Language Competence Across Populations

Toward a Definition of Specific Language Impairment

Yonata Levy,Jeannette C. Schaeffer

• Publisher: Psychology Press
• Year: 2003
• ISBN:
  9781135642549
• Edition: 1
• Topic: Personal Development
• Subtopic: Writing & Presentation Skills

This unique, edited book bridges studies in language disorders and linguistic theory with timely contributions from leading scholars in language development. It presents an attempt to define Specific Language Impairment, relating it to children of normal and disordered language capabilities. The chapter presentations examine language development across a variety of populations of children, from those with Specific Language Impairment to second language learners. The contributors discuss criteria for the definition of SLI, compare and contrast SLI with profiles of children with other disorders and dialects, and offer a comprehensive look at the Whole Human Language, which ties together spoken and signed languages. Methodological concerns that affect the credibility and generalizability of the findings are discussed and controversies between opposing linguistic approaches to language acquisition are presented. The conceptual thread that gradually reveals itself as the chapters unfold is a theoretical issue of central importance to cognitive theory, as well as to our understanding of the biological correlates of language--it concerns the variability that linguistic competence can manifest in children under different biological conditions and life circumstances. Language Competence Across Populations: Toward a Definition of Specific Language Impairment is an essential volume for advanced students and scholars in linguistics and psychology who have an interest in language acquisition and language disorders, as well as for the clinical professionals dealing with children with language impairments.

Part A

Language Competence Across Populations

Section I

The Characterization of Specific Language Impairment

An Introduction

Jeannette Schaeffer

Ben-Gurion University of the Negev

Within the general theme of the present volume, language competence across populations, this first section takes a special focus. It concentrates on one particular population, namely, children with specific language impairment (SLI). Specific language impairment has been studied for several decades, and is viewed by many researchers as a disorder distinct from other (language) disorders. However, as several contributions in this volume show, it is not always clear in what sense the language characteristics of children with SLI are different from those of children with other disorders. Thus, it is important to investigate the differences and similarities in linguistic behavior between the various populations. This provides more insight into the possible specificity of the disorders, as well as into the general constraints on language development. In order to maintain the hypothesis that SLI is a distinct disorder, it is necessary to create a set of exclusionary and inclusionary criteria on the basis of which the disorder can be isolated. The chapters in this section contribute to this goal.

Another focus concerns the crosslinguistic aspects of SLI. Originally, descriptions and theories of SLI were based mostly on data from English SLI. However, a recent wave of crosslinguistic research has challenged certain traditional characterizations of SLI. Severed chapters in this section add in an important way to the recharacterization of SLI because languages other than English (e.g., Dutch, French, Hebrew, and Inuktitut) come under investigation.

A few questions help in organizing thoughts on the characterization of SLI: How can anyone know if a child is specifically language impaired? And what does specific language impairment mean exactly? Despite decades of research on language impairment, there is still no consensus on a definition of SLI.

A first step is to decide in what kind of terms to define SLI. Would it be useful to formulate a definition on the basis of theoretical models of linguistics and/or cognition? Should there be a list of symptoms (description)? Or should the definition have some explanatory value as well? Should the aim be for one general, crosslinguistic definition, or should the definition be customized according to (groups of) target languages? Are there subtypes of SLI? If yes, on the basis of which factors should these subtypes be distinguished?: Comprehension versus production? The different components of language, for example, the lexicon, syntax, semantics, morphology, phonology, pragmatics?

Leonard (1998) listed several inclusionary and exclusionary criteria for the diagnosis of SLI. Although diagnostic criteria are not the same as a definition, they could help in formulating a definition. The inclusionary criterion of a significant deficit in language ability is the most straightforward. A diagnosis of a language problem can usually be made with confidence. The trick is to distinguish SLI from other disabling conditions of which language problems are a part. Furthermore, a “language problem” can manifest itself in many different ways (recall the different language components: lexicon, syntax, etc., and the specifics of the target language). The exclusionary criteria require normal behavior in areas other than language, such as nonverbal IQ, hearing, brain structure, speech organs, and physical and social interactions.

