Acquiring Tarifit-Berber by children in the Netherlands and Morocco

Tilburg University

Acquiring Tarifit-Berber by children in the Netherlands and Morocco

E-Rramdani, Y.

Publication date:

2003

Document Version

Publisher's PDF, also known as Version of record

Link to publication in Tilburg University Research Portal

Citation for published version (APA):

E-Rramdani, Y. (2003). Acquiring Tarifit-Berber by children in the Netherlands and Morocco. Aksant Academic

Publishers.

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal Take down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

-IN THE NETHERLANDS

AND MOROCCO

bh l i r l> 51 T F i I O * VAN Ti LBL I'.8 |

r---Sl___ i

BBLIOTHEEK

7... -i

STUDIES

IN MEERTALIGHEID

/

STUDIES

IN

MULTILINGUALISM

The aim of thisseries isto disseminatetheoreticalandapplied linguistic knowledge and results_{of empirical}research inthefieldofindividualand societal_{multilingualism.} Editors

Guus Extra & Ton Vallen

Babylon, CentreforStudiesofMultilingualism in the MulticulturalSociety Tilburg University, TheNetherlands

Editorialboard

RenOAppel, University of Amsterdam Hans Bennis, MeertensInstitute Amsterdam Kees de_{Bot, University}of_Groningen

Kris vanden Branden, University ofLeuven DurkGorter, FryskeAkademy Leeuwarden

Roeland van _{Hout, University of Nijmegen/Tilburg University} lacomine Nortier, Utrecht University

Previouslypublished

No. 1 leanne Kurvers (2002)

Met ongeletterde ogen. Kennis van taal en schrift van analfabeten No. 2 Chefena_{Hailemariam (2002)}

Language and Education in Eritrea. A Case Study of Language Diversity, Policy and Practice

No. 3 _{YahyaE-rramdani (2003)}

ACQUIRING

TARIFIT-BERBER

BY CHILDREN IN

THE NETHERLANDS

AND

MOROCCO

UNIVERSITEIT * * VAN TILBURG BIBLIOTHEEK

- 1 1 ... . r.,n 1EL_L. Ur\U

PROEFSCHRIFT

ter verkrijging van de graad vandoctor

aan deUniversiteitvan_Tilburg, op gezag vande rector magnificus, prof. dr. F.A. van derDuyn Schouten,

inhet openbaarteverdedigen ten overstaan van een door het college voor promoties aangewezencommissie

in de aula vande Universiteit

op woensdag 24september 2003 om 16.15 uur

door_YahyaE-rramdani,

promotor: prof. dr.

G. Extra

copromotor: dr. A. El Aissati

ISBN 90-5260-114-3

© 2003 Aksant Academic Publishers, Amsterdam

All_{rights reserved. No part ofthis publication may}be_reproducedwithout_{permission of the} publisher.

Coverdesign:_JosHendrix, Groningen Lay-out: Anneke Smits and CarineZebedee

Printed andbound inTheNetherlands on acidfree_paper

Aksant Academic Publishers

Table

of contents

Transcription symbols Tarijit ix

Preface xi

Introduction I

Rationale ofthestudy I

Overview ofthestudy 2

I Tbe Status of Tarijit in Morocco and in tbe Netherlands 5 I.I Status

of

Berberin_{Morocco S}

I.I.I Historical background f

I.I.2 Languagesituation in

Morocco 6

I.I.3 Research on Berber acquisition inMorocco 7

I. 2 Status

of

_{Berber in the Netherlands 9}

I.2.I Demography 9

I.2.2 Berber in the Netherlands IO

I.2.3 Researchon Berberacquisition in theNetherlands I3 I.2.4 Research on Berber acquisition inneighbouring countries 4 I.3 Acquisition of morphology andsyntax IS

I.4 The present study 20

2 Design of tbe study 25 2. I Conceptualisation 27

2. I.I Specification of thetargetgroups 27 2. I.2 Language of the test 27

2.I.3 Selection of the task domains 29 2.I.4 Selection of thetaskcontents JI

2.2 Operationalisation 32

2.2. I Plural formation 32 2.2.2 Casemarking 37 2.2.3 Gender-numberdistinction 40 2.2.4 Person

inflection 42

2.2.5 Perfectiveformation 43 2.2.6 Wordorderconstruction 46

vi Acquiring

Tarifit-Berber I

3 Tbe pilot study SI

3.I Design of the pilot study SI

3.I.I Settingand_{informants SI} 3.I.2 Data collection 52

3.I,3 Dataanalysis procedure 53 3.2

Validity

analysis 54

3.2.I Pluralformation 54

3.2.2 Casemarking 58

3.2.3 Gender-numberdistinction 60

3.2.4 Personal inflection 63

3.2.5 Perfectiveformation 63

3.2.6 Wordorderconstruction 66

3.3 Conclusions 70

4 Sociolinguistic projile of tbe children of tbe main study 71

4.I Grade I children 7.r

4.I.I Coregroup in_{the Netherlands 7I}

4.I.2 Referencegroupin Morocco 78 4.2 Grade8_{Children 78}

4.2.I Coregroup in _{the Netherlands 78}

4.2.2 Referencegroupin_{Morocco 84}

5 Plural formation 85

5.I Grade I children 85

S.I.I Taskdescription anddatacollection procedure 85 5.I.2 Dataanalysis 88

5.I.3 Conclusions anddiscussion 96

5.2 Grade8children Ioo

5.2. I Taskdescription anddatacollection procedure Ioo

5.2.2 Data analysis /03

5.2.3 Conclusions anddiscussion III

6 Case marking II7 6.1 Grade I children II7

6.I.I Taskdescription anddatacollectionprocedure 117

6.I.2 Data_{analysis .rIS}

6.I.3 Conclusionsanddiscussion I20

6.2 Grade8_{children I.23}

6.2.I Task description anddatacollection procedure I.23 6.2.2 Dataanalysis I.25

7 Gender and number distinction I3I 7.I Grade I children I3.r

7.I.I Task description anddatacollectionprocedure IJI

7.1.2 Dataanalysis I32

7.I.3 Conclusionsand discussion I35

7.2 Grade8 _{children I37}

7.2.I Task description and data collectionprocedure I37 7.2,2 Data analysis I39

7.2.3 Conclusion and discussion I44 8 Peifective formation I47

8.I Grade I children I47

8.I.I Task description anddatacollectionprocedure I47

8.I.2 Data analysis I49

8.I.3 Conclusionsand

discussion Iii

8.2 Grade8_{children I56}

8.2.I Task description and data collection procedure 156 8.2.2 Dataanalysis I57

8.2.3 Conclusions anddiscussion I61

9 Word order construction I65 9.I Grade I children /65

BI.I

Task description anddatacollection procedure I65

9.I.2 Data analysis I67

9.2.3 Conclusion and discussion I71

9.2 Grade 8 children I74

9.2.I Task description anddatacollection procedure I74 B 2.2 Data analysis I76

9.2.3 Conclusions and discussion I78 Io Conclusions and discussion ISI

IO.I Acquisition

of

morphologicaldevices ISI IO.2 Acquisition

ofword

orderdevices I90 IO.3 LanguageacquisitioninContext I92 IO.4 Perspectivesfor furtherresearch I99

Viii AcquiringTarifit-Berber

Appendices 209

Appendix I: Language testfor grade I children 209 Appendix 2:Language test

of

grade8 _{children 215}

Appendix3: Questionnaireconducted withthe mothers

of

grade I children 223 Appendix4:Questionnaire conducted

with

grade8_{children 229}

Transcription

symbols

Tarifit

The transcription systemfollowed in

writing Tarifit in this

book conforms more or less totheUtrechtconvention

ofI996

(Adrar-NieuwsbriefI997/98).Thetranscription

is phonologically based. As such, words pronounced as _[ada:] (bend down), for example, are written as /adar/.the_{ones pronounced} as _Izudjl(swear) are written as /zull/, and words pronounced as

_Irkazil

_{(window) written} as /lkazi/. Pharyngealized

segmentsarewritten with a dotunderneath as d, s,t. Geminate segments are written as a sequence oftwosegements ascc. Dashes are usedto segmentwordsintovarious morphemes, depending onthenature of the taskunderinvestigation, as in i-dar-en, the plural

ofdar(foot),or

i-kker.theperfective of kker (stand up), with i marking the gender andnumber ofthe subject.

