Verbs in American Norwegian: Tense/Finiteness morphology and V2-syntax

Alexander Lykke, ILN

Introduction: This abstract and talk departs slightly from the conference talk genre. I will start out with discussing a concluded study on the possible interplay between Finiteness Morphology and V2-syntax, which answers most of the questions in the project proposal for my PhD-project. For the last third of the talk, I will turn to work in progress research on Heritage Norwegian tense morphology. The overarching theme is linguistic variation and change.

The first study explores morphologically marked finiteness (FIN) and verb-second word order (V2), in North American Heritage Norwegian (HN), and a possible link between them. A theory put forth by Eide (2009), claims that in English the loss of the productive FIN-marking on main verbs of the weak inflection, has led to a categorical loss of morphological FIN, causing a subsequent loss of V2 in main clause declaratives. These changes separate English from other Germanic languages, e.g. Norwegian. The present study investigates morphological finiteness marking on verbs, and verb placement in main clause declaratives in the Norwegian speech of speakers of HN, whose dominant language is English. In the present material, the link between FIN-morphology and V2 is not apparent, and both categories seem close to the relevant baselines. In a broader perspective, these findings bear on the questions on what causes syntactic change: Is syntactic change caused by preceding morphological change or is it rather driven by other factors like pragmatics and/or semantics? In this case, morphology appears to play no role.

Data: The survey is based on experimental data from nine participants, who each produce approx. 15 main clause declaratives with topicalization. The participants show some variability with regards to production of V2. Three participants produce no non-baseline word order. A second more tenuous grouping of five participants, produce a majority of target-like syntax, but have non-V2 in 7%–33% of their sentences. The last participant, Fargo_ND_01gm produces non-target word order in 10 out of 15 instances (67%). With regards to FIN-morphology, no participant shows a system clearly different from Homeland Norwegian. Firstly, there are no instances of unambiguously inconsistent or non-target-like use of the morphological forms, i.e. there is no use of an unambiguous infinitive for the present, perfect participle for the preterite or vice versa. Secondly, the observed verbal paradigms are almost formally identical to what is found in the baseline Norwegian dialects.

Discussion: The results do not show a clear correlation between (absence of) FIN-morphology and (absence of) V2 in these experimental data. E.g., two speakers with no V2 violations, Sunburg_MN_03gm and Sunburg_MN_12gk, display different morphological FIN-systems: Sunburg_MN_03gm has a system with the FIN-distinction even in the least formally differentiated class (the a-class), whereas sunburg_MN_12gk has a system without the finiteness distinction in this same class. Indeed, there are Norwegian baseline dialects, notably from the Southern Gudbrandsdal area, which lack the morphological FIN-distinction in the a-class (Eide & Hjelde 2015: 78–79), where the status of V2 has never come into question. Sunburg_MN_12gk has ancestral ties to the Gudbrandsdal area. Fargo_ND_01gm (with 67% non-V2) has the same morphological system as sunburg_MN_03gm, who has 100% target-like V2.

I have not found other linguistic factors correlating with non-V2. Eide & Hjelde (2015), studying V2 in HN, report a higher degree of V2 violations when the topic is more syntactically complex. Fargo_ND_01gm, the only participant with a majority of non-V2 clauses, does not show any such tendency. Furthermore, there is no apparent correlation between the type of topic (e.g. adverbial/object etc.) or the type of subject (e.g. pronominal/nominal) and the non-V2 syntax.

Conclusion to the FIN–V2-story: The finiteness and/or tense morphology appears to follow the homeland dialects for most of the participants. This is in keeping with a reported tendency for tense morphology to be stable in heritage grammars, even though morphology is vulnerable to change (Benmamoun et al. 2013: 141–144). Furthermore, my study shows that V2 (i.e. V-to-C movement) in main clause declaratives is relatively stable despite the fact that syntax pertaining to the CP layer is reportedly susceptible to change in heritage grammars (see e.g. Benmamoun et al. 2013: 148–149). My results are in line with Håkansson (1995), who shows that V2 is at baseline level in data from five heritage speakers of Swedish.


The way onwards: Change and stability in tense morphology


Regular Weak verbs in Nynorsk

Larger weak class (a-class)

Smaller weak class (te-type)

Smaller weak class (de-type)

Smaller weak class (dde-type)


hopp-e ‘jump’


kjøp-e ‘buy’


prøv-e ‘try’ /prø:ve/

nå-Ø ‘reach’ /no:/





























In my ongoing research, my data are retrieved from the Corpus of American Nordic Speech (CANS, Johannessen 2015). The CANS consists of the recorded speech of American Norwegian speakers, which is transcribed orthophonically, transliterated orthographically. The corpus is also tagged for several grammatical categories, e.g. part of speech, by an automatic tagger. At the present the corpus contains 251315 tokens/words, distributed across 69 speakers.

Given the reported tendency for tense morphology to be a stable category in an unstable domain (morphology), I ask: What changes, if any, occur in American Heritage Norwegian, and how can we explain them? Additionally, how can this cross-linguistic stability of tense morphology in heritage varieties be explained?

For the present, I will report some tendencies in the data from the CANS. The main tendency is indeed that tense morphology is stable, but we do see some variation and change in use of inflectional affixes. One tendency is overregularization, e.g. the irregular verb ‘walk’ – gikk (pret.) getting a regular preterite gådde. An explanation for such a levelling is probably found in decreased exposure to Norwegian language.

Another tendency is variable use of inflectional affixes with one root, e.g. plukke ‘pick’ being inflected as both plukte and plukka in the preterite by the same informant. I have found two speakers, Chicago_IL_01gk and Coon_Valley_WI_12gm, who produce tense morphology which at the very least shows that they have uncertain knowledge of the inflectional endings.



Benmamoun, E., S. Montrul & M. Polinsky (2013): «Heritage languages and their speakers: Opportunities and challenges for linguistics.», In: Theoretical Linguistics 39 (3–4), DOI: 10.1515/tl-2013-0009, pp. 129–181

Eide, K. M. (2009): «Finiteness: The haves and the have-nots.», In: A. Alexiadou, J. Hankamer, T. McFadden, J. Nuger, F. Schäfer (red.): Advances in Comparative Germanic Syntax, John Benjamins Publishing Company, Amsterdam / Philadelphia, pp. 357–389

Eide, K. M. & A. Hjelde (2015): «Verb Second and Finiteness Morphology in Norwegian Heritage Language of the American Midwest.», In: R. B. Page & M. Putnam (red.): Moribund Germanic Heritage Languages in North America, Brill, Leiden / Boston, pp. 64–101

Håkansson, G. (1995): «Syntax and morphology in language attrition: A study of five bilingual expatriate Swedes.», In: International Journal of Applied Linguistics 5, pp. 153–171

Johannessen, J. B. (2015): «The Corpus of American Norwegian Speech (CANS).», In: B. Megyesi (red.): Proceedings of the 20th Nordic Conference of Computational Linguistics NODALIDA 2015, Linköping University Electronic Press, pp. 297–300, 23. sept. 2015



Publisert 27. apr. 2018 10:41 - Sist endret 27. apr. 2018 10:41