The Dutch-French Language Border in Belgium

Home , Brussels Periphery, Dutch in Belgium, Flemish, French Flanders, Voeren

Roland Willemyns Vrije Universiteit Brussel, Germaanse Talen, Pleinlaan 2, B-1050 Brussels, Belgium

Thisarticle is restricted to adescriptionof languageborder fluctuations in Belgium as faras itsDutch-French portion isconcerned.After a briefdescription of theso-called ‘languagequestion’ in Belgium thenotion of languageborder is discussedin general. Then comesan overviewof thestatus and function of thelanguage border in Belgium and of theactual language border fluctuations as they haveoccurred up to thepresent day. Two problem areas:the ‘ Voerstreek’and theBrussels suburban region are discussedin moredetail. Afterwards language shift and changethrough erosionin Brusselsare analysed as wellas thepart played in thatprocess by linguisticlegislation, languageplanning and sociolinguisticdevelopments. Finally a typology of language borderchange is drawn up and thepatterns of changeare identified in orderto explain and accountforthea lmostunique natureoftheBelgianportion of the Romance-Germanic language border.

1. Introduction Belgium (approximately10 million inhabitants) is a trilingualand federal country,consisting of four different entitiesconstituted on the basisof language: the Dutch-speaking community(called Flanders;58% of the population),the French speaking one (called Wallonia;32%), the smallGerman speaking community (0.6%)and the Dutch-French bilingual communityof Brussels(9.5%). Since regionalgovernments have legislative power, the frontiersof their jurisdiction, being language borders, are defined in the constitution (Willemyns, 1988). The Belgian portionof the Romance-Germaniclanguage borderis quite remarkablefor mainly two main reasons: (1) itsstatus and function have changed considerablysince the countrycame into existence; (2) itspresent status andfunction arealmost unique ascompared to all the otherportions under consideration.Because of that it has frequently caughtthe attention(and imagi- nation)of scientists of variousdisciplines (although,for a long time,mainly of historians;Lamarcq & Rogge,1996). It often servedas a parspro toto for the Romance-Germanicborder as a whole andmany researchers have tried to explain itsgenesis basedon itsBelgian portion(see Van Durme in thisspecial issue). Arealbreakthrough hasonly been achieved fromthe momentlinguists have entered the debate.Maurits Gysseling in particularis to be mentioned in this respectbecause ofhis idea to use the oldestlinguistic sourcesavailable (including toponyms)and for the skill displayed in doing so.His work is now being continued –withno less skill –by hisformer student Luc vanDurme (there are references on the work of both in Van Durme’s article in this issue). Thisarticle is restricted to a descriptionof language borderfluctuations in Belgium asfar as its Dutch-French portionis concerned. Language contactin Brusselsis treated in Treffers-Daller, andthe trilingualcontact between German,

36 Dutch-French Language Border in Belgium 37

French andDutch in Belgium’s easternpart in Nelde andDarquennes, bothin this issue.

2. The Language Border and the so-called ‘Language Question’ in Belgium The ‘language struggle’which was going todominate Belgian politicallife startedshortly after 1830, the yearin which Belgium hadbecome anindependent constitutionalmonarchy with a parliamentarysystem dominated by the bour- geoiselite, andwhich secured itsposition by adoptinga poll-taxsystem (out of 3.5million people, only 46,000had the rightto vote; Witte & Van Velthoven, 1998).Although the new constitutionprovided for ‘ linguisticfreedom’ , itwas obviousthat this ‘ freedom’was profitable only tothe richand the powerful, i.e. tothe bourgeoisie fromWallonia and Flanders, all of whomwere French speakers.For this bourgeoisie, French wasanaturalchoice as the language ofthe state.The governmentappointed only French-speaking civilservants and the discriminationof Dutch throughoutthe 19thcentury wasgeneral andvery delib- erate,despite the factthat Dutch speakersconstituted the majorityof the population.There isnoroom here toelaborate on the genesis ofthis situation (for more information see Van de Craen & Willemyns, 1988). Aso-calledFlemish Movement startedup almostimmediately and fought a long-lastingbattle for cultural and linguistic rightsfor Dutch speakers.It took until 1889for the ‘gelijkheidswet ‘todeclare Dutch andFrench the twoofficial languagesof the country. Afterwardsthings developed faster:two sets of lawsin 1932and 1963 guaranteed what had been the ultimategoal of the Flemish Move- ment,i.e. the officialand complete ‘Dutchification’of Flanders.The Walloons havingbeen opposedto widespread bilingualism throughoutthe country, Belgium graduallyturned tothe territorialityprinciple modelto accommodate itsvarious linguistic groups.It officialised the language frontieras adomestic administrativeborder, made it virtually unchangeable andaccomplished the linguistic homogeneity of the language groups and regions. The Belgian language struggle hasnever been anexclusively linguistic problem but hasalways been intertwined withsocial and political issues as well. Yet,a considerablechange in natureis to be discerned fromthe early1960s onwardwhen the language problemswere replaced by so-called‘ community problems’and the borderbetween Walloniaand Flanders ceased to be amere linguistic one in orderto become asocialone aswell. Thiscan be accountedfor by majordomestic economic changes. From the late1950s onwards a dramatic industrialdevelopment waswitnessed in Flanders,turning thisformally agri- culturalterritory into a highly industrialisedregion, largely dominatingthe domesticpolitical, social and economic scene. Atthe same,time the outdated industrialequipment ofWallonia was slowly breaking down,giving wayto a seriouseconomic recession from which it has not yet recovered.In 199674.5% of the industrialgross added value wasgenerated inthe Flemish region (58%of Belgium’s population).Consequently, the culturaland linguistic balanceof powershifted towardsFlanders (Willemyns, 1992).The present-day socialand economicimbalance between Flanders,Brussels and Wallonia is to be consid- ered potentiallydisruptive forthe continuationof Belgium’s existence, since it 38 Journal of Multilingual and Multicultural Development requires aconsiderableamount of so-called‘ solidaritytransfers’ from Flanders toWallonia (for 80%)and from Flanders to Brussels (for 20%).Most of these transfers occur in the field of social security financing (De Boeck, 1999). Revisionsof the constitutionin 1970and 1980 provided for cultural autonomy anda considerableamount of self-determinationfor the linguisticallydivided partsof the country.Subsequent constitutionalchanges in 1988and 1993 finally turned Belgium into the federal country it is now (Alen & Suetens, 1993).

