Heilbron
Heilbron
Heilbron
çt
t!^t
Johan Heilbron
CENTRE LILLoTS D'ÉTUDES ET DE RECHËRCHES SOCIoLoGIQUES ET
É,coNoMIquEs (cLERsE), FRANcE
Abstract
This article argues that the translation of books may be fruitfully understood
as constituting a cultural world-system. The working of this system, based on
a core-periphery structure, accounts for the uneven flows of translations
between language groups as well as for the varying role of translations
within language groups. The final part outlines how this general sociological
model may be further developed.
Key words
r globalization r translation studies ¡ transnational cultural exchange
r world-systems theory
of ranslations which I will explore here, tfie conceptual shift from source-text to
terget-context offers a f¡uitfr¡l but insufficient point of departure.
They are therefore by no means self-evident, a fact which the termi¡ology itself
recalls. In classical Greek, for example, the¡e is no proper word for translation,
only hermeneuein,but that also means to interpret, to explain. TheLeun tøn¡-
htio is cioser to the cur¡ent meaning, but is a more general term as well, reFer-
ring to the va¡ious forms of ransfer, including the tra¡lsfer of power, as in
trqnsktio imperü. The more specitc and modern serue of the word'ranslatiorÍ
emerged only in the Renaissa¡ce, rilhen halian humanists started to distinguish
ben¡¡een tanshtio and *øductio. The later term, and the coresponding verbs in
Italian and French, tøducere and tradaire, referred speciÊcally to the translation
of texrs Êom one'language into another, and especially into the ve¡nacular.
Tianslations into the vernacula¡ had existed well befo¡e the Renaissance, but
'the printing press gave tÌ¡e vernaculars, and uanslations in the verna.ulr¡, att
entireþ new social significance. rFith dre formation oFnational states, staldard
languages were codi6ed and much of the tra¡sladon acriviry in early modern
Europe was bound m tle evolving relations of cooperacion and confl.ict be¡ween
nadon states.
The actual practice of ua¡slation obviously exists in a greet variety of forms
and contexts: interpreting in political and diplomatic semings; subtiding and
dubbing in the media; high-culture litera¡y translation; as well as a range of rnore
standa¡dized, technical and professional translations in law, technology and
cornmerce. If meaning is determined by use, as dre pragmatist adage says, trens-
lation practice musi be analysed specifically witåin the teld o¡ the subñeld in
which it ecÈually funcdons.
ln this anide I em concerned with one major form of ua¡slation: r}re trans-
lation of boofts. Book ua¡slations represent an identiâable and broad cffegory:
they are published and disributed in a simila¡ manner; they are registered,
counted and classified as a particula¡ category of cultu¡al goods; and they are
destined for a wide variety of audiences. Sociologically such ransladons can be
srudied from various angles. By analysing transl¿dons, questions can be raised
about the way in q'hich cultural goods circulate ouside tireir context of produc-
ii^-
rÀv¡¡
l'R^;ttJí.rì
\.yve¡sÙ¿,^¿
I QAfì\
¿ vrt ^nê -ãn rñ' r^ rrñrrr'êl tl'. rolati^-.lri^ h.ntr..n Åiffer.n¡
countries and cultu¡es (Schoneveld, i983), srudy the role of intermediary centres
(Dirlcx, t995), decþher the complocities of cross-cultu¡al (mis)understanding
(Oz-Salzberget 1995), consider translators as a professional group (Heinich,
I9B4), or analyse the evolution of the system of rransnational communication
itself, for enample by studying túe social organization of the market for trans-
larion rights, the role of literary agents or the functioning of international .book
fairs (Sorá,,1'998). .
This anicle focuses on what is perhaps the most general issue in the sociolog¡r
of uanslation: the uanslation of books considered as an inte¡national system. The
objecdve is to present a suuctu¡al analysis of the international flows of translated
books, and to demonsuare why such an analysig is indispensable for under-
standing the actud ranslation process. Two more speciÊc questions ete cenûel
in this respect. How can one account for the uneven flows of book uanslations
bet'¡seen va¡ious language groups? ,A¡d how can one explain the varying role of- -.
ranslations within different language groups? In proposing ari answer to bor}r--
432 European Journal of Social Theory 2(4)
As the basic units of ú¡e world-sysrem of uansiation are language groups, the
object of analysis is the süucrure of the tra¡sladon flows berween these language
groups. Language groups do not always coincide with nation srares: some of the
more central languages - English, German, French, Spanish - have a suprana-
tional cha¡acter. The flow of boolc uanslations berween tlese language groups can
be analysed by using book statisÈics, which regularly intlude 6gures for tra¡sla-
tions. In reþing on these figures, however, a g{e3r deal ¿rf caurion is required.
