Adjusting for speaking rate when perceiving speech in background noise.

Speech perception is a very relevant concept occurring every day. Acoustic context effects such as temporal contrast effects (TCEs) influence perception significantly. For instance, when a faster context sentence is spoken, the participant should perceive the following target word as slower and more like /t/ in “tier”; when a slower context sentence is spoken, the participant should perceive the following target sound as faster and more like /d/ in “deer”. Recent work by Bosker et al. (2020) concluded that selective attention (directing attention to a specific stimulus while ignoring surrounding stimuli) had no effect on TCEs, suggesting they were automatic and low-level. However, their paradigm was not an ideal test; the voices heard contained different talkers with one presented to each ear, making them easy to perceptually separate. Here, the paradigm was designed to eliminate talker variability (acoustic variability among talkers) by using the same male talker speaking one sentence to both ears, two sentences simultaneously to both ears (diotically) or one to each ear (dichotically). Two experiments tested these effects of presentation mode on TCEs. In each experiment, TCE magnitudes were similar across presentation modes. These results are consistent with Bosker et al.’s (2020) claims of TCEs being automatic and low-level. Potential neural mechanisms contributing to TCEs are discussed