Citation Context

... Ye s 10 ;5 –1 2; 11 Tu rk is h /D u tc h co m m u n it ie s si b lin g s N o (C o rn ip s et a l., 2 00 6) Fr en ch /D u tc h ( H u lk , 2 00 7) E xp at ri at e fa m ili es Fa m ily M id d le c la ss Ye s 2; 07 .1 0– 3; 10 .7 M o re H ee rl en d ia le ct / s ta n d ar d ‘O ld ’ t yp e b ili n g u al Fa m ily M id d le c la ss Ye s 2– 7 D u tc h co m m u n it y (C o rn ip s an d H u lk , 2 00 6) Ta rg et lik e pe er -0 05 70 74 5, v er si on 1 - 1 M ar 2 01 1 272 Factors of success and failure II Factors for success–failure: age of onset: a new critical age? At least since the seminal work by Lenneberg (1967), differences between L1 acquisition and (adult) L2 acquisition have been related to the age of onset of acquisition, although this continues to be a much debated issue. The main idea is that there is a sensitive period for grammatical development. Whereas Lenneberg thought that the critical age was around puberty, since then it has generally been assumed that the end of such an optimal period lies around age seven or eight (for a recent overview, see Hyltenstam and Abrahamsson, 2003). The implicit assumption that before age seven the age of onset has no influence on either the development or ...

Citation Context

...e a greater awareness of gender distinctions in general and of neuter gender-marking in particular, and this may help them in their acquisition of gender in standard Dutch. These children had, however, also had a lengthy and intensive exposure to standard Dutch, in contrast to the other groups of bilingual children. Therefore, we have to be cautious in interpreting the decisiveness of the role of the other language here. 4 Summary: yes, there is cross-linguistic influence, but … Cross-linguistic influence is a well-known factor for success in bilingual language acquisition (see, among others, Hulk and Müller, 2000; Müller and Hulk, 2001). Nevertheless, the data from the literature studied here show that for the other language to contribute to the success of the acquisition of grammatical gender on D in bilingual Dutch, the conditions seem to be very narrowly defined: only the bidialectal children with another language which morphosyntactically overlaps with standard Dutch appeared to experience some positive cross-linguistic influence. Clearly, further research about the exact amount of overlap necessary for success in this domain is needed. 5 Summary: the role of the input Among all the different bili...

Citation Context

... Finally, the bidialectal children (of ‘old’ bilingual communities) and the children raised bilingually from birth do best. Given the possible interaction with other factors of success and failure, this hierarchy cannot be used to support the role of the age of onset as a decisive success factor, but it certainly does not constitute counterevidence for such a role either. At the very least, it supports the idea that the role of input is an important factor. Another aspect of the input is considered in the next section. III Factors for success–failure: quantity of the input: threshold effects? Sorace (2005) has argued that interface domains present an inherent problem for acquisition since the acquisition of constructions/phenomena requiring the integration of syntactic knowledge and knowledge from other domains is more complex than those involving syntactic knowledge only. Interestingly, she also discusses the role of the input in relation to such integration problems. She suggests that quantitative and qualitative differences between the input of monolinguals and bilinguals may have ‘destabilizing’ effects and may affect emerging grammars in different ways. Her hypothesis is that quantitative ...

Citation Context

...Haznedar (2003), Unsworth (2005) and others explicitly examines various morphosyntactic phenomena in child second language (L2) acquisition and raises important questions concerning issues such as age of first exposure and the influence of the children’s other language (for an overview of the literature on child L2 acquisition, see Paradis, 2007). In this article, we are interested in the acquisition of grammatical gender in Dutch, which is, as yet, a relatively unexplored area in child language (but see Müller, 2000; Gathercole et al., 2001; Möhring, 2001; Kupisch et al., 2002). According to Franceschina (2005), the child L2 acquisition of grammatical gender shows a development that is very similar to what happens in first language (L1) acquisition, even in cases where one of the two languages does not have gender. Recent studies on the acquisition of grammatical gender in Dutch with learners belonging to different groups of child L2 learners (see below) have, however, shown that the development of grammatical gender in the Dutch of bilingual children does sometimes differ from what is reported for monolingual children. The aim of this article is to examine and compare the following factors of succe...

