the largest number, 11 (50%), chose their own generation, two more (9.1%)
chose their own in combination with a younger generation, and two informants
chose their own generation in combination with an older generation (at 9.1%).
Two informants chose a younger generation by itself, again at 9.1%, and five
informants chose an older generation.

The third group had 13 respondents. Five of the informants chose ‘all’

(38.5%). Four informants, or 30.1%, chose their own generation in combination

99

with a younger generation. Three, or 23.1%, chose simply their own generation,
and one person, or 14.3%, chose his own in combination with an older
generation.
The oldest group had seven respondents. The largest group for them was
’all,’ which had three informants, or 42.9% of the vote. One person chose his
own generation(4.3%), and two chose their own in combination with a younger
generation (28.6%). Finally, one person chose a younger generation by itself
(14.3%).
4.1.4.3 Vedan analysis
First it is important to state that Verlan, like Arabic, does not have enough

lexical items in this study to make any sweeping generalizations, only
suggestions for further research. However, if we look at the responses to the
questions, we do still find patterns similar to those in Roma, Creole, and Arabic.
Usage between the genders is predominantly viewed as shared, which follows
most of the vocabulary we have seen thus far, and which is unusual in a non-
prestige variety. If we separate the responses for those who chose both genders
and concentrate only on usage attributed to men alone as the primary users or
to women alone as the primary users, keum breaks the typical gender traditions
for non-prestige dialects. When comparing the infonnants' views between those
who chose men and those who chose women, women were selected as the
primary users of the lexical item at a higher percentage than men. It may be

because the word does not have any violent or illicit connotations and is

100

considered mainstream vocabulary. I observed it being used on an RER subway
billboard when l was there for my fieldwork.

The education level of the lexical items in Verlan was well above
expected. For both tokens, those who chose the lowest level of education were
few, 13% for keum and only 2.2% for sky. In addition, those who chose all levels
of education were high as well, especially for sky, at 43%. An additional 7% said
sky was used primarily by those with high levels of education. This is completely
contrary to what we would expect in a non-prestige variety, as illustrated by the
Roma discussion.

For keum and sky, most informants in all age groups included their own
generation and/ or a younger generation. This seems to support the idea that
these two items are very well integrated into banlieue vocabulary. Through the
perceptions of the education level of the user, as well as the gender, we see that
there seems to be little stigma attached to the use of these items.

4.1.5 Angot
This group contains a myriad of lexical items that belong to banlieue language
as well as popular French. Some of the lexical entries in this category, such as
damn and its gender pair, daronne, date from the early 19th century.
4.1.5.1 Binouse

This lexical item is a noun that means ‘beer.’

101

4.1.5.1.1 C
Thirty-lhre
hose males as
than users a
41.5.1.2
Only 27
12 (44%) chos
educated by ‘
the inlormar
One Petson (
educated.’
4.1.5. ‘.

The in

 

daimed the

claimed th

 

4.1.5.1.1 Gender

Thirty-three informants answered this question. Of those, 20 (60.6%)
chose males as the primary users. One informant (3%) chose females as the
primary users and the remaining 12 (36.3%) chose both sexes.

4.1.5.1.2 Education

Only 27 informants responded to the question of education, and of those,
12 (44%) chose all three categories of education, though no one chose ‘very
educated’ by itself. The second largest category was ‘somewhat educated’ with
nine informants (33.3%), and five informants (18.5%) chose ‘not very educated.’
One person (3.7%) chose a combination of ‘not very educated’ and ‘somewhat
educated.’

4.1.5.1.3 Age

The first age group had only four respondents. Of those, three (75%)
claimed that the word was from an older generation. Only one person (25%)
claimed that the word was from his own generation.

The second age group had 12 informants. Of those, 50%, or six
informants, chose their own generation. Five informants chose an older
generation (41.7%). Only one informant (8.3%) chose ‘all.’

In the third age group, 10 informants replied to this question. Of those,
four chose their generation in combination with an older generation (40%). Two
chose their generation in combination with a younger generation (20%). Two
(20%) chose only their generation, and one informant each (10%) chose an

older generation and ‘all.’

102

Finally, in the oldest group, from six responses, we see that one person
chose his own generation, at 16.7%. Two chose their own generation in
combination with a younger generation, at 33%. One chose an older generation
(16.7%), and two chose ‘all’ (33%).

This word is very likely to have an aging user base since no one in any of
the categories chose a younger generation. In fact, in all of the age groups,
there was a choice of an older generation, which is rare. In addition, the
youngest group does not seem to recognize it.
4.1.5.2 Damn

This lexical item is a noun, and it is another term for ‘father,’ though it
would be closer in register to ‘pop’ or ‘the 'rents’ for parents. It is never used as a
form of address, but rather with a possessive adjective or a definite article, and
the term dates back to the 1900s (Delvau 1883).

4.1.5.2.1 Gender

Out of 5 total responses, 52 informants (94.5%) chose males and females
using daron equally. Only one informant (1.8%) chose females as primary users,
and the remaining two (3.6%) chose males.

4.1.5.2.2 Education

Eighteen informants (37.5%) chose all three categories of education. An
additional four (8.33%) chose ‘very educated,’ and two informants (4.16%)
chose a combination of ‘very educated’ and ‘somewhat educated.’ At the other

end, only three people (6.2%) chose ‘not very educated.’ The second largest

103

group was ‘somewhat educated’ with 33.3% of the vote (16 informants), and five
people (10.4%) chose a combination of the latter two categories.

4.1.5.2.3 Age

The youngest group had nine respondents for this word. Of those, the
overwhelming choice was of their own generation, at seven informants or 77.8%.
The two remaining informants were split (11.1% each) between a combination of
the latter and a younger generation and ‘all.’

The second age group had 24 respondents. Almost half of them, 11
(45.8%), claimed that it belonged to their own generation. Another six, or 25%,
said that it belonged to their own generation in combination with a younger
generation. Two others (8.3%) claimed the opposite, saying that it was their
generation and an older generation that shared the word. Two informants (8.3%)
chose ‘all.’ Three informants, or 12.5%, claimed that it belonged to a younger
generation by itself.

The third age group had 13 informants. Of those, three informants (23.1%)
said that the word was from their own generation, and five others chose their
own generation in combination with a younger generation (at 38.5%). Four
respondents chose ‘all’ (30.8%). The remaining informant chose his own
generation in combination with an older generation.

The last group had eight informants respond to this question. Of those,
only one person, or 12.5%, chose his own generation alone, but four chose the

latter with a younger generation (50%). The remaining three (37.5%) chose ‘all.’

104

We see that this word is pretty well integrated into the banlieue language.
Of 54 responses, only three informants, a mere 5.5%, chose a younger
generation. The rest chose their own generation by itself or in combination with
other generations.
4.1.5.3 Damnne

This lexical item, a noun, means ‘mother.’ The same linguistic constraints
apply to damnne as to its pair, damn. It has also existed from the late 19th
century (Delvau 1883).

4.1.5.3.1 Gender

Again, both sexes were seen at the primary users of the word, with 48 out
of 56 respondents (85.7%). Only one informant, (1.78%) chose females as the
primary users, identical to that of its pair. However, here seven informants
(12.5%) chose males alone as the primary users.

4.1.5.3.2 Education

Out of 46 respondents, 17 informants (37%) chose all three levels of
education and one (1.78%) informant chose both ‘not very educated’ and ‘very
educated.’ No one, however, chose ‘very educated’ alone, in contrast with four
people who made that choice with damn. The largest category was ‘somewhat
educated’ with 18 informants, or 39%. Five informants chose ‘not very educated’
(10.86%), and four informants chose a combination of the latter two categories

(8.7%).

105

4.1.5.3.3 Age

The youngest age group had 10 responses. The largest category was ‘my
own generation’ at seven informants (70%). One informant (10%) chose his own
in combination with a younger generation, and another (10%) chose simply a
younger generation. A final informant (10%) in this group chose ‘all.’

The second group had 22 informants. Most of these, 13 informants
(59.1%), chose their own generation. Another three (13.6%) chose their own in
combination with a younger generation. Three (13.6%) chose a younger
generation by itself. Two informants, or 9.1%, chose all generations, and one
person chose his own generation in combination with an older generation
(13.6%).

The third group had 12 respondents. Of those, the largest group, six or
50% of the informants, chose a younger generation in combination with their
own. Three chose their own generation by itself (25%), and two people chose
‘all’ (16.7%). A final informant chose an older generation (8.3%).

Seven informants in the last group responded. Of those, four chose their
generation in combination with a younger generation (57.1%). One informant
each (14.3%) chose the following categories: ‘all’, ‘my generation,’ and ‘my
generation and an older generation.’

The reader may notice a disparity between the results on damn and
damnne. In theory, one would expect them to be identical since they belong to
the same language, have the same part of speech, and are simply gender pairs.

However, the participants are more reticent with damnne. It is marked with more

106

males as being the primary users and a lower education level overall. To
interpret this, it is necessary to look at the culture of the banlieue. Women in
general may not be very respected in the banlieue, but there is an element of
respect that resurfaces when it’s a question of motherhood, sometimes taken to
extremes.

The theory that talking about one’s mother as “ma damnne’ is less
respectful than talking about one’s father as ‘mon damn’ was confirmed by a
native speaker in the area, who said, ‘pour damnne effectivement cela s'emploie
moins facilement que damn. C'est pas forcement plus vulgaire mais une maman
echappe plus aux mots de la banlieue.’

‘For damnne, indeed, it’s not used as easily as damn. It’s not so much that it’s

vulgar, but a mother escapes banlieue words more easily.’

4.1.5.4 Savater

Like marave and goumer, this lexical item is a verb meaning to fight/ beat
up. It is simple slang and reflects normal French overt flexion.

4.1.5.4.1 Gender

Forty-nine people answered this question. Of those, 29, or 59.2%, say that
both sexes use it equally. 40.8%, or 20 informants, report that men are the
primary users. This lexical entry seems to follow the masculine-based pattern of

the other words and phrases that are used for fighting.

107

4.1.5.4.2 Education

Of 47 respondents, 12 (25.5%) claim that people from all levels of
education are the primary users of this lexical item, though ironically, no one
chose “very educated’ alone or in any combination other than all three
categories. Instead, 10 people (21.3%) chose ‘not very educated,’ 11 (23.4%)
chose ‘somewhat educated,’ and the remaining 10 informants (21.3%) chose a
combination of the two. This is one of the lower perceived levels of education
among the lexical items.

4.1.5.4.3 Age

When we break the results down by age group, we find that the youngest
group had seven informants. Of those informants, one chose his own generation
(14.3%). Two chose the latter in combination with a younger generation (28.6%).
One informant chose simply a younger generation (14.3%), and one other chose
the latter in combination with his own generation (14.3%). Two informants chose
an older generation (28,6%).

The second group had 20 informants. The largest category was ‘my
generation’ at 40%, or eight responses. Another five responses combined their
own generation with a younger generation (25%). Three informants chose a
younger generation by itself (15%), and another three informants (15%) chose
an older generation. The remaining informant (5%) chose the latter in
combination with his own generation.

The third age group had 12 respondents. Five of those chose their

generation in combination with a younger generation (41.7%). Four chose the

108

opposite, their own generation with an older generation (33.3%). One person
relegated the lexical item to a younger generation by itself (8.3%), and two
informants chose all generations (16.7%).

The oldest age group had eight informants. Three of those informants
(37.5%) chose all generations, and another three (37.5%) chose their own
generation in combination with a younger generation. Only one person each
(12.5%) chose an older generation and their own generation alone.

It is interesting to note how comparatively few people chose their own
generation alone, though about 78% of the informants identify the item in some
way as being part of their generation.
4.1.5.5 Argot analysis

Damn and damnne, gender pairs, are in the 80th and 90th percentile as
being used by both genders, 5% and 15% higher than the average for this
category, respectively. Binouse and savater, however, are viewed as exclusively
male at at 60% and 40% respectively. In both cases, we can say that the same
analysis in terms of gender applies to this category as to Roma and the other
languages. In what should be a variety spoken primarily by males, we see that is
not the case in general.

In terms of why the statistics in this category among the lexical items may
differ, it seems to be established that the meaning of the lexical item has an
effect on whether nor not a lexical item is viewed as being used by males or
females. So far, most items pertaining to violence are viewed as male. I would

like to propose, based on my observations during fieldwork and residence in

109

France, that more women tended to drink wine (kir) and the men tended to drink
beer (binouse) or whisky (sky). Both of those were viewed as used more by
males. This is by no means definitive, but may have an effect. The results of
savater are very similar to those of marave, which also has the same definition
of violence.

The results for education show the same split. Binouse and savater were
attributed to those of low education by 18 to 21%, which seems to be a default
result among the lexical items.8 Damn was at 6% and damnne at almost 11%.
Again I remind the reader that even the lexical item most attributed to a low level
of education is only 21 %, which would seem to counter claims that banlieue
language is used by those who are uneducated, such as with the ideas set out
by Lepoutre (1997) that I discussed in the analysis of Roma, and which bears
repeating.

For savater, damn and damnne, most informants, regardless of age
group, attributed the items to their own generation by itself or in combination with
a younger generation. Binouse was viewed as an older token, with no choice of
‘a younger generation’ by itself, and only the oldest group used it in any
combination. The fact that these lexical items were well known, much more so
than with Creole, for example, may well bear out Tejedor de Felipe’s (2004)

theory that banlieue language is indeed a French-based phenomenon. More

 

8 I see this as a positive in a non-standard variety that only around 20% say the
lexical items are used by those with little formal education. This means that
around 80% (depending on the word, the language, etc.) of the informants are
saying that the words are used by those with mid or high levels of education, a
fact that defies previous research.

110

research should be conducted on the lexical category of argot in the future to
confirm this hypothesis.
4.1.6 Lexical items

When the linguistic items are separated by language, we see that in each
linguistic category, some lexical items do not conform to any pattern, but general
tendencies can indeed be observed. First of all, the lexical items in Verlan and
argot were very well known in general. They had not only the highest
percentages of people who were familiar with the items but also the highest
percentage of admitted usage. These are both categories that have been noted
as contributors to banlieue language, as mentioned by many scholars (T ejedor
de Felipe 2004, Goudailler 1997, Séguin and Teillard 1996), which suggests
that Tejedor de Felipe (2004) may be correct in his idea that this is indeed a
French phenomenon.

Arabic is represented quite well in banlieue language in the present study.
The speakers know the lexical items and admit to using them. However, it would
be beneficial to continue the study with more tokens from this particular lexical
category. Though Arabic was mentioned by other researchers, Roma also is
prolific in this study. We see this is not the case, however, with Creole, which
seems to be fading out of favor in banlieue language, perhaps because of the
rapid and complete assimilation of its bilingual speakers into French society.

In all language categories we notice that the previous literature on gender
is somewhat indicative of the reality of banlieue language and somewhat not.

Most lexical items are attributed to both sexes equally as the primary users of

111

the words, which challenges the previous research on non-prestige varieties, in
which men would usually be the primary speakers of the variety. Milroy (1980)
attested that women used more of the non-prestige variety when their own
community's non-prestige dialect had a prestige in its own area. I believe that
with the importance of identity as a factor in banlieue language, as attested by
Doran (2002), Melliani (2000), and others, that identity may supersede gender
vanafion.

When we investigate only men versus women and the perceived usage
between them, women alone are rejected almost categorically as being the
primary speakers of the variety, which follows previous research on gender
studies (Lavob 1972, Trudgill 1972). Males alone are regarded as using the
vocabulary, generally in the area of 40% for most lexical items, and in higher
percentages where the definition of the lexical item involved violence.

In general, the informants evaluated the perceived education level of
people who would use the lexical items as being somewhere in the middle. They
attribute these lexical items neither to those with very little education nor to those
with a very high level of formal education. The notable exception to this rule was
the category of tokens that seem to have been viewed as foreign to the
informants, such as choucrane, where the educational level was perceived as
very high. The definition of the lexical item seems to have some bearing on the
perceived educational level of someone who would use the vocabulary as well,

since all three words meaning ‘to beat someone up’ were attributed to a lower

112

level by the informants than the other lexical items in the study, which seems to
suggest that a thematic investigation is necessary.

Finally, if we take into account the age of the informants, we find that there
was a division between the youngest groups and the oldest groups.
Comparatively across the age groups, all groups chose their own generation
more than any other choice, whether by itself or in combination with a younger
or older generation. If this were a youth language, we would expect to see more
choices of a younger generation since the informants were from 18 to 36 years
of age. This would support Zerling’s (1999) assertion that banlieue language is
not limited to the young.

We do see that the youngest group attributed the most lexical items to a
younger generation than theirs, effectively distancing themselves from those
who would use the vocabulary. This is perhaps due to the younger informants
being from either the Mission Locale where they were actively seeking
employment, and thus did not want to admit to vocabulary seen as non-
professional, or at the university where perhaps they would be rejecting the
vocabulary as from another social setting (Doran 2002). It may also be simply
that the vocabulary had been heard, but since it was not from their group, they
attributed it to those in a younger generation. Especially with the Creole
vocabulary in the study, the younger groups did not attribute the lexical items to
their own age groups, but rather the other groups in general, with no consistency
among the peer group. It was as if they had heard the lexical items, but rarely

enough to not be able to pinpoint the actual age group who might use the words.

113

Usage of Verlan (the syllable reversal language game) was first studied in
the late 1980s, which may also help understand the category of age. The oldest
group chose their own generation or all generations for the majority of the lexical
items and knew the most vocabulary. That oldest group would have been in high
school during the late 1980s. Newer generations could be changing the
vocabulary, as one student in the Séguin and Teillard (1996) study indicated.
Some indication that this may be the case is given by the three Maghrebi males
at the Mission Locale who stated that they did not use bédo and that it was from
an older generation. We will see in 4.2 that these results are also supported

when we look at the age group of informants who know a particular lexical item.

4.2 Global Results

After analyzing the results for each lexical item by itself, some general
analysis is necessary. I examine the main factors, such as gender, age, and
education in two separate manners. The first section in each subsection
discussed the infonnants' collective perceptions about who uses the lexical
items. However, in contrast to the section 4.1, we will not be looking at each
individual lexical item, but rather the lexical items as a whole. I will then talk
about each of the factors in terms of admitted usage, i.e. which gender, age
group and education level (by informants) admits to the most usage of the lexical
items as a whole. Finally, I compare the results for any discrepancies, i.e. do
women say that both sexes use the lexical items equally and do they use the

lexical items they say other women use, or is there a difference in their

114

perceptions and what they admit to using? Do members of one age group
attribute usage to their own group and not admit using that same vocabulary?
Two categories are only investigated in terms of admitted usage since they have
no counter part. Those two factors are venue and language.

