Search

Acceptability task that compares scores for adequate and inadequate connectives as well as implicit relations across native speakers and learners with and without possible transfer effects.

Experiment that manipulates the structure of the segments and the frequency of the connective across native and non-native participants.

In this reading task, participants will have to read segments of sentences connected by a contrastive connective or with an implicit contrastive relation, and answer comprehension questions. We manipulate the use of a parallel or non-parallel structure across the two segments of the discourse relation, and the presence or absence of a connective.

Offline task to measure accuracy of disambiguation with or without parallel structure in native and non-native participants.

Discourse connectives are lexical items like "but" and "so" that are well-known to influence the online processing of the discourse relations they convey. Yet, discourse relations like causality or contrast can also be signaled by other means than connectives, such as syntactic structures. So far, the influence of these alternative signals for discourse processing has been comparatively under-researched. In particular, their processing in a second language remains entirely unexplored. In a series of three self-paced reading experiments, we compare the reading patterns of contrastive relations by native French-speakers and non-native speakers of French with English as a first language. We focus on the effect of syntactic parallelism and how it interacts with different types of connectives. We test whether native and non-native readers equally recruit parallelism to process contrast in combination with or without a connective (Experiment 1), with a frequent vs. infrequent connective (Experiment 2) and with an ambiguous vs. unambiguous connective (Experiment 3), thus varying the explicitness and ease of retrieval of the contrast relation. Our results indicate that parallelism plays an important role for both groups of readers, but that it is a more prominent cue for non-native speakers, while its effect is modulated by task difficulty for native participants.

Reading and understanding a text in L2 can be a cognitively demanding task. While the impact of discourse connectives and alternative signals is rather well documented for L1 speakers, the literature about the reading processing of connectives of language learners is yet relatively sparse. Among the few studies that assessed the processing of discourse relations in L2, the recent findings by Zufferey and Gygax (2017) and Recio Fern��ndez (2020) demonstrate that even highly proficient L2 readers fail to detect incoherencies during online reading when they have to process implicit discontinuous relations (e.g., concessive and confirmation relations), as predicted by the continuity principle (Segal et al., 1991; Murray, 1995, 1997) as well as the causality-by-default assumption (Sanders, 2005). However, results from these experiments also raise intriguing questions that we address in a new set of experiments. First, we do not know yet at which proficiency level learners acquire a native-like ability to process continuous and discontinuous discourse relations. Second, learners��� ability to process continuous and discontinuous relations might depend on the frequency and register of the connective that conveys it. It can be assumed that connectives which are not common in speech might represent, especially for untrained speakers and readers, an added complexity to text processing. Since working memory is used for the encoding these unusual elements, less proficient L2 learners should not able to detect a loss of coherence within the already demanding discontinuous relations. Finally, previous experiments showed no evidence for an interaction effect between the type of relation (causal or concessive) and the nature of the link (explicit or implicit). In other words, there is no evidence that for example a discontinuous relation is processed more slowly by non-native speakers when the link is implicit compared to a continuous relation. In a set of three experiments, we will assess the way German-speaking learners of French process relations by means of a self-paced-reading-task, by comparing coherent and incoherent relations (Experiment 1), explicit and implicit relations (Experiment 2) and relations that are conveyed by a frequent or an infrequent connective (Experiment 3). Whereas both Zufferey and Gygax (2017) as well as Recio Fern��ndez (2020) tested highly frequent connectives in French and Spanish, we focus in Experiment 3 on infrequent connectives that are bound to the written mode. By doing so, we gain further insight into the impact of the connectives��� characteristics on online text processing in L2. We will further assess the participants��� language proficiency using the French version of the Lextale-task (Brysbaert, 2013) which showed in previous experiments high correlations with an offline sentence-completion-task with discourse connectives, as well as with a grammar task (Zufferey & Gygax, 2020). We will also link L2 learners��� ability to process French connectives to their degree of exposure to print in their native language, as this factor was found to be relevant in offline tasks with native (Zufferey & Gygax, 2020) and non-native speakers (Wetzel, Zufferey & Gygax, 2020).

