Cross-talker generalization in the perception of nonnative speech: A large-scale replication.

Abstract

Speech perception depends on the ability to generalize previously experienced input effectively across talkers. How such cross-talker generalization is achieved has remained an open question. In a seminal study, Bradlow & Bent (2008, henceforth BB08) found that exposure to just 5 min of accented speech can elicit improved recognition that generalizes to an unfamiliar talker of the same accent (N = 70 participants). Cross-talker generalization was, however, only observed after exposure to multiple talkers of the accent, not after exposure to a single accented talker. This contrast between single-and multitalker exposure has been highly influential beyond research on speech perception, suggesting a critical role of exposure variability in learning and generalization. We assess the replicability of BB08's findings in two large-scale perception experiments (total N = 640) including 20 unique combinations of exposure and test talkers. Like BB08, we find robust evidence for cross-talker generalization after multitalker exposure. Unlike BB08, we also find evidence for generalization after single-talker exposure. The degree of cross-talker generalization depends on the specific combination of exposure and test talker. This and other recent findings suggest that exposure to cross-talker variability is not necessary for cross-talker generalization. Variability during exposure might affect generalization only indirectly, mediated through the informativeness of exposure about subsequent speech during test: Similarity-based inferences can explain both the original BB08 and the present findings. We present Bayesian data analysis, including Bayesian meta-analyses and replication tests for generalized linear mixed models. All data, stimuli, and reproducible literate (R markdown) code are shared via OSF.

Distribution-dependent representations in auditory category learning and generalization

“…However, across the three experiments, we demonstrate that II learners can resist changes in F0 for generalization tests performed immediately and one week after training, suggesting that auditory II category representations are not tightly linked to low-level dimensions irrelevant to discriminating category members. Changes in F0 analog changes in talker (e.g., Bradlow and Bent, 2008;Xie et al, 2021). This finding is consistent with previous auditory and speech category learning studies that showed talker-independent category representations emerged at both behavioral and neural levels during training (Bradlow and Bent, 2008;Feng et al, 2019Feng et al, , 2021cJohnson and Sjerps, 2021;Xie et al, 2021).…”

“…Changes in F0 analog changes in talker (e.g., Bradlow and Bent, 2008;Xie et al, 2021). This finding is consistent with previous auditory and speech category learning studies that showed talker-independent category representations emerged at both behavioral and neural levels during training (Bradlow and Bent, 2008;Feng et al, 2019Feng et al, , 2021cJohnson and Sjerps, 2021;Xie et al, 2021). In speech perception, native listeners can maintain perceptual constancy while facing various types of acoustic and perceptual variabilities (Bradlow and Bent, 2008;Feng et al, 2018Feng et al, , 2021a, suggesting exceptional generalization ability to overcome the "lack-of-invariance" challenge.…”

Using Rational Models to Interpret the Results of Experiments on Accent Adaptation

“…It is also shared with the majority of work on talker-specific accent adaptation: both experiments analyzed here employed a single non-native accented talker. There is now evidence that the results of such experiments can depend on the specific talker (for evidence and discussion, see Xie et al, 2021b). Moving forward, the same models employed here for talker-specific adaptation can be used to understand adaptive changes in listeners' perception and categorization following exposure to multiple talkers, or listeners' ability to generalize previously experienced input to unfamiliar talkers (for discussion and model development, see Kleinschmidt and Jaeger, 2015, Part II.…”

Does high variability training improve the learning of non-native phoneme contrasts over low variability training? A replication

“…To mitigate any talker-specific effects in this new sample of talkers (see e.g. Xie et al, 2021), we also counterbalanced the talkers across the taskswhich was not something that was done in the original. Finally, our statistical analyses differed substantially from the original (details below).…”

“…While the original studies did not use counterbalancing but instead used a fixed talker for single-talker tasks, we counterbalanced the assignment of talkers and items to avoid unwanted talker-or item-specific effects (see e.g. Xie et al, 2021). In the HV condition, stimuli from the five talkers were presented on separate days in line with the original studies.…”