Phonetic recalibration of speech by text

Hearing is believing: Lexically guided perceptual learning is graded to reflect the quantity of evidence in speech input

There is wide variability in the acoustic patterns that are produced for a given linguistic message, including variability that is conditioned on who is speaking. Listeners solve this lack of invariance problem, at least in part, by dynamically modifying the mapping to speech sounds in response to structured variation in the input. Here we test a primary tenet of the ideal adapter framework of speech adaptation, which posits that perceptual learning reflects the incremental updating of cue-sound mappings to incorporate observed evidence with prior beliefs. Our investigation draws on the influential lexically guided perceptual learning paradigm. During an exposure phase, listeners heard a talker who produced fricative energy ambiguous between /ʃ/ and /s/. Lexical context differentially biased interpretation of the ambiguity as either /s/ or /ʃ/, and, across two behavioral experiments (n = 500), we manipulated the quantity of evidence and the consistency of evidence that was provided during exposure. Following exposure, listeners categorized tokens from an ashi - asi continuum to assess learning. The ideal adapter framework was formalized through computational simulations, which predicted that learning would be graded to reflect the quantity, but not the consistency, of the exposure input. These predictions were upheld in human listeners; the magnitude of the learning effect monotonically increased given exposure to four, 10, or 20 critical productions, and there was no evidence that learning differed given consistent versus inconsistent exposure. These results (1) provide support for a primary tenet of the ideal adapter framework, (2) establish quantity of evidence as a key determinant of adaptation in human listeners, and (3) provide critical evidence that lexically guided perceptual learning is not a binary outcome. In doing so, the current work provides foundational knowledge to support theoretical advances that consider perceptual learning as a graded outcome that is tightly linked to input statistics in the speech stream.

Want to quickly adapt to distorted speech and become a better listener? Read lips, not text

When listening to distorted speech, does one become a better listener by looking at the face of the speaker or by reading subtitles that are presented along with the speech signal? We examined this question in two experiments in which we presented participants with spectrally distorted speech (4-channel noise-vocoded speech). During short training sessions, listeners received auditorily distorted words or pseudowords that were partially disambiguated by concurrently presented lipread information or text. After each training session, listeners were tested with new degraded auditory words. Learning effects (based on proportions of correctly identified words) were stronger if listeners had trained with words rather than with pseudowords (a lexical boost), and adding lipread information during training was more effective than adding text (a lipread boost). Moreover, the advantage of lipread speech over text training was also found when participants were tested more than a month later. The current results thus suggest that lipread speech may have surprisingly long-lasting effects on adaptation to distorted speech.

Recognition of Interrupted Speech, Text, and Text-Supplemented Speech by Older Adults: Effect of Interruption Rate

PURPOSE

Studies of speech and text interruption indicate that the interruption rate influences the perceptual information available, from whole words at slow rates to subphonemic cues at faster interruptions rates. In young adults, the benefit obtained from text supplementation of speech may depend on the type of perceptual information available in either modality. Age commonly reduces temporal aspects of information processing, which may influence the benefit older adults obtain from text-supplemented speech across interruption rates.

METHOD

Older adults were tested unimodally and multimodally with spoken and printed sentences that were interrupted by silence or white space at various rates.

RESULTS

Results demonstrate U-shaped performance-rate functions for all modality conditions, with minimal performance around interruption rates of 2-4 Hz. Comparison to previous studies with younger adults indicates overall poorer recognition for interrupted materials by the older adults. However, as a group, older adults can integrate information between the two modalities to a similar degree as younger adults. Individual differences in multimodal integration were noted.

CONCLUSION

Overall, these results indicate that older adults, while demonstrating poorer overall performance in comparison to younger adults, successfully combine distributed partial information across speech and text modalities to facilitate sentence recognition.

