You are viewing the SNL 2018 Archive Website. For the latest information, see the Current Website.


Poster B13, Thursday, August 16, 3:05 – 4:50 pm, Room 2000AB

Testing the engagement of dorsal stream motor areas during the perception of phonemes, words, and environmental sounds

Kelly Michaelis1, Andrei Medvedev1,2, Peter Turkeltaub1,2,3;1Georgetown University Interdisciplinary Program in Neuroscience, 2Georgetown University Medical Center, Department of Neurology, 3Medstar National Rehabilitation Hospital

Recent research has shown that frontal lobe motor areas typically responsible for speech production are also active during speech perception, especially when speech is noisy or ambiguous. However, the broader literature on speech perception clearly implicates temporal lobe pathways, not motor systems, as the major structures underlying speech perception. While the evidence for motor involvement in perception continues to increase, it has not been fully incorporated into a neurobiological model of speech processing and several critical questions remain unanswered. In the context of the dual stream model (Rauschecker & Scott, 2009), we hypothesize that unambiguous words are processed solely in the ventral stream within the temporal lobe, and speech production pathways in the dorsal stream are flexibly engaged when this ventral stream processing is insufficient to identify a word (i.e. during sublexical perception or high noise conditions) or when additional information, such as seeing the speaker, obligatorily engages motor speech systems. This study provides one of the first tests of motor involvement in word perception, and tests whether motor engagement is specific to perception of pronounceable auditory signals, as theories of motor involvement predict. We tested these hypothesis by examining EEG signatures of motor activity during the perception of auditory phonemes, auditory single words, and audiovisual single words in an adaptive four alternative forced choice task that manipulated signal-to-noise ratios (SNRs) to achieve two levels of difficulty for each stimulus type (80% accuracy and 50% accuracy). To confirm that motor activity is speech-specific, we also examined perception of meaningful auditory environmental sounds in the same task context. Word stimuli consisted of matched lists of unique English words, with near phonological neighbors as foils. Phoneme stimuli were CVC tokens. EEG was recorded using a 128-channel Geodesic sensor net. Data from seven electrodes corresponding to left premotor and motor cortex was preprocessed in EEGLAB and subjected to time-frequency analyses. We tested for differences in sensorimotor-linked μ (8-12Hz) and low beta (13-20Hz) power between stimulus conditions using nonparametric permutation-based paired t-tests. Participants were 10 healthy adults (mean age 24.6, 5 female). Each condition showed a pattern of μ/beta power suppression (indicating greater motor activity) in the pre- and post-stimulus periods, with greater μ/beta suppression in the three speech conditions (auditory words/phonemes, audiovisual words) than for the environmental sounds (p<.01). In line with our hypotheses, the audiovisual and phoneme conditions showed marginally greater pre- and post-stimulus μ/beta suppression than the auditory word condition (p<.05). Our results suggest that motor involvement in perception is specific to speech, and that sublexical processing and visual input preferentially engage dorsal stream motor areas, even when task difficulty is held constant. These findings suggest a role for the motor system in the perception of sublexical speech and in audiovisual contexts, and support a model in which the dorsal stream is flexibly engaged for perception depending on the nature of the speech stimuli. Future analyses will evaluate dorsal stream connectivity by testing for differences in beta band coherence between motor and posterior temporal speech regions.

Topic Area: Perception: Speech Perception and Audiovisual Integration

Back