Font Size: a A A

Neural Encoding Of The Chunk Structure In Speech And Other Complex Sequences

Posted on:2022-04-06Degree:DoctorType:Dissertation
Country:ChinaCandidate:P Q JinFull Text:PDF
GTID:1480306512454284Subject:Electronic information technology and instrumentation
Abstract/Summary:PDF Full Text Request
The living environment of human is full of dynamic sequences,e.g.,speeches and videos,how the brain encodes speech and other dynamic sequences is thus a core question in cognitive neuroscience.Currently,two hypotheses have been proposed: One hypothesizes that the brain recursively combines basic units(e.g.,words)into large chunks(e.g.,phrases),while the other hypothesizes that the brain does not construct the chunk at all but instead processes information across the basic units.Here,this thesis explores(i)whether the brain encodes speech by chunking the sequential information or not,using electroencephalography(EEG),magnetoencephalographic(MEG),and electrooculogram(EOG),and(ii)whether the neural encoding of chunks is soley generated in sensory and language system,or also in attention and motor system as well.Recent EEG and MEG studies provide important evidences for the neural encoding of chunks,it shows that low-frequency neural activity tracking chunk-related structures has been observed in the auditory and language circuits of the brain.It has been argued that,however,such neural activity can also be explained as the neural encoding of word properties.To address this debate,the first study designs novel word sequences and artificial rules,to dissociate word properties from the chunks.MEG responses show that the low-frequency neural activity is phase-locked to the chunk structures,instead of the word properties,indicating that low-frequency neural activity reflects encoding of chunk structures of the speech sequence.The second study explores whether the neural encoding of chunks is soley generated in sensory and language system,or also in attention and motor system as well.We show that during speech comprehension not only the EEG activity is synchronized with the chunk structure,but also the EOG activity,in particular.Critically,such EOG tracking of chunks is independent of visual stimuli.A further eye-tracker experiment shows that such EOG tracking of chunks comes from eye blinks.These results indicate that the motor system is involved in neural encoding of sequential chunk structures of speech,and its activation results in ocular muscle activity.The second study finds that the motor system is engaged in chunk processing,it remains unclear,however,what the role the motor system plays.Previous studies have proposed that the motor system and attention system share parts of neural circuits,therefore the third study hypothesizes that the neural activity in the motor system synchronsized with the chunk structrure may regulate the attentional focus within the chunk.The study directs attentional focus within the chunk and finds that the EOG and blink activities are phase-locked to the attentional focus.A further study shows that such phase-lock character exists in non-speech auditory sequences and visual sequences.These results indicate that the neural activity that synchronizes with the chunk structures in motor system is related to the attentional focus within the chunk.To sum up,this thesis indicates that the brain can generate neural activity that synchronizes with the chunk structure in speech sequences,and such neural activity can not be explained by basic units,such as word properties.Critically,the neural activity is widespreadly generated in the sensory-,language-,and motor-related areas of the brain,and regulates attentional focus within the chunk.Given that the brain indeed encodes chunk structure,these findings provide novel insights into neural mechanism underlying speech comprehension and promote the construction of human-brain-like algorithm for sequence processing.
Keywords/Search Tags:speech, language, cognitive neuroscience, magnetoencephalographic, prosodic boundary
PDF Full Text Request
Related items