Time: Thursday, December 1st, 2016, 10:00–12:00
Venue: C307, Södra huset, Frescati

Professor Valéria Csépe, Brain Imaging Centre, Research Centre for Natural Sciences of the Hungarian Academy of Sciences, Hungary, will be opponent at Hatice Zora's public defence of her PhD thesis "Mapping prosody onto the lexicon" on December 2nd 2016.


Languages differ in the way they organize syllables at a prosodic level. While fixed-stress languages mandatorily assign stress to a certain syllable position, some fixed-stress languages use the word level. In Hungarian and Finish, for example, stress always falls on the initial syllable of a word, in Polish it falls on the penultimate syllable, and in Turkish on the final syllable. Other fixed-stress languages use the phrase level. In French, for example, the final syllable of a phrase is mandatorily stressed. By contrast, free-stress languages, such as Dutch, English, German, Italian, Japanese, or Spanish, allow the position of the stressed syllable to vary between words. Therewith, stress becomes contrastive. It appears that a fixed-stress native language (L1) is associated with a reduced ability to exploit free stress for recognizing words in a second language (L2). Therefore, one of the questions addressed in recent neuroscientific studies is how to test the origin of this difficulty.

Recent neurocognitive research made it clear that native listeners are implicitly using the mandatory stress pattern of a fixed-stress L1. Mismatch responses for speech stimuli that deviate from the fixed native stress pattern were found in Event-Related Potentials (ERPs) of Hungarian listeners (Honbolygó & Csépe, 2013; Honbolygó et al, 2004; Ragó et al, 2015 ); Polish listeners (Domahs et al, 2012); Turkish listeners (Domahs et al, 2013); and French listeners (Astésano et al, 2004; Magne et al, 2007). Those findings implicate that native listeners of fixed-stress languages are able to extract prosodic cues from speech, that they store mandatory prosodic templates of their native language, and that they match prosodic templates with the input.  Native listeners of a free-stress language, however, rapidly use stress for implicit aspects of word recognition. It appears that ERPs reflect facilitated processing of multiple lexical hypotheses more precisely than response latencies (Friedrich et al, 2013). unrelated controls (Friedrich et al., 2008) as well as for targets that were responded to more slowly than unrelated controls (Friedrich et al., 2013).

The presentation will highlight recent neuroscientific data on word stress representation and processing acquired by the author’s research group. These, in comparison with recent publications available, provide a new insight and call for new paradigms as well as for a new theoretical frame that includes the role of competing cues and templates.

This seminar is open to the public and is free of charge.

Hatice Zora