The mastery of connectives like n��anmoins, en effet and donc in French is a well-known problem in corpus-based studies of learner language (e.g., Lamiroy, 1994; Degand & Hadermann, 2009;). However, the cause of these difficulties remains highly debated as only few experiments have so far been conducted on this topic (e.g., Zufferey et al. 2015; Wetzel et al., 2020). In order to investigate the role of crosslinguistic influence, we compared the ability of two groups of learners to master the connective alors in French. On the one hand, a group of German-speaking learners who typically misuse this connective to convey a specification as in (1), a misuse that could come from a crosslinguistic influence of the German also. On the other hand, a group of English-speaking learners who do not produce this misuse in corpus data. (1) * L���arbre, alors le sapin que j���ai plant��, est vieux. * ���The tree, CONNECTIVE, the fir I planted, is old���. In an on-going set of experiments, we compare the ability of these two groups of learners as well as a control group of native speakers to discriminate between correct (consequence) and incorrect (specification) uses of alors across an offline judgement and an online reading task. In the judgement task, our results indicate that even though both groups of learners provide similar judgements of the incorrect uses of alors, the group of German-speaking learners have a lower ability than English-speaking learners to assess correct uses of alors, and also rate lower the correct way to produce a specification in French with the connective c���est-��-dire. In the online reading task using self-paced reading, we expect that because of their lower sensitivity to the correct uses of alors, German-speaking learners will have different reading patterns of correct and incorrect sentences compared to the other groups. Finally, we will also assess whether the sensitivity to detect the misuse despite crosslinguistic influences can be predicted by the overall language proficiency of the learners.

Connectives are linguistic items that indicate discourse relations like cause and condition between discourse segments and thus represent crucial elements for coherence (Halliday and Hasan, 1976; Sanders, Spooren, & Noordman, 1992). Despite their high frequency in most discourse genres, the ability to handle connectives typical of the written mode has been found to be quite variable even among adult native speakers (Zufferey & Gygax, 2020), and this variability has been linked to the degree of exposure to print that people have. In this presentation, we investigate the ability of German-speaking learners of French to use 12 French connectives typical of the written mode. We also investigate the role of three variables that could account for the variability between learners: language proficiency; exposure to print in the second language (French) and exposure to print in the first language (German). In a sentence completion task performed online, we tested 151 German-speaking learners of French as well as a control group of 63 French native speakers. In order to assess learners��� proficiency level in written French, we gave them a written language competence task (Zufferey and Gygax, 2020) and a vocabulary knowledge-test (Lextale, Brysbaert, 2013). In addition, we tested the exposition to print in both their L1 (Art-Ger, Grolig, Tiffin-Richards & Schroeder, 2020) and L2 (Art-F, Zufferey and Gygax, 2020). Finally, in a self-assessment task, we measured learners��� perceived importance of 10 given factors (e.g. school, friends, reading, Internet) for their acquisition of French. Results indicate that for the non-native speakers, a higher vocabulary knowledge and a better grammar mastery predicted a better score in the main task. More intriguingly, a higher exposition to print in L1, but not in L2, also predicted a better mastery of connectives in L2. These results thus tend to indicate that L2 acquisition research should focus more on the first language competence of the participants as predictors of their ability to use connectives, as results show that there are several factors that facilitate or hinder learners��� mastery that are linked to their native language. These factors raise important questions about a possible interaction between the native and foreign languages (see also Sparks, Patton, Ganschow, Humbach, & Javorsky, 2006; Sparks, Patton, Ganschow, & Humbach, 2012). Finally, participants that assigned a higher importance to the factor reading for their acquisition of French were more likely to score higher at the main task. Beside supporting our finding of the link between connective mastery and the exposition to print, this finding shows that subjective impressions of participants are reliable predictors of their level of competence.