Symbols Example Closs

a aman water b baba _papa c cek _you d din there d dar foot f fafa lookfor

g ge\ n,ved take by hand

7 ayrum bread h hwa go down h henna grandmother i _{ini say} i iull swear k kenniw you (pl) 1 Imeyrib Morocco m mayemmi why n anu a well p paltu overcoat q qawit peanuts r ru cry r tarut lung s sevv drink s asemmid cold t tini dates t batata potatoes k _kamyun loary u uccen wolf

w wenni that one

x xali uncle

y yemma mother

z anzar nose

z anzar rain

Preface

I read once that_{promoveren is een eenzame}zaak_{(doing a PhD study is a lonely job),}

now I realize what this means. And once I was asked whether I was _{capable of} conducting a PhD study me and mycomputer alone, I said yes at the time full of trust, andtoday I submittheproofi

If the work is done, itiSthanks to many people. First of all, I would like to thank the people who made the data available for this study, namely the schools for their cooperation, the mothers and the children in the Netherlands andMoroccofor their

availabilityforinterviews and testing, and thedata collectors MinaEl hajoui, Radia Azghari and Naima Madaya in the Netherlands and Driss Mousaddaq, Rachid Bouzeggou, Mohammed Zianiin Morocco.

Besides mysupervisors Guus Extra andAbderrahman El Aissati are thanked for

theirguidanceduring thisstudy, and fortheir dedication tothe success of this work. Het moet een goed boek worden Grshould be a good book) was the sentence Guus Extra often repeated in thelastyear. For this success, they did more than what they ought to do, eachintheir own way.

Ad Backus, Jan JaapdeRuiter and Otman

Ait

Ouarasseoffered some o

f

their time

toread partsof this book, and to comment onandimprove them. Tim vanderAvoird

wasalways availableformethodologicaldiscussionsandadvice. JeanneKurvers made the effort

of

translating the summaryinto Dutch.CarineZebedeeandAnneke Smits took charge of the lay-out of the book, and

with

success. I ammuchindebted to all of

them.

Colleagues in Babylonarethanked for thegezelligbeidand for makingthe sphere pleasant to work in. Anne Vermeer, ex-office neighbourin building B, isthanked also

forhis readinesstoanswer_myquestions when unexpectedlydropping intohisoffice.

He previewed also one of the articles I published.

Last butnotleast, I expressmygratitude to myfamilytherein Morocco, especially

my mother, whojust couldnotbelieve that Iwasstill going on with mystudies, and

my family here in the Netherlands; my wife

Hetty for taking some of

my family

responsibilities over, and my two sons, AminandKhalil born duringthis period, who had to miss me manytimes; nottoforgetmybrotherMohammed,who

kindly

looked after

Amin

whenever possible, and mydearfriendJaial Hassani forshowing interest

in what Iwas_doing.

It was agreatlearning experience togothrough, and it isagreat pleasuretofinish

Introduction

During the last decennia, a number

of

_{empirical studies have} been carried out on

language use withintheMoroccancommunity inthe Netherlands.These Studies have been directed both towards the acquisition ofand proficiencyin Moroccan Arabic

(Nortier, I989; El Aissati, I997; Bos, I997) or towards the learning and teaching of

StandardArabic in the context

of

Arabic languageinstruction inelementaryschool

(Driessen, I990; Van de Wetering, I990; Saidi, 2001). Very few studies have been conducted

with

_{respect to Berber acquisition andproficiency in} the Netherlands or

Morocco,and especiallywith respect to youngerchildren. Thestudies available are

De Ruiter (I989) and E-rramdani (I998) for the _{Netherlands, Boukous (I982) and}

El Kirat (2000) forMorocco, Bouhjar (I993) and Hassani (200I) for Belgium, and Mehlem (I998) for Germany. A number

of

other studies were carried out in the

Netherlands, focussed _{on reported language proficiency, language} use, _language attitudes, language preference and so on

of

Berberophones among others. _{After 40}

yearsof migration, and at the time thattheMoroccancommunityiscountingitsthird

generation, many questionsemerge as to the process

of

Berber acquisition andstatus quo

of

proficiency

of

children in the Netherlands incomparison

with

theirpeers in Morocco.

Rationale of

the

_study

With respectto languageacquisition, thefocus of thepresentstudy, it isworthwhile

to take aclose look at the Status of thelanguageamong childrengrowing up in the

Netherlands. Thereisclear evidencethatchildren

of

minority

groups in thiscountry donotreachnative-like mastery inthe language

of

theirparentsortheirprimary home language. The proficiency

of

thesechildrenseemstodeviate fromestablishednorms of native speakers in the country oforigin. Deviations from such norms

implicitly

refer

to inaccurateor incorrect grammaticaloutput(Boeschoten, I990). Such deviations can be temporary, related to aslow-down in the order

of

acquisition, or enduring and

permanent as a result of incompleteacquisition.

The first step towards thestudy oftheacquisition

of

Berber, or more specifically

Tarifit

amongchildren inthe Netherlands, was set up in myMaster's thesis (I998) at

Tilburg

University. That was the trigger to go

further in

this direction, and to

undertake

botha

broader anddeeperinvestigation. Threefactors have beenhelpful in

pursuingmyresearch onthistOpiC.First, the fact that Iammyselfanative speaker of

Tarifit isanadvantage fortherealisation of theproject,especially because the topic has

notbeen_{exploredyet.Second, I was born and grew up in}anArabic_{dominant city in}

2 AcquiringTarifit-Berber

Netherlands, acquiring both

Tarifit

and Dutch. This fact made me aware of the

experiencethesechildrengothrough,both cognitivelyandlinguistically.

A

thirdfactor is my interest in languageacquisition in general. An illustrating description of the

child's language acquisition phenomenon is thefollowing:

„

Children master the intricacies

of

their nativelanguagebefore they are able to tie a knot, jump a rope, or drawa_{decent-looking circle. Thisachievement is so routine and so expected}

thatmostpeoplerarely give itasecondthought. But,its significance has not been lost on linguists, who are only too aware ofthe complexity oflanguage and the mysteries that surrounditsacquisition."(0'Grady, _1997:I).

Indeed,languageacquisition istaken forgranted just like the

birth of

thechilditself. Few _{people wonder how language acquisition} is _{possible. How do human beings}

withinfewyearsafterbirthbecome able tospeakbythemselves

without

anyexplicit

guidance,transcendingboththeirlimitedexperienceand biologicallimitations How

is it possible thatachildis capable

of

learning any language, or even more than one

language easily Neither linguists norbiologists have afforded definitive answers to thesequestions yet. It isindeedthemiraculousnature

of

languageacquisitionwhich

makesthe_{study ofthis phenomenon challenging.}

Overview of

the

_study

The present study consists of Io chapters. Chapter I is devoted to background

information.

It

deals with the status

of

Berber in Morocco, gives a briefhistorical

background, anddiscussesthe circumstances in whichother languages have entered

the country. It also provides general information on thelanguage situation at the

moment, in terms ofthe status and use of each ofthe languages that make up the

sociolinguistic profile

of

Morocco, and provides a summary of the few empirical

studiesconducted on Berber_{languageacquisition in Morocco.}The_{second part of the}

chapter is devoted to

Berber in

the Netherlands, and includes demographic

information on the Moroccancommunity in the Netherlands and the status of its languages. Afterareview ofsomeempirical studies on Berber intheNetherlands and theadjacent countriesBelgiumandGermany, the chaptercloses with anoutline of the

study at hand, and a formulation ofthe researchquestions tobeaddressed.

