#t to superhell to live with the emma bots etc

[dancedance-resolution]

Loosing grammatical gender in Dutch: the result of bilingual acquisition and/or an act of identity?

Leonie Cornips, International Journal of Bilingualism (Vol 12, Issue 1-2) March-June 2008, Sage Publications Ltd. (UK). Document type: Report.

@ekingston apologies for the formatting, I just copy pasted everything lol

Abstract

In the Dutch sociolinguistic literature it is often noted that the overgeneralization of common gender, that is, the use of the definite article de where het is required, characterizes the Dutch variety(ies) of immigrants. Recent sociolinguistic research shows that this overgeneralization also occurs in the speech of subsequent generations born in the Netherlands who acquire Dutch as one of their first languages. The aim of this paper is to discuss whether the overuse of de can exclusively be ascribed to the effects of bilingual (2L1/child L2) acquisition. It is argued that although the overgeneralization of de constitutes a linguistic resource for every bilingual child acquirer (and even for monolingual acquirers), it only becomes meaningful in the indexing and reproducing of an (immigrant) "allochtone" identity versus the dominant (indigenous) "autochtone" one. It is used according to a set of language use norms defined by generations of that social group of speakers.

Key words

bilingual

acquisition

grammatical

gender

identity

1 Introduction

Grammatical gender in Dutch distinguishes between neuter and common nouns. This gender distinction is marked morphologically on the determiner if it has the features singular and definite: neuter nouns take het and non-neuter nouns take de. In contrast, no gender distinction is reflected on the indefinite, plural, and indefinite article in Dutch, as illustrated in Figure 1. (1)

In the Dutch sociolinguistic literature, it is often noted that overgeneralization of de characterizes the Dutch variety(ies) of immigrants and their native-born descendants (Aissati, Boumans, & Cornips, et al., 2005; Cornips, 2002, 2004; Cornips & de Rooij, 2003). The immigrants and their descendants use, in other words, de where het is required. According to Muysken (1984) and Snow, Van Eeden, and Muysken (1981), the Turkish and Moroccan immigrants entering the Netherlands in the 1960s frequently deleted the definite determiner and/or overgeneralized common gender. (2) More specifically, Muysken (1984) describes the spontaneous speech of a 21-year-old Moroccan who had been living in the Netherlands for three years. His proficiency was above average. However, his utterances showed that articles were often deleted or that they were erroneously used for example, de instead of het.

Recent sociolinguistic research shows also that this overgeneralization does not disappear; that is, the subsequent generations of the immigrants who are born in the Netherlands and acquire Dutch as one of the languages from birth onwards show this overgeneralization as well, as illustrated in (1) (cf. Cornips 2002; Cornips, 2004; Cornips & de Rooij, 2003): (1) a. zitten we in de laatste jaar are we in the final year [the-common year-NEUTER, Cengiz, Turkish descent, Utrecht, adolescent] b. de meest serieuze type the most serious type [the-common type-NEUTER, Abdelkhalek, Moroccan descent, Berber, Utrecht, adolescent] c. Hij had de juiste merk aan 'he was wearing the right brand' [the-common brand-NEUTER, Surinamese descent, Rotterdam, adolescent] d. interviewer: "deze vrouw snijdt de tomaat en deze vrouw ..." 'this woman cuts the tomato and this woman ...' Ayni: "snijdt de brood." cuts the bread [the-common bread-NEUTER, Ayni, Bolivian descent, Amsterdam, 6 yrs]

Thus, the bilingual (and monolingual) descendants of the former immigrants have a potential role in contact effects as observed in creole and other languages. Hence, it is not only the case that grammatical gender is disappearing in Dutch contact varieties in the Netherlands but it has already disappeared in contact situations outside the Netherlands. It took place in Dutch lexifier contact languages such as Negerhollands (Muysken, 2001, p. 165), Berbice Dutch, Afrikaans (Donaldson, 1993; Ponelis, 2005), Curacaos-Dutch (Joubert, 2005), Surinamese-Dutch (Cornips, 2005), and Indisch-Dutch (de Vries, 2005).

Interestingly, the overgeneralization of one gender takes place in immigrant communities with other languages too, such as the suburbs of Sweden where immigrants and their children learn Swedish as the target language (Kotsinas, 2001, p. 150). Also Quist (this volume) describes the extended use of the common gender where standard Danish nouns require neuter articles.

Taken together, grammatical gender constitutes a vulnerable domain for variation and change in language contact situations The aim of this paper is to discuss to what extent the overgeneralization of the definite determiner de is the result of processes of (early) child second language and/or bilingual acquisition (Cornips & Hulk, 2008; Hulk & Cornips, 2006b). In addition I will consider what other factors should be taken into consideration to explain this phenomenon.

The bilingual acquisition of grammatical gender in Dutch by Moroccan and Turkish children (Cornips et al. 2006)

2.1

At the end of the 1990s, a considerable amount of spontaneous speech of Moroccan, Turkish and Surinamese teenagers and adolescents in Utrecht was collected (approximately 8 hours. In Cornips (2002), the spontaneous speech of six male speakers who were mutual friends in the Utrecht neighborhood Lombok/Transvaal was examined. These speakers fell into three groups according to their language background: four Moroccan Arabic/Berber/Dutch, three Turkish Dutch bilinguals and one Surinamese (Hindi) Dutch. All speakers were in their early twenties. This group of friends is special in that they use Dutch to communicate with each other. Note that without Dutch, the young Moroccans are not able to communicate with each other since some speak Arabic and others Berber.

