The theme of this special issue of the Nordic Journal of Linguistics was intended to solicit contributions on Nordic languages being acquired as non-native languages as well as work on Nordic-language native speakers learning a second language, not necessarily a Nordic one. We adopted a wide definition of the term ‘second language’ (L2), i.e. a non-native (second, third, etc.) language acquired in late childhood, adolescence or adulthood, in a naturalistic or an instructed setting.

Among the Scandinavian languages (Danish, Faroese, Icelandic, Norwegian and Swedish) and the other Nordic languages (Finnish, Greenlandic and Saami) only some are regularly studied as a foreign language outside the Nordic countries. None of these languages is widely taught outside the Nordic area, and thus the number of foreign language learners of Nordic languages is relatively modest. Within the Nordic countries however, there are sizable numbers of naturalistic and instructed L2 learners of Scandinavian languages, largely due to migration and bilingual education.

Conversely, for many native speakers of a Nordic language, a second language is a near-must. Higher education is not always available in one's first language in the Nordic region, and communication outside one's home area or community often relies on a non-native language – typically Danish, Norwegian, Swedish or English. L2 use is also commonplace in areas of language contact, such as the Arctic area of Scandinavia.

However, systematic research on L2 acquisition in the Nordic countries only dates back a few decades. New theories of L2 learning from North America and mainland Europe were embraced at Nordic universities and teacher training colleges at a time when some of the Scandinavian countries, Sweden in particular, saw a major influx of immigrants who would soon be acquiring the language of the host country. In the 1970s and 1980s, researchers began collecting production data from these adult learners, often in cross-sectional studies. Some data were also collected from school-age instructed L2 learners, particularly native speakers of Finnish learning Swedish.

As in other countries at that time, a renewed interest in language universals led to a search for typical L2 features and universal L2 developmental sequences, especially concerning syntax and morphology. This search for L2 universals continues today within a number of theoretical frameworks (e.g. Klein & Perdue Reference Klein and Perdue1997; Schwartz & Sprouse Reference Schwartz, Sprouse and Archibald2000; Carroll & Lambert Reference Carroll, Lambert, Dimroth and Starren2003; Pienemann Reference Pienemann2005; von Stutterheim Reference von Stutterheim, Foster-Cohen and Doehler2003; Hammarberg Reference Hammarberg, Hyltenstam and Lindberg2004; White Reference White, Ritchie and Bhatia2009). But proposals of universal L2 sequences and the theories behind them have also been challenged by recent studies that compare L2 learners whose native languages (L1s) are typologically different. Many studies suggest that the properties of the L1 influence the course of L2 acquisition to a considerable extent (e.g. Hammarberg Reference Hammarberg, Cenoz, Hufeisen and Jessner2001; Westergaard Reference Westergaard, Foster-Cohen and Doehler2003; Bohnacker Reference Bohnacker, den Dikken and Tortora2005, Reference Bohnacker2007; but see Håkansson, Pienemann & Sayehli Reference Håkansson, Pienemann and Sayehli2002). How much such findings corroborate earlier proposals of transfer and whether transfer effects show up to the same extent in all linguistic domains (e.g. phonology, morphology, syntax, discourse pragmatics) remains a matter of vivid debate, and more empirical studies are needed.

Another major avenue in L2 research are comparisons of child L2 and adult L2 acquisition. However, effects of age on the developmental course and eventual outcome of L2 acquisition have not been investigated much for the Nordic languages, and there are remarkably few studies of child L2 acquisition of the Scandinavian languages or any of the other Nordic languages. We thus do not know whether child L2 learners of Nordic languages and Nordic child L2 learners of other languages invariably outperform late L2 learners, and if so, whether performance with increasing age declines gradually or abruptly (and thus, in the latter case, might be indicative of a critical period).

The four articles in this volume address some of these issues and illustrate the diversity of the L2 research carried out in the Nordic domain today. Our call for papers inspired eight rather wide-ranging contributions, four of which we were able to accept for publication after multiple reviews. We would like to thank the eighteen referees for their time and effort spent in reviewing these manuscripts.