3. Language Borders The notionof ‘ language border’which is essential in thispaper andin this volume isnot easy to define. Asis often the casewith, for example, the related notionof ‘ dialectborder’ , itmighteven beargued thatlanguage bordersdo not actuallyexist, since itis obvious that language areasare but seldomseparated by aclear-cutline. Usually,there issome kind oftransitionalzone between them, anda demarcationline, therefore, will alwayshave a somewhatarbitrary char- acter.Moreover, it is obviousthat in transitionzones a socialvariable, rather than ageographic one,may be decisive forlinguistic ‘ affiliation’. Dialect-geographers arevery familiarwith such problems and to cope withthem they tend tomake use notso much ofatheoreticalbut ofapracticalsolution, which may differ from one region toanother.In thispaper too,the variouskinds of language contact under investigationwill be decisive forthe particularuse whichis made of the concept of ‘language border’. Yet,both in ahistoricaland a contemporarysense, it may be necessaryto refer towhatGoossens (1968) calls an ‘intuitive consensus’on language borders.In the caseof French-Flanders forexample, there isa general consensusamong scholars toconsiderthe isoglossused in dialect-geographicstudies as the language border between the Romanceand the Germanicdialects in the region (it isreproduced in,among others, Pé e, 1957).The sameapplied toBelgium up to1963, the yearin which the language borderwas laid down by law.From then onwardthe notion of‘language border’is used in asociolinguisticsense, meaning thatit separates tworegionsin which either Dutch orFrench isthe officiallanguage, disregarding anypossible bilingual communicationwhich mayactually occur in the transition zone:the language bordercoincides with the borderseparating two administrative entities. Asregards the changeswhich are discussed in thispaper, twoessentially different types haveto be discerned (Willemyns, 1996):(1) language shift resulting in achange ofthe locationof the border,meaning thatplaces which used tobe partof the transitionzone have, in the courseof time, definitely moved intothe monolingualzone on either side of the border;(2) language shift resulting in ‘erosion’, meaning thatthe contactsituation has decisively been changed in the courseof history although the ‘language border’(in the tradi- tionalsense) hasnot changed itscourse. Since in Belgium the constitutional notionof language bordernot only refers tothe demarcationlines between monolingualterritories but alsoto the demarcationlines between monolingual andofficially bilingual zones,‘ erosion’is also used tocharacterise a decisive change of the situation within bilingual zones. Dutch-French Language Border in Belgium 39 4. Status and Function of the Language Border in Belgium

4.1 General observations Although the language borderhas existed for centuries, no solidinformation priorto the 19thcentury isavailable as faras the territoryof present-day Belgium is concerned. In 1846,the recently establishedKingdom of Belgium startedconducting censusesincluding aquestionon language usage which provided statistical informationuntil 1947(De Metsenaere, 1998).For various reasons the informa- tiongathered this way isoften inaccurate:the exactwording of the questionswas changed fromone censusto another and, more importantly, two basic require- mentsfor reliable informationgathering, namely honestintentions and scientific support,were hardlyever met,as hasbeen convincingly demonstratedby Gubin (1978). The mostimportant insight yielded by the firstcensus (1846) is that the admin- istrativedivision of the countryinto provinces, ‘ arrondissementen’(counties) andeven communeshad been carriedout without taking into account the language borderand had never intended toprovide formore or lesslinguisti- callyhomogeneous administrative entities. On the otherhand, the information yielded perfectly allowedto draw a language mapshowinga borderline neatly separatingthe French- (i.e. Walloondialect) speaking andthe Dutch- (i.e. Flemish dialect)speaking communities.For almost a century (andin spite ofthe deficient methodology)there were (with the exception ofBrussels)no significant differences fromone censusto another (Martens, 1975), a factdemonstrating the remarkable stability of Belgium’s linguistic communities.

