The goal of the present investigation was to identify the reading processes that are impaired in children whose digit naming speeds are slow. Continuous digit naming speed was assumed to measure the automaticity with which character codes may be accessed in memory, and the automaticity of this process was assumed to be a prerequisite for the accurate performance of higher level reading processes. In Study I it was found, for children in grades 1 through 3, that digit naming speed was reliably correlated with reading of both irregularly spelled words and pronounceable nonsense words, and that most of the variance shared by irregular-word and nonsense-word reading was accounted for by digit naming speed. These results were interpreted as evidence that character-identification automaticity is equally important to the direct-access and speech-recoding routes of word recognition. Models of how character-identification automaticity might affect each of these routes are discussed. In Study 2 it was found, for children in grades 4 through 10, that the correlation of digit naming speed with reading comprehension was significantly smaller than its correlation with word recognition, and that the former correlation was completely accounted for when either WISC-R verbal IQ or word-recognition accuracy was controlled. These results are contrasted with comparable results reported by other investigators, and methodological differences are noted.

The acquisition of words referring to three types of actions was examined in normally developing and language-impaired children using a nonsense word paradigm. Fourteen language-normal and 10 language-impaired children whose speech was limited to single-word utterances served as subjects. The children were presented with 12 experimental words in 5 sessions over a period of approximately 3 weeks. The experimental words referred to actions that were classified as intransitive, transitive and specific to a particular object, or transitive but performed on 4 different objects. The children in both groups produced few of the action words. However, the groups differed in their comprehension of the three action word types. Specifically, the language-impaired children did not exhibit differences in comprehension across the different types of actions. The language-normal children, however, comprehended fewer words for intransitive actions than for the other types. The implications of these findings for characterizations of early lexical acquisition and for the nature of specific language impairment are addressed.

This study tested the hypothesis that certain sound substitutions in older, normally developing children are perceptually based substitutions, typified by poor discrimination, while others are phonetic substitutions – phonemic distinctions that are maintained by the child in a phonetically non-adult fashion. The perception and production of English voiceless fricatives in twelve normally developing monolingual children aged 3;2–5;6 was investigated using a picture-pointing task, audio recordings, and acoustic analysis. Results include significant correlations between production and perception scores for /θ/ but not /s/ and significant acoustic differences in substitutions of [θ] for /s/ (versus productions of [θ] for /θ/) but not in substitutions of [f] or [s] for /θ/. It is suggested that substitutions for /θ/, with the correlated poor discrimination, may indicate a non-adult phonemic representation, while substitutions for /s/ tend to have a motoric basis in older children. Stages in the acquisition of /θ/ are hypothesized.

The early evidence pertaining to the development of phonological segmentation abilities and their relation to reading was collected with English-speaking subjects. Although data from other languages have been obtained, explicit cross-language comparisons have not been made. It was considered that since languages vary in their phonological structures, they may also vary in the demands they make on the beginning reader. The present study compared the segmentation abilities of Italian children with those of English-speaking (American) children using the same methods of assessment and the same subject-selection criteria. At the preschool level, though the Italian children manifested a higher level of performance overall, their pattern of performance paralleled that obtained earlier with American children. In both groups, syllable segmentation ability was stronger than phoneme segmentation. After school entrance, this pattern remained unchanged in American children but was reversed in Italian beginning readers. In both language groups, however, phonemic segmentation ability distinguished children of different levels of reading skill. The discrepancies between the language groups were seen as reflecting phonologic and orthographic differences between the languages.

In this study we compared two methods for teaching grammatical exceptions in the foreign language classroom. Thirty-nine students in two sections of an introductory college French course served as subjects. Eight target structures, exemplifying “exceptions to a rule,” were randomly assigned to one of two teaching conditions for a section taught in the spring; each structure was assigned to the opposite teaching condition for a section taught the following fall. In one condition we simply taught the students the exception as an exception. In the other – what we called the Garden Path condition – we presented canonical exemplars encouraging students to induce the rule; we then asked them to generate the form (which we knew to be an exception) and then corrected their resulting overgeneralization error. Analysis of subsequent formal testing showed that students learned the exception better in the Garden Path condition and that this advantage persisted throughout the semesterlong course. We hypothesized that this technique helped students to focus attention both on the rule and on the features of the particular structure that marked it as an exception.

The effect of language of instruction on verbal reasoning of balanced bilingual children was investigated. Subjects were 36 sixth-grade French-English bilinguals who were receiving primary instruction in either French or English. Eight types of verbal analogies were created by varying the language of the analogy stem (French or English), the language of the analogy solution (French or English), and level of difficulty (easy or hard). The primary analysis was conducted through ANOVA with program type as between-subjects variable and item types as repeated measures factors. The results showed a significant interaction between language of instruction and language of presentation. In addition, there were different language effects for the two groups. The introduction of the noninstructional language (French) had a significant effect on the mean response time of the English Language Program Group at both levels of difficulty. The French Language Program Group showed a more complex pattern of results: there was a statistically significant advantage for the All-English items as compared with the All-French items on the easy analogies but a trend towards poorer performance upon introduction of English on the hard analogies. These results are discussed in terms of the differences between the two groups of subjects and the effect of academic experience on language-related thought processes and fluency. The distinction between contextualized and decontextualized language is also discussed.

