The study, led by Seung-Eun Kim of Yonsei University together with Matthew Goldrick, Joseph Keshet, and Ann R. Bradlow of Northwestern University and the Technion-Israel Institute of Technology, tackled a puzzle that has dogged the field for years. When native English listeners adapt to one non-native talker, does that adaptation help them understand a different non-native talker they have never heard before? The literature is inconsistent. Some experiments report robust generalization from a single training voice; others find that the learning stays stubbornly locked to the specific talker. The new work was designed to adjudicate between two competing theoretical accounts, and to do so with an unusually rigorous, data-driven methodology.

The first hypothesis centers on intelligibility. Perhaps, the researchers reasoned, generalization is driven by how easy the training talker is to understand. A highly intelligible talker might provide cleaner, more reliable evidence about how English sounds are realized in non-native speech, allowing the listener’s perceptual system to extract general rules rather than talker-specific quirks. Under this account, training with a clear, easily understood accented speaker should produce stronger benefits for understanding other accented speakers than training with a hard-to-understand one. The second hypothesis centers on acoustic similarity. Perhaps generalization depends on how closely the training talker resembles the test talker in the physical details of their speech. If the perceptual system tunes to specific acoustic properties, then a training voice that sounds similar to a test voice should yield better transfer than a dissimilar pairing. This account echoes earlier findings by Xie and Myers, who showed that acoustic similarity between talkers constrains how far foreign-accent adaptation spreads.

Testing these hypotheses properly required something the field has rarely had: precise, quantitative measurements of both intelligibility and acoustic similarity for a large pool of second-language talkers. The team drew on the ALLSSTAR archive, a Northwestern University repository of recordings from more than 100 speakers of English as a second language, each of whom had been assessed with empirical intelligibility measures. For the similarity dimension, the researchers turned to a cutting-edge computational tool: self-supervised speech models, neural networks of the same family as HuBERT that learn rich representations of speech sound by being trained on massive amounts of audio without labels. Distances between talkers in the representational space of such models have been shown in prior work by the same group to capture perceptually meaningful acoustic similarity, providing a principled way to measure how alike any two speakers sound to a machine that has learned the statistics of human speech.

From this large dataset, the researchers constructed twenty training-test talker pairs, deliberately engineered so that the pairs spanned a wide range of intelligibility levels and acoustic similarity values. Each pair was then evaluated by ten native English listeners in a sentence-in-noise transcription task. The design was an exposure-test format: listeners first heard sentences produced by the training talker, giving their perceptual system a chance to adapt, and then transcribed sentences from the test talker, a different accented speaker they had not encountered before. Accuracy on those test sentences, measured as the proportion of words correctly transcribed against the noise, served as the index of cross-talker generalization. The inclusion of background noise was not incidental. Noise is one of the most ecologically common and cognitively demanding listening conditions, and prior work has shown that understanding accented speech in noise is substantially harder than in quiet, making any adaptation benefit both more meaningful and more measurable.

The headline result was genuinely encouraging: most of the twenty talker pairs showed cross-talker generalization. In other words, brief exposure to a single second-language talker was frequently sufficient to improve listeners’ comprehension of a completely different second-language talker. This finding matters practically. It suggests that single-talker training paradigms, which are far cheaper and simpler to implement than multi-talker regimes, can still deliver real-world listening benefits, whether in language education, in preparing listeners for communication in multilingual workplaces, or in clinical contexts where listeners must adjust to unfamiliar voices.

But then comes the twist that gives the study its title. When the researchers asked whether the size of the generalization effect was predicted by the intelligibility of the training talker, the answer was no. When they asked whether it was predicted by the acoustic similarity between training and test talkers, again the answer was no. Neither variable reliably drove the magnitude of transfer across the twenty pairs. The effects, in statistical terms, refused to line up with either theoretical account. The researchers had planned a replication strategy, intending to test additional talker triplets after the initial set, but because no clear patterns emerged in the first batch, the planned follow-up analyses were not pursued, and the untested triplets’ data were deposited in the project’s Open Science Framework repository for other researchers to examine.

This null result is more provocative than it might first appear. In science, a clean failure of two leading hypotheses is a map of where not to dig, and it forces the field to consider what else could be carrying the load. One possibility is that generalization depends on properties of the listener rather than the talker: individual differences in working memory, attention, or prior experience with accented speech are known to shape how people cope with difficult listening conditions, and such listener-side factors could swamp talker-side predictors in samples of ten listeners per pair. Another possibility is that the relevant similarity is not global acoustic distance but something more structured and linguistic, such as overlap in specific error patterns, for example whether two talkers share the same substitutions of particular English vowels or consonants. Two speakers could be acoustically distant overall yet confusable in exactly the phonemic categories that matter for transcription, a hypothesis that coarse representational distances might miss.

