Search

Cross-linguistic studies on time reference in highly inflected languages have shown that tense inflection is particularly vulnerable in agrammatic speakers. According to the PAstDIscourseLInking Hy...

This chapter provides a tutorial on how to analyze phonetic data with Generalized Additive Mixed Models (GAMs). It presents analyses for three example datasets, all of which address aspects of the phonetics of Taiwan Mandarin. The first dataset contains response time data from an auditory priming experiment. For the second dataset, the chapter focuses on the realizations of Mandarin high-level tones by different speaker groups in Taiwan. The third dataset addresses the ongoing merging of two sets of sibilants in Taiwan Mandarin. GAMs can model time-varying response variables. The linear mixed model can allow the intercept and slope of regression lines for individual subjects to vary, but it cannot handle the case in which functional relations are not linear but instead wiggly. In phonetics, a lot of features that speech clinician measure change over time. These include pitch, intensity and formants. Time is a numeric variable ranging from one to ten in normalized time.

This chapter presents examples of acoustic indices that have been discussed in the literature of speech disorders and have been proven useful in clinical settings. These include both segmental and suprasegmental acoustic measures. Acoustic phonetics is the study of physical properties of speech. The essence of acoustic phonetics is the conversion of sound waves to visual representations that can be measured and related to phonetic features and physiological components of speech production. Overall, the goal of speech recording is to capture a speech signal that is clear, and has minimal distortion and background noise. Generally, the acoustic signal can be analyzed as an amplitude-by-time display, an amplitude-by-frequency display or an amplitude-by-time-by-frequency display. Vowel formants for monophthongs and formant trajectories for diphthongs can be readily measured by identifying the darker bands in a wideband spectrogram. Suprasegmental characteristics refer to aspects of speech that are not restricted to individual segments, but extend over larger units, that is, syllables, words, phrases or sentences.

Phonological awareness is closely related to reading acquisition and it is often the focus of the therapy services provided by speech-language pathologists and special education teachers. In this study, we investigate whether sonority theory can account for the developmental patterns of phoneme awareness skills in Greek. To that end, 40 preschool and first grade children carried out an offline metaphonological task that involved the initial consonant deletion of two-consonant clusters. Overall, children's performance was in line with the sonority sequencing principle (SSP); consonant clusters that display a maximal rise in sonority were easier to manipulate compared to clusters with minimal sonority difference or SSP-violating clusters. Affricates generated the highest number of errors, a finding that strengthens the singleton status of /ts/ and /dz/ in Greek. Increased error rates were also found for /ps/, /pc/, /ks/, a fact attributed to both to their spelling and sonority characteristics. Conclusively, it is claimed that the developmental error patterns reported in this study can be used to inform appraisal and treatment protocols of phonological awareness in Greek, by organizing metaphonological tasks based on the presumed level of difficulty of the items tested and/or treated.

This brief article reflects on some pitfalls inherent in the learning and teaching of segmental phonetic transcription. We suggest that a gestural interpretation to disordered speech data, in conjunction with segmental phonetic transcription, can add valuable insight into patterns of disordered speech, and that a gestural orientation should form part of the understanding and interpretation of the process of transcribing and the use of transcripts. We illustrate our discussion with data from a case of moderately to severely unintelligible speech produced by a boy aged 5;9.

Purpose: Accents may have an impact on how a speaker’s personality is perceived, and such dynamics may impact the therapeutic relationship. This study examined how speech-language pathology students in the Acadiana (Cajun) area of Louisiana perceive speakers of Cajun English (CE), Standard American English (SAE), and Standard Southern English (SE) on 2 dimensions of personality perception: solidarity (social closeness) and competence (education and intelligence). Method: An online survey was completed by 73 of the 285 undergraduate and master’s students at the University of Louisiana at Lafayette. Participants were presented with a standardized reading sample in each of the target accents. Target accents were not disclosed to participants. Students were asked to rate the speakers on a 6-point Likert scale for each of the 5 items in either dimension. Results: The CE speaker was rated higher than the SAE and SE speakers on 4 out of 5 measures of solidarity (Sociable, Likable, Friendly, and Kind), whereas the SAE speaker was rated higher than the CE or SE speakers on 2 out of 5 measures of competence (Educated and Intelligent). These results are in line with previous findings regarding the perception of personality traits in speakers of standard versus nonstandard, in-group accents. Conclusion: Despite the fact that training in linguistics is part of their curriculum, speech-language pathology students in the Acadiana area showed stereotypical patterns of accent perception. It is suggested that speech-language pathology education might benefit from targeting students’ preconceptions in order to help students appreciate the uniqueness of each client.