Barbara Wise

Specific Deficits in Component Reading and Language Skills Genetic and Environmental Influences

Phonological coding, measured by the oral reading of nonwords, and orthographic coding, measured by the discrimination of words from homophonic nonwords (e.g., rane, rain), were compared for pairs of older children with reading disabilities (RD) and younger nondisabled readers matched on word recognition. Phonological coding was substantially lower for most children with RD, indicating a unique developmental deficit in phonological coding rather than an equal developmental lag across all component reading skills. Data from identical and fraternal twins indicated that the phonological coding deficit of the children with RD was highly heritable and accounted for most of the heritable variance in their word recognition deficits. The deficits of the twins with RD in segmental language skills (rhyming and phoneme segmentation) were related to the heritable variance in their phonological coding deficits. Orthographic coding was not significantly heritable, and it accounted for much of the environmental variance in word recognition deficits. Implications of the results for the remediation of reading disability are discussed.

Genes, Environment, and the Development of Orthographic Skills

Researchers have taken two somewhat different views regarding the nature and development of orthographic and phonological knowledge in spelling and reading. A strongly integrative view holds that development in both areas depends on a common underlying knowledge base. For example, Ehri (1989, 1992) has argued that alphabetic “orthographic images” involved in both reading and spelling are amalgamated with the phonological information pertaining to the word (see also, Barron, 1986; Goswami & Bryant, 1990; Stuart & Coltheart, 1988). A more separatist view is represented in “dual-route” theories that emphasize the independence of two routes to the lexicon; an indirect phonological-decoding route operating through the reader’s knowledge of grapheme-phoneme correspondences, and a direct “visual” route that uses orthographic knowledge to access the lexicon without phonological mediation (Coltheart, 1978; Morton, 1969). The “dual-route” view has been very influential in attempts to account for individual differences in acquired and developmental reading disorders (Boder, 1973; Castles & Coltheart, 1993; Coltheart, Curtis, Atkins, & Haller, 1993; Mitterer, 1982; Patterson, Marshal, & Coltheart, 1985; Seymour, 1986), as well as individual differences across the normal continuum of reading ability (Baron, 1979; Baron & Strawson, 1976; Freebody & Byrne, 1988; Treiman, 1984). In this chapter we selectively review earlier research and present new behavioral-genetic evidence on the degree of developmental independence between disabled readers’ skills in the indirect (phonological) and direct (orthographic) routes for the identification of printed words.

Reading instruction and remediation with the aid of computer speech

In this paper, we describe the application of new computer and speech synthesis technologies for reading instruction. Stories are presented on the computer screen, and readers may designate words or parts of words that they cannot read for immediate speech feedback. The important contingency between speech sounds and their corresponding letter patterns is emphasized by displaying the letter patterns in reverse video as they are spoken. Speech feedback is provided by an advanced text-to-speech synthesizer (DECtalk). Intelligibility data are presented, showing that DECtalk can be understood almost as well as natural human speech by both normal adults and reading disabled children. Preliminary data from 26 disabled readers indicate that there are significant benefits of speech feedback for reading comprehension and word recognition, and that children enjoy reading with the system.

Computer-Based Phonological Awareness and Reading Instruction

Reading with Orthographic and Segmented Speech (ROSS) programs use talking computers to deal with deficits in word recognition and phonological awareness. With ROSS, children read stories on a computer screen. Whenever they encounter a word they find difficult, they can request assistance by targeting the word with a mouse. The program highlights the word in segments and then pronounces the segments in order. In previous studies, children improved in reading, but children with relatively lower initial phonological awareness (PA) gained less than the others. In order to maximize the benefits from ROSS for all children, the current study aimed to improve PA before and while reading with ROSS, by using some programs based on theAuditory Discrimination in Depth method (Lindamood and Lindamood 1975), and others focusing on phoneme manipulation with speech feedback for all responses. The study compared the effects of this training with training in Comprehension Strategies (CS) based on Reciprocal Teaching techniques (Palincsar and Brown 1984), among second- to fifth-grade students with problems in word recognition. While both groups received equal instructional time in small-groups and with the computer, the groups differed in how much time they spent reading words in context. Whereas PA children spent half their computer time on PA exercises involving individual words and half reading words in context with ROSS, the CS group spent all their computer time reading words in context with ROSS. Both groups made significant gains in decoding, word recognition, and comprehension; however the PA groups gained significantly more than the CS group on all untimed tests of phoneme awareness, word recognition, and nonsense word reading. The CS children performed better on a test of time-limited word recognition; they also achieved higher comprehension scores, although only while reading with a trainer. The PA children’s improved decoding skill led to greater accuracy, but slower responses with difficult words, after one semester’s training.

Implementing a long-term computerized remedial reading program with synthetic speech feedback: Hardware, software, and real-world issues

This paper discusses hardware choices, software developments, implementation issues, and preliminary results from an ongoing long-term remedial reading study. Reading-disabled children read books on microcomputers linked to speech synthesizers, obtaining speech feedback on difficult words at whole-word, syllable, or subsyllable levels of segmentation. Word-recognition ability and attitude about reading improved for children using the system. In addition, segmented feedback especially benefited phonological word-decoding skills for most of the children.