It should be noted that in this study women make up almost 60% of the
participants. For this reason, in order to ‘norrnalize’ the data, results are
presented first in an overall format, and then percentages are calculated
separately for each gender. All mathematical operations were performed using
Systat.

Overall there were 552 unexpected answers and non-answers out of 1407
possible answers. An unexpected answer is defined as any answer other than
the definition provided by the original informant who helped create the list.
Sometimes the unexpected response may be the result of a language change in
progress, and the definition of the word in question may have evolved.
Sometimes it may simply be a guess on the part of the informant. Many times it
was also due to fatigue on the part of the informant since the questionnaire was
quite long. In any case, the total number of the combination of non-answers and
unexpected answers represents only 24.16% of the total 1407 possible.

Percentages from this point on are compiled excluding unexpected and
non-answers, though these numbers are discussed at the beginning of every
section. It should be noted at this point that each category contains a high level
of non-answers or unexpected answers and that many of these non-answers or

unexpected answers resulted from not knowing the lexical item; or when the

115

informant knew the lexical items, they chose the expected definition but did not
elaborate on the questions that followed due to fatigue.

It should also be noted that if a participant chose more than one answer,
the responses were not collapsed, but the selection of multiple responses
created a separate category. Thus because of the choices of both ‘my
generation’ and ‘a younger generation,’ a category was created for that dual

choice.

4.2.1 Gender

l have already discussed what happens with language in terms of covert
and overt prestige. I now examine in detail the perceptions of both sexes about
the gender of someone who would use the vocabulary. I follow this by
discussing the answers the informants themselves gave and comparing the two
sets of data.
4.2.1.1 Perceived gender usage by gender of informant

Examining the non-answers in general terms by gender, we see that
women, as might be expected in a non-prestige dialect, know fewer words than
men (Eckert 1999, Labov 1972, Trudgill, 1972). However, when tallying the non-
answers, (as mentioned above, 552/1407), we find that that difference is
marginal. Women are just slightly less likely to know the lexical items, at 40% of
no/unexpected answers (563) versus 37% (521) for men. It may also be noted at
this time that the word ouaiche was counted in these percentages because all

three possible answers were correct in some way, as discussed above.

116

The table below shows the number of tokens for each category, along with
the corresponding percentage. Therefore, counting the responses of all
informants regarding the gender of the perceived users of the lexical items, 208
tokens were recognized as being used by ‘males,’ at 24.33%. Only 36 tokens
were attributed to females, which corresponds to 3.04% of the tokens. 72.63%,
or 621 of the 854 tokens were attributed to both men and women equally.

Table 1: Perceived gender usage by gender of informant

 

 

% N
ALL INFORMANTS
Male 24.33 208
Female 3.04 26
Both 72.63 621
Total 100 854
FEMALE
INFORMANTS
Male 22.6 1 13
Female 2.4 12
Both 75 375
Total 100 500
MALE INFORMANTS
Male 26.76 95
Female 3.94 14
Both 69.38 246
Total 100 355

 

The expected result is that women would not be the primary users of the
lexical items in question. This is borne out quite nicely in the percentages for
each gender separately, with men attributing more of the vocabulary to females
than females do, but neither sex attributing more than 4% to females. This
means that the vocabulary is seen mainly as either a shared vocabulary or a
masculine vocabulary. Most importantly, it is not perceived as a feminine

vocabulary, which is what we would expect in a non-prestige variety.

117

Banlieue language is not a variety of overt prestige, but rather one of
covert prestige. It should be, then, according to general previous sociolinguistic
research on gender and linguistics, a variety where change is led by men rather
than women (Labov 1972, Trudgill 1972). Furthermore, since men tend to under-
report their use of the standard variety and women tend to over-report their use
the same (Labov 1972, Trudgill 1972), we might expect that men would want to
claim banlieue language, a non-prestige variety, as a masculine variety. We did
see that males were slightly more likely than women to claim that the lexical
items were indeed used mainly by men. However, the difference is minimal,
22.6% for females and 26.76% for males, but even taking the men's perception.

What is intriguing is that when we separate responses by the gender of
the informant, we find that both males and females attribute usage of the lexical
items to both sexes by a large degree. Women are especially prone to do so, at
almost 75% versus almost 70% for men. It may be that usage between the
sexes is less common than usage among the sexes separately. Men may be
less likely to attribute the lexical items to both sexes since it may be viewed as a
linguistic blending of gender roles.

I would like to propose that Doran’s (2002) ‘We’re not bourge’ theory may
be applicable to explain why over 70% of both sexes attributed the words and
phrases to both sexes. It is very possible that this social distinction between
classes, the mentality of les bourges versus les banlieuesards, the identity factor
underlined by so many scholars (Calvet 1994b, Doran 2002, Goudailler 1997,

Melliani 2000), is more important than gender distinctions.

118

4.2.1.2 Admitted usage by gender of informant

Now that we have seen the perceptions of who uses these words, broken
down by gender, we should look at the respondents’ own admitted usage. In
doing so, I omit the percentage of people who either did not answer or gave an

unexpected answer. Table 2 below presents these results.

Table 2: Admitted usage by gender of informant

 

 

% N
ALL INFORMANTS
Often 28.8 252
Rarely 36 315
Never 34.97 306
Oftenand rarely .23 2
Total 1 00 875
FEMALE
INFORMANTS
Often 22.61 116
Rarely 33.33 171
Never 43.66 224
Oftenand rarely .39 2
Total 1 00 51 3
MALE INFORMANTS
Often 37.57 136
Rarely 39.78 144
Never 22.65 82
Oftenand rarely 0 0
Total 100 362

 

Results including all of the informants show that almost 35% of the
informants say that they never use the lexical items. This means that 65% of the
informants admit to using the lexical items in question to some extent. In fact,
28.8% of the participants say they use the lexical items in question often, and an

additional 36% admit to using them rarely.

119

By gender of the informant, we see that the distribution of male versus
female self-reported usage seems to be what we would expect from a non-
prestige dialect, since the women do not admit to using it as much as the men
do. In fact, 43.66% of the female informants say they never use the lexical items
in question, while only 22.61% of the males said they never used the lexical
items in question.

It is the direct comparison of admitted usage between the sexes that
becomes important. When we examine admitted usage, starting with those who
claim to use the words often, we see something we would expect to see in terms
of non-prestige dialects. Females admit to using the vocabulary often at 22.6%,
while men admit to 37.57%. For those who claim to use the vocabulary rarely,
women come in at 33.33% and men at 39.78%. If we create a total admitted
usage, women say that to some extent, they use 55.94% of these words. Men
show and even higher average at 77.35%. This is partially in accordance to what
we would expect with a non-prestige variety. Men admit to speaking it more than
women do. However, as l have mentioned before, there is a high percentage of
women who admit to using the vocabulary , which leads me to believe that
perhaps identity with the banlieue is stronger than gender.
4.2.1.3 Comparative analysis

In the graph below, for perceived usage, the numbers were calculated by
separating what was attributed to the opposite gender alone from anything that
was attributed to the informant's gender. For example, for men, perceived usage

includes both what was attributed solely to men and what was attributed to both

120

genders, but excludes what was attributed to females only. For women the
opposite is true, and therefore numbers were compiled by adding those
attributed solely to females along with those attributed to both sexes. Admitted
usage includes all admitted usage, and therefore both ‘often’ and ‘rare’ usage
was combined. Only ’never’ was excluded. As a reminder, these figures are only

for expected answers.

 

perceived vs. admitted usage by gender

120

 

 

 

I males
‘ females

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

perceived usage admitted usage

 

 

 

Figure 1: Perceived vs. admitted usage by gender

Comparing perceived usage to admitted usage, women claimed that men
use the items exclusively at only 22.6% (113 tokens). That leaves 77.4% of the
items as being perceived either strictly women's lexicon or shared. However,

adding up all of their admitted usage (both responses of ‘rarely’ and ‘often’), that

121

number is clearly lower (56.3%). It could be that the women were under-
reporting their usage of these terms, which would be explainable by covert and
overt prestige. In addition, the women as a group were quite well educated. In
fact, there were only four women with a BEP or equivalent, versus 12 men of
that educational level.

What is unexpected is that the men admitted to using the lexical items less
than they perceived their gender as using the same vocabulary. In theory, men
should be underreporting their use of the standard and over-reporting their use
of a non-prestige variety. The perception among the male informants is that
96.1% of the tokens are viewed as strictly male or shared between their own
gender and females. They admit to only 77.4% usage, however.

Investigating this phenomenon, we find that men in the older groups were
most likely to state that the vocabulary was from their own generation, and they
were also most likely to admit to using the vocabulary. These men were
employed and relatively financially stable. It is possible that this group felt more
comfortable telling me that they used the vocabulary since I had already
witnessed this when we had socialized in the past, since they were at a private
gathering. They knew me and trusted me, and therefore would not expect such
an admission to adversely affect them in any way. As for the younger men at the
Mission Locale, while I was able to assure them of their anonymity, it is possible
that they were hesitant to admit to using banlieue language in the fear that
somehow the results might be released to employers or that said vocabulary

might somehow affect their job employment prospects. At the university, some

122

students may not have wanted to admit to the vocabulary since they were not in
an environment in which that vocabulary would be appreciated. Even the
younger participants in Doran’s (2002) study avoided use of Verlan in an
academic setting. These university students may simply not have wanted to
admit usage to a stranger. Additionally, it may be that for some students,
banlieue language was not an identity they embraced. I did not investigate the
cite and its staircases of unemployed youth, but rather places where people
were either employed or actively seeking employment or the skills to attain it. My
goal was not to confirm the stereotype, but rather to question the stereotype by
including rarely studied groups.

Finally, this empirical data allows us to say that these lexical items are
indeed used, regardless of gender. There may be difference among the
genders, both in perceived usage and admitted usage, but both genders
admitted to using the vocabulary. The women in this study admitted to using
fewer banlieue language tokens than men, which is what we would expect in a
non-prestige variety. Though, again, we see that women did admit to usage
rates of over 50%, which may indicate that identity with the banlieue, and
solidarity, may be a more important factor than gender variation, a phenomenon
witnessed previously in Milroy's (1980) study.

4. 2.2 Education

46% of the respondents did not answer this question. The question of the

education of the user was toward the end for each lexical item, and fatigue

seemed to play a part. There were also several qualitative remarks stating that

123

education was not a factor, so it may be the case that people did not want to
answer that particular question because they did not feel that the question of
education was valid, especially given the stereotypes of the banlieue. One of the
informants told me directly that education level was irrelevant and questioned
the legitimacy of the question. One simply placed an exclamation mark next to
each of the questions concerning education level. Another wrote that the
education level of the informant, in her opinion, was irrelevant to whether or not
that person would use the vocabulary, but rather usage of banlieue language
depended on where the person was raised.

It is important to remember that we are discussing perceptions in this
section. I avoided actual categories of education such as ‘someone who has the
baccalaureal’ or ‘someone who has a DEA,’ because I felt that I would elicit true
language attitudes better by asking questions aimed at more general categories
such as ‘not very educated’, ‘somewhat educated’ and ‘very educated.’ Actual
degree levels were also rejected due to the changing nature of educational level
need in society. Professions that several decades ago required only a high
school degree may need higher levels now, and in fact, as we have mentioned,
the DEUG is no longer a current degree program. Attitudes toward this particular
degree may change over time since it has been eliminated. In addition, including
every possible educational level and combinations thereof would have resulted
in extreme cell paucity among the factors. Finally, choosing three categories for
the educational levels allowed for consistency. All attitude questions for age,

gender, and education level had three choices.

124

This section on education is organized by the educational level of the
informants, starting with those who have BEP or a BTS. The scholastic track
degrees were kept separately, so though the BTS represents more education
than the Baccalauréat, it was placed between the BEP and the Baccalauréat.
The DEUG has been eliminated recently in order to streamline French education
with European education, but many of my informants were educated at a time
when it was still valid. It was collapsed with the Licence, the three-year degree
that replaced it, but which has also historically been an option. The final
grouping is ‘Maitn’se and up,’ any of the multiple degrees after that educational
level, including directed studies and the doctorate.

4. 2. 2. 1 Perceived education-level usage by education of informant

Table 3: Perceived education-level usage by education of informant

 

 

% N

ALL INFORMANTS

Not very educated 13.74 100

Not very and 10.71 78
somewhat

Somewhat educated 33.52 244

Somewhat and very 1.51 11

Very educated 4.81 35

All 33.10 241

Other 1 .37 1 0

Total 100 728
BEP

Not very educated 6.35 8

Not very and 16.67 21
somewhat

Somewhat educated 35.71 45

Somewhat and very .79 1

Very educated 8.73 11

All 31.75 40

Other 0 0

Total 100 126
BTS

125

Table 3 (cont’d)

Not very educated 18.52 5
Not very and 11.11 3
somewhat
Somewhat educated 0 0
Somewhat and very 0 0
Very educated 3.7 1
All 66.67 3
Other 0 0
Total 100 27
Bac
Not very educated 17.99 25
Not very and 10.07 14
somewhat
Somewhat educated 25.90 36
Somewhat and very .72 1
Very educated 5.04 7
All 38.85 58
Other 1.44 2
Total 100 1 39
DEUG and licence
Not very educated 16.28 49
Not very and 4.98 15
somewhat
Somewhat educated 43.52 131
Somewhat and very 1.99 6
Very educated 5.32 16
All 27.57 83
Other .33 1
Total 100 301
Master and higher
Not very educated 10.32 3
Not very and 19.84 25
somewhat
Somewhat educated 25.4 32
Somewhat and very 2.38 3
Very educated 0 0
All 36.51 46
Other 5.56 7
Total 100 126

 

Regardless of their own level of education, we still see that only 13.74% of

the informants attributed the lexical items to those who are ‘not very educated,’

126

and only 10.71% of the informants said that it is used by both those who are ‘not
very educated’ in combination with those who are ‘somewhat educated.’ The
vast majority, in fact, chose ‘somewhat educated,’ at 33.52% or ‘all’ education
groups, at 33.1%. Very few informants chose either ‘highly educated’ alone or
‘highly educated’ in combination with ‘somewhat educated,’ at only 1.51%. This
may be explained by the fact that banlieue language is a non-prestige variety,
and therefore does not represent change from above. It is, however, notable that
in general, they rarely attributed usage solely to those who are ‘not very
educated’. This goes in direct opposition with Lepoutre (1997: 335) and the idea
that those who retain the dialect are part of a ‘sous-culture délinquante et
déviante,’ ’deviant and delinquent sub-culture.‘

Another interesting phenomenon, found when informants are separated
according to their own level of education, is that among the educational groups,
the perception of ‘not very educated’ never reached any higher than 18.52%.
This percentage was held by the BTS group and was closely followed by the
Baccalauréat group at 17.99%. The BEP group, that with the lowest education,
was the least likely to attribute usage to lower education, at only 6.35%, but
interestingly enough, the group with the most formal education was actually
closest in opinion to those who held a BEP. Even among those with the most
education, however, we see that only 10.32% of them attributed the use of this
vocabulary to those who are ’not very educated.’ If it were truly a case in which
only those with little or no formal education use banlieue language, we would

expect those numbers to be reflected, with a much higher percentage of each

127

educational group choosing ‘not very educated’ as their response and with a
slight difference in condemnation according to the education of the informant.
This is not the case.

Let us now examine those who responded ‘all’ educational levels. The
DEUG and Licence group were at the lower end of the scale with 27.57%.
Conversely, the BTS group was at 66.67%. The other groups were between 33
and 38% on this issue, with the most educated group falling toward the higher
end.

The group that is the most salient in terms of responses is BTS. They
were very polarized in their answers, attributing usage either to all groups
(66.67%) or to the lower educational groups, represented by their choice of
18.52% for the lowest educational level and 11.1 1% for the ‘not very educated’
in combination with ‘somewhat educated.’ Some of the differences may be
attributed to the fact that they had completed a post secondary education
degree, but as we have mentioned, that is not the same as the DEUG because
of its technical nature. The BTS group may consider itself fairly educated, as
indeed it is, but because it is not on the same scholarly track those who hold the
degree may well find themselves in less-formal job environments. The paucity of
the cells may also contribute to this effect as this reflects fewer than 30 answers
over the entire questionnaire, which means two informants, since there were a

total of 21 lexical items per informant.

128

4.2.2.2 Admitted usage by education of informant
Here we discuss admitted usage of banlieue language among the

informants, divided by their own level of education.

Table 4: Admitted usage by education of informant

 

 

% N
ALL INFORMANTS
Often 29.47 247
Rarely 34.96 293
Never 35.32 296
Often and rarely .24 2
Total 100 838
BEP
Often 32.72 54
Rarely 36.36 60
Never 30.91 51
Often and rarely 0 0
Total 100 165
BTS
Often 41.18 14
Rarely 52.94 18
Never 5.88 2
Often and rarely 0 0
Total 100 34
Bac.
Often 39.87 63
Rarely 20.89 33
Never 38.61 61
Often and rarely .63 1
Total 100 158
DEUG and licence
Often 26.5 93
Rarely 39.89 140
Never 33.33 1 17
Often and rarely .28 1
Total 100 351
Masters and higher
Often 21.74 30
Rarely 34.06 47
Never 44.2 61
Often and rarely 0 0
Total 100 1 38

 

129

Education does indeed seem to matter in terms of self-reported usage. It
is, however, worthy of note that according to much of the research, the age and
education levels of these informants would inhibit use of banlieue language
(Doran 2002, Lepoutre 1997). Additionally, since education is often viewed as
one of the determining factors for socio-economic group (T eevan 1985), the
higher levels of education should not be admitting to usage by virtue of their
educational level, either. We find, however, that this is not the case.

