Multiple Phonemes Research Articles

Transcranial Magnetic Stimulation Facilitates Neural Speech Decoding.

Transcranial magnetic stimulation (TMS) has been widely used to study the mechanisms that underlie motor output. Yet, the extent to which TMS acts upon the cortical neurons implicated in volitional motor commands and the focal limitations of TMS remain subject to debate. Previous research links TMS to improved subject performance in behavioral tasks, including a bias in phoneme discrimination. Our study replicates this result, which implies a causal relationship between electro-magnetic stimulation and psychomotor activity, and tests whether TMS-facilitated psychomotor activity recorded via electroencephalography (EEG) may thus serve as a superior input for neural decoding. First, we illustrate that site-specific TMS elicits a double dissociation in discrimination ability for two phoneme categories. Next, we perform a classification analysis on the EEG signals recorded during TMS and find a dissociation between the stimulation site and decoding accuracy that parallels the behavioral results. We observe weak to moderate evidence for the alternative hypothesis in a Bayesian analysis of group means, with more robust results upon stimulation to a brain region governing multiple phoneme features. Overall, task accuracy was a significant predictor of decoding accuracy for phoneme categories (F(1,135) = 11.51, p < 0.0009) and individual phonemes (F(1,119) = 13.56, p < 0.0003), providing new evidence for a causal link between TMS, neural function, and behavior.

A systematic review of auditory processing abilities in children with speech sound disorders

Children with speech sound disorder (SSD) have incomplete phonological acquisition without organic alteration. These children’s speech exhibits multiple phoneme omission, substitution, and distortion errors. Auditory processing disorders (APD) are associated with speech sound disorders. Since APD prevents the development of a stable representation of phonemes in the brain, it may cause speech issues and make semantics, syntax, and phonology challenging to acquire. The review summarises behavioral findings of auditory processing abilities in children with SSD. The study used a literature search of electronic databases (Google Scholar, Science Direct, Research Gate, PubMed) from 2007 to 2022. The retrieved articles were assessed in two stages: title and abstract screening, followed by a full-length article review. Six articles were selected after the full-length review of 12 shortlisted articles. The review of the articles showed that temporal processing abilities are affected in children with SSD. Thus, temporal processing is critical for developing phonological awareness because it is vital for perceiving and rapidly changing acoustic cues in speech. The close relationship between temporal processing and phonological awareness in children with SSD highlights the importance of including temporal processing assessment in these children. Appropriate auditory training which taps into temporal processing skills may improve phonological awareness in these children. The effect of auditory processing training on phonological skills in SSD children needs further study.

Short-term perceptual reweighting in suprasegmental categorization

Segmental speech units such as phonemes are described as multidimensional categories whose perception involves contributions from multiple acoustic input dimensions, and the relative perceptual weights of these dimensions respond dynamically to context. For example, when speech is altered to create an “accent” in which two acoustic dimensions are correlated in a manner opposite that of long-term experience, the dimension that carries less perceptual weight is down-weighted to contribute less in category decisions. It remains unclear, however, whether this short-term reweighting extends to perception of suprasegmental features that span multiple phonemes, syllables, or words, in part because it has remained debatable whether suprasegmental features are perceived categorically. Here, we investigated the relative contribution of two acoustic dimensions to word emphasis. Participants categorized instances of a two-word phrase pronounced with typical covariation of fundamental frequency (F0) and duration, and in the context of an artificial “accent” in which F0 and duration (established in prior research on English speech as “primary” and “secondary” dimensions, respectively) covaried atypically. When categorizing “accented” speech, listeners rapidly down-weighted the secondary dimension (duration). This result indicates that listeners continually track short-term regularities across speech input and dynamically adjust the weight of acoustic evidence for suprasegmental decisions. Thus, dimension-based statistical learning appears to be a widespread phenomenon in speech perception extending to both segmental and suprasegmental categorization.

