Segmentation of coarticulated speech in perception

The research investigates how listeners segment the acoustic speech signal into phonetic segments and explores implications that the segmentation strategy may have for their perception of the (apparently) context-sensitive allophones of a phoneme. Two manners of segmentation are contrasted. In one, listeners segment the signal into temporally discrete, context-sensitive segments. In the other, which may be consistent with the talker’s production of the segments, they partition the signal into separate, but overlapping, segments freed of their contextual influences. Two complementary predictions of the second hypothesis are tested. First, listeners will use anticipatory coarticulatory information for a segment as information for the forthcoming segment. Second, subjects will not hear anticipatory coarticulatory information as part of the phonetic segment with which it co-occurs in time. The first hypothesis is supported by findings on a choice reaction time procedure; the second is supported by findings on a 4IAX discrimination test. Implications of the findings for theories of speech production, perception, and of the relation between the two are considered.     

Speech perception, lexicality, and reading skill

This study examined the interaction between speech perception and lexical information among a group of 7-year-old children, of which 26 were poor readers and 36 were good readers. The children's performance was examined on tasks assessing reading skill, phonological awareness, pseudoword repetition, and phoneme identification. Although good readers showed clearly defined categorical perception in the phoneme identification task for both the /bif/-/pif/ and the /bis/-/pis/ continua, the category boundary for /bif/-/pif/ was at longer VOTs than the boundary for /bis/-/pis/, which characterizes the classic lexicality effect. Poor readers showed less sharply defined categorical perception on both continua. Although poor readers did not show the classic lexicality effect, lexicality did affect the overall rate with which phonemes were identified as /b/ or /p/ at each VOT. These findings suggest that the lexicon may operate as a compensatory mechanism for resolving ambiguities in speech perception. Furthermore, statistical correction for group differences in phoneme identification made group differences in phoneme deletion disappear, suggesting that deficits in speech perception may play a causal role in the phonological core deficit associated with reading failure.     

Speech perception and memory coding in relation to reading ability

Previous work has demonstrated that children who are poor readers have short-term memory deficits in tasks in which the stimuli lend themselves to phonetic coding. The aim of the present study was to explore whether the poor readers' memory dificit may have its origin in perception with the encoding of the stimuli. Three experiments were conducted with third grade good and poor readers. As in earlier experiments, the poor readers were found to perform less well on recall of random word strings and to be less affected by the phonetic characteristics (rhyming or not rhyming) of the items (Experiment 1). In addition, the poor readers produced more errors of transposition (in the nonrhyming strings) than did the good readers, a further indication of the poor readers' problems with memory for order. The subjects were tested on two auditory perception tasks, one employing words (Experiment 2) and the other nonspeech environmental sounds (Experiment 3). Each was presented under two conditions: with a favorable signal-to-noise ratio and with masking. The poor readers made significantly more errors than the good readers when listening to speech in noise, but did not differ in perception of speech without noise or in perception of nonspeech environmental sounds, whether noise-masked or not. Together, the results of the perception studies suggest that poor readers have a perceptual difficulty that is specific to speech. It is suggested that the short-term memory deficits characteristic of poor readers may stem from material-specific problems of perceptual processing.     

Categorical speech perception deficits distinguish language and reading impairments in children

We examined categorical speech perception in school‐age children with developmental dyslexia or Specific Language Impairment (SLI), compared to age‐matched and younger controls. Stimuli consisted of synthetic speech tokens in which place of articulation varied from ‘b’ to ‘d’. Children were tested on categorization, categorization in noise, and discrimination. Phonological awareness skills were also assessed to examine whether these correlated with speech perception measures. We observed similarly good baseline categorization rates across all groups; however, when noise was added, the SLI group showed impaired categorization relative to controls, whereas dyslexic children showed an intact profile. The SLI group showed poorer than expected between‐category discrimination rates, whereas this pattern was only marginal in the dyslexic group. Impaired phonological awareness profiles were observed in both the SLI and dyslexic groups; however, correlations between phonological awareness and speech perception scores were not significant. The results of the study suggest that in children with language and reading impairments, there is a significant relationship between receptive language and speech perception, there is at best a weak relationship between reading and speech perception, and indeed the relationship between phonological and speech perception deficits is highly complex.     