All participants, irrespective of the other languages which they speak, reveal a tendency to use the definite determiner de in cases where the definite determiner het is required, as illustrated in (2) (2) a. de geloof [Guray, Turkish descent] the-COMMON faith-NEUTER b. de hele gebouw [Hassan, Turkish descent] the-COMMON entire building NEUTER c. de laatste jaar [Hassan, Turkish descent] the-COMMON last year-NEUTER d. de boek [Badir, Moroccan descent] the-COMMON book-NEUTER e. de kanaal [Naraen Surinamese descent] the-COMMON channel-NEUTER

This tendency is much stronger than the other way round in which the definite article het is used for de: (3) a. bet nadruk [Guray, Turkish descent] the-NEUTER emphasis-COMMON b. het muziek [Badir, Moroccan descent] the-NEUTER music-COMMON

More evidence that neuter nouns are considered as common is the use of the deictic pronouns die and deze that combine with neuter nouns instead of the standard Dutch dat and dit which must precede neuter nouns: (4) a. die optreden [Naraen, Surinamese descent] this COMMON performance NEUTER b. deze gedeelte [Naraen, Surinamese descent] this COMMON part NEUTER c. die nummer [Hasan, Turkish descent] this COMMON number NEUTER d. die eten [Naraen, Surinamese descent] this COMMON food NEUTER

Strikingly, also diminutives--although they are always neuter in standard Dutch--may be treated as common nouns: (5) a. de filmpje [Badir, Moroccan descent] the COMMON movie NEUTER+DIM b. de vaste groepje [Youssef, Turkish descent] the COMMON steady group NEUTER+DIM

However, it is important to emphasize that it is only in the DP domain that these speakers differ from standard Dutch speakers. From a morphosyntactic perspective, the Moroccan and Turkish adolescents are standardlike with respect to other phenomena in Dutch such as negation, expletive constructions and word order (Cornips, 2002). Furthermore, in Cornips (2002) it is shown that the Moroccans and Turks in the group described above who overuse common gender most frequently, also use local substandard features most frequently. This is again an indication of their native speaker status, hence, proficiency in Dutch.

2.2

Experimental research

This section will describe the experimental research by Hulk and Cornips (2006) by focusing on one specific research question: Do children of Turkish and Moroccan descent raised in an ethnic minority community in The Netherlands reveal a nontarget-like acquisition of grammatical gender of the Dutch definite determiner, and if so, is this "fossilization-effect" also present in their monolingual peers?

In order to examine whether bilingual children reveal a fossilization effect, Cornips, Van der Hoek, and Verwer (2006) selected older children of Moroccan and Turkish descent, varying in age between 10;5 and 12;11 years old who all attend one primary school located in an ethnic minority neighborhood in Rotterdam. Furthermore, in order to determine the presence or absence of cross-linguistic influence, the authors carefully included both Moroccan-Arabic- and/or Berber- and Turkish-speaking bilingual children whose other language instantiates a gender distinction in its noun/determiner system or without gender altogether, respectively. Therefore, a group of 28 children was made up of (i) 12 bilingual children from Turkish descent, (ii) 12 bilingual children from Moroccan descent speaking Berber and/or Moroccan-Arabic and 4 Dutch monolinguals (of Dutch descent) as a control group. A language profile was obtained from these 24 Moroccan and Turkish children that provided not only information about competences in and preferences for languages spoken in the home environment but also about the language(s) spoken to and by parents, older and younger siblings, grandparents, and friends (see appendix). It is important to note that this language profile reveals that all children are born and raised in the Netherlands. Consequently, these children are exposed to Dutch from birth onwards although more research has to be done to find out the details about the quantity and quality of the input to which they were exposed. Strikingly, Dutch is not spoken by the parents in only three Moroccan (out of 12) and five Turkish families (out of 12). Although those parents don't speak Dutch at all in the home domain, these eight children report speaking Dutch with their siblings. An interview fragment with Nabil of Moroccan descent, selected in other research, nicely illustrates this:

Nabil 10;5 Moroccan descent

Interviewer: Wat spreken je ouders?

Nabil: Marokkaans.

Nabil: Maar ik spreek altijd met mijn zus Nederlands.

Interviewer: What language do your parents use?

Nabil: Moroccan.

Nabil: But I always speak Dutch to my sister

Moreover, it is striking that the majority of the 24 children prefer to exclusively speak Dutch with their friends and consider Dutch to be their best language. Thus, even if the parents do not speak Dutch to the child, this does not mean that the home-situation is predominantly monolingual Turkish or Moroccan/Arabic or Berber. Thus the role of the siblings seems to be crucial in the exposure of Dutch from birth onwards (Hazen, 2002, Milroy & Li Wei, 1995). According to Hazen (2002, p. 516) the differences between older and younger siblings may point to the fluid and complex nature of language variation in migrant families. The children who most closely follow the language patterns of their parents are those in families more recently arrived. As the social integration of the family changes, the effects on older and younger siblings will differ. Whereas an older sibling may be more connected to the family when the family arrived recently, a younger sibling a few years later may have more opportunity for peer group interaction (Hazen, 2002, p.516)

2.3

Definite determiner: Fossilization and crosslinguistic influence?

2.3.1

The bilingual children

The results regarding the (non)target use of the definite determiners de and het are presented in Table 1 below. It reveals that both the monolingual and bilingual children (i) produce a higher score of the common determiner de than the neuter determiner het and (ii) overgeneralize de, namely 18.7% and 48.96% respectively. It is important to point out that experimental and spontaneous data have shown that monolingual acquisition of neuter gender of the Dutch definite determiner is a long-lasting process in the sense that the monolingual ethnic Dutch children do not acquire a target grammar until relatively late compared with other target language properties, namely age six. (cf. references in Hulk & Cornips, 2006a,b). Overall, the bilingual children reveal lower percentages compared to the monolingual children, especially with respect to the target-like use of het (42.01% vs. 68.8%, respectively). Note that this target-like use is below chance level. (3)

Since the bilingual children in this experiment are already relatively old, that is, between 10;5 and 12;11 years, it is likely that their grammar has "fossilized" with respect to the acquisition of the neuter definite determiner (cf. Hulk & Cornips, 2006a,b). However, only if we were to follow these children longitudinally would we be able to conclude that the results in Table 1 display their ultimate attainment.

