Purpose: Auditory–motor synchronization, the ability to align motor output with external auditory input, reflects neural coupling between auditory and motor cortices and supports tracking of the speech envelope and linguistic cues, essential for speech perception and production. In aphasia, disrupted neural coupling may impair synchronization and affect speech and language processing. While domain-general synchronization (i.e., tapping to metronome or musical stimuli) has been studied previously, its relationship to speech-specific synchronization (i.e., speech-to-speech) remains underexplored, particularly in the context of aphasia-related impairments. This study examined the feasibility of assessing auditory–motor synchronization in individuals with aphasia and with advancing age, particularly focusing on individual variability. Method: Sixteen individuals with aphasia and 32 age-matched neurotypical controls completed both tapping and speech-to-speech synchronization tasks. Synchronization abilities and their relations to auditory, musical, and linguistic prosodic abilities were analyzed using linked models. Results: Despite the speech-specific demands for individuals with aphasia, adaptations to data preprocessing and cleaning retained feasibility of the speech-sync task in our sample. Individuals with aphasia showed reduced synchronization abilities across both domain-general and speech-specific synchronization compared to neurotypical controls. In both groups, better auditory, musical, and linguistic prosodic abilities were associated with enhanced speech-specific synchronization. Conclusions: These findings highlight the relevance of auditory–motor synchronization for impaired and unimpaired speech and language processing. Results underscore the importance of accounting for individual variability, considering both challenges and resources in synchronization abilities, potentially informing individualized therapeutic interventions in aphasia rehabilitation. Supplemental Material: https://doi.org/10.23641/asha.34914330
Purpose: Second-language (L2) speech perception is strongly influenced by prior language experience, but the neural mechanisms supporting this process remain incompletely understood. This study examined how language experience modulates the pre-attentive encoding of acoustic and phonological information in Mandarin tone perception and whether these cues are processed serially or in parallel. Method: Twenty native Mandarin speakers and 20 novice Mandarin learners with a tonal first language (L1: Thai) completed a behavioral tone identification task and a passive listening multifeature oddball event-related potential paradigm. Stimuli were drawn from an 11-step Tone 1–Tone 2 continuum with Stimulus 5 (S5) as the standard and Stimuli 1, 3, 7, and 9 as deviants. Deviants varied in fundamental frequency distance (large vs. small acoustic distance from S5) and category status (across- vs. within-category) relative to each participant's own phonetic boundary for the continuum. Temporal principal component analysis was applied to separate early and late mismatch negativity (MMN) components. Results: Behavioral identification did not differ across groups. However, MMN results revealed that both acoustic and phonological factors modulated early and late responses, supporting parallel processing. Native Mandarin speakers exhibited stronger and more sustained phonological effects across both time windows bilaterally. Nonnative speakers showed greater acoustic sensitivity, particularly over the right-hemisphere electrodes, and weaker phonological effects, with reduction most evident over the left-hemisphere electrodes. Conclusions: The findings show that Mandarin tone perception involves parallel extraction of acoustic and phonological information at the pre-attentive level, with language experience influencing their relative weighting. The neural–behavioral dissociation suggests that tonal L1 experience supports perceptual decisions without fully automatized phonological representations transferred to L2. Supplemental Material: https://doi.org/10.23641/asha.34925781
Purpose: Paired digit stimuli can effectively measure speech recognition threshold (SRT) in typically developing English-speaking children and in adults who are learning English as an additional language (EAL). We evaluated paired digits (digit SRT [D-SRT]) and pediatric words (word SRT [W-SRT]) for determining SRT in children with EAL backgrounds and children with language disorders (LDs) compared to a control group of typically developing, monolingual English-speaking children. SRT efficacy was assessed by comparison to pure-tone average (PTA) thresholds. Method: Participants ( N = 74) were normal-hearing children, with a mean age of 7 years, in three groups. Sixteen children had LD; 21 were from EAL backgrounds; and 37 were age- and gender-matched, typically developing, monolingual English-speaking peers. Experimental sessions included the audiometric protocol and speech-language assessments. Results: There were no statistically significant group differences for PTA and W-SRT thresholds; however, mean D-SRT was significantly higher (poorer) in children with LD compared to the EAL and control groups. For EAL children, D-SRT was correlated with PTA and W-SRT with excellent predictive value. D-SRT reliably measured the hearing threshold for speech in EAL and typically developing, monolingual English-speaking children; however, for children with LD, PTA and W-SRT revealed the strongest correlation. Conclusions: Digit pairs were effective in measuring SRT in both EAL and typically developing, monolingual English-speaking children. D-SRT, therefore, can fill a need for audiologists serving children with limited English familiarity. In children with LD, however, popular pediatric word stimuli produced a closer match to PTA than digit pairs, highlighting the predictive nature of popular pediatric words for English speakers.