Phonetic feature size in second language acquisition: Examining VOT in voiceless and voiced stops

Featural approaches to second language phonetic acquisition posit that the development of new phonetic norms relies on sub-phonemic features, expressed through a constellation of articulatory gestures and their corresponding acoustic cues, which may be shared across multiple phonemes. Within featural approaches, largely supported by research in speech perception, debate remains as to the fundamental scope or ‘size’ of featural units. The current study examines potential featural relationships between voiceless and voiced stop consonants, as expressed through the voice onset time cue. Native English-speaking learners of Spanish received targeted training on Spanish voiceless stop consonant production through a visual feedback paradigm. Analysis focused on the change in voice onset time, for both voiceless (i.e. trained) and voiced (i.e. non-trained) phonemes, across the pretest, posttest, and delayed posttest. The results demonstrated a significant improvement (i.e. reduction) in voice onset time for voiceless stops, which were subject to the training paradigm. In contrast, there was no significant change in the non-trained voiced stop consonants. These results suggest a limited featural relationship, with independent voice onset time (VOT) cues for voiceless and voices phonemes. Possible underlying mechanisms that limit feature generalization in second language (L2) phonetic production, including gestural considerations and acoustic similarity, are discussed.

Detecting Effect of Levodopa in Parkinson's Disease Patients Using Sustained Phonemes.

Background: Parkinson’s disease (PD) is a multi-symptom neurodegenerative disease generally managed with medications, of which levodopa is the most effective. Determining the dosage of levodopa requires regular meetings where motor function can be observed. Speech impairment is an early symptom in PD and has been proposed for early detection and monitoring of the disease. However, findings from previous research on the effect of levodopa on speech have not shown a consistent picture. Method: This study has investigated the effect of medication on PD patients for three sustained phonemes; /a/, /o/, and /m/, which were recorded from 24 PD patients during medication off and on stages, and from 22 healthy participants. The differences were statistically investigated, and the features were classified using Support Vector Machine (SVM). Results: The results show that medication has a significant effect on the change of time and amplitude perturbation (jitter and shimmer) and harmonics of /m/, which was the most sensitive individual phoneme to the levodopa response. /m/ and /o/ performed at a comparable level in discriminating PD-off from control recordings. However, SVM classifications based on the combined use of the three phonemes /a/, /o/, and /m/ showed the best classifications, both for medication effect and for separating PD from control voice. The SVM classification for PD-off versus PD-on achieved an AUC of 0.81. Conclusion: Studies of phonation by computerized voice analysis in PD should employ recordings of multiple phonemes. Our findings are potentially relevant in research to identify early parkinsonian dysarthria, and to tele-monitoring of the levodopa response in patients with established PD.

Feature Acquisition in Second Language Phonetic Development: Evidence From Phonetic Training

AbstractThis study employed a targeted phonetic instruction to explore the mechanisms that underpin second language (L2) phonetic acquisition. Broadly, two general approaches to phonetic acquisition have been previously proposed. A segmental approach suggests that learners acquire a series of individual, discrete phonemes (e.g., Flege, 1995), while a featural approach posits that L2 phonetic development occurs at the subsegmental level of the feature, which may be shared across multiple phonemes (e.g., de Jong, Hao, &amp; Park, 2009). This study extended this line of research, using a visual feedback paradigm to train English speakers on one of the three voiceless stop consonants in Spanish. Analysis focused on the change in voice onset time across three testing sessions (pretest, posttest, delayed posttest). Results demonstrated a significant change in voice onset time for trained and nontrained phonemes, suggesting that featural changes generalize to related phonemes. Theoretical and pedagogical implications are discussed.

Interactions Between Item Set and Vocoding in Serial Recall.