Speech perception and short-term memory deficits in persistent developmental speech disorder

Children with developmental speech disorders may have additional deficits in speech perception and/or short-term memory. To determine whether these are only transient developmental delays that can accompany the disorder in childhood or persist as part of the speech disorder, adults with a persistent familial speech disorder were tested on speech perception and short-term memory. Nine adults with a persistent familial developmental speech disorder without language impairment were compared with 20 controls on tasks requiring the discrimination of fine acoustic cues for word identification and on measures of verbal and nonverbal short-term memory. Significant group differences were found in the slopes of the discrimination curves for first formant transitions for word identification with stop gaps of 40 and 20 ms with effect sizes of 1.60 and 1.56. Significant group differences also occurred on tests of nonverbal rhythm and tonal memory, and verbal short-term memory with effect sizes of 2.38, 1.56, and 1.73. No group differences occurred in the use of stop gap durations for word identification. Because frequency-based speech perception and short-term verbal and nonverbal memory deficits both persisted into adulthood in the speech-impaired adults, these deficits may be involved in the persistence of speech disorders without language impairment.     

Contingent feature analysis in speech perception

A contingent adaptation effect is reported for speech perception. Experiments were conducted to test the effects of an alternating sequence of two adapting syllables, [da] and [t^hi], on the perception of two series of synthetic speech syllables, [ba]-[p^ha] and [bi]-[p^hi]. Each of the test series consisted of 11 stimuli varying in voice onset time, a cue which distinguishes voiced from voiceless stop consonants in word-initial position. The [da]-[t^hi] adapting sequence produced opposite shifts in the loci of the phonetic boundaries for the two test series. For the [ba]-[p^ha] series, listeners made fewer identification responses to the [b] category after adaptation, while for the [bi]-[p^hi] series, listeners made more responses to the [b] category. The opposing shifts indicate that the perceptual analysis of voicing in stop consonants is carried out with respect to vowel environment.     

Chronometric analysis of speech perception

The relationship between the phonological properties of speech sounds and the corresponding semantic entries was studied in two experiments using response time measures. Monosyllabic words and nonsense words were used in both experiments. In Experiment I. Ss were each presented with individual items and were required, in three different conditions, to respond positively if (1) the item contained a particular final consonant, (2) the item was a real word, (3) the item contained either a particular consonant or was a real word. Latencies indicated that separate retrieval of phonological and lexical information took about the same time. but that their combined retrieval was longer, indicating a serial or overlapping process. In Experiment II, Ss were presented with pairs of items, and they responded positively if (1) the two items were physically identical, (2) the two items were lexically identical (both real words or both nonsense words). Response latencies were longer for lexical than for physical matches. Lexical matches were’ significantly slower than physical matches even on the same pair of items. The results imply differential accessibility to separate loci of phonological and semantic features.     

Contextual facilitation in children's oral reading in relation to grade and decoding skill

This research investigated children's use of context to facilitate word recognition and comprehension-monitoring processes in the oral reading of connected prose as a function of grade level and decoding skill. Results indicated no overall contextual facilitation of word recognition accuracy, even in less skilled decoders, although there was evidence that less skilled decoders were assisted by context in decoding some content words. Children read word lists 50% more slowly than comparable selections of prose. The adoption of different and compensatory reading speed strategies in children's reading of prose and word lists renders the oral reading task an insensitive test of the contextual facilitation of word recognition accuracy. A qualitative analysis of the errors made in reading the prose passage showed that skilled decoders made (relative to less skilled decoders) a lower proportion of reading errors which, as first uttered, violated prior context, and a higher rate of contextually obligatory self-corrections, thus making a higher overall rate of contextually acceptable oral reading errors. These data were interpreted as suggesting that children's oral reading incorporates processing that occurs after lexical access, and that skilled decoders use context more effectively to monitor comprehension. In an oral reading task, this may counteract the tendency of less skilled decoders to rely more on context in the process of word recognition.     