Contrary to tle normal primary use of official statistics, the statistical material
itsehchas to be cridcally oramined before it can be used.
International uanslation staristics have been produced since the 1930s. The
Irutitute for Intellectual Collaboration, which formed a part of the League of
Nations, sta¡ted during the interwar years widr annual publications about
t¡anslated books, rJrre Indcx trdnslationam (193240). The activiry was part.of the
Johan Heilbron A Sociology of Tia¡rslation 433
initiatives after the First \Øorld \Øa¡ ro promote inte¡narional collabo¡acion and
the ¡nuual understanding berween narion.s. After rhe Second'!Øorld SØa¡, Unesco
resurned publication of the uanslarion stadsrics; they have been published in *¡e
Ijnesco series of StøtisticdlYearboohs ever since. A closer look at these statiscics
reveals that they a¡e not very reliable. The mosr obvious problem is one of defi-
nition. \Øhat is considered to be a 'book' or a 'dde' va¡ies from country to
country. Certain publicacions qualify as a 'tide' or a 'book' in one country
¡¡¡hereas they are considered ro belong to rhe'grey' lirerature in other countries
and a¡e therefore eliminared from of;Êcial book statistics. Such is the case for
docto¡al dissertations, school boola, governmental, parliamentary a¡d adminis-
trative documenß, and annual reporrs from enterprises. W'tren it is reponed that
21 percent of rhe published books in Spain in 1982 were translations, th.is
percentage does not have dre same meaning in orler countries. Rigorous
comparisons berween translarion ratios e¡e the¡efore impossible on the basis of
the Unesco figures.
\Øhen analysing the Unesco staristics for one country alone, precisely in order
to avoid such deânirion problems, it turns our, frrftherrnore, that they exhibit
great fluduetions over the years. According to the Unesco figures, 14 peicent of
the published books in the Netherlands were tra¡rslations in L979, a percentage
.i...:1j ..j
which has risen ro 34 percenr five years later. Such a fluctuation is highly improb-
:. ..:: I j . ; .-. .:1:f.1 :.
abfe and it does not correspond with dre data provided by the Dutch agenry, the
Stichting Speanaerk, which produces rhe national book statistics for råe Netåer-
lands. Their figures indicate råat the percentage of transladons in national book
production is more regular, varying berween 22 percent and 25 percent between
1979 and 1984 (Heilbron,l995a). The Unesco data are therefore not very reii-
ablq it remains unclea¡ to what extent they are actually comparable, and even for
single countries they exhibit very improbable flucruations. IJnfortunatel¡ these
figures a¡e the only international data which a¡e readily available. I will therefore
use some of them, but merely in an indicative manner to highlight srructural
patterns. I will ref¡ain from giving any rables and-b¡eakdowns in categories and
lub-categories of bools, bec'¿use in the form in which they are rrsrio-üy ¡;ublishccl
drey suggest a degree of accuracy which is misleading if not unfounded.
If one cauriously combines r}re inrernational uanslation statistics with some
of the more reliable national data a¡d wiúr what is lcnown from several case
studies, a coherent model ca¡ be constructed of the srucrural dynamics of the
inrcrnational translation system. I will outline somewhat schemarically its main
properties and illusrate their significance for understanding uanslation practices.
indicate rlat English is byfar the most cenual language in rhe international trans-
lation system. Mot than 40 perceff of ali the ranslated boolcs worldwide were
a¡ound 1980 uanslated from English (Cu¡wen, 1986: 2l; Venuti, 1995: 14).