Furthermore, the results in Table 2 reveal that having a noun/determiner system with a gender distinction in the other language (Moroccan-Arabic and/or Berber) or not (Turkish) is not crucial. The Moroccan children do not differ significantly from the Turkish children or vice versa:

2.3.2

The monolingual children

Interestingly, Table 1 presented above also shows that the monolingual children do not produce a 100% correct score for both the definite determiners de (83.3%) and het (68.8%) at an age between 10;5 and 12;11 years old. It seems that the monolingual children in a primary school located in an ethnic minority community are sensitive to peer groups effects with bilingual acquirers of Dutch in the sense that their standard-like use of het is rather low. More evidence that monolingual children are sensitive to input from their bilingual peers comes from a study by Brouwer, Cornips, & Hulk (2008). They conducted an experiment to "tap the knowledge" of monolingual children (and bilinguals) in two age groups (6-8 and 11-13 year-olds) about grammatical gender in Dutch. It appears that the location of the primary school in a monolingual or multilingual neighborhood is an important factor in determining the effects of the social stratification of the target-like awareness of grammatical gender in Dutch. Brouwer et al. distinguished two sociolinguistic contexts for the monolinguals, namely, monolinguals in a 100% and in a 75% monolingual school. It appears that the two monolingual groups differ significantly with respect to the awareness of grammatical gender in Dutch, as presented in Table 3: (4)

Thus, the acquisition of grammatical gender, in addition to internal linguistic factors, is (Cornips & Hulk, 2008) extremely sensitive to input so we may assume that it is sensitive to peer-group effects as well.

Until now, the experimental results show that the Moroccan and Turkish children do not differ significantly. So, the overgeneralization of de is not restricted to one group with a specific national and/or cultural, linguistic origin, that is, only the Turkish or Moroccan children. Subsequently the concept of ethnolect or ethnicity as an independent variable assuming a fixed link between an observed linguistic variant and ethnic identity is extremely doubtful (Cornips & de Rooij, 2004; de Rooij, 2005).

In order to find out if, and to what extent, the overgeneralization of de is restricted to children of Moroccan and Turkish descent, let us consider a completely different group of child second or child bilingual acquirers, namely English-Dutch bilingual children from expatriate families (cf. Unsworth 2006).

3 Bilingual children from a "one parent-one language" setting

The children of Turkish and Moroccan descent who participated in the experiments described above differ considerably from those who have generally been examined in "one parent--one language" research. The bilingual children from ethnic minority communities (i) belong to lower-class families, (ii) have been raised bilingually in a bilingual, that is, ethnic minority community, (iii) have one or both parents who are not native speakers of Dutch, and (iv) have been raised bilingually within a family in which language choice "rules" do not reflect a "one parent, one language" setting at all. Instead, according to Auer (1995, p, 115), in many if not in all "new" bilingual communities that have come into being in Europe as a consequence of work-related migration, language choice is often indecisive.

However, in "one parent--one language" research the bilingual children such as the English-Dutch children we will discuss below (i) belong to middle-class families, (ii) have been raised bilingually from birth in an otherwise monolingual community, and (iii) have parents who speak their respective varieties as native (L1) speakers (Meisel, 1989). Clearly, with respect to linguistic behavior, these families function as small, isolated groups in which the language choice rules are not at all reflected in the surrounding community.

In order to find out whether overgeneralization of de is restricted to bilingual child acquirers descending from immigrant communities, I will present the results of a study of bilingual English-Dutch children by Unsworth (2006). Unsworth (2006) examined 58 bilingual English-Dutch children who all attended international schools in The Netherlands at the time of testing, and some of them had previously attended Dutch pre-/primary schools. The children were predominantly from middle class families who had moved to The Netherlands for professional reasons. Their age at the time of testing ranged from 5;3 to 17;4 (mean 10;5, SD 3;8), their age of first exposure to Dutch from birth to 7;3 (mean 4;10; SD 1;9) and their length of exposure from 0;11 tot 15;2 (mean 5;6; SD 3;8). Unsworth raises two research questions: (i) do English-speaking bilingual children also overgeneralize the definite determiner de and (ii) if so, do they fossilize in this stage? In order to find out, she collected semispontaneous data using a picture description task. The results are presented in Table 4:

Table 4 displays that the English-Dutch children overgeneralize de for neuter nouns to a high extent, namely 74.6%. (5) Thus, the question whether bilingual children from a different type of bilingual population also show loss of grammatical gender of the definite determiner can only be answered affirmatively. What is important to point out is that Unsworth (2006) convincingly discusses that the use of target-like het with neuter nouns increases significantly with (i) quantity of exposure (length) to Dutch, (ii) with the quality of exposure (intensity) to Dutch and with (iii) proficiency in Dutch but in all cases the different groups are still only at chance level, that is, the target-like score of het varies between 34.3% and 51%.

Thus, both Moroccan/Turkish immigrant and English expatriate children fail to produce het with neuter nouns in Dutch in a target-like way, even when they are rather old. From this, we have to conclude that overgeneralization of de is not simply due to crosslinguistic influence (or so-called interference, substrate effects) from the other language having a gender/noun distinction in its grammar (Moroccan-Arabic/Berber) or without this distinction (Turkish, English). Importantly, it appears that the overgeneralization of de is a resource for every bilingual child acquirer in Dutch and even for monolingual children acquiring Dutch in a multilingual context.