4.2 The territoriality principle Adramaticchange occurredfrom 1932 onward, the yearin whichthe language borderbecame apoliticalissue. A century ofstruggle by the Vlaamse Beweging [‘Flemish Movement’] in favourof the promotionof Dutch in a countryup tothen dominatedby French speakershad finally resulted in exten- sivelinguistic legislationbringing aboutthe de facto acceptanceof the territorialityprinciple (McRae,1975), which implied thatFlanders was to be governed exclusively in Dutch andWallonia exclusively in French. Toimple- mentthis decision, though, a precise legal descriptionof the delimitationof these territories,in otherwords of the language border,was needed. Although the 1932laws did notprovide sucha description,it held aprovisionthat communes witha linguisticminority of atleast30% were tobe governed bilingually and that,should a minoritybecome the majority,the linguistic statusof the commune wasto change accordingly.This seems to be fairenough, were itnot that the only meansof acquiring the informationneeded wasthe censuswhich thus, unfortu- nately, acquired important political significance. The firstcensus with these politicalimplications was scheduled for1940 but waspostponed because ofWorldWar II andwhen in 1947it was finally carried outit resulted inanoutburst of politicalcommotion. Contrasting heavily with the stabilitythe returnshad shown for more than a century, itappeared thatthis timenot only notoriousshifts were registered but thatthey allwent in the same direction:many Dutch-speaking villagesappeared toharbourso manyFrench 40 Journal of Multilingual and Multicultural Development speakers,that they turned intoeither bilingual oreven French-speaking communes(Martens, 1975). Since fraudulent manoeuvringby (localand/ or national)authorities was very apparent,the Flemish reactionwas extremely vigorousand the governmentwas finally forcedto skip language questionsfrom future censusquestionnaires altogether and to lookfor a politicalsolution which might,once and for all, determine the language borderbetween the communities.A lawto thiseffect cameinto being on1September 1963and since itsunder- lying philosophy wasto produce linguisticallyhomogeneous administrative entities,several adjustments had to be made,transferring 25 communes with 87,450inhabitants from Flanders to Wallonia and 24 communes with 23,250 inhabitantsfrom Wallonia to Flanders (detailed informationin Martens,1975 and Deweerdt, 1998).

4.3 Problem areas Anotherprovision of the 1963law was the instalmentof communes with so-called‘ faciliteiten’[linguistic facilities](De Schryver, 1998),meaning thatif a communityharboured a considerablelinguistic minority (on 1September 1963), provisionswere tobe madeenabling thisminority to communicatein itsown language withcommunal authorities and to obtain limited possibilitiesfor instructionin itsown language. Thisstatus was allottedto arestrictednumber of communeson bothsides of the language border.The majorprovision, though, meantto put mindsat rest, was that after 1 September 1963,changes in the linguisticstatus of communesand provinces became virtuallyimpossible and couldonly be brought aboutthrough a very complicatedprocedure ofchanging the constitution.On top of atwo-thirdsmajority required forany constitutional change,those with ‘ linguisticimplications’ require amajoritywithin both language factionsof the Belgian parliament.Almost everywhere thispeace of mind wasindeed brought about;two notorious exceptions, the so-called ‘Voerstreek’and the Brusselssuburban region (the so-called‘ Randgemeenten’) will now be treated in some more detail. 4.3.1 The Voerstreek The Voerstreek ispart of the so-called‘ Landvan Overmaas’ , asmallterritory situatedbetween the majorcities of Aachen (Germany), Maastricht(The Nether- lands)and Liè ge (Wallonia,Belgium). During the ‘ancien régime’ Dutch wasthe language ofinstruction and administration in the whole Overmaasterritory (Goossens,1998), but subsequently botha ‘Germanifying’and a ‘Frenchifying’ tendency hadbecome apparent.Also, it has always been almostimpossible to distinguishbetween ‘Dutch’and ‘ German’dialects in the region, onthe basisof purely linguistic criteria (Nelde, 1979: 41). Fromthe end ofWorld War Ionward,when the Eupen region,a formerPrus- sianpossession, was annexed by Belgium, the 17‘Overmaas’communes were partof three different linguistic regimes (Goossens,1998). Two of them (called Altbelgien and Neubelgien)will notbe dealtwith here, since they areanalysed by Nelde and Darquennes in this issue. Thisleaves us with the sixcommunes of theVoerstreek whichgained political celebrity anda wretchedreputation during recent decades(Murphy, 1988). Subject toboth the Frenchificationprocess which also affected the other Dutch-French Language Border in Belgium 41