Two papers discuss the L2 acquisition of a Nordic language, Swedish in the paper by Bohnacker and Norwegian in the paper by Busterud, while the other two, by Thomas and Gao, deal with Swedish learners’ L2 acquisition of French and Chinese, respectively. The papers also address a variety of syntactic constructions, namely the nature of the clause-initial position in verb-second (V2) languages, specifically Swedish vs. German (Bohnacker), anaphoric binding in Norwegian (Busterud), verbal morphology in French (Thomas) and classifiers in Chinese (Gao). They investigate the acquisition of these syntactic constructions in connection with other domains, such as information structure and discourse pragmatics (Bohnacker), aspectual semantics (Thomas) and lexical semantics (Gao).

Major issues in L2 research are dealt with in these papers, e.g. the difference between adult and child learners (Gao), the nature of the input (Busterud) and the importance of input frequency (Thomas, Bohnacker), and the question of L1 transfer (Bohnacker). Furthermore, Busterud discusses methodological challenges related to the investigation of ambiguous structures. We now turn to a more detailed description of the four articles in the order that they appear in the volume.

The first paper, by Ute Bohnacker, is entitled ‘The clause-initial position in L2 Swedish declaratives: Word order variation and discourse pragmatics’. It discusses discourse-related linearization preferences concerning the prefield, i.e. the clause-initial position in V2 clauses. Extending an earlier study of informal written texts (Bohnacker & Rosén Reference Bohnacker and Rosén2008) to informal speech data, Bohnacker documents a significantly stronger tendency in Swedish than in German to start a V2 clause with a lightweight element, particularly with an expletive subject. New information (the rheme) is placed postverbally more often in Swedish than in German. Bohnacker suggests that Swedish has stricter discourse-pragmatic constraints than German on what can (or does) occur in the prefield. She then investigates naturalistic production data from six L1-German advanced learners of Swedish after 3, 6 and 9 years of L2 exposure. These learners exhibit German-style frequency patterns (e.g. a low proportion of expletives in the prefield) and non-Swedish information-structural patterns (e.g. heavy rhematic elements in the prefield), particularly in the earliest data at 3 years of exposure. From a Swedish perspective, they ‘overuse’ the prefield, a finding which can be related to Bohnacker & Rosén's (Reference Bohnacker and Rosén2008) study of the reverse combination of L1-Swedish learners of L2 German. From a German perspective, the L2 learners in this study underused the prefield, restricting it largely to elements of no or low informational value. Bohnacker interprets her L2 Swedish data at 3 years as evidence for L1 transfer of information-structural or discourse-pragmatic preferences, and thus makes a contribution to the current transfer debate in L2 acquisition theory, which has so far been largely concerned with phonological and syntactic phenomena. The L2 data also indicate a gradual development towards the target, in the sense that the learner productions at 6 and 9 years show substantially higher rates of clause-initial expletive subjects, clefts and lightweight given elements (e.g. pronominal object det ‘it’). This suggests, Bohnacker argues, that syntax–pragmatic difficulties can be overcome as proficiency develops. These findings are related to current generative theorizing on the syntax–pragmatics interface, where it is often maintained that the interaction of syntax and discourse pragmatics is one of the hardest areas to master. A commonly cited reason for this is that it involves a grammar-external interface, where multiple types of information have to be integrated with each other (e.g. Hulk & Müller Reference Hulk and Müller2000; Sorace & Filiaci Reference Sorace and Filiaci2006; Belletti, Bennati & Sorace Reference Belletti, Bennati and Sorace2007; Rothman Reference Rothman2009). By contrast, Bohnacker argues that it is not the involvement of an interface as such that makes the L2 acquisition of certain structures difficult, but rather the fact that language-specific constraints can have subtle consequences for information distribution in discourse, and these consequences are often not categorical, but gradual in nature. Native speakers have preferences for information selection and distribution in particular contexts and discourse types, and when advanced non-native speakers neglect to adopt these preferences, this may lead to a ‘discourse accent’.