4 Ethnolect: Process of "ethnic" identification and stereotyping

Until now, subjects are categorized as "Moroccan," "Turkish," and "English" on the basis of their other individual language(s). However, these categories do not imply that the individuals within these categories are homogenous with respect to their experiences as immigrants, newcomers to Dutch society, gender, education, SES and neighborhoods where they live, religious and cultural associations, and so forth (see also Cutler, this volume; Fought, 2006). There is no necessity that the outsiders' categorization as "Moroccan," for instance, signals group membership from the insiders' view. For example, Koole and Hanson (2002) discuss a classroom discourse in the Netherlands in which a primary school teacher uses the national identity category 'Moroccan' to identify pupils inside and outside the classroom and attributes meanings to this category in terms of category-bound activities such as "Moroccan women wear headscarves for religious purposes." However, although the teacher uses "Moroccan," many children belonging to this category from an outsiders' view do not adopt Moroccan to refer to themselves. Social groups such as ethnic ones need not be defined beforehand. The concept of ethnolect focusing on migrant communities categorized on the grounds of region of origin, ethnicity, and religion might essentialize groups and ignore their overlapping networks. Hence, as is the case in the Dutch context, young native-born descendents of immigrants meet each other in very diverse urban contexts--as do other newcomers--creating a linguistic diversity or using linguistic variants which are not necessarily "theirs" through family heritage or they may use variants which "are not generally thought to belong to them" for purposes of creating alternative solidarities (Rampton, 1995). As Fought (2006, p. 38) discusses, we cannot say that members of an ethnic group will construct an ethnic identity in the same linguistic way all the time. There are different types of contexts for the construction of ethnicity through language, including at the very least: (a) the situational context which can vary from one conversation to the next or even within a single conversation, (b) the broader context of an individual's life history, and (c) the social context of how the community, or a segment of the community, views ethnicity at a particular point in time. Essential in the concept of ethnolect is the important role of both self-identification and the perceptions and attitudes of others in the construction of ethnic identity.

The meaning of ethnicity and/or ethnic groups (ethnic variety) in the Netherlands developed in the context of a theory that focused on immigrants of Moroccan and Turkish descent on a former chain-immigration setting. Moroccans and Turks are treated as a single entity by the dominant ideology (cf. Fought, 2006, p. 15). What Turks and Moroccans in The Netherlands have in common are their size, cultural backgrounds, socioeconomic status, and the same historical demographic patterns of immigration. Turkish and Moroccan men who immigrated to the Netherlands in the sixties came to work as unskilled laborers in the industrial sector. The majority of these laborers originated from the countryside where the educational facilities were very poor. Over 90% of the Moroccan and about 55% of the Turkish inhabitants in The Netherlands between 50 and 64 years old did not finish primary education (Jongenburger & Aarssen, 2001). According to Dabene and Moore (1995, p.21), the early migrants encouraged family and village members to travel to the host county and join them in work. Newly arriving migrants often had friends and acquaintances in the closed group of people they already knew, directly or via other people, in their place of origin. Subsequently, specific migratory chains rapidly emerged, leading new arrivals to settle in specific city neighborhoods in the host country, and voluntary clustering and social encapsulation are strengthened. This explains the emergence of ethnic enclaves in cities with large migrant populations. This clustering, reinforced by the arrival of women and children in later years, allows groups to function as a micro-society or an extended family.

In the Dutch context, the arrival of work-related immigrants resulted in the construction of an "us" versus "them" opposition, to be more specific, an "autochtone" versus an "allochtone" identity, respectively. (6)

Although the overgeneralization of the definite determiner de constitutes a linguistic resource for every bilingual child acquirer (and even for monolingual acquirers), it only becomes meaningful in a process of ethnic identification (de Rooij, 2005), that is to say, it is an element in the indexing and reproducing of an (immigrant) "allochtone" identity versus the dominant "autochtone" one. There are several pieces of evidence.

First, as a group and even as individuals, the English-Dutch children from expatriate families are highly invisible in Dutch society. They are not openly commented upon and there is not something like an English/American community or neighborhoods in The Netherlands. Second, Unsworth's study shows that the English-Dutch children show progress in the target-like use of het with length (and extent) of exposure to Dutch and with proficiency. However, the children, teenagers, and adolescents from Turkish and Moroccan (and Surinamese) descent are exposed to Dutch from birth onwards and their proficiency in Dutch is target-like (§ 2.1). Third, the young people from Moroccan and Turkish descent are highly visible in Dutch society: many of them live in ethnic minority communities in cities and they are frequently commented upon by members of Dutch society and politicians. Finally, the overgeneralization of de has the status of a stereotype (Labov, 1994). The use of this variant is sufficient to stereotype young people as "allochtone" (immigrant) or non-Western originally. This single feature of language use suffices to identify someone's membership of a given group (Tabouret-Keller, 1997, p. 317).

Fragment 1:

Jeroen: egen wie ben je aan 't praten, tegen wie ben je aan 't praten?

Martje: egen die cling [ ... ]

Jeroen: ho are you talking to, who are you talking to?

Martje: to that thing [ ... ]

meta: die ding [that.sub.COMMON] [thing.sub.NEUTER]

Fragment 2:

Jeroen: Wij even wat inspreken op deze bandje?

Jeroen: We say something on this tape?

meta: deze bandje [this.sub.COMMON] tape-[DIM.sub.NEUTER]

Cornips and de Rooij (2004) discussed young Surinamese who imitate or stereotype Antillians in Rotterdam by means of nontarget like use of the common demonstrative pronoun. They use common die instead of neuter dat 'that'. Ronald, Vincent, and Gerard in the fragment below are of Surinamese descent and belong to a core-group of four males who are mutual friends (cf. Cornips, 2004).