communesin the region andto Walloonimmigration, they turned outa linguisti- callymixed region witha French-speaking minority.For that reason the 1963law transferredthem from the Walloonprovince ofLiège tothe Flemish province of Limburg. Since, however,they arenot geographically linked tothe latterprov- ince andare/ were dependent, forvarious economic functions, on Liège, someof itsinhabitants were notvery happy withthis transfer. This uneasiness has been exploited by Walloonactivists, causing political commotion ever since.Yet, both Flemish determinationand recent changesin the constitutionwhich put the ‘Voerstreek’firmly under the authorityof the autonomousgovernment of Flan- dersseem tohave lessened politicaltension a gooddeal. Wynants (1980) explains why,even forautochthonous inhabitants, the establishedtriglossic situation of old– Dutch dialectfor informal communication; Standard Dutch in primary schoolsand in church;French in secondaryeducation and part of the administration– wasfinally disturbed andeventually changed because ofpolitical attitudes.From 1964 to 2000 the politicalfaction advocating a return tothe Francophoneprovince of Liège managedto secure an,ever diminishing, majorityin the localcity council. As a resultof the latestcommunal elections (October2000) though, the oppositefaction came into power. Consequently, a return toLiè ge isno longer onthe politicalagenda of Voeren’s citycouncil and of the majorityof itsinhabitants. A recent andextensive analysisof the present situation as well as its genesis is to be found in Vandermeeren (1996). 4.3.2 The Brussels suburban region The 1963law also affected the statusof somesuburbs inthe Brusselsregion where the officially bilingual territoryis restricted to 19 communes which togetherconstitute Brussels as a politicalentity. Ongoing Frenchificationof Brusselswhich isdiscussed below, also affected someof itssuburbs. Mostly because ofimmigrationof French speakersbut alsopartly because ofupward socialmobility behaviour affecting partof the autochthonouspopulation some ofthese communeslost their formerexclusively Dutch-speaking characterand pressure wasput onconsecutive governments to annex themto bilingual Brussels(Sieben, 1993).Yet, surrendering toFrancophonedemands was politi- callyunfeasible and,afterwards, made constitutionally impossible. Some of these suburbs, though,appeared toharbournot only importantFrench-speaking minoritiesbut in somecases even de facto majorities.Six ofthem, Drogenbos, Kraainem,Linkebeek, Sint-Genesius-Rode, Wemmel andWezenbeek-Oppem, officially received a‘faciliteiten’-system (De Witte1975; Witte 1993a) but remained partof Flandersand, consequently, officially Dutch-speaking. This waythe riskof Francophone overspill to otherthan these sixcommunes had been considerablydiminished andrecent evolutionshows a significantdecrease of Francophoneinfluence in allof the hinterlandcommunes, as wasdemonstrated in researchby Deschouwer andMariette (1993). The mostrecent constitutional change of1993 also provided forthe splitof the province ofBrabantas from1 January1995, cutting the Brusselsperiphery forgood from the capitalitself (Detant,1998). This may very well bring toa conclusionin the nearfuture a processwhich isto be discerned allalong the language border,namely the increasinghomogeneity ofthe language territorythrough assimilation of minority language islands. 42 Journal of Multilingual and Multicultural Development

4.4 Pattern of fluctuation Whatwe havebeen witnessingin the ‘Voerstreek’and the Brusselshinterland isnot a fluctuationof the language border,but anerosion within Flemish borders,due tothe sociologicalmechanism of ‘upwardsocial mobility and inte- gration’and to Francophone immigration. Actually, the villagesmentioned earlier arethe only incidence ofplacesreally shifting fromone communityto the other.

4.5 Language contacts across the border Asfar as linguistic parallels and mutual influence onboth sides of the linguisticborder are concerned, some research has been done in the courseof time,but the subject hasnever been extremely popular.The influence ofFrench onthe evolutionof Dutch isundeniable anda logicalconsequence ofa language incontactsituation which hasexisted for centuries. The oppositephenomenon, althoughless prominent, has been establishedas well. An overviewof research inbothfields isto be found in Willemyns (1997a).There isalso a (limited) tradi- tionof researchinto mutual influence ofdialectsand dialect phenomena in Flan- dersand Wallonia, most of it conducted before WorldWar II. In morerecent timesA. Weijnen isone ofthe mostprominent names in thisfield (e.g. Weijnen, 1964).A recent overviewwith a discussionof casestudies and possible explana- tionsfor similar linguistic developments onbothsides of the Flemish-Walloon dialect borders is given in De Schutter (1999). Bothcases mentioned pertainto the field ofhistorical linguistics, that is, the studyof mutualinfluence due tothe extremely long durationof language contact anddemonstrable in language change thatoccurred a long timeago. The same goesfor so-called ‘ gallicisms’in Dutch and‘ flandricisms’in French, asthey are spottedand denounced by puristson bothsides of the border.Those too have been existingfor ages but itis notknown to whatextent the phenomenon isstill productive. Apartfrom the bilingual Brusselsregion (see Treffers-Daller in thisissue) there hasnot been, tomy knowledge, anysubstantial research on linguistic contact phenomena in the present time.

5. Brussels

5.1 General observations Language shiftin Brussels,spectacular though itmay have been, isalso a shift notaffecting the language borderas such. Here too,the patternof fluctuation showsan erosionwithin the Dutch linguistic territoryeventually rendering acity whichused tobe partof Flanders ( àpartentiè re )intoa bilingual citywith Francophonedominance. From a judicial pointof view,though, a shiftdid never- thelessoccur, since Brusselschanged itsstatus from ( de facto)monolingualDutch into (de jure)bilingual. Yet,there isnot and there cannotbe ananswer to the apparently simple questionof how many speakers are to be attributedto the Dutch orFrench speaking groupsrespectively (BaetensBeardsmore, 2000). The portraitof Brusselsis one ofimmense complexity(Mackey, 1981)involving not only linguistic backgroundand competence but alsoattitudes, social status, job Dutch-French Language Border in Belgium 43

conditions,circumstances of discourse,feelings towardsthe interlocutor,etc.; in awordall of the sociolinguisticvariables which are known to determine linguistic interactionin multilingual settings(Willemyns, 1997and Treffers-Daller in this issue).

