Foreword

Spoken languages, qua formal objects, are not restricted to a single phenotypic type. Rather, their grammars exhibit considerable structural variation that can be characterized, in traditional terms, as primarily isolating, agglutinating, or inflecting. Similarly, when languages are expressed in more or less permanent visual form (in cuneiform, or Cyrillic script, for example), their orthographies differ widely in the level or levels of linguistic form that are represented explicitly and (more or less) systematically.

Just as current syntactic theory attempts to specify parametric variation within the constraints of Universal Grammar, so a theory of orthography should characterize both the universal constraints and allowable variants that define the script-types accessible to human psychobiology. Just as the child can acquire any natural language (upon adequate exposure), so can he or she learn to use a variety of culturally evolved writing systems. Despite this fact, it is often argued that the acquisition of literacy is, in some sense, an “unnatural” skill. Scholars are prone to remind us that the history of “true” writing systems can only be traced back to about 3500 B.C.E. (although Schmandt-Besserat, 1980, has argued cogently for the existence of precursors dated some 5000 years earlier). The history of mass literacy (in the Western world) is considerably briefer, and over considerable areas of the globe literacy is even now the prerogative of a small, privileged class; even in societies where the educational system aims to achieve 100% literacy, a small but distressing minority of children experience great difficulty in learning to read and never achieve the fluent command of visual language skills that are necessary to participate fully in a technological society.

Small wonder, then, that linguists should so often have regarded visual language as the disadvantaged relative of spoken language, and that educators should periodically contemplate orthographic reform as a means of fitting the medium to the message. Without wishing to downplay the very real problems involved in achieving the goal of global literacy, there is a more positive and, to my mind, more interesting way of looking at the issue. Given adequate exposure, sufficient motivation, and decent teaching, the vast majority of children are perfectly capable of acquiring good reading and writing skills. Furthermore, they are capable of this feat when exposed to the orthographies of French or English, Arabic or Hebrew, Korean or Navajo, or any of the other script-types that can be found in five continents. Some children achieve an understanding of written language before formal schooling, and many have mastered the basic skills of literacy after only a year or two of instruction. No account of the psychobiology of written language can afford to ignore such basic facts. Written language cannot be dangerously close to the epistemological boundary of what the human brain is incapable of comprehending and manipulating.

Much of what we know about the structural organization of reading and writing skills and their material substrate in the brain has come from detailed studies of the fractionation and partial loss of written-language competence in previously literate adults who have sustained damage to the central nervous system (Geschwind, 1962). More recently, the paradigm of information-processing psychology has allowed for a considerable integration of studies of normal and impaired reading whereby pathological data are interpreted as the selective loss of particular modular components and mechanisms that underlie fluent, mature skill (Coltheart, 1981; Newcombe & Marshall, 1981; Patterson, 1981). Introduction of the technique of split visual-field presentation, whereby a stimulus can be projected initially to either the left or the right half-brain (Franz & Davis, 1933) has likewise served to revive theoretical interest in differential hemispheric specialization for the linguistic and visuospatial aspects of written language (Zaidel, 1983).

For obvious reasons, most of this work has been conducted with people whose native orthographies are the alphabetic scripts of West Europe and North America: English, French, Italian, and German, in particular. Whilst there are significant differences in the ways in which these languages are mapped into written form the structural characteristics of their orthographies fall within a very narrow band of the typological variants of writing systems. Outside of Europe and North America, only one other culture—that of Japan—has so far supported an extensive research tradition in which the patterns of selective loss and preservation of visual-language skills consequent upon brain damage have been evaluated in any detail. Much of the documentation of these deficits, written for the most part in Japanese, has, of course, been unavailable to Western students of the acquired dyslexias and dysgraphias. Much of the information that has percolated into Western neurolinguistics has been grossly oversimplified or downright misleading, as Westerners have typically failed to appreciate the full complexity of the Japanese writing system(s). Most of us know that Japanese is a mixed system, syllabic (kana) and “ideographic” (kanji), but we have all too often glossed over the richness of structural information that hides behind these seemingly innocent descriptions.

The development of a truly universal neurolinguistics of written language will for long be in debt to Michel Paradis, Hiroko Hagiwara, and Nancy Hildebrandt. For the first time in English, they have provided us with a full and lucid analysis of the fine grain of the Japanese writing system, and have summarized all the clinical reports of dyslexic fractionation in Japanese back to the turn of the century. This major monograph, linguistically informed, clinically astute, and organized with an eye to theory construction within an information-processing framework, is the firm foundation on which future work will build. It will, one hopes, serve as a model for further studies in which a similarly detailed analysis is undertaken of the neurolinguistic structure of other non-Western orthographies.

One particularly valuable aspect of Paradis, Hagiwara, and Hildebrandt’s work is that they have brought together in one monograph clinical studies of acquired dyslexia and experimental investigations of hemispheric specialization for written language in normal subjects. Our journals bear horrendous witness to how easy it is to run ill-thought and badly controlled experiments on laterality effects. The temptation to display a motley collection of kana and kanji in the two visual fields and to interpret the results in terms of some overreaching, oversharp dichotomy has often proved overwhelming. Western (and, to some extent, even Japanese) scholars have succumbed to conclusions that are simply unfounded because the stimulus materials have been confounded. Paradis, Hagiwara, and Hildebrandt provide a masterly summary and critique of this literature that should, at very least, provoke a dramatic increase in the level of sophistication that future experimentalists may reasonably aspire to.

Modern information-processing accounts of normal reading and pathologies of reading have a provenance of only a little over a decade; they were constructed on the basis of a very limited range of data from an even more limited set of writing systems. Despite the narrowness of this empirical base, it is most heartening to see that those models do appear to provide a theoretical framework within which to interpret the pathologies of reading and writing seen in the radically different orthographic system of Japanese. Whilst current models are computationally much underspecified, they do seem to be a step in the right direction. The systematization of Neurolinguistic Aspects of the Japanese Writing System that Paradis, Hagiwara, and Hildebrandt have achieved brings forcibly to our attention the absolute necessity of developing these models within the context of a universal theory of reading and writing. And universal in a linguistic context means, of course, that language- and culture-specific variation must be accounted for. There is every reason to hope and expect that the next decade of neurolinguistic research on writing systems will be even more exciting than the last, and that our theories will be considerably deepened by insisting that they be made responsive to data on a fair sample of the phenotypically distinct script-types that children can and do master “naturally.”