Chapter 2 deals with the issue

of

conceptualisation and operationalisation. It

describes the successive steps involved in the procedure

of

constructing language proficiencytasks,involvinglanguagevariation in the homecountry,theissueofwhich

varietyto focus on in thisparticularstudy, and taking decisions about thetasks and their contents.

Chapter3 reports on the

pilot

study conducted in Morocco andtheNetherlands

for the validation of tile tests.

It

focusses on the objectives and design ofthe pilot

Chapter4 presents socio-biographicalinformation about thechildrentaking part

in this Study, i.e. grade I and 8 children in the Netherlands and Morocco, with a

detaileddescription ofthebackground of the children in the Netherlandswithrespect

totheirparents' migration history, theireducation, birthcountry, reportedlanguage proficiency in

Tarifit

andDutch, _{language choice, language preference, and language}

attitudes.

Chapters5through 9 make up the main body of this study. They present the results

in two domains

of

morphology and syntax. The morphology tasks concern plural

formation (Chapter 5), case marking

of

nouns (Chapter 6), gender and number

distinction(Chapter 7), and perfectiveformation(Chapter8). Chapter9presents the results ofword orderconstruction taSks.Thechaptersfollowasimilarstructure. Each

of

them

opens with a description of the

task under investigation and the dara

collection procedure, followed by thedata analysis. Both quantitative and qualitative analyses will be undertaken. Each chaptercloseswithconclusions anddiscussion. The results

of grade I children will

be presented first, and those

of

grade 8 children

afterwards.

Chapter io is devoted to a general discussion and conclusions. It sums up the findings ofthis study

with

respect to each ofthedomains investigated, andprovides answers tothe research questionswithrespect tothe acquisition

of

morphology and word orderin

Tarifit.

There will also beadiscussion of thepossible

link

between the acquisition

of

Tarifit in the Netherlands and the background factors of the children andtheirparents.Finally, this chaptercloses up byexploringsome avenuesfor future

research, paving the way forfurtherstudies in order to get adeeper insight on the

CHAPTER I

The status

of Tarifit

in

Morocco and in the

Netherlands

Thischapter providesasocio-historical background forthe presentstudy. It is made

up offour

mainparts. Part I.1isdedicated to the status

of

Berberin Morocco.

It

gives first ahistorical account of the circumstancesin whichseverallanguage varieties have entered the country, followed by a description of the language

situation at the

moment, and areview ofsome studieson BerberinMorocco. Part I.2 is devoted to Berber in the Netherlands, includingdemographic Statistics, the status

of

Berber in daily life, andstudiesonBerber in the Netherlands andin neighbouring countries, i.e. Belgium and Germany. Part I.3 presents a general framework

of

studies on the acquisition

of

morphologyand syntax,and majortheoretical issues in this field. The last part I.4presents anoverview of the actual study, andits research questions.

1.1 Status of Berber in Morocco

I.I.I _{Historical background}

Berbers are the _{first indigenous inhabitants}

of

Northern Africa in _{general, and} Morocco _{in particular. They lived in an} area_stretchingfrom Egypt to the Atlantic

ocean, and from the Mediterranean coast to the Niger river. ThetermBerbercomes

fromthe derogatory Greek word for non-Greeks andwastaken into Latin, yielding the English term Barbarian. Berbers identify themselves commonlyas _Imazigben,which nneans Free men.

From 600

BC, Berber lands were _{invaded by various groups, including} Carthaginians, Romans, and Vandals. There has never been a _{homogeneous and}

organised Berber society or state. Berbers have always been organised under tribal

systems (Obdeijn etal., 2000'I5).Berbertribescould neverunitelongenough to rid

themselves

of

theirconquerors. Asaresult, Berber history can onlybe followed as a

history

of

individual tribes (Montagne

&

Seddon, I973)·

The turning point in the histOry

of

Northern Africawasmarked by thearrival of Muslims inthe_{second half of the 76}century.Therewereapparently fewArabs from theArab peninsula amongthesoldierswhoconquered Morocco. Most of them were Berbers ofthe Eastern part ofthe Maghreb who had become themselves Muslims (Obdeijn et al., 2000:IS).

The strongest Arab migration from the East took place in the I2'h and the I3'h

the East. In the IS'h centuryanotherinflux

of

refugeesarrivedfromSpainafter having been driven away,andsettledin cities likeFes,Rabat, Said, andTetouan. From then on, theArabiclanguageandculture graduallybecamepredominant intheplains and themoreaccessibleparts

of

NorthernAfrica, whilethe Berberlanguageand culture survivedprimarilyinrelativelyinaccessible areas in theAures inAlgeria, Rif,andAtlas mountainsinMorocco. Islam spread all over the countrywithoutexception.

th

The20 centurymarkedanotherturning point inthehiStory oftheregion, due to the conquests

of

Franceand Spain, starting from the end of the I9'h century. In that period, a new lifestylewas introduced

in

Morocco at the political, social,economic and_linguisticlevels. _{French and Spanish}becametheofficial _{languages, each in their}

own area. The effects ofthat periodare still apparent andinfluence thedaily life in

Morocco_today.

In the course ofthehistory

of

Morocco, the Berber dynasties oftheAlmoravids (1063-II47)andAlmohads (II47-1269) were the only tribes

of

Berberorigin to rule the country, besides_{Andalusia (Spain), Algeria, Tunisia,}and Senegal inthe south. Berber speakers, who today _{number more than IS million, are distributed through} Libya, Tunisia, Algeria,Morocco, andMauritania (Boukous, 1995).The number

of

Berber speakers hasalways beenarough estimate, and noofficiallanguage statistics have ever beencollected.Theirdensityincreases generally from east to West,butBerberlanguage varietiesarestill retreatinginfavour ofArabic as the populations of thepresent

nation-statesbecomegraduallyhomogenized(Montagne

&

Seddon, 1973)·

I.I.2 _Languagesituation inMorocco

The languageprofile

of

Morocco is basically made up of fourlanguages, i.e. Berber, Arabic, French, and _Spanish. Berber is _{spoken in three main} zones, with three subsequent language varieties.

Within

Arabic,adistinction ismade betweenMoroccan Arabic and Standard or literary Arabic. French and Spanish are former colonial languages. French still has the status ofthe firSI foreign language in the country, enjoying ahigh prestige in the intellectual andpoliticalarena.

Berber isoneofthetwomajormother tongues in Morocco.

It

comprisesthree main regional varieties, known as

Tarifit in

the Northeast

of

Morocco, Tamazight in the MiddleAtlas mountains, andTashelhit in theHighAtlas,Anti-Atlasmountains, and the Sous-valley. Divergence among these varieties occurs

primarily at the level of

phonology and lexicon _{(Ennaji, I985:8-9), which} happens to make mutual

intelligibility

hardlypossible. _{Moreover, each}

of

thesethree Berbervarieties is made

up of a

wide spectrum

of

local sub-varieties which are _{relatively mutually} comprehensible. Theissue

of

languagevariation and how thisaffectsthe presentstudy

will

be taken up in Chapter 3, where the choice ofa particular languagevariety is

motivated. All in

all, Berber varieties function primarily as oral

mediums of

communication, although they have been gaining territory at the level of written

The statusofTarifit inMorocco and intheNetherlands 7

At

theofficiallevel,rwoeventsareworth mentioning.Thefirsteventconcerns the declaration by the lateKingHassan the Second on the ZI"of August I994, when he announced thatBerber willbe_givena_{place at school. The introduction}

of

Berber at schoolstarted in the schoolyear 2002/2003.Theotherimportantevent concerns the

decision ofthe present kingMohammed 6'h in October 2001 onthe creation of the Royal Institute fortheAmazigh (Berber) Culture.Amongtheduties of this institute is the implementation of mother tongue literacy in cooperation with theconcerned ministries.Theimportance ofthesedecisions lies also intheimplicationsbehind them, inthesense that theyareconceived asanofficialrecognition

of

Berber atthenational level as one oftheexisting languages ofthe country, and meant togetBerber out of

the shadow

of

marginalisation.