The study also carries a methodological lesson that extends well beyond speech perception. The use of self-supervised model representations to quantify acoustic similarity represents a broader shift in the cognitive sciences, where large-scale neural models trained on natural data are increasingly used as measurement instruments for psychological constructs. The same approach has recently been used by other groups, including work presented by Jin, Zhu, and Jaeger, to predict generalization of adaptation across talkers from latent speech representations. Yet the current findings temper the enthusiasm: even a sophisticated, data-driven similarity metric grounded in state-of-the-art speech models did not predict human generalization in this paradigm. The representations may capture acoustic similarity faithfully, but what the human perceptual system extracts from a brief encounter with one accented voice may be organized along dimensions that neither overall intelligibility nor global acoustic distance adequately describes.

What remains is a robust empirical anchor and an open question. The empirical anchor is that single-talker training works more often than not: listeners who adapt to one second-language voice typically carry some of that benefit to new voices, even under the adverse condition of background noise. The open question is the mechanism. The authors are candid that the predictors remain elusive, and that candor is itself a contribution, redirecting the field away from tidy two-factor stories and toward richer models that incorporate listener variability, linguistic structure, training design, and the temporal dynamics of perceptual learning. For the millions of people who navigate multilingual environments daily, the practical takeaway stands: a little exposure goes a long way. For the scientists, the accent on the horizon is clear even if its coordinates are not. The data and materials from the study are publicly available, ensuring that the next generation of hypotheses about how the brain learns to understand unfamiliar speech can be tested against the same evidence.

Subject of Research: Cross-talker generalization of perceptual adaptation to second-language speech after single-talker training

Article Title: Generalization of perceptual adaptation to second-language speech in a single-talker training paradigm: Predictors remain elusive

Article References: Kim, S.-E., Goldrick, M., Keshet, J., & Bradlow, A. R. (2026). Generalization of perceptual adaptation to second-language speech in a single-talker training paradigm: Predictors remain elusive. Attention, Perception, & Psychophysics, 88(7), Article 192. https://doi.org/10.3758/s13414-026-03333-5

Image Credits: AI Generated

DOI: 10.3758/s13414-026-03333-5

Keywords: speech perception, perceptual learning, second-language speech, cross-talker generalization, single-talker training, speech intelligibility, acoustic similarity, self-supervised representations, accented speech, speech-in-noise, bilingualism, auditory adaptation

The study, led by Cristian Pérez-Granados and an international team of ornithologists, set out to do something rarely attempted: compare how songs and calls, the two main categories of bird vocalizations, respond to individual, population, and geographic factors within the same species. Songs are typically complex, learned, and tied to sexual selection, while calls tend to be shorter, more innate, and used for immediate communication such as territorial defense. Because these signals follow different developmental pathways, the team reasoned they might respond differently to habitat fragmentation, population size, and social isolation, providing complementary windows into population health.

Dupont’s lark proved an ideal model. Males sing and give territorial calls at high rates during the hour before dawn, and the species’ communication system is already known to be shaped by landscape structure. Songs are learned after juveniles disperse from their natal areas, by copying nearby adult males, and typically consist of three to eight discrete types. Territorial calls, by contrast, are learned before dispersal during a short learning phase in summer, likely from parents and neighbors in the natal patch, and consist of just one to three short whistles repeated regularly. This difference in timing means that songs and calls should carry information about movement and social interaction at different life stages.

Between March and June of 2024 and 2025, the researchers surveyed 27 habitat patches spanning the two main strongholds of the species in Spain, the Ebro Valley and the Iberian System. Patch areas ranged from 14 to 1,903 hectares and local male abundance from three to 77 individuals. Working in the pre-dawn darkness, they recorded 188 singing males with directional microphones and high-resolution digital recorders, capturing at least two minutes of song per male, sufficient to document the complete individual repertoire. For calls, they recorded a minimum of three calls per individual, ultimately sampling 74 calling males across 11 populations. When no birds were detected during repeated visits, autonomous recorders were deployed to confirm genuine absence.

