Introduction

Matthysse and Parnas (1992) demonstrated that studies which only investigate overt schizophrenia as the informative phenotype are bound to result in failures to detect linkage, even if linkage is present. We suggested the use of the spectrum concept of schizophrenia and of putative extraclinical measures, such as eye tracking disorder, in order to enhance the likelihood of detecting existing linkage. The question is how to delimit the range of phenotypes which may be informative in genetic studies of schizophrenia. Many researchers, including the author (Parnas et al., 1993), subscribe to the diathesisstress model of schizophrenic etiology (e.g., Meehl, 1962, 1989), elegantly formulated by Minkowski (1927) half a century ago:

Introduction

Few today would contest that genes are implicated in the etiology of schizophrenia, but general agreement ends at that broad conclusion. The extent of the genetic influence, the mode of genetic transmission, and the identification of specific etiologically relevant genes are still uncertain and are all areas of active research. Holzman and Matthysse (1990) have recently reviewed the genetics of schizophrenia with a perspective that reflects concerns that should be incorporated in research studies on schizophrenia. Others have also recently reviewed the genetics of schizophrenia (e.g., Diehl and Kendler, 1989; McGue and Gottesman, 1989; Gelernter and Kidd, 1990; Gottesman, 1991). Briefly, genetic studies have evolved from studies designed to show that genes are involved to studies designed to illuminate how genes and environment interact in the etiology of schizophrenia (see Figures 24.1 and 24.2). Unfortunately, the studies based on distribution of illness among relatives of schizophrenics, although giving results consistent with a substantial genetic component to etiology, have failed to provide definitive evidence of what the genetic factors are. As the diverse genetic hypotheses have become more sophisticated, it is either impossible to discriminate among them or any apparent discrimination is based on some assumption that, however plausible, remains unproven. For example, Risch (1990) concludes that genetic variation at a single locus cannot explain the recurrence risks observed in various classes of relatives.

Introduction

Schizophrenia is a major mental illness that affects 1% of the general population and is extremely costly to the patient, family, and larger community. Unfortunately, its etiology is as yet unknown, and for this reason it is categorized as a “functional” psychosis rather than an “organic” psychosis, a category which implies that it arises from no known structural or pathological alteration of the brain. The role of brain dysfunction, however, in the etiology of schizophrenia seems likely in light of recent brain structural and functional studies, and has been suspected since Kraepelin (1919/1971) and Bleuler (1911/ 1950) first delineated the syndrome(s). Kraepelin, in fact, believed that the symptoms of schizophrenia, which he called Dementia Praecox, would ultimately be linked to abnormalities in both the frontal and temporal lobes. He believed that the frontal lobes were responsible for the disruption in reasoning so clearly evident in schizophrenia, while the temporal lobes were responsible for auditory hallucinations and delusions. Other workers at the close of the nineteenth century, such as Crichton-Browne, quoted above, as well as Alzheimer (1897), Kahlbaum (1874), and Hecker (1871) also believed that to understand the etiology of severe mental illnesses, such as schizophrenia, an understanding of the brain was necessary.

The history of scientific disciplines is marked by one or more theoretical transformations that were made possible by the introduction of novel sources of evidence. Physics provides the best examples. The invention of apparatus that permitted measurements of the varied energy profiles emitted by different forms of matter motivated the creation of quantum mechanics and new conception of both the infinitesimal space of the atom and the near infinite expanse of the cosmos. The invention of the amplifier, electrode, and chemical stains for neurons provided surprising information about the brain that required concepts very different from those used by nineteenth-century scholars.

Developmental psychology and the psychology of personality have not often enjoyed the extraordinary advantages of new sources of evidence. The primary data in most studies of the development of personality and pathology remain verbal replies, spoken in interviews or written on questionnaires, that rely on the nontechnical language of everyday conversation. The use of physiological data, direct behavioral observations, or performance on specialized laboratory procedures is relatively rare in developmental studies. Even in psychiatric nosology, referential meanings of diagnostic categories are usually restricted to phenomenological information. The concept social phobia, for example, means that a patient has told a clinician that he experiences severe discomfort when he has to meet new people, avoids such social situations as much as possible, and, in addition, is unhappy with this aspect of his life.

Detailed reviews of the literature on the development of speech, language, and communication skills in persons with DS and other mental handicaps are readily available (e.g., Gunn, 1985; Rondal, 1985a, 1988a, 1988b; J. Miller, 1987; C. B. Mervis, 1988; Dodd & Leahy, 1989; Barrett & Diniz, 1989). The following presentation, therefore, will only summarize major data. (For an update on the general problems of DS envisaged in a developmental-organizational perspective, see the contributions gathered by Cicchetti & Beeghly, 1990, including their own review chapter.)

A general indication in the specialized literature has sometimes been for DS to be more detrimental to language development than are other syndromes leading to mental retardation (e.g., Zisk & Bialer, 1967; Gibson, 1981). Exact reasons and detailed supporting analyses for this claim have never been provided, however, and there are opposing views (e.g., Evans & Hampson, 1968). Nevertheless, a number of observations may be recalled (also J. Miller, 1987) in relation to the important speech and language deficits in DS persons. None of them are pathognomonic of DS, but their accumulation in this syndrome is remarkable.