MoroccanArabic is the mother tongue

of

non-Berberspeakers, as well as alarge number

of

Berber-Arabic bilinguals. ThisArabic dialect comprises a set

of

regional sub-dialectswhich varyslightlybetween each_other,

without

affectingtheirmutually

intelligibility.

During the laSI decade, adrastic development occurred in Moroccan Arabic, mainly due tomassmediaand school.Thisdevelopment led to the emergence of what some linguistscall Middle Moroccan Arabic(YOUSSi, 1992). This variety is a

mixtureofMoroccan ArabicandStandardArabic. Thephonologybelongsprimarily

toMoroccanArabicwhile the lexicon is thatofStandardArabic.Morphologyisshared somehow between the two.

StandardArabic and French aretaughtatschool, starting from the first and third

grades

of

primary school, respectively. These two languages are the dominant languagesparexcellenceinformal daily life,andfunctions aresharedbetween them. Spanish was the official language in the

North

and South

of

Morocco during the

colonizationperiod. In thetwonorthernenclaves

of

Mellitaand Ceuta,SpanishiSStill

the official language. Due to this historical background and theproximity

ofthe

north of Morocco to Spain, Spanish remains a vital communicative

medium in the Rif

region (Hassani, 1995:Io).

I.I.3 ResearchonBerberacquisition inMorocco

As to empirical studiesconducted on children's use

of

Berberin Morocco, there are two studies to mention in this respect. Boukous ₍₁₉₈₂₎ investigated the language proficiencyof8 bilingualBerberophone children (Tashelhit-Berber/Moroccan-Arabic) of parents

of

Berberorigin, living inthe cities

ofAgadir

(4), Inezgane (2), and

Tiznit

(2). Theirlanguageproficiencywascompared tomonolingualBerberophone children

(5), living in theruralareas

of

TaddartandBiougra, also in the Southern parts. The

children, aged _{between 5 and I3 years,} were tested in their lexical repertoire with

respect to reference to colours, body parts, animals, and kinship, and in their

Table 1.1: Correct scores (%) of bilingual and _{monolingual Berberophone children in Morocco} (adapted from Boukous, 1982)

|Domains ' Reference tasks. - - - Bilingual childrin Monolingual_{children |}

Lexicon colours 4 95

bodyparts 66 100 animals 41 70

kinship 80 100

Grammar _{personalpronouns 55 80}

perfective 44 80

imperfective 44 80

mperative 62 80

The results show thatmonolingualTashelhit speaking children performed better than bilingual ones on alltasks. Lower proficiency ofthebilingualchildren can be seen at all levels. _Shortcomings in _{lexical knowledge} were _{compensated by recourse to} strategiesasgeneralisation, paraphrasing,andborrowing from Arabic. Atthelevel of personal pronouns, the children happened to neutralize the difference between feminine and masculine in the plural form, in favour

of

masculineforms. For verb

conjugation, there weretemporal confusions on the part of the bilinguals, who used the imperfective instead of the perfective orviceversa.There was also a neutralization of gender and number, by using masculine forms instead

of

feminine ones, and singular formsinstead

of

plural ones.

Atthesyntacticlevel,thebilingualspeakers hadproblems with the construct form, which was neutralised by usingthe free form asin_{lqism ntarbat (the girl's classroom)} where tarbat (the girl) is used instead of the correct constructform terbat.

Boukous_(I982) concluded that thebilingual informants inthecities did not have

aperfectcommand

of

Berber. This wasalsoclearfrom theircommunicativestrategies, marked by the use

of

otherlanguages_{(Moroccan Arabic}andFrench) whenspeaking Berber, a phenomenon that hecalled_{transitional pluralism.}

The second _{empirical srudy on the acquisition}

of

Berber has been _{conducted by} El Kirat (2001). This study investigated the status of BeniIznassen Berber,orIznassni Berber, avarietyspokenaroundBerkane in theNortheast

of

Morocco.

Two

groups were involved:onegroupliving intheurban area, and the other in the ruralarea,split over two main generations, i.e. an old generation (I6 informants) and a young generation(25 informants). Theage

ofthe

informants varied between 40 and99 years for theold generation, and between 7 and39years for theyoung one. Data collection

was done_bymeans

of

_participantobservations, freeconversations, and interviews. ElKiratfocussed on the effect

of

_languagecontact inMoroccoonIznassni Berber.

The investigation concerned 3 areas. The first area was concerned

with

language

The statusofTarifit inMorocco and in theNetherlands 9 groups shared these attitudes, irrespective

of

their age or rural/urban background. Moreover, the youngestgroups in bothareasdeniedtheirBerberorigin,andviewed

it

asastigma. The second area

of

investigation dealt with language use

within the

community.The studyrevealed thatthedomainsofuse ofBerber hadbeencompletely submerged by Moroccan Arabic in the urban area in alldomains. Berber was more

dominantamong the informantsabove₄₀years andless dominantamong the ones

under the age of 25, in that only Io% of the latter used

the language. This

phenomenon

of

languageshift wasalsowitnessed intheruralarea,where Berber is the

dominant_{language. Berber was usedamong adults, but became}less_{spoken with and}

amongspeakersunder the age of26.There wasatendencytowards the taking over of

Moroccan Arabic, being used more and more at home too. Some parents had even decided to stoptalking Berber to their children andused MoroccanArabicinstead.

The third

aspect studied was language proficiency in Iznassni Berber. There was a considerabledifference inproficiencyamongtheolder and the younger groups.Fluent

speakers in the urban area were aged 40 years or older. For the younger ones, the Berber proficiencylevelvaried between those whowere relatively abletoconduct a conversationthough not inaperfect way (aged 26-39), and theones_{whose proficiency} was limited to afewwords andphrases (aged 7-25). The rural groupswerebilingual in all agegroups. Theywere fluent in both Berber and MoroccanArabic, with the exception o f the youngest ones (7-25), some of whom had only receptive skills in Berber.

1.2 Status of Berber inthe Netherlands

I.2.I _Demography

In x963,theNetherlandsreached anagreementwithMoroccofor recruitingpeople to come to work in the country, asaresult oftheacuteneeds forworkers ontheDutch

labour marker. This was the starting point of the Moroccanmigration. In 200I, there lived more than 260,000 people

of

Moroccanorigin in theNetherlands (Dominigez

& Veenman, 2001).

Moroccanmigration totheNetherlandshasknownthreeinfluxes.The first influx

took place between I963-1973. Men left Morocco and came to the Netherlands to work

for a certain period of time. They were referred to as_{guest workers, and}had_strong

cultural and social contacts with their home country. In 1973, the

number of

Moroccans in The Netherlands rose to 14,000 (Muus, I993). Thesecondinfluxstarted in I973, when increasing numbersofworkersbegantobring theirfamilies over to the Netherlands. The stay in theNetherlands seemed to take longer than they had thought

at first. This circle migration changed w permanent migration. Family reunion reached its top in I981 with lo,400 family members brought over from the home

was mainlygenerated by marriages. The children

of

guest workers have meanwhile become adults, and have decided to

build

Up their life further in the Netherlands.

Most of them_getmarried

with

_{partners from Morocco. This type}

of

_{migration made}

up to 10% of the migration balance in the seventies, and 40% in I992 (De Beer et al., 1996).