Serial recall of digits is frequently used to measure short-term memory span in various listening conditions. However, the use of digits may mask the effect of low quality auditory input. Digits have high frequency and are phonologically distinct relative to one another, so they should be easy to identify even with low quality auditory input. In contrast, larger item sets reduce listener ability to strategically constrain their expectations, which should reduce identification accuracy and increase the time and/or cognitive resources needed for identification when auditory quality is low. This diminished accuracy and increased cognitive load should interfere with memory for sequences of items drawn from large sets. The goal of this work was to determine whether this predicted interaction between auditory quality and stimulus set in short-term memory exists, and if so, whether this interaction is associated with processing speed, vocabulary, or attention. We compared immediate serial recall within young adults with normal hearing across unprocessed and vocoded listening conditions for multiple stimulus sets. Stimulus sets were lists of digits (1 to 9), consonant-vowel-consonant (CVC) words (chosen from a list of 60 words), and CVC nonwords (chosen from a list of 50 nonwords). Stimuli were unprocessed or vocoded with an eight-channel noise vocoder. To support interpretation of responses, words and nonwords were selected to minimize inclusion of multiple phonemes from within a confusion cluster. We also measured receptive vocabulary (Peabody Picture Vocabulary Test [PPVT-4]), sustained attention (test of variables of attention [TOVA]), and repetition speed for individual items from each stimulus set under both listening conditions. Vocoding stimuli had no impact on serial recall of digits, but reduced memory span for words and nonwords. This reduction in memory span was attributed to an increase in phonological confusions for nonwords. However, memory span for vocoded word lists remained reduced even after accounting for common phonetic confusions, indicating that lexical status played an additional role across listening conditions. Principal components analysis found two components that explained 84% of the variance in memory span across conditions. Component one had similar load across all conditions, indicating that participants had an underlying memory capacity, which was common to all conditions. Component two was loaded by performance in the vocoded word and nonword conditions, representing the sensitivity of memory span to vocoding of these stimuli. The order in which participants completed listening conditions had a small effect on memory span that could not account for the effect of listening condition. Repetition speed was fastest for digits, slower for words, and slowest for nonwords. On average, vocoding slowed repetition speed for all stimuli, but repetition speed was not predictive of individual memory span. Vocabulary and attention showed no correlation with memory span. Our results replicated previous findings that low quality auditory input can impair short-term memory, and demonstrated that this impairment is sensitive to stimulus set. Using multiple stimulus sets in degraded listening conditions can isolate memory capacity (in digit span) from impaired item identification (in word and nonword span), which may help characterize the relationship between memory and speech recognition in difficult listening conditions.

Judgments of self-identified gay and heterosexual male speakers: Which phonemes are most salient in determining sexual orientation?

While numerous studies have demonstrated that a male speaker's sexual orientation can be identified from relatively long passages of speech, few studies have evaluated whether listeners can determine sexual orientation when presented with word-length stimuli. If listeners are able to distinguish between self-identified gay and heterosexual male speakers of American English, it is unclear whether they form their judgments based on a phoneme, such as a vowel or consonant, or multiple phonemes, such as a vowel and a consonant. In this study, we first found that listeners can distinguish between self-identified gay and heterosexual speakers of American English upon hearing word-length stimuli. We extended these results in a separate experiment to demonstrate that listeners primarily rely on vowels, and to some extent consonants, when forming their judgments. Listeners were able to differentiate between the two groups of speakers for each of the vowels and three of the seven consonants presented. In a follow-up experiment we found evidence that listeners' judgments improved if they were presented with multiple phonemes, such as a vowel and /s/. These results provide important information about how different phonemes can provide discriminant information about a male speaker's sexual orientation.

Effects of Phonetic Context on Relative Fundamental Frequency

The effect of phonetic context on relative fundamental frequency (RFF) was examined, in order to develop stimuli sets with minimal within-speaker variability that can be implemented in future clinical protocols. Sixteen speakers with healthy voices produced RFF stimuli. Uniform utterances consisted of 3 repetitions of the same voiced sonorant-voiceless consonant-voiced sonorant speech sequence; moderately variable sentences contained speech sequences with a single voiceless phoneme (/f/, /s/, /ʃ/, /p/, /t/, or /k/); highly variable sentences were loaded with speech sequences using multiple phonemes. Effects of stimulus type (uniform, moderately variable, and highly variable) and phoneme identity (/f/, /s/, /ʃ/, /p/, /t/, and /k/) on RFF means and standard deviations were determined. Stimulus type and the interaction of vocal cycle and stimulus type were significant for RFF means and standard deviations but with small effect sizes. Phoneme identity and the interaction of vocal cycle and phoneme identity on RFF means and standard deviations were also significant with small to medium effect sizes. For speakers with healthy voices, uniform utterances with /f/ and /ʃ/ have the lowest standard deviations and thus are recommended for RFF-based assessments. Future work is necessary to extend these findings to disordered voices.

Primary Research on Arabic Visemes, Analysis in Space and Frequency Domain

Visemes are the unique facial positions required to produce phonemes, which are the smallest phonetic unit distinguished by the speakers of a particular language. Each language has multiple phonemes and visemes, and each viseme can have multiple phonemes. However, current literature on viseme research indicates that the mapping between phonemes and visemes is many-to-one: there are many phonemes which look alike visually, and hence they fall into the same visemic category. To evaluate the performance of the proposed method, the authors collected a large number of speech visual signal of five Algerian speakers male and female at different moments pronouncing 28 Arabic phonemes. For each frame the lip area is manually located with a rectangle of size proportional to 120*160 and centred on the mouth, and converted to gray scale. Finally, the mean and the standard deviation of the values of the pixels of the lip area are computed by using 20 images for each phoneme sequence to classify the visemes. The pitch analysis is investigated to show its variation for each viseme.

