Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Poster Slams
Neural dynamics of anticipatory coarticulation
Poster C35 in Poster Session C, Friday, October 7, 10:15 am - 12:00 pm EDT, Millennium Hall
Monica Lancheros1, Anna Marczyk1, Marina Laganaro1; 1Faculty of Psychology and Educational Science, University of Geneva
Models of speech motor control propose different encoding processes allowing the transformation of an abstract linguistic message into speech motor codes that will be later articulated. During motor planning, the first post-linguistic speech motor process, motor plans (or speech sound maps in the DIVA model, Gunther, 2016) are suggested either to be retrieved from memory or to be assembled on-line according to the frequency at which a speech target is produced. Upon the retrieval and/or assembling of the motor goals, movements for each sound composing the speech item are sequentially organized and coordinated. In parallel, motor plans are adapted to the phonetic context in which they will be produced, creating the potential for coarticulation of those sounds within a unit. In the present study we are interested in this last aspect. Specifically, we explore the brain dynamics of anticipatory coarticulation, in which the phonetic realization of a given phoneme is influenced by subsequent speech sounds. Assuming that the phenomenon of coarticulation takes place during motor speech encoding processes preceding articulation and considering that the delayed production task permits to target those non-linguistic processing stages, participants were asked to produce a series of disyllabic (CCV1-CV2) non-words in a delayed fashion while their brain activity was recorded with electroencephalography (EEG). Half of the non-words involved some degree of vowel-to-vowel (VtV) coarticulation (i.e. /traki/), while the other half did not (i.e. /traka/). As expected, preliminary acoustic analyses on six participants revealed systematic shifts of formant values in the coarticulation condition (/traki/) as compared to the neutral condition (/traka/). Specifically, the compact V1 /a/ exhibited significantly more diffuse spectrum—as measure by the F1-F2 difference—under the influence of the high frontal V2 /i/ (W = 125, p = .000). Contrarily, no difference between conditions was found in terms of accuracy (94% for both coarticulated and non-coarticulated non-words) and reaction times (618ms and 619ms for non-words requiring or not coarticulation, respectively). Event-related potential (ERP) results will be analysed when completing the group of participants (n=20) in order to determine whether anticipatory VtV coarticulation involve the recruitment of different brain networks between items that exhibit some degree of coarticulation and those which do not.
Topic Areas: Speech Motor Control, Language Production