Interviewer: kunjij ook Antilliaans.

Ronald: dumpen+ ...

Interviewer: hah?

Gerard: gewoon straattaal Antilliaans xx.

Interviewer: xx eenvoorbeeld?

Gerard: ja dat, ik schop die ding terug.

meta: they are imitating young Antillian, he laughs

Ronald: ik heb gezien dat je xx.

meta: they are imitating young Antillians

Vincent: ik rende voor die bus, maar die bus rende weg.

meta: they are imitating young Antillians

Ronald: ik heb gezien dat je die jongen hebt geschiet.

Vincent: ik schiet je met die mes.

meta: they are imitating young Antillians

Ronald: jongen met die broek met die korte mouwen.

meta: they are imitating young Antillians

Interviewer: maar dat is geen Antilliaans?

meta: she laughs

Vincent: zo praten ze.

Interviewer: zo praten ze.

Interviewer: are you able (to speak) Antillian [Papiamentu]

Ronald: throw away

Interviewer: hat?

Interviewer: n example?

Gerard: es, that, I kick that thing back

meta: hey are imitating young Antillians, he laughs

%morph die ding [that.sub.COMMON] [thing.sub.NEUTER]

Ronald: I saw that you xxxx

meta: they are imitating young Antillians

Vincent: I tried to catch the bus but the bus ran away

meta: they are imitating young Antillians

Ronald: I saw that you shot that boy

I shoot you with that knife

meta: they are imitating young Antillians

%morph die mes [that.sub.COMMON] [knife.sub.NEUTER]

Ronald: boy with those trousers and short sleeves

meta: they are imitating young Antillians

Interviewer: but that's no Antillian [Papiamentu]?

meta: they laugh

Vincent: that's the way they talk

Interviewer: that's the way they talk

Furthermore, an interview with Brian of Surinamese descent reveals that by means of the overgeneralization of common de, he creates an opposition between Rotterdam and Amsterdam and between Surinamese and Antillians young people (Cornips, de Rooij, & Reizevoort, 2006):

Fragment 4:

[Brian, 33 years, Surinamese descent, Amsterdam, Steve Bikoplein, date 05-11-2004]

Interviewer: Zeg je ook, of zeg jij ... zeggen ze ook: "wil je even wat inspreken op deze bandje"?

(...)

Interviewer Maar ook dingen als die boek en zo? Dat is vaak toch, die boek?

Brian: Maar aaa ... alles is bijna die.

(...)

Brian: (...) eh ... die komt meestal voor in Rotterdam.

Interviewer: Ja?

Brian: ... of je nou blank bent of niet, alles is daar is die. Alles, toch, bij alles gewoon die. En Amsterdam ... Amsterdammers hebben dat gewoon minder. Maar Surinamers ... nee, die is echt een probleem bij de Antilliaanse mensen. Bij hun is alles die.

Interviewer Ja.

Brian Die meisje ... die ... bij Surinamers dat valt nog mee.

Interviewer: Do you say also, do you say also' Do you want to say something on this tape"?

%morph: deze bandje [this.sub.common] [tape.sub.DIM+neuter]

(...)

Interviewer and things as this book and so? It is often that book right?

%morph: die boek [that.sub.common] [book.sub.neuter]

Brian: But ... everything is almost [die.sub.common]

(...)

Brian: die often occurs in Rotterdam

%morph: die [that.sub.common]

Interviewer: yes?

Brian: whether you are white or not, everything there is die. Everything, everything just die. And Amsterdam, Amsterdam people have that less. But Surinamese people ... no die is really a problem for Antillian people. Everything is die for them.

Interviewer: yes

Brian: that girl.... that ... it's not so common among Surinamese people

%morph: that common girl neuter

Importantly, Nortier and Dorleijn (this volume) point out that teenagers and adolescents Moroccan descent are developing both linguistic norms and norms of stylistic appropriateness, and overgeneralization of common gender is crucially involved, as is illustrated by the following quotation from a conversation with an informant (S) from Rotterdam (J and M are the interviewers):

S: Dat is het slechte Nederlands

J: En heeft dat ook een naam?

S: Ja, niet echt, maar 't is in principe dan eh lidwoorden die gebruik je dan expres verkeerd

M: Ja ja die gebruik je dan expres verkeerd, net als

S: Ja, dus

J: Die meisje

S: Die huis zeg ik dan. Terwijl ik weet ik bedoel ik weet heus wel dat het dat huis is maar 't staat zo dom als ik dat op straat zeg, als ik zeg

M: Ja

S: Als ik zeg dat huis

M: Jaja

S: 't Is gewoon die huis. Maar als ik met jullie spreek dan wordt 't gewoon dat huis.

S: That is the bad kind of Dutch.

J: Does it have a name?

S: No, not really, but in principle you uhmm ... just use the articles Deliberately in the wrong way.

M: Right! So you use them in the wrong way deliberately? Just like

S:--Yes, like

J: Die meisje ('that girl'--dat meisje in standard Dutch)

S: I would say: Die huis ('that house--dat huis in standard Dutch). At the same time I know, I mean, I am very well aware of the fact that it should actually be het huis, but it would make a dumb impression if I would say....

M: Yes

S: If I would say dat huis out on the street

M: Yes, yes.

S: It is just die huis. But when I speak with you two (the interviewers, both Dutch and middle-aged JN / MD) it is just dat huis.

"The speaker in the quotation explicitly says that you have to make errors, deviations from the standard norm, in order to be recognized as someone who is hanging out with friends" (cf. Nortier & Dorleijn, this volume). This quote suggests that the "optional" nature that goes with a minority ethnic identity affects its linguistic expression. For the older generation, it was a learner characteristic; the younger generations born in The Netherlands are able to downplay or to put it forward (Fought, 2006, p. 93).

