Representation and selection of determiners with phonological variants.

Abstract

The aim of this study is to contribute to a better understanding of cross-linguistic differences in the time course of determiner selection during language production. In Germanic languages, participants are slower at naming a picture using a determiner + noun utterance (die Katze "the cat") when a superimposed distractor is of a different gender (gender congruency effect). In Romance languages in which the pronunciation of the determiner also depends on the phonology of the next word, there is no such effect. This difference is traditionally assumed to arise because determiners are selected later in Romance languages (late selection hypothesis). It has further been suggested that in a given language, all determiners are either selected late or early (maximum consistency principle). Data on French have challenged these 2 hypotheses by revealing a gender congruency effect when participants name pictures using the definite singular determiner le-la (l' before vowels) and a noun, at positive stimulus onset asynchrony (SOA), that is, when there is a delay between the presentation of the picture and that of the distractor. We examined this finding further and investigated whether it generalizes to the indefinite determiner un-une. Results of 4 picture-word interference experiments reveal that gender congruency effects in French are not restricted to the definite determiner or positive SOAs, but can be hard to detect in experiments which do not account for the variability in reading and naming times across participants and trials. We discuss the implications of these results for the modeling of determiner selection across languages. (PsycINFO Database Record

Assessing effect sizes, variability, and power in the web-based study of language production

“…Whereas the sample for the lab experiment mostly consisted of undergraduate students with backgrounds in linguistics or psychology, the background of participants in the on-line study was more diverse. The two populations might differ in the strategies they used to perform the task, in the extent to which they noticed the similarities between distractors and target words, in their ability to ignore the distractor, or in the relative speed with which they processed the two words (Bürki et al, 2019;Bürki & Madec, 2022). It is important to note, however, that this nding cannot be generalized to other paradigms or effects without further investigation.…”

When words collide: Bayesian meta-analyses of distractor and target properties in the picture–word interference paradigm

“…Before concluding, we highlight the methodological implications of the present findings. While the majority of picture–word interference studies report comparisons involving the same list of distractors counterbalanced across conditions, not all of them do (e.g., Bürki et al, 2019; Finkbeiner & Caramazza, 2006; Foucart et al, 2010; Mahon et al, 2007; Rizio et al, 2017). The finding that naming latencies are influenced by word length and by orthographic neighbourhood (see Supplementary Material 3) calls for a strict control of these properties if different distractor lists are to be used.…”

A cautionary note on the studies using the picture-word interference paradigm: the unwelcome consequences of the random use of “in/animates”

“…The close inspection revealed that from a total of 193 works, only 12 did not mix animate and inanimate stimuli randomly within the target and distractor pairs and the experimental conditions (Ehri, 1976;Guttentag and Haith, 1979;Schnur et al, 2006;Foucart et al, 2010;Muehlhaus et al, 2013;Hwang and Kaiser, 2014;Dank and Deutsch, 2015;DiBattista, 2015;Shin, 2016;Bürki et al, 2019;Deutsch and Dank, 2019;Sá-Leite et al, 2021). Additionally, among these 12, only three of them explicitly stated that they controlled animacy (Foucart et al, 2010;Shin, 2016;Bürki et al, 2019;i.e., "only inanimate stimuli were used") and only one included animacy as a factor to check its impact on the effect sizes (Sá-Leite et al, 2021). Importantly, only Shin (2016) and Sá-Leite et al ( 2021) explicitly mention and discuss animacy theoretically as a potential factor affecting the results.…”

“…However, the most ingenious element of the PWI paradigm, i.e., the use of language comprehension to study language production in its oral and written form (Bonin and Fayol, 2000;Bürki et al, 2019), involves a complex process whose outcomes can be misleading. In the PWI paradigm trials consist of a target that is both a picture and a noun, along with a written or oral distractor noun.…”