Acoustic representations for tonal phonological categories.

The learnability of tonal categories was investigated by studying how category separability is affected by different phonetic representations of tone. Gauthier et al. [(2007)] found that f 0 velocity contours densely sampled over time were sufficient for near-perfect categorization of Mandarin tones. Since infant learners begin as “citizens of the world” before developing language-specific representations of sound categories in response to ambient language input [Kuhl (2004)], cross-linguistic data were studied. Based on tonal production data from Bole, Igbo, Mandarin, Cantonese, and White Hmong, we found that (1) densely sampled f 0 velocity contours are insufficient for learning tones, but (2) coarse temporal sampling of phonetic features can produce well-separated tonal categories, and moreover, that (3) phonetic features for tonal representation necessarily extend beyond f 0 to voice quality features, and (4) features for tonal representation are language-specific. Results (2) and (3) were also supported by tonal perception experiments in Cantonese. It was found that listeners can maintain tonal identification accuracy under coarse sampling (down to three to five samples per syllable), with the speech signal interrupted with noise as in multiple phoneme restoration experiments. It was also found that listeners are biased in tonal identification by creaky voice in the speech signal.

The effects of syllable instruction on phonemic awareness in preschoolers

Purpose Preschooler instruction for speech sound awareness typically teaches a progression of speech units from sentences to phonemes, ending at simple first phoneme activities. This study investigates the effects of teaching advanced tasks of phoneme blending and segmenting with and without the larger speech unit of the syllable. Method Thirty-nine 4–5-year-old typically developing children received twice-weekly small-group instruction in three conditions: two weeks of syllable tasks then four weeks of multiple phoneme tasks (SP), four weeks of multiple phoneme tasks only (MP), or an active control condition of first phoneme instruction (FP). Results The conditions SP and MP showed large significant gains on blending and segmenting and no significant differences on first phoneme isolating compared to the FP condition. A comparison of SP and MP did not show significant differences on phoneme blending and segmenting, but SP showed significantly more confusion during early sessions of phoneme instruction. Conclusion This preliminary evidence suggests that preschoolers can improve understanding of phoneme blending and segmenting, without first being taught syllable blending and segmenting, and with no negative effects on first sound awareness. These findings support a more efficient way of teaching preschoolers awareness of the individual sounds of speech. Replication with a larger sample, including children at-risk for literacy difficulties, is recommended before firm conclusions should be drawn.

Speech technology‐based assessment of phoneme intelligibility in dysarthria

Currently, clinicians mainly rely on perceptual judgements to assess intelligibility of dysarthric speech. Although often highly reliable, this procedure is subjective with a lot of intrinsic variables. Therefore, certain benefits can be expected from a speech technology-based intelligibility assessment. Previous attempts to develop an automated intelligibility assessment mainly relied on automatic speech recognition (ASR) systems that were trained to recognize the speech of persons without known impairments. In this paper automatic speech alignment (ASA) systems are used instead. In addition, previous attempts only made use of phonemic features (PMF). However, since articulation is an important contributing factor to intelligibility of dysarthric speech and since phonological features (PLF) are shared by multiple phonemes, phonological features may be more appropriate to characterize and identify dysarthric phonemes. To investigate the reliability of objective phoneme intelligibility scores obtained by three types of intelligibility models: models using only phonemic features (yielded by an automated speech aligner) (PMF models), models using only phonological features (PLF models), and models using a combination of phonemic and phonological features (PMF + PLF models). Correlations were calculated between the objective phoneme intelligibility scores of 60 dysarthric speakers and the corresponding perceptual phoneme intelligibility scores obtained by a standardized perceptual phoneme intelligibility assessment. The correlations between the objective and perceptual intelligibility scores range from 0.793 for the PMF models, over 0.828 for PLF models to 0.943 for PMF + PLF models. The features selected to obtain such high correlations can be divided into six main subgroups: (1) vowel-related phonemic and phonological features, (2) lateral-related features, (3) silence-related features, (4) fricative-related features, (5) velar-related features and (6) plosive-related features. The phoneme intelligibility scores of dysarthric speakers obtained by the three investigated intelligibility model types are reliable. The highest correlation between the perceptual and objective intelligibility scores was found for models combining phonemic and phonological features. The intelligibility scoring system is now ready to be implemented in a clinical tool.

