Dataset Abstract

This dataset contains data from an experimental study on third language acquisition of verb placement. All participants are first language (L1) speakers of Norwegian, with English as theirs second language (L2) who acquire either French or German as a third language (L3). There are two participant groups: 1) L1 Norwegian-L2 English-L3 French (n=125), 2) L1 Norwegian-L2 English-L3 German (n=154). Data are collected using a written acceptability judgment task (AJT) with a 1-4 point Likert scale. The AJT tested placement of finite lexical verb as the second (Verb-2) and third (Verb-3) constituent, respectively. The tasks consisted of 48 sentences: 24 test and 24 filler items. Items were semi-randomized, with four items per page. Test items were 12 non-subject initial main clauses with a topicalized element (Non-SU-I) and 12 subject-initial main clauses with a short adverb (SU-I). For each sentence type, word order was manipulated with the finite verb in either the second (verb-2) or third (verb-3) position, resulting in 6 items in each condition. Fillers were 6 grammatical and 6 ungrammatical wh-questions, as well as 6 grammatical and 6 ungrammatical simple declaratives, the latter with missing arguments. The participants was tested for the same structures in L3 (French/German) and L2 English.

The data set contains information about participant's proficiency in L2 English and L3 French or German, raw data for judgments on AJT, discrimination score, test sentences (including filler sentences), and the replication R code is additionally shared together with the output as an html-file.

The file '0_README_L3 acquisition of verb placement' contains information and description of the dataset and files.

Article Abstract - Related Publication

Authors: Busterud, G., Dahl, A., Kush, Dave & Listhaug, Kjersti F.: Verb placement in L3 French and L3 German: The role of language-internal factors in determining cross-linguistic influence from prior languages

Abstract: This article explores cross-linguistic influence and the relationship between surface structure and underlying syntactic structure in L3 acquisition of verb placement in L1 Norwegian L2 English learners of L3 German or French, respectively. In these languages, verb placement varies systematically. Previous research has found transfer from both L1 and L2 in similar language combinations. Using an acceptability judgment task, we tested verb placement in non-subject-initial and subject-initial sentences. Findings indicate that L3 French learners performed better on non-subject-initial sentences compared to subject-initial sentences, whereas the opposite was the case in L3 German.

We argue that our findings can be explained by a generative account of verb movement and are compatible with an analysis where verbs do not move, or do not move far enough, in the L3 learners’ underlying syntactic representation. Following the assumption that verb movement is a costly operation, we argue that the syntactic operation verb movement is constrained by principles of economy in L3 acquisition, and that economy plays a role in determining cross-linguistic influence in multilingual acquisition. Our account is compatible with a uniform analysis of the acquisition of verb movement in L1, L2 and L3, and underlines the qualitative similarities in different acquisition processes.