During thelast decennia, theCOntributiOn ofthemigrationfactor totheMoroccan

population intheNetherlandshasbecome lower, with 7,I28persons in I989 and 2,041 in I996. Instead, birth intheNetherlandshasbecomeacrucialfactor indetermining

the absolutenumber oftheMoroccancommunity. TheaverageMoroccan family has more than fourmembers (Van de

Heijdt

&

Harsen, I996). From January 1990 to January 1992,thenumber ofthefirstgeneration rosefromII5,000 tOI3I,000 persons,

withan_{increase of I3%,}whilethat

ofthe

secondgeneration, born intheNetherlands, rose_{from fI,000}to_{63,000 persons, with}anincrease of 25% (Van de Heijdt

&

Harsen, I996). In 2000,thenumber

ofthe

secondgeneration reachedalmost110,000 persons,

withan increase_{of42% (Dominigez}

&

_{Veenman, 2001). Table}I.2givesalongitudinal

and comparative

overview of

the top-IO immigrant minority

groups in the

Netherlands, based on the combined birth

country of

the person, father, and/or mother, according totheCentral Bureau

of

Statistics (2003).

Table1.2:_{Top-10 ofimmigrant minoritygroups in the Netherlands between 1995-2002, according}

tobirth countryof person, father,and/ormother (x 1000) (CBS, 2003)

[CountrVelblith .' -" 1995 - --- 2000 - 2 01 - - -2002 i

Turkey 264 309 320 331 Surinam 276 303 309 315 Morocco 219 262 273 284 AntilleanIslands 86 107 117 125 Former_{Yugoslavia 49 67 71 75} Italy 31 34 35 35 Spain 28 30 30 31 China 22 30 32 36 Somalia 17 29 30 29 Hong Kong 17 18 18 18

TableI.2showsastrong increase overtime among most ofthegroups,including the

Moroccan _community.

I.2.2 Berber in theNetherlands

The Central Bureau

of

Statistics commissioned a

study in

_{1984 among I,098}

Moroccans in theNetherlands to findout about theirbirth place inMorocco. 60% originated fromthe North

of

Morocco, IS% from the South, Io% fromthe Centre,

8% from the Northeast, and 7% from thearea along theAtlanticcoast. The people

The statusof TarifitinMorocco and inthe Netherlands 11

Berber in daily life

Within the Moroccan community at large, Moroccan Arabic is used asalingua#anca.

This is

at least relevant for the first generation. Among people of the second generation, Dutch becomesthelinguaji·ancabetween Arabic and Berber speakers, and even between speakers of the same language. Dutch is increasingly used at home betweensiblings and, to a lesserextent, between parents and children. All in all, the

use

of

Berberislimited toasmall set

of

informaldomains.

Berber in religion

Inreligiousmatters,StandardArabicisprevalent fortheMoroccancommunity in the

Netherlands. _{Prayers take place}in Standard Arabic.The_{Quran is read in its original}

version in

_{Arabic too, as is the tradition all over}

the world. Friday sermons are sometimes exceptions in this matter.Normallyspeaking, thespeeches areinStandard

Arabic, but they may also be _{given in Moroccan Arabic, Berber or} even Dutch.

Because the public attending the prayers are not all Berber speakers, recourse to StandardArabicand/or MoroccanArabic is_preferred,

with

_{explanations sometimes}

in

Tarifit,

depending on the imam's proficiency in thelanguage.During thelast years, there is alsoatendency to Use Dutch, due to theemergence

of

attendants from the secondgeneration. As aresult, the status

of

Berber in this domain is very weak.

Berber in tbe mass media

The NPS radioforMoroccansbroadcastsfrom Monday throughFriday,between I9:45 and 20:30. The first 10 minutes are reserved for news in

Tarifit. The rest of the

programme ispresented in Arabic (Standard andMoroccan Arabic). Still, Tarifit is used from time to time in thecourse of the programme, mainly duringinteractions

withthe audience, which uses

Tarifit

as language

of

communication. Thischoice is

made possible by the fact thatthe programmehosts arealsoproficient in Tarifit. This

is the onlydomain of use

of

Berber/Tarifit in the Netherlands in thepublicmedia. The

NMO

(Dutch Muslim Broadcastingservice) radio and television broadcasts in Arabic, Turkish, and Dutch. In addition, Moroccans in the Netherlands with a

satellite dish can follow the news in Tarifit as well as in other Berber varieties, broadcasted_{by Moroccan radio}andtelevision stations in Rabat. TheuseofBerber in

Moroccan television programmes is verylimited, whileradio programmes have four hours per day for each of the threemajor Berbervarieties.

Berber in tbe literature

The last decenniawitnessed the emergence ofa Berberliterature, i.e. novels, plays, stories, andpoetry in

Tarifit,

whichare_{becoming more and morepopular. However,} there are some problems in the way

of

spreading this literature, in particular with

Netherlands andWestern Europe ingeneral, themajority

ofwriting

in

Tarifit

or other Berber varieties takes place in a Latin alphabet adapted to accommodate a few

additionalsymbols needed towrite Berber.

Besideswritten literature,theatreproductions in

Tarifit

havebecomequite popular during the _{last years.} This type of art enjoys morepublic thanthewritten literature,

among other things, due IO itsoral nature.Theactorsoriginate mostlyfromMorocco, and tourin Europeespeciallywherelarge Berbercommunitiesarefound.

In addition, a number of education programs have been developed, like DeSchatkist (Ldle Uitgeverij, I996) (thetreasure_{box), translated}from Dutch into

Tarifit. In the

framework of the Comenius Action II-programme for tbe European Union, acurriculum

hasbeendeveloped inTarifit (nexttoother languages),consisting

of

threebooklets for childrenbetween 4 and8years and two onesforpupils aged I2 toISyears(E-rramdani

et_{al., I998c, 2000; Benhakia, 200I).}

Berber at scbool

In the Netherlands, childrenstartschool at the age of4.Dutchelementary schooling is made up of8grades,divided into three parts.The lowerpart includes the first two years, the medium part consists

of

grades 3,4 and 5, and the higher part

of

grades 6, 7 and 8. Instruction in the homelanguagefor immigrant

minority

children inDutch

elementary schools was introduced in 1974,

referred to at that time as OETC

(Education in Own Language and Culture), later on as

OET

(Education in Own

Language since 1987), and most recently as

OALT

(Educationin Allochthonous Living

Languages since I998).

Inall contexts,the languagetaughttoMoroccanchildrenhasalways been Standard Arabic, theofficiallanguage

of

Morocco (cf. Saidi, 200I:28-33)· The last decennium has witnessedmore requests forteaching

Tarifit to

childrenwhoseparentsoriginate from the Ri f. Yet,manyparents do not seemtosupport suchdemands. Most of them arein favour ofthe teaching

of

Standard Arabic. They claimthat theirchildren can learn Berber at home. The need for the learningand teaching ofStandardArabic is

linked to Islam. Thoseparentswant their_{children IO be able to read the Quran, and}

have access to theofficial language

of

Morocco.

However,some changeseems to take place in this respect. In thelanguage survey

The statusofTarifit inMorocco and in the Netherlands 13

I.2.3 ResearchonBerberacquisition in theNetherlands

Anearlier study ofE-rramdani (1998, 1999)wasdedicated to the development of both

Tarifit

and Dutch in theNetherlands, among Moroccan children ofthe first three grades in primary _{school, aged 4-6} years. The study investigated the domains of

phonology, lexicon (receptive andproductive), syntax and textcomprehension. The best scores for all groupswere realised in thephonology taskfor Tarifit, and in the

receptivevocabulary task for Dutch. The lowestscores were achieved in syntax in

Tarifit,

and_{productive vocabulary in}Dutch.

With respect to theperformances on the tasks of the two languages, there was a tendencyin favour ofthetasksinDutch forthefirstgroup (4-year old).Yet, there was nosignificantdifference, with the exception

ofthe

textcomprehensiontask,where the grade I childrenscoredsignificantly better in Dutch. The children

of

grade 2 (5-year old)scoredbetterinDutch on thetwotasks

of

receptive vocabularyand syntax,while

the grade3children(6-yearold) realisedbetterscores in Dutch on all tasks, with the

exception of the task oftext comprehension.

With respecttolanguage_dominance,the grade Ichildrenturned out tobebalanced bilinguals, while theones of grade 2 and 3 weredominantin Dutch. The development

of the

two languages showed two different patterns.