Citation Context

...of the ethnic minority communities, see Cornips and Hulk (2006). V Factors for success–failure: the other language: conditions for cross-linguistic influence The discussions in the literature on adult L2 acquisition of grammatical gender all address the role of the L1. Globally speaking there are two main perspectives: ● learners can acquire grammatical gender in their L2 regardless of whether this feature is present in their L1 (Bruhn de Gavarito and White, 2000; White et al., 2001); ● the functional feature make-up of the L1 prevents (a subset of) L2 learners from becoming fully targetlike (Hawkins and Franceschina, 2004; Franceschina, 2005). Interestingly, Sabourin (2001) argues that just having gender in the L1 may not be sufficient, and that the (gender system in the) two languages should also be closely related in a way to be specified. Her data on the L2 acquisition of Dutch by three different groups of L1 speakers – German, French and English – shows that the German group performed best and that the French group did better than the English, suggesting that the morphosyntactic overlap between the Dutch and German gender systems has had a positive, cross-linguistic influence. Let us now consider whether t...

Citation Context

...ilingualism, Dutch, quality and quantity of input, sociolinguistic context © 2008 SAGE 10.1177/0267658308090182 Address for correspondence: Leonie Cornips, Meertens Institute, Royal Netherlands Academy of Sciences, Joan Muyskenweg 25, PO Box 94264, 1090 GG Amsterdam, The Netherlands; email: leonie.cornips@meertens.knaw.nl Second Language Research 24,3 (2008); pp. 267–295 pe er -0 05 70 74 5, v er si on 1 - 1 M ar 2 01 1 Author manuscript, published in "Second Language Research 24, 3 (2008) 267-295" DOI : 10.1177/0267658308090182 268 Factors of success and failure I Introduction Recent work by Schwartz (2003), Haznedar (2003), Unsworth (2005) and others explicitly examines various morphosyntactic phenomena in child second language (L2) acquisition and raises important questions concerning issues such as age of first exposure and the influence of the children’s other language (for an overview of the literature on child L2 acquisition, see Paradis, 2007). In this article, we are interested in the acquisition of grammatical gender in Dutch, which is, as yet, a relatively unexplored area in child language (but see Müller, 2000; Gathercole et al., 2001; Möhring, 2001; Kupisch et al., 2002). According t...

Citation Context

...ge seven the age of onset has no influence on either the development or the ultimate attainment of the child’s acquisition has recently been questioned. This issue can be related to the distinction which is made between child L2 acquisition, on the one hand, and 2L1 acquisition, on the other. The age of onset for child L2 acquisition is generally assumed to be between age four and seven (Schwartz, 2003; Unsworth, 2005), whereas for 2L1 acquisition the age of onset is ideally the moment of birth (see De Houwer, 1990). The period between birth and age four is as yet an unexplored area, although Meisel (2007) considers this period to belong to 2L1 acquisition. He suggests that at around age three to four, changes in the acquisition patterns are brought about by neurological maturation, resulting in modification of the human ‘language making capacity’. In other words, according to Meisel, age three to four would constitute another critical age for language acquisition. Let us now consider what role this factor has been claimed to play in the success–failure of each of the four groups of bilingual children, i.e. children whose first exposure to Dutch (1) widely varies, but is after age four and (2) ...

Citation Context

...,3 (2008); pp. 267–295 pe er -0 05 70 74 5, v er si on 1 - 1 M ar 2 01 1 Author manuscript, published in "Second Language Research 24, 3 (2008) 267-295" DOI : 10.1177/0267658308090182 268 Factors of success and failure I Introduction Recent work by Schwartz (2003), Haznedar (2003), Unsworth (2005) and others explicitly examines various morphosyntactic phenomena in child second language (L2) acquisition and raises important questions concerning issues such as age of first exposure and the influence of the children’s other language (for an overview of the literature on child L2 acquisition, see Paradis, 2007). In this article, we are interested in the acquisition of grammatical gender in Dutch, which is, as yet, a relatively unexplored area in child language (but see Müller, 2000; Gathercole et al., 2001; Möhring, 2001; Kupisch et al., 2002). According to Franceschina (2005), the child L2 acquisition of grammatical gender shows a development that is very similar to what happens in first language (L1) acquisition, even in cases where one of the two languages does not have gender. Recent studies on the acquisition of grammatical gender in Dutch with learners belonging to different groups of child L2...