Perceptual learning of multiple talkers: Determinants, characteristics, and limitations

Research suggests that listeners simultaneously update talker-specific generative models to reflect structured phonetic variation. Because past investigations exposed listeners to talkers of different genders, it is unknown whether adaptation is talker specific or rather linked to a broader sociophonetic class. Here, we test determinants of listeners' ability to update and apply talker-specific models for speech perception. In six experiments (n = 480), listeners were first exposed to the speech of two talkers who produced ambiguous fricative energy. The talkers' speech was interleaved during exposure, and lexical context differentially biased interpretation of the ambiguity as either /s/ or /ʃ/ for each talker. At test, listeners categorized tokens from ashi-asi continua, one for each talker. Across conditions and experiments, we manipulated exposure quantity, talker gender, blocked versus interleaved talker structure at test, and the degree to which fricative acoustics differed between talkers. When test was blocked by talker, learning was observed for different but not same gender talkers. When talkers were interleaved at test, learning was observed for both different and same gender talkers, which was attenuated when fricative acoustics were constant across talkers. There was no strong evidence to suggest that adaptation to multiple talkers required increased quantity of exposure beyond that required to adapt to a single talker. These results suggest that perceptual learning for speech is achieved via a mechanism that represents a context-dependent, cumulative integration of experience with speech input and identity critical constraints on listeners' ability to dynamically apply multiple generative models in mixed talker listening environments.

Evidence For Selective Adaptation and Recalibration in the Perception of Lexical Stress

Individuals vary in how they produce speech. This variability affects both the segments (vowels and consonants) and the suprasegmental properties of their speech (prosody). Previous literature has demonstrated that listeners can adapt to variability in how different talkers pronounce the segments of speech. This study shows that listeners can also adapt to variability in how talkers produce lexical stress. Experiment 1 demonstrates a selective adaptation effect in lexical stress perception: repeatedly hearing Dutch trochaic words biased perception of a subsequent lexical stress continuum towards more iamb responses. Experiment 2 demonstrates a recalibration effect in lexical stress perception: when ambiguous suprasegmental cues to lexical stress were disambiguated by lexical orthographic context as signaling a trochaic word in an exposure phase, Dutch participants categorized a subsequent test continuum as more trochee-like. Moreover, the selective adaptation and recalibration effects generalized to novel words, not encountered during exposure. Together, the experiments demonstrate that listeners also flexibly adapt to variability in the suprasegmental properties of speech, thus expanding our understanding of the utility of listener adaptation in speech perception. Moreover, the combined outcomes speak for an architecture of spoken word recognition involving abstract prosodic representations at a prelexical level of analysis.

Longitudinal changes in cortical responses to letter-speech sound stimuli in 8-11 year-old children

While children are able to name letters fairly quickly, the automatisation of letter-speech sound mappings continues over the first years of reading development. In the current longitudinal fMRI study, we explored developmental changes in cortical responses to letters and speech sounds across 3 yearly measurements in a sample of 18 8-11 year old children. We employed a text-based recalibration paradigm in which combined exposure to text and ambiguous speech sounds shifts participants' later perception of the ambiguous sounds towards the text. Our results showed that activity of the left superior temporal and lateral inferior precentral gyri followed a non-linear developmental pattern across the measurement sessions. This pattern is reminiscent of previously reported inverted-u-shape developmental trajectories in children's visual cortical responses to text. Our findings suggest that the processing of letters and speech sounds involves non-linear changes in the brain's spoken language network possibly related to progressive automatisation of reading skills.

Boosting lexical support does not enhance lexically guided perceptual learning

A challenge for listeners is to learn the appropriate mapping between acoustics and phonetic categories for an individual talker. Lexically guided perceptual learning (LGPL) studies have shown that listeners can leverage lexical knowledge to guide this process. For instance, listeners learn to interpret ambiguous /s/-/∫/ blends as /s/ if they have previously encountered them in /s/-biased contexts like epi?ode. Here, we examined whether the degree of preceding lexical support might modulate the extent of perceptual learning. In Experiment 1, we first demonstrated that perceptual learning could be obtained in a modified LGPL paradigm where listeners were first biased to interpret ambiguous tokens as one phoneme (e.g., /s/) and then later as another (e.g., /∫/). In subsequent experiments, we tested whether the extent of learning differed depending on whether targets encountered predictive contexts or neutral contexts prior to the auditory target (e.g., epi?ode). Experiment 2 used auditory sentence contexts (e.g., "I love The Walking Dead and eagerly await every new . . ."), whereas Experiment 3 used written sentence contexts. In Experiment 4, participants did not receive sentence contexts but rather saw the written form of the target word (episode) or filler text (＃＃＃＃＃＃＃＃) prior to hearing the critical auditory token. While we consistently observed effects of context on in-the-moment processing of critical words, the size of the learning effect was not modulated by the type of context. We hypothesize that boosting lexical support through preceding context may not strongly influence perceptual learning when ambiguous speech sounds can be identified solely from lexical information. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Cortical responses to letters and ambiguous speech vary with reading skills in dyslexic and typically reading children

•

Text recalibrates ambiguous speech perception in children with and without dyslexia.