Although the exclusionary criteria narrow down the number of candidates for SLI considerably, and are therefore very useful, the one single inclusionary criterion remains vague, and brings things back to the main question: Which parts/forms of language must be impaired in order to be diagnosed with SLI? When there are answers to these questions, then an adequate set of test batteries can be developed to serve as selectional criteria. The chapters in this section show that by keeping alive the interaction and exchange of information between models of linguistic theory and crosslinguistic data studies on SLI, it is possible to refine the inclusionary criterion and therefore come closer to a more detailed definition of SLI.

First consider linguistic theoretical models that distinguish different components within language. At least three major components of language can be identified: the lexicon, the computational system (Chomsky, 1993), and the pragmatic system. The computational system can in turn be subdivided into syntax, semantics, morphology, phonology (grammar), and the mechanism that processes their information (parser). Approaching SLI from this angle, then, a limitation in language ability can be lexical, syntactic, semantic, morphological, phonological, or pragmatic in nature, or a combination of two or more of these components.

The studies presented in this section all concern the question of whether these linguistic components are the exact ones that can help in an understanding and investigation of SLI, or whether the pie should be cut slightly differently. Whereas Wexler, Crago & Paradis, Van der Lely, and Schaeffer approach the problem from a (generative) linguistic theoretical angle, Rice, de Jong, and Ravid and colleagues are more concerned with the description of the symptoms of SLI. Furthermore, several authors propose to divide SLI up into subgroups, either according to the target language (Ravid et al. and Crago & Paradis) or to homogeneity (Van der Lely).

In addition, there is the crosslinguistic issue. Some authors in this section focus on English, the traditional language of investigation—namely, Wexler, Rice, Van der Lely. On the other hand, other chapters discuss other languages: Schaeffer and de Jong investigate Dutch; Crago and Paradis deal with French and Inuktitut; and Ravid and colleagues look at Hebrew. This provides for an interesting comparison between the manifestations of SLI in different languages.

The progress made in this section concerns the original inclusionary criterion of “language deficit.” It is an attempt to formulate in much more detail what kind of language deficit characterizes children and adults with SLI, both in terms of symptoms and linguistic theory. A closer look at the individual chapters illustrates these issues.

Wexler (chap. 1) concentrates on the relation between linguistic theory and SLI. He argues that, in principle, the study of linguistic theory and the study of language development (including disordered language development) could and should mutually influence each other. If something goes wrong with language development (different from “normal”), then it should reveal something about the nature of normal language development, which in turn tells something about linguistic theory. Conversely, without a description of normal linguistic abilities given by linguistic theory it would be difficult to ask the question of what goes wrong.

The empirical issues that Wexler discusses include the rates of finiteness (optional infinitives, or OIs), properties of agreement and case, and the development of object clitic pronouns. He shows that errors concerning these phenomena occur both in normal child language, and in the language of children with SLI, with one difference—namely, the error stages last much longer for children with SLI. Wexler suggests a syntactic constraint of a specific type that is genetically determined and is accountable for the observed delays. The upshot of his account with respect to the search for a definition of SLI is linguistic-theoretical: SLI is caused by a maturational delay of a specific syntactic constraint.

From a more bird’s-eye point of view, Rice (chap. 2) also emphasizes the importance of the mutual influence between the study of (impaired) language development and linguistic theory. In most children, the emergence of morphology and syntax occurs at the same time as other advances in social and cognitive development. The co-occurrence of multiple developmental dimensions complicates the determination of the extent to which these dimensions are truly distinct or codependent. Children who display unexpected discrepancies in their acquisition of language provide valuable insights into the ways in which language can be spared or affected, relative to other developmental dimensions. The exact nature of the discrepancies can further an understanding of the linguistic system, and how it emerges in young children.

The two types of discrepancies Rice discusses are timing in the emergence of language acquisition and the trajectory of change over time, and expected and unexpected variation across children. On the basis of data with English-speaking children (normal and SLI), she concludes that children with SLI are delayed in the onset of speech and their (morphosyntactic) developmental stages are protracted (timing), and they show high variability on linguistic measures that do not detect variation for unaffected children (variation). Thus, Rice suggests that the definition of SLI should include the terms timing and variation.