Over the years, from 1960 rc abóur 1987, this percentage seems to have gone
up, despite the fact that the percentage of English bools in the total number of
bóoks worldwide has decreased (Mélitz, 19982 3Ç7)' On the European conri-
nent the position of English.is even more predominantwith about 50-70 percent
of the published uanslations being made from Engiish.G
Fo[ãwing rhe ranking downwards, drree languages alsg have a cenüal ro.le,
although rhèir sha¡e is significandy smaller tha¡ that of English: French, German
and Russia¡r. Each of these languages had a þroponion somewhere between 10
percent and 12 pefcent of the international ma¡ket for transladons a¡ound 1980.
it follows ftom r]tese 6gures that *uee-guarters of the total number of ua¡slated
books worldwide were uanslated from four languages only. The international
uansladon Сstem is thus ma¡ked by a very uneyen disuibution and is firrnly
dominated by English.
Afrer dtese cenual languages, with English in a kind a hyper-central role,
approximately six languages have a semi-peripheral role, to use Imma¡uel'W'ailer-
si"io's terminology, eãch *ith a proportion of 1-3 percent of the total number
of translated books. In ¡978,for example, these languages were Spanish, Italian,
Danish, swedish, Polish and czech.T these semi-peripheral languago, however,
cannot be separated very clearly from dre perþheral ones. Contrary to tj¡e
distinctions bèm".n hyper-central, cenËral and semi-perþheral, which are rela-
tively clear cut, the diffeiences beween semi-peripheral and perþheral languages
f- *or. gradual. One might say provisionally and for analytical purposes that
"r.
all hrrguages with a sha¡e of less tban one Percent of rlre world market occuPy a
perþh-.rJposidon in the inrernadonal ranslation rysït. Among these_perþh-
rhen, are Chinese, Japanese, fuabic and Portuguese, each repre-
"rrli*gn"g"s,
serrtingã vãry hrg. number of speakers, yer occupying a_ peripheral position in
the t¡alsladon sys-rem. The size oll*gu"ggtoups is dearly not decisive for tüeir
degree of cenualiry ir: -ulre tia¡sla'úon sy'stem.
But the stfucnue of the international transladon system is obviously not a static
but a dynamic constellation, The position of language grouPs changes. over time,
cenral languages may lose somedtirrg of their share, more-perþheral- languages
.* improi" tñeir poiitions in the iniernational ranking. The traruladon qfstem
is a hisiorical ryrrot, marked by a specific genesis and rninor and major uans-
formadons over time. The major-changes ar.e long-term processes. \Øhen
considering the relations benveen English, French and German, for example, one
observes diat both the hegemony ofEngliqh and the relative decline of French
have a long history. prerich wæ tlre mosr central rnodern languasj in early
modern Er-irop., more importanr than English or German. The first major
change in the constellation occurred at the end of the eighteendr. century. -For
Johan Heilbron A Sociology of Thanslation 435
geo-polirical and geo-culrural reasons, French lost some of its cenuality, as is indi-
cated in the uanslation starisrics for the Netherlands. The proportion of books
uanslated from French declined fai¡ly rapidly during the last decades of rhe eight-
eenrh a¡d the beginning of the nineteenr:h centlrry (Korpel, 1992). German
especially profited &om tüe French decline; English also gained but the growing
sha¡e of tra¡sladons from English was a relatiyely slow process for quite sorne
rime. The breakthrough of English probably occurred only after dre Second
TØodd'War, when tlre hegemony of rhe US gave English a. decisive advantage over
its main rivals.s
Changes in the internadonal position of languages do not generally occur
abrupdy. They require a cultu¡ai reorientation which talçes at least a change of
generation, and ofren more than that. Chaages in the position of languages and
language groups occur suddenly oniy ifthe position ofa language depends closeþ
on the politiel power of a regime. The central position of Russia¡, for example,
which is clea¡ from the Unesco statistics for the 1980s, will undoubtedly have
declined rapidly since 1989. Its predomina¡rt rolg in the qystem of international
uanslations was based on dre domination of the Soviet Union over Eastern
Europe, impþing obligatory and quasi-obhgttory tra¡slations in nearly all telds,
not merely those which were bound to r]¡e Ma¡xist-Leninist orthodory. Since the
fall of the Soviet ernpire, tåe use of Russian has declined sharply in Eastern
Europe, just as, undoubtedl¡ translations from Russian have.
which was common practice during dre sevenreenrh and eighteenrh centuries,
fell into disrepute at råe end of the old regime, when narionalism became a polid-
cal and cultural force. German and English literarure gained wider recognidon
and ra¡slations into French losr rheir exemplary role.