•

Dyslexia and poorer reading skills are linked to reduced left fusiform activation.

•

Poorer letter-speech sound matching is linked to higher superior temporal activation.

One of the proposed issues underlying reading difficulties in dyslexia is insufficiently automatized letter-speech sound associations. In the current fMRI experiment, we employ text-based recalibration to investigate letter-speech sound mappings in 8–10 year-old children with and without dyslexia. Here an ambiguous speech sound /a?a/ midway between /aba/ and /ada/ is combined with disambiguating “aba” or “ada” text causing a perceptual shift of the ambiguous /a?a/ sound towards the text (recalibration). This perceptual shift has been found to be reduced in adults but not in children with dyslexia compared to typical readers. Our fMRI results show significantly reduced activation in the left fusiform in dyslexic compared to typical readers, despite comparable behavioural performance. Furthermore, enhanced audio-visual activation within this region was linked to better reading and phonological skills. In contrast, higher activation in bilateral superior temporal cortex was associated with lower letter-speech sound identification fluency. These findings reflect individual differences during the early stages of reading development with reduced recruitment of the left fusiform in dyslexic readers together with an increased involvement of the superior temporal cortex in children with less automatized letter-speech sound associations.

Lexical Information Guides Retuning of Neural Patterns in Perceptual Learning for Speech

A listener's interpretation of a given speech sound can vary probabilistically from moment to moment. Previous experience (i.e., the contexts in which one has encountered an ambiguous sound) can further influence the interpretation of speech, a phenomenon known as perceptual learning for speech. This study used multivoxel pattern analysis to query how neural patterns reflect perceptual learning, leveraging archival fMRI data from a lexically guided perceptual learning study conducted by Myers and Mesite [Myers, E. B., & Mesite, L. M. Neural systems underlying perceptual adjustment to non-standard speech tokens. Journal of Memory and Language, 76, 80-93, 2014]. In that study, participants first heard ambiguous /s/-/∫/ blends in either /s/-biased lexical contexts (epi_ode) or /∫/-biased contexts (refre_ing); subsequently, they performed a phonetic categorization task on tokens from an /asi/-/a∫i/ continuum. In the current work, a classifier was trained to distinguish between phonetic categorization trials in which participants heard unambiguous productions of /s/ and those in which they heard unambiguous productions of /∫/. The classifier was able to generalize this training to ambiguous tokens from the middle of the continuum on the basis of individual participants' trial-by-trial perception. We take these findings as evidence that perceptual learning for speech involves neural recalibration, such that the pattern of activation approximates the perceived category. Exploratory analyses showed that left parietal regions (supramarginal and angular gyri) and right temporal regions (superior, middle, and transverse temporal gyri) were most informative for categorization. Overall, our results inform an understanding of how moment-to-moment variability in speech perception is encoded in the brain.

Interaural recalibration of phonetic categories

Recalibration is a learning process in which perceptual boundaries between speech-sounds adjust through exposure to a supplementary source of information. Using a dichotic-listening methodology, the experiments reported here establish interaural recalibration-in which an ambiguous speech sound in one ear is recalibrated on the basis of a clear sound presented to the other ear. This demonstrates a previously unknown form of recalibration and shows that location-specific recalibration occurs even when people are unaware of location differences between the sounds involved.

Written sentence context effects on acoustic-phonetic perception: fMRI reveals cross-modal semantic-perceptual interactions

This study examines cross-modality effects of a semantically-biased written sentence context on the perception of an acoustically-ambiguous word target identifying neural areas sensitive to interactions between sentential bias and phonetic ambiguity. Of interest is whether the locus or nature of the interactions resembles those previously demonstrated for auditory-only effects. FMRI results show significant interaction effects in right mid-middle temporal gyrus (RmMTG) and bilateral anterior superior temporal gyri (aSTG), regions along the ventral language comprehension stream that map sound onto meaning. These regions are more anterior than those previously identified for auditory-only effects; however, the same cross-over interaction pattern emerged implying similar underlying computations at play. The findings suggest that the mechanisms that integrate information across modality and across sentence and phonetic levels of processing recruit amodal areas where reading and spoken lexical and semantic access converge. Taken together, results support interactive accounts of speech and language processing.