To conclude, the overgeneralization of de cannot be ascribed to bilingual acquisition effects alone, but is governed by linguistic norms of the members of a group (Le Page & Tabouret-Keller, 1985).

5 Conclusion

The aim of this paper was to discuss whether the overuse of the singular definite determiner de instead of het, could exclusively be ascribed to the effects of bilingual (2L1/child L2) acquisition. It is argued that although the overgeneralization of de constitutes a linguistic resource for every bilingual child acquirer (and even monolingual acquirers), it only becomes meaningful in the indexing and reproducing of an (immigrant) "allochtone" identity versus the dominant "autochtone" one. Loosing grammatical gender plays a crucial role in the process of stereotyping "the other." On the one hand, from an outsiders' perspective, it is a characteristic ascribed to young people who are descendants from immigrant groups (excluding the English-Dutch speaking group). On the other hand, it indexes group membership and in-group speech and may serve as an ethnic marker or stereotype. It is this indexing group membership that explains why the older Moroccan and Turkish children fossilize in a nontarget-like stage although they have been exposed to Dutch from birth onwards. Their "incomplete" learning of the Dutch determiner system is incorporated into their language and "incomplete" learning becomes meaningful, not a mistake, and is used according to a set of language use norms defined by generations of that social groups of speakers. Thus, they acquire a different variety of Dutch successfully, that is to say, an insiders' variety of Dutch. Appendix Language profiles--reported behaviour by the children: Moroccan and Turkish bilingual children in a multilingual setting (NL = The Netherlands; M = Moroccan Arabic/Berber; T = Turkish) country country country of birth of birth name age of birth mother father Ihsan 11;0 NL NL M Hicham 12;8 NL M M Malika 12;7 NL M M Hakim 12;10 ? M M Loubna 11;5 ? M M Mohamed 12;5 NL M M Ali 11;6 NL M M Jihad 12;10 NL M M Kaoutar 12;5 NL M M Latifa 12;8 NL M M Yousra 11;10 NL M M Soufiane 12;1 NL M M Emre 11;0 NL T T Merve 12;11 NL T T Niyazi 12;2 NL T T Serdal 12;5 NL T T Tolga 12;9 NL T T Elif S 11;8 NL T T Faruk 11;8 NL T T Melis 12;5 NL T T Elif A 11;9 ? T T Ramazan 12;9 ? T T Berna 11;2 ? T T Cem 10;9 NL NL T Cemre 11;0 NL T T home mother father grandfather name language speaks speaks speaks Ihsan M M M M Hicham M+D M M M Malika M+D M M M Hakim M+D M M+D Loubna M M M+D M Mohamed M+D M D+M M Ali M+D M+D M+D M+D Jihad M+D M+D M M Kaoutar M+D M+D M+D Latifa M+D M+D M M+D Yousra M+D M+D M M Soufiane M+D D D M Emre T+D T T T Merve T T T T+D Niyazi T T T T Serdal T+D T T T Tolga T+D T T T Elif S T+D T+D T T Faruk T+D T+D T T Melis T+D T+D T+D T+D Elif A T+D T+D T+D Ramazan T+D T D T Berna T+D D T T Cem T+D D T T Cemre T+D D T T grandmother older younger name speaks siblings siblings Ihsan M M+D Hicham M M+D D Malika M D Hakim M M+D D Loubna M M+D Mohamed M M+D M+D Ali M+D M+D M+D Jihad M M+D M+D Kaoutar M D D Latifa M Yousra M M+D M+D Soufiane M D D Emre T T+D Merve T+D Niyazi T T+D Serdal T T+D T+D Tolga T T+D Elif S T T+D Faruk T D D Melis T+D T+D Elif A T+D Ramazan T D D Berna T D Cem D+T D+T Cemre T D lang. most with best favourite name friends language language Ihsan M+D M+D D Hicham M+D D M Malika D D D Hakim M+D M+D M+D Loubna D M+D Mohamed M+D M+D M+D Ali D Jihad M+D D M Kaoutar D D D Latifa D D M+D Yousra D M D Soufiane D D M Emre T+D T+D T+D Merve T T+D T Niyazi T+D T T+D Serdal T+D D D Tolga T+D T+D D Elif S T+D D T+D Faruk T+D D T Melis T+D T+D T+D Elif A T+D T+D T+D Ramazan D D D Berna T+D D T Cem D+T D T Cemre D T+D T+D

Acknowledgment

I would like to thank two anonymous reviewers for their valuable comments.

References

AISSATI, A. El-, BOUMANS, L. & CORNIPS, L., DORLEIJN, M., & NORTIER, J., (2005). Turks- en Marokkaans Nederlands [Turkish and Moroccan Dutch]. In N. van der Sijs, (Ed.), Wereldnederlands. Oude en jonge varieteiten van het Nederlands (pp. 149-183). The Hague: SDU.

AUER, P. (1995). The pragmatics of code-switching: A sequential approach. In L. Milroy, & P. Muysken (Eds.), One speaker, two languages: Cross-disciplinary perspectives on code-switching (pp. 115-135). Cambridge: Cambridge University Press.

BROUWER, S., CORNIPS, L., & HULK., A. (2008). Misrepresentation of Dutch target-like neuter gender in older bilingual children? In B. Haznedar & E. Gavruseva (Eds.), Current trends in child second language acquisition: A generative perspective. Amsterdam: John Benjamins.

CORNIPS, L. (2002). Ethnisch Nederlands [Ethnic Dutch]. In H. Bennis, G. Extra, P. Muysken, & J. Nortier (Eds.), Een Buurt in Beweging. Talen en Culturen in het Utrechtse Lombok en Transvaal (pp.285-302). Amsterdam: Aksant.