In fact, over all, results including all informants show that only 35% say
they never use the vocabulary in question. The group with the highest self-
reported non-usage is the ‘Maitn'se and higher’ group at 44.2%, followed by the
‘Baccalaureat’ group at 38.61%. Most of the other groups are in the 30th
percentile for ‘never’, which means that the majority does admit to using the
vocabulary. 65% of informants, in fact, still admit to some form of usage. ‘Often’
ranges between 21.74% for the highest level of education to over 41% with the
’BTS’ group, which is again closely followed by the ‘Baccalauréat’ group at
39.87%. The ‘DEUG and Licence’ group is slightly higher than the ‘Maitn'se and
up’ group at 26.5%. The ‘BEP’ group admits to using the vocabulary often at
only 32.72%, toward the lower end.

4. 2.2.3 Comparative analysis

First and foremost, the reader should be reminded that while no study has
ever specifically been done on the educational level of those who speak
banlieue language, Doran (2002) posited that the variety was dropped once a

student decided to go the scholastic track, and Lepoutre (1997) relegated

130

banlieue language in late teens and older to derelicts. Therefore we might have
expected that the informants, all older than 16, and all having at least the
baccalauréat or equivalent, would not admit to speaking banlieue language at all
based on those two studies. This isnot the case.

In fact, in the section on individual lexical items, we see that no word was
exclusively seen as used by those with low education. In 4.2.2.1 this is
highlighted, since even those informants with the most formal education say that
the lexical items in this dissertation are used by people of all educational
backgrounds.

However, just as with gender, there is a difference between infonnants'
perceptions of what educational level uses the vocabulary most and what they
themselves admit using by education level. We see in 4.2.2 that education level
of the speaker does play a role in how often one admits to using the lexical
items. This would conform to what we would expect in a non-prestige variety.
However, while those with higher education admit to using banlieue language
less than those with a lower level of education, they still admit to using the
lexical items. This allows me to suggest that education may not have as strong
an effect as previously thought, and may, like gender, be less important than
solidarity and identity with the area. Labov's Martha's Vineyard study (1972)
proved that those with strong feelings of identity and who wished to stay in the
area adopted a more regional accent then those who wished to leave the area. I
propose that this may have an effect in the banlieue and that the covert prestige

(Labov 1972) and symbolic power (Bourdieu and Thompson 1991) of using the

131

variety may play an important role in the high levels of admitted usage among
my informants.
4. 2.3 Age

Here I examine the perceptions of what age group would use the lexical
items and compare those results to what the informants in this group admit to
using.

As a reminder, Doran (2002) said that use of the dialect ends around age 16
when students decide whether they will be continuing the educational track or
moving to a more technical track, which developed the theory set out by
Lepoutre (1997).

For analysis, I innovated a little. Typical sociolinguistic research asks the
informant to choose among specific age groups or a system that includes
‘young’, ‘middle aged’ and ‘old’. I argue that this is in some way more
amorphous than my system since judgments on what age is considered ‘old’ or
‘young’ already depends on the age of the informant and his or her perceptions
of that age. Someone who is 18 may view ‘young’ very differently than someone
who is 36.

Furthermore, I have been very consistent about keeping the labels that
informants used to describe themselves throughout this dissertation whenever
possible since the factor of identity plays such a strong role in who will or will not
use banlieue language. Therefore, allowing subjects to self-identify forced them
to look at those who they believe to use the lexical items in relation to

themselves. I believe that allowing them to use their own generation as a focal

132

point permitted the patterning of identity of one group versus another. It focused
the issue on the informants instead of imposing upon them my own categories.

In addition, it allowed informants to select multiple categories, yielding
more complete results. Therefore, if an informant viewed himself or herself to be
on the cusp of two generations, perhaps because his or her cohort or peer group
was younger than the informant, that possibility could be expressed by a
combination of the categories ‘a younger generation and mine’ and ‘mine and an
older generation,’ and finally ‘all generations.’ I believe that this gives us a more
accurate picture of what the informants believe to be generational boundaries. It
is also important to note that responses were not counted twice. Therefore, if an
informant circled all three responses, his or her response fell under ‘all’ and not
under any of the separate responses.
4.2.3.1 Perceived usage by age of informant

Here we examine the perceptions of age in relation to banlieue language.
lnforrnants decided if the lexical items belonged to their own generation, an older
generation, or a younger generation. This section is analyzed in terms of the
infonnants' age group.

Table 5: Perceived usage by age of informant

 

 

% N

ALL INFORMANTS

Younger 10.26 82
generation

Younger and 22.03 176
mine

My generation 37.42 299

Mine and older 4.51 36

Older generation 10.03 80

All 15.64 125

133

Table 5 (cont’d)
Total
18- and 19 year-olds
Younger
generation
Younger and
mine
My generation
Mine and older
Older generation
All
Total
20- to 24-year-olds
Younger
generation
Younger and
mine
My generation
Mine and older
Older generation
All
Total
25- to 29-year-olds
Younger
generation
Younger and
mine
My generation
Mine and older
Older generation
All
Total
30- to 35-year-olds
Younger
generation
Younger and
mine
My generation
Mine and older
Older generation
All
Total

100
11.92
11.02

51.97
3.15
14.96
7.87
100

15.62
14.11

49.25
2.7
12.9
5.41
100

5.26
33.97

23.44
8.61
4.78
23.92

3.85
33.85

15.38
3.85
6.15
36.15
100

134

798
14
14
66
19
10
127
52
47
164
43
18
333
11
71
49
18

10
50

44

20

47
130

Striking among the responses for age is the paucity of ‘a younger
generation’ responses. The overall responses have just 10% of the informants
saying that the vocabulary belongs to a younger generation, even though the
informants range in age from 18 to 36. What is even more interesting is that
among the oldest group, fewer than 4% said the lexical items belonged to a
younger generation. Instead, it was the younger speakers who attributed
banlieue language most to people younger than they, and even there it never
exceeds 15.62%. This may be in accordance with Doran’s (2002) theory that the
students who are arriving at the university are loathe to use the lexical items in
order to avoid a perceived stigma. However, it may also be due to an aging of
the lexical items, since the lexical items were given to me by someone who was
31 years of age at the time of the questionnaire creation.

Also in direct contrast to what one might expect, it is the oldest group that
claimed most usage of ‘all generations’ at 36.15%, (47/130 tokens), which is
12% above the next highest group. The second highest group is, interestingly
enough, the second-oldest age group, that of 25-29. They chose ‘all’ at 23.92%
(50/209). The younger two groups are both under 10%.

Again, the overall statistics have almost 40% (37.42%) of the respondents
choosing their own generation as speakers of the variety, with ages ranging from
18 to 36. This is unexpected. At the other end of the spectrum, it was more
expected that only 4.51% (36l1407) of the informants would say that the lexical
items belonged to someone of an older generation in combination with their own

generation, since this was originally viewed as a language of youth. We should

135

El‘
3f
i954
la“

4

ed

ati

3‘31

 

 

 

 

also underscore that only 10.03%, or 80 tokens, were viewed as the lexical
items of an older generation. When looking at the overall statistics, 15.64% of
the tokens were viewed as belonging to all generations, which leads us to
investigate the responses by age group.

I now discuss their responses by age groups, with the oldest age group
analyzed first. We see that though they were between 30 and 36 years of age,
this group viewed only 3.15% of the vocabulary items as belonging to a younger
generation, which is a mere four tokens. They were fairly evenly divided
between ‘younger and mine’ at 33.85% and ‘all’ at 36.15%. They seemed to
view these items as vocabulary they share with younger generations, since only
15.38% claimed the vocabulary as belonging to strictly their generation. This
may indeed open the question of retention among older users.

The 25- to 29-year old group is almost identical on the ‘younger and mine’
combination, but claims ‘my generation’ by itself more often, at 23.44% and ‘all’
is slightly reduced at 23.92%. This may be because this group considers itself
on the older edge of those who would use banlieue language.

A great age divide becomes apparent between those of the first two age
groups and those of the oldest two age groups. Both of the younger groups
claimed their own generation at the largest percentage in their category, 20- to
24-year olds at 49.25% and 18- and 19-year olds at 51.97%, while the two older
groups said that all generations use it. What may be a factor is the changing
nature of the banlieue. The banlieue is changing so quickly that one set of three

related informants said that the banlieue had changed between the oldest

136

informant
younger.
An
insaying
group at
han do
42.3.2
I
educatir

“110 L154

Table E

Tu

18

 

 

informant's generation and that of the middle informant, only three years
younger, and yet again for the youngest informant, another three years younger.
Another unexpected element is that both of the younger groups are higher
in saying that the lexical items belong to an older generation. The youngest
group at almost 15% and the 20- to 24-year olds at almost 13%, this is more
than double that of the highest percentage of the two older groups.
4.2.3.2 Admitted usage by age of informant
I shall next examine at the admitted usage by age. We have seen that in both
education and gender, admitted usage has differed from what they perceive as the people

who use the lexical items in question.

Table 6: Admitted usage by age of informant

 

 

% N
ALL INFORMANTS
Often 29.47 247
Rarely 34.96 293
Never 35.32 296
Oftenand rarely .24 2
Total 100 838
18- and 19-year-olds
Often 23.48 31
Rarely 22.73 30
Never 53.03 70
Oftenand rarely .76 1
Total 100 132
20- to 24-year-olds
Often 29.33 105
Rarely 38.27 137
Never 32.12 1 15
Oftenand rarely 0.0 0
Total 100 358
25- to 29-year-olds
Often 31.78 68
Rarely 31.78 68
Never 35.98 77

137

Table 6 (cont’d)

Oftenand rarely .47 1
Total 100 214
30- to 35-year-olds

Often 31 .85 43
Rarely 42.96 58
Never 25.19 34
Oftenand rarely 0.0 0
Total 1 00 1 35

 

In examining the data we notice that informants, even among the age
groups that have been signaled as not using banlieue language (Doran 2002,
Lepoutre 1997), admitted to using the vocabulary items often. Instead of the
youngest group using the vocabulary the most and a sliding scale of downward-
progressing percentages among the older informants, which is what we might
expect, the opposite is true. In fact, at each age group the admitted usage in
terms of the ‘often’ category rises. The youngest group admitted to using
banlieue vocabulary ‘often’ the least of all four groups (23.48%, or 31/132), with
the second group at 29.33% (or 105/358), the third group at 31.78% (68/214),
and the oldest group admitting to using it the most (31.85% or 43/135). Admitting
to using vocabulary items ‘rarely’ does not show as progressive a continuation,
since though the oldest group still has the highest usage (42.96% or 58/135) and
the youngest group still has the lowest usage (22.73% or 30/132), the two inner
groups are more to what we would have expected. The 20— to 24-year olds admit
to using the vocabulary more under the category of ‘rarely’ (38.27% or 137/358)
and the 25- to 29-year-olds admit to using the vocabulary items at a rate of

31.78% (or 68/214).

138

When one investigates the responses of those who said that they never
use the lexical items, we might expect a sliding scale from the youngest to the
oldest, with the latter admitting to the least usage. Again, this is not the case. We
have a sliding scale that is the exact opposite of what we would expect. We see
that in the category of ‘never,’ the highest age group to claimed it is actually the
youngest age group, at 53.03% or 70/132. The second group is at 32.12%
(115/358), the third group at 35.98% (77/214), and the oldest group at a mere
25.19% (34/135). I would like to take this time to remind the reader that if the
informant did not know the lexical item, but still answered the questions on
perceptions, those answers were not included. Therefore, those who admit to
‘never’ using the items in question did indeed know the words.

We turn to an examination of admitted usage in general, or more
specifically, adding those who admitted to rare usage to those who admitted to
using it often. This means that 74.8% of our oldest group admitted to using the
vocabulary, while only 45% of the youngest group admitted to using it. Again,
the 20- to 24-year-old group performs more to what we would expect (compared
to the other age categories and not to what we would expect in terms of previous
research), at slightly less in terms of percentage. 67% of the informants admitted
to usage compared to 63.6% in the 25- to 29-year-old group. It is extremely
interesting that even the group that admitted to using the vocabulary the least
still admitted to some kind of active vocabulary usage at 63%.

It is worth mentioning here again that the list was created by a native who

was 31 at the time, which would put him in the oldest group. The holiday

139

gaterin
dealer
ddnot
niuenr
deata‘
datlnc
hanpr

4.2.3.3

diose
order I
ttzni
diner
cats

inc

gatherings venue does indeed contain several social groups of the native
speaker who created the list, as well as their significant others. Those women
did not grow up with the list-creator. So while we may be seeing a generational
influence, it is not simply in-group/ out-group with the speaker who helped to
create the list. What is important to us, however, is that regardless, empirical
data now proves that the vocabulary is being used, and by much older speakers
than previously thought.
4. 2. 3.3 Comparative Analysis

In the graph below, I counted the number of informants in each group who
chose their own generation by itself, or in combination with any other group in
order to represent the perceived usage by age group. As to the admitted usage,
the numbers were computed as in section 4.1. This means that there was no
difference between admitting to using the words rarely or often to create a binary
category. An informant either was counted as admitting to using the lexical item

in question or not admitting to using the vocabulary in question.

140

 

Figure 2

group

tier,

 

 

perceived vs. admitted usage by age
groups

-l—perceived usage
-0-admitted usage

 

18&19 20-24 25-29 30-36

 

 

 

Figure 2: Perceived vs. admitted usage by age groups
As with the other categories there is some discrepancy between what a

group perceived as those who would use the vocabulary in question and what
they, themselves, admitted to using, though the 24- to 29-year-old group is very
close. In each case, however, we see that the perceived usage of a group's
generation was higher than that generation admitted to using. The most salient
is the youngest group, which attributed the vocabulary to their generation at
74.1%, but only admitted to 46.9% usage. These differences may be due to the
women who made up the larger part of the study at the university venue (which
was both a younger venue and comprised mostly of women who claimed to use
it less overall), where not all informants were from Val d'Oise or Seine-Saint-
Denis. It may be that some informants, particularly at the younger venue of the

university, did not identify with the area and its language, or Doran's (2002) idea

141

of separation from the variety at the university level may explain the difference.
Additionally, as l have mentioned before, while I told informants that the survey
was anonymous and they were relatively familiar with the process due to
frequent surveys on campus, they may have underreported usage to a stranger
because of the ‘we code’ (Gumperz 1983) nature of banlieue language.
Therefore it is extremely possible that not only is age a factor but that since
there is no way to control for age and gender or age and venue at the same
time, there is a slight crossover effect.

Even more important is the visual representation of both the usage and
the perceived usage. Both increase steadily as the ages of the people within the
group increases, not what we would have expected since so many scholars had
originally delineated it as a youth phenomenon (Boyer 1985, Doran 2002,
George 1986, Lepoutre 1997, Tejedor de Felipe 2004, Walter 1997). In addition,
it is necessary to underscore that no group admitted to using less than 46.9%.
More importantly, the oldest group admitted to using 74.8%. Therefore we are
able to challenge the established idea of an age limit, put forth by Zerling (1999)
and Calvet (2006), to banlieue language with quantitative, empirical data.

As to the underreporting, the aforementioned hypotheses apply. It is very
possible that the two youngest groups were in environments in which banlieue
language is not as appreciated. In particular, those at the Mission Locale may
have decided to underreport due to fears that it may adversely affect their job
possibilities, despite reassurances of anonymity. Another possibility is that

perhaps some of them were indeed not using the vocabulary. Perhaps they did

142

Mdmh
sudias'

someonr

issues
ngme
educat

usage

metro
numb.
anare

num

 

not identify with minority struggle. While I fully recognize certain AAVE tokens
such as ‘wazzup’ as part of my generation, and would not necessarily categorize
someone who used it as having a low level of education, I would not use it.

4. 2.4 Language of the lexical item

The original language of the lexical item is another factor that has to be
discussed. However, since the informants did not make any correlated
judgments, that is, perceived judgments such as those of age, gender, and
education, the variable of language will be discussed in terms of self-reported
usage alone.

However, to make the results more meaningful and to provide another
method of comparison, for each category of language, there are two sets of
numbers provided. The percentages of non-answered and unexpected answers
are reinvestigated here since they are relevant. Therefore, the first set of
numbers is the percent and number of all non-answered/unexpected answers
per language. The remaining percentages are based on only those with
expected answers, to remain consistent with the rest of the thesis. This
distinction was justified due to a statistical significance between Creole and
some of the other languages.

4. 2. 5 Admitted usage by the original language of lexical item

Table 7: Admitted usage by original language of lexical item

 

 

 

% N
ALL INFORMANTS
Non-answers 37.74 531
Often 29.47 247
Rarely 34.96 293

143

Table

Aral
Non-

Arg
Nor

 

 

Table 7 (cont’d)
Never
Oftenand rarely
Total

Arabic

Non-answers
Often
Rarely
Never
Oftenand rarely
Total

Argot

Non-answers
Often
Rarely
Never
Oftenand rarely
Total

Creole

Non-answers
Often
Rarely
Never
Oftenand rarely
Total

Roma

Non-answers
Often
Rarely

Never
Oftenand rarely
Total
Verlan
Non-answers
Often
Rarely
Never
Oftenand rarely

Total

35.32
.24
100

34.33
37.88
34.85
26.52
.76
100

24.25
34.48
33
32.02
.49
100

67.16
18.18
42.18
38.64
0

100

35.45
21.97
38.15

39.6
.29
100

20.15
37.38
29.91
32.71
0

100

296
2
1 407

69/201
50

46

35

1

1 32

65/268
70

67

65

1

203

180/268
16

38

34

0

88

1 90/536
76
1 32

1 37
1
346

27/134
40

32

35

0

107

The first interesting, noticeable difference is among the percentage of

unexpected and non-answers given by language. While Roma and Arabic had

35.45% (190l536) and 34.33% (69/201) of non/unexpected answers

ie

(I)

respectively, 67.16% (180/268) of the informants did not know or did not answer
the Creole questions. At the other end of the spectrum, there is Argot with only
24.25% (65/268) and Verlan with only 20.15% (27/134) of non/unexpected
answers.

All responses were ‘normalized’ as well in order to analyze simply those
who gave an expected answer. This means that once it has been established
what percentage of non-answers and unexpected answers existed for each
language, those non-answers and unexpected answers were excluded from the
rest of the statistics, much as I did for gender. In this manner, we see a true
comparison among the languages.