Language-related deficiencies in Down syndrome

Two series of problems may be envisaged: first, organic malformations and difficulties affecting speech; second, central processes of a cognitive nature.

The nature of the relationship between language and thought has been of interest to philosophers, psychologists, and linguists for a long time. One will recall the position identified in recent times as the Whorfian hypothesis, associated with the names of Sapir (1921) and Whorf (1956). According to this position, and roughly put, the language that we speak directly affects the way we think. Over the years, the Whorfian hypothesis has taken two basic forms. The stronger form asserts that the categories and boundaries of language determine the categories and boundaries of perception and cognition. Consequently, people speaking unrelated languages should perceive and conceive the world in different ways. The weaker form posits that language orients our thinking in certain directions but does not actually force our thoughts to take specific forms. The Whorfian hypothesis has generated a great deal of research over the past decades in anthropological, linguistic, and psychological circles, leading to the conclusion that only its weaker form is at least partially acceptable (see R. Brown, 1968; see Bloom, 1981, for a different point of view, and R. Brown, 1986, for a reply). But even this weaker form did not go unchallenged. The opposite viewpoint, that thought processes are ontogenetically prior to language and that they constitute developmental prerequisites for language development, has found a grounding in the work of Piaget and followers. I will not consider here in detail the Vygotsky-Luria position on the problem because it pertains more particularly to the role of language and language representations in cognitive development and functioning.

The preceding review of studies on language development in DS may have left the impression of a relative uniformity in the patterns of development. Indeed, research on language development in DS children has most often been conducted from an implicit theoretical stance assuming basic uniformity. Few data bear on the question of variation in language acquisition and functioning in DS subjects (see J. Miller, 1988, for a significant exception). In so doing, this subfield of psycholinguistics, as it applies to DS, mirrors the general tendencies that can be observed in developmental psycholinguistics at large. Much research on language in the NR children has been concerned with the supposedly invariable course of development, not only for particular languages, but also at a more abstract level and between languages (e.g., Slobin, 1985). A so-far limited number of studies have shown, however, that there is considerable inter – and intraindividual variation in rate or quantitative aspects of language acquisition as well as noticeable differences in patterns and/or styles of acquisition (see Peters, 1983; Goldfield & Snow, 1985; Bates, Bretherton, & Snyder, 1988; Bates & MacWhinney, 1987, for reviews and discussions). Observing and explaining variations in patterns of development is important because it may indicate the extent to which language acquisition may be constrained by a number of factors, such as the characteristics of the target language, the physical and the verbal environment in which development takes place, and of course, the child.

As reviewed in Chapter 3, there is a small number of documented cases of mentally handicapped individuals with exceptional language capacities despite otherwise severe cognitive limitations. The case of Françoise is particularly striking among these cases, and so far it is undoubtedly one of the most thoroughly studied. This case, as do the others, begs for an explanation. How is such a state of affairs possible? Before trying to approach this difficult question, let me review the extent to which linguistic development of the type exhibited by Françoise is truly exceptional in comparison with other (typical) mentally handicapped persons.

Exceptional and nonexceptional language development in mental handicap

As I have demonstrated in what precedes, Françhise's nonlinguistic cognitive abilities in general are only moderately better than what is known of most Down syndrome persons. Her performance IQ on the WAIS is 60 (64 at the second administration) and she is credited with a nonverbal MA of 5 years and 8 months on the EDEI scale. This performance IQ is higher than the reported mode of 45–50 for the typical trisomy 21 DS population (Moor, 1967; Gibson, 1981). However, her nonverbal MA is almost compatible with the average MA of the DS population, which is about 5 years (again according to Gibson, 1981). Although cursorily presented as nonverbal, the performance section of the Wechsler Scales actually is saturated with speech (e.g., the instructions in each subtest are given verbally), which, of course, may favor those subjects with a good receptive command of the language. This may account, at least partially, for the difference between Françhise's nonverbal IQ and the population modal IQ for typical trisomy 21 DS.

I met Françoise, my Down syndrome (henceforth DS) subject, a few years ago, at La Fermette, an occupational center for DS adults managed by APEM (Association de Parents d'Enfants Mongoliens), to which I am a consultant. La Fermette is located 30 km from Liège, in the eastern part of Belgium. Upon conversational contacts in the center, Françoise's productive language as well as her apparent excellent comprehension struck me as unusually close to normal for a DS person. In searching the specialized literature, I found that the case appeared unique enough to justify undertaking a full-scale study.

Françoise was recorded during 2 hours of free conversation with one member of my research team. She was also given a very large number of tests, verbal and nonverbal intelligence, perceptual, mnemonic, and so forth. Retrospective information was obtained on her development, education, and life since birth. This book is the product of this compilation and of the analysis and interpretations made of the data. It addresses several central questions.