Citation Context

...tant consequences for the quality of the input to which these immigrant children are exposed. However, children of Turkish and Moroccan descent differ from children from Surinamese descent in this respect. The command of Dutch by the first-generation Surinamese immigrants is standard-like due to the long period of contact with Dutch since colonial times. Furthermore, Surinamese households in which Sranan is spoken are generally oriented towards Dutch and none of the Surinamese participants claim to speak Sranan better than Dutch (Jongenburger and Aarssen, 2001). It is also reported elsewhere (Appel and Schoonen, 2005) that children in Surinamese families often acquire Dutch as their first language. These differences may account for the fact that the youngest 3-year-old Surinamese children in Hulk and Cornips’ study (2006b) performed much better and revealed an increasingly correct use of het between three and five years old, in contrast to the Moroccan and Ghanaian children of the same age. Interestingly, sociolinguistic research has shown not only that adult L2 speakers of Dutch widely overgeneralize de with all nouns (Muysken, 1984), but also that the use of de in some adolescent groups of these ethnic c...

Citation Context

...actors for success–failure: the other language: conditions for cross-linguistic influence The discussions in the literature on adult L2 acquisition of grammatical gender all address the role of the L1. Globally speaking there are two main perspectives: ● learners can acquire grammatical gender in their L2 regardless of whether this feature is present in their L1 (Bruhn de Gavarito and White, 2000; White et al., 2001); ● the functional feature make-up of the L1 prevents (a subset of) L2 learners from becoming fully targetlike (Hawkins and Franceschina, 2004; Franceschina, 2005). Interestingly, Sabourin (2001) argues that just having gender in the L1 may not be sufficient, and that the (gender system in the) two languages should also be closely related in a way to be specified. Her data on the L2 acquisition of Dutch by three different groups of L1 speakers – German, French and English – shows that the German group performed best and that the French group did better than the English, suggesting that the morphosyntactic overlap between the Dutch and German gender systems has had a positive, cross-linguistic influence. Let us now consider whether the L1 plays a role in the development of the bilingua...

Citation Context

... Hulk 275 (24%, 68/288), albeit to a lesser extent than the use of de with neuter nouns (54%, 141/262). Monolinguals, in contrast, only revealed a 4% (2/48) overgeneralization of het to common nouns (Cornips et al., 2006). The 24% overgeneralization of het by the bilinguals seems to suggest that they have some active knowledge of het. It raises the intriguing question as to whether the initial and persistent overgeneralization of de could be, at this older age at least, the effect of a speech production problem, rather than a lack of knowledge regarding the gender of the definite determiner.4 Brouwer et al. (2008) try to shed light on this question by using an experiment designed to ‘tap the knowledge’ of monolingual (n 18) and bilingual children (n 24) between 11- and 13-years-old. Interestingly, the results show that the bilingual children did not really misrepresent the gender specification of de as compatible with both common and neuter nouns. Nevertheless, it is certainly not the case that the bilingual children possessed the right knowledge (yet) about gender specification in Dutch either: the accuracy rates were too low (about 40%) and significantly lower than the accuracy rates of the monolin...

Citation Context

...ker. pe er -0 05 70 74 5, v er si on 1 - 1 M ar 2 01 1 276 Factors of success and failure children as child L2 learners.5 In Hulk and Cornips (2006a; 2006b) a slightly different perspective is chosen. There, we provided substantial evidence for the claim that, nowadays, in Moroccan and Turkish families Dutch is always spoken at home (see Jongenburger and Aarssen, 2001). For these families, the heritage language is used more with mothers than with fathers, and Dutch is used more between siblings (for a more elaborate discussion of the languages used by these children in different contexts, see Cornips and Hulk, 2006a). Taken together, these results suggest that ‘age of onset’ does not appear to be a decisive factor here. 3 Children growing up bilingually from birth in expatriate families, in a ‘one parent – one language’ context Hulk (2007) considers longitudinal, spontaneous production data from two French–Dutch girls, Anouk and Annick, who were raised bilingually from birth (from the Amsterdam-corpus of Hulk and Van der Linden). All the grammatical gender errors of Anouk and Annick between ages 3;1.4 and 3;10.7 consisted of the use of (the common) de with a neuter noun instead of het. Anouk was conside...

