Dan J.  Stein, M.D.






Dept of Psychiatry

College of Physicians and Surgeons, Columbia University,

and the New York State Psychiatric Institute,

722 W 168 St

NY, NY 10032







Running Title:  Cognitive Science and Psychiatry


Tel:  212-960-2355


































     Cognitive science is a multidisciplinary field, comprising cognitive


     psychology, artificial intelligence, linguistics, neuroscience, and


     anthropology.  In recent years, cognitive science has become a


     predominant paradigm in studies of the mind.  This paper reviews work


     at the emerging interface between cognitive science and psychiatry.


     It is argued that cognitive science has significant potential as an


     integrative framework for theorizing and researching psychiatric


     disorders and their treatment.














































     Cognitive science is a relatively new field that incorporates concepts


     and methods from philosophy, cognitive psychology, artificial


     intelligence, linguistics, neuroscience and anthropology (Gardner,


     1985; Posner, 1989).  Whereas behaviorism dominated the psychological


     sciences during the first part of this century, cognitive science has


     become a central paradigm of the latter part.  This "cognitive


     revolution" was fostered by the promise of cognitive science as an


     integrated and fertile approach to the mind.  Indeed, departments of


     cognitive science have become important loci of interdisciplinary


     research and have generated a prodigous literature.




     During the period of the cognitive revolution, significant changes


     also took place in psychiatry (Lipowski, 1989).  During the 1950s,


     psychoanalysis comprised a dominant school in many departments of


     psychiatry.  In subseqent decades, however, a community-focused model


     was widely touted.  Most recently, neurobiological approaches have


     been given pride of place.  Certainly, there have been calls for a


     synthetic or "biopsychosocial" approach to psychiatric disorders and


     treatment.  Nevertheless, there has been little progress on the


     construction of an integrative framework for such an approach.




     On the whole, cognitive scientists have paid little attention to


     clinical phenomena, while clinicians have in turn only occasionally


     employed the concepts and methods of cognitive science.  In recent


     years, however, dialogue between the two fields has increased (Breger,


     1969; Mahoney and Freedman, 1985; Ingram, 1986; Horowitz, 1988a;


     Williams, Fraser, MacLeod, and Mathews, 1988; Magaro, 1991; Stein and


     Young, 1992a).  In this paper I review research at this intersection,


     and discuss its potentials and limitations.  I suggest that cognitive


     science can provide an important integrative framework for psychiatry.








     What exactly is cognitive science?  The following three sections


     attempt to address this question.  I briefly discuss descriptions and


     definitions of cognitive science, its origins and subdisciplines, and


     then contrast and compare cognitive science with behaviorist,


     psychoanalytic and neurobiological models (Stein, 1992a).




     a)  Descriptions and Definitions




     In general, cognitive scientists are interested in mental structures


     and processes, their representational significance, and their physical


     instantiation (Stillings, Feinstein, Garfield, Rissland, Rosenbaum,


     Weisler, and Baker-Ward, 1987).  Several authors have attempted more


     rigorous definitions of cognitive science.  A cognitive


     information-processing account, for example, views the mind as an


     information-processing system that selects, transforms, encodes,


     stores, retrieves, and generates information and behavior (Lachman,


     Lachman, and Butterfield, 1979).  A computational view, on the other


     hand, emphasizes that, "Cognitive science, sometimes explicitly, and


     sometimes implicitly, tries to elucidate the workings of the mind by


     treating them as computations, not necessarily of the sort that is


     carried out be the digital computer, but of a sort that lies within


     [the] broader theory of computation" (Johnson-Laird, 1988, p. 9).




     Restrictive definitions of cognitive science may, however, include


     only one or other of the divergent models that cognitive scientists


     have developed.  Early cognitive scientists viewed the mind as a


     sequential processor, similar to the early digital computer.  The mind


     was seen as a passive recipient of information, which was registered


     in a short-term memory, and perhaps encoded in a long-term memory.


     More recent cognitive scientists have, however, pointed out that the


     mind is a parallel processor (Rumelhart, Hinton, and the PDP Research


     Group, 1986), and have emphasized that mental structures are active


     and that they occur within a particular context (Neisser, 1976;


     Lakoff, 1987).  Such work may be excluded by a definition of cognitive


     science that focuses solely on information-processing and computation.