Reading-Induced Shifts in Speech Perception in Dyslexic and Typically Reading Children

One of the proposed mechanisms underlying reading difficulties observed in developmental dyslexia is impaired mapping of visual to auditory speech representations. We investigate these mappings in 20 typically reading and 20 children with dyslexia aged 8–10 years using text-based recalibration. In this paradigm, the pairing of visual text and ambiguous speech sounds shifts (recalibrates) the participant’s perception of the ambiguous speech in subsequent auditory-only post-test trials. Recent research in adults demonstrated this text-induced perceptual shift in typical, but not in dyslexic readers. Our current results instead show significant text-induced recalibration in both typically reading children and children with dyslexia. The strength of this effect was significantly linked to the strength of perceptual adaptation effects in children with dyslexia but not typically reading children. Furthermore, additional analyses in a sample of typically reading children of various reading levels revealed a significant link between recalibration and phoneme categorization. Taken together, our study highlights the importance of considering dynamic developmental changes in reading, letter-speech sound coupling and speech perception when investigating group differences between typical and dyslexic readers.

Maintaining information about speech input during accent adaptation

Speech understanding can be thought of as inferring progressively more abstract representations from a rapidly unfolding signal. One common view of this process holds that lower-level information is discarded as soon as higher-level units have been inferred. However, there is evidence that subcategorical information about speech percepts is not immediately discarded, but is maintained past word boundaries and integrated with subsequent input. Previous evidence for such subcategorical information maintenance has come from paradigms that lack many of the demands typical to everyday language use. We ask whether information maintenance is also possible under more typical constraints, and in particular whether it can facilitate accent adaptation. In a web-based paradigm, participants listened to isolated foreign-accented words in one of three conditions: subtitles were displayed concurrently with the speech, after speech offset, or not displayed at all. The delays between speech offset and subtitle presentation were manipulated. In a subsequent test phase, participants then transcribed novel words in the same accent without the aid of subtitles. We find that subtitles facilitate accent adaptation, even when displayed with a 6 second delay. Listeners thus maintained subcategorical information for sufficiently long to allow it to benefit adaptation. We close by discussing what type of information listeners maintain-subcategorical phonetic information, or just uncertainty about speech categories.

A Selective Deficit in Phonetic Recalibration by Text in Developmental Dyslexia

Upon hearing an ambiguous speech sound, listeners may adjust their perceptual interpretation of the speech input in accordance with contextual information, like accompanying text or lipread speech (i.e., phonetic recalibration; Bertelson et al., 2003). As developmental dyslexia (DD) has been associated with reduced integration of text and speech sounds, we investigated whether this deficit becomes manifest when text is used to induce this type of audiovisual learning. Adults with DD and normal readers were exposed to ambiguous consonants halfway between /aba/ and /ada/ together with text or lipread speech. After this audiovisual exposure phase, they categorized auditory-only ambiguous test sounds. Results showed that individuals with DD, unlike normal readers, did not use text to recalibrate their phoneme categories, whereas their recalibration by lipread speech was spared. Individuals with DD demonstrated similar deficits when ambiguous vowels (halfway between /wIt/ and /wet/) were recalibrated by text. These findings indicate that DD is related to a specific letter-speech sound association deficit that extends over phoneme classes (vowels and consonants), but – as lipreading was spared – does not extend to a more general audio–visual integration deficit. In particular, these results highlight diminished reading-related audiovisual learning in addition to the commonly reported phonological problems in developmental dyslexia.

Multisensory integration of speech sounds with letters vs. visual speech: only visual speech induces the mismatch negativity

Numerous studies have demonstrated that the vision of lip movements can alter the perception of auditory speech syllables (McGurk effect). While there is ample evidence for integration of text and auditory speech, there are only a few studies on the orthographic equivalent of the McGurk effect. Here, we examined whether written text, like visual speech, can induce an illusory change in the perception of speech sounds on both the behavioural and neural levels. In a sound categorization task, we found that both text and visual speech changed the identity of speech sounds from an /aba/-/ada/ continuum, but the size of this audiovisual effect was considerably smaller for text than visual speech. To examine at which level in the information processing hierarchy these multisensory interactions occur, we recorded electroencephalography in an audiovisual mismatch negativity (MMN, a component of the event-related potential reflecting preattentive auditory change detection) paradigm in which deviant text or visual speech was used to induce an illusory change in a sequence of ambiguous sounds halfway between /aba/ and /ada/. We found that only deviant visual speech induced an MMN, but not deviant text, which induced a late P3-like positive potential. These results demonstrate that text has much weaker effects on sound processing than visual speech does, possibly because text has different biological roots than visual speech.

