Does grammatical gender in German influence eye movements of 3th and 4th grade children in word-picture matching?

Neitzel, I. (2017)

Introduction: Variable forms of grammatical gender systems exist across languages (Corbett, 1991). All nouns in these languages are divided into gender categories and the noun can be used as a point of reference for other (agreeing) words. In German, the gender category of a noun (female, male or neuter) cannot be deduced from semantic or phonological features of the noun itself in most cases. Consequently, during language acquisition, children have to categorize nouns mostly by the gender feature of agreeing words (e.g. article and adjective), so-called noun-external gender cues, such as ‘eine kleine Biene (‘a small bee’). The psycholinguistic function of gender for language comprehension has not been clarified in conclusion. However, it has been shown for several languages that gender information can have an effect of facilitation on lexical retrieval in adult speakers. For example, nouns can be recognized faster after the presentation of gender-agreeing priming words (gender-congruency-effect, see Heim, Friederici, Schiller, Rüschemeyer & Amunts, 2009; Schiller & Caramazza, 2003; Schriefers & Teruel, 2000). The reason for this effect might be the accelerated activation of parallel gender nodes as described in the model of language production (Levelt, Roelofs & Meyer, 1999) in contrast to different gender nodes.
Aim of this study: A possible influence of noun-external gender cues on eye movements of German-speaking 3rd and 4th grade school children was inspected in this study. We aimed to investigate a possible facilitation of lexical access by markings on indefinite article and adjective in auditory language comprehension as a marker of systematic usage of the grammatical gender system in German native speakers with completed language acquisition.
Method: Eye movements of 32 children (8 to 9 years old, attending 3th or 4th grade, monolingual German, age-appropriate language development) were recorded during a
word-picture matching task. Eye tracking recordings can be used as an effective tool to understand the processing of auditory stimulus files, as eye movements seem to reflect nearly immediately the mental leaps of participants. The so-called ‘visual world paradigm’, in which an auditory stimulus is presented to the participants during an eye movement recording, has been shown to be very sensitive to many language processes (Huettig, Rommers & Meyer, 2011). Two pictures of objects from equal or unequal gender paradigms (e.g. male-male vs. male-female) were presented to the children. Nominal phrases (article, adjective and noun, e.g.’ein kleiner Teller’ – ‘a small plate’) were presented as an auditory stimulus. Children were instructed to look at the target picture as fast as possible. Only in the experimental condition with unequal gender, it is possible to recognize the target already after the auditory presentation of the article or adjective. In the control condition (equal gender), children must wait for the presentation of the noun.
Results: Total fixation time on the target picture was significantly higher in the experimental condition (unequal gender) in comparison to the control condition directly after the presentation of the adjective and even before the noun was given.
Conclusion: Our results show that German-speaking school children notice grammatical gender in auditory language comprehension and indicate that eye tracking can be used as an effective way of investigating specific language processes. Therefore, this method should replenish existing diagnostic tools to evaluate receptive outcomes of grammar therapy especially in clinical contexts. ‘Hidden disabilities’ such as auditory comprehension deficits might be investigated more effectively using ‘visual world paradigms’. Our results show mental processing strategies of monolingual children with completed language acquisition in the field of grammatical gender. A comparable sensitivity of populations with difficulties in gender processing, especially participants with specific language impairment (SLI) or bilingual children acquiring German, should be further investigated. Our data might be used as comparative data for future studies.