     On the hand, too broad a definition of cognitive science may prevent a


     constrast between cognitive science models and other models often used


     by clinicians.  It may be useful to at least specify the broad


     categories into which cognitive science models fall.  Cognitive


     science models typically specify cognitive architecture in one of two


     ways, symbolic and connectionist.  The elements of symbolic systems


     are symbols, which are stored in associative structures.  Symbolic


     systems include levels-of-processing models (Craik and Lockhart,


     1972), spreading activation constructs (Collins and Loftus, 1975), and


     schema approaches (Neisser, 1976).  The elements of connectionist


     systems are simplified and schematized neurons which are


     interconnected in a network.  Again, a variety of these parallel


     processing models have been developed (Rumelhart et al., 1986).  In


     subsequent sections the application of both symbolic and connectionist


     architectures to clinical theory and practice is reviewed.




     b)  Origins and Subdisciplines




     Further understanding of cognitive science can be gained by discussing


     its origins.  The development of cognitive science models was


     encouraged by various factors, including the failure of behaviorism,


     the invention of the computer, and various theoretical advances


     (Gardner, 1985).  One of the most important of these theoretical


     advances was the development of computer science.  The father of this


     field was Turing (1936), a British mathematician who described a


     simple machine (the Turing machine) that executed instructions in


     binary code, and proposed that such a machine could in principle


     perform any computational task.  He also described the idea of a


     universal Turing machine which takes a coded version of other Turing


     machines as input and then emulates their behavior.




     Computer science can immediately be seen as relevant to psychology,


     for the question arises of whether the human mind can be instantiated


     on a universal Turing machine.  If so, computer science would


     constitute a basic science for exploring the hypothesis (Craik, 1943)


     that the mind is a symbol manipulating device.  Turing (1950)


     suggested it might be possible to program a machine so that a user


     communicating with the machine and with a person would be unable to


     differentiate the two (the Turing machine test).




     A number of other developments consolidated the importance of


     computational constructs for psychology.  Thus, it was proposed that


     certain concepts were useful in explaining both computing machines and


     human brains.  This was clearly formulated at a seminal meeting in


     1948, the Hixon Symposium on "Cerebral Mechanisms in Behavior"


     (Jeffress, 1951) at which McCulloch, a professor of psychiatry, and


     Von Neumann were the opening speakers.  For example, the all-or-none


     property of neuronal activation can be compared with the determination


     of Boolean statements as either true or false.  Furthermore, neuronal


     networks and Boolean statements can be described in elecrical terms as


     current that passed or failed to pass in a circuit.




     In addition, it was argued that certain constructs were useful in


     explaining both computing machines and human minds.  Hebb (1949), for


     example, suggested that synaptic strengthening in neuronal networks


     led to learning.  Wiener (1947) proposed that machines and minds that


     had feedback mechanisms that displayed purposefulness.  Central to


     control and communication engineering was the notion of the message,


     whether this was transmitted by electrical, mechanical, or neural


     means.  Information, Shannon (1938) showed, was independent of its


     physical instantiation.




     It has been suggested that cognitive science was founded in September


     1956, at the Symposium on Information Theory at MIT (Gardner, 1985).


     Significant research presented at the conference included papers by


     the cognitive psychologist, Miller, artificial intelligence workers,


     Newell and Simon, and the linguist, Chomksy.  In his paper, Miller


     illustrated the importance of empirical studies of cognition by


     discussing work on the constraints of short-term memory processes.


     Work by other psychologists in the 1950s, including Broadbent and


     Cherry in Britain, and Bruner and colleagues at Harvard, gave further


     impetus to the development of cognitive psychology, and established


     the discipline and its empirical methodologies as a cornerstone of


     cognitive science.  Newell and Simon argued that artifial intelligence


     was possible, and drew comparisons between artificial and human


     problem solving processes.  Together with Minsky and McCarthy they


     pioneered the field of artificial intelligence, and helped establish


     the importance of a computational methodology for cognitive science.