Segmenting two-phoneme syllables: developmental differences in relation with early reading skills

This study explored developmental differences in children's segmentation skills of VC and CV syllables (e.g., /af/ and /fa/) in relation to their early reading abilities. To this end, we followed a subgroup of Dutch speaking prereaders who participated in, and replicated the segmentation task in first grade, at the outset of phonics reading instruction. Reading abilities were assessed after 6 and 9 months. First, we confirmed that VCs offer an easier context to isolate phonemes than CVs. Second, matching analyses showed that this development from VC to CV segmentation posed comparatively increasing difficulties for poor segmenters. Third, this qualitatively different development was reflected in early reading performance. Our data emphasize the importance of phonetic factors and instruction-based experiences in phonological development.     

Categorical perception of speech stimuli in children at risk for reading difficulty

Children determined to be at risk (n = 24) or not at risk (n = 13) for reading difficulty listened to tokens from a voice onset time (VOT) (/ga/-/ka/) or tone series played in a continuous unbroken rhythm. Changes between tokens occurred at random intervals and children were asked to press a button as soon as they detected a change. For the VOT series, at-risk children were less sensitive than not-at-risk children to changes between tokens that crossed the phonetic boundary. Maps of group stimulus space produced using multidimensional scaling of reaction times for the VOT series indicated that at-risk children may attend less to the phonological information available in the speech stimuli and more to subtle acoustic differences between phonetically similar stimuli than not-at-risk children. Better phonological processing was associated with greater sensitivity to changes between VOT tokens that crossed the phonetic boundary and greater relative weighting of the phonological compared to the acoustic dimension across both groups.     

Schematic concept formation: psychophysical analysis of early reading skill

The relation between the perception of class structure and reading skill was studied within the context of a schematic concept-formation task. Computer generation techniques were utilized to construct visual pattern-classification problems with controlled levels of inter-class pattern similarity of class discrimination difficulty. Children with reading problems were less efficient in their ability to assign patterns to classes than were children without reading problems. It was possible to assess reliably individual subjects' reading skill on the basis of their ability to perform the task.     

Inner speech as a learned skill

Three experiments are reported which address the problem of defining a role for inner speech. Experiments 1 and 2 establish that inner speech is acquired by normally developing readers between the ages of 8 and 11, and that both slow and fast readers show a similar pattern of acquistition, but do so at a different rate from normal readers. We suggest that the development of inner speech accompanies a strategy of reading aloud “with expression”; and that it is a manifestation of the need to prestructure oral utterances. These will thus contain the lexical items visible on the page within an appropriate prosodic envelope. Both segmental and suprasegmental phonemes contribute to meaning of spoken and, by analogy, written language. Experiment 3 showed that children at this critical point in learning to read comprehended text better when certain prosodic features were made visible on the text. Prosodic restructureing may thus be an important skill acquired by young readers as they progress toward fluent, silent adult reading.     

An analysis of speech perception in word deafness

A patient with a rather pure word deafness showed extreme suppression of right ear signals under dichotic conditions, suggesting that speech signals were being processed in the right hemisphere. Systematic errors in the identification and discrimination of natural and synthetic stop consonants further indicated that speech sounds were not being processed in the normal manner. Auditory comprehension improved considerably however, when the range of speech stimuli was limited by contextual constraints. Possible implications for the mechanism of word deafness are discussed.     

Visual pattern perception: a multidimensional analysis of development of children's reading skills

This study was designed to assess the extent to which there are visual pattern processing deficits which are related to differential reading skill of young children. Discrimination problems were constructed using visual patterns with features known to be processed differentially by the central nervous system based upon neurophysiological data. Analysis of variance and multidimensional scaling techniques were used to analyze the discrimination latencies of kindergarten through third grade children (12 per grade) who were classified either as good or poor readers. The results show that there are no pattern-specific effects which discriminate good from poor readers. Rather, good readers process all patterns more efficiently than do poor readers, and efficiency improves with age for all readers studied. This suggests that the locus of the reading deficit occurs at levels in the nervous system beyond where pattern-feature processing occurs.