. . . discussions of reliability no longer refer back to Edgeworth (1888, 1890), or of [the] validity of essay-marking to Sir Phillip Hartog (Hartog and Rhodes, 1935, 1936), or of the problem of scaling to the elderly Thorndike's dream of an absolute scale of language proficiency (Monroe, 1939), nor do our criticalists cite the impassioned attacks on the 'encroaching power' of examinations expressed by Henry Latham (1877). (p. 12)Dismissing the notion of overall language proficiency as a "will-o-the-wisp" (p. 10) or "chimera" (p. 14), Spolsky refrains from mentioning theoretically what it might mean to know a language. He lauds the efforts by the Council of Europe to develop a linguistic framework of reference, yet concludes, "In practice of course, it is not more validated than any other scale is . . . and is as easily translated into rigidity" (p. 15). The central question in Spolsky's essay - which resonates throughout this book - seems to be "how to value results and translate them into interpretations" (p. 16). At this point widespread consensus regarding that quandary still seems elusive.
[ p. 17 ]
focus on any one of these approaches . . . to the exclusion of others, will lead to potential weakness in assessment itself, or to limitations on the uses for which the assessment is appropriate. (p. 41)Regarding theoretical foundations, Bachman reminds readers of the interactive nature of language testing: assessment seldom occurs in isolation and the notion of "competence is itself co-constructed and shared by participants, and context-bound" (p. 60). With almost clinical precision, Bachman points out the shortcomings of current interactionist perspectives. The nature between interaction, construct, and performance is still under debate. Part of the difficulty in arriving at a balanced perspective has to do with at times conflicting agendas within the language testing field: (1) to promote theoretical research, and (2) to develop useful real-life assessments (p. 66). In other words, applied linguists seeking to explore theoretical constructs often find themselves at odds with others more concerned about cranking out and practically validating actual proficiency tests.
[ p. 18 ]A stronger predictor of future academic performance, according to Davis, is present academic performance. This forces us to examine what tests such as the IELTS might in fact be measuring. Reputedly a measure of academic language proficiency, Davies acknowledges that the theoretical underpinning of what is known as "academic English" is somewhat shaky:
While academic language is taken for granted as a construct, attempts to describe it as a single domain raise even greater doubts than those which query the unitary nature of academia. Do science, music, the humanities, engineering, and dentistry all share some idea of knowledge and investigation or do we just assume they do because all are studied and researched in universities? And for us, the harder question: do they all have a language in common which is different from other languages? (p. 74)
[ p. 19 ]Shohamy contends that language tests are never neutral since they shape instructional priorities and language hierarchies (p. 150). She points out some of the problems with the U.S. No Child Left Behind program as well as the disturbing nature of many citizenship exams. Although Shohamy urges the creation of tests which "are in line with broader and more realistic language constructs" that "incorporate multilingualism and multi-modal realities" (p. 141), she avoids offering concrete details about how such a test might actually be constructed.
The idea behind . . . [national citizenship tests] is the belief that language proficiency, as exemplified through these tests, is an expression of loyalty and patriotism and should be a requirement for residency, and especially for citizenship" (p. 149)Shohamy seldom leaves space for alternative hypotheses and the lack of evidence for some of her assertions left me rather askance. However, if readers regard her writing as a persuasive essay designed to foster action rather than a dispassionate analysis of some applied linguistic point removed from any controversy, the realization may come that we are fortunate to have someone like Shohamy in our field.
– Reviewed by Tim Newfields