CORNIPS, L. (2004). Straattaal: Sociale betekenis en morfo-syntactische verschijnselen [Streetlanguage and morphosyntactic phenomena]. In J. de Caluwe, G. de Schutter, M. Devos, & J. Van Keymeulen (Eds.), Taeldeman, man van de taal, schatbewaarder van de taal (pp. 175-188). Gent: Academia Press.

CORNIPS, L. (2005). Het Surinaams-Nederlands in Nederland [Surinamese-Dutch]. In N. van der Sijs (Ed.), Wereldnederlands. Oude enjonge varieteiten van bet Nederlands (pp. 131-147). The Hague: SDU.

CORNIPS, L., & HULK., A. (2006). External and internal factors in bilingual and bidialectal language development: Grammatical gender of the Dutch definite determiner. In C. Lefebvre, L. White, & C. Jourdan (Eds.), L2 acquisition and creole genesis. Dialogues (pp. 355-378). Amsterdam / Philadelphia: John Benjamins.

CORNIPS, L., & HULK., A. (2008). Factors of success and failure in the acquisition of grammatical gender in Dutch. Second Language Research.

CORNIPS, L., Van der HOEK, M., & VERWER, R. (2006). The acquisition of grammatical gender in bilingual child acquisition of Dutch (by older Moroccan and Turkish children). The definite determiner, attributive adjective and relative pronoun. In B. Los & J. van de Weijer (Eds.), Linguistics-in-The Netherlands 2006 (pp.40-51). Amsterdam: John Benjamins.

CORNIPS, L., & De ROOIJ, V. (2003). Kijk, Levi's is een goeie merk: Maar toch hadden ze 'm gedist van je schoenen doen 'm niet. Jongerenvarieteit heeft de toekomst. In J. Stroop (Ed.), Het Nederlands van Nu en Straks. Waar gaat het Nederlands naar toe? (pp. 131-142). Amsterdam: Prometheus.

CORNIPS, L., & De ROOIJ, V. (2004, October). The concept of "neger" (negro) in a Rotterdam youth language and culture (The Netherlands). Talk presented at NWAV 33, University of Michigan, Ann Arbor, U.S.A.

CORNIPS, L., & De ROOIJ, V., & REIZEVOORT, B. (2006). Straattaal: Processen van naamgeving en stereotypering. Toegepaste Taalwetenschap in Artikelen 76. Thema's en trends in de Sociolinguistiek 5(2), 123-136.

CUTLER, C. (2008, this volume). Brooklyn style: Hip-hop markers and racial affiliation among European immigrants in New York City. International Journal of Bilingualism, 12(1,2), 7-24.

DABENE, L., & MOORE, D. (1995). Bilingual speech of migrant people. In L. Milroy & P. Muysken, (Eds.), One speaker, two languages: Cross-disciplinary perspectives on code-switching (pp. 17-44). Cambridge: Cambridge University Press.

De HOUWER, A. (1990). The acquisition of two languages from birth: A case study. Cambridge: Cambridge University Press.

De ROOIJ, V. (2005, November). De studie van etnolecten: Overdenkingen vanuit een etnografisch perspectief [The study of ethnolects. Reflections from an ethnographic perspective]. Talk presented at the Workshop Etnisch Nederlands, Utrecht University, Utrecht, The Netherlands.

De VRIES, J. W. (2005). Indisch-Nederlands. In N. van der Sijs (Ed.), Wereldnederlands. Oude en jonge varieteiten van het Nederlands (pp. 59-77). The Hague: SDU.

DONALDSON, B. (1993). A grammar of Afrikaans. Berlin: Mouton de Gruyter.

DORLEIJN, M., & NORTIER, J. (2006). "Het is een grappige accent weet je": Het opvallende prestige van de Marokkaanse tongval. Onze Taal, 75(2/3), 47-49.

ECKERT, P., & RICKFORD, J. (2001). Style and sociolinguistic variation. New York: Cambridge University Press.

FOUGHT, C. (2006). Language and ethnicity. Cambridge: Cambridge University Press.

HAZEN, K. (2002). The family. In J. Chambers, P. Trudgill, & N. Schilling-Estes (Eds.), The handbook of language variation and change (pp. 500-525). Malden, MA, and Oxford: Blackwell.

HEIDELBERGER FORSCHUNGSPROJEKT PIDGIN-DEUTSCH (HPD) (1978). The unguided learning of German by Spanish and Italian workers. UNESCO Report SS-78/Conf. 801/4.

HEWITT, R. (1986). White talk, black talk. Cambridge: Cambridge University Press.

HULK., A., & CORNIPS, L. (2006a). Between 2LI and child L2 acquisition: An experimental study of bilingual Dutch. In C. Lieo (Ed.), Interfaces in multilingualism: Acquisition, representation, and processing (pp. 115-138). Hamburg.

HULK, A., & CORNIPS, L. (2006b). The acquisition of definite determiners in child L2 Dutch: Problems with neuter gender nouns. In S. Unsworth, T. Parodi, A. Sorace, & M. Young-Scholten (Eds.), Paths of development in LI and L2 acquisition (pp. 107-134). Amsterdam: John Benjamins.

IRVINE, J., & GAL, S. (2000). Language ideology and linguistic differentation. In P. Kroskrity (Ed.), Regimes of language: Ideologies, polities and identities (pp. 35-84). Santa Fe, New Mexico: School of American Research Press.

JONGENBURGER, W., & AARSSEN, J. (2001). Linguistic and cultural exchange and appropriation. A survey study in a multi-ethnic neighbourhood in the Netherlands. Journal of Multilingual and Multicultural Development, 22(4), 293-308.