     Finally, Chomsky described his theory of a grammar based on linguistic


     transformations.  This work, together with his scathing review


     (Chomsky, 1959) of Skinner's work on verbal behavior buttressed the


     shift from behaviorism to cognitivism, and established linguistics as


     a useful frame from which the issues of cognitive science could be






     Other core disciplines of cognitive science include neuroscience,


     anthropology, and philosophy.  Neuroscience comprises a lower limit


     for cognitive science, for human cognitive structures and processes


     are ultimately based in and constrained by neuroanatomy and


     neurochemistry.  Anthropology and social psychology constitute an


     upper limit for cognitive science, for cognitive structures and


     processes vary from place to place and time to time.  Finally, many of


     the questions studied by cognitive scientists were first raised by


     philosophers.  Philosophers have also addressed the foundational


     principles of the discipline, and more recently, as cognitive science


     has advanced, have been faced with the meta-cognitive question of what


     the mind and the world are like for this scientific advance to have


     taken place.




     Developmental psychology is not usually listed as a central discipline


     of cognitive science.  Nevertheless, one of the most important


     pre-cognitivists was the developmental psychologist Piaget.  Piaget


     detailed the transformation of mental structures from sensorimotor


     reflexes to the operations of formal thought (Piaget, 1952).


     Contemporary dialogue between developmental psychology and cognitive


     science is increasing (Sternberg, 1984; Meltzoff, 1990; Globerson and


     Zelniker, 1989).






     c)  Contrasts and Comparisons




     Cognitive science models can be contrasted and compared with the


     clinically familiar models of behaviorism, psychoanalysis, and


     neurobiology.  Behaviorism, for example, self-consciously makes the


     mind into a "black box", asserting that only observable stimuli and


     responses can be studied.  In contrast, the cognitive model holds that


     what is most interesting are the mental structures in the "black box"


     and the processes (operations) whereby they generate cognitive


     products (thoughts and feelings).




     Nevertheless, it is possible to see a continuity between behaviorist


     and cognitive science models.  While the behaviorist model is limited


     to inputs (stimuli) and outputs (responses), the cognitivist model is


     concerned with inputs, processing, and outputs.  Like behaviorists,


     cognitivists have adopted an empirical stance, which emphasizes the


     importance of careful measurement and laboratory experimentation.




     The classical psychoanalytic model of the mind was an energy-based


     one.  For Freud (1894), "in mental functions something is to be


     distinguished - a quota of affect or sum of excitation - which


     possesses all the characteristics of a quantity (though we have no


     means of measuring it), which is capable of increase, dimunition,


     displacement and discharge, and which is spread over the memory-traces


     of ideas somewhat as an electric charge is spread over the surface of


     a body".  Freud described how the forces of the unconscious are


     expressed, transformed, or repressed, resulting in everyday behaviors


     and psychopathology.




     Again, however, it is possible to demonstrate a continuity between


     psychoanalysis and cognitive science.  Cognitive science and


     psychoanalysis both focus on the structures of the mind, and the way


     in which these determine mental phenomena.  Furthermore, in Freud's


     later work he describes affect not in terms of energy, but in terms of


     its role as a signal.  Finally, post-Freudian psychoanalysts,


     including the self psychologists, the objects relations school, and


     the interpersonalists, have increasingly employed cognitively oriented


     constructs such as self and other representations (Stein, 1992b).




     Neurobiological models attempt to explain the biological basis of


     mental processes and products.  Proponents of these models may state


     that all psychological explanations are reducible to neurobiological


     ones.  This view contrasts with the cognitive science argument that


     cognitive phenomena necessarily require psychological explanations.


     Psychological events are emergent phenomena, and psychological


     explanations are not reducible to biological ones.




     Once again, however, it is possible to see a continuity between the


     biological model and cognitive science.  Earlier neuroscience was


     described as comprising a lower limit of cognitive science.  While


     some cognitivists are interested in information-processing in only the


     most abstract sense, many others are interested specifically in how


     information-processing occurs in the human neurobiological substrate.


     This interest is particularly apparent in connectionist work informed


     by neuroscience (Rumelhart et al, 1986).








     For the clinician, the question immediately arises of whether


     cognitive science models are useful in accounting for clinical


     phenomena.  However, standard texts of cognitive science make little


     reference to psychiatric disorders and treatment.  Nevertheless, a


     number of developments in cognitive science have been instrumental in


     allowing a bridge to clinical phenomena.  This section briefly reviews


     some of these bridges from cognitive science to the clinic, including


     research on emotional and unconscious processing, and relevant work in


     each of the subdisciplines of cognitive science.