Tarifit

seemed to develop

significantly betweenthe first and the second grade

of

primary school, and to slow down bythe

third

grade,whilefor Dutchthe oppositepatternoccurred.

Reportedlanguagechoicepatterns of the children_{showed that 85% of them}spoke

Tarifit

with theirparents,while thetwolanguages were usedequallywithsiblings. The groups differed finally in their language use

with

friends in that 66%

of

grade I childrenspoke

Tarifit,

while the ones of thetWOhighergrades usedDutchexclusively.

De Ruiter (I989) dedicated part ofhisstudy to theproficiencyin

Tarifit, in

a

quasi-longitudinal design, among 40 Berberophones in the Netherlands. The study was conducted amongfourcohorts

of

informants, aged 7, II, I4, and21_{years. Part of the}

study wasalsodesigned to test communicative and grammaticalproficiency in both Dutch and

Tarifit.

Data_{collection was done by} means

of

_{picture description tasks.}

Concerning the first topic, the subjects turned out to express themselves better in

Dutch than in Tarifit, with

the exception ofthe oldestgroup. Some oftheyounger subjects were not at all able to express themselves in Tarifit. As far as grammatical competenceinDutchisconcerned, thesubjectshappenedtomatch tosomeextent the performance

of

Dutch nativespeakers inthereference groups.

In a follow-upstudy, De_{Ruiter (I990) compared the proficiency in}

Tarifit of the

group intheNetherlandswithBerberophonepeersin Morocco(CityofNador). They

were 6intotal,aged between 8 and26_years.Again,apicturedescription test was used. Data analysiswasconductedwithrespectto _lexicon,morphologyand syntax.

With respect tothefirstdomain, thegroup inthe Netherlands usedlessvocabulary

Netherlandsshowed inaccuracieswith_respectto_gender,number,andconstructstate. In syntax, theanalysis focussed on the use

of

simpleversuscomplexsentences, and the mean length

of

utterances (MLU). The younger groups of 7 and II years in the

Netherlandsobtainedsimilarresultsastheirpeersin Moroccowithrespect to the use

ofsimpleandcomplex sentences and MLU. Thesubjects aged_I4showedcomparable scores in the use

of

simple and complex sentences. In the case of MLU, theonesliving

in the Netherlands performed better. Finally, the oldest group in the Netherlands obtainedlowerresults than the one inMoroccowithrespect to the use of simple and complex sentences and MLU.

In

general, the core group in the Netherlands used simple syntactic structures and simple forms compared to the reference group in

Morocco, which used more rich and complexsentences.

Basedonreportsonlanguageuse,Tarifit came out as themostchosenlanguage in

interaction withtheparents for thefourgroups intheNetherlands.Theaveragechoice varied between 74% with thefather and 92% with the mother forthe group aged II

years, to Ioo% with the _{mother and 94% with} the father in the case ofthe oldest group.

With

siblings, the choice

of Tarifit

regressed drastically to 54% in the case of the IIyearsold group.Thechoiceof Tarifit waseven lowerwithfriends atanaverage of 23% for thethree youngestgroups, and 42% for the oldest group.

In_{the period between I997 and 2000,}alarge-scale language survey wasconducted by Extra etal. (2002) in thirteen municipalities in theNetherlands among children in primaryschools_{(99,460) and secondary}schools(39,45I) in theagerange of 4-I7_years. The reported top-20 ofhomelanguagesother than Dutchisgiven_{in Table I.3.} Table1.3:_{Reportedtop-20 of homelanguages}otherthanDutch (source: Extra eta/.,2002:52) ILari*iahes (1-10) - - - Rep-6,tZ-d freq1i6ncy Langlid#bs (11-20) Reported frequency i

1 Turkish _8,686 11 Chinese 1,062 2 Arabic 6,755 12 Kurdish 1,052 3 Berber 6,302 13 Somali 692 4 English 5,153 1 4 Italian 690 5 Hind(ustan)i 5,037 15 _{Moluccan/Malay} 657 6 _{Papiamentu 1,572} 16 Urdu/Pakistani 644 7 French 1,534 17 _Portuguese 559 8 German 1,449 18 Serbian/Croat./Bosn. 534

9 Sranan_Tongo 1,426 19 javanese 481

10 _{Spanish 1,270} 20 Farsi 400

The majority (76%) of thechildren

ofthe

Berber group were born in the Netherlands,

while most

of

their parents (88-92%) were born in _{Morocco. 96% of}the children

The statusofTarifit in Morocco and intheNetherlands 15

younger ones (28-47%) and _{older ones (25-49%), as well as}

with

Berber speaking friends (I4-45%)·

With respect to language dominance, children aged 4-5 years were dominant in Berber (52%), whilethe older ones _(6-I7years)weredominant inDutch. Regarding languagepreference, 45% of theyoungestchildren(4-5 years) preferred Berber. This percentage decreased withage,reachingthelowestscore with the ones aged Io-II years,

with _25%· After this age, the percentage

of

children preferring Dutch decreased,

reaching 36% with the ones aged _I7 years. There were also informants without a preference for any of the two languages. Their number varied between 5 and 25%, increasing withage (Extra etal.,2002:86-87).

The general picture that emerges from these studies is that there is a pattern of

languageshift takingplace betweendifferentagegroups. This shifttakesplace at the levels

of

_{language proficiency, language choice, language} dominance, and language preference. Shift takes place within thesamegeneration, and as such isexpected to speed up betweengenerationstoo. Languageshift

of

BerberophonesinMorocco also takes place

within

onegeneration, in particular when aBerberophonefamilymoves to anArabicspeaking area.Thechildren bornandgrowingup there acquire Moroccan Arabic inwhich theybecomedominantandwhich they also usewith theirparents. Proficiencyin Berberisreducedto receptive skills.

The overall impression is that Berberophone children in the Netherlands are dominant in Berber by the time

of

startingschool. By the time they reach the last grade

of

primaryschool, children have becomedominant in Dutch. Berber is then used extensively

with

parentsand younger siblings,while Dutchisnearlyalways used

with_{older siblings and friends.}

I.2.4 Researchon Berberacquisition inneighbouringcountries

Outside The Netherlands, empirical studies ontheacquisition

of

Berberingeneral or

ofTarifit

inparticular are verylimitedin Europe, i.e.twostudiesinBelgium and one studyinGermany. Bouhjar (I993) studiedlanguageproficiency in the mothertongues

of

Moroccan youngsters in Brussels. Her sample consisted of9 female Moroccan Arabicspeakers and 6 female

Tarifit

speakers, aged I4 to 20 years. Theywereeither

born in Belgium or had been living there for a

period of

at least Io _years. Data collectionwas basedon spontaneous conversations. Domains ofanalysiswerelexicon,

morphologyand syntax.Theresultsindicated that I0% ofthe youngsters'lexicon was borrowed from French, revealing strong gaps in their language repertoire. This outcomewasreached_{after having distinguished}real_borrowingsfromFrench (French words not _integrated in Arabic), from French loanwords _{integrated in and having} become part

of

Tarifit, and as such used in Morocco too. For morphology, the

feminine. _{Surprisingly, the} same _{happened with} the reference _{group in} Nador

(Morocco) too. Bouhjar referred to this phenomenonas neutralisation of_{the gender}

difference.

In syntax andmorphology, there wasadifferencebetweentheinformants born in

Morocco and the ones born

in

Belgium regarding the free-constructstate. The latter group usedthefree state form of nouns instead of the construct one. This last form is

marked bymodifyingthe

initial

vowel

of

nouns. Besides,the utterances ofthegroup in Belgiumwere marked by the use

of

simplesentenceformsasopposedtocomplex ones. The meanlength

of

utterances (MLU) was3items,whilethereference group in

Morocco had an MLU of5items.Mostdifferencesbetween thegroupinBrussels and

the one in Morocco, however, were witnessed at the level ofthe lexicon (Bouhjar, I993:I68).