5.2 Language shift Until the 19thcentury Brusselsshared its linguistic fate with other Flemish cities:an importantpart of the socialelite wasbilingual andused French formost ofthe traditionalculture language functions.The competence ofthe majorityof the populationwas restricted to aDutch dialect;other varieties of Dutch were only atthe disposalof the few (De Vriendt &Willemyns, 1987).This situation existedin mostFlemish citiesand there wasneither atypologicalnor a quantita- tivedifference between the situationin Brusselsand in othercities such as Antwerp,Ghent orBruges. Consequently, since Frenchificationwas stopped andeventually reversed in Flandersbut notin Brussels,an explanationcan only be found in factors specific to the Brussels situation. 5.2.1 Historical development Ever since the startof the Burgundian period in the 15thcentury (De Vries et al.,1995:50 ff.), Brusselshas been acapitaland consequently the number ofcourt- iers,noblemen andinfluential governmentofficials and civil servants has always been largerthan elsewhere. Itis precisely in these groupsof people thatthe influence (andusage) ofFrench hasalways been the mostimportant (Witte, 1988, 1993). Frenchificationafter the annexationby Francewas more intensive here than elsewhere in Flanders,not the leastbecause of the presence of aninfluential group ofFrench immigrants(Deneckere, 1954).After 1830,Brusselsemerged asa symbolof Belgium andhere the ‘one country,one language’principle appeared tobe moreappropriate still than elsewhere. The stronglycentralising Belgian policy,moreover, resulted in adisproportionalhigh concentrationof the country’s financialand industrial power in the Francophone‘ milieu’of the capital. Since powerand wealth essentially derived fromWalloon industry it is hardly surprising that the elite particularly favoured this region and its language. Atrathershort notice Brussels became apole ofattractionto numerousimmi- grantsfrom both the Dutch andthe French-speaking partsof the country.In Brusselsas well asin itssuburbs there wasan explosion of the population. Between 1830and 1840 the populationquadrupled (De Metsenaere &Witte, 1990:3). Flemish immigrantsmostly consisted of lower-classand poor people, whereasWalloon immigrants mostly consisted of upper-working-class and middle-classpeople (De Metsenaere, 1988).The latterimmediately fortified the Francophonepopulation. As to the former:‘ the pressure fromthe topsocial stratumto adoptits French language filtered downthrough the middle-classes andfrom them into the “labouraristocracy” of skilled workers,but generally stopping shortat the lowestcategories of servicepersonnel andday labourers, madeup toa largeextent in the 19thcentury ofFlemish immigrantsto the capital’ (BaetensBeardsmore, 1990: 2). Consequently, until farin the 20thcentury being Flemish (andspeaking Dutch) used tobe associatedwith being pooror even being socially and culturally retarded. 44 Journal of Multilingual and Multicultural Development

An additionalhandicap for both immigrant and autochthonous Dutch-speaking ‘Brusselers’(the English language seemsto haveno appropriate termto designatethe ‘inhabitantof Brussels’; ‘Brusseler’will be used tofill this gap) wasthat their habituallanguage wasa dialect,i.e. a varietywith a very limitedsocial prestige. Consequently, the majorityof the lowermiddle and workingclasses tried to acquiremastery in the only language whichappeared to makeupward social mobility at all possible. Hence, the attractivenessof the French educationalsystem was immense inaperiod ofrapid development of mass education. The unprofessionaland fraudulent censusesin Brussels(Gubin, 1978)showed anenormous increase of the ‘statistical’amount of allegedly French-speaking inhabitants, and the judicial consequences of censuses were very real. 5.2.2 Linguistic legislation An additionalreason why the development in Brusselswas different from Antwerp orGhent wasthe factthat either mostof the linguistic legislationdid notapply toBrussels or had to be paidfor by concessionsintensifying the Frenchificationof the capital.This situation changed assoon as the major struggle in Flanderswas over and the Flemish Movement couldstart paying attention to the capital as well (Witte & Van Velthoven, 1998). The turning pointappears to have been when Flemings agreed togive upthe advantagesof their numericalmajority in the countryat largein favourof parity in administrationfor Brussels. This implied thatDutch-speaking ‘Brusselers’, even afterhaving become aminoritygroup, were nevertheless allottedhalf ofthe high ranking civilservants in the administrationof Brussels’ 19 communes (Willemyns, 1997). Several measurestaken on the level ofthe nationalgovernment guarantee Dutch speakersin Brusselsa positionon allkinds of levels, whichthey never couldhave extorted by virtue oftheir solenumerical strength. A very eloquent example istobe found in the schoolsystem. Although the number ofpupils in the Dutch schoolsystem had very much deterioratedin the 1950sand 1960s, a combinationof measures accounts for a constantincrease of the populationof Dutch schoolsfrom the late1970s onward (Baetens Beardsmore, 1990), as opposedto the decreaseof the schoolpopulation in the countryat largeand in French schools in Brussels in particular. Othermeasures, then, haveentailed consequences whichwere completely unpredictable. The reinstalmentof the so-called‘ freedom ofthe headof the family’, meaning thatDutch-speaking familiescould choose French education fortheir children andvice versa,was very much feared by supportersof the Dutch causein Brusselswho expected language shiftto be increasedby it. Completely unexpectedly, itappears to be the ‘heads’of French-speaking fami- lies who,to anever increasingextent, use their ‘freedom’to choose Dutch educa- tionfor their children. Consequently, pupils in the Dutch schoolsincreasingly originatefrom linguistically mixed orhomogeneously French-speaking house- holds (D’hondt, 1999). Also,as Baetens Beardsmore (1990: 5) points out, Flanders’ increasing economicresources made it possible toput up structuresin Brusselswhich enabled ‘the individualto function asamonolingual.Schools, hospitals, welfare Dutch-French Language Border in Belgium 45

