“Cross-linguistic influence in grammatical gender processing: Polish in contact with German and Ukrainian”
Kamil Długosz (Adam Mickiewicz University in Poznań)
Grammatical gender processing in bilinguals involves the integration of bottom-up and top-down sources of information, such as morphophonological cues and lexical gender representations. Previous research has shown that gender information can be co-activated across languages, often resulting in facilitated processing when nouns share gender across a bilingual’s two languages (the gender congruency effect). At the same time, bilingual speakers can rely on morphophonological cues such as noun endings when processing gender and may transfer these cues from the first language to the second. However, much less is known about how these mechanisms interact across different bilingual contexts and whether influence can also extend from the second language to the native language.
In this talk, I present two lines of research on gender processing involving Polish in contact with German and Ukrainian. First, I report results from eye-tracking and production studies with late Polish–German bilinguals that examine whether an L2 acquired later in life can affect gender processing in the native language. Focusing on representational stability, agreement context, and cognateness, the findings show that native-language performance is not immune to L2 influence and that this influence is modulated by these three factors.
Second, I examine microvariation in gender production in L2 Polish among L1 Ukrainian children. Using timed production data, I show that children transfer the distributional frequency of gender cues from the L1 to the L2, affecting the retrieval of gender during production.
Together, these findings shed light on how two gender systems interact in the bilingual mind and highlight the need for research designs that jointly examine lexical information and morphophonological cues in bilingual gender processing.
“When Do Effects Emerge? Resampling Methods for Latency Comparisons in the Visual World Paradigm”
Serge Minor (UiT The Arctic University of Norway)
Over the last three decades, eye tracking in the visual world paradigm (VWP) has been a valuable tool in language processing research. In a typical VWP experiment, researchers compare looks to particular pictures or objects in a scene while manipulating some aspect of the linguistic stimulus. Usually, the aim of the analysis is to compare the strength of preference for particular pictures within a set time window of interest. However, it is often also of theoretical interest to determine when an effect starts, e.g., when the participants first begin to exhibit a preference for a particular picture, and to compare the onsets of the effect between groups or conditions.
In this workshop, we will discuss resampling-based methods for comparing effect latencies in the VWP. We will focus on the divergence point analysis procedure proposed by Stone et al. (2020), which uses non-parametric bootstrapping to estimate the uncertainty around effect onsets, as well as permutation-based testing for latency differences (Minor et al., 2022). We will systematically walk through each step of these analytic procedures. Particular attention will be given to the “researcher degrees of freedom” involved in these methods (i.e., the choices that we as researchers have to make) and how these analytical choices may impact statistical power and false-positive rates, based on recent findings by Minor (2026).
“Modeling multilingual grammatical acquisition and attrition: the role of similarity, cue salience, and learner’s age “
Natalia Mitrofanova (UiT The Arctic University of Norway, Tromsø)
The talk will focus on cross-linguistic influence (CLI) in multilingual children aged 6–12 years speaking East Slavic home languages (Ukrainian or Russian) and residing in Germany, Spain, Poland, and the Czech Republic. We will explore how structural similarity between properties, cue salience, age-of-L2-onset and their interaction predict the direction and strength of grammatical CLI in multilingual contexts. The findings draw on two projects that investigate bidirectional CLI in the domains of nominal case and verbal aspect – areas known to be vulnerable in heritage speakers.
The first project focused on Russian-speaking children growing up bilingually with German or Spanish as their societal language. We used Visual World eye-tracking and tested the children in both their heritage (Russian) and societal languages. The bilingual groups were matched by age, overall exposure, and lexical proficiency in the heritage language. The selection of the populations was motivated by structural overlap between the languages: while both Russian and German mark grammatical case (nominative vs. accusative) overtly in noun phrases, Spanish does not employ overt case marking. On the other hand, Spanish and Russian share similarities in their use of verbal aspect (perfective vs imperfective) to distinguish between completed vs ongoing events in the past, while German doesn’t grammatically mark aspect on the verb. We make comparisons across the two bilingual groups in the heritage language as well as with monolingual controls in the societal language. The results highlight the role of cue salience in shaping the direction and extent of CLI (societal to heritage and heritage to societal language).
The second project extends this investigation to multilingual Ukrainian refugee children residing in Poland and the Czech Republic and focusses on structural variations within Slavic languages. Specifically, we investigated the use of accusative plural marking for animal referents (which is syncretic with genitive in Ukrainian but with nominative in Polish) and the use of perfective aspect is habitual contexts (grammatical in Czech but not Ukrainian or Polish). By analysing production data and acceptability judgements paired with pupillometry, we investigated how structural (dis)similarity between languages, age-of-L2-onset, and their interaction predict the direction and strength of grammatical CLI from the L1 to the L2 and vice versa.
“Early acquisition and comprehension of polysemous words in Croatian”
Ana Werkmann Horvat (University of Osijek)
Polysemy (i.e., words having multiple related meanings) is one of the most productive ways of creating new meanings across languages. Despite being a linguistic universal, the ambiguity of polysemous words can create issues in comprehension. Nevertheless, natural languages use the ability to encode multiple meanings in a single form to their advantage. This tendency appears to be beneficial, as it lessens the burden of remembering and learning new forms (Piantadosi et al., 2012). Since the main generators of polysemy are metaphor and metonymy (Bowdle & Gentner, 2005), the comprehension and acquisition of polysemous words become even more complex due to variation in the dominance and abstractness of meanings. The human need to make communication effortless and to reduce cognitive load, as well as the overwhelming presence of polysemy in everyday language, make polysemy an important part of child language from the very beginning of language acquisition. Furthermore, since new meaning is coded differently in different languages (Sweetser, 1990; Malt, 2010; Traugott, 2014), learning that one form can be related to various conventional meanings represents an important part of learning a language (Floyd, 2020).
While there has been some research on this topic (e.g., Dautriche et al., 2016; Floyd et al., 2020; Werkmann Horvat et al., 2024), the details of which linguistic factors influence the acquisition and comprehension of polysemous words remain largely under-researched. This topic is complex due to the interactions between linguistic factors such as the number of meanings, types of meanings, dominance of meanings, and the frequency of words and meanings. The conventionality of meanings also plays a role, since the figurativeness of meanings is often not recognized. However, conventionality is not guaranteed, and new meanings can emerge at any time. As a result, understanding how children navigate these interacting factors remains an important open question in research on language acquisition.
In this talk, I discuss open research questions related to the acquisition of polysemy in L1 and suggest possible ways of addressing them. I also introduce a newly funded Croatian Science Foundation project POLYACQ that aims to examine the status of the aforementioned linguistic factors and the role they play in language development and early language comprehension. Finally, I consider the implications of this project for the cross-linguistic diversity of current language acquisition research.
“Czeching the Baby Brain: Rhythm, Words, and Speech Sounds in Early Language Development”
Kateřina Chládková (Charles University, Prague)
TBA