Language (General)

Introduction

The practice of language assessment, both in children and adult populations, has been undertaken from various perspectives, and its evolution mirrors to some extent the development of Psychology as a scientific discipline from the first decades of the twentieth century. In this regard, there are two most influential lines of thought, generally known as the psychometric and the cognitive approaches. Since there are a number of background theoretical issues concerning the nature of language as an object of scientific inquiry that have an important bearing on either approach to language assessment, we will begin by briefly addressing these theoretical issues. This will lead us to describe the main strategies used for language assessment purposes, and to review the main components or processing levels of language and the tasks that are used in adult language assessment for each component, together with the major variables that should be taken into account in the assessment process. Finally, we will review some problems, both theoretical and methodological, that assessment procedures have to face.

Theoretical Background

Language is a very peculiar object of study. It is both a declarative body of knowledge possessed by adult competent speakers, and a set of procedures (or abilities) by which such knowledge is put to use in a variety of ways in linguistic activities. Furthermore, language can be viewed primarily as a means of communication among conspecifics (human language being the most developed and sophisticated code), but also as a means of representing and conveying thoughts and intentions, as a symbolic tool or device relating sound and meaning. In this regard, the psychological study of language is at least a twofold enterprise, for it must address (1) a wide array of information types and processing levels involved in understanding and speaking (which in principle can be selectively impaired); and (2) the intimate connection between the [Page 534]speakers' linguistic knowledge and abilities, on the one hand, and their cognitive and communicative capacities at large, of which linguistic skills are but a subset.

The psychometric approach to language assessment (e.g. Burt, 1940; Carroll, 1941; Hakstian & Cattell, 1978; Thurstone, 1938; Vernon, 1950) views language as a set of performance skills that rest on a number of underlying, more or less permanent, abilities. Although this view of language seems to parallel the competence-performance distinction proposed by Chomsky (1965), it does not carry any commitment to a rule-based account of linguistic competence or an information-processing view of the cognitive operations underlying linguistic performance. Rather, it defines ‘verbal’ (as opposed to ‘linguistic’) abilities in crudely operational terms; that is, as a direct reflection of the subject's performance in a number of standardized linguistic tasks. Language, more properly called ‘verbal ability’, is thus seen as a factor (or number of factors) of intelligence along which subjects may show quantitative variations. Verbal intelligence is alternatively viewed as a unitary ability, or as a set of distinct factors (e.g. ‘verbal comprehension’, ‘verbal fluency’, etc.) that can be independently evaluated. From this perspective, the targets of language assessment, or the components of verbal ability, are defined by crossing the main modalities of language use (spoken vs. written language) with the major linguistic tasks (comprehension and production), rendering the four basic language skills, namely listening, speaking, reading and writing.

...