services,cultural instances, recreational facilities have all been setup toservice either communityin itsown language. Hence the institutionalpressures to Frenchificationhave been eliminated and… the minorityspeaker (wasenabled) tomaintain his ethnolinguistic identity’ . Itenabled him alsoto profitmaximally from the gain in prestige the language had acquired in the country at large. Aninternalfactor which has to be stressedis the shiftin linguisticbehaviour andattitudes away from dialects in the directionof the standardlanguage. Amongsupporters of the Dutch causethere hasalways been ageneral awareness thatmeeting the French challenge wasonly possible by increasingthe impor- tanceand usage of Standard Dutch, the only variety(if any) able toequal the socialprestige ofFrench. Thisawareness was particularly strong among inhabit- antsof the bilingual Brusselsregion. Inquiries havedemonstrated that here indeed, the shiftfrom Dutch dialectsto Standard Dutch startedearlier andhas been moremassive than elsewhere in Flanders(De Vriendt &Willemyns, 1987: 224–225). This factor is tobe addedto those mentioned before andis to be consid- ered one ofthe mostimportant contributions from the Dutch-speaking ‘Brusselers’ themselves to the dramatic change of the Brussels linguistic scene. Finally, the rapidly expanding populationof foreign originaccounts for the factthat for probably one-third of the capital’s citizensnone of Belgium’s languagesis their mothertongue. Yet,for the overwhelming majorityof those, French is their first ‘Belgian’ language.

6. Typology of language border change The analysisof these variousand differing instancesof language shiftwill help toexplain myinitialassumption concerning the ‘language border’concept. Historyproves that demarcation lines between dialectsof different languages canremain remarkably stable over centuries, but alsothat changes in the political and/orsocial constellation may account for dramatic alterations leading to language shiftand eventually language loss.The pointthat I wouldlike to emphasisethough, is that only (consciousor semi-conscious)language planning initiativesmay entaildurable, irreversible change.Planned politicalinterference ofvariouskinds (and linguistic legislation is only one ofthem) accountsfor a processof sociallydetermined shift,resulting in language erosionon one side of the language borderand eventually in complete loss.Since language planning initiativesmay be ofacontradictorynature – either toencourage the ‘offensive’ language ortosupport the ‘defensive’one –politicaland sociological factors will decisively determine the eventual outcome.The factthat language planning methodsin Franceand in Belgium areso completely different in naturewill help to clarify that matter.

6.1 The French type Franceis an officially monolingualcountry where French isthe only official language. Eversince the French Revolutionconstant and determined pressure hasbeen put onthe ‘allophone’regions to accommodate, both officially and privately,to the officialpolicy. This had led toamassiveshift of whichFrench Flandersand Alsace are only afew examples(Bister-Broosen, 1998; Ryckeboer, 1997).As aconsequence ofthe monolingualassumptions of the centralgovern- 46 Journal of Multilingual and Multicultural Development mentlanguage borderswere neither protectednor was their mere existence acceptedin anylegal orofficial way and so there hasbeen noofficial protection or even function forminority languages in France;their usagehas, on the contrary, constantlybeen fought andbanned. Consequently they havedeteriorated everywhere, in French Flandersalmost to the pointof complete extinction,emptying the stillexisting ‘ language border’of almost all practical relevance nowadays. Yet,even in thisvery restrictedsense the language boundaryhas shifted consid- erably in anortherndirection over the lasttwo centuries. A quantitativelyand qualitativelyvery restrictedform of bilingualism and/ordiglossia, and what Dorian(1982) calls ‘ semi-speakers’is all that has survived the aggressive language planning activities(namely linguistic legislationand social pressure) devised by the French government and establishment.

6.2 The Belgian type The evolutionin Belgium hasbeen completely different, mainlybecause ofa very different historical evolution: Priorto 1794 (annexation of the Belgian territoriesby France) there has never been aconsistentlinguistic policy (mainly because there hasnever been a central government!) . Boththe French (up to1814; Deneckere, 1954)and the King ofthe ‘United Netherlands’(from then till1830; De Jonghe, 1967)legislated on linguistic mattersin the ‘one country– one language’-sense, albeitit with completely different intentions. The ‘founding fathers’of independent Belgium meantto appease linguistic unrestby constitutionallydeclaring ‘ the use ofthe languagesoptional’ . In a nationdominated by anindustrialisedand powerful Walloonpart and a mainlyFrench-speaking Flemish nobility andbourgeoisie, thismeant the perpetuation ofthe dominanceof French overthe majority,i.e. the Dutch speaking, yet politically powerless Flemings. During allthis time the internallanguage borderbetween Dutch andFrench continued toexistpractically unchanged andunchallenged since itwas simply aninformalline ondialectologists’maps having no official or politicalimplica- tionswhatsoever. The struggle ofthe so-calledFlemish Movement forcultural andlinguistic rightsfor Dutch speakersgradually changed the picture. Bitterly fought, yet only graduallyimplemented linguisticlegislation resulted, in the 1930s,in the de facto acceptanceof the territorialityprinciple legally acknowl- edging the existence oflanguage communities.The nextstep, consequently, had tobe the officialdelimitation of these communities;in otherwords the official determinationof the language border.Subsequent constitutionalreforms finally transformedBelgium intoa federal statewith regional governments having extensive legislativepower within their territoriesconfined by language borders.These borderswere laiddown in the constitutionand made virtually unchangeable. Consequently, eachBelgian townor village hasbeen allotteda specific linguisticstatus and the officiallanguage ofeach individual isnot a matter of personal choice but of the territory she or he lives in. Dutch-French Language Border in Belgium 47