First, is normal or quasi-normal language development and functioning possible in spite of severe central cognitive limitations, particularly conceptual ones? The answer, as it would appear, is partially a positive one. Normal phonological and grammatical functioning are indeed possible in such cases.

Second, how can this be the case? One possible answer, and the one that is favored in what follows, is that phonology and grammar may be acquired and function relatively independently from other central processes. Current indications in the specialized literature suggest the hypothesis of a loose tie between general cognition and particular components of the language system, such as phonology and syntax.

This intriguing study of a Down syndrome subject by Professor Rondal has a wellgrounded basis in the very early studies of thought and language and their interconnections by the two outstanding psychologists Piaget (1923) and Vygotsky (1929/1962) who both addressed the problem of the relationship between language and thought. At that time, the interaction of the two was seen in simple developmental terms in such a manner that, for example, thought could at first occur in the absence of language and subsequently speech or language could develop in the child in such a way that at a certain stage “thought could become speech and speech thought,” as Vygotsky expressed verbal development in children. However, two aspects to these kinds of studies distinguish them from more recent studies of the interaction of speech and thought. They do not reflect the complexity of current psycholinguistics, on the one hand, and they do not study abnormalities of language development, on the other. An exception to the latter point are the studies of Luria (1961), in which linguistic deficiencies in the control of behaviour are referred to, and in some instances Luria suggests that weakness in the control of motor behaviour in some oligophrenic children can be compensated for and brought under better control through the supplementary use of speech. In this kind of study, Luria makes use of Pavlov's differentiation of the primary (conditioning) and secondary (speech) signalling systems. The major difference between the two systems is, of course, the additive but slow action of conditioning as compared with the immediacy and generality of the secondary (speech) system.

The subject

Framçoise was 32 to 36 years old during the time of the study. Her IQ and MA are supplied in Section 5.8. She was born on April 5, 1955, 3 weeks before term, weighing 2.750 kg. Delivery was relatively rapid and normal but for some degree of cyanosis due to the umbilical cord being wrapped around the neck. Additional oxygen was given on an intermittent basis for 24 hours following birth.

The Down syndrome condition was not diagnosed until Françoise was 3 years old, at which time the parents consulted at the Pediatric Clinic of the Liège University Hospital for her slow development and complete lack of speech. A karyotype was made at the time revealing a genotype 47, XX, + free 21 in each of the metaphases studied (standard trisomy 21). In the course of the present study, suspecting a possible subetiology of mosaicism, I requested a new karyotype. It was made in July 1988 and confirmed the first one.

The mother of Françoise was 35 years old at the time of birth, and the father 32. They already had two normal children, one boy and one girl, respectively aged 5 and 7 years. The socioeconomic level of the family is upper middle class. Françoise made her first attempts at walking unassisted around 2 years. She was toilet trained at 4 years. The parents did not report any sensory problem at the time, but no systematic examination of her visual and auditory acuity was made. The only word she was able to pronounce at 4 years was /to/ for couteau – knife. Little information is available on her language receptive ability at the time.

The productive and receptive language functioning of Françoise, our DS subject, was thoroughly examined and found to be normal or close to normal in its phonological and grammatical aspects. It proved more compatible with what can reasonably be expected from a moderately MR person with regard to other language subsystems, such as lexicon, semantics, and pragmatics. Françoise's awareness of aspects of her receptive and productive language (including phonological and grammatical aspects) were extremely limited. Corresponding data available from studies conducted by Curtiss, Yamada, Cromer, Bellugi, and others, were also examined. The empirical evidence clearly supports Chomsky's (1981) proposal according to which computational aspects of language, that is, phonology and syntax, are autonomous components, largely independent from general cognitive abilities, whereas conceptual aspects of language, that is, lexicon, semantics, and pragmatics, are more dependent on cognitive functioning. As established, converging evidence exists in the socalled delay-difference literature on language development in regular mentally handicapped subjects when various aspects of language development are systematically compared with respect to mental age.

The data reported and analyzed concerning language-exceptional MR individuals are strong counterindications for any theory attempting to causally relate advanced phonological and grammatical development to general cognitive variables. In particular, the theoretical suggestions coming from the work of Piaget as to the existence of specific ties between operational development and grammatical development were found wanting in many respects. Another cardinal indication to be dismissed on the basis of the present work is the idea that children, particularly MR children, develop linguistically to the extent that they have been trained and taught.

Coming to know about the physical world is a long process that begins early in infancy and is not completed for years. The child's contact with the physical world comes via the senses, and thus the first set of questions has to do with the quality of the infant's perceptual capabilities. Perception is not the end of the matter, though – the infant and child also have to acquire the skills to engage in motor and locomotor interaction with the environment. Throughout, the child gradually gains the ability to conceptualize the physical world through thought, rather than just acting upon it. A large and important part of this growing ability to conceptualize is the perception and understanding of the spatial aspects of the physical world. In Part I we address each of these major issues in separate chapters.