A spelling-age match design was used to test the hypothesis that deficits in phonologically related skills may be causally related to difficulties in acquiring basic spelling knowledge. Poor grade 5 spellers, average grade 3 spellers, and good grade 2 spellers matched on a standardized spelling test, and a group of good grade 5 spellers matched by chronological age with the poor grade 5 spellers were administered a phonemic segmentation test containing nondigraph pseudowords and an experimental spelling test containing words of the following four types: exception, ambiguous, regular, and pseudowords. Consistent with the hypothesis, it was found that when compared with the poor spellers, the average and good spellers performed better on the phonemic segmentation task, made fewer errors in spelling pseudowords, and made spelling errors that were more phonetically accurate.

The relation between reading comprehension and linguistic awareness is discussed in terms of a set of processing skills common to both these activities. Two metalinguistic skill components are identified and examined for their relative contribution to reading comprehension. The general hypothesis is that these skill components explain a significant portion of the variance in reading comprehension. The two skill components are operationalized in terms of a battery of metalinguistic tasks, each of which is claimed (by means of task analysis) to rely primarily on one or the other processing component for its solution. A study is reported in which 8-year-old children are tested with these metalinguistic tasks, general measures of intelligence, and a test of reading comprehension. The results show that the relation among performance on the metalinguistic tasks is strongest for those tasks relying on the same processing skill component and that one of these components is most significant in determining the child's level of reading comprehension. These findings are interpreted in terms of the underlying linguistic processes involved in reading.

The spontaneous speech of both English-speaking (E) and Italian-speaking (I) children with specific language impairment (SLI) was examined to determine (a) whether phonological factors influence the grammatical morpheme use of ISLI children, as has been found for ESLI children, and (b) whether ESLI and ISLI children show similar syntactic abilities at the same level of mean utterance length as measured in words. The results indicated that word-final consonants adversely influenced the ISLI children's tendency to use articles – the only Italian grammatical morphemes in which word-final consonants are required. There was no evidence of syntactic differences between the ESLI and ISLI children. However, both groups of children seemed to have a problem using morphemes that constituted unstressed elements in a sentence even though the grammatical and semantic function of these elements varied across the two languages. The findings suggest that a speech production or perception component may be playing a greater role than previously believed in contributing to SLI children's well-documented expressive grammatical difficulties, though the specific effects of this factor will vary as a function of the surface characteristics of the language being acquired.

The relationship between bilingualism and metalinguistic awareness was explored in Puerto Rican Spanish- and English-speaking children. All subjects were from low-income backgrounds and were enrolled in a transitional bilingual education program in the United States. Two longitudinal studies were conducted. The first study examined the abilities to note and correct ungrammatical sentences in Spanish. Subjects were 104 children in first and second grade at the beginning of the study. They were followed over a period of 2 years. The second study looked at the ability to detect ambiguity in sentences, and to paraphrase the different meanings. There were 107 subjects who were in fourth and fifth grades, and were also followed over a 2-year period. The results from both studies indicated that native language proficiency as well as the degree of bilingualism affected metalinguistic awareness. The results also indicated that these effects interacted with the types of items in the metalinguistic tasks. This suggests that both subject-specific and task-specific factors are important in understanding the relationship between bilingualism and metalinguistic awareness.

On the basis of prior research and a theoretical analysis of picture naming, we examined three correlates of child picture naming difficulty: age of name acquisition, picture-to-name uncertainty, and name generality (instance or category names). Study 1 demonstrated that late age of acquisition and high adult name uncertainty were associated with increased naming difficulty for both single instance and category items from the Expressive One-Word Picture Vocabulary Test (EOWPVT) (Gardner, 1979). Age of acquisition was a particularly strong predictor for instance items. Moreover, category pictures were more difficult to name than single instance pictures with names of comparable age of acquisition. In Study 2, analysis of incorrect responses to category pictures demonstrated that children made more instance errors than adults, but not more errors of other kinds. Study 2 also showed that a child uncertainty measure correlated more strongly with EOWPVT item difficulty than the adult measure of Study 1. The findings suggest that picture naming scores reflect both the availability of the names in lexical memory and their accessibility, which, in turn, partly depend on the amount of interference from competing responses. Some implications for the interpretation and remediation of picture naming deficits are suggested.

This study investigated students' knowledge of derivational morphology and the relationship between this knowledge and their ability to spell derived words. The subjects (fourth, sixth and eighth graders) were given the Wide Range Achievement Test, Spelling subtest, and experimental tests of their ability to generate base and derived forms orally, to spell the same base and derived words, and to apply suffix addition rules. The results indicate strong developmental trends in both the mastery of derivational morphology and the spelling of derived words; however, spelling performances lagged significantly behind the ability to generate the same words. Success generating and spelling derived words depended on the complexity of transformations between base and derived forms. Further, mastery of phonological and orthographic transformations most strongly distinguished the three grades in both spelling and generating derived words. Indications that the older students were using knowledge of morphemic structure in spelling derived words were found in analysis of the spelling of base and derived word pairs and the application of suffix addition rules.

