ISB 6. The 6th International Symposium on Bilingualism. June 2, 2007 - University of Hamburg DIMINUTIVES AND GENDER AWARENESS IN BILINGUAL ACQUISITION Leonie Cornips Meertens Institute Royal Netherlands Academy leonie.cornips@meertens.knaw.nl 1. Aafke Hulk University of Amsterdam aafke.hulk@uva.nl Susanne Brouwer Max Planck Institute Nijmegen susanne.brouwer@mpi.nl Problems with the acquisition of neuter definite determiner het Table 1: Dutch morphology of definite determiners Definite determiners Gender of noun Singular Plural common de de neuter de het Production data bilingual children (Blom et al. 2006, this talk); (Hulk & Cornips 2006a,b); (Cornips & Hulk 2006); (Cornips et al. 2006): Age of ultimate attainment acquisition het unclear/after age 13; Quantity and quality of input seem to be crucial. Production data monolingual children: Van de Velde (2004), Policenska (p.c., control groups), Hulk & Cornips (2006a,b), Cornips et al. (2006), Blom et al. (2006): Initially almost only de, later massive overgeneralization of de both with neuter and common nouns; Het comes in slowly and late: no targetlike production before age 6 at the earliest, errors persist until a more advanced age. This is different from what happens in the L1 acquisition process of other languages with gender morphology on the definite determiner. Especially, the difference with German is striking: in German monolingual children produce articles with the correct gender around age 3 (Mills 1986). 2. A linguistic analysis of the acquisition of grammatical gender Hawkins & Franceschina (2004): D has an un-interpretable, grammatical gender feature which has to be erased/checked by the lexical gender feature of the noun: Initially children do not have this un-interpretable gender feature on D in their grammar. Their (non targetlike) use of different forms of the determiner is not based on grammatical concord with the noun, but on other strategies, related to the phonological cues; A point in the development comes when the grammatical u-gender feature on D is triggered – thereafter their use is targetlike. 1 Possible triggers according to Hawkins & Franceschina (2004): Paradigmatic link between definite and indefinite articles; Input above a certain threshold. What about Dutch – what could be the trigger? Not much evidence in the input: No paradigmatic link between definite and indefinite article wrt grammatical gender, no gender on indefinite articles, no gender on plural definite articles but gender on single definite articles and demonstratives; Other gender marked elements: confusing. Gender marking on pronominal elements are partially based on other features [± count, ± animate]; Frequency differences between neuter (25%) and common (75%) nouns. HOWEVER One salient trigger for grammatical gender: the diminutive suffix In Dutch any noun, regardless of its lexical gender, “becomes” neuter in taking the diminutive suffix -(t)je: (1) a. b. c. (2) a. b. c. de tafel the (COMMON) table het tafeltje the (NEUTER) table + DIM de tafel-tje-s the (PLUR) table (DIM) PLUR het boek the (NEUTER) book het boekje the (NEUTER) book + DIM de boek-je-s the (PLUR) book (DIM) PLUR 3. Diminutives and the acquisition of grammatical gender 3.1 Experimental production data Van de Velde (2003, 2004): monolingual children In an experiment eliciting the definite determiner, Dutch children age 3 make significantly less errors with diminutives than with other neuter nouns; for 4 and 6 year olds, this difference is clearly present but not statistically significant. 2 Van Ginkel (2006:49): monolingual production: similar results Table 2: The determiner production by monolingual children, age 6;0 – 7;6, for neuter nouns and diminutives in an elicitation task (picture description task based on Blom 2006), target in grey PRODUCTION MONOLINGUALS n=8 age 6;0-7;6 neuter nouns diminutives de 39 40.6 % 19 14.8 % het 47 49 % 103 80.5 % een ‘a’ 10 10.4 % 6 4.7 % Ø 0 0 % 0 0 % total 96 100 % 128 100 % Table 3: The determiner production by monolingual children, age 10;3 – 11;5, for neuter nouns and diminutives in an elicitation task (picture description task based on Blom 2006), target in grey PRODUCTION MONOLINGUALS n=8 age 10;3-11;5 neuter nouns diminutives de 15 15.6 % 6 4.7 % het 76 79.2 % 117 91.4 % een ‘a’ 4 4.2 % 4 3.1 % Ø 1 1 % 0 0 % total 96 100 % 128 100 % HOWEVER, Van Ginkel (2006:49): bilingual – Turkish-Dutch production results Table 4: The determiner production by Turkish-Dutch children, age 6;0 – 7;6, for neuter nouns and diminutives in an elicitation task (picture description task based on Blom 2006), target in grey PRODUCTION BILINGUAL – TURKISH-DUTCH n=10 age 6;0-7;6 