Complexity, frequency and structural similarity in bilingual Norwegian-English acquisition and attrition
In this paper we compare findings from Norwegian-American heritage speakers to data collected from Norwegian-English bilingual children as well as Norwegian monolingual children. The focus is on two aspects of the Norwegian DP: word order in possessive constructions and modified definites requiring double definiteness. Our findings are discussed in relation to factors such as complexity, frequency and structural similarity in the acquisition and attrition processes. The data from the heritage speakers indicate that, unlike the children, they are more influenced by the frequency than the complexity of a structure. Furthermore, while the bilingual children’s production is affected by structural similarity with English, the heritage speakers seem to be more sensitive to the structural difference between the two languages.
Norwegian DPs are relatively complex. For example, possessives may be pre- or postnominal (1a, b), and definiteness is marked twice in modified definite noun phrases (2b), by a suffixal article, also found in unmodified structures (2a), and a free determiner (double definiteness, DD).
(1) a. min stol b. stol-en min
my chair chair.def my
(2) a. hus -et b. det gamle hus -et
house.def def old house.def
‘the house’ ‘the old house’
These structures enable us to consider the relative impact of frequency, complexity, and structural similarity: Postnominal possessives are more frequent than prenominal ones (75%), but are also more complex, involving both definiteness marking (1b) and syntactic movement, see Anderssen & Westergaard (2010), Lødrup (2012). Prenominal possessives are structurally similar to their English counterparts. DD is both complex and infrequent. The suffixal article is extremely frequent, while the prenominal one is infrequent but structurally similar to English.
Anderssen &Westergaard (2010) show that monolingual Norwegian children use both possessive word orders, but have a preference for prenominal ones early on, i.e. the least complex and least frequent structure. This suggests that the children do not simply pay attention to frequency, but rather to structural complexity. Anderssen & Westergaard (2012) investigate possessives in Norwegian-English bilingual children and Norwegian heritage speakers. The results reveal that, while the bilingual children have a stronger and more long-lasting preference for prenominal possessives, the heritage speakers almost exclusively use postnominal ones. Thus, the less complex prenominal possessives are the preferred order in language acquisition, while the high frequency of the postnominal possessive seems to protect it against language attrition. Similar findings have recently been made with respect to adjective-noun word order in Italian-German bilingual acquisition and attrition (Kupisch forthcoming), where the heritage speakers are argued to pay attention to frequency and structural difference.
Against this backdrop, we re-evaluate the data in Anderssen &Westergaard (2012) and argue that frequency is a more important factor in attrition than complexity or structural similarity. We provide data on DD (2b) from a bilingual child (Bentzen 2000; Anderssen & Bentzen 2013) and a number of heritage speakers, and compare them to monolingual acquisition (Anderssen 2007, 2012). Given that DD is both complex and infrequent, we expect it to be vulnerable in both bilingual groups, as it is in monolinguals, who omit the prenominal determiner for an extended period of time, cf. (4). Our findings show that the bilingual child tends to omit the suffix (55.6%), cf. (5), while the heritage speakers omit the prenominal determiner at 57%, despite its structural similarity to English, cf. (6).
(4) store trollet (Ina, 2;1.0)
big troll.def (Target: det store trollet)
(5) den stor ball (Emma 2;7.10)
the big ball (Target: den store ballen)
(6) engelske skolen (coon_valley_WI_sep_03gm)
English school.def (Target: den engelsk skolen)
These results support the hypothesis that complexity plays a more important role than frequency in acquisition. Furthermore, structural similarity may be a factor in bilingual acquisition. In attrition, on the other hand, we argue that frequency and structural difference are the most important factors, accounting for the heritage speakers’ lack of prenominal determiners in DD as well as their overuse of postnominal possessives in Anderssen & Westergaard’s (2012) data.
Anderssen, M. & K. Bentzen. 2013. Cross-linguistic influence outside the syntax-pragmatics interface: A case study of the acquisition of definiteness. Studia Linguistica.
Anderssen, M. & M. Westergaard. 2010. Frequency and economy in the acquisition of variable word order. Lingua 120.11, 2569-2588.
Anderssen, M. & M. Westergaard. 2012. Tospråklighet og ordstilling i norske possessivkonstruksjoner. Norsk Lingvistisk Tidsskrift [Norwegian Journal of Linguistics] 30, 170-197.
Anderssen, M. 2007. The Acquisition of Compositional Definiteness in Norwegian. In M. Anderssen and M. Westergaard (eds.). Papers from the Language Acquisition Workshop, SCL 2006 Nordlyd 34:3, 252-275.
Anderssen, M. 2012. A spanning approach to the acquisition of definiteness in Norwegian. Iberia: an international journal of theoretical linguistics (Special Issue in honour of Andrew Radford’s retirement). 4.1, 1-34.
Bentzen, K. 2000. ‘I like not it like du like it’: A Case Study of Language Transfer in Bilingual First Language Acquisition. University of Tromsø. Master’s thesis.
Kupisch, T. Forthcoming. Language influence in developing vs. end-state grammars and the concept of “cross-linguistic overcorrection”. Bilingualism: Language and Cognition.
Lødrup, H. 2012. Forholdet mellom prenominale og postnominale possessive uttrykk. H.-O. Enger, J. T. Faarlund & K. I. Vannebo (eds.), Grammatikk, bruk og norm, 189-203. Oslo: Novus.