Subsyllabic Units in Computerized Reading Instruction: Onset-Rime vs. Postvowel Segmentation

Previous studies suggest that adults and children divide spoken syllables into subsyllabic onset-rime units more readily than into any other kind of subsyllabic unit. We asked whether this same onset-rime segmentation might also be beneficial in teaching children to read. That is, can children learn more words segmented at the onset-rime boundary (e.g., CL-AP, D-ISH) than words segmented after the vowel (CLA-P, DI-SH)? In three experiments, first-grade students studied single words presented by a computer connected to a high-quality speech synthesizer. Experiment 1 used words of four letters but only three phonemes apiece (e.g., WHIP, DISH). In some of these words the onset-rime segmentation corresponded to the initial bigram (e.g., WH-IP); in some it did not (e.g., D-ISH). Experiments 2 and 3 used words of four letters and four phonemes (e.g., CLAP, CORN). In all three experiments, onset-rime segmentation proved more helpful than postvowel segmentation in short-term learning of the words.

Accurate visible speech synthesis based on concatenating variable length motion capture data

We present a novel approach to synthesizing accurate visible speech based on searching and concatenating optimal variable-length units in a large corpus of motion capture data. Based on a set of visual prototypes selected on a source face and a corresponding set designated for a target face, we propose a machine learning technique to automatically map the facial motions observed on the source face to the target face. In order to model the long distance coarticulation effects in visible speech, a large-scale corpus that covers the most common syllables in English was collected, annotated and analyzed. For any input text, a search algorithm to locate the optimal sequences of concatenated units for synthesis is described. A new algorithm to adapt lip motions from a generic 3D face model to a specific 3D face model is also proposed. A complete, end-to-end visible speech animation system is implemented based on the approach. This system is currently used in more than 60 kindergartens through third grade classrooms to teach students to read using a lifelike conversational animated agent. To evaluate the quality of the visible speech produced by the animation system, both subjective evaluation and objective evaluation are conducted. The evaluation results show that the proposed approach is accurate and powerful for visible speech synthesis.

Accurate automatic visible speech synthesis of arbitrary 3D models based on concatenation of diviseme motion capture data

We present a technique for accurate automatic visible speech synthesis from textual input. When provided with a speech waveform and the text of a spoken sentence, the system produces accurate visible speech synchronized with the audio signal. To develop the system, we collected motion capture data from a speakers face during production of a set of words containing all diviseme sequences in English. The motion capture points from the speakers face are retargeted to the vertices of the polygons of a 3D face model. When synthesizing a new utterance, the system locates the required sequence of divisemes, shrinks or expands each diviseme based on the desired phoneme segment durations in the target utterance, then moves the polygons in the regions of the lips and lower face to correspond to the spatial coordinates of the motion capture data. The motion mapping is realized by a key‐shape mapping function learned by a set of viseme examples in the source and target faces. A well‐posed numerical algorithm estimates the shape blending coefficients. Time warping and motion vector blending at the juncture of two divisemes and the algorithm to search the optimal concatenated visible speech are also developed to provide the final concatenative motion sequence. Copyright

Computer Speech and the Remediation of Reading and Spelling Problems.

A high-quality speech synthesizer (DECtalk, by Digital Equipment Corporation) is very intelligible to children with reading disabilities. Linking the DECtalk to a microcomputer yields a “talking computer” that provides a powerful tool for research and remediation of reading and spelling problems. Two clear and related findings about children with “specific reading disability” (dyslexia) have emerged from previous research: 1) deficits in word recognition primarily cause these childrens problems in reading comprehension, and 2) inherited deficits in analytic language processes underlie their difficulties in word recognition. These two findings suggest that speech support for words in text could help these children. In several studies at the University of Colorado, children with reading problems have read stories and books on talking computers for about 30 minutes a day, usually for 3–4 days per week during most of a semester. The childrens word recognition and phonological decoding have improved, compared to the skills of similar students who spent the time in regular classroom language arts instruction. The studies suggest that accurate word feedback supporting the reading of text helps these readers. One goal of the research has been to compare the benefits of presenting unknown words as wholes or in segments for different students. That goal has remained somewhat elusive, with interesting interactions that have been significant but not always stable. The paper also describes a different line of study using the DECtalk in a spelling program that allows children to explore English sound-spelling patterns as they compare pronunciations of their own spelling attempts and those of the test words. The paper concludes with descriptions of a current home-based reading study and a future study exploring the benefits of computer-based phonemic awareness training prior to the reading instruction.

How poor readers and spellers use interactive speech in a computerized spelling program

The ‘Spello’ program was designed to use interactive speech feedback on a talking-computer system to improve childrens spelling and phonological skills. In two versions of the program, the synthesizer pronounced the word to be spelled and the student tried to type in the word correctly. Both versions of the program showed the students which letters were correct in their spelling attempts. One version pronounced only the target word, as often as the child requested. The other version also provided intermediate speech feedback for childrens spelling attempts, so they could hear how their own attempts sounded, and compare them to the target word. Twenty-eight children aged seven to fourteen studied 16 words they had misspelled on pretests and 16 words of related word structure. For children ten years or older, training with intermediate speech feedback led to greater benefits in phonological coding skills than training with word-only feedback, reflected in the ability to read nonsense words related in structure to the trained words. Intermediate speech feedback also led to a marginally significant advantage in spelling the trained words. When the groups switched conditions, however, there was no difference in their tested skills after a second week of training.