Facilitating morphosyntactic and semantic prediction among second language speakers of German

ABSTRACT

Using two visual priming experiments, the present study investigates whether presenting facilitative semantic (i.e. colour) and morphosyntactic (i.e. grammatical gender) information in a prime image prior to reading a target sentence facilitates naming of a sentence-final target image among L1 German and L1 English-L2 German speakers. In Experiment 1, L1 and L2 German speakers used both semantic and gender cues to predict the sentence-final target image. In Experiment 2, a new group of L2 German speakers used gender cues to predict, but this effect was stronger when gender information was provided via a gender-marked indefinite article and adjective in the prime, than when the prime contained only the gender-marked article. These results suggest that if L2 speakers are able to overcome unstable and often inaccurate L2 gender assignment, they can use gender cues in a native-like manner for prediction, but that multiple gender-marked cues may be necessary for such prediction to occur.

KEYWORDS:

• Predictive processing
• grammatical gender
• processing resources
• German
• second-language acquisition

Acknowledgements

Thank you to Joe Kelly who helped with data collection. Data associated with this study were presented at the following conferences: CUNY Conference on Human Sentence Processing (March, 2016); International Conference on Multilingualism (October, 2013); German Linguistics Annual Conference (April, 2013). We are grateful to all conference attendees for their valuable feedback. We also thank the three reviewers for their helpful comments and suggestions.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 For all t-tests, when the assumption of homogeneity of variance was not met, the Welch–Satterthwaite method was applied to the degrees of freedom in SPSS and the corrected p-value is reported.

2 Although every image only appeared once throughout the experiment as a target image, due to a programming error 11 images appeared 2 times over the course of Experiment 1 (and Experiment 2) as a foil image. A second set of analyses for both Experiment 1 and Experiment 2 that excluded the 11 items in which these images were the target image revealed the same pattern of results as those reported in the main manuscript here.

3 An analysis including only the L2 speakers similarly revealed that no two- or three-way interactions involving target gender and the variables gender prime or semantic prime significantly improved the model fit (all ps > .4).

4 We ran the same analyses reported here using gender naming accuracy from the picture naming task, as this measure has been used to measure gender assignment accuracy in previous studies (e.g. Hopp, 2013, 2016). However, these analyses revealed no significant effects or interactions between gender naming accuracy and the other factors (all ps > .1). An additional analysis including L2 proficiency (as measured by the 30-point multiple choice German grammar task) revealed that neither the inclusion of L2 proficiency as a main effect, nor any interaction terms between L2 proficiency and gender prime or semantic prime improved the model fit (all ps > .3).

5 In Experiment 2 the ratio of items for which participants assigned the correct gender to both the target and foil word on the gender assignment task versus the incorrect gender for at least one word was sufficiently imbalanced (n = 154 versus n = 522) that the results of any analysis including item-specific gender assignment accuracy as an additional variable would be unreliable. Finally, because the L2 speakers in Experiment 2 were older, on average, than the L2 speakers in Experiment 1, we ran an additional analysis with the subset of L2 speakers in Experiment 2 who were in the same age range as the L2 speakers in Experiment 1 (n = 17). While the results were not statistically significant due to a loss of statistical power, the overall pattern of results was the same.

Additional information

Funding

This research was funded in part by the National Science Foundation under grant OISE-0968369 (PI: J. F. Kroll; co-PIs: P. E. Dussias and J. G. van Hell) and a grant to Dr Carrie Jackson from the Humboldt Foundation.