The acoustic analysis was meticulous. A single expert scanned all recordings through spectrogram inspection in Raven Pro, delineating each vocalization and classifying it into types based on morphology, temporal patterns, frequency characteristics, and note structure. When classifications were uncertain or types appeared across populations, additional experienced ornithologists reviewed the material, and types were only confirmed when at least two researchers agreed. The final catalog contained 396 distinct song types drawn from 4,208 annotated songs, and 60 call types from 634 annotated calls. These were converted into presence-absence matrices allowing the team to compute repertoire sizes at both individual and population levels, and to quantify acoustic similarity between every pair of males using the Jaccard index.

The results revealed a striking divergence between the two vocalization types. Individual song repertoire size did not depend on population size, but varied significantly between regions, with Ebro Valley males singing larger repertoires than those in the Iberian System. At the population level, however, song repertoire size increased significantly with the number of territories, likely because larger patches host multiple local dialects among neighboring groups of males. Call repertoires, whether measured per individual or per population, showed no relationship with population size, geography, or sampling effort. Songs, in short, behave as flexible cultural traits molded by social environment, while calls remain remarkably conservative and spatially stable.

The spatial analysis sharpened this contrast. Acoustic similarity between males declined steeply with distance for both signal types, but calls were always more similar than songs at every scale examined. Neighboring males within 500 meters shared substantial portions of both repertoires, with mean song similarity of 0.48 and call similarity of 0.60, both far above random expectations. Between populations, similarity collapsed to almost nothing. Fine-scale analysis in 100-meter distance classes uncovered a telling threshold: song similarity declined steadily and leveled off beyond roughly 500 meters, while call similarity remained relatively constant up to a kilometer. Because adults disperse only about 100 to 150 meters, the wider coherence of calls hints that juveniles move at least a kilometer within their natal patches before dispersing, offering a rare behavioral clue about a life stage that has remained almost entirely invisible to researchers.

The most consequential result came from a two-decade natural experiment. The team revisited 50 populations originally monitored between 2004 and 2007, spanning most of the European range, and asked whether historical mean song repertoire size predicted population fate in 2025. A binomial logistic regression delivered a clear answer: populations with larger individual song repertoires had significantly higher persistence probabilities. Extinct populations had averaged 4.9 song types per male, compared with 5.7 in those that survived. Only one extinct population, five percent, had exceeded a mean repertoire size of 6.0, whereas 43 percent of surviving populations did, and not one of the nine populations with means above 6.25 was lost over the following twenty years. This threshold, the authors suggest, may mark the point below which extinction risk rises markedly.

The researchers are careful to note that population persistence depends on many interacting pressures, from wind farm development and grazing regimes to genetic erosion, habitat quality, and extreme weather events. Yet the strength of the acoustic signal is hard to ignore, and it validates, on a far broader geographic scale and a much longer timescale, an idea first proposed in 2008: that a cultural trait can serve as a predictor of population viability. Because vocal repertoire size can be measured non-invasively from field recordings, even with autonomous units, it offers managers a practical, low-cost indicator for tracking population health before declines become irreversible.

Beyond its conservation implications, the study fills a fundamental gap in behavioral ecology. Most research on avian vocal behavior has focused on songs and alarm calls, rarely comparing multiple signal types within the same species. By showing that songs and calls follow partially independent ecological rules, shaped by different learning schedules, dispersal stages, and social contexts, the work demonstrates how much information is layered into a dawn chorus. For a species clinging to fragmented fragments of Spanish steppe, those layers may now determine whether conservationists hear its song for decades to come, or only its echo.

Subject of Research: Acoustic communication, vocal repertoire size, and population viability of the endangered Dupont's lark

Article Title: Singing and Calling Behavior of a Threatened Passerine With Spatially Variable Vocalizations: Individual, Population and Geographic Patterns

Article References: Pérez‐Granados, C., Alonso‐Moya, C. D., Barrero, A., Sáez‐Gómez, P., Bota, G., Lahoz‐Monfort, J. J., Laiolo, P., Méndez, M., Serrano, D., Tella, J. L., Vögeli, M., & Traba, J. (2026). Singing and Calling Behavior of a Threatened Passerine With Spatially Variable Vocalizations: Individual, Population and Geographic Patterns. Ecology and Evolution, 16(9), Article e74380. https://doi.org/10.1002/ece3.74380

Image Credits: AI Generated

DOI: 10.1002/ece3.74380

Keywords: Dupont's lark, birdsong, vocal repertoire, bioacoustics, habitat fragmentation, population viability, conservation, cultural transmission, acoustic similarity, dialects, steppe birds, early-warning signal