Citation Context

...antity of input, sociolinguistic context © 2008 SAGE 10.1177/0267658308090182 Address for correspondence: Leonie Cornips, Meertens Institute, Royal Netherlands Academy of Sciences, Joan Muyskenweg 25, PO Box 94264, 1090 GG Amsterdam, The Netherlands; email: leonie.cornips@meertens.knaw.nl Second Language Research 24,3 (2008); pp. 267–295 pe er -0 05 70 74 5, v er si on 1 - 1 M ar 2 01 1 Author manuscript, published in "Second Language Research 24, 3 (2008) 267-295" DOI : 10.1177/0267658308090182 268 Factors of success and failure I Introduction Recent work by Schwartz (2003), Haznedar (2003), Unsworth (2005) and others explicitly examines various morphosyntactic phenomena in child second language (L2) acquisition and raises important questions concerning issues such as age of first exposure and the influence of the children’s other language (for an overview of the literature on child L2 acquisition, see Paradis, 2007). In this article, we are interested in the acquisition of grammatical gender in Dutch, which is, as yet, a relatively unexplored area in child language (but see Müller, 2000; Gathercole et al., 2001; Möhring, 2001; Kupisch et al., 2002). According to Franceschina (2005), the child L...

Citation Context

...ctors of success and failure I Introduction Recent work by Schwartz (2003), Haznedar (2003), Unsworth (2005) and others explicitly examines various morphosyntactic phenomena in child second language (L2) acquisition and raises important questions concerning issues such as age of first exposure and the influence of the children’s other language (for an overview of the literature on child L2 acquisition, see Paradis, 2007). In this article, we are interested in the acquisition of grammatical gender in Dutch, which is, as yet, a relatively unexplored area in child language (but see Müller, 2000; Gathercole et al., 2001; Möhring, 2001; Kupisch et al., 2002). According to Franceschina (2005), the child L2 acquisition of grammatical gender shows a development that is very similar to what happens in first language (L1) acquisition, even in cases where one of the two languages does not have gender. Recent studies on the acquisition of grammatical gender in Dutch with learners belonging to different groups of child L2 learners (see below) have, however, shown that the development of grammatical gender in the Dutch of bilingual children does sometimes differ from what is reported for monolingual children. The aim ...

Citation Context

...ors of success and failure II Factors for success–failure: age of onset: a new critical age? At least since the seminal work by Lenneberg (1967), differences between L1 acquisition and (adult) L2 acquisition have been related to the age of onset of acquisition, although this continues to be a much debated issue. The main idea is that there is a sensitive period for grammatical development. Whereas Lenneberg thought that the critical age was around puberty, since then it has generally been assumed that the end of such an optimal period lies around age seven or eight (for a recent overview, see Hyltenstam and Abrahamsson, 2003). The implicit assumption that before age seven the age of onset has no influence on either the development or the ultimate attainment of the child’s acquisition has recently been questioned. This issue can be related to the distinction which is made between child L2 acquisition, on the one hand, and 2L1 acquisition, on the other. The age of onset for child L2 acquisition is generally assumed to be between age four and seven (Schwartz, 2003; Unsworth, 2005), whereas for 2L1 acquisition the age of onset is ideally the moment of birth (see De Houwer, 1990). The period between birth and age four ...

Citation Context

...ngenburger and Aarssen, 2001). It is also reported elsewhere (Appel and Schoonen, 2005) that children in Surinamese families often acquire Dutch as their first language. These differences may account for the fact that the youngest 3-year-old Surinamese children in Hulk and Cornips’ study (2006b) performed much better and revealed an increasingly correct use of het between three and five years old, in contrast to the Moroccan and Ghanaian children of the same age. Interestingly, sociolinguistic research has shown not only that adult L2 speakers of Dutch widely overgeneralize de with all nouns (Muysken, 1984), but also that the use of de in some adolescent groups of these ethnic communities functions as a kind of group ‘identifier’ (see Cornips, 2008). 3 Bidialectal children Contrary to the bilingual children in communities of the ‘new’ type, the (grand)parents of the bidialectal children are all bidialectal themselves and their Dutch does not differ from standard Dutch with respect to grammatical gender. Thus, the quality of the standard Dutch input to which these children are exposed is excellent. 4 Summary Summarizing, the quality of the input appears to be a significant factor in communities o...

