Purpose: This study investigated how listeners’ native language affects their weighting of acoustic cues (such as vowel quality, pitch, duration, and intensity) in the perception of contrastive word stress. Method: Native speakers (N = 45) of typologically diverse languages (English, Russian, and Mandarin) performed a stress identification task on nonce disyllabic words with fully crossed combinations of each of the 4 cues in both syllables. Results: The results revealed that although the vowel quality cue was the strongest cue for all groups of listeners, pitch was the second strongest cue for the English and the Mandarin listeners but was virtually disregarded by the Russian listeners. Duration and intensity cues were used by the Russian listeners to a significantly greater extent compared with the English and Mandarin participants. Compared with when cues were noncontrastive across syllables, cues were stronger when they were in the iambic contour than when they were in the trochaic contour. Conclusions: Although both English and Russian are stress languages and Mandarin is a tonal language, stress perception performance of the Mandarin listeners but not of the Russian listeners is more similar to that of the native English listeners, both in terms of weighting of the acoustic cues and the cues’ relative strength in different word positions. The findings suggest that tuning of second-language prosodic perceptions is not entirely predictable by prosodic similarities across languages.
Purpose: The current study investigates how individual differences in cochlear implant (CI) users’ sensitivity to word–nonword differences, reflecting lexical uncertainty, relate to their reliance on sentential context for lexical access in processing continuous speech.
Method: Fifteen CI users and 14 normal-hearing (NH) controls participated in an auditory lexical decision task (Experiment 1) and a visual-world paradigm task (Experiment 2). Experiment 1 tested participants’ reliance on lexical statistics, and Experiment 2 studied how sentential context affects the time course and patterns of lexical competition leading to lexical access.
Results: In Experiment 1, CI users had lower accuracy scores and longer reaction times than NH listeners, particularly for nonwords. In Experiment 2, CI users’ lexical competition patterns were, on average, similar to those of NH listeners, but the patterns of individual CI users varied greatly. Individual CI users’ word–nonword sensitivity (Experiment 1) explained differences in the reliance on sentential context to resolve lexical competition, whereas clinical speech perception scores explained competition with phonologically related words.
Conclusions: The general analysis of CI users’ lexical competition patterns showed merely quantitative differences with NH listeners in the time course of lexical competition, but our additional analysis revealed more qualitative differences in CI users’ strategies to process speech. Individuals’ word–nonword sensitivity explained different parts of individual variability than clinical speech perception scores. These results stress, particularly for heterogeneous clinical populations such as CI users, the importance of investigating individual differences in addition to group averages, as they can be informative for clinical rehabilitation.
This study investigated methodological and theoretical aspects of using mean length of utterance (MLU) and its alternatives in cross-linguistic research, and in particular its applicability to Russian – a language with opaque grammatical paradigms and rich system of derivational morphology.
Audio recordings of spontaneous speech samples were collected from 27 Russian-speaking children aged between 2;9 and 5;7 (years;months) over individual play sessions. For each participant, the first 100 complete utterances were transcribed and coded for several types of utterance length measurements, including their length in morphemes (grammatical and derivational), words and syllables. In addition, the average number of unique grammatical forms produced by each child was calculated.
A combination of Pearson correlation analysis and Bland-Altman difference plots established that MLU can be reliably used in Russian-speaking children aged around 3;0 years. In contrast, the average number of unique grammatical forms remains a sensitive measurement of language capabilities even in older children aged over 3;6. In addition, it was demonstrated that two quantitative measurements – MLU in syllables and morphemes – show good agreement, suggesting that these measurements can be used interchangeably across studies. Sample size analysis revealed that samples under 75 utterances do not provide sufficient reliability for estimating a child’s MLU.
This paper demonstrated that MLU can be used in young Russian-speaking children under 3;0–3;6 years. Also, we showed that the classical morpheme calculation approach can be substituted with counting syllables, which is much more time-efficient in the absence of automated parsers and is potentially more appropriate for some (e.g., polysynthetic) languages. Importantly, the proposed alternative to MLU – the average number of grammatical forms in a sample – appears to be a more sensitive measurement of language capabilities even in older children. Theoretical and practical implications of these findings are discussed.
Corpus analyses of spontaneous language fragments of varying length provide useful insights in the language change caused by brain damage, such as caused by some forms of dementia. Sample size is an important experimental parameter to consider when designing spontaneous language analyses studies. Sample length influences the confidence levels of analyses. Machine learning approaches often favor to use as much language as available, whereas language evaluation in a clinical setting is often based on truncated samples to minimize annotation labor and to limit any discomfort for participants. This article investigates, using Bayesian estimation of machine learned models, what the ideal text length should be to minimize model uncertainty.
We use the Stanford parser to extract linguistic variables and train a statistic model to distinguish samples by speakers with no brain damage from samples by speakers with probable Alzheimer's disease. We compare the results to previously published models that used CLAN for linguistic analysis.
The uncertainty around six individual variables and its relation to sample length are reported. The same model with linguistic variables that is used in all three experiments can predict group membership better than a model without them. One variable (concept density) is more informative when measured using the Stanford tools than when measured using CLAN.
For our corpus of German speech, the optimal sample length is found to be around 700 words long. Longer samples do not provide more information.
Purpose: The aim of this paper was to explore how the type of allomorph (e.g., past tense buzz[d] vs. nod[əd]) influences the ability to perceive and produce grammatical morphemes in children with typical development (TD) and with Specific Language Impairment (SLI). Method: The participants were monolingual Australian-English-speaking children. The SLI group included 13 participants (mean age=5;7); the controls were 19 TD children (mean age=5;4). Both groups performed a grammaticality judgment and elicited production task with the same set of nonce verbs in 3rd person singular and past tense forms. Results: Five-year-olds are still learning to generalise morphophonological patterns to novel verbs, with syllabic /əz/ and /əd/ allomorphs significantly more challenging to produce, particularly for the SLI group. The greater phonetic content of these syllabic forms did not enhance perception. Conclusions: Acquisition of morphophonological patterns involving low-frequency allomorphs is still underway in TD 5-year-olds, and is even more protracted in SLI population, despite these patterns being highly predictable. Children with SLI will therefore benefit from targeted intervention with low-frequency allomorphs.