JOUBERT, S. M. (2005). Curacaos-Nederlands. In N. van der Sijs (Ed.), Wereldnederlands. Oude en jonge varieteiten van het Nederlands (pp. 31-58). The Hague: SDU.

KOOLE, T., & HANSON, M. (2002). The category "Moroccan" in a multi-ethnic class. In S. Hester & W. Housley (Eds.), Language, interaction and national identity. Studies in the social organisation of national identity in talk-in-interaction (pp. 211-232). Burlington: Ashgate.

KOTSINAS, U. (2001). Pidginization, creolization, and creoloids in Stockholm, Sweden. In N. Smith & T. Veenstra (Eds.), Creolization andcontact (pp. 125-156). Amsterdam: John Benjamins.

LABOV, W. (1994). Principles of linguistic change: Internal factors. Oxford: Blackwell.

LABOV, W. (2000). Principles of linguistic change. Social factors. Oxford: Blackwell.

Le PAGE, R., & TABOURET-KELLER, A. (1985). Acts of identity. Cambridge: Cambridge University Press.

MEISEL, J. (1989). Early differentiation of language in bilingual children. In K. Hyltenstam & L. Obler (Eds.), Bilingualism across a lifespan: Aspects of acquisition, maturity and loss (pp. 13-40). Cambridge: Cambridge University Press.

MILROY, L., & LI WEE (1995). A social network approach to code-switching: The example of a bilingual community in Britain. In L. Milroy & P. Muysken (Eds.), One speaker, two languages: Cross-disciplinary perspectives on code-switching (pp. 136-157). Cambridge: Cambridge University Press.

MUYSKEN, P. (1984). Attitudes and experiences of discrimination: The Netherlandic of Moroccan foreign workers. In K. Deprez (Ed.), Sociolinguistics in the low countries (pp. 333-356). Amsterdam: John Benjamins.

MUYSKEN, R (2001). The origin of creole languages: The perspective of second language learning. In N. Smith & T. Veenstra (Eds.), Creolization and contact (pp. 157-173). Amsterdam: John Benjamins.

NORTIER, J. (2001). Murks en straattaal; Vriendschap en taalgebruik onder jongeren. Amsterdam: Prometheus.

NORTIER, J., & DORLEIJN, M. (2008, this volume). A Moroccan accent in Dutch: A sociocultural style restricted to the Moroccan community? International Journal of Bilingualism, 12(1,2), 125-142.

PONELIS, F. (2005). Nederlands in Afrika: Het Afrikaans. In N. van der Sijs (Ed.), Wereldnederlands. Oude enjonge varieteiten van het Nederlands (pp. 15-30). The Hague: SDU.

QUIST, P. (2008, this volume). Sociolinguistic approaches to multiethnolect: Language variety and stylistic practice. International Journal of Bilingualism, 12(1,2), 43-61.

RAMPTON, B. (1995). Crossing. Language and ethnicity among adolescents. Longman: London.

SNOW, C., Van EEDEN, R. & MUYSKEN, P. (1981). The interactional origins of foreigner talk: Municipal employees and foreign workers. International Journal of the Sociology of Language, 28, 81-92.

TABOURET-KELLER, A. (1997). Language and identity. In F. Coulmas (Ed.), The handbook of sociolinguistics (pp. 315-326). Oxford: Blackwell.

UNSWORTH, S. (2006). Age and input in early child bilingualism: The acquisition of grammatical gender in Dutch. In A. Belikova, L. Meroni, & M. Umeda (Eds.), Galana 2, Proceedings of the Conference on Generative Approaches to Language Acquisition North America 2 (pp.448-458). Somerville, MA: Cascadilla Press.

Leonie Cornips

Meertens Institute, Royal Netherlands Academy of Sciences

(1) Not only determiners, but also demonstrative determiners and personal and relative pronouns agree in gender with the accompanying noun in the singular. These elements are outside the scope of this paper, as are attributive adjectives, which also vary morphologically according to the gender of the noun, under certain conditions (cf. Cornips, Van der Hoek, & Verwer, 2006).

(2) Consider also the results for the acquisition of German nouns found by the Heidelberger Forschungsprojekt "Pidgin-Deutsch'" (1978, p. 14).

(3) See Cornips et al. 2006 for a more extensive discussion of the results in Table 1. Cornips and Hulk (2008) discuss the variability of the grammatical gender feature in Dutch.

(4) ANOVA: Accuracy items x 2 (age) x 4 (group)

Age: F: 18.952; df= 1;p = 0.00 (sign.)

Group: F=26.431, df=3;p=0.00 (sign.)

Age*Group: F=1.719; dr= 3; p =.171 (n.s.)

Post hoc analysis: Group (Bonferroni)

L1 100%--all sign. (p<.05)

L1-75%--all sign. (p<.05)

L2 100%--all sign, except n.s. L2-25% (p = 1.00)

L2 25%--all sign, except n.s. L2-100% (p=1.00)

(5) This is a much higher percentage compared to the Dutch-Moroccan and Dutch-Turkish children: this is due to a later age of onset of first exposure to Dutch, length of exposure to Dutch, and a younger age of testing.

(6) An "allochtone" is an individual if he has at least one parent who is born abroad. He is "autochtone" if both parents are born in the Netherlands. Someone who belongs to the first generation of "allochtone" is born abroad and has at least one parent who is born abroad. Someone who belongs to the second generation of "allochtone" is born in the Netherlands and has at least one parent who is born abroad. Note that in this definition, it may take more generations to become an "autochtone." See for these definitions (2000). Index 10 (november / december). <http.//www.cbs.nl//nl//publicaties/publicaties/algemeen/ index/index1119.pdf>.