The ¡et¡anslation of ranslations, indirect or second-hand ranslation, has
become much less common. To a cenain ercenr, however, rhe phenomenon
persists. Even when dre uansladons rhemselves are far more often made direcdy
from tlre original language, the decision to publish a translarion from a periph-
eral language still depends on the existence of their translation in a cenral
language. Literaty ranslations from Spanish into Dutch afte¡ the Second'\J7brld
'I7ar,
for example, were neariy always preceded by their ranslation into one of
the cenrral languages. This was panicularly rhe c¿se for the most prominent
authors (Borges, Cortázar, Ga¡cla Márquez, Vargas Llosa), who were all uans-
latêd into French or English before being published in Dutch (Steenmeije¡
1989). Afthough these books were translated from Spanish, many signs indic4ted
that the English or French translation had acuaily served as the example. The
choice of the ride, che text on the cover, the quoted praise from reviews, all
revealed the exemplary role of the English or French translation.
There were only a few cases in which Dutch publishers published a translation
before their English or French counterpâ¡ts. But they paradoxically confirm the
dominant roie of central languages. Not only were drese uanslated autho¡s
'minor' writers, who were discovered by Dutch specialist, but their translations
into Dutch were not well received, either by t}re critics or by the public. They
illustrate, A contario, that peripheral and semi-peripheral language groups tend to
follow the example of the internadonal centres, including boolcs, etc., that are
imponed into the cenües.
Much of the international comrnunication about boolcs worla in this manner
and is dependent on rhe role of the leading centres of the international system.
Once a boolc is uanslated into a central language by an authoriadve publisher,
it immediately catches the attention of publishers in other parts of the globe. The
simple fact that an ,A.merican or English publisher will publish an author from a
semi-peripherai language is used extensively by the original publisher, because it
is the best recommendation for publishers elsewhere to acquire the uanslation
rights. The international recognition of Dutch literature is a good example of the
Ieading role ofliterary centres in dre transladon business.
Tlansladons from Dutch and Flemish have been made for centuries..{lthough
a few literary figures in the sixteenth and seventeenth centuries seem to have had
.
a certain inteiíaûoäal ¡erìown! none of tl¡em entered the canon of world litera-
ture (Schenkeveld, 199i). In the eighteenúr and nineteenth centuries very few
books were uanslated from Dutch, and it was only from the end of the nine-
teenth century onwards, following the European recognidon of Russian and
Scandinavian literature, that *re number of books tra¡slated f¡om Dutch started
to rise. It increased more or less regularþ with a total number of approúmately
500 or 600 tides per year since dre l960s.e
Despite the relatively steady growh in the number of translations durigg-the
Johøn Heilbron A Sociology of Tianslation 437
¡¡¡entieth century Dutch literarure has remained largely unknown. Undl recently
nor a single Dutch wrirer was inrernadonaily acknowledged as a major literary
figure. Fiãancial supporr and susrained rranslacion efforts proved insufficient, and
th"e lack of lirerary recognition outside rhe Nedrerla¡ds seerned an inescapable
fate (van Noesel andJan-ssen, 1985; Vanderauwera, 1985; Paul, i990)' Thelack
of success was artributed ro rhe doubú.il qualiry of the translarions, to dre fact
thar they were pubiished by small a¡d often marginal publishing houses, and to
the virtual absence of good translators, who would not only produce.ProPer üans-
Iations, bur who ,}o inform publishers, wrire reviews a¡d train future
"o,r-id
translators.
The change sta¡ted during rhe 1980s when a few Dutch authors were
published by weil-established literary publishers; some of t}re uansiators won
iir.r"ry prirå.10 The emerging inrereir r¡/as nor restrioed ro a single country; the
breakiuough occumed iri 6[t**¡ and from there spread to orher literary
centres ,, iell as to mote peripheial language grouPs' Compared with other
cent¡es in the internationi uanslarion tytt.*, German publishers were best
prepared for Dutch authors, since Germany was the only country ïú " cenrral
potitiott which had a tradidon of uanslating Dutch lirerarure and incorporating
it into a national cultu¡al sffate*. Since t¡e end of rhe Napoleonicwars,_Germa¡-
ists had come to consider the iow Countries as a cultural province of Germanic
culrure, a¡¡d as a minor but not irreleYant ally against F¡ench civilization (Kloos,
1992). Dutch and Flemish were perceived as a kind of P|a.tt-Deøtsch and popular
novels especially were translatedlnto German ro meet the rising demand.