Purpose: This longitudinal study examined the development of listening and spoken language skills in 322 deaf and hard of hearing children who participated in a multidisciplinary early intervention program. Method: Listening skill acquisition was assessed using the Functional Listening Index–Paediatric (3,911 assessments) in children from 0 to 6 years of age accessing the program between 2011 and 2022 and was analyzed against standardized language scores (1,015 assessments) measured using the Preschool Language Scale–Fourth/Fifth Edition and the Clinical Evaluation of Language Fundamentals Preschool–Third/Fourth Edition. Listening scores were compared with previously reported normative data for age-matched typically hearing controls. Results: Results indicated that listening scores significantly correlated with language scores with similar development trajectories. When listening and language scores were normalized using z scores to enable accurate comparison, listening scores proved more sensitive to the effects of hearing loss. Time in the early intervention program, diagnosed additional needs, language background of the family, and age at commencement of intervention were found to be significant factors influencing outcomes. Commencement of early intervention before 5 months of age resulted in the highest scores when compared with 20 months after intervention. There was no interaction found for mean bilateral hearing thresholds on listening or language outcomes, indicating that the range of audiological devices available is providing appropriate levels of access to sound for children to achieve similar results regardless of type of device required. Conclusion: These results suggest that listening measures can provide complementary information on skill development that standard language scores do not incorporate.
Purpose: Patient speech samples provide insights into cognitive and linguistic status, yet their utilization has been limited by labor-intensive manual transcription. Advances in automated speech recognition (ASR) with large language models (LLMs) potentially save time, but performance on impaired speech, such as due to poststroke aphasia, remains understudied. Method: We transcribed 4,535 trials of the Philadelphia Naming Test (PNT) administered to 26 people with aphasia (PWA) using OpenAI's Whisper speech-to-text LLM. These ASR transcripts were compared with gold-standard transcriptions by an experienced speech-language pathologist (SLP). Clinician–LLM match or mismatch in transcription was analyzed across speech-error types (semantic vs. phonemic), patient severities, and word frequency. Results: For each 175-trial PNT, transcriptions by SLP required 180 ± 40 min versus 2.3 ± 0.5 min by Whisper. Compared to > 95% match rate with SLP transcripts for trained humans, Whisper achieved an overall match rate of 75.52%, subdivided as 88.83% match on correct PNT trials, 92.22% on semantic errors, and 23.05% for trials with phonemic errors by PWA. Western Aphasia Battery–Aphasia Quotient scores correlated with match rate ( R 2 = .196, p = .023) and hallucination rate ( R 2 = .291, p = .0048). Low-frequency words showed higher mismatch than medium- or high-frequency words (analysis of variance F = 14.26, p < .001). Conclusions: Whisper reliably transcribed semantic but not phonological errors, suggesting automation is best for fluent aphasia types with less phonemic distortions. This study benchmarks LLM transcription for impaired speech, offering a scalable alternative to manual transcription that may streamline clinical and research workflows. Supplemental Material: https://doi.org/10.23641/asha.34790292
