Currently funded projects
(limits of) Variability in advance planning (DFG, 320’660 €, 4 years, P.I. A. Bürki)
Started May 2021, this project is part of the second phase of the DFG funded SFB 1287 on Variability in language.
Our project examines variability in advance planning. Using a virtual driving simulator, we investigate whether advance planning (i.e., how much of an utterance is prepared before the onset of the articulation) depends on available cognitive resources and how the interplay between planing and cognitive resources changes with age.
New insights into the speaker’s lexicon: Computational modelling and experimental data (DFG, 239’000 €, 3 years, P.I. A. Bürki).
Started in 2022. Funded by the
This projects investigates
Dominant psycholinguistic models of language production assume that words that are produced more often can be accessed more quickly in subsequent utterances. In these models, each word of the speaker’s language is represented in long term memory. According to these models, the system does not keep track of co-occurrences. If the same utterance is produced on day 1 and on day 2, the processes required to combine them do not remember that these words have been produced together before.
Given that some utterances are more frequent than others, a system that does not keep track of co-occurrences does not seem optimal. (Note also that the ability to track probabilities is precisely what makes systems like chatGPT so efficient). Studies (including from our lab) have shown that more frequent utterances can be prepared faster for production. In one of our studies, participants named pictures in German, using an adjective and a noun. Upon seeing the picture. We asked them to produce frequent adjective-noun combinations and less frequent ones. They were faster to prepare the frequent ones for production (see Jeong et al, 2021).
In this project, we test several explanations of this finding. One of them assumes that co-occurrences are represented by means of weighted connections between word representations. We simulate the data of several computational models (some of them using naive discriminative learning) and thest their predictions with on-line behavioral experiments.
Past projects
Planning utterances across languages
In this project, funded by the DFG-Deutsche Forschungsgemeinshaft (2019-2021) we investigate how speakers plan utterances and in particular, how utterance planning (i.e., how many words are planned before the onset of articulation, in which order are words encoded within planning units). To do so, we capitalize on cross-linguistic differences in how words constrain the pronunciation of one another. See here for updates.
Intra- and inter-speaker variability in word production
In this project, funded by the DFG-Deutsche Forschungsgemeinshaft (2017-2021 “Variability and its limits in the time course of language production” within the CRC 1287 “Limits of Variability in Language; Cognitive, Grammatical, and Social aspects, 287’900 €) we examine inter- and intra-individual variability in the time course of encoding processes during isolated noun production, using ERPs, behavioral, and eye-tracking experiments. See here for updates
Determinants of efficient word learning in a second language
In this project, conducted together with Elsa Spinelli (Université Grenoble-Alpes, France) and Pauline Welby (Aix-Marseille Université, France), we examine the roles of orthography and variability in L2 word learning. This project is partly supported by the University of Potsdam International cooperation program (KoUP, 2018, 4427 €).
Role of talker variability in adult speech processing
This project results from a collaboration between two teams in the SFB 1287 and is funded by the DFG-SFB 1287 (19‘157 €, 2018-2019). Together with Barbara Höhle, Rowena Garcia, Adamantios Gafos, Natalie Boll-Aveytisan, Sylvain Madec, and Marc Hullebus, we examine the role of talker variability in speech processing.
Production-perception link
In this project, funded by the Swiss National Science Foundation (2015-2016, “All speakers are listeners, and vice versa: On the interface between production and comprehension processes” 119’562 CHF) we examined the representations involved in producing and perceiving speech at the phonological and phonetic encoding levels.