Reading-induced shifts of perceptual speech representations in auditory cortex

Learning to read requires the formation of efficient neural associations between written and spoken language. Whether these associations influence the auditory cortical representation of speech remains unknown. Here we address this question by combining multivariate functional MRI analysis and a newly-developed ‘text-based recalibration’ paradigm. In this paradigm, the pairing of visual text and ambiguous speech sounds shifts (i.e. recalibrates) the perceptual interpretation of the ambiguous sounds in subsequent auditory-only trials. We show that it is possible to retrieve the text-induced perceptual interpretation from fMRI activity patterns in the posterior superior temporal cortex. Furthermore, this auditory cortical region showed significant functional connectivity with the inferior parietal lobe (IPL) during the pairing of text with ambiguous speech. Our findings indicate that reading-related audiovisual mappings can adjust the auditory cortical representation of speech in typically reading adults. Additionally, they suggest the involvement of the IPL in audiovisual and/or higher-order perceptual processes leading to this adjustment. When applied in typical and dyslexic readers of different ages, our text-based recalibration paradigm may reveal relevant aspects of perceptual learning and plasticity during successful and failing reading development.

Cross-modal noise compensation in audiovisual words

Perceiving linguistic input is vital for human functioning, but the process is complicated by the fact that the incoming signal is often degraded. However, humans can compensate for unimodal noise by relying on simultaneous sensory input from another modality. Here, we investigated noise-compensation for spoken and printed words in two experiments. In the first behavioral experiment, we observed that accuracy was modulated by reaction time, bias and sensitivity, but noise compensation could nevertheless be explained via accuracy differences when controlling for RT, bias and sensitivity. In the second experiment, we also measured Event Related Potentials (ERPs) and observed robust electrophysiological correlates of noise compensation starting at around 350 ms after stimulus onset, indicating that noise compensation is most prominent at lexical/semantic processing levels.

Contributions of Letter-Speech Sound Learning and Visual Print Tuning to Reading Improvement: Evidence from Brain Potential and Dyslexia Training Studies

We use a neurocognitive perspective to discuss the contribution of learning letter-speech sound (L-SS) associations and visual specialization in the initial phases of reading in dyslexic children. We review findings from associative learning studies on related cognitive skills important for establishing and consolidating L-SS associations. Then we review brain potential studies, including our own, that yielded two markers associated with reading fluency. Here we show that the marker related to visual specialization (N170) predicts word and pseudoword reading fluency in children who received additional practice in the processing of morphological word structure. Conversely, L-SS integration (indexed by mismatch negativity (MMN)) may only remain important when direct orthography to semantic conversion is not possible, such as in pseudoword reading. In addition, the correlation between these two markers supports the notion that multisensory integration facilitates visual specialization. Finally, we review the role of implicit learning and executive functions in audiovisual learning in dyslexia. Implications for remedial research are discussed and suggestions for future studies are presented.

Watching Subtitled Films Can Help Learning Foreign Languages

Watching English-spoken films with subtitles is becoming increasingly popular throughout the world. One reason for this trend is the assumption that perceptual learning of the sounds of a foreign language, English, will improve perception skills in non-English speakers. Yet, solid proof for this is scarce. In order to test the potential learning effects derived from watching subtitled media, a group of intermediate Spanish students of English as a foreign language watched a 1h-long episode of a TV drama in its original English version, with English, Spanish or no subtitles overlaid. Before and after the viewing, participants took a listening and vocabulary test to evaluate their speech perception and vocabulary acquisition in English, plus a final plot comprehension test. The results of the listening skills tests revealed that after watching the English subtitled version, participants improved these skills significantly more than after watching the Spanish subtitled or no-subtitles versions. The vocabulary test showed no reliable differences between subtitled conditions. Finally, as one could expect, plot comprehension was best under native, Spanish subtitles. These learning effects with just 1 hour exposure might have major implications with longer exposure times.