Ap.tt from iop"l"r writers, some of the more esrablished literary figures were
aho åa¡slated,ioì th.it role remained a very minor one. It was only in r}re 1980s
that lead,ing German publishers starred to publish tra¡slations of major Dutch
writers: su"hrkamp p"btirh.a cees Nooteboom, Klem-cona Hugo claus and
Hanser H"rry M"liich. Their bools were favourably received-by_literary crirics
and some sold quite well. More transladons followed and German critics
acclaimed Cees hîooteboom, in parricular, as an outsta¡ding European writer.
rwt!-:-
wllcll
:- '!ocì2 !L-
ll¡ L))J
\l-.!"--!^-le ]..ror the f^c,,c of úe inte¡nationa-l book fair at
ÚLt lrLu¡Llr4¡s
Frankfurt, the conditions were favourable for the German recognition to snow-
ball. Since then the number of ranslated Dutch authors has substantially
increased, as has rhe number of languages in which thei¡ work is t¡anslated.
The Durch case illustrates the esiential role of prominent cuitu¡al centres in
the international diffirsion of literarures from rhe semi-periphery. International
culrural cenffes are nor only interested in rhe diffi.rsion of rheir own goods, they
also have a vested inte¡esr ín m.¡rrit uade and the benefits this offers. Symbolic
and econom ic tansit prof.ts are an essenrial component of the worlcing of rhe
internadonal cultu¡al system.
The example of Dutch literature also shows that dependelry on the inter-
national centies works rhe other way around. Once a peripheral literature has had
some degree of internationrl ,""ognition, the recognition abroad will contribute
to a¡rd rãay indeed interfere withlndigenous reputations, In the Netherlands it
'Willem
,r", to speak of rhe 'big rtriee' of poi*rr Dutch literature:
"o*-or
438 European Journal of Social Theory 2(4)
Frede¡ik Hermaas, Gera¡d Reve and Harry Mulisch. Some would add a founh,
Hugo Claus. Fo¡ decades their reputation was nor seriously threatened by anyone,
not even by Hella Hrrsse and Nooteboom. Bur since Haasse and Nooteboom
enjoy a growing international fame, rvhereas Herrnans and fuve do not, rhe
indigenous canon is undermined. Especially in small counüies, the process of
canonization is increasingly affeced by the international market place.
same period Uanslations from German increased from 1.4 percentrc 4.3 Pe¡centr
,r*rl-*iorr, from French from 0.6 Percent to 2.2 percent, and translations from
'other languages'from 1.2 percent ro 2] percent (Heilbron, 1995a)' Thus
aldrough È"giirtt has proñred fa¡ more frorn rhe growth of ransladon¡ than any
orher ioreign languagO these other languages have also increased rlei¡ share in
national book production.
Percent.
These figures, although incomplete and at best indicative, clearly suggest en
inverse rela:tionship b.r*eett the ienualiry of a language in the inte¡national
uanslation qrsrem and the proponion of translations in national book produc-
tion. The more cenÊral úe édtural producdon of a country is, fhe more it serves
as an e.xafnple m other countries, and the less it is iself concerned with the
cultural prod,rccion from other countries. Instead of assuming that transladons
,normally' occupy a marginal posidon (Even-Zohar, 1990: 50), it is fa¡ more
accurete to ,"y th.t theiriole varies sigrritcand¡-and that rhe va¡iation depends
on the degree of centraliqv in the international transladon system' The core of an
inrcrnatioîal cultural system has the highest stetus; it is carefi:Ily observed,
followed and emulated, and at the same time it is mt¡ch less oriented towa¡ds
products and producers from outside the centres.
The ,*. årr be observed in internarional exchange emong the sciences. ,As
indicated by citation patterns, scientific research in the US is the most cenüal
and most prestigious þ"tt of the sciendtc world-system. But sciendûc produc-
tion in the US isäro the lowest percentage of foreign refetences,
"h"t""terwedbyabroad. The percentages offoreþ references
foreign co-authors and publications
in scîentific a¡ticles ,rrå for"ign publications in US citarions are bot}r about 25
percenr. In Japan and the E"ropãrrr counrries the figurelies.somewhere between
Z0 p"r""rr.-rrrd 71 percent; for råe developing_counrries it varjes between 70
p.r..rrt 92 percent (Schort, l99I).Insread of an equilibriurn between import
"nd
td oporr, th.ie"lity of ransnational exchange is a process of uneven exchange,
Fo, .uåry book tbar is tr nslared from Dutcll, for ocample, t-here are six boola -.