6.3 Patterns of change The changesthat have occurred can be classifiedinto different types or patterns: (1) ‘Monolingualisation’of formerly bilingual orbicultural villages as is the case of some of the language border communities. (2) ‘Bilingualisation’of formerly mostlymonolingual villages, i.e. ‘Frenchification’, forexample, in the bordervillages of Brussels, some of which haveindeed been annexed tothe Brusselsbilingual communityin the course of time. (3) A specific evolution in Brussels itself, mainly of the type mentioned in (2). Politicalevolution in recent decadeshas stabilised the language borderand made drastic changes virtually impossible in the future. It appears that: changeshave become ‘definitive’over time mainly by securing the linguistic homogeneity of administrative entities; ongoing shifthas been frozenby firmly embedding shifting villagesinto a monolingualcommunity. Investigations have shown that Frenchification notonly seemsto havestopped but isbeing slowlyreversed (Deschouwer & Mariette, 1993). Asa resultwe arenow in the presence ofa firmly monolingualisedBelgium, divided intoautonomous communities based on linguistic homogeneity and determined toreduce the politicalconsequences of language contactat their bordersto astrictminimum. In sodoing the language borderhas become the mostimportant internal boundary to which all pre-existing administrative delimitationswere subordinated.The only interesting place left (froma socio-linguisticpoint of view) isBrussels. The principal ofterritorialitydoes not apply here; there isno official demarcation line between speakersof both languagesand no way ofofficially controllinglanguage contactor potentialshift either. In Belgium atlargethe language borderis no longer amere linguisticnotion but alegal,administrative and political reality. This evolution has completely changed the natureof the coexistenceof the country’s variouslinguistic communities– firmly embedded in their ownmonolingual structures – andhas also demonstratedhow decisive the implicationsof language planning activitiescan be.

Correspondence Any correspondenceshould be directedto ProfessorRoland Willemyns, Vrije UniversiteitBrussel, Germaanse Talen, Pleinlaan 2,B-1050 Brussels, Belgium ([email protected]).

References Alen, A.and Suetens, L.P.(eds) (1993) Hetfederale België na devierde staatshervorming. Brugge: Die Keure. Baetens Beardsmore, H.(1990)The evolution and current statusof Brussels asa bilingual city.In H.Baetens Beardsmore Bilingualismin Education:Theory and Practice. Brussels Pre-prints in Linguistics 11 (3). 48 Journal of Multilingual and Multicultural Development

Baetens-Beardsmore, H.(2000)Bruxelles. In W.F.Mackey (ed.) Espacesurbaines et coexistencedes langues .Terminogramme 93–94 (pp.85– 102). Qué bec: Les Publications du Québec. Bister-Broosen, H.(1998) Sprachkontakteund Sprachattitüden Jugendlicher im Elsaßund in Baden. Frankfurt: Peter Lang. De Boeck, E. (1999)Regionale groeiverschillen in België. EconomischFinancië le Berichten KBC, 24 September 1999. De Jonghe, A. (1967) De taalpolitiek van Willem I. Sint-Andries: Darthet. De Metsenaere, M.(1988) Taalmuur:Sociale muur? De19de-eeuwse taalverhoudingen in Brussel. Brussel: VUPress. De Metsenaere, M.(1998)Talentellingen. In NieuweEncyclopedie van deVlaamseBeweging (pp. 2950–2956 ). Tielt: Lannoo. De Metsenaere, M.and E.Witte (1990)Taalverlies en taalbehoudbij de Vlamingen te Brussel in de negentiende eeuw. BMGN 105 (1), 1–38. Deneckere, M.(1954) Histoirede lalangue franç aise dans lesFlandres. Gent: Rijksuniversiteit. Detant, A.(1998)Randgemeenten. In NieuweEncyclopedie van de Vlaamse Beweging (pp. 2536–2547). Tielt: Lannoo. Deschouwer, K.and Mariette, D.(1993)De westerse migranten in Brussel en in de rand. In E. Witte (ed.) De Brusselse rand (pp. 423–449). Brussel: VUBPress. De Schryver, R.(1998)Faciliteiten In NieuweEncyclopedie van de Vlaamse Beweging (pp. 1105–1107). Tielt: Lannoo. DeSchutter,G.(1999)Fonologischeparallellenaanweerszijdenvande Germaans-Romaanse taalgrens. Taal en Tongval 51, 111–130. De Vriendt, S.and Willemyns, R.(1987)Linguistic Research on Brussels. In E.Witte and H. Baetens Beardsmore (eds) TheInterdisciplinary Study of Urban Bilingualism in Brussels (pp. 195–231). Clevedon: Multilingual Matters. De Vries, J.,Willemyns, R.and Burger, P.(1995) Hetverhaal van een taal. Negen eeuwen Nederlands. Amsterdam: Prometheus. Deweerdt, M.(1998)Taalgrens. In NieuweEncyclopedie van de Vlaamse Beweging (pp. 2956–2962). Tielt: Lannoo. De Witte, K.(1975)Randgemeenten In Encyclopedievan de Vlaamse Beweging (pp. 1286–1299). Tielt : Lannoo. D’hondt, A-S.(1999) Taalproblematiek in het Brusselse onderwijs. InP.Frantzen (ed.) NederlandstaligeBrusselaars in eenmulticulturele samenleving (pp.151– 172). Brussel: VUBPRESS. Dorian, N.(1982)Defining the speech community toinclude its working margins. InS. Romaine (ed.) SociolinguisticVariation in SpeechCommunities (pp.25– 33). London: Edward Arnold. Goossens, J. (1968) Wat zijn Nederlandse dialecten? Groningen: Wolters Noordhoff. Goossens, J.(1998)Overmaas. In NieuweEncyclopedie van de Vlaamse Beweging (pp. 2367–2368). Tielt : Lannoo. Gubin,E. (1978)La situation des langues àBruxelles auXIXme siècle àlalumiè re d’un examen critique des statistiques. Taal en Sociale Integratie 1, 33–79. Lamarcq,D. and Rogge, M.(eds) (1996) Detaalgrens. Van deoude tot de nieuwe Belgen . Leuven: Davidsfonds. Mackey,W.F. (1981) Urban language contact:Common issues in Brussels and abroad. Taal en Sociale Integratie 3, 19–38. Martens, P.(1975)Taalgrens. In Encyclopedievan de Vlaamse Beweging (pp.1552– 1562). Tielt: Lannoo. McRae,K. (1975)The principle ofpersonality and the principle ofterritoriality in multilingual states. International Journal of the Sociology of Language 4, 35–54. Murphy, A.(1988) TheRegional Dynamics of Language Differentiation in Belgium. University of Chicago: Geography Research Paper 227. Nelde, P.H.(1979) Volksspracheund Kultursprache.Die gegenwä rtige Lage des sprachlichen Übergangsgebietesim deutsch -belgisch-luxembur gischenGrenzraum. Wiesbaden: Zeitschrift für Dialektologie und Linguistik, Beihefte 31. Dutch-French Language Border in Belgium 49