Citation Context

...uisition of the definite determiner result from the frequency of de in the Dutch input, because there are many more de-words than het-words. Although this may play an important role in explaining the delay in the development of (most) bilingual children from the literature studied, it cannot be the whole story. On the one hand, it is especially unclear why most of the older bilingual children seem to fossilize in a non-target stage regarding the acquisition of Dutch grammatical gender, when this is not a general characteristic of their performance in other phenomena of Dutch morphosyntax (see Hulk and Cornips, 2005). On the other hand, there are clear differences in the rate of (ultimate) success of the different groups of bilingual children. If frequency were the crucial factor, we would not expect the different groups of bilinguals to differ in the extent of the delay they experience compared to their monolingual peers. Therefore, we would like to offer some speculative suggestions related to the acquisition of grammatical features in more theoretical terms in the next section. 7 Here we only consider grammatical gender in Dutch definite determiners; but see Blom et al. (this issue) for grammatical gen...

Citation Context

...ese communities: Blom et al. adopt age four as the age of onset, and therefore they considered these 4 Such a mapping problem could be similar to Lardiere’s (2005) findings concerning the L2 acquisition of English verbal morphology by a fossilized Chinese speaker. pe er -0 05 70 74 5, v er si on 1 - 1 M ar 2 01 1 276 Factors of success and failure children as child L2 learners.5 In Hulk and Cornips (2006a; 2006b) a slightly different perspective is chosen. There, we provided substantial evidence for the claim that, nowadays, in Moroccan and Turkish families Dutch is always spoken at home (see Jongenburger and Aarssen, 2001). For these families, the heritage language is used more with mothers than with fathers, and Dutch is used more between siblings (for a more elaborate discussion of the languages used by these children in different contexts, see Cornips and Hulk, 2006a). Taken together, these results suggest that ‘age of onset’ does not appear to be a decisive factor here. 3 Children growing up bilingually from birth in expatriate families, in a ‘one parent – one language’ context Hulk (2007) considers longitudinal, spontaneous production data from two French–Dutch girls, Anouk and Annick, who were raised bili...

Citation Context

...ure I Introduction Recent work by Schwartz (2003), Haznedar (2003), Unsworth (2005) and others explicitly examines various morphosyntactic phenomena in child second language (L2) acquisition and raises important questions concerning issues such as age of first exposure and the influence of the children’s other language (for an overview of the literature on child L2 acquisition, see Paradis, 2007). In this article, we are interested in the acquisition of grammatical gender in Dutch, which is, as yet, a relatively unexplored area in child language (but see Müller, 2000; Gathercole et al., 2001; Möhring, 2001; Kupisch et al., 2002). According to Franceschina (2005), the child L2 acquisition of grammatical gender shows a development that is very similar to what happens in first language (L1) acquisition, even in cases where one of the two languages does not have gender. Recent studies on the acquisition of grammatical gender in Dutch with learners belonging to different groups of child L2 learners (see below) have, however, shown that the development of grammatical gender in the Dutch of bilingual children does sometimes differ from what is reported for monolingual children. The aim of this article...

Citation Context

...8090182 268 Factors of success and failure I Introduction Recent work by Schwartz (2003), Haznedar (2003), Unsworth (2005) and others explicitly examines various morphosyntactic phenomena in child second language (L2) acquisition and raises important questions concerning issues such as age of first exposure and the influence of the children’s other language (for an overview of the literature on child L2 acquisition, see Paradis, 2007). In this article, we are interested in the acquisition of grammatical gender in Dutch, which is, as yet, a relatively unexplored area in child language (but see Müller, 2000; Gathercole et al., 2001; Möhring, 2001; Kupisch et al., 2002). According to Franceschina (2005), the child L2 acquisition of grammatical gender shows a development that is very similar to what happens in first language (L1) acquisition, even in cases where one of the two languages does not have gender. Recent studies on the acquisition of grammatical gender in Dutch with learners belonging to different groups of child L2 learners (see below) have, however, shown that the development of grammatical gender in the Dutch of bilingual children does sometimes differ from what is reported for monol...

