Production of plural nouns in German: evidence from non-fluent aphasia

Abstract

This paper investigates the production of German plural nouns by two aphasic participants with non-fluent speech production. Experimental tasks included two production tasks: (1) picture naming of single and multiple objects, and (2) an elicitation task of singular and plural nouns. Materials were controlled for regularity and predictability of plural form, and for dominance of number, referring to the relative difference in word form frequency between a singular and its corresponding plural form. Both regularity and number dominance have been shown to affect plural noun production in both unimpaired and aphasic speakers, but the underlying functional origin of these effects is still a matter of debate. The results point to differences in the lexical representation and processing of regular and irregular German plural nouns. Thus, the data are in line with the dual mechanism account.

Keywords:

• aphasia
• plural nouns
• dual mechanism account
• plural dominance
• regularity
• inflection

Acknowledgements

We are indebted to MM and RA for participating in the study. We thank Anja Jennings for her help in the preparation of materials and data collection. We also thank Lyndsey Nickels, David Caplan and two anonymous reviewers for helpful comments on earlier versions of this paper, and we thank Pienie Zwitserlood for interesting discussions of the results. Thank you also to Catherine Mason for reading of proofs.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Sonnenstuhl & Huth (2002) reported the following percentages for type frequency of the different forms, counted in CELEX (Baayen et al., 1993): -n: 48%, -e: 27%; -Ø: 17%, -er and –s: both 4% (cf. Sonnenstuhl & Huth, 2002, Table 1, pp. 278).

2. Note that the (e)n-suffix can also occur with masculine and neuter nouns, and with nouns not ending with schwa. In these cases, however, it's occurence is not fully predictable (see Sonnenstuhl & Huth, 2002 for a description of four different subclasses of the German n-plural).

3. Regular German participles are using a t-suffix, such as in kaufen [to buy] → ge-kauf-t [bought], irregular participles are using an -en-suffix, such as in laufen [to walk] → ge-lauf-en [walked].

4. Note that Biedermann et al. (2012, 2013) controlled their materials differently to Baayen et al., 2007; Biedermann et al. used two subsets, matched for surface frequency across plural subsets (singular-dominant vs. plural-dominant), but the surface frequency across singular subsets (inevitably) differed. In addition, stem frequency between singular dominant and plural dominant items were matched. In contrast, Baayen et al. (2007) used two subsets of items with different frequency of singular and plural surface form, but matched stem frequency (see also Baayen et al., 1997).

5. Inflectional entropy is a measure that reflects the frequency distribution of an inflected form, such as the relative distribution of singular and plural forms of a noun. Relative entropy considers the distributional properties of all words that fall into the same inflectional class in a language, such as all German singulars and plurals (cf. Baayen et al., 2007).

6. In German, case is overtly marked on determiners (e.g., SVO: Der_Nom Sohn sucht den_Acc Vater. [The son is seeking the father.] versus: OVS: Den_Acc Sohn sucht der_Nom Vater.), and number is (additionally) marked on verbs and nouns (e.g., SVO: Die Frauen_plural suchen_plural das Kind_singular [The women are seeking the child] versus OVS: Das Kind_singular suchen_plural die Frauen_plural).

7. Two er-plurals, and 30 e-plurals were included in the irregular set; overall 15 plurals involved additional vowel changes (Umlauting), e.g., Zahn → Zähn-e [teeth → tooth], Wurm → Würmer [worm → worms].

8. The different numbers of participants resulted from collecting stimulus pictures from different sets.

9. The determiner “die” can also refer to the nominative singular of the feminine determiner. Since RA almost exclusively omitted the determiner in this task (the task was to name the object with one word, not with a complete determiner-noun phrase), we assumed that he used “die” in order to mark the plural of the noun (die Zebras = the zebras).

10. In a grammaticality- judgement task with gender-marked determiners and nouns, MM showed completely unimpaired performance, indicating preserved access to grammatical gender at lemma level (Lorenz & Zwitserlood, 2014), whereas RA was mildly impaired in this task (Lütteken, 2011).

11. Recorded by Lütteken (2011).

Additional information

Funding

Britta Biedermann was funded by an Australian Research Council Postdoctoral Fellowship (ARC APD), [grant number DP110100799].