Benjamin V. Tucker

This study uses a corpus of sociolinguistic interview speech from Western Canadian English to examine individual-level patterns in random forest classification models for fricatives. The models include 23 spectral, durational, and amplitudinal measures commonly used for group-level analyses. Results reveal that measures such as midpoint standard deviation, spectral peak frequency, peak power, and segment duration can capture meaningful distinctions within individual speakers, with some acoustic cues playing a greater role in distinguishing fricatives within speakers than across them. Limiting models to the ten most important predictors slightly increased classification error, suggesting the importance of broader acoustic information.

Article Frontmatter was published on February 1, 2026 in the journal Phonetica (volume 83, issue 1).

Purpose: This study aims to examine how language background and grade level influence the voiceless sibilant fricative production of two groups of children enrolled in a Mandarin–English bilingual school in Canada. One group of children was exposed to Mandarin at home as heritage language (HL), and the other group had little to no exposure to Mandarin until they started learning it as a second language (L2) at school. Method: Eighty-two children in Grades 1, 3, or 5 as well as 12 teachers engaged in picture-naming tasks in both English and Mandarin. Their speech production of voiceless sibilant fricatives was analyzed using both transcription and acoustic methods. The center of gravity (CoG) and F2 onset of the vowel following the fricatives were measured in the acoustic analysis. Results: Both groups of children exhibited high accuracy rates in producing English /s/ and /ʃ/ and were able to distinguish the two English fricatives based on the CoG regardless of their grade level. For Mandarin fricatives, children in the higher grades achieved higher accuracy rates. Mandarin HL speakers had a higher accuracy rate than Mandarin L2 speakers, with their productions more closely resembling teachers' productions. The CoG distinguished all three Mandarin fricatives regardless of grade level and language background. However, significant differences in F2 onset were only observed in the HL group. Conclusions: The finding that unshared fricatives were more difficult than shared ones, along with the significant impact of language background—but not grade level—on the accuracy of fricatives unique to Mandarin, suggests that students, even in higher grades, can benefit from more explicit teaching. The developmental patterns and error trends identified in this study can serve as a reference for speech assessment and aid in planning treatment or instructional sessions.

We present SingleMALD, a large-scale auditory lexical decision study in English with a fully crossed design. SingleMALD is freely available and includes over 2 million trials in which 40 native speakers of English responded to over 26,000 different words and over 9000 different pseudowords, each in 67 balanced sessions. SingleMALD features a large number of responses per stimulus, but a smaller number of participants, thus complementing the Massive Auditory Lexical Decision (MALD) dataset which features many listeners but fewer responses per stimulus. In the present report, we also use SingleMALD data to explore how extensive testing affects performance in the auditory lexical decision task. SingleMALD participants show signs of favoring speed over accuracy as the sessions unfold. Additionally, we find that the relationship between participant performance and two lexical predictors – word frequency and phonological neighborhood density – changes as sessions unfold, especially for certain lexical predictor values. We note that none of the changes are drastic, indicating that data collected from participants that have been extensively tested is usable, although we recommend accounting for participant experience with the task when performing statistical analyses of the data.

Article Welcome was published on February 4, 2025 in the journal Phonetica (volume 0, issue 0).

During conversation, speakers produce reduced speech, and this can create homophones: ‘we were’ and ‘we’re’ can both be realized as [ɚ], and ‘he was’ and ‘he’s’ can be realized as [ɨz]. We investigate the types of information non-native listeners (Dutch L1-English L2) use to perceive the tense of such verbs, making comparisons with previous results from native listeners. The Dutch listeners were almost as successful as natives (average percentage correct for ‘is’/’was’ in the most accurate condition: 81% for Dutch, 88% for natives). The two groups showed many of the same patterns, indicating that both make strong use of whatever acoustic information is available in the signal, even if it is heavily reduced. The Dutch listeners showed one crucial difference: a minimal amount of context around the target, just enough to signal speech rate, did not help Dutch listeners to recover the longer forms, i.e., was/were, from reduced pronunciations. Only the full utterance context (containing syntactic/semantic information such as ‘yesterday’ or another tensed verb) helped Dutch listeners to recover from reduction. They were not able to adjust their criteria based on the surrounding speech rate as native listeners were. This study contributes to understanding how L2 learners parse information from spontaneous speech in a World Englishes setting with inputs from multiple dialects.

Given an orthographic transcription, forced alignment systems automatically determine boundaries between segments in speech, facilitating the use of large corpora. In the present paper, we introduce a neural network-based forced alignment system, the Mason-Alberta Phonetic Segmenter (MAPS). MAPS serves as a testbed for two possible improvements we pursue for forced alignment systems. The first is treating the acoustic model as a tagger, rather than a classifier, motivated by the common understanding that segments are not truly discrete and often overlap. The second is an interpolation technique to allow more precise boundaries than the typical 10 ms limit in modern systems. During testing, all system configurations we trained significantly outperformed the state-of-the-art Montreal Forced Aligner in the 10 ms boundary placement tolerance threshold. The greatest difference achieved was a 28.13 % relative performance increase. The Montreal Forced Aligner began to slightly outperform our models at around a 30 ms tolerance. We also reflect on the training process for acoustic modeling in forced alignment, highlighting how the output targets for these models do not match phoneticians’ conception of similarity between phones and that reconciling this tension may require rethinking the task and output targets or how speech itself should be segmented.

It is well-known that children with expressive communication difficulties have the right to communicate, but they should also have the right to do so in whichever language they choose, with a voice that closely matches their age, gender, and dialect. This study aimed to develop naturalistic synthetic child speech, matching the vocal identity of three children with expressive communication difficulties, using Tacotron 2, for three under-resourced South African languages, namely South African English (SAE), Afrikaans, and isiXhosa. Due to the scarcity of child speech corpora, 2 hours of child speech data per child was collected from three 11- to 12-year-old children. Two adult models were used to "warm start" the child speech synthesis. To determine the naturalness of the synthetic voices, 124 listeners participated in a mean opinion score survey (Likert Score) and optionally gave qualitative feedback. Despite limited training data used in this study, we successfully developed a synthesized child voice of adequate quality in each language. This study highlights that with recent technological advancements, it is possible to develop synthetic child speech that matches the vocal identity of a child with expressive communication difficulties in different under-resourced languages.

Research on speech communication investigates how speech is produced and how it is perceived. Researchers and practitioners who engage with speech communication come from various disciplines, such as audiology, computer science, engineering, linguistics, psychology, and speech pathology. Speech communication researchers seek to understand the acoustic signal as produced by a speaker or perceived by a listener. Research under the speech communication technical committee overlaps with other technical committees. In this presentation, I provide brief illustrations of various topics within speech communication, focusing on how people perceive and produce speech to communicate in the world's languages.

This study constitutes an investigation into the acoustic variability of intervocalic alveolar taps in a corpus of spontaneous speech from Madrid, Spain. Substantial variability was documented in this segment, with highly reduced variants constituting roughly half of all tokens during spectrographic inspection. In addition to qualitative documentation, the intensity difference between the tap and surrounding vowels was measured. Changes in this intensity difference were statistically modeled using Bayesian finite mixture models containing lexical and phonetic predictors. Model comparisons indicate predictive performance is improved when we assume two latent categories, interpreted as two pronunciation variants for the Spanish tap. In interpreting the model, predictors were more often related to categorical changes in which pronunciation variant was produced than to gradient intensity changes within each tap type. Variability in tap production was found according to lexical frequency, speech rate, and phonetic environment. These results underscore the importance of evaluating model fit to the data as well as what researchers modeling phonetic variability can gain in moving past linear models when they do not adequately fit the observed data.