The rest of the statistics in this section should be viewed as ‘of those who
gave an expected answer, X many vocabulary tokens were admitted to by the
informants as a whole.’ In this case, numbers represent the percent of
responses that were ‘often,’ ‘rarely,’ ‘never,’ or a combination, out of the number
of total expected responses. For example, with Creole, there were a total of 88
expected responses out of 67 surveys and four lexical entries per survey. Here
again, we find that Creole distances itself from the other languages. Of those
who identified the expected answer, a mere 18% (16) admitted to regular use,
while 42% (38) said that they rarely used the lexical items. Finally 38.7% (34)
said that they never used the lexical items. When we compare this to the other
languages, we see that in Arabic, the informants admitted to using 34.3% (50)
often, another 37.9% (46) rarely, and only 26.5% (35) never. In Argot, the

informants admitted to using 34.5% (70) often, 33% (67) rarely, and only 32%

145

(‘65) never. '
39.6% (137
(32) rarely,

In inv
may be ger
program wi
fonctionnai
hat occasir
influx of Ca
slowed dov
Iallguage o
Creole wou
assimilatior
title since [
Who does r

Al th
noll/unexp.
and my 3:
of admittec

Argo
34% (70) c
additional 1

We °nly :

(65) never. In Roma these numbers are 22% (76) often, 38.2% (132) rarely, and
39.6% (137) never. Finally, in Verlan the numbers are 37.4% (40) often, 29.9%
(32) rarely, and 32.7% (35) never.

In investigating this phenomenon, I am able to suggest that much of this
may be generational. In fact, during the 1970s and 19803, France instituted a
program with the DOMTOM, and those who wished to work in France as a
fonctionnaire (highly envied state jobs) were allowed to participate in a program
that occasionally paid for voyages to their former homes. The result was an
influx of Caribbean ‘immigration’ to the Metropole in fonctionnaire positions. This
slowed down with time. Since they are French departments, French is the
language of education in those areas. Therefore, since people who speak
Creole would also speak French, it should be reasonable to expect that
assimilation would occur more quickly than with other groups. This is especially
true since many people who speak Arabic or Roma have at least one parent
who does not speak French very well (Carpin, Tavan, and Dugué 2005).

At the other end of the spectrum, there is Verlan, with only 20% of
non/unexpected answers and 37% (40) responses of ‘often,’ 30% (32) of ‘rarely,’
and only 32% (35) of ‘never.’ It is the most popular language category in terms
of admitted usage.

Argot is almost identical to Verlan for self-reported usage percentages.
34% (70) of informants admitted to using the lexical items frequently, and an
additional 33% (67) said they rarely used them. In the ‘never’ category, there

were only 32% (65).

146

Many
high percer
iere we se
this sectior
answers, I
rare usage

Aga
frequently
Doran 20
though it
to identih
Speech a

Rc
banlieue
identical
(35.5%)
Usage. |
lSpossi
CI‘llltit: I

H
phenon
languas

Many scholars (Calvet 1997, Doran 2002, etc.) have commented on the
high percentage of Arabic items in the language spoken in the banlieues, and
here we see that this is supported by the empirical data. Much of the statistics in
this section are influenced by kiffer and ouaiche. Of the total of expected
answers, 38% (50) admitted to using the vocabulary often. 35% (46) admitted to
rare usage. 32% said they never used it.

Again, many speakers are of Maghrebi heritage in the banlieue, and
frequently at least one of their parents is not fluent in French (Amara 2002,
Doran 2002, Jablonka 2003, Melliani 2000). Arabic is often spoken in the home,
though it is not ‘classical Arabic.’ Melliani (2000) and Doran (2002) have pointed
to identity factors as a reason for introducing Arabic lexical items into their
speech and the frequency has led to it being absorbed into the local parlance.

Roma is unexpectedly prolific in its contribution to the vocabulary in the
banlieue as a language of immigration. Its roots are nomadic, but it is almost
identical to Arabic in percentages of those who gave unexpected answers
(35.5%). 22% (76) admitted to using it regularly, and 38% (132) admitted to rare
usage. Much of the borrowed vocabulary relates to theft, drugs, and police, so it
is possible that borrowing from a lesser-known language would promote the
cryptic function, originally the idea behind argot and the malfaiteurs.

However, we must emphasize the fact that Verlan and argot, both French
phenomena, are the most widely known. They are also the categories of
language for which the informants admitted usage. Tejedor de Felipe's (2004)

theory of banlieue language as a mostly French idea seems to bear true.

147

The differer
nionnant I:
equivalent
lower as a
people, so
had settler
the Missio

generally

Table 8; A

\

m

Frie

n9fi1f‘3—Jfl77‘1n

4.2.6 Venue
The differences among the venues are such that in each case a very different
informant base was solicited. In the university setting all had the Baccalauréat or
equivalent in order to be enrolled. At the Mission Locale, the education level was
lower as a whole by the very nature of that venue. Finally, the last group of
people, solicited in private gatherings, were rather well-educated people who
had settled into careers and adulthood. Therefore, the hypothesis would be that
the Mission Locale would be the group to use the vocabulary the most, given the

generally lower education level.

Table 8: Admitted usage by venue

 

 

°/o N
ALL INFORMANTS
Often 28.77 252
Rarely 35.96 315
Never 24.93 306
Often and rarely .34 3
Total 100 876
L’Université de Saint
Denis
Often 28.21 123
Rarely 35.32 154
Never 35.78 156
Often and rarely .69 3
Total 100 436
La Mission Locale
Often 27.86 39
Rarely 31.43 44
Never 40.71 57
Often and rarely 0 0
Total 100 140
Ffiends
Often 30 90
Rarely 39 1 17
Never 31 93
Often and rarely 0 o

148

Table 8 I
Tote

’

Whe
Locale info
investigate
with unexp
answers is
answers vr
299/735, <
answers v

Adc
results frc
had the h
ifthey dic
Comparis
The resu
Usage_ Ir
the Missi
at 31_43r
DllVate g
{rarely}

ls

gatherin

Table 8 (cont’d)

r Total 100 300

When examining the results, however, we find that it is not the Mission
Locale informants who admitted to using the lexical items the most. I first
investigate this data by way of those questions left unanswered or answered
with unexpected responses. In terms of percentages, the highest group of non-
answers is actually found in the Mission Locale. Their ratio of non-answers to all
answers was 112/252, or 44%. The university was the middle group, with
299/7 35, or 40%. Finally, the lowest percentage of non-answers or unexpected
a"Iswers was the friends group, at 120/420, or 28.6%.

Additionally, when we look at self-admitted usage, we find that, again, the
resu Its from the Mission Locale are not what we would expect. In fact, this group
had the highest number of responses in the ‘never’ category, meaning that even
if they did know it, they claimed not to use 40.7% of it (57/140). This is in
comparison to 35% of the university setting and only 31% of the private venue.
The results are also reversed of what one might expect in terms of general
usage. In terms of categories, both ‘often’ and ‘rarely’ were admitted to least in
the Mission Locale. The category of ‘often’ is at 27.86% (39/140) and ‘rarely’ is
at 3 1 .43% (44l140) in the university setting. The highest percentages were in the
priVete gatherings venue, at 30% (90/300) for ‘often’ and 39% (117/300) for
.ra rely}

I suggest that there may be a few factors at work here. First, private

gatherings were filled with peers of the list-maker, though it is important to note

149

hat their sig
group. While
surrounding
were older 2
shown, whic
informants,
close. persc
knowing me
me persona
3de to usi
with its mer
A few
Vocabulary
are Slill USI
Additionall
Teillard (1.
the infoirn
just as it v
the Wild
was admi

“new. h:
(Labov 1g

that their significant others were also present and also were informants in this
group. While the core group of males may have belonged to Garges and the
surrounding areas, many of the females did not. In addition, these speakers
were older and tended to be more familiar with the Creole vocabulary, as I have
shown, which is another contributing factor. Thirdly, l was in-group with these
informants, and either I was a close, personal friend or was introduced by a
close, personal friend of theirs. Because of this trust factor, that of either
knowing me personally or being introduced to be in a group of peers who knew
me personally and trusted me, these informants may have felt more free to
admit to using banlieue language. This group was also well-established socially,
With its members enjoying full-time employment.

A few informants, all male, at the Mission Locale mentioned that the
V<1>carr:lbulary was of an older generation, which may mean that younger people
are still using banlieue language, but said language is changing rapidly.

Add itionally, the younger groups may well be innovating as the Séguin and
Teillard (1996) informant suggests. When referring to the Mission Locale alone,
the informants there were told that I was a student studying banlieue language.
just as it was explained in every other venue. However, at the Mission Locale,
the participants were actively searching employment at the very place where l
was administering my questionnaire. Banlieue language, like any non-standard
Va riety, has a stigma in the job world. It is possible that the Observer's Paradox

(La bov 1972) would be more strongly felt in this venue and that the informants

150

may have felt that they must behave differently in the Mission Locale so that
they could find employment.
At the university, the students were very used to questionnaires since the
social science students administer surveys and questionnaires for classes
regularly. They were, however, in an environment where the stigma would be
more apparent, as Doran (2002) has indicated.
Finally, another important remark about these results is that while one
9" 011p may have a slightly higher percentage than another group, it is important
to note that they are still all fairly close. When the categories are normalized and
solely expected answers are analyzed, there is very little difference among the
groups in terms of how much banlieue language they admitted to using. All of

these groups admitted to using the lexical items in this dissertation, regardless of

age . education, and gender.

4. 2. 7 Overall statistics analysis
In general, we find that among the different categories—gender, education, and
age‘the perceptions the informants harbored about who uses the vocabulary
differed from what the informants themselves admitted to using. While both men
and women said that both sexes use the lexical items equally, we find that
indeed, men did admit to using the words more than the women. As previously
mentioned, this may be due to the sociolinguistic tendencies among the genders

ll". both usage of standard language and reporting of their usage of standard

lat”Oguage.

151

In terms
signal a mlddl
generally tenr
we might hav
then the int:
group was th
language to
dwhmep
was the fact
vocabulary '
the vocabul
education r
categories,
see a sllgh
the vocabt
results we;
speaking I

As t
the lexical
”1°39 Whr
Older. In E
generatio

some of l

In terms of education, we find that on average, informants preferred to
signal a middle level of education for those who use the variety. They did not
generally tend to attribute the variety to those with a low level of education, as

we might have expected, since banlieue language is a non-prestige variety.
When the informants are broken down into groups of similar education, the BTS
group was the most distinct and the most generous in attributing banlieue
language to all educational levels, though more investigation may be needed
due to the paucity of informants. The most important element of this category
was the fact that even among the most educated group, only 10.3% of the
Vocabulary was attributed to those of low levels of education, and over 36% of
the vocabulary was attributed to all educational groups. In general, all groups of
ed ucation had the most responses in the ‘all’ or ‘somewhat educated”
categories. When we contrast this with their admitted usage, however, we do
see a slight disconnect. Those with the most education did indeed admit to using
the Vocabulary less themselves than the other groups. Again the BTS group's
reSl-Ilts were very salient since those informants overwhelmingly admitted to
Speaking the variety.

As to age, when we investigate the perceived age level of those who use
the lexical items in the variety, we see that there is an inverse proportion of
those who attributed the variety to a younger generation as the age groups get
o'der. In addition, the two youngest groups attributed more items to the older
generations than either of the two older groups, which leads me to believe that

sothe of the lexical items, especially those from the linguistic category of Creole,

152

may be vocat
acquire and L
lino chose ‘3
generations.
This 34
groups adm'
peroentage
concentratir
of admitted
Final
results are
Vocabulary
0h M
slightly mr
closely to
negative,
the result
more Sin
i”Ormar
W

the Miss
those m

least lik.

may be vocabulary that only the older generations were exposed to enough to

acquire and use. Finally, this supposition is also supported by the fact that those
who chose ‘all generations’ with the most frequency were those in the oldest
generations.
This seemingly backwards distribution is not negated by what the age
groups admitted to using. In fact, the older generation reported the highest

percentage of admitted usage among the age groups and the lowest

concentration of ‘never,’ and the youngest group reported the lowest percentage

of admitted usage and the highest percentage of ‘never.’

Finally in this category I investigate both language and venue. Language
resu Its are consistent with the earlier section in which informants admit to using
Vocabulary from Roma, Arabic, argot, and Verlan at over 60% in each category.
only 20.2% of the Verlan vocabulary received an unexpected answer, and only
slightly more (24.3%) in argot. In Arabic, that figure increased to 34.3%, very
(“08er followed by Roma at 35.4%. Creole, at 67.2% of the responses in the
heQative, indicated that many people were not familiar with the vocabulary. Once
the results were normalized, however, the percentages of admitted usage were
moi‘e similar, with almost 60% of each category of languages being used by the

informants who knew the vocabulary.

When contemplating venue, we might have expected at first that those in
the Mission Locale venue would speak the most banlieue language and that
th ose most firmly established in society or in an educational setting would be the

'east likely to utilize the vocabulary. This was disproved. In fact, we see that the

153

Mission Locale was the least likely to know and to use the vocabulary, and those

of the friends venue were the most likely to admit usage of the variety.

4.3 Statistically significant factors influencing knowledge of lexical items
Tests of statistical significance for knowledge of the lexical items were run on
Goldvarb.
4.3.1 Factors tested
The Goldvarb program forces the researcher to make many choices about which
items are significant. In the first run, all factors were tested separately, even
tho ugh they may have been combined in other sections. This allows us to
examine the significance of individual factors, regardless of any social prejudice.
A researcher may find, for example, that arguably socially similar factors such as
the DEA and the Maitrfse had completely opposite weights. Once the program is
"J n . as I mentioned before, the researcher combines any factors that are of
Similar social significance as well as similar weights. For example, one could not
combine the BEP and the DEA even if the weights were identical, since they are
at Such opposite ends of the educational scale.

Each of the factors was tested separately, but I will group them here for
the sake of facility of discussion. The first factors are gender, age, education,
and residential history. Therefore, for gender, both males and females were
teSted. For age, each of the four age groups was tested. For education, each
particular educational level was tested, from BEP, BTS, the Baccalauréat,

lDEUG, License, Maitn’se, DEA, DESS, and the one older informant was used

154

and had no t
residential hi
(91). Essonr
Natior
whether or I
nationality. '
specific nat
French anc
French, Fn
Ahenan,F
into hNO oz
by the Ma
and Easte
An
The read.
The ques
Check ‘yE
many on
Tt
WhethEr
Categdri
Spanisi

who had no education, to see where he lined up with the other groups. In
residential history, there were Paris (75), Seine-et-Mame (77), Hauts-de-Seine
(9 1), Essonne (92), Seine-Saint—Denis (93), and Val D'oise (95).
Nationality, birthplace, and ethnicities follow. I tested the significance of
whether or not the informants were born in France. Then I tested their
nationality. Those reporting two nationalities had a code indicating their two
specific nationalities. Therefore someone who was French and Bulgarian or
French and Portuguese was given separate codes. The original categories were
French, French and Bulgarian, French and Portuguese, French and Polish,
Algerian, Portuguese, and Romanian. Ethnicities were originally broken down

into two categories. For this category, we had France; Africa, which was divided

by the Maghreb; West Africa and Eastern Africa; Central and South America;

and Eastern Europe.

An informant's ties to religion were also tested for statistical importance.
The reader should be reminded here that these categories were self-assigned
The questionnaire simply provided a blank line for religion and then a place to
check ‘yes’ or ‘no’ if the informant practiced the religion. Therefore, there were
many original labels; Catholic, Muslim, Berber, Christian, Protestant, and Kabyl.

The informants were also asked to provide the languages they spoke and
whether or not that language was spoken fluently. Originally there were three
caliegories for language. The first language included French, Arabic, Kabyle,
Sdbanish, Creole, Polish, Portuguese and Soninke. The second language

consisted of French, English, Arabic, Spanish, Italian, German, Berber, Creole,

155

and Portuguese. The third language contained responses of English, Spanish

Portuguese, German, Arabic, French, and Italian.

Word origin was also tested as a linguistic factor, which included Arabic
argot, Verlan, Creole, and Roma.

4.3.2 Factors eliminated

The following factors were not elected by Goldvarb as significant.9 What
follows is a discussion of each of those factors.

4- 3.2.1 Gender

Gender was eliminated by Goldvarb as not significant in terms of
knowledge of the lexical items. Men do not show significantly more knowledge of
th Is vocabulary, which seems somewhat unexpected in a non-prestige variety
Th Is seems to support the perceived opinion by the informants in this study that

I"I“Ieri and women used most lexical items equally. We will see if this is borne out

in terms of admitted usage.

4- 3.2.2 Age

Goldvarb eliminated age as insignificant as well. This seems very counter-

"1tuative since other researchers have strongly stressed its importance and

s'nce we did see that the youngest generation seemed to know the least

\

:There were interactions, but they were deemed unimportant because they were
He simply to cell paucity, since there were no females from Hauts-de-Seine, nor
ales from Essonne. Some interactions are inherent in the factors themselves,
such as age and education. While someone in his 403 could land anywhere
Z'Ong the educational spectrum, someone who IS 18 IS likely to have only a
b :ccalaureat or equivalent technical high school degree. For example, it would

th extremely difficult to have a DEA at age 18. This appears as an interaction
at the researcher must reject.

156

 

 

vocabulary; but this is where percentages are not as important as weights and
significance. It is possible that the total number of youth was less than the total
number of informants in those of the older generations.
4. 3.2.3 Languages spoken
The languages an informant spoke did not have a statistically significant
effect on knowledge of the lexical items. Since languages spoken is tied to
ethnicity, and those of Maghrebi heritage have been touted as speakers of this
variety and indeed have claimed it as their own (Doran 2002, Melliani 2000), it is
interesting that the factor is not significant in terms of knowledge.
4. 3.2.4 First ethnicity
Also surprisingly not significant was first ethnicity. This is especially
important since not only do young people of Maghrebi heritage claim banlieue
language as their own (Doran 2002, Melliani 2000) but the major focus of many

banlieue studies is Maghrebi youth.

4. 3.3 Statistically significant factors kept for knowledge
Before considering the results, further mention must be made of how they
were compiled. What is seen in the tables is the information for the people who
claimed one of the significant identity factors. Again, an informant had the choice
of not answering a question for any reason. Therefore he or she might have
filled out most questions with expected responses, but may not have indicated
his or her ethnicity or nationality. In the above case, his or her responses would

not have been counted in terms of ethnicity or nationality.

