The acquisition of number interpretation in German (with Ulrike Domahs) The project aims at investigating the developmental path of German plural from a comprehension perspective under the framework of usage-based morphology. With the help of corpus data, data from visual world paradigm, electrophysiological data and modeling techniques we examine preschool German children.
The role of productivity factors in state-of-the-art inflectional models (with Akhilesh Kakolu Ramarao, Kevin Tang) The project brings together human and machine learning techniques and investigates learning of irregular inflectional patterns. Learning behavior of humans in artificial language learning experiments is compared to learning behavior of neural networks.
Learning S – Duration as a key to morphology? (with Dominic Schmitz) As a natural continuation of the project Final S in English we are investigating whether the distinction of morphological categories based on subphonemic cues is learnable in an artificial language.
Typing S (with Julia Muschalik, Dominic Schmitz) The project explores whether written output modality resembles findings from oral output. Is morphological structure not only leaking through articulation but also through typing?
L1 und L2 Effects im Artificial Language Learning (with Kevin Tang) In this project we introduce a technique on how to control for language-specific effects during artificial language learning using corpus data and modeling techniques. The method can facilitate to disentangle langue-specific effects during artificial language learning experiments, which is useful especially when designing cross-linguistic studies.
Vowel and consonant contrasts in German IDS and ADS (with Charlotte von Kries) In a picture description task we investigate child-directed vs. adult-directed speech in German mothers and fathers. We focus on whether consonants and vowels are hypo- or hyperarticulated in IDS. Hypo- or hyperarticulation shows up depending on the acoustic detail under investigation and varies with age of the child indicating that child-directed speech may be customized more than previously assumed.
Syllabification strategies in German (with Ruben van de Vijver) The project investigates experimentally, how the onset maximization principle and sonority at the syllable boundary influence syllabification behavior in German adults. Results show that onset maximization principle plays a major role and that the sonority at the syllable boundary is of much lower importance to German speakers.
Substative Bias: The relevance for phonetic detail The project aims at investigating the role of very fine phonetic detail on learning phonological alternations. It uses artificial language learning paradigm to compare languages that differ only minimally.
Final S in English: The role of acoustic detail in morphological processing (with Ingo Plag, Dominic Schmitz) In this DFG-funded project we found that final Ss in English, which have different morphological functions, come with a unique duration. Non-morphemic S is longest, followed by suffix S and clitic S is shortest. A carefully designed production study thus confirms previous corpus findings. In two comprehension studies we have shown that the differences also affect comprehension.
Replication studies in phonetic research (with Timo B. Roettger) The project shows how classroom project investigations can contribute to academic discourse, a useful combination of teaching and research. The debate about incomplete neutralization is taken as a testbed for the approach.
Frames in Phonology (with Prof. Ruben van de Vijver) The project explored how frames can be utilized in phonology.
Learners’ Little Helper – Substantive Bias (PhD Dissertation) In three artificial language learning studies with German adults I demonstrated the impact of a substantive bias which is at work during the acquisition of phonological patterns. I investigated how the bias interacts with other factors during the learning process and characterize it as phonetic.
Acquisition of alternations in German (with Prof. Ruben van de Vijver) We studied the acquisition of German voicing and vowel alternations in German children. Comparing productions of 5 year-olds, 7 year-olds and adults we found that initially voicing alternations are less error-prone and more likely to be generalized to pseudowords and vowel alternations are more error-prone and less likely to be generalized. The pattern develops until adulthood when Germans generalize both alternations to the same extent, that is the same extent which is found in the lexicon. We explain this development by the phonetic motivation of the voicing alternation.
Dichotic Listening and Semantic Priming (with Jennifer Aydelott, Prof. Frederic Dick, Maciej Trzaskowski, Prof. Robert Leech) We have examined the cocktail party effect: how does dichotic listening (backward and forward) affect semantic priming?