Crago and Paradis (chap. 3) illustrate the interaction between linguistic theory, acquisition theory, and the characterization of SLI by comparing varying language learners (L1, L2, and SLI) acquiring a variety of languages (French, English, Inuktitut) on a variety of grammatical topics. They address the following questions: Does the type of language learned affect the pattern of impairment seen in children with SLI? Are impaired learners different from all other learners of L1 and L2? Conversely, are L2 learners distinct from L1, both normal and impaired? Or, do commonalities exist across all learners at certain stages of development? What can account for the shared characteristics?

Crago and Paradis show that just as data from children learning structurally different languages continue to reshape language acquisition theories and linguistic theory, a broader range of cross-learner data (e.g., SLI in languages other than English and normal L2 development) leads to further refinements of these theories. This chapter underscores the significance of testing acquisition theories over a wider set of learners in order to develop such theories, as well as to better understand the groups of children on which they are based. Furthermore, comparing the end state grammars of the SLI population to the grammars of adults learning a second language helps elucidate both maturational claims regarding language development as well as certain recent claims in the L2 literature that second language learning adults are “impaired” speakers. Specifically, Crago and Paradis show that nonimpaired second language learners of French produce OI during a period of around 3 years, after age 7. This is not predicted by Wexler’s claim that the protracted period of OI in SLI speech (extended optional infinitives, EOI) is accounted for by the delayed maturation of certain UG principles, as maturation of these principles is not supposed to take place after age 7 in normally developing populations.

They also show that SLI in a morphologically rich language such as Inuktitut cannot be explained by Leonard’s processing account, because the Inuktitut child with SLI studied by Crago and Paradis fails to produce systematically even stressed morphemes and those that have a CVC structure. Thus, Crago and Paradis’ study is a major example of the importance of testing hypotheses that emerged from studies on English SLI in other languages.

As Van der Lely (chap. 4) points out, current research on SLI shows that the language deficits found are rather heterogeneous, even if only one language— English—is considered. She examines the relation between SLI in English-speaking children who show relatively pure language impairments with those who evince coexisting (subtle) cognitive deficits. In addition, she considers the linguistic differences in children with SLI. At the broader level, the distinction between different SLI forms appears clear; for example, children with pragmatic SLI show little overlap in linguistic characteristics with those children with grammatical impairment (cf. Bishop, 1997, who argued that pragmatic disorders may also coexist with grammatical impairment). At a narrower level, the distinction between different forms of SLI become less clear, that is, with respect to the relation between variable deficits within grammar (e.g., syntax, semantics, morphology, or phonology). Although there appear to be some children with primary phonological impairment, thorough investigation of all these areas within grammar are not typically carried out with the same rigor in the same population of children. Even at the narrowest level of characterization, that is, within syntax or morphology, differences between studies are also apparent with respect to impairments in tense, agreement, or more general core aspects of the human syntactic faculty that involve movement. These differences are subsequently reflected in the resulting hypotheses characterizing or accounting for the SLI performance. Whereas some differences between studies may reflect the focus of the investigations in different labs, or the interpretation of data rather than the data themselves, there do appear to be some genuine, qualitative differences that call for theoretical discussion.

Van der Lely formulates the heterogeneity problem as follows: What is the relation between the different forms of SLI? Are there multiple disorders that result in different impairments that can variably coexist? Or is there a single disorder that, for some unexplained reason, variably manifests itself in children? As she notes, there is some empirical and theoretical justification for considering that deficits in phonology versus morphosyntax have a different origin—particularly if consideration is given to the adult language disorder literature. However, it is by no means clear that this model is appropriate when considering development, as Van der Lely sensibly points out.

By studying highly selected subgroups with clearly defined linguistic characteristics, Van der Lely addresses some of the problems mentioned here. She argues that the core deficit responsible for the grammar of children with grammatical-SLI (G-SLI) is in the syntactic mechanism of “Move,” the movement of syntactic constituents from their base-generated position to their surfac...