157

 

 

Statistical importance is determined by weight. A value of 0.500 is neutral.
Anything above 0.500 promotes the knowledge or usage of banlieue language
(in this case) and anything below 0.500 demotes the same.
4.3.3.1 Education

The first significant factor in terms of knowledge was education. 1°

Table 9: Weight of education

 

 

 

weight percentage

Education
DESS/DEA 0.817 76
None 0.652 81
DEUG 0.625 76
Bac 0.519 65
Maitrise 0.500 80
BTS 0.499 69
License 0.460 63
BEP 0.349 59

Range 0.468

In education, only the DESS and the DEA could be successfully
combined. All other factors were either too disperse in terms of weights or not
socially compatible enough in terms of the comparability of their educational

level to collapse.

'0 The reader will notice that educational groups are divided differently between
this section and the section on education in 4.2. In section 4.2, I am discussing
groups which possess a similar level of education. I argue that when people look
at educational levels, the levels in 4.2 are those that come to mind in terms of
length of time spent in the scholastic sector and scholastic versus trade-related
tracking. In this section however, it is significance of factors that becomes
important and therefore the levels have been, in effect, chosen by the logistic
regression program. The researcher then looks at factors that can be collapsed
in terms of both similar weights and similar social significance. It is for this reason
that the educational groupings are so different.

158

Again I am discussing knowledge of the word base, and not admitted
usage, but it is interesting that the DEA/DESS was the highest promoter,
followed directly by the group with no education. When observing the
educational group, one sees that the educational levels are interspersed in what
seems to be a random manner. It is not what we would expect according to the
French media or prior research, meaning it does not show an inversely
proportional relation of knowledge to education level. This hypothetical scenario
would be someone with no education, or someone with a BEP knowing the most
words and those with a DEA/DESS knowing the fewest.

This is not the case and could lead one to conclude that other factors are
at work instead of just educational level. What it also indicates is that someone
from the banlieue, familiar with the language, may still be able to master
standard French well enough to become quite educated. It says that upward
mobility is possible to some extent, but as others have suggested, perhaps it
comes at the cost of leaving some banlieue language behind. This can again be
a type of age grading in which people who are of the age of building a career will
be prone to using more standard language.

4. 3. 3. 2 Residential history

Table 10: Weight of residential history

‘

 

_ weight percentage
Residential history
Paris 0.590 64
Seine-Saint-Denis 0.518 69
and Seine-et—Marne
Val d'Oise 0.480 67
Essonne 0.291 49

 

159

Range 0.371

Residential history was more complex. If one looks at the area in which
these dépan‘ements lie, they are all dispersed around Paris. Using a clock
analogy, the 91, or Hauts—de-Seine is at 7:00. The 92, or Essonne, is directly
above that at about 9:00 to 11:00. The 93, or Seine-Saint-Denis, is from 11:00 to
3:00. The 77, or Seine-et-Marne, is in the second concentric circle around Paris.
It goes from 1:00 to 6:00. The 95, or Val d'Oise, is also in the second concentric
circle around Paris, from 10:00 to 1:00. Decisions were based on not only similar
weights but also geographical proximity. From the original six categories, we
were able to collapse Seine-Saint-Denis and Seine-et-Marne due to their similar
weights and their location.

Living in Paris is a promoter for knowledge of the lexical items in question,
with Seine-Saint-Denis/Seine-et-Mame being somewhat neutral and Val d’Oise
actually being a ‘demoter’. Essonne, located in He de France toward the south,
was the largest demoter.

Looking at the results, some explanation is in order. The Mission Locale‘s
effect may be seen here with Val d'Oise being a slight demoter. As we saw, the
Mission Locale was the group that least embraced banlieue language, and all of
the informants at that venue are from Val d'Oise. As to Essonne, it may well be
the paucity of informants. While Goldvarb does account for differences in cells, a
few more informants from this area would have been welcome in order to

confirm the statistical significance in a greater population. Paris may be a

160

promoter i
from Paris
4.3.3.3 Bir

Table 11:

Birthplai
born in ’
France
observe
we see
4.3.3.4

Table 1

Na'

Categ

promoter because of the proximity to the banlieue or because the informants
from Paris were from ethnically diverse neighborhoods.
4. 3. 3. 3 Birthplace

Table 11: Weight of Birthplace

 

 

weight percentage ’
Born in France
Yes 0.518 70
No 0.407 58

 

Range 0.111

Birthplace is significant in terms of knowledge, and it is indeed those who are
born in France who are more likely to know the lexical items. Not being born in
France is a fairly strong demoting factor. This follows Tejedor de Felipe's (2004)
observation that banlieue language is primarily a French phenomenon, though
we see by the range of 0.111 that this factor is relatively weak.

4.3.3.4 Nationality

Table 12: Weight of nationality

 

 

 

weight percentage
Nationality
French 0.517 67
French and another 0.458 67
European ethnicity
Eastern European 0.339 38
Maghrebi 0.324 55
Range 0.193

Nationality further supports his theory since the only promoter in the

category of nationality is a French nationality by itself. Interestingly, once that

161

nationality is combined with another European nationality, it becomes a demoter,
though still fairly weak. This weakness may be caused by the tendency for
Western European cultures to quickly assimilate (Carpin, Tavan, and Dugué
2004). However, being Eastern European or Maghrebi is a strong demotor for
banlieue language knowledge. These are also two of the slowest to assimilate
according to Carpin, Tavan, and Dugué (2004). Therefore, if this is indeed, as
Tejedor de Felipe (2004) suggests, a French phenomenon, it would be logical
that the more French one is, the more one has been exposed to it. More
importantly, with empirical data, we have questioned the fact that the variety is
reserved to Maghrebi youth.

4. 3. 3. 5 Second ethnicity

Table 13: Weight of second ethnicity

 

 

Weight percentage
Second Ethnicity
African 0.712 74
Polish 0.674 74
Spanish 0.520 71
Portuguese 0.498 76
Maghrebi 0.481 73
German 0.256 67
South/Central 0.093 62
American
Range 0.619

While first ethnicity was not significant in terms of knowledge of the lexical items
in the questionnaire, second ethnicity was, and at 0.619, is a fairly strong factor
for knowledge. The highest promoter was African, which did not include the

Maghreb. At 0.712, it is a strong weight. The next promoter in terms of

162

knowledge was Polish. Spanish and Portuguese followed closely, though the
former was a weak promoter and the latter was relatively neutral. More
surprisingly was Maghrebi, as a slight demoter. German was a strong demoter
at 0.256, and Central/South American was a very strong demoter at 0.093.

Possible analysis of these results may also be that those of Maghrebi
heritage, people from an ethnicity which claims banlieue language (Doran 2002,
Melliani 2000), are innovators. Perhaps the reason they are demoters with the
current word list is that they have continued to change the language rather than
use the vocabulary that is also used by an older generation. Creole vocabulary
also may have affected these results. The younger generations are clearly
unaffected by Creole's influence and may therefore have chosen other linguistic
contributors such as Arabic or Roma as the basis for newer lexicon in banlieue
language. Those of Polish heritage, as well as those of Spanish and Portuguese
heritage, are more likely to have been quickly assimilated into French
mainstream culture, as historically they have done (Carpin, Tavan, and Dugué
2004); and therefore Tejedor de Felipe (2004) may have been correct in his
assertion that banlieue language is indeed a French phenomenon. Cell paucity
may have been a factor since the Polish heritage is unique to one informant. It
does, however, allow us to underscore the need for studies involving more
ethnicities.

The African second ethnicity was the highest promoter. l have not seen
research specifically on this ethnic group in the banlieue, though many of the

recent banlieue language authors are African. It may well be due to cell paucity.

163

4. 3. 3. 6 Religion

Table 14: Weight of religion

 

 

 

weight percentage
Religion
Non- practicing 0.516 72
Practicing 0.458 64
Range 0.058

Religion was very complicated. It was apparent that no matter which religion
was claimed, the weight of the factors was similar enough to combine them and
simply look at whether a person claimed to practice the religion or not.
Therefore, religious categories were eliminated.

In terms of knowledge, practicing a religion, any religion, was a demoting
factor, while atheists and those who did not practice their religion tended to know
and use more of the lexical items. One possible explanation for this is that since
banlieue language is in-group/out-group, those who practiced their religion were
part of another group, that of their church, synagogue, temple, mosque, or other
place of worship. Practicing a religion is often group related. It is also possible
that the linguistic components of banlieue language that are sexual and illegal in
nature would be less attractive to someone who practices his or her religion. It
is, however, a small range, so further investigation is needed.

4. 3. 3. 7 Word origin

Table 15: Weight of word on'gin

 

 

weight percentage
Word Origin
Verlan 0.769 86
Argot 0.644 78
164

Table 15 (cont’d)

 

Arabic 0.568 72

Roma 0.492 66

g Creole 0.209 38
Range 0.560

The origin of the lexical item was important as seen by the range of 0.560.
Verlan was far and away the highest promoter, with argot following closely.
Arabic is a weaker promoter, although Roma is a weak demoter. As we see with
Creole, it is the largest demoter of the group. Much of this follows Tejedor de
Felipe’s (2004) corpus study analysis, which underlines that this may indeed be
a French phenomenon.

This is different from Table 7 in that we would expect Roma to still be a
promoter, since non-answers to this category only amount to 35.5%. The rest of
the languages in the section on significance in terms of knowledge coincide with
what we would expect from the first table. Verlan, argot, Arabic, Roma and
Creole are known proportionally in that order by the informants.

4. 3. 4 Statistically significant factors for admitted usage

In this section, I examine the results for admitted usage. In this case, the
informants who admitted to using a lexical item ‘rarely’ or ‘often’ were
consolidated, and those who either said that they never used the lexical item in
question or did not know the vocabulary were counted as not using it.

I would argue that while knowledge is an important factor, admitted usage
is much more important. In this section, I am no longer investigating what
socioeconomic factors promote or demote knowledge of the lexical items, but

rather which of those factors promote or demote usage.

165

Bo
hrknovv
exanghe
conduct
434.1l

H
hoissi
wthzac

ornott

Secor
bonni
43xl
Beth

those

Tabk

ll

hst

Both sections are useful because we see that sometimes a large promoter
for knowledge is not a promoter for usage, which can be tied to my earlier
example of ‘wazzup.’ Most studies on non-prestige varieties have been
conducted using either admitted usage or attested usage.11

4. 3. 4. 1 Factors eliminated
If a socioeconomic factor is not deemed significant in terms of knowledge
but is significant in terms of usage, the only logical conclusion is that participants
With a certain socioeconomic factor (gender, age group, etc.) have chosen to use
or not to use banlieue vocabulary.
Only two factors were not selected as significant with admitted usage.

Second ethnicity was eliminated as well as whether or not an informant was

born in France.
4. 3.4.2 Factors retained
Below is a discussion of each factor significant in terms of usage, particularly

those factors that promote or demote admitted usage of banlieue language.

4. 3. 4. 2. 1 Gender

Table 16: Weight of gender

 

 

 

weight percengge
Gender
Male 0.654 49
Female 0.394 34
Range 0.260

 

” With the exception of Boyer's (2001) study in which students created a word
list.

166

Importantly, while gender was not significant in terms of knowledge, it was
indeed significant in terms of admitted (self-reported) usage, with the predictable
results of women being demoters and men being promoters of the usage of
banlieue language. Since banlieue language is a language of covert prestige,
these results are somewhat in accordance with all the previous research and
assumptions. However, it is also important to look at the range. At 0.260, gender
is not as strong a factor as those that follow.

The results in 4.1 and 4.2 where informants attributed equal usage to men
and women at between 60% and 75%, depending upon the language of origin
and the meaning of the lexical item, seem to contradict the results presented
here. However, if we compared simply usage attributed to men versus usage
attributed to women, we saw that, indeed, men were viewed as speaking the
variety more than women. Additionally, the results of section 4.2 support these
findings, since men and women attributed usage to both genders at 69.38% and
75% respectively, their admitted usage did not match that at 77.35% and
55.94% respectively. It may be, therefore, that gender is not as polarizing a
factor in the banlieue and that the importance of identification with the area and
its struggle may supersede gender.

4. 3.4. 2.2 Age

Table 17: Weight of age

 

 

weight percentage
Age
25-29 0.653 40
30-36 0.567 53
20-24 0.528 41
18and 19 0.194 26

 

167

Range 0.459

Age is another factor that was not selected as significant in terms of
knowledge but was selected as significant in terms of admitted usage. What is
most interesting is that, again, what we would expect—younger people using the
lexical items the most and the older people using the vocabulary the least—is
not the case.

In fact, the age group that most strongly tends to use banlieue language is
the 25- to 29-year-old group. The oldest group, that of the 30- to 36-year-olds,
was the second-highest in terms of promoting, followed closely by the 20- to 24-
year-old group. Interestingly enough, the 18- and 19-year-old group was a very
strong demoter with a weight if 0.194. We saw similar results in section 4.1 and
section 4.2. The range is fairly strong at 0.459, which attests to the strength of
the factor.

There could be several reasons for this seeming inversion in terms of age
and admitted usage of banlieue language. It is possible that the words are
indeed that of an older generation and being phased out, and that new
generations are recreating the language and using new vocabulary. In fact, the
idea of aging vocabulary may be supported by the fact that the only informants
who chose ‘an older generation’ for the lexical item bédo were three young,
Franco-Maghrebi males. This idea of changing vocabulary would perhaps help
explain why there was no statistical difference among the age groups in terms of

knowledge, but that the younger groups choose not to use, or at least not to

168

admit to using the vocabulary. Calvet (1996) and Mela (2000) both discuss the
changes in banlieue language that have occurred since their last articles on the
subject and emphasize the frequently changing nature of banlieue language,
and Séguin and Teillard's (1996) informant was quoted as saying that they
would change the language if it became too well known.

Another hypothesis is a type of age grading for non-reference features.
The oldest group is the most well established, employed and living on their own.
The youngest group is almost all at the university, or at the Mission Locale
where they are not well established and do not have employment. I would like to
propose that the older group is more relaxed or more secure about admitting to
their usage. In the university setting, some of Doran’s (2002) theory about the
choice to leave behind Verlan or banlieue language during college because they
are all on the educational track may also be a possible explanation for the
relative lack of usage among the group at the university.

4.3.4.2.3 Education

Table 18: Weight of education

 

 

 

weight percentage

Education
BTS 0.894 76
None 0.765 67
DEUG 0.655 49
BEP 0.522 39
DEA/DESS 0.480 35
Bac 0.477 38
License 0.476 35
Maitrise 0.241 32

Range 0.653

169

 

As we see with the fairly large range of 0.653, education is indeed a very
important factor. We have seen a difference between what informants say about
those who use the lexical items in question and whether or not they admit to
using the vocabulary themselves in 4.1 and 4.2. The biggest admitted user of
the banlieue language items in question is the BTS group. This group, again, is
either that of a technical trade or those who will be going into engineering
schools. There is a very strong reclamation of using these lexical items. This
group is followed closely by the group with no formal education. However, what
is surprising is that the DEUG is also a strong promoter. In fact, as we examine
all categories of education, we see that the only strong demoter is the maitn'se.
The other demoters are not strong, because their weights are too close to 0.5.

In analyzing the information, it is possible that the BTS represents such
disparate results for two reasons. The first is cell paucity, since there were only
two informants with a BTS. The other reason may be their unique status in the
French educational system. The holder of a BTS alone is educated formally in a
technical subject, so he or she may be prone to reject the notion of banlieue
vocabulary as used by the uneducated. At the same time, his or her job would
not have the same status as that of an engineer. The pressure to avoid use at
work might be lower than for that of a more formally educated person.

Those with a Maitrise in this study are concentrated at the university and
may be geared toward more academic positions in which banlieue language,
because of its covert prestige, would not be as welcome. This would not,

however, be the case for everyone with a Maitrise; and as we see, people with

170

even hig
items. It
themsel
peers. l
banlieue
lead the
heme

4

Table II

~

Res

R3797:

import
Promo
and th.
Which
be the
Usage

Outsid,

even higher levels of education (DEA/DESS) are admitting to using the lexical
items. It is therefore possible that the informants who do not admit to using it
themselves are either underreporting their usage or observe it only among their
peers. However, since most informants, regardless of education, admit to using
banlieue language, there is only one real demoter: the Maitrise. All of this would
lead the author to believe that education is not as big a factor as research and
the media would have us believe, but that other factors are at work.

4.3. 4. 2.4 Department of residence

Table 19: Weight of department of residence

 

 

weight percentage
Residential history
Seine-et-Marne 0.879 55
Hauts-de-Seine 0.633 38
Seine-Saint—Denis 0.527 41
Paris 0.499 30
Val d'Oise 0.455 42
Essonne 0.398 27

 

Range 0.481

By the range of 0.481, we see that the department of residence is a fairly
important factor. A cursory glance gives us Seine-et-Marne as the largest
promoter, followed by Hauts—de-Seine. Seine-Saint-Denis is a weak promoter
and the Val d'Oise, contrary to all supposed logic, is a demoter, along with Paris,
which is really neutral at 0.499. As with knowledge of the vocabulary, it may well
be the effect of the Mission Locale, where the many informants did not admit to
usage. In addition, it seems that those who admit to using the vocabulary are

outside of what one would think of as the ‘epicenter.’ I suggest that two factors

171

are at wort
normally b
instead it i
has sprea
perhaps t
to keep it

language

4.:

Table 20:

 

are at work. First of all, since the supposed ‘epicenter’ of the vocabulary would
normally be Seine-Saint-Denis and the Val d'Oise, results that indicate that
instead it is the surrounding areas that promote suggests that the vocabulary
has spread and more segments of society are using it, and at the same time
perhaps those who are at the center are innovating and creating new vocabulary
to keep themselves from being understood, as is often the case with secret
languages.

4. 3. 4. 2.5 Ethnicity

Table 20: Weight of ethnicity

 

 

 

weight percentage

Ethnicity
Italian 0.978 71
DOM/T OM 0.863 33
Maghrebi 0.554 38
African 0.478 40
French 0.406 45
Portuguese 0.406 10
Eastern European 0.182 1 0

Range 0.796

The factor of ethnicity is very important, given its large range of 0.796. It is
important to remind the reader, however, that this category is different from
second ethnicity, which was selected as significant in terms of knowledge. This
category of first ethnicity was not selected as significant in terms of knowledge
and therefore we can not compare them equally, but rather can only discuss
them individually.

