Amlap | Bilbao | September 2nd 2016 Tracking lexical garden-path resolution with MEG: Phonological commitment and sensitivity to subphonemic detail are independent Laura Gwilliams, Tal Linzen, Kyriaki Neophytou, David Poeppel & Alec Marantz Laura Gwilliams | NYU | Sept 2nd 1
Challenge of Speech Comprehension Speech is an inherently noisy and ambiguous signal • To fluently derive meaning, listeners must integrate top- • down contextual information to guide their interpretation Top-down input occurring after an acoustic signal can be • integrated to affect the perception of earlier sounds (Connine et al., 1991; Samuel, 1981; Szostak & Pitt, 2013; Warren & Sherman, 1974) Laura Gwilliams | NYU | Sept 2nd 2
Challenge of Speech Comprehension (this is a parakeet) Categorical Perception 1.00 p ? a r ə k i: t 0.75 % /p/ responses 0.50 b ? a r ə k e ɪ d 0.25 0.00 1 2 3 4 5 6 7 8 9 10 11 Continuum (this is a barricade) “ P oint o f D isambiguation” (POD) Laura Gwilliams | NYU | Sept 2nd 3
Today’s Questions • How does the auditory system respond to phonological ambiguity? • What mechanism(s) uphold subsequent context integration to resolve bottom-up ambiguity? • Is there a time-limit on how late subsequent context can be received? Laura Gwilliams | NYU | Sept 2nd 4
Technique: MEG 5 Laura Gwilliams | NYU | Sept 2nd
Design & Materials Selection ~ Continuum 1.00 /p/ 0.75 Selection Proportion a r ə k i: t Ambiguity 65 x T ? 0.50 a r ə k e ɪ d 0.25 /b/ 0.00 0 25 50 75 100 Continuum (see also McMurray et al., 2009) Laura Gwilliams | NYU | Sept 2nd 6
Experiment 1: Design & Materials /pa/ /ba/ 0 25 50 75 100 VOT (31 pairs) {p-b, t-d, k-g} and PoA (22 pairs) {t-k, p-t} • Laura Gwilliams | NYU | Sept 2nd 7
Experiment 2: Design & Materials /pa/ /ba/ 0 25 50 75 100 parakeet barakeet 0 25 50 75 100 Laura Gwilliams | NYU | Sept 2nd 8
Experiment 2: Design & Materials parakeet barakeet 0 25 50 75 100 50 100 75 25 0 parricade barricade Point of Disambiguation (PoD) ranged 3-8 phonemes / 150-750 ms • Laura Gwilliams | NYU | Sept 2nd 9
Procedure & Analysis Data Collection & Preprocessing parakeet barakeet 25 native English participants in each experiment • 0 25 50 75 100 Low pass filter online at 200 Hz • MNE-Python used for preprocessing and source localisation • 50 100 75 25 0 Analysis Multiple regression spatio-temporal cluster test ( Eelbrain ) • parricade barricade (for more information see Gwilliams, Lewis & Marantz, 2016 - NeuroImage) • Point of Disambiguation (PoD) ranged 3-8 phonemes / 150-750 ms Corrected for multiple comparisons (Maris & Oostenveld, 2007) • • Laura Gwilliams | NYU | Sept 2nd 10
Today’s Questions How does the auditory system respond to phonological ambiguity? Laura Gwilliams | NYU | Sept 2nd 11
Ambiguity at Onset Experiment 1: Syllables Experiment 2: Words • Time-window: 0-200 ms after syllable/word onset • Region: Heschl’s Gyrus and superior temporal gyrus bilaterally Left Heschls Gyrus: Word Onset Left Heschls Gyrus: Syllable Onset 2 2.5 0 dSPM dSPM 0.0 − 2 − 2.5 − 4 − 5.0 0 250 500 750 1000 0 250 500 750 1000 Time (ms) Time (ms) /p/ /a/ /r/ / ə / /k/ /i:/ /t/ 0 50 100 150 Laura Gwilliams | NYU | Sept 2nd 12
Ambiguity at Onset Experiment 1: Syllables Experiment 2: Words PoA: 40:60 ms VOT: 80:100 ms PoA: 40:80 ms VOT: 40:80 ms PoA, Word Onset: 40 − 60 ms VOT, Word Onset: 80 − 100 ms 0.6 0.6 0.6 0.4 0.4 dSPM dSPM 0.4 0.4 dSPM dSPM 0.2 0.2 0.2 0.2 0.0 0.0 0.0 0.0 0 25 50 75 100 0 25 50 75 100 0 25 50 75 100 0 25 50 75 100 Continuum Continuum Continuum Continuum Left Heschls Gyrus: Word Onset Left Heschls Gyrus: Syllable Onset 2 2.5 0 dSPM dSPM 0.0 − 2 − 2.5 − 4 − 5.0 0 250 500 750 1000 0 250 500 750 1000 Time (ms) Time (ms) /p/ /a/ /r/ / ə / /k/ /i:/ /t/ 0 50 100 150 Laura Gwilliams | NYU | Sept 2nd 13
