Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Linguistic tone is used in automatic lexical activation: an MMN study of Swedish nouns with word accents
There is a Poster PDF for this presentation, but you must be a current member or registered to attend SNL 2023 to view it. Please go to your Account Home page to register.
Poster E73 in Poster Session E, Thursday, October 26, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
Renata Kochančikaitė1, Yury Shtyrov2, Mikael Roll1; 1Lund University, Sweden, 2Aarhus University, Denmark
The Dual Route Model [1] posits that morphologically irregular and frequent words are stored in the lexical memory as full-form representations while regular and infrequent words are decomposed into morphemes. However, to date, it is unknown which role lexically specified tones (e.g. word accents) play in automatic word activation, i.e. whether tones are processed as decomposable components or as an inherent feature of the word’s full form. We used the mismatch-negativity (MMN) paradigm in an oddball ERP experiment to investigate this question. Full-form storage manifests as a lexical MMN, where valid deviant stimuli elicit greater negativity than invalid ones [2]. Decomposition manifests as a syntactic MMN, where valid morpheme combinations presented as deviants elicit smaller negativity than invalid, rule-violating deviants [3]. In our study, we used Swedish words with valid and invalid combinations of tone and suffix. Swedish has two word accents, the low tone (accent 1) and the high tone (accent 2), which are assigned to the stem by the following suffix, e.g. singular definite suffix requires accent 1 on the stem while plural indefinite suffix requires accent 2 on the stem. These compositional properties make it possible to vary tone and validity without changing the actual word meaning (as would be the case in tone languages like Mandarin). We recorded ERP responses in 17 healthy native speakers of Swedish: They watched a silent movie and were instructed to ignore the stimuli played to them binaurally. The stimuli were four versions of the word “krok” – two valid and two invalid combinations of stem tone (high/accent1 or low/accent2) and suffix (sg.def “-en” or pl.indef “-ar”) – krok1-en, krok2-ar, *krok2-en and *krok1-ar. Each word was presented as a standard and as a deviant in four blocks, where validity and suffix varied orthogonally in a fully counterbalanced fashion. Identity MMN was calculated [4] using suffix onsets for time locking of ERP responses. Permutation analysis revealed an early increased negativity (p = 0.033) at 80-180 ms post-onset for valid combinations compared to invalid ones. This negativity had a frontocentral scalp distribution typical of MMN. We estimated cortical sources based on GFP peaks observed in this time window and found that during the early peak (120-140 ms), significant activity occurred in the right auditory cortex, generally associated with suprasegmental and pitch-related processing, while during the late peak (160-180 ms), significant activity occurred in the left mid-posterior temporal lobe, typically associated with retrieval of lexical units from long-term memory. Together, the findings suggest a lexical MMN, i.e. full-form storage, for tones. The results indicate an early and automatic full-form lexical retrieval of valid tone-bearing words. Moreover, this process appears to have two stages that engage different brain areas: first the right auditory cortex where the pitch contour of the word accent is established, and then the left lexical memory area, where the memory trace of the appropriate lexical unit is activated. REFERENCES: [1] https://doi.org/10.3765/bls.v17i0.1624 [2] https://doi.org/10.1016/s1053-8119(03)00261-1 [3] https://doi.org/10.1162/jocn.2007.19.6.971 [4] https://doi.org/10.1016/j.pneurobio.2006.04.004
Topic Areas: Speech Perception, Phonology