Pée, W.(1957) Anderhalveeeuw taalgrensverschuiving en taaltoestandin Frans-Vlaanderen. Amsterdam: Bijdragen en Mededelingen van de Dialectencommissie 17. Ryckeboer, H.(1997) HetNederlands in Noord-Frankrijk.Sociolinguï stische, dialectologische en contactlinguïstische aspecten . Gent: Universiteit. Sieben, L.(1993)Gemeenten opde wip tussen Vlaanderen en Brussel. Een bijdrage totde ontstaansgeschiedenis van de randgemeenten (1900–1940). In E.Witte (ed.) De Brusselse rand (pp. 145–166). Brussel: VUBPress. Van de Craen, P.and Willemyns, R.(1988)The standardization ofDutch in Flanders. The International Journal of the Sociology of Language 73, 45–64. Vandermeeren, S.(1996)Language attitudes on either side ofthe linguistic frontier: A sociolinguistic survey in the Voeren/Fouron-areain Old Belgium North. In U.Ammon and M.Hellinger (eds) ContrastiveSociolinguistics (pp.157– 174). Berlin-New York: W. de Gruyter. Weijnen, A.(1964)Fonetische en grammatische parallellen aanweerszijden van de Frans-Nederlandse taalgrens. Tijdschrift voor Nederlandse Taal-en Letterkunde 80, 1–15. Willemyns, R.(1988)Belgium. In U.Ammon, N.Dittmar and K.Mattheier (eds) Sociolinguistics (pp. 1254–1258). Berlin-New York: Walter de Gruyter. Willemyns, R.(1992)Linguistic legislation and prestige shift. In U.Ammon and M. Hellinger (eds) StatusChange of Languages (pp.3– 16). Berlin-New York: W.de Gruyter. Willemyns, R.(1996)Language borders in Northern France and in Belgium: Acontrastive analysis. In U.Ammon and M.Hellinger (eds) ContrastiveSociolinguistics (pp.229– 249). Berlin-New York: W. de Gruyter. Willemyns, R.(1997)Toward aplurilingual urbanenvironment: Language policy and language planning in Brussels. In M.Pütz (ed.) LanguageChoices. Conditions , Constraints, and Consequences (pp. 179–193). Amsterdam: John Benjamins. Willemyns, R.(1997a)Interferenz Niederländisch-Franzö sisch. In H.Goebl,P. Nelde, S. Zdenek and W.Wölck (eds) Kontaktlinguistik.Ein internationalesHandbuch zeitgenössischer Forschung (pp. 1123–1130). Berlin and New York: W. de Gruyter. Witte, E.(1988)Bilingual Brussels asanindication ofgrowing politicaltensions. InE. Witte and H.Baetens Beardsmore (eds) TheInterdisciplinary Study of UrbanBilingualism in Brussels (pp. 47–74). Clevedon: Multilingual Matters. Witte, E.(1993)De plaatsvan Brussel in het gefederaliseerde België. Internationale Spectator 47, 560–565. Witte, E.(1993a)Faciliteiten voor taalminderheden in de Brusselse rand. InE.Witte (ed.) De Brusselse rand (pp. 168–209). Brussel: VUBPress. Witte, E.and van Velthoven, H.(1998)Taalpolitiek en -wetgeving. In NieuweEncyclopedie van de Vlaamse Beweging (pp. 2994–3043). Tielt: Lannoo. Wynants, A.(1980)Taalovergang en taaltrouwin de Voerstreek sinds de aansluiting bij de provincie Limburg. InP.Nelde (ed.) Sprachkontaktund Sprachkonflikt (pp.467– 474). Wiesbaden: F. Steiner Verlag.