L1 Gender Influences Lexical Competition in L2 Listening

Phonological competition in L1 spoken-word recognition has been confirmed by numerous eyetracking studies. Moreover, it is known that during L2 noun recognition, onset-overlapping L1 nouns are also considered. But what role does L1 morphosyntactic knowledge play in L2 listening? For L1, [1] demonstrated that preceding gender information can constrain lexical access: Following gender-marked articles, native listeners do not activate gender-mismatching competitors.

Three eyetracking experiments examined effects of L1 gender during L2 listening. Proficient second-language learners were asked to mouseclick on target pictures in displays containing a competitor and 2 unrelated distractors. In the instructions, gender-marked articles preceded L2 targets. After every experiment, participants completed a questionnaire ensuring they had good knowledge of L2 gender in general, especially of relevant nouns.

In Experiment 1, German-speaking participants heard French instructions. There were 2 conditions, depending on the competitor's gender: In the "same-gender" pairs (1), target and competitor shared gender in both languages, so that neither French nor German gender constrained the initial set of lexical candidates. In the "different-gender" pairs (2) however, target and competitor differed in gender in French, but not in German. Thus, French gender excluded the competitor, but German gender did not. The German participants fixated competitors more than unrelated objects in both conditions: In "different-gender" trials, they were unable to use French gender to suppress competitor activation. By comparison, native participants showed no competition in these trials.

In Experiment 2, the same materials were presented in German to French-speaking learners. Here, the competitor's gender in the different-gender pairs only differed from the target in the participants' L1. French participants showed competition in same-gender trials, but not in different-gender trials. This suggests that they inappropriately used L1 gender in L2 noun recognition. A German control group showed competition in both conditions.

Competitor nouns in Experiments 1 and 2 were cognates, and L2 gender could be stored differently for cognates and non-cognates: In visual-word recognition, for example, [2] obtained an effect of L1 gender for cognates but not for non-cognates. Experiment 3 therefore investigated whether L1 gender interference could also be found with non-cognate L1 competitors. German learners of French received French instructions. The competitor's German name overlapped in onset with the French target, but its French name did not. Moreover, in German, the competitor shared gender with the L1 target in the same-gender pairs (3), but not the different-gender pairs (4). The German listeners looked more often at competitors than at distractors in same-gender trials, but not in different-gender trials, confirming the results of Experiment 2. French controls showed no competitor activation in either condition.

These results contribute to explain the extra effort involved in L2 listening. Just as it is known that L2 listeners cannot inhibit L1 phonology (see, e.g., [3]), it also seems they cannot inhibit L1 gender. However, not using gender to eliminate incorrect word candidates means that the competitor set remains larger; worse, if correct L2 candidates were excluded at first by L1 interference, they may then need to be later recovered at additional cost.

Example items in Experiments 1 and 2:

                                  | French         | German         | English
-----------------------------------------------------------------------------
(1) Same-gender      | target     | perle_[fem]    | Perle_[fem]    | 'pearl'
                     | Competitor | perruque_[fem] | Perücke_[fem]  | 'wig'
-----------------------------------------------------------------------------
(2) Different-gender | target     | cassette_[fem] | Kassette_[fem] | 'tape'
                     | Competitor | canon_[masc]   | Kanone_[fem]   | 'canon'

Example items in Experiment 3:

                                  | French         | German         | English
-----------------------------------------------------------------------------
(3) Same-gender      | target     | table_[fem]    |                | 'table'
                     | Competitor | sapin_[masc]   | Tanne_[fem]    | 'fir'
-----------------------------------------------------------------------------
(4) Different-gender | target     | radis_[masc]   |                | 'radish'
                     | competitor | fusée_[fem]    | Rakete_[fem]   | 'rocket'

Carrier sentences:
French: "Cliquez sur le_[masc]/la_[fem]..."
        ('Click on the...')
German: "Wo befindet sich der_[masc]/die_[fem]/das_[neuter]...?"
        ('Where is the...?')

References:

[1] Dahan, D., Swingley, D., Tanenhaus, M., & Magnuson, J. (2000). Linguistic gender and spoken-word recognition in French. Journal of Memory and Language, 42, 465-480.
[2] Lemhöfer, K. (2004) A ROSE is een ROOS ist eine ROSE. Effects of cross-linguistic similarity in multilingual word recognition. Doctoral thesis, Radboud Universiteit Nijmegen, The Netherlands.
[3] Weber, A., & Cutler, A. (2004) Lexical Competition in non-native spoken-word recognition. Journal of Memory and Language, 50, 1-25.