Guro Busterud's paper, ‘Methodological problems related to L2 research on Norwegian anaphors’, focuses on the methodological challenges involved in investigating anaphoric binding in Norwegian as a second language. Norwegian anaphors can be bound both locally and non-locally, and since anaphors vary cross-linguistically, it is of interest to establish whether and where L2 speakers of Norwegian allow such target-like local and non-local binding. Truth-value judgment tasks are often employed to elicit speakers’ intuitions of anaphoric binding in sentences with two possible antecedents (e.g. ‘Line heard Anne tell about SELF’). Such sentences might be ambiguous for L2 speakers. However, in native Norwegian, long-distance binding cannot cross a finite clause boundary, and the long-distance anaphor cannot be locally bound. Because of this, Busterud argues that the truth-value judgment task may sometimes be less adequate for testing all relevant binding structures in Norwegian, and suggests that an interpretive judgment task be used as a complement. Busterud also notes that there is dialectal variation in Norwegian that affects long-distance binding. This poses additional challenges for the study of the L2 acquisition of anaphors, as the nature of the input to L2 learners will need to be taken into account.

Anita Thomas’ paper, ‘Lexical aspect versus input frequency in adult L2 beginners of French’, presents a study of the variation found in the oral production of verbs in L2 French by adult beginners with Swedish as L1. The study deals with the production of the two main forms of regular verbs in spoken French, a short form /parl/ and a long form /parle/, in present-tense and infinitive contexts. As has been found in several other studies of adult L2 learners of French, Thomas’ learners at beginner level use these two forms invariantly in both finite and non-finite contexts. Thomas investigates the choice of the invariant verb forms made by the learners for twelve verb types. This is done by testing the influence of input frequency as well as the influence of the lexical aspect of the verb. Thomas compares the distribution of short and long forms in the data from input sources (i.e. teacher–learner dialogues, classroom interaction, spoken colloquial French, and textbook material) with the learner data collected in two tasks, a free production task and an imitation test. Although for many of the verbs studied, the two factors tested cannot be easily teased apart, Thomas’ results suggest that the learners’ choice of verb form is influenced more substantially by input frequency than by lexical aspect.

Finally, Helena H. Gao's paper, ‘A study of Swedish speakers’ learning of Chinese noun classifiers’, addresses the acquisition of a complex feature of Chinese grammar. Chinese noun classifiers (CL) are an obligatory category associated with nouns and constitute categorical linguistic indicators dividing real-world entities into different cognitive sets, e.g. entities with a rectangular shape, entities with a longitudinal shape, certain animals (e.g. yì gēn zhēn ‘a cl needle’, yì zhī yáng ‘a cl sheep’). No such linguistic categories exist in Swedish. Gao examines the acquisition of classifiers by native Swedish adult L2 learners of Chinese during a three-month intensive language course, using a repeated-measure elicitation task and follow-up interviews. Their learning results are compared to a group of bilingual Swedish–Chinese children. The children do relatively well; for the adult L2 learners, however, a full understanding and good mastery of Chinese noun classifiers proves hard to achieve. While a few high-frequency classifier–noun combinations are produced target-consistently, the adults’ learning of classifiers lags behind their general proficiency level of Chinese. Some L2 learners seem to adopt a grammatical rule-based strategy and overuse the general classifier gè, thus producing syntactically correct but semantically inappropriate noun phrases. Other L2 learners consciously overextend classifiers used with a particular noun (e.g. ‘bed’) to related nouns (e.g. nouns denoting other pieces of furniture), based on groupings of real-world entities which they are familiar with from their native Swedish, but which do not correspond to Chinese classifier cognitive sets. Gao thus argues that the adult learners have not acquired the complex semantic and cognitive meanings embedded in the classifiers.