Anotherstudy conducted in _{Brussels is the one by Hassani (200I) on} language loss among second generation Moroccans in Brussels. There were I6 informants, aged between I4 and 29 years. Theywere divided into two groups, one group speaking Moroccan _{Arabic (8) and} the other one speaking

_Tarifit

_{(8). Besides, data were} collectedin Morocco with twosimilarreferencegroups.

Patterns

of

_{language loss}were studied inthedomains o

f

_{phonology, morphology,}

lexicon,andsyntax. The findingswith respect to the Berbergroup in Brussels show inaccuracies in alldomains.

In

phonology, thefollowingareencountered: reduction

of

geminate segments, sound substitution, and to a lesser extent, problems with

pharyngealised sounds.

With

respecttomorphology,theinformantshappened to omit or misuse personal affixes,andshoweddifficulties inplural formation. Atthelevel of syntax, irregularities are observed with respect to the deletion or

substitution of

prepositions and conjugations. As far as the lexiconisconcerned,someinformants had

difficulties in retrieving the appropriate lexical items during _{conversations, due to}

lacunasin theirlanguagerepertoire(Hassani,200 I. 4II-4I2).

The socio-cultural orientation (friends' network) ofthe informants explained to some extenttheobserved degree

of

languageloss. Informants with anon-Moroccan

social network hadarestricted use o

f

Tarifit, and thus were more subjecttolanguage

loss. The ones with a _{strong Moroccan} social network had generally a very good retention

ofthe

language._{However, this was}notalways the case,and depended on the

proportion ofuse

of Tarifit

amongtheir network_{friends (Hassani, 200I:474-475)·}

For Germany, the study

of

_{Mehlem (I998) is the one available to us.}

It

deals with language use and languageproficiency

of

Moroccanchildren in theagerange of Io-If

years. Moroccan migration to Germany has been more limited compared to the

The statusofTarifit inMorocco and in the Netherlands 17 status

of

their children in the German school system, in

particular in the city of

Dortmundwhere he carried OUthisstudy.

It

shouldbenoted thatin conformity with Tilmatine (I994:22), Mehlem refers to Berber people and their language as Masirel

Masirin and Masiriscb, respectively. These concepts relate to the common self-definition in terms

of

_{Amazigb (person) and Tamazigbt (the language). Mehlem's} Berber dataare basedon

Tarifit

_{speaking children. The} research_{sample consisted of}

acoregroup of 28 children who hadbeen living for more than_{4 years}in _Germany;

18 of them were

Tarifit

speakers, andIo spokeMoroccan Arabic. Besides there was a

control

group of

Io children who were less than 4 years in _{Germany (Mehlem,}

1998:36).Datacollectionwasbased onboth reportedlanguagebehaviour andlanguage proficiencytasks.Thelatter consisted

ofa

seriesofpictureswhich had tobedescribed orally, first in German and then in

Tarifit or

in Moroccan Arabic. In a laterstage, another picture description taskwas conducted forwritten text production in both

German and

Tarifit

or MoroccanArabic. In the analysis, these two-part data have been taken as awhole. Here, our focus is on the outcomes of the core

group of

I8

Tarifit

speakingchildren.

In interaction with

the

_{father, 4 out of}

IB children chose almost only

_Tarifit,

5 predominantly Tarifit, 8

Tarifit

and German equally, and one child chose

pre-dominantlyGerman.In interaction withthemother, II out ofI8childrenchosealmost only Tarifit, 5predominantly Tarifit, and2

Tarifit

andGerman equally.

With

siblings, 7 onesused

Tarifit

andGermanequally, 5predominantlyGerman, and5almost only

German. No child _spoke

_{Tarifit only or}

at least _{predominantly.} These _{patterns of} languagechoice showa_process

of

_{graduallanguage}shift.

When asked why theparentswould talk German to them, thechildren gave the

following reasons: because theyknowGerman well one time, because they want to facilitatethelearning

of

German4times,because they knowtharchild'sTarifit is not

good enough6times,or otherreasons/doesnotapply₇times.Whenaskedmirrorlike whytheparentswould talk Tarifit to thechildren, variousreasonsweregiven as well, i.e. because they want to maintain

Tarifit

8 times, because rileyspeakTarifit better

than German6times,orbecause

ofboth

reasons 4 times.Reporting ontheirlanguage preference, 2 children were in favour

of

Tarifit, 8 ones in favour

of

German, and

6 ones infavour ofboth_languages.

The outcomes of the oral picture description task led to the construction of a cumulativelanguageproficiency index for bothGerman and

Tarifit. For

this index, a selection of thefollowingparameters was takenintoaccount:

- Text length, in terms ofthetotal number

of

finiteverbforms;

-Lexical variation, in terms

of

spatio-temporal adverbs, content_words,

full

_verbs, word types, andprepositions;

- Lexical strategies, in terms

of

code-switching, paraphrasing, generalising, and

- Mean length of utterance;

- Sentencecomplexity, interms

of

subordinate constructions.

Here, we can not go intomany detailS oftheanalyses that have beencarried out. For such details, we refertoMehlem _{(I998:171-195, 229-235)·}

1.3 _{Acquisition ofmorphology andsyntax}

Before giving a presentation ofthe research questions for the present study, an

overview of the majorissuesraisedinlanguageacquisitionstudies isin place. Some of theseissues willbeaddressed in thediscussion ofthefindings inthisStudy.

Morphology is the area oflinguistics concerned with tile internal structure of

stratified words. Early grammatical development involves the acquisition of the major devices used to express or understand grammatical relations in a given languages,

including _{grammatical morphemes. Questions raised in this} context are: how do

childrenrecognize thestructuralproperties of thelanguage they have been exposed to, and how dothey produce grammatical morphemes,i.e.acquirewhichforms to use to

fulfil

whichfunctions2

Following Chomsky's universal grammar theory (Chomsky, I98I, I988), a good

number

of

researcherspostulatethatchildren's acquisition

of

grammarisgoverned by

a set

of

innate principles andparameters, and maintain thatgrammar acquisition is basically determined by Strictly linguistic factors. By contrast, functionalist and

constructive approaches _{(Elman et al., I996)} argue in favour ofan interactionist account where language acquisition is the result

of

interaction between biological, cognitive, andenvironmental constraints. Intheir view, theemergence

of

grammar is relatedtocommunicative andlanguage processingconstraints, and

it

interacts with the development

of

otherlanguage capacities, suchasphonological andlexicalskills.

Within

functionalist-constructive frameworks, attention hasbeen devoted to two main types

of

variation. First, thereare cross-linguisticstudies addressingquestions such as why is theTurkish inflectionalsystem somucheasierforchildrentolearn than

the English one, or why is French so much harder for children to segment than Chinese.i(Slobin,19976:I36). Secondly, thereisvariationacrossindividuallearners of

alanguagebecausechildren, even when acquiring thesamelanguage, vary in the rate as well asthe style

of

acquisition of awiderange

of

language capacities.

Grammatical morphemesemerge at different times and aremastered at different

rates _{(Brown, I973)· The frequency}ofoccurrence

of

inflections in theinputmakes a difference in the time

of

occurrence inthe_{speech of}achild. Whenaninflectionoccurs frequently, its grammatical function

will

generally behypothesized sooner, and its representation in the paradigm will be strengthened (Leonard

&

Swanson, I999).

The statusof TarifitinMorocco and in theNetherlands 19

(Pinker, 1984; Slobin, I985). Notions high in this hierarchyare thosethat appear in many words andhaveclearsemanticcorrelates (Leonard

&

Swanson, I999).

Bybee (I995) suggestsadifferenceinacquisitionbetweenregular and irregularverbs. Regular forms such as wasb-washedare_{stored as one lexical item, whereas irregular} forms likebuy-bought arestored as two separate lexical items, which Bybee refers to as dual processing model Itisfrequency in the daily use that is important, and not the rule itself. The frequency ofoccurrence ofwords like good-bestin English is high, and thus thesewordsareacquired atanearlierstagethan other regularlow-frequentwords.