440 European Journal of Social Theory 2(4)
t¡anslated into Durch. Imbala¡ces of rhis kind cha¡acterize the very srucru¡e of
transnational o<change.
In order to understand rhe srucrure of rhe inrernarional flows rhere is no need
to invoke the peculiarities of national cul¡ural uaditions. Comparing dre propor-
rion of 10-12 percent rranslarions in France, for exa.mple, with rhe significandy
higher percentage fo¡ Sweden, one might deplore the relative closu¡e of French
cultu¡e. Some even consider this ro be a speciÊc fearu¡e of French cultu¡e. \Øasnt
Chauvin a Frenchma¡? Comparing rhe French widr dre lIKt much iower
percentage of uanslations, one might be equally tempted ro invoke rhe opposite
argument, narnely thar ir resriÍ.es to rhe very richness of the French cultural
tradirion. The tradidonally high esreem for culcure dren appears as being reflecred
in a high level of uanslations.
Neither one of rhese arguments is necessary for ercplaining the level of cultural
importation. The proponion of translarions in France corresponds to tåe inrer-
national posirion of French in the world-sfstem of sansladon, and is perfecdy
comparable with the role of German and the proporrion of uanslations in
Germany. As argued above, it is not so much the national tradition, but rather
the international position of national cuhures, which determines the level of
cultu¡al imponation.
Notes
grouping ofVenud (L995), who has raken the Scandinavian languages, as well as
Greek and latin, together.
8 Book translations from English have an ever growing sharq in rhe number of boola
published in che Netherlands. ln 1946, 39 percent of all tra¡slated books were rrans-
lations Êom English; in 1990 the propordon was up to 65 percent (Heilbron, 1995a).
9 These approximate Êgures are based on the bibliography of ranslations ûorn Dutch
which is produced by rJre Royal Libra¡ies of The Hague and Brussels. T'he absolute
numbers a¡e less significanr tlen the trend they indicate (see Heilb¡on, 1995a).
i0 Important ua¡slation prizes were awarded ro Philippe Noble for his French trans-
lation of E. Du Perron's Zc pays d'orígine (1980) and to,A.drienne Dixon for her rrans-
lation of Rina& (1983) by Cees Nooteboom.
References
Vanderauwe¡a, Ria (1985) Dutch Noue[s Ti,an¡kted ínto English. Amsterdam: Rodopi.
varr Bonenburg, Maarren (L994) Verborgen coTnpetitie: ouer de uiteenlopende popuhríteit
aan tl)orten. Amsrerdam: Ben Bakker.
van Hoof, Henri (1991) Histoire de h trdductíon en Occident. Paris: Duculot.
van Noesel, Ma¡ion and Janssen, Ans (1985) De Nederknùc literataur in Franse ærtøling.
Utrechc Frans en Occitaans Instiruut,
Venud, Lawrence (1995) The Tianskmr! Inuisíbility: A History of 7iørulation.
LondonlNew York Roudedge.
Von Staclcelberg, Jürgen (1984) Uebersetzangen aus zueiter Hand: Rezeptionsuorgaege in
der earopaeischen Literatur aom 14. bís ,/8. Jahrhundert. Berli,l;,lNew York; lfl'alter de
Gruyter,
tüØallerscein,
Immanuel (1991) Gn?olitics and Geocultare: Essdys on the Changing V'orll
System. Cambridge/Paris: Camb¡idge Universiry Press/Édidons de la Maison des
Sciences de I'Homme.
I Johan Heilbron published The Rise of Socíal Theory U995) and Iåe Rise of the
SocÍal Sciences and the Formation of Modernity (1998, co-ed¡ted with Lars
Magnusson and Björn Wittrock). His research is in the sociology of art and culture,
intellectual history and economic sociology. Address: Centre Lillois d'Études et de
Recherches Sociologiques et Économiques (CLERSE), Faculté des sciences
économiques et sociales, Université de Lille 1, 59655 Villeneuve d'Ascq Cedex,
France. [email: johan.heilbron@univ-lillel .frl