Shortlist B: A Bayesian model of continuous speech recognition.

A Bayesian model of continuous speech recognition is presented. It is based on Shortlist (D. Norris, 1994; D. Norris, J. M. McQueen, A. Cutler, & S. Butterfield, 1997) and shares many of its key assumptions: parallel competitive evaluation of multiple lexical hypotheses, phonologically abstract prelexical and lexical representations, a feedforward architecture with no online feedback, and a lexical segmentation algorithm based on the viability of chunks of the input as possible words. Shortlist B is radically different from its predecessor in two respects. First, whereas Shortlist was a connectionist model based on interactive-activation principles, Shortlist B is based on Bayesian principles. Second, the input to Shortlist B is no longer a sequence of discrete phonemes; it is a sequence of multiple phoneme probabilities over 3 time slices per segment, derived from the performance of listeners in a large-scale gating study. Simulations are presented showing that the model can account for key findings: data on the segmentation of continuous speech, word frequency effects, the effects of mispronunciations on word recognition, and evidence on lexical involvement in phonemic decision making. The success of Shortlist B suggests that listeners make optimal Bayesian decisions during spoken-word recognition.

Assessment of information processing in children with functional articulation disorders

The aim of this study was to investigate the information processing skills in children who have functional articulation disorder by using a cognitive-linguistic test battery. Thirty-three children with functional articulation disorder were enrolled in this study. They ranged in age from 6 to 10 years. A control group without articulation disorders was composed of 160 children with similar properties. In the first stage of this study, both groups were administered the Ross Information Processing Assessment Test-Primary. In this test, the scores obtained from eight subtests; namely immediate memory, recent memory, recall of general information, spatial orientation, temporal orientation, organization, problem solving and abstract reasoning, was calculated for both groups. The scores of the eight subtests are combined to form the four composites. These composites are memory quotient, orientation quotient, thinking and reasoning quotient and information processing quotient. Information processing quotient is the best and the most comprehensive estimate of a child's overall information processing ability. In the second stage of the study the articulation-disordered group was divided into two subgroups according the number of the mistakes, which was done at the level of the phoneme. The scores obtained from both subgroups were compared with the scores of the control group. The overall statistical analysis of the scores revealed that; the study group had significantly lower scores than the control group from memory quotient, thinking and reasoning quotient and information processing quotient. When compared to control group, the subgroup who cannot pronounce one phoneme had similar scores from all four composites. Moreover, the children who cannot pronounce multiple phonemes had significantly lower scores from memory quotient, thinking and reasoning quotient and information processing quotient. The results obtained from this study seem to be suggesting that information processing skills of children with functional articulation disorder are significantly low as compared to normal children. These results are revealing that the information processing skills of children with functional articulation disorder should be investigated in a detailed manner. According to the results obtained from this investigation these children should be put on deficit oriented education programs in addition to articulation therapy.

The Significance of Word Lists (review)