The blocking-plus-retrieval-failure model ofMarcus (I995) predicts that noun plurals have an

initial

periodofcorrectusebeforethechildhasinducedthedefault plural rule,

followedlater byaperiod

of

overregularizationafterthechildhas induced theplural

rule. Children's early stage

of

correct performance comes from using memorized forms. Overregularizationsonlybegin when thechildhas_{mastery of}a_{regular default}

rule. In this sense, overregularizations are repairs

of

missing knowledge, created by overlapping a regular rule to inappropriate regular forms and irregular rules. The

irregular forms areretrieved from memory and blockthe application ofthe default

rule. Ifachildfails to retrieve thecorrect irregular form, the regular rule applies by

default andthechildproducesthe_{overregularization (Marcus et al., I995)·}

Connectionist models

of

inflection (cf. Rumelhart & McClelland, I986) offer an alternative account

of

overregularization. Overregularizations occur when irregular formsareattractedtoregular forms. Forexample,the overregularizationtbinkedmight

occur by analogy to blinked. Thus, other things being equal, the more regular forms there areascomparedtoirregular forms, the moretheseforms willbeovergeneralized.

The study oftheacquisition

of

morphologygenerallymakesadistinctionbetween

the acquisition

of

nouns and the acquisition

of

verbs. More nouns than verbs are acquired in the earlieststages

of

children'slanguage development as Gentner (1982) found.OtherStudiesshowed justthe opposite, i.e. more verb-dominance in thechild's

earlylanguage use_{(Le6n, /999b).}

By most developmental accounts, children learningalanguageacquire names for concreteentitieseasily and do soinpretty much the sameway. Gentner (I982)found

that children haveadominance o

f

nominalelementsin theirearly utterances,leading her to propose a primordial cognitive readiness in young language learners to lexicalized, perceptually bound, andconcrete entities. Thefunctionalist/constructive

approach considers thatthecategory verbisacquiredorconstructedprogressively and, therefore, that its mental representation changes during the process

of

acquiring a

language andapproaching adult competence.

Tomasselo (I992) proposes whathe calls tbe VerbIslandHypothesis,according to

which, in early stages, children use verbs as individual islands

of

verb-by-verb organization. The mastery

of

argument structure and a generalised morphological

as well as inthe_{infinitive, in}ordertoconveytheirmessage. Later on, they extend verb morphological markers to other tenses, and acquire a greater repertoire

of

personal referenceother thanthirdpersonsingular.

With respecttosyntax, it is generally suggested that people rely on multiple cues for decoding sentences. The competition model

of

Bates and MacWhinney (1989) providesanexhaustivedescription inthis field.Theassumption is that different cues enterin competitionwhenprocessingsentencesandmapping form onto function, i.e.

when determiningthemeaning ofasentence. Cues vary and includeintra-sentential ones asword order, agreement, animacy, context, case marking, or extra-sentential onesasstress andtopic.Theweight of the cuesvaries fromone language toanother. An example of this istheinterplay of wordorder and subject-verbagreement as cues to sentencemeaninginterpretation inItalianand English. Italian has morewordorder options than Englishwhich hasarelatively rigid wordorder (SVO).Thisresults in a difference in thereliability

ofword

order as a cueinagentindicationbetweenEnglish speakersfavouringwordorder,andItalian_{speakersfavouringsubject-verbagreement}

(Bates

&

MacWhinney, I989:II). Relianceon secondarycues occurs when an

intra-sentential cue isnoteffectiveinsentence_{interpretation. Stress,}forinstance,wasfound

to have an impact on Italian processing when the most important cue (i.e. verb

agreement) was unavailable. Forspeakers

_of

_German, reliance on stress takes place when verb agreement is ambiguous, or when agreement and word order are in

competition (Su, ZOOIa).

1.4 The present study

The subject ofthe presentstudy is theacquisition

of

Tarifit

morphologyandsyntax

by children born and/orgrowing up in theNetherlands, i.e. in amigration context

whereDutch isthedominantlanguage, compared to the ones in Morocco, living in

a_{Tarifit monolingual context.}

Morphologytesting will be conducted

with

respecttonounsand verbs. Nouns in Tarifit maybe affixed for number _{and gender,} as _{subjects, and objects. They agree}

withthemarkers on the verb. Forverbs,therearethreeverbclassesin

Tarifit:

finite, non-finiteandauxiliary (Cadi, I987; McClelland,_I987).Non-finiteverbs, including

infinitives,copulas andauxiliaries,haverestricted_{inflections (McClelland, 2000:25).} The domains

of

investigation under thecategory

of

nouns willbepluralformation

and case marking, the latter also referred to as free and construct state. Plural

formation in Tarifit

is _{manifested mostly by simultaneous processes}

of

_prefixation

The statusof TarifitinMorocco and intheNetherlands 21

-yen, -ten) and fominine nouns ( -in, -win, -awin, -iwin, .jin, -tin). Internal Stem modification takes place at the level of vowels as in aserd-u-n/iserd-a-n (mule).

The domain ofcasemarking

will

concerntheshift of a noun form from the free to theconstructstate, i.e.after prepositions and when the subject of a verb ispostverbal. This changeinstate takes form by means of thealternation oftheinitial of the noun.

It is characterised either by vowel modification, like a becomes u, as in _aryaz/uryaz

(man), or by the insertion

ofa

glide w

ory

before a-, u- and i-, or even bythedeletion of the vowel following the initial gender marker t- for feminine nouns, as in_{taI enjin 1}

tbenjirt(girl). Yet, there are a number of nouns that are invariableas _{baceklit (bike).}

The shift to

the construct state concerns nominal stems, masculine or feminine, singularorplural. The structure of the noun in thefreestate_hardlygivesanindication

about the form the nominalstemtakes intheconstruct state.Thereare,however, two regular principles: (a) the vowel

ofthe

feminine plural

i-

whichcomesafterthe gender i ndex t-, as in t-i-myar-in(women), or of the singular form a which comes after the gender index t-, as in t-a-mYar-t (woman) drops in the construct state, and (b) the

vowel uinsingular or plural formsisneverdeleted intheconstructState(Bouylmani,

I999). Saib (I982:IBI) observes that feminine nounsas _{taflinast(cow), being derived}

from masculine ones as _afunas (bull), are characterised by the deletion of the vowel

whichfollows the initial femininemarkert-, whereas thosewhich have thefeminine property (without a masculine counterpart)as tara (fountain) maintain the vowel, and

remain unchanged. However, this is not always predictable, and the learner has to acquire theetymology of the nouninquestion.

The domains studied

within

verb morphologyaregender-number agreement, and perfective.

With

respect to thefirstdomain,the conceptofgenderinvolves thebinary

opposition

of

masculine and femininefeatures, while that

of

number concerns the

dichotomy

of

singular and plural

forms. In

the present study, gender-number

morphology willbestudied in relationtoverbs,bymeansofpersonal inflections. They areobligatorilyadded to theverb, except when the verb is in the participle form. They occupy either the suffix position as infirstpersonsingularstem -7, thirdpersonplural masculine stem-n, and

third

personpluralfemininestem-nt, or the prefix position as inthirdpersonsingular masculine i-andfeminine t-Stem, or boththeprefix and the

suffix positions as insecond person singular masculine and feminine t-stem-d and second person pluralmasculine t-stem-mandfemininet-stem-n-t. Insecond person plural and

third

person singular and plural, the persons' infiections are complex features, containing gender markers (i-/t-stem, and stem-t) and/or plural markers (stem-n/-m).

In_complexsentencesin

Tarifit,

some_{interrogatives}andrelativeclauses_{require the} participle form o f the verb. The crucialproperty ofthis participle form is that it does not agree with the subject, neitherin number nor ingender. It isan invariable form