Citation Context

... to be cautious in interpreting these results, but the tendency was clearly different from the one found regarding the other bilingual children (for details, see Cornips and Hulk, 2006). The bidialectal children in question were raised in a sociolinguistic context that is both different from and similar to the bilingual children growing up in bilingual communities of the ‘new’ type. It is different because the parents of the bidialectal children are not adult L2 speakers of Dutch but are balanced bilinguals themselves, speaking both the local dialect and standard Dutch as native speakers (see Cornips, 1998). It is similar, however, in that the bidialectal children have also been raised bilingually, both in the family and community domains. There is no ‘one parent – one language’ setting per se but language choice patterns depend on, for example, particular interlocutors. Irrespective of any differences between these bidialectal children in this domain, the claim that they were exposed to standard Dutch before they went to school is uncontroversial. 5 Summary Notwithstanding possible interaction with other factors, there are quantitative differences in the correct use of het between the different...

Citation Context

...h, quality and quantity of input, sociolinguistic context © 2008 SAGE 10.1177/0267658308090182 Address for correspondence: Leonie Cornips, Meertens Institute, Royal Netherlands Academy of Sciences, Joan Muyskenweg 25, PO Box 94264, 1090 GG Amsterdam, The Netherlands; email: leonie.cornips@meertens.knaw.nl Second Language Research 24,3 (2008); pp. 267–295 pe er -0 05 70 74 5, v er si on 1 - 1 M ar 2 01 1 Author manuscript, published in "Second Language Research 24, 3 (2008) 267-295" DOI : 10.1177/0267658308090182 268 Factors of success and failure I Introduction Recent work by Schwartz (2003), Haznedar (2003), Unsworth (2005) and others explicitly examines various morphosyntactic phenomena in child second language (L2) acquisition and raises important questions concerning issues such as age of first exposure and the influence of the children’s other language (for an overview of the literature on child L2 acquisition, see Paradis, 2007). In this article, we are interested in the acquisition of grammatical gender in Dutch, which is, as yet, a relatively unexplored area in child language (but see Müller, 2000; Gathercole et al., 2001; Möhring, 2001; Kupisch et al., 2002). According to Franceschina (2...

Citation Context

...ion Recent work by Schwartz (2003), Haznedar (2003), Unsworth (2005) and others explicitly examines various morphosyntactic phenomena in child second language (L2) acquisition and raises important questions concerning issues such as age of first exposure and the influence of the children’s other language (for an overview of the literature on child L2 acquisition, see Paradis, 2007). In this article, we are interested in the acquisition of grammatical gender in Dutch, which is, as yet, a relatively unexplored area in child language (but see Müller, 2000; Gathercole et al., 2001; Möhring, 2001; Kupisch et al., 2002). According to Franceschina (2005), the child L2 acquisition of grammatical gender shows a development that is very similar to what happens in first language (L1) acquisition, even in cases where one of the two languages does not have gender. Recent studies on the acquisition of grammatical gender in Dutch with learners belonging to different groups of child L2 learners (see below) have, however, shown that the development of grammatical gender in the Dutch of bilingual children does sometimes differ from what is reported for monolingual children. The aim of this article is to examine and comp...

Citation Context

...es with respect to the correct use of het; for details of a more sociolinguistic explanation of the (quality of the) input in some of the ethnic minority communities, see Cornips and Hulk (2006). V Factors for success–failure: the other language: conditions for cross-linguistic influence The discussions in the literature on adult L2 acquisition of grammatical gender all address the role of the L1. Globally speaking there are two main perspectives: ● learners can acquire grammatical gender in their L2 regardless of whether this feature is present in their L1 (Bruhn de Gavarito and White, 2000; White et al., 2001); ● the functional feature make-up of the L1 prevents (a subset of) L2 learners from becoming fully targetlike (Hawkins and Franceschina, 2004; Franceschina, 2005). Interestingly, Sabourin (2001) argues that just having gender in the L1 may not be sufficient, and that the (gender system in the) two languages should also be closely related in a way to be specified. Her data on the L2 acquisition of Dutch by three different groups of L1 speakers – German, French and English – shows that the German group performed best and that the French group did better than the English, suggesting that the mor...

Citation Context

...groups of bilingual children, i.e. children whose first exposure to Dutch (1) widely varies, but is after age four and (2) children whose first exposure is at birth, and children growing up in bilingual families and surrounding bilingual communities of the (3) ‘new’ and (4) ‘old’ type. In what follows, we first very briefly summarize in each subsection in what respect the bilingual children studied are found to be successful or not in the acquisition of grammatical gender on the determiner category (D). 1 Bilingual children from expatriate families with Dutch input after age four Two studies (Hulk, 2007; Unsworth, 2007) examine bilingual children growing up in expatriate families in which most of the children pe er -0 05 70 74 5, v er si on 1 - 1 M ar 2 01 1 Leonie Cornips and Aafke Hulk 273 were exposed to Dutch between age four and seven at school. These children would generally be characterized as child L2 learners of Dutch. Importantly, in an elicited production task the bilingual children in both studies showed a similar development to monolingual children in that they initially produced de with both neuter and common nouns (for more detailed information about the data collection method...