This article reports a single case study of a brain-damaged subject who presented with impaired performance on tasks involving production of verbal numbers (i.e., numbers in the form of words, such as eighty-six). Lexical errors occurred in spoken production but not in written production. Syntactic errors, however, occurred in both production modes. This pattern of performance provides empirical support for assumptions about the componential structure of the normal number processing system.

The relation between speech perception and reading ability was investigated. Identification and discrimination tests with stimuli on a stop consonant place of articulation continuum (/PA/-/IA/) were presented to various groups of good and poor readers (normal first graders on a reading continuum in Study 1, moderately and severely dyslexic and high- and low-achieving normal children in Study 2). In both studies, reading-related perception differences were especially marked in a comparison of actual and predicted (from identification test) discrimination scores, suggesting that the poor readers have special difficulty with a discrimination-specific task demand. In Study 3, reading-group differences still show up if the stimulus pairs of the discrimination test are simply to be repeated (eliminating the similarity judgment factors), and are also very pronounced on a nonspeech discrimination test (with 130 ms pure tones of slightly different frequencies). In Studies 1 and 3 identification slope and phoneme boundary differences between reader groups were found as well. Though these results do not definitely prove the auditory perception hypothesis, they support it. The conditions that should be satisfied by auditory perception tasks in future reading acquisition research are briefly discussed.

This study explored the conceptual status of metalinguistic ability by determining whether or not metalinguistic ability can account for variation in early reading achievement independently of more general language abilities. First-grade children were given a test battery assessing phonemic awareness, syntactic awareness, receptive vocabulary, syntactic proficiency, word decoding ability, and reading comprehension ability. Strong zero-order correlations were observed among all experimental measures. However, multiple regression analyses revealed that metalinguistic ability did not contribute to the prediction of early reading achievement when general language ability effects were statistically controlled.

An African gray parrot, previously taught by an interactive modeling technique to vocalize labels for several objects and instances of color and shape, was similarly taught to employ the vocalization “want” before an object label so as to discriminate functional “labeling” from “requesting.” The parrot's use of “want” subsequently generalized to novel situations. There was no decrease in the accuracy of labeling ability during training, a problem that arose when nonmodeling techniques were employed with chimpanzees. The results suggest that the procedure might be applicable to intervention programs for humans with specific communicative deficits.

Results of recent studies comparing the spelling errors of children with varying discrepancies between their reading and spelling skills have yielded conflicting results. Some studies suggest that good readers-poor spellers (mixed) are characterized by a set of deficits that differentiates them from poor readers-poor spellers (poor). Other studies fail to find differences between groups of poor spellers who differ in their reading skills. The present study attempted to determine the degree to which these discrepant results reflected differences in methods of subject selection and of error analysis. Two different sets of criteria were used to identify poor spellers-good readers. Subjects were selected on the basis of standardized reading comprehension and spelling test scores or on the basis of standardized single-word-recognition and spelling-test scores. The phonetic accuracy of the spelling errors was assessed using two different scoring systems – one that took positional constraints into account and one that did not. In addition, children were identified at two different age levels, allowing for developmental comparisons. Regardless of age or reading ability, poor and mixed spellers had difficulty converting sounds into positionally appropriate graphemes. Only older children with good word recognition but poor spelling skills provided some evidence for a distinct subgroup of poor spellers. These children had relatively good visual memory for words and, unlike other poor spellers, showed relatively good use of rudimentary sound-letter correspondences.

The goal of this article is to investigate whether the acquisition of some morpho-syntactic aspects in Italian deaf adolescents is simply delayed with respect to hearing children, or whether it follows significantly different developmental patterns. Twenty-five deaf students (age range: 11–15 years) and a group of 125 hearing controls (age range: 6–16 years) performed four tests, administered in written form, relative to different grammatical aspects: plurals, articles, and clitic pronouns. Results showed three different patterns of development depending on the grammatical aspect considered. Deaf children compared to hearing controls showed normal development in the pluralization task, delayed development in the pronoun task, and a qualitatively different pattern in the article task.

A longitudinal study was conducted of the effect of mothers' communication mode on the language development of 149 children with severe and profound hearing losses. The relationship of several subject and family characteristics to mode selection was also investigated. The children were between 3 and 5 years of age at the commencement of the 4-year study. Mothers' mode was coded from videotapes of structured and semistructured interactions. Findings indicated that children whose mothers used oral communication had higher scores on measures of spoken language, whereas children whose mothers used manual communication had higher scores on measures of receptive language and mother-child communication. Within the oral group, children whose mothers made extensive use of gestures had lower scores than children whose mothers did not. Within the manual group, children receiving limited sign input had lower scores on most measures than children receiving extensive input.