We see that Italian was a very strong promoter in terms of admitted

usage, and I take the time to remind the reader that it was a knock-out in terms

172

of kn
rangi
infor
whic
inter
edu
Oihi
the
usi
na“

me

be

of knowledge. This means that while the category has an extremely important
range, cell paucity is indeed a factor, and the results are based on one
informant. Other promoting ethnicities are those informants from the DOM/T OM,
which include the Caribbean and many of the islands. They are extremely well
integrated into mainland French society, and, as previously mentioned, are
educated in French since they are actual départements with equal status of any
other—such as Paris—or they are territories. It would seem logical that since
they are French, informants from that area would be linguistically similar, by
using a linguistic phenomenon that my results show to be primarily French in
nature. In addition, many people from the DOM/TOM-are black, which still
makes them a minority in France, and in such a case, those who live in the
banlieue would be likely to cling to the identity of the minority struggle in the
banlieue. People from these areas may be French, but they are not descendants
of the Gauls. They are therefore French by birth or on paper, but minority due to
the color of their skin.

The last promoter is those of Maghrebi heritage, but as we see, it is one of
the weaker promoters. With the research focused primarily on those of Maghrebi
heritage, we might have expected the weight to be higher. However, as l have
indicated before, it is possible that they have innovated and that their current
vocabulary is not included in this study.

The African ethnicity is a weak demoter, almost neutral. French and
Portuguese are identical in weights and are also weak demoters. The rate of

assimilation of those from Portugal is quite high and quite fast (Carpin, Tavan.

173

 

and Dugué 2005), and we see that their weight of usage is identical to the
French. Eastern European, again, is the group least quick to assimilate because
most arrived more recently than others (Carpin, Tavan, and Dugué 2005), and
since this is a hypothesized French phenomenon, one could postulate that they
have not been able to integrate into the banlieue culture in order to pick up and
use banlieue language, or more importantly, perhaps they do not identify with
the banlieue. These informants were all from the university and Mission Locale.
4. 3.4. 2. 6 Religion

Table 21: Weight of religious affiliation

 

 

 

 

weight percentage
Religion
Protestant, non- 0.924 81
practicing
Atheist 0.583 33
Muslim, non- 0.566 51
practicing
Catholic, practicing 0.534 37
Muslim, practicing 0.455 39
Catholic, non- 0.416 51
practicing
Christian, non- 0.415 27
practicing
Kabyle, non- 0.266 38
practicing
Range 0.658

Religion was chosen as significant in both knowledge and usage and
indeed, with a range of 0.658, we see religion is an important factor. The results
of religion, however, are not as simple with usage as they were with knowledge.
Here the highest promoters are indeed non-practicing, but there are also some

non-practicing informants who are demoters.

174

It is
demotes
are all m
holds tn
importai
inlorma
Chhsha
are net

sell-se
necess
answe
quesh
four C

CE“ p;

Table

It is interesting to see that Muslims pattern with the original ‘practicing
demotes and non-practicing promotes’ hypothesis, and the top three promoters
are all non-practicing; but I did not have enough informants to see if the rule
holds true for practicing protestants, or those practicing the Kabyle faith. It is
important to note here again as a reminder that I did not place the labels on the
informants. They were all self-applied titles. One could argue that non-practicing
Christian is non-practicing Catholic, especially in France, and as we see, they
are nearly identical in weights; but I did not want to collapse them, as they were
self-selected labels. What these results do say is that more investigation is
necessary. Thirty-one of the total number of informants, or 46.6%, chose not to
answer the question on religion. Among the informants who did answer the
question on religion, there were 17 Muslim informants, 13 Catholic informants,
four Christians, and one each of Protestant and Kabyle. Therefore we know that
cell paucity was a factor in both the highest promoter and the largest demoter.

4. 3. 4. 2. 7 Languages of informants

Table 22: Weight of infonnants' language

 

 

 

weight percentgge
Languages
Portuguese 0.733 38
Scholastic 0.557 41
Spanish 0.500 37
Kabyle 0.477 33
Arabic 0.468 39
Polish 0.274 38
Table 22 (cont’d)
Irish 0.182 43
Creole 0.066 24
Range 0.667

175

 

assi
thOL
(Te.
oth
dei

inr

0T

Ta

Portuguese is the highest promoter, perhaps due to the more rapid rate of
assimilation. Those who speak a scholastic language are considered promoters,
though weak. This would follow the theory that this is a French phenomenon
(T ejedor de Felipe 2004), since those speaking scholastic languages and no
others would tend to be French citizens. Arabic and Kabyl are very weak
demoters. This may be due to the idea of the vocabulary aging and subsequent
innovation by younger speakers, which has been raised earlier. As to Creole,
Irish, and Polish, cell paucity may very well have had an effect. There was only

one informant of each of the last two languages. ‘

 

4.3. 4. 2.8 Language of the lexical item.

Table 23: Weight of word origin

 

 

 

weight percentage
Word Origin
Verlan 0.664 54
Argot 0.636 51
Arabic 0.595 48
Roma 0.490 39
Creole 0.247 20
Range 0.417

Once again, we see the same frequency ranking as with knowledge.
Verlan is most used, though not much of the corpus, followed closely by simple
argot. Arabic is also a promoter. Roma, though almost neutral, is on the
demoting side, and Creole is by far the strongest demoter. This just confirms
Tejedor de Felipe's (2004) claim that banlieue language may, indeed, be a

French phenomenon.

176

4.4 Analysis of statistically significant factors

In sections 4.1 and 4.2, we saw which factors were selected by Goldvarb
as statistically significant and their weights in importance in terms of knowledge.

One of the more interesting results is that age and gender are not
significant in terms of knowledge of the word, but they are significant in terms of
usage. With gender, we can argue that even in terms of usage, it is one of the
weaker factors with a narrow range, and that it follows the data from sections 4.1
and 4.2. It may be that identification with the banlieue is more important than
gender

Age is significant, but the pattern is not as we might have expected based
on previous research. The older groups are actually promoters, which is contrary
to almost everything that has been published. This can be taken to mean that
the variety is being retained to some degree (and as we see in the 4.2.3.2, at
almost 75%) among older speakers. We see that this also follows the results
from 4.1 and 4.2. It may well be that the younger generations did not want to
admit to usage due to venue; or it may be that many at the university venue,
where most younger informants are not from the banlieue, do not identify with
minority struggle. We can say that even the youngest group shows that the
variety is retained to a larger extent than has ever been empirically attested.
This emphatically contradicts Lepoutre (1997). It may also be simply that the
younger generation is inventing new vocabulary and does not want to be

associated with this older vocabulary.

177

US

ad

IE

We see that area of residence is a significant factor in both knowledge and
usage, as some of the qualitative remarks would suggest. The results in
admitted usage seem counter-intuitive, since val d'Oise, a supposed epicenter,
is a weak demoter. However, I hypothesize that the word list given may not
reflect innovations from the area and that in the Mission Locale, informants may
have been loathe to admit to usage—though they were informed that all
questionnaires were anonymous—since banlieue language is not appropriate in
the professional world of France.

Nationality and birthplace are only significant in terms of knowledge, not in
terms of usage. It is interesting that being born in France has a strong influence
on whether or not a person will be familiar with the vocabulary of banlieue
language but not whether or not an informant is inclined to use it. This would go
against Tejedor de Felipe's (2004) hypothesis that banlieue language is a
French phenomenon.

Second ethnicity is very important in knowledge, but first ethnicity
influences whether or not an informant will actually use the vocabulary, which
may have to do with rate of assimilation to French culture, if Tejedor de Felipe's
(2004) hypothesis is correct. More investigation is necessary to separate these
factors more clearly.

Religion is significant in both knowledge and usage, which is what we
might expect from the sociological studies in F rance—a place where religion is
more important than socioeconomic status (Brulé 1966, Michelat and Simon

1977). For usage, we see that internal factors cannot be simplified to whether or

178

not an informant practices a religion, though the top three promoters are those
who do not practice one. This may reflect either an informant who practices a
religion being reluctant to embrace a linguistic variety in which so many
scatological and illicit references are a focal point, or that those informants
identify first and foremost with their religious group.

With languages, we see that word origin is always a factor but that the
languages a person speaks only affect the usage of banlieue vocabulary. Creole
is such a demoter in both knowledge and usage that in future studies this
language should probably be eschewed. We see also that French is the most
popular source of banlieue language, which would support Tejedor de Felipe's
(2004) hypothesis that banlieue language is a French phenomenon. Both Verlan
and old argot were the most popular in terms of knowledge and usage. As to the
languages spoken by the informant, we see that the scholastic languages as
well as Portuguese and Spanish were the only promoters, which, given the quick
rates of assimilation of those two cultures (Carpin, Tavan, and Dugué 2005),
may again support the hypothesis that banlieue language is a French-based
phenomenon.

To summarize, there are two major categories of important factors. The
first category consists of those factors that are not significant in terms of
knowledge—meaning that all parties know the vocabulary more or less
equally—but that are, however, significant in terms of admitted usage. These
factors are gender, age, and languages spoken. This means that women do not

have a significantly lower knowledge of the vocabulary but admit to using it less

179

 

th

than men do at a statistically significant weight. This follows what we might
expect from a covert-prestige variety. Age is also only significant in terms of
usage, but as we have discussed, it may be due to innovation on the part of the
younger generations, who may be choosing to use newer lexicon; or it may be
due to the overlap of venue in which those at the university may not identify with
banlieue culture. With languages, those that are scholastic or quickly assimilated
(Portuguese and Spanish) were the highest promoters, which underscores
Felipe de Tejedor's (2004) assertion that banlieue language is a French
phenomenon. The second category of important factors includes the factors that
were significant for both knowledge and usage. These include place of
residence, religion, education, and word origin.

The factors that are the significant in both knowledge and usage are more
what we might expect as researchers. It seems to stand to reason that if a factor
is significant in terms of knowledge (whether or not a person knows the items), it
would also be significant in terms of usage. You cannot use what you do not

know.

180

orig

infr

kr

CHAPTER 5: CONCLUSION

5.1 Study synopsis

In this dissertation, I studied 21 lexical items from five different linguistic
origins, all of which are reported to be used in the banlieues of Paris. This
information was given to me by an informant in the banlieue, who at the time
was very immersed in the culture. In order to study the factors influencing the
knowledge and usage of these 21 lexical items, 69 informants completed a
questionnaire. The questionnaire, which consisted of three sections, asked first
for the informant's personal and socioeconomic information. lnfonnants were
then asked if they knew each of the 21 lexical items, if they used them, and
finally their perceptions of the type of person who would use the lexical items—
i.e. the perceived educational level, gender and age of someone who would use
the vocabulary.

Results were analyzed in three ways. First, each individual lexical item
was investigated in terms of the gender of perceived users, their educational
level and their age group or generation. Secondly, the results were analyzed by
the major categories of gender, education, and age. In gender, the data was
compiled first by the what the lnfonnants as a group felt was the gender of
someone who would use the vocabulary and then by what the informants in the
study admitted to using themselves, analyzing females and males separately.
Results were then compared in order to investigate discrepancies between the
infonnants' perceptions of who would use the vocabulary in question and

whether or not they themselves admitted to using the vocabulary. The same was

181

don
a lo

bot

Th
hi

di

done for education and age. Finally, the socioeconomic factors were loaded into
a logistic regression program to see which factors were statistically significant for

both knowledge of the lexical items and admitted usage.

5.2 General commentary
The focus of this study was to open more dialogue on banlieue language. I
hypothesized that studies on banlieue language needed to be viewed from
different angles than they had been previously, specifically in terms of age,
education, and gender.

In terms of age, I believe that the statistics in this dissertation have
successfully challenged, with empirical data, the original belief that banlieue
language was a youth language (Boyer 1997, Doran 2002, George 1986,
Lepoutre 1997), and opened up the matter for further study. As I mentioned,
some scholars had already initiated the conversation on age (Calvet 2006,
Zerling 1999), but neither had published a study specifically pertaining to age
with empirical data. Banlieue language may be at the very least changing from
its reported static ‘youth’ status to something more. As the speakers grow older,
some aspects of the language seem to be retained, as Doran (2002) and Calvet
(2006) have suggested. Words like daron, keum, kiffer, schmidts, and marave
were indicated in the Séguin and Teillard (1996) study, but they were still known
and used by some of the older speakers in this study. It is yet to be determined if

those speakers pass anything on to their children, and to what extent.

182

It is important to remember that the informant who created the list was
older, in his early 30s, which may mean that the vocabulary does not reflect
what is currently used by younger generations. This is especially true of the
youngest students in the study, who frequently signaled that lexical items
belonged to an older generation. What this study does underscore, however, is
that the older generation is not abandoning this vocabulary completely as it
ages.

In terms of gender, we saw that while the results adhere somewhat to
what one would expect, women speak the variety less than men do; the results
showed that both genders admit to usage at over 50%. In addition, we saw that
both genders perceived equal usage at about 70%. This may suggest that
identification with the area may be more important to these speakers than
gender in a non-prestige variety. More research should be done to see if the
younger generations are also showing this solidarity with new banlieue language
vocabulary, or even specifically Verlan innovations.

Education results were similar to those of age and gender. The informants
did not attribute this variety primarily to those with little formal education. Most
vocabulary items were viewed as being used by those who were somewhat
educated, with very few exceptions. These exceptions included the lexical items
connoting the physically violent act of ‘beating someone up.’ Other exceptions
included those lexical items that may well have been viewed as foreign tokens,
such as ouaiche and the greetings in Creole. These items were viewed either as

being from all levels of education or from the higher levels. Finally, words such

183

as bédo and bédave were also viewed ovenrvhelmingly as being used by people
of all educational backgrounds.

While the informants viewed the lexical items as not being predominantly
the language of those who have little formal education, their admitted usage was
indeed somewhat affected by their own level of education. The group that has
the highest percentage of ‘I never use this lexical item’ is the ‘Maitn'se and
higher’ group; and, inversely, this group also has the lowest admitted usage.
Most groups (BEP, BTS, DEUG/License) were over 60% for admitted usage.
The ‘Maitn'se and higher’ group admitted to 56% usage, however. These results
should suggest that we should challenge the belief that this variety is for those
who are uneducated (Lepoutre 1997). They may even imply that perhaps this
variety is coming to be seen as a badge of identity, similar to the situations of
Sheng and Camfranglais, and that socioeconomic factors that would usually
diminish the number of speakers are less important than the informant's sense
of identity with the region. This hypothesis is consistent with Labov's (1972‘)

Martha's Vineyard study.

5.3 Factors explaining knowledge and usage
5. 3.1 Word on'gin
We have seen that the origin of the word is very important. The results for
Creole stood out from those of the other languages as the least known.
Additionally, the lexical items derived from French, whether from old slang/argot

or from syllable reversal were not only the most known but the highest

184

promoters in terms of usage. Arabic and Roma followed quite closely in terms of
popularity. Those results would tend to support Tejedor de Felipe's (2004)

assertion that banlieue language is a French phenomenon.

5. 3.2 Social factors
5.3.2.1 Gender
Gender is a significant factor for usage, with males promoting and females
demoting (following the stereotype), but we have seen that women do admit to
using the vocabulary—and at a much larger degree than previously thought.
While the overwhelming majority of both males and females report that for most
lexical items both males and females use them equally, their own admissions for
usage contradict this. The women in the study admitted to using the banlieue
language items less than men and much less than what they attributed as usage
to women or both men and women equally.
5. 3. 2.2 Age
Age results had only the youngest group as a very strong demoter. The
other groups all promoted to some extent, with the 25- to 29-year-olds as the
highest promoters, followed by the oldest group. This may indicate that the
youngest group is over conscious about using the vocabulary since they are
mostly at the university venue, which would be supported by the fact that most of
the responses with ‘a younger generation’ were from the youngest two
categories. It may also indicate that some of the lexical items in general are part

of the lexicon of the older group and their siblings, and that a new lexicon

185

bdong

tousu

usagr
adnn
shit
Iheh
lexh

53.

tho
hh
the

EC

belongs to the younger generation. Regardless, however, the informants admit
to using the vocabulary at over 50%.

It is also important to remind the reader that all groups admitted some
usage of the lexical items, and even in the lowest group, 23.5% of informants
admitted to using lexical items often and 22.7% admitted to using them once in a
while (rarely). In addition, though the youngest group members were the least
likely to admit to usage, they were very quick to indicate that, indeed, most of the
lexical items were part of their generation.

5. 3. 2. 3 Education

Education results were also contrary to the stereotype that allows only for
those with very little education to know and use the lexicon. The DEA, one of the
highest levels of education, is the largest promoter in terms of knowledge; and at
the opposite end of the spectrum, the BEP, the lowest level with ‘some
education,’ is the largest demoter. We see that only the BTS and the license are
demoters, since the BTS and Maitrise are neutral. What this means in terms of
knowledge is that education does not have any bearing on whether or not an
informant will be exposed to banlieue language, and indeed knowing and using
this vocabulary is not indicative of scholastic failure.

When we look at the same theme in terms of admitted usage, we see a
different story. While there is still a certain perceived randomness in that the
BTS, no education, and the DEUG are all strong promoters, we do see that the
higher levels of education become slight demoters, with the Maitrise being a

very strong demoter. What this shows us is that even though the users have the

186

knowledge, some of the language is left behind, just as in normal age grading
when the speakers reach a level of professionalism that requires a different
vocabulary. This has been indicated by scholars, but at a very different age level
(Doran 2002, Lepoutre 1997). One of my informants told me that he was not
aware of how much he used banlieue language until he was taking advanced
math classes in the Maitrise at a local university. It was there that he realized,
through peer pressure, that banlieue language was not as appropriate at that
venue. When with his friends, however, he does indeed use the vocabulary. It is
important to notice, however, the fact that most higher levels of education are
still fairly neutral in terms of demotion. And, indeed, if we look at admitted usage,
55% of this group still admitted to using some of the lexical items in question.
5. 3. 2. 4 Residence

Area of residence is a factor that was significant on both knowledge and
usage. Obviously only areas that in theory had already been exposed to
banlieue language were questioned. Of those areas, in terms of knowledge, we
see that Paris is the highest in terms of knowledge, followed closely by Seine-
Saint-Denis and Marne. The Val d’Oise is a slight demoter in the group, and
Essonne is a very strong demoter. In terms of admitted usage, however, many
of these areas could not be collapsed as they were in the first category.

