Ongoing projects

Decoding phonetic cue weighting from EEG responses to naturalistic speech
(Advised by Dr. Lisa Sanders)

L2 learners’ cue weighting may be suboptimal due to interference from their native language. While cue weighting is adjustable, a prerequisite for targeted phonetic training is a thorough understanding of native and non-native listeners’ cue weighting.

To facilitate research on cue weighting, this project develops a measure based on representational similarity analysis of neural activities and speech stimuli, which has advantages over the typical behavioral methods:

separating perceptual processes from post-perceptual influences
maintaining ecologically validity in naturalistic listening (e.g., audiobooks)
enhancing temporal resolution for research on real-time cue integration

Phonetic cue weighting under divided attention
(Advised by Dr. Alexandra Jesse and Dr. Lisa Sanders)

Much of our knowledge of speech perception came from experiments conducted in highly controlled environments. However, in real life, listeners’ attention is not always entirely dedicated to speech, differing from the typical setup of perceptual experiments.

This study aims to investigate how divided attention affects speech perception, such as in some classroom scenarios. We ask participants to perform a secondary task and a speech perception task to answer the following questions:

Does distraction interfere with the early perceptual processing of speech or the retention of speech percepts in memory?
What characteristics of distractors influence the nature and extent of such interference?

Testing the roles of post-consonantal F0 perturbation and noise in tonogenesis
(Advised by Dr. Lisa Sanders)

F0 perturbation caused by adjacent consonants is thought to be a possible trigger of tonogenesis, i.e., the emergence of tone in initially non-tonal languages. During this process, the F0 difference presumably becomes larger in magnitude and longer in duration, making it perceptually more salient. Thus, listeners may possibly increase the perceptual weight of F0 while decreasing the weight of consonant phonation.

Using English stop voicing as a testing case, this study examines cue weightings of VOT and F0. Specifically, we ask:

Does the temporal extent of F0 perturbation contribute to voicing perception?
As F0 is more robust to noise, does the presence of noise modulate the F0 perturbation effects?