Reviewed by: The significance of word lists by Brett Kessler Magnus Widell The significance of word lists. By Brett Kessler. (Dissertations in linguistics.) Stanford, CA: CSLI Publications, 2001. Pp. x, 277. $23.00. This is a significant revision of a Ph.D. dissertation submitted in the Stanford Department of Linguistics in 1999 (available from UMI dissertation services under the title Estimating the probability of historical connections between languages). It should, however, be emphasized that the book under review is a rather different animal from a typical doctoral dissertation. In Ch. 1 (1–16) Kessler presents the main purpose of the book and provides a general introduction to the basic methodology and problems connected to statistical evaluations of language connectivity. Ch. 2 (17–41) leads the reader through what can best be described as a crash-course in statistics and, in the final ten pages, how statistics have and may be applied to historical linguistics. The following seven chapters (43–144) are concerned with various aspects and problems connected to statistical measurements of language correspondence. K offers a number of important insights and suggestions on how existing methods may be improved or modified to reduce the risks of misleading or incorrect results. To illustrate his theoretical arguments, K uses a data set of 200 words in English, French, German, Latin, Albanian, Turkish, Hawaiian, and Navajo (38f.). Ch. 10, the longest in the book (145–80), explores the χ2 metric and some problems connected to it as a statistical method. K concludes that there are strong arguments in favor of recurrence (R2) metric as an alternative to the χ2 metric. Different phonetic issues as well as tests involving multiple phonemes are addressed, and, while not entirely dismissive, K argues against multilateral comparisons to show language relatedness. Ch. 11 (181–97) summarizes the book and offers what is, regardless of whether one chooses to employ K’s methodology or not, useful advice to anyone attempting to embark on a statistical study of languages. A lengthy appendix (199–257) lists the 200 different words used in the book’s many methodological examples. Shorter etymological notes (with [End Page 442] references) are provided. The book ends with a list of references (259–66) and a general index (267–77). According to K, the book is targeted for ‘linguists who are curious about statistical argumentation but do not necessarily wish to devote an inordinate amount of attention to its more tortuous byways’ (180), and the book can indeed be successfully studied by a more general readership. As stated by the author (19ff.), the establishment of true relatedness between languages represents only a minor part of historical linguistics, the main concern being to evaluate and investigate the already established types of relations. The author, however, rather than trying to establish the nature of the historical connection of a set of languages—or presenting any method on how it could be established—attempts to fill a gap in the field of historical linguistics by offering a statistical method to establish historical correspondence between languages. Obviously, a major issue is to decide what can be considered true correspondence as contrasted to correspondence generated by chance. Whether K’s measure of relatedness is going to be more universally accepted and employed by historical linguists remains to be seen. However, the numerous tests performed on the 200-word lists are convincing, and since K invites other experts to apply the methodology to other languages, the reviewer sees no reason why it should not. On the whole, K’s book is a most thorough study of statistical methodology that can be most enthusiastically recommended to both students and more established scholars interested in all forms of lexicostatistical analyzing. Magnus Widell Northeast Normal University, PRC Copyright © 2003 Linguistic Society of America

Activation of Multiple Phoneme Associates of Graphemes in Visual Word Recognition

One general issue in the domain of visual word recognition is to delineate the nature of readers' knowledge of the print-sound mapping. A more specific question is to determine whether multiple grapheme–phoneme associations are available and activated during the phonological transcoding of a letter string. Evidence for the activation of irregular associations during print-to-sound transcoding, independently from lexical influences, was assessed in a letter detection task by examining performance on target-absent pseudowords. We contrasted two types of pseudowords that could be considered homophone with a real word by application of either grapheme–phoneme correspondence rules or of multiple phonemic activation. Performance on both types of homophones was compared to nonhomophone control pseudowords, strictly equivalent in terms of orthographic similarity to the base words. The finding of a homophony disadvantage for the homophones by multiple activation was interpreted as evidence for multiple phonemic activation in the print-to-sound conversion system.

The modification of multiple articulation errors based on distinctive feature theory.

The effectiveness of articulation remediation procedures based on distinctive feature theory was evaluated through the administration of an articulation program designed for this purpose. Two preschool children with multiple phoneme errors which could be described by a distinctive feature analysis were the subjects. Both children substituted stop phonemes for most continuant phonemes. Each child was individually administered the distinctive feature program which is described in full. Data are presented which indicate the adequacy of the treatment program, the acquisition of correct articulation of the two directly treated target phonemes, and the concurrent improvement of five other nontreated error phonemes. Such across-phoneme generalization was predicted by distinctive feature theory. Certain modifications in the treatment program are suggested and theoretical/empirical questions regarding articulation remediation from a distinctive features viewpoint are discussed.

Auditory perception of sequenced words in apraxia of speech.

In a test of auditory perception of phonemes, 10 patients with apraxia of speech were compared with 10 matched normal control subjects on a task of pointing to pictures designated by sequences of auditory word stimuli. The test items were 190 sequences of two and three words with minimal variation of initial consonant, final consonant, medial vowel, or multiple phonemes. The apraxic patients made significantly more perceptual errors as a group, but varied considerably in their level of performance. The achievement of some patients was within the range established by the control subjects; that of others was unequivocally inferior. The major deficit in the apraxic group appeared to be impairment of ability to retain the second- and third-syllable consonant elements in three-word sequences. Apraxia of speech can occur in relatively pure form in the absence of auditory perceptual impairment. It can also occur in conjunction with reduced auditory retention span, which impairs the perception of sequences of words which are minimally varied phonemically.