When we look at admitted usage, location of the département may affect
admitted usage. Those départements furthest from Seine-Saint—Denis and Val
d'Oise seem to admit to it the least. The factor of identification with the area in

which the variety is spoken seems to be extremely important. It may be that

187

those who live furthest from Seine-Saint-Denis and Val d'Oise do not identify
themselves with that area and therefore would be less likely to want to break
social norms by using banlieue language. Additionally, since I questioned people
who were older than the rest of the participants, it may be that those speakers
who lived furthest from where the variety was spoken when they were young
picked up the fewest lexical items.

Being born in France is significant in terms of knowing the lexical items,
though not significant in terms of usage. Firstly, banlieue language seems to be
a French phenomenon as Tejedor de Felipe (2004) suggested, and being born
in France would facilitate knowledge of this vocabulary simply because peers
and older siblings would be using it. However, it would seem that the factor of
identification with the banlieue might also be important. Even if an immigrant
came to France as a young child, the peer pressure toward usage of the
vocabulary and identification with his or her environment would outweigh being
born outside of the Hexagon.

5. 3. 2. 5 Birthplace, nationality, and ethnicity

Tied to being born in France are issues of nationality and ethnicity. We
see that for nationality and knowledge, the only promoter is a French nationality.
Once French nationality is coupled with another ethnicity, the factor becomes a
demoter. Ironically, Eastern European nationalities and Maghrebi nationalities
were strong demoters. Much of this can be linked to Tejedor de Felipe's (2004)

study and his conclusion that banlieue language is indeed a French

188

phenomenon. Interestingly, nationality is not significant in terms of actual,
admitted usage.

A second ethnicity was significant in terms of knowledge, with Italians
being a knockout, explained by the fact that there was only one Italian and he
knew all the lexical items. Africans were then the second highest promoters,
followed by the Polish. All other second ethnicities were demoters, with Eastern
Europeans being the largest demoters. In terms of admitted usage, the Italian is
again the largest promoter, followed by those in the DOMTOMs. This may well
have to do with the fact that the Italian in question is very well integrated into the
banlieue, and while his second ethnicity is Italian, he is French by culture. As we
saw in the first place with knowledge, a double ethnicity was a demoter; so it
stands to reason that in terms of actual usage, French as a second ethnicity
would also be a demoter. The same thing rings true for those Eastern European
ethnicity, since they are one of the least integrated minorities (INSEE 2005) in
France. The maghrebi ethnicity is a promoter in this category, which may well
come from the wealth of Arabic terms in the questionnaire, as well as the Verlan
and Argot.

5. 3. 2. 6 Religion

Religion affected both knowledge and usage. In knowledge, those who
practiced their religions, regardless of which religion they practiced, were less
likely to know the banlieue language lexical items. The opposite was also true—
not practicing a religion was a promoting factor. One could argue that a similar

effect is shown by admitted usage as well in that the highest promoters are

189

indeed non-practicing. However, we see that Kabyle, Christians, and Catholics
who are non-practicing are demoters, while practicing Catholicism is a promoter.
This may well indicate that, again, it is a mostly French phenomenon, as
Catholicism is one of the main religions in France. While the results in this
category are far from conclusive, they may suggest that further research should
be done, targeting religion as a factor. One might, for example, collect
informants at several places of worship.
5. 3. 2. 7 Languages

The languages spoken by the informants were not significant in terms of
knowledge, but they were significant in terms of usage. We see that the biggest
promoter is Portuguese, which is a culture with a very high level of integration
and a very quick level of assimilation into French culture (INSEE 2005).
Scholastic languages are next, which would indicate someone who is French, so
again, a factor that is reinforced throughout the significance tables is Tejedor de
Felipe’s (2004) suggestion that this is indeed a French phenomenon. Spanish
seems to have a discrepancy between knowledge and usage, but this apparent
discrepancy is due partly to the fact that Central and South America were very
high demoters in the knowledge category. Speaking Arabic is a slight demoter,
which may come from being less integrated. Creole speakers are the least likely

to use banlieue language.

190

5.4 Conclusions
The most important conclusion of this dissertation is that we must, as
researchers, begin to examine more fully the established stereotypes of the
speakers of banlieue language. More empirical research should be done on this
group, especially in the domain of age, gender, and educational level of the
speaker. We must begin to entertain the intertwined hypotheses that this variety
may be gaining momentum, and that banlieue language may indeed be a
French phenomenon (T ejedor de Felipe 2004), not just born out of immigrants in
the banlieue but rather born out of the diversity of peoples in the banlieue who
call themselves French.

Banlieue language does not seem to be tied to one ethnic group or origin,
but it seems to be used among those who have assimilated to some extent into
French culture, though they may still be very marginalized politically and/ or
socially. Peer groups in these areas seem to be taking banlieue language with

them as they age, regardless of education level and employment.

5.5 Limitations
The greatest limitation to this study was the paucity of informants. With 67
informants, it would be impossible to fill all the cells relating to personal data,
such as sex, religion, languages spoken, age and educational level to make as
detailed a comparison as I would have liked. The logistical regression program
helps in that respect, but a more comprehensive informant base could be

instrumental in backing up these results.

191

Ott
questioni
small nu
represei

study in

I
and dc
secon
they I
intere
grout

banl‘

com

are
the
adi
c0

0T

0’

Other limitations include the use of terms that have aged based on the
questionnaire developer, insufficient inclusion on Verlan lexical items, the fairly
small number of lexical items included, which may have skewed the
representativeness of the list. Finally, though I had three different venues, the

study may not be representative due to the overlap in the settings in the venues.

5.6 Areas for further study

It would be interesting to go back to the same generations of informants
and do a similar format that tested simply Verlan knowledge and usage with a
second portion of the questionnaire asking the lnfonnants to elaborate on when
they use the lexical items. There is obviously a limit to even this, but it would be
interesting to see what qualitative data could be gleaned from this in an older
group. It is very hard to witness banlieue language in a public setting since
banlieue language is, as established, primarily and in-group form of
communication.

It is important to note again that the informants were all residents of the
area in which banlieue language is spoken. Since we saw that the further from
the geographic epicenter of banlieue language we were, the less the informants
admitted to knowing and using, it would be interesting to conduct the study in
concentric circles that generate further and further from the 93 and the 95 in
order to compare results more clearly.

Another possible area of further study would be to take recordings of

spontaneous conversations of a group of employed individuals over the age of

192

25 ar
langi
resu‘

USBC

not
art
to t
the
chi
va

inl

25 and do a corpus-style study on the transcripts, quantifying the banlieue
language items used during the tapings and then statistically analyzing the
results. In this manner, one would be able to discern if indeed lexical items were
used spontaneously in an older group.

Additionally, I would like to observe couples with small children who have
not yet gone to school (so that I can be sure that they have not yet interacted
extensively with peers) to see if banlieue language lexicon has been passed on
to them by their parents. I would not expect the parents to teach it directly to
their children, but even among their own peers, if banlieue language is used, the
children are likely to acquire it in some measure. If the parents who know the
variety are not using it with their children, that information would also be very
interesting because it would mean that said parents were monitoring their
language in order that their children would not pick it up. This type of behavior
would relegate banlieue language to the same level as scatological language or
swearing. It would also be beneficial to study certain language groups in a more
in-depth fashion. I would like to redo the questionnaire for Verlan by itself to see
at what level it is retained over a large informant base. It would also be
interesting to see if the perceptions of banlieue language are changing among
older residents.

Finally, a true perceptual study in which a scale is used by informants to
indicate their linguistic attitudes—be they positive or negative—toward the
variety would be informative. Judging from the responses regarding the

educational level of those who would use the vocabulary, in which few

 

informants labeled the lexical items as belonging to people with a low level of
education, it seems as though the perceptions are at least somewhat positive. I
believe we have made the case, therefore, for a perceptual study in which
informants are asked directly if speaking banlieue language is viewed positively
or negatively. Those eventual results could then be compared to what seems to
have been indicated here, that the stigma for those who speak it is not that

exaggerated.

194

 

Appendix .

1 Questio
Veuillez n

Age_
Quel est
Cl

B:

D

Licer
Mail!
Doct
Auu
Autr
Depuis
Eslce .
Estce
Est-Ce
Esme
non
Est

    

R
E

 

 

Appendix A
1 Questionnaire
Veuillez répondre aux questions suivantes :

Age Sexe
Quel est votre plus haut niveau d’éducation?
CAP/BEP/ diplome professionnel
Bac
DEUG
Licence
Maitrise
Doctorat
Aucun
Autre :
Depuis combien d’années est-ce que vous habitez Garges ?
Est-ce que vos parents habitent Garges aussi ? oui non
Est-ce que vous avez des freres et soeurs qui habitent Garges? oui non
Est-ce que vous avez des oncles et tantes qui habitent Garges ? oui non
Est-ce que vous avez des grands-parents qui habitent/ont habité Garges ? oui
non
Est-ce que vous étes né(e) en France ? oui non
Quelle(s) est (sont) votre (vos) nationalité(s) ?

 

 

 

 

 

Quelle profession cherchez-
vous ?
Origine(s) familiale(s)(vous pouvez encercler plus d’un choix):

France

Cara'ibes

Afrique du nord

Au sud du Sahara/ Afrique noire

Afrique de l’ouest

Espagne

Portugal

Europe de I’est

Asie

Amérique du sud, central

Amérique du nord

Autre
Religion ?
Etes-vous pratiquant(e) ? oui non

 

 

 

195

Langues parlée(s) et niveau (debutant, moyen, avancée, langue maternelle)

 

 

 

Entourez la réponse qui convient :

196

Marave Je connais ce
mot et je Je ne
l’utilise : a. souvent b. rarement c. jamais connais
pas ce
mot
Ce mot veut
dire : a.marcher b. massacrer c. discuter Autre :
Ce mot est
principaleme a. les b. les c. les deux
nt utilise par : hommes femmes
a. les gens b. les gens c. les gens
peu moyennemen tres
scolarisés t scolarisés scolarisés
a. une b. ma c. une
generation generation generation
plus jeune plus vieille
a. entre b. en public c. partout
amis
Je connais ce
mot et je Je ne
I’utilise : a. souvent b. rarement c. jamais connais
pas ce
mot
Ce mot veut
dire : a. fumer b. danser c. avoir faim Autre :
Cemmea
principaleme a. les b. les c. les deux
nt utilise par : hommes femmes
a. les gens b. les gens c. les gens
peu moyennemen tres
scolarisés t scolarisés scolarisés
a. une b. ma c. une
generation génération generation
plus jeune plus vieille
a. entre b. en public c. partout
amis
We Je connais ce
mot et je Je ne
l’utilise : a. souvent b. rarement c. jamais connais

 

Ce mot veut
dire :

Ce mot est
principaleme
nt utilisé par :

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Ce mot est
principaleme
nt utilise par :

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Ce mot est
principaleme
nt utilisé par :

a. voler

a. les
hommes
a. les gens
peu
scolarisés
a. une
generation
plus jeune
a. entre
amis

a.souvent

a. un bébé

a. les
hommes
a. les gens
peu
scolarisés
a. une
generation
plus jeune
a. entre
amis

a.souvent

a. une
mmme

a. les
hommes

b.coufir

b. les
femmes

b. les gens
moyenneme
nt scolarisés
b. ma
generation

b. en public

b. rarement

b. un homme

b. les
femmes

b. les gens
moyennemen
t scolarisés

b. ma
generation

b. en public

b. rarement

b. l’herbe

b. les
femmes

197

c. fumer
c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille
c. partout

c. jamais

c. un joint
0. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille
c. partout

c. jamais

c. un enfant

c. les deux

pas ce
mot

Autre :

Je ne
connais
pas ce
mot

Autre :

Je ne
connais
pas ce
mot

Autre :

Binouse

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Cemdea
principaleme
nt utilisé par :

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :
Cemdea
principaleme
nt utilisé par :

a. les gens
peu
scolarisés
a. une
generation
plus jeune
a. entre
amis

a.souvent

a. un gros
nez

a. les
hommes
a. les gens
peu
scolarisés
a. une
generation
plus jeune
a. entre
amis

a.souvent

a. le ciel

a. les
hommes
a. les gens
peu
scolarisés
a. une
génération
plus jeune

198

b. les gens
moyennemen
t scolarisés

b. ma
generation

b. en public

b. rarement

b. une biere

b. les
femmes

b. les gens
moyenneme
nt scolarisés
b. ma
generation

b. en public

b. rarement

b. le ski

b. les
femmes

b. les gens
moyennemen
t scolarisés

b. ma
generation

c. les gens
tres
scolarisés
c. une
generation
plus vieille
c. partout

c. jamais

c. un pince-

nez
c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille
c. partout

c. jamais

c. un whisky

c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille

Je ne
connais
pas ce
mot

Autre :

Je ne
connais
pas ce
mot

Autre :

Daronne

\

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Cemdefi
principaleme
nt utilise par :

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Cemdea
principaleme
nt utilise par :

Je connais ce

mot et je

a. entre
amis

a.souvent

a. une
maladie
d’enfance

a. les
hommes
a. les gens
peu
scolarisés
a. une
génération
plus jeune
a. entre
amis

a.souvent

a. un
poisson

a. les
hommes
a. les gens
peu
scolarisés
a. une
generation
plus jeune
a. entre
amis

199

b. en public

b. rarement

b. massacrer

b. les
femmes

b. les gens
moyennemen
t scolarisés
b. ma
generation

b. en public

b. rarement

b. un cousin

b. les
femmes

b. les gens
moyenneme
nt scolarisés
b. ma
generation

b. en public

c. partout

c. jamais

c. un mégot

c. les deux

c. les gens
tres
scolarisés
c. une
génération
plus vieille
c. partout

c. jamais

0. un pere

c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille
c. partout

 

Je ne
connais
pas ce
mot

Autre :

l.
a
3
3

Je ne
connais
pas ce
mot

Autre :

Je ne

1 .

Sa ou fé

Choucrane

I’utilise :

Ce mot veut
dire :

Ce mot est
principaleme
nt utilisé par :

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Ce mot est
principaleme
nt utilise par :

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Ce mot est
principaleme
nt utilise par :

a.souvent

a.une
cousine

a. les
hommes
a. les
gens peu
scolarisés
a. une
generation
plus jeune
a. entre
amis

a.souvent

a. Ca va?

a. les
hommes
a. les
gens peu
scolarisés
a. une
generation
plus jeune
a. entre
amis

a.souvent

a. de rien

a. les
hommes
a. les

200

b. rarement

b. une mere

b. les
femmes

b. les gens
moyennemen
t scolarisés

b. ma
generation

b. en public

b. rarement

b. en effet

b. les
femmes

b. les gens
moyennemen
t scolarisés

b. ma
generation

b. en public

b. rarement

b. va-t-en

b. les
femmes
b. les gens

c. jamais

c. un
poisson

c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille
c. partout

c. jamais

c. mais oui
c. les deux
c. les gens
tres
scolarisés
c. une
generation

plus vieille
c. partout

c. jamais

c. merci
c. les deux

c. les gens

connais
pas ce
mot

Autre :

Je ne
connais
pas ce
mot

Autre :

Je ne
connais
pas ce
mot

Autre :

lynx-«fl I! l-. '
l

 

 

 

Sa ka
maché

Keum

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Ce mot est
principaleme
nt utilise par:

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Ce mot est
principaleme
nt utilisé par :

gens peu
scolarisés
a. une

generation
plus jeune

a. entre
amis

a.souvent

a. une
maison

a. les
hommes
a. les gens
peu
scolarisés
a. une
génération
plus jeune
a. entre
amis

a.souvent

a. un
homme

a. les
hommes
a. les gens
peu
scolarisés
a. une
generation
plus jeune

201

moyenneme
nt scolarisés
b. ma
generation

b. en public

b. rarement

b. Ca va ?

b. les
femmes

b. les gens
moyennemen
tscolarisés
b. ma
génération

b. en public

b. rarement

b. une voiture

b. les
femmes

b. les gens
moyennemen
t scolarisés
b. ma
generation

tres
scolarisés
c. une
generation
plus vieille

c. partout

c. jamais

c. C’est bon.

c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille
c. partout

c. jamais

c. une féte

c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille

Je ne
connais
pas ce
mot

Autre :

Je ne
connais
pas ce
mot

Autre :

I: It!)

Se (Ial

natchave

schmidts

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Cemmem
principaleme
nt utilise par :

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Ce mot est
principaleme
nt utilise par :

a. entre
amis

a.souvent

a. voler

a. les
hommes
a. les gens
peu
scolarisés
a. une
generation
plus jeune
a. entre
amis

a.souvent

a. les
Allemands

a. les
hommes
a. les gens
peu
scolarisés
a. une
generation
plus jeune
a. entre
amis

202

b. en public

b. rarement

b. s’échapper

b. les
femmes

b. les gens
moyenneme
nt scolarisés
b. ma
generation

b. en public

b. rarement

b. les
ballerines

b. les
femmes

b. les gens
moyennemen
t scolarisés

b. ma
génération

b. en public

0. partout

c. jamais

c. s’habiller
bien

c. les deux

c. les gens
trés
scolarisés
c. une
generation
plus vieille
c. partout

c. jamais

c. les
gendannes

c. les deux

c. les gens
tres
scolarisés
c. une
génération
plus vieille
c. partout

Je ne
connais
pas ce
mot

Autre :

Je ne
connais
pas ce
mot

Autre :

1‘an (A. .x

:Qscii

 

 

kifier

Ouaiche

Makoumé

 

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Ce mot est
principaleme
nt utilise par :

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Cemmefl
principaleme
nt utilise par :

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :
Ce mot est

a.souvent

a. aimer

a. les
hommes
a. les gens
peu
scolarisés
a. une
generation
plus jeune
a. entre
amis

a.souvent

a.Qu’est-
ce qui se
passe ?

a. les
hommes
a. les gens
peu
scolarisés
a. une
génération
plus jeune
a. entre
amis

a.souvent

a. bien fait

203

b. rarement

b. diviser

b. les
femmes

b. les gens
moyennemen
t scolarisés
b. ma
génération

b. en public

b. rarement

b. Ca va?

b. les
femmes

b. les gens
moyenneme
nt scolarisés
b. ma
generation

b. en public

b. rarement

b. marcher

c. jamais

c. manger
c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille
c. partout

c. jamais

c. oui

c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille
c. partout

c. jamais

C. U”

homosexuel

Je ne
connais
pas ce
mot

Autre :

Je ne
connais
pas ce
mot

Autre :

Je ne
connais
pas ce
mot

Autre :

 

‘0.