neuter nouns diminutives de 63 52.5 % 75 46.9 % het 55 45.8 % 77 48.1 % een ‘a’ 1 0.8 % 3 1.9 % Ø 1 0.8 % 4 2.5 % missing 0 0 % 1 0.6 % total 120 99.9 % 160 100 % 3 Table 5: The determiner production by Turkish-Dutch children, age 10;0-11;10, for neuter nouns and diminutives in an elicitation task (picture description task based on Blom 2006), target in grey PRODUCTION BILINGUAL – TURKISH-DUTCH n=11 age 10;0-11;10 neuter nouns diminutives de 53 40.2% 58 33% het 77 58.3% 114 64.8% een ‘a’ 2 1.5% 2 1.1% Ø 0 0% 1 0.6% missing 0 0% 1 0.6% total 132 100% 176 100.1% The production data of the bilingual children in Table 4 & 5 differ from the monolingual results in Table 2 & 3. QUESTIONS Is there awareness of grammatical gender at all? More specifically, do the bilingual children recognize the ‘diminutive’ as trigger for grammatical gender? 4. Knowledge awareness experiments 4.1 Experimental design (see Brouwer 2006) An experimental design is set up to “tap the knowledge” in monolingual and bilingual children with respect to their gender system of definite determiners of diminutives. Spoken items are paired with an identical image on the screen of a laptop monitor. The subjects have to decide as soon as possible whether the spoken item is correct or incorrect in Dutch. They responded by pressing buttons on a button box. One button had a happy smiley face that the subjects pressed to indicate they believed the sentence to be correct Dutch, and another button had a sad smiley face that the subjects pressed if they believed the sentence to be incorrect Dutch. After the practice session, two lists with each 32 spoken experimental items, are offered (see table 6). Every experimental item consists of correct: Ik zie het hondje ‘I see theN dog-DIMN; and incorrect: Ik zie *de hondje ‘I see theC dog- DIMN. In other words, each presented diminutive is paired with a correct and an incorrect definite determiner. Moreover, a distinction is made between 8 common and 8 neuter nouns. 4 Table 6: Overview of all presented items Grammatical form Practice items 3 Experimental items 16 common noun 8 e.g. Ik zie het lepeltje ‘I see the spoonC-DIM’ neuter noun 8 e.g. Ik zie het paardje ‘I see the horseN-DIM’ Fillers 4 Ungrammatical form *3 * 16 *8 * Ik zie de lepeltje * ‘I see the spoonC-DIM’ *8 * Ik zie de paardje * ‘I see the horseN-DIM’ *4 Total 6 32 16 16 8 Examples: Ik zie het hondje; *Ik zie de hondje ‘I see the dog-common-DIM’ 4.2 Ik zie het schaapje; *Ik zie de schaapje ‘I see the sheep-neuter-DIM’ Results diminutives (Van Ginkel 2006) Table 7: Accuracy rates (% correct) of *DE+DIM and HET+DIM by monolingual children in two age groups, target in grey input L1 L1 n=8 n=8 age 6;0-7;6 10;3-11;5 correct incorrect correct incorrect judgment judgment judgment judgment correct 116 12 123 5 HET+DIM 90.6% 9.4% 96.1% 3.9% incorrect 56 72 97 31 DE+DIM 43.7% 56.3% 75.8% 24.2% Table 8: Accuracy rates (% correct) of DE+DIM and HET+DIM by bilingual TurkishDutch children in two age groups, target in grey input bilingual bilingual n=10 n=11 age 6;0-7;6 10;3-11;5 correct incorrect correct incorrect judgment judgment judgment judgment correct 145 15 153 23 HET+DIM 90.6% 9.4% 86.9% 13.1% incorrect 26 34 96 80 DE+DIM 16.2% 83.8% 54.5% 45.5% 5 • Bilinguals of both age groups reveal much lower accuracy rates for incorrect *DE+DIM than monolinguals; high accuracy rates for correct HET+DIM may be due to ‘yes’-strategy (cf Brouwer et al.); Question: is diminutive a trigger? However: • Bilinguals reveal development between age groups 6-7;6 and 10;3-11;5. The accuracy-rates for incorrect *DE+DIM increase from 16.2% to 54.5%, respectively. 5. Discussion bilingual children age 6;0 – 7;6 do not produce more correct het with diminutives (45.8%) than non-diminutives (48.1%; see Table 4); are not ‘aware’ of neuter gender with respect to diminutives (accuracy rate of incorrect *de + diminutives is 16.2 %, see Table 8). • • - bilingual children age 10;3 -11;5 produce some more correct het with diminutives (64.8%) than with non-diminutives (58.3%; see Table 5); show some ‘awareness’ of neuter gender with respect to diminutives (accuracy rate of incorrect *de + diminutives is 54.5 %, see Table 8). Although 54.5% is around chance level, it’s a much higher percentage than in the younger bilingual group (16.2%). JUST DELAY? However, with respect to the acquisition of neuter gender in Dutch, even 11-13 years old bilinguals are still non-targetlike and significantly different from monolinguals of the same age both in production (see Cornips et al. 2006) and ‘awareness’ (see Brouwer et al. 2006). QUESTION • Do bilinguals need another trigger than diminutives? OR • Do they fossilize? 