Citation Context

...nouns but they infrequently (incorrectly) produce het with common nouns. The overgeneralization in one direction and the late age at which the monolingual children are targetlike makes Dutch different from other languages – such as German, French and Spanish – where the L1 acquisition of grammatical gender is not a late phenomenon and it does not show similar unidirectional overgeneralization properties (Franceschina, 2005). German is particularly relevant in this respect since it also has neuter gender determiners, and these have not been reported to present specific problems in acquisition (Mills, 1986). The question we are interested in is whether bilingual children show the same pattern of development in the acquisition of grammatical gender in Dutch as monolingual children. To be more specific, we seek to determine whether and for what reasons the neuter gender definite determiner het has been found to be more problematic for bilingual children as it is for monolingual children. Table 1 presents a schematic overview of the different types of bilingual children, revealing their increasing rate of success in the acquisition of gender of the Dutch definite determiner, as reported in the lite...

Citation Context

... as het appel (the-neuter applecommon). Their monolingual peer controls (age-matched) were also found to make this error but only in 4% (2/48) of cases. This is a significant difference. Interestingly, Unsworth (2007) mentions that six of her bilingual (child L2) children produce more than 20% of common words with het.13 Let us go back to the hypothesized development in the acquisition of the gender feature specification and speculate on the question of how 11 For some suggestions about the grammatical feature count being involved in the (acquisition of the) Dutch definite determiner, see Roodenburg and Hulk, 2007. 12 In that article, we suggested that it could be the case that these children have acquired the correct specific neuter value of the gender feature, but that they somehow have problems with its correct morphological spell-out as het on the definite determiner, using the default form de instead. Such an explanation is compatible with the slightly different line we suggest here. 13 It is not entirely clear in her work at what stage of acquisition these bilingual children are in. pe er -0 05 70 74 5, v er si on 1 - 1 M ar 2 01 1 288 Factors of success and failure children progress beyond this ...

Citation Context

...tion is compatible with the slightly different line we suggest here. 13 It is not entirely clear in her work at what stage of acquisition these bilingual children are in. pe er -0 05 70 74 5, v er si on 1 - 1 M ar 2 01 1 288 Factors of success and failure children progress beyond this unstable stage. Although we did not discuss it in any detail and there are not many data on the overgeneralization of het to common nouns for monolingual Dutch children, we assume that the monolingual children have left this stage around age six or seven, when they are reported to behave in a targetlike fashion (Bol and Kuiken, 1988; Van Kampen and Wijnen, 2000: 275; Van der Velde, 2003; 2004; but see Blom et al., this issue). It is plausible to assume – following the suggestion of Hawkins and Franceschina (2004) and discussions in the literature – that this happens after a certain threshold in the input has been reached. Interestingly, it could also be the case that the specification of the value of the gender feature on D in Dutch and its independent status as a grammatical feature is due to a specific characteristic of the input, as mentioned in the literature on Dutch monolingual children, namely: the awareness of th...

Citation Context

...f gender distinctions in general and of neuter gender-marking in particular, and this may help them in their acquisition of gender in standard Dutch. These children had, however, also had a lengthy and intensive exposure to standard Dutch, in contrast to the other groups of bilingual children. Therefore, we have to be cautious in interpreting the decisiveness of the role of the other language here. 4 Summary: yes, there is cross-linguistic influence, but … Cross-linguistic influence is a well-known factor for success in bilingual language acquisition (see, among others, Hulk and Müller, 2000; Müller and Hulk, 2001). Nevertheless, the data from the literature studied here show that for the other language to contribute to the success of the acquisition of grammatical gender on D in bilingual Dutch, the conditions seem to be very narrowly defined: only the bidialectal children with another language which morphosyntactically overlaps with standard Dutch appeared to experience some positive cross-linguistic influence. Clearly, further research about the exact amount of overlap necessary for success in this domain is needed. 5 Summary: the role of the input Among all the different bilingual child learners, th...