 

Savater

principaleme
nt utilise par :

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Cemdem
principaleme
nt utilise par:

Je connais ce
mot et je
l’utilise :

Ce mot veut
dire :

Cemdea
principaleme
nt utilise par:

a. les
hommes
a. les gens
peu
scolarisés
a. une
génération
plus jeune
a. entre
amis

a.souvent

a.
massacrer

a. les
hommes
a. les gens
peu
scolarisés
a. une
generation
plus jeune
a. entre
amis

a.souvent

a. le bled

a. les
hommes
a. les gens
peu
scolarisés
a. une
generation
plus jeune

204

b. les
femmes

b. les gens
moyennemen
t scolarisés
b. ma
génération

b. en public

b. rarement

b. manger

b. les
femmes

b. les gens
moyennemen
t scolarisés

b. ma
generation

b. en public

b. rarement

b. un joint

b. les
femmes

b. les gens
moyenneme
nt scolarisés
b. ma
generation

c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille
c. partout

c. jamais

c. aimer

c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille
c. partout

c. jamais

c. un
saucisson

c. les deux

c. les gens
tres
scolarisés
c. une
generation
plus vieille

Je ne ..
connais
pas ce
mot

”F" Ei’J'AA‘hI x

Autre :

Je ne
connais
pas ce
mot

Autre :

 

 

a. entre b. en public c. partout
amis

 

 

205

REFERENCES

MIPES: Mission d'information sur la pauvreté et l'exclusion sociale en lle—de-
France. M.d.i.s.l.p.e.l.e.s.e. lle-de-France: 2005. www.mipes.org.

Abdulaziz, Mohamed H. and Ken Osinde. 1997. Sheng and Engsh: Development
of Mixed Codes among the Urban Youth in Kenya. International Journal of
Society and Language 125.43-63.

Aitsiselmi, Farid. 1997. Arab Language and Identity in Hexagon. International
Journal of Francophone Studies 1.41-52.

Albert, Steve J. 2001. Linguistic Anthropology and the Study of Contemporary
France. The French Review 74.1165-75.

Amara, Fadela. 2003. Ni putes ni soumise Paris: Editions la Découverte.

Antoine, Fabrice. 1998. Des mots et des oms: Verlan, Truncation and Formal
Recycling in Contemporary Slang. Cahiers de Lexicologie 72.41-70.

Ayres-Bennett, Wendy. 1996. A History of the French Language Through Texts.
New York: Routledge.

Azra, Jean-Luc and Véronique Chéneau. 1994. Language Games and
Phonological Theory. Verlan and the Syllabic Structure of French. Journal
of French Language Studies 4.147—70.

Baccaini, Brigitte. 1998. Recent Periurban Growth in the lle-de-France: Forms
and Causes. Population: an English Selection 10.349-84.

Bachmann, Christian and Luc Baisier. 1984. Le verlan, argot d'école ou langue
des keums. Mots 8.169-87.

Baker, Collin. 1992. Attitudes and Language Philadelphia: Multilingual Matters.

Battye, Adrian and Marie-Anne Hintze. 1992. The French Language Today. New
York: Routledge.

Billiez, Jacqueline (1992) "Le parler véhiculaire interethnique de groupes
d'adolescents en millieue urbain," in Jean Calvet, ed., Des Villes et des
langues. Dakar, Paris: Didier Erudition, 117-126

Boocock, Sarane Spence. 1974. Youth in Three Cultures. The School Review
8393-1 1 1.

206

Bortot, Christine. 1998. French, the Language of Moliere?
Fremdsprachenunterricht 42.127-29.

Bourdieu, Pierre. 1982. Ce que parler veut dire: l'économie des échanges
Iinguistiques. Paris: Fayard.

Bourdieu, Pierre and John B. Thompson. 1991. Language and symbolic power.
Cambridge, MA.: Harvard University Press.

Bouzini, Roja. 2000. Variation in Language Usage of Maghrebian Youth in La
Goutte D'Or Paris XVlllth. Language and Society 9283-84.

‘1

Boyer, Henri. 1997, Nouveau francais, parler jeune ou langue des cites? Langue
francaise 114.6-15.

‘—

‘ W, I“ a—A‘

Boyer, Henri. 2001. Young People's French as Lived and Seen by Students. lit-
Surveys in Montpellier, Paris, Lille. Langage & Société 95.75-87.

Brulé, Michel. 1966. L'appartenance religieuse et la vote du 5 décembre 1965.
Sondages 28.9-23.

Bucholtz, Mary. 2002. Youth and Cultural Practice. Annual Review of
Anthropology 31 .525-52.

Bullock, Barbara E. 1996. Popular Derivation and Linguistic Inquiry: Les
Javanais. The French Review 70.180-91.

Calvet, Louis-Jean. 1989. On Contacts between Languages. Le Francais dans le
Monde 200.23.

Calvet, Louis-Jean. 1993. A Verlan Kit. Le Francais dans le Monde 256.42.

Calvet, Louis-Jean. 1994. Les voix de la ville. Paris: Payot.

 

Calvet, Louis-Jean. 1994. L'argot. Que sais-je? Paris: PUF.

Calvet, Louis-Jean. 1996. Ask for veull It's a New Release. Le Francais dans le
Monde 281.42.

Calvet, Louis-Jean. 1996. Les Politiques Iinguistiques. Paris: Presses
Universitaires de France.

Calvet, Louis-Jean. 1997. A Linguistic "Melting Pot". Le Francais dans le Monde
292.23.

Calvet, Louis-Jean. 1997. La Prof de sefran, elle soque que tchi. Le Francais

207

dans le Monde 288.21.

Calvet, Louis-Jean. 1998. The Instant and the Long Term. Le Francais dans Ie
Monde 294.21.

Calvet, Louis-Jean. 1998. Language wars and linguistic politics. Oxford and
New York: Oxford University Press.

Calvet, Louis-Jean. 2006. Linguistic Fractures. Bulletin suisse de linguistique
appliquée 83.29-38.

Calvet, Louis-Jean. 1981. Les langues véhiculaires. Paris: Presses Universitaires
de France.

Carpin, Jean-Michel, Chloe Tavan, and Agnés Dugué (ed.) INSEE. 2005. Les
imigrés en France (Statistique Publique.) Paris: Jouve.

Certeau, Michel de. 1986. Heterologies: discourse on the other. Minneapolis:
University of Minnesota Press.

Cervenkova, Marie. 2001. The Influence of Argot on the Common Language and
Its Formation Processes in Contemporary French. Etudes Romanes de
Brno 31 .77-86.

Chumbow, Beban Sammy and Augustin Simo Bobda (2000) "French in West
Africa: A Sociolinguistic Perspective." In Sociolinguistics in West Africa,
Ayo Bamgbose (ed.), 39-60. Berlin and New York: Mouton de Gruyter.

Crystal, David. 2003. A Dictionary of Linguistics and Phonetics. In A Dictionary of
Linguistics and Phonetics, D. Crystal (ed.). Malden: Blackwell.

Delvau, Alfred. 1883. Dictionnaire de la langue verte. Paris: Editions Dentu.

Doran, Meredith Christine. 2002. A Sociolinguistic Study of Youth Language in
the Parisian Suburbs: Verlan and Minority Identity in Contemporary
France. Dissertation Abstracts International, A: The Humanities and Social
Sciences 62.4143-A.

Echchaini, Nabil. 2001. We are French too, but different. Gazette 63.295—310.

Eckert, Penelope. 2003. Language and Adolescent Peer Groups. Journal of
Language and Social Psychology 23.112-18.

Eckert, Penelope and Sally McConnell-Ginet. 1999. New generalizations and

explanations in language and gender research. Language in Society
28.185-201.

208

 

Eriksen, Thomas. 1992. Hegemony and Minority Resistance. Journal of Peace
Research 29.313-32.

Feral, Carole de. 1994. The Adaptation of French in Southern Cameroon.
Langue Francaise 104.37-48.

Fink, Teresa. 2005. Attitudes towards Languages in Nairobi: University of
Pittsburg Master's thesis.

Gadet, Francoise. 2003. La variation sociale en francais. Paris: Ophrys.

George, K. E. M. 1986. The Language of French Adolescents. Modern
Languages 67.137-41.

Ginthinji, Peter. 2008. Sexism and (mis)representation of women in Sheng.
Journal of African Cultural Studies 20.15-32.

Githiora, Chege. 2002. Sheng: Peer Language, Swahili Dialect or Emerging
Creole? Journal of African Cultural Studies 15.159—81 .

Goudailler, Jean-Pierre. 1997. Comment tu tchatchesl: dictionnaire du francais
contemporain des cites. Paris: Maisonneuve et Larose.

Goudaillier, Jean-Pierre. 2002. From Traditional Argot to Contemporary Urban
French. La Linguistique 38.5-23.

Guiraud, Pierre. 1963. L'argot. Paris: Presses Universitaires de France.
Gumperz, John J. 1983. lnterethnic Discourse. Society 20.64-69.

Haliday, Michael. 1978. Antilanguages. Languages as a Social Semiotic, The
Social Interpretation of Language and meaning. London: Edward Arnold.

Hargreaves, Alex. 1997. Immigration and identity in Beur fiction: voices from the
North African immigrant community in France. New York: Berg.

Heath, J. G. 1984. Language Contact and Language Change. Annual Review of
Anthropology 13.367-84.

Houbedine-Gravaud, Anne-Marie. 2003. Trente ans de recherche sur la
diférence sexuelle, ou Ie Iangage des femmes et la sexuation dans la
langue, les discours, les images. Langage & Société 106.33-61.

Hudson, RA. 1980. Sociolinguistics London: Cambridge University Press.

209

Kie

K0

K0

 

Jablonka, Frank. 2003. Suburban Sociolinguistics. Cahiers de I'institut de
linguistique 29.165-77.

Kiessling, Roland. 2005. Bak mwa me do-Camfranglais in Cameroon. Lingua
Posnaniesis 47.87-107.

Kouega, Jean-Paul. 2003. Word Formative Processes in Camfranglais. World
Englishes 22.511-38.

Kouega, Jean-Paul. 2003. Camfranglais: A Novel Slang in Cameroon Schools.
English Today 19.23-29.

Labov, William. 1972. Sociolinguistic Patterns. Philadelphia: University of
Pennsylvania Press.

Labov, William. 1994. Principles of Linguistic change. Oxford: Blackwell.

Lafont, R. 1997. Quarante ans de sociolinguistique a la périphérie. Paris:
L'Harrnattan.

Lambert, Wallace E.; Fathali M. Moghaddam, Jean Sorin, and Simone Sorin.
Assimilation vs. Multiculturalism: Views from a Community in France.
Sociological Forum 5.387-411.

Lefebvre, Anne. 1991. Le Francais dans la région lilloise. Paris: Publications de
la Sorbonne.

Lefl<owitz, Natalie Jeanne. 1988. Talking Backwards, Looking Forwards: The
French Language Game Verlan. Dissertation Abstracts International, A:
The Humanities and Social Sciences 48.3101-A.

Lefkowitz, Natalie Jeanne. 1989. Verlan: Talking Backwards in French. The
French Review 63.312-22.

Leila. 2004. Mariée de force. Paris: Oh! Editions.

Lepoutre, David. 1997. Coeur de banlieue, codes, rites et Iangage. Paris:
Editions Odile Jacob.

Lijphart, Alain. 1971. Class Voting and Religious Voting in the European
Democracies: a preliminary Report. Glasgow: The University of
Strathclyde Press.

Lobé-Ewane, Michel. 1989. Camaroune: le camfranglais. Diagonales 10.33-34.

Marconot, Jean-Marie. 1989. Aspects of Youth Language Knowledge in the

210

Domain of French-Occitan Diglossia. Revue des Langues Romanes
93.189-92.

Mathews, Peter H. 2008. The Concise Oxford Dictionary of Linguistics: Oxford
University Press.

Mela, Vivienne. 1991. Verlan or Mirror Language. Langages 25.73-94.

Mela, Vivienne. 1997. Verlan 2000. Langue Francaise 114.16-34.

Méliane, Loubna. 2003. Vivre Libre. Paris: Oh! editions.

Melliani, Fabienne. 2000. La langue du quartier: appropriation de l'espace et
identités urbaines chez des jeunes issus de I'immigration maghrébine en
banlieue rouennaise. Paris: L'Harmattan.

Merle, Pierre. 1997. Argot, Verlan et Tchatches. Toulouse: Editions MILAN.

Merlin, Pierre. 1998. Les banlieues des villes francaises. Paris: La
Documentation francaise.

Michelat, Guy and Michel Simone. 1997. Classe religion et comportement
politique. London: Frank Cass.

Milroy, Lesley. 1980. Language and Social Network. Belfast Working Papers in
Language and Linguistics 2.1-11.

Milroy, Lesley.1982. Social network and linguistic focusing. In Sociolinguistic
Variation in Speech Communities (ed.) Suzanne Romaine, 141-52.
London: Edward Arnold.

Mondesir, Jones E. 1992. Dictionary of St. Lucian Creole. In Trends in Linguistics
Documentation 7 (ed.) L.D. Carrington. New York: Mouton de Gruyter.

Myers Scotton, Carol and John Okeju. 1973. Neighbors and Lexical Borrowings.
Language 49.871 -89.

Nicolas, Isabelle. 2005. The Slang of the Inner-City [les Banlieues] as a
Reflection of Social Fracture. Fremdsprachenunterricht 5.51-52.

Offord, Malcolm H. 1996. A reader in french Sociolinguistics. Clevedon,
Philadelphia: Multilingual Matters.

Pecheur, Jacques. 1988. How the French View Their Language. Le Francais
dans le Monde 2928-29.

211

Petitpas, Thierry. 1998. A Descriptive Analysis of Three Processes of French
Morphological Coding: Javanais. Verlan and Largonji. Cahiers de
Lexicologie 73.149-66.

Platen, Christoph. 1994. Talking Backwards, Looking Forwards. The French
Language Game Verlan. Zeitschrift fur Romanische Philologie 110.577-
79.

Plenat, Marc. 1995. A Prosodic Approach to the Morphology of Back Slang.
Lingua 95.97-129.

Pooley, Thomas. 2003. The Man-Woman Differentiation When Speaking
Regional Languages in France. Langage & Société 106.9-31.

Preston, Dennis R. 2000. Some Plain Facts about Americans and their
Language. American Speech 75398-401.

Preston, Dennis R. 2002. Language with an Attitude. The Handbook of Language
Variation and Change, ed. by J.K. Chambers; Trudgill, Peter; Schilling-
Estes, Natalie 40-66. Malden: Blakwell.

Robb, Graham. 2007. The Discovery of France; a Historical Biography. New
York: W.M. Norton and Company.

Schwob, Marcel. 1999. Etudes sur l'argot francais. Paris: Editions Allia.

Séguin, Boris and Frederique Teillard. 1996. Les céfrans parlent aux francais:
Chronique de la langue des cites. Paris: Calmann-Lévy.

Sewel, Penny and Chris Payne. 2000. Stop Bugging Me, You're Really Giving
Me a Headache with all Your Crap. Francophonie 21 .8-1 1.

Simo-Bobda, Augustin. 1994. Aspects of Camaroon English Phonology. Bern:
Peter Lang.

Suleiman, Yasir. 2006. The Linguistic Construction of National Identity. Paper
presented to the CASID, Michigan State University, 2006.

Teevan, James J. 1985. Socioeconomic status is significantly related to....
Sociological Perspectives 28.241-47.

Tejedor de Felipe, Didier. 2004. A propos de la "folklorisation" de l'argot des
jeunes. Francais des banlieues, francais populaire?, ed. by M.-M.D.D.
Bertucci. Cergy-Pointoise: Université de Cergy-Pointoise.

Tiayon-Lekoubou, Charles. 1985. Camspeak: a speech reality in Camaroon:

212

University of Yaounde, Camaroon. Unpublished Master's dissertation.

Trudgill, Peter. 1972. Sex, Covert prestige and Linguistic Change in the urban
British English of Norwich. Language and Society 1.179-95.

Trudgill, Peter. 1974. Sociolinguistics : An introduction. Hannondsworth:
Penguin. .

Trudgill, Peter. 1984. Applied sociolinguistics. London: Orlando Academic Press.

Urciuoli, Bonnie. 1995. Language and Borders. Annual Review of Anthropology
24.525-46.

Valdman, Albert. 2000. The Language of the Inner and Outer Suburbs: From
Argot to Popular French. The French Review 73.1179-92.

Vermeij, Lotte. 2004. "Ya Know What I'm Sayin'?" The Double Meaning of
Language Crossing among Teenagers in the Netherlands. lntemational

Journal of Society and Language 170.141-68.

Vermes, Genvieve (ed.). 1988. Vingt-cinq communautés linguistiques de la
France (logiques sociales 1). Paris: Harmattan.

Vermes, Genvieve (ed.). 1988. Vingt-cinq communautés linguistiques de la
France (logiques sociales 2). Paris: Harmattan.

Walter, Henriette. 1988. Le francais dans tous les sens. Paris: Editions Robert
Laffont.

Walter, Henriette.1997. The Lexis of the very young. Langue Francaise 114.41-
55.

Walter, Henriette.1997. L'aventure des mots francais. Paris: Editions Robert
Laffont.

Weinreich, Ulrich. 1953. Languages in Contact New York: Linguistic Circulation
Publication.

Wieviorka, Michel. 1996. Identity and difference: reflections on the French non-
debate on multiculturalism. SAGE.49-71.

Zerling, Jean-Pierre. 1998. Verlan 2000 at the Bastille: A Phonetic Approach.
Travaux de l'institut de Phonétique de Strasbourg 28.203-32.

Zerling, Jean-Pierre.1999. A Small Bilingual Lexicon of Urban Verlan: Verlan-
F ranais and Cefran-Verlan. Travaux de l'institut de Phonétique de

213

Strasbourg 29.95-106.
Zerling, Jean-Pierre.1999. The Syllabic and Morphological Structure of Verlan: A

Phonetic Analysis of Urban Verlan Vocabulary. Travaux de l’institut de
Phonétique de Strasbourg 2973-93.

214