6 References Berkum, J.J.A. van (1996). The psycholinguistics of grammatical gender. Studies in language comprehension and production. Nijmegen: Nijmegen University Press. Blom, E. (2005). Dutch adjectival inflection: L1, child L2 and adult L2 learners. Talk presented at the Radboud University. Blom, E., D. Polisenská & F. Weerman. (2006). Effects of age in the acquisition of gender: a three-way distinction between child L1, child L2 and adult L2 acquisition. Talk presented at Amsterdam Gender Colloquium, Vrije Universiteit, 16 September 2006. Bol, G.W., and Kuiken, F. (1988). Grammaticale analyse van taalontwikkelingsstoornissen. Phd. Dissertation. University of Amsterdam. Brouwer, S. (2006). An experimental study about possible dissociation between representation and processing definite determiners of L1 and child L2 Dutch in different sociolinguistic contexts. Trainee report Meertens Institute/University of Utrecht. Brouwer, S., L. Cornips & A. Hulk. (2007). Misrepresentation of Dutch neuter gender in older bilingual children? ms. University of Utrecht, Meertens Instituut & University of Amsterdam. Cornips, L. & A. Hulk. (2006). Bilingual and bidialectal language development: grammatical gender in Dutch. Lefebvre, C., L. White & C. Jourdan (eds).The Montreal Dialogues. Amsterdam: John Benjamins. Cornips, L., M. van der Hoek & R. Verwer (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. Linguistics-in-The Netherlands, 2006. B. Los & J. van de Weijer (eds): 40-51. Gillis, S., Schaerlaekens, A.M. (eds.). Kindertaalverwerving. Een handboek voor het Nederlands. Groningen: Martinus Nijhoff. Ginkel, B. van (2006). “De verwerving van lexicaal en morfologisch bepaald geslacht. Een onderzoek naar tweede taalverwerving bij eentalige Nederlandse kinderen en tweetalige Turks-Nederlandse kinderen”. Master thesis Meertens Instituut/University of Utrecht. De Houwer, A. and S. Gillis. 1998. The acquisition of Dutch. Amsterdam: Benjamins. Hawkins, Roger & Florencia Franceschina. (2004). “Explaining the acquisition and nonacquisition of determiner-noun gender concord in French and Spanish”. The acquisition of French in Different Contexts ed. by P. Prévost & J. Paradis, 175–205. Amsterdam & Philadelphia: John Benjamins. Hulk, A. to appear. Deviance in early child bilingualism. LSRL Hulk, A. & L. Cornips (2006a) ‘Neuter gender and interface vulnerability in child L2/2L1 Dutch. In: Sharon, U. et. al. (eds.), Paths of Development in L1 and L2 acquisition: In honor of Bonnie D. Schwartz. 2006. p. 107–134.’ Hulk, A. & L. Cornips. (2006b). Between 2L1 - and child L2 acquisition: an experimental study of bilingual Dutch. Interfaces in Multilingualism: Acquisition and representation. Hamburg Studies on Multilingualism Vol 4. Amsterdam: John Benjamins. C. Lleo (ed), pp. 115-137. Hulk, A. C. J. and N. Müller 2000. Bilingual first language acquisition at the interface between syntax and pragmatics'. Bilingualism: Language and Cognition 3 (3): 227-244. Meisel, J. 1989. "Early differentiation of language in bilingual children." In 7 Bilingualism across a lifespan: Aspects of acquisition, maturity and loss, K. Hyltenstam and L. Obler (eds.), 13-40. Cambridge: Cambridge University Press. Mills, A.E. (1986). Mills, A. (1986) The acquisition of gender. A Study of English and German. Berlijn-Heidelberg: Springer-Verlag,. Müller, N. & A. Hulk 2001. "Cross-linguistic influence in bilingual first language acquisition: Italian and French as recipient languages." Bilingualism: Language and Cognition 4 (1): 1-21. Sabourin, L.L., and Haverkort, M. (2003). Neural substrates of representation and processing of a second language. In: R. van Hout, A. Hulk, F. Kuiken, R. Towell (eds.). The lexicon-syntax interface in second language acquisition. Amsterdam: John Benjamins. Sabourin, L.L. (to appear). Grammatical gender agreement in L2 processing. Research Proposal. University of Groningen. Unsworth, S. (2006). "On the acquisition of noun gender in 2L1A/child L2 Dutch." Talk presented at the Linguistic in the Netherlands day, 4 February. Unsworth, S. (in press). Age and input in early child bilingualism: The acquisition of grammatical gender in Dutch.’ In A. Belikova, L. Meroni and M. Umeda (eds.). Galana 2Proceedings of the Conference on generative Aprroaches to langzame Acquisition North America 2. Somerville, MA: Cascadilla Press. Velde, M. van der (2004). L'acquisition des determinants en L1: une etude comparative entre le Francais et le Néerlandais. Acquisition et interaction en langue etrangere, 21, 9 - 46. Velde, M. van der (2002). Determiner acquisition in Dutch and French. TIN-dag (Linguistics in the Netherlands). University Utrecht. 8