Ambiguity at Onset Experiment 1: Syllables Experiment 2: Words PoA: 100:130 ms VOT: 100:130 ms PoA: 100:130 ms VOT: 100:130 ms PoA, Word Onset: 100 − 130 ms VOT, Word Onset: 100 − 130 ms PoA, Syllable Onset: 100 − 130 ms VOT, Syllable Onset: 100 − 130 ms 0.0 0.0 0.0 0.0 − 0.3 − 0.1 − 0.1 − 0.5 dSPM dSPM dSPM dSPM − 0.6 − 0.2 − 0.2 − 1.0 − 0.3 − 0.9 − 0.3 − 0.4 Voiced Voiceless Bilabial Labio − Velar Voiced Voiceless Bilabial Labio − Velar dental dental Left Heschls Gyrus: Word Onset Left Heschls Gyrus: Syllable Onset 2 2.5 0 dSPM dSPM 0.0 − 2 − 2.5 − 4 − 5.0 0 250 500 750 1000 0 250 500 750 1000 Time (ms) Time (ms) /p/ /a/ /r/ / ə / /k/ /i:/ /t/ 0 50 100 150 Laura Gwilliams | NYU | Sept 2nd 14
Today’s Questions What mechanism(s) uphold subsequent context integration to resolve bottom-up ambiguity? Laura Gwilliams | NYU | Sept 2nd 15
Ambiguity at POD • Time-window: 0-200 ms after POD onset • Region: Heschl’s Gyrus and superior temporal gyrus bilaterally Left Heschls Gyrus: POD Onset 2 0 dSPM − 2 − 4 − 500 − 250 0 250 500 Time (ms) /p/ /a/ /r/ / ə / /k/ /i:/ /t/ 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 ms Laura Gwilliams | NYU | Sept 2nd 16
Ambiguity at POD PoA: 40:80 ms VOT: 40:80 ms PoA, POD Onset: Heschl's Gyrus: 50 − 100 ms VOT, POD Onset: Heschl's Gyrus: 50 − 100 ms 0.00 0.0 Late Disambiguation (> 450ms) − 0.25 PoA VOT − 0.3 − 0.50 dSPM dSPM − 0.6 − 0.75 0.00 − 0.9 − 1.00 − 0.05 0 25 50 75 100 0 25 50 75 100 dSPM Continuum Continuum − 0.10 Left Heschls Gyrus: POD Onset 2 0 − 0.15 dSPM − 2 50 100 50 100 Continuum − 4 − 500 − 250 0 250 500 Time (ms) /p/ /a/ /r/ / ə / /k/ /i:/ /t/ 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 ms Laura Gwilliams | NYU | Sept 2nd 17
Ambiguity at POD PoA: 40:80 ms VOT: 40:80 ms PoA, Syllable Onset: 100 − 130 ms PoA: 40:80 ms VOT, Syllable Onset: 100 − 130 ms VOT: 40:80 ms PoA, POD Onset: Heschl's Gyrus: 50 − 100 ms VOT, POD Onset: Heschl's Gyrus: 50 − 100 ms 0.00 0.00 0.00 0.0 − 0.25 − 0.05 − 0.3 − 0.05 − 0.50 dSPM dSPM dSPM dSPM − 0.6 − 0.10 − 0.10 − 0.75 − 0.9 − 0.15 − 1.00 − 0.15 0 25 50 75 100 0 25 50 75 100 Bilabial Labio − Velar Voiced Voiceless Continuum Continuum dental Left Heschls Gyrus: POD Onset 2 Onset phoneme is “heard 0 again” at disambiguation dSPM − 2 − 4 − 500 − 250 0 250 500 Time (ms) /p/ /a/ /r/ / ə / /k/ /i:/ /t/ 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 ms Laura Gwilliams | NYU | Sept 2nd 18
Today’s Questions Is there a time-limit on how late subsequent context can be received? Laura Gwilliams | NYU | Sept 2nd 19
Commitment Before POD parakeet barricade band pan barrister paranoid back paddle ballet past /pa/ cohort /ba/ cohort (75% /p/) (25% /b/) /p/ /a/ /r/ / ə / /k/ /i:/ /t/ 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 ms Laura Gwilliams | NYU | Sept 2nd 20
Commitment Before POD • Time-window: 0-400 ms parakeet pan after POD onset paranoid • Region: Heschl’s Gyrus, paddle superior and middle past temporal gyrus bilaterally /pa/ cohort (75% /p/) /p/ /a/ /r/ / ə / /k/ /i:/ /t/ 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 ms Laura Gwilliams | NYU | Sept 2nd 21
Commitment Before POD Lexicality*Latency parakeet 0.975 ** 200:230 ms pan paranoid 0.950 paddle 0.925 dSPM past 0.900 /pa/ cohort 0.875 (75% /p/) 0.850 Early Early Late Late Word Non − Word Word Non − Word /p/ /a/ /r/ / ə / /k/ /i:/ /t/ 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 ms Laura Gwilliams | NYU | Sept 2nd 22
Conclusion • Sensitivity to phoneme ambiguity ~50 ms 0.6 0.4 after onset in primary auditory cortex 0.2 0.0 0 25 50 75 100 • Subphonemic detail is maintained over long 0.0 time-scales (+700 ms) and re-evoked at 0.3 0.6 point of disambiguation 0.9 0 25 50 75 100 • Phonological commitment resolves ~450 ms ** after phoneme onset in superior temporal 0.950 0.925 gyrus 0.900 0.875 0.850 Early Early Late Late Laura Gwilliams | NYU | Sept 2nd 23
Take Home Message maintain subphonemic detail across long time-scales p b b a r ə k i: t p p commit to most likely phonological interpretation ~450 ms Laura Gwilliams | NYU | Sept 2nd 24
laura.gwilliams@nyu.edu @GwilliamsL With big thanks to: • My supervisors, Alec Marantz & David Poeppel • Everyone in the Neuroscience of Language Lab and Poeppel Lab Funding: G1001 Abu Dhabi Institute
laura.gwilliams@nyu.edu @GwilliamsL Eskerrik Asko!
Recommend
More recommend