     Cognitive science has been characterized by a disregard for emotion


     (Gardner, 1985).  Nevertheless, humans are not only the most


     intelligent of the animals, they are perhaps also the most emotional


     (Hebb, 1946).  Furthermore, it may be suggested that this relationship


     between intelligence and emotionality is a necessary one.  In a


     seminal paper, Simon (1967) argued that emotion in humans is


     comparable to the prioritized interruption of different processes by


     one another in complex artificial intelligence systems with multiple


     goals and limited resources.  The idea that cognitive design problems


     are solved by emotional processes remains popular in contemporary


     cognitive science attempts to theorize emotion (Oatley and


     Johnson-Laird, 1987; Sloman, 1987), and has helped generate a


     rapidly growing empirical literature on cognitive-affective


     processing.  Cognitivist work on emotion has subsequently been


     employed in order to understand affective experience and change in the


     clinic (Mandler, 1975; Greenberg and Safran, 1990).




     Another important area within cognitive science that has clinical


     relevance is the study of unconscious processing.  Research on the


     unconscious is not new, and includes the pioneering work of Helmholtz,


     Janet, and Freud (Ellenberger, 1970).  However, it was not until the


     emergence of cognitive science that the "cognitive unconscious"


     (Kihlstrom, 1987) became a respectable area of study.  Research has


     focused on such areas as selective attention, subliminal perception,


     implicit memory, hypnotic suggestion, and dreaming (Erdelyi, 1985;


     Foulkes, 1985; Kihlstrom, 1987).  While this work immediately brings


     psychoanalysis to mind, there appear to be important differences


     between the cognitive unconscious and the psychoanalytic unconscious


     (Bowers and Meichenbaum, 1984).  According to Freud, the unconscious


     is a set of drives, affects, and motives that has an organization and


     content more primitive than those of the consciousness, but that


     nevertheless has latent the potential for interacting with other


     elements of the psyche.  According to cognitive science, the


     unconscious is a set of cognitive processes, including attitudes and


     predispositions, that act prior to consciousness, so actively


     organizing and ordering experience and behavior.  Nevertheless,


     contemporary work on unconscious processing establishes a dialogue


     between cognitive and clinical scientists (Bowers and Meichenbaum,


     1984; Eagle, 1986; Horowitz, 1988b; Safran and Greenberg, 1986; Uleman


     and Bargh, 1989; Singer, 1990; Prigatano and Schachter, 1991; Cloitre,


     1992; Stein, 1992b).




     A number of early cognitivists have been interested specifically in


     psychiatric disorders.  Ruesch and Bateson (1968) were among the first


     to apply cybernetic concepts to psychopathology and psychotherapy.


     Within cognitive psychology, Hilgard (1977), whose work will be


     discussed later, employed Janet's notion of dissociation and developed


     a neodissociative model of the dissociative disorders.  This work


     exemplified the potential utility of employing cognitive psychology


     constructs to conceptualize psychiatric disorders.




     In artificial intelligence, Colby (1981) pioneered the investigation


     of clinical phenomena.  He developed a program, PARRY, which


     incorporated a model of the paranoid process, and succeeded in


     simulating a paranoid patient.  The difficulty of such work, and


     Colby's achievement, is illustrated by the limited amount of


     subsequent research in this area (Tomkins and Messick, 1963;


     Clippinger, 1977).  More recently, however, a number of authors have


     employed computer implemented neural networks to model psychopathology


     and psychotherapy (Hoffman, 1987; Hestenes, 1991; Cohen and


     Servan-Schreiber, 1992l; Williams and Oaksford, 1992; Caspar,


     Rothenfluh, and Segal, unpublished manuscript).




     Linguists, neuroscientists, anthropologists and social psychologists


     have long been interested in clinical phenomena, and some of this work


     has fallen within the realm of clinical cognitive science.


     Cognitivist models of the aphasias have their origins in the work of


     the "diagram-makers", who constructed modular models of the mind at


     the turn of the century (Morton, 1984).  Current cognitive studies of


     language may contribute to various clinical areas, such as treatment


     of dyslexia (Swerling and Sternberg, 1992), or understanding clinical


     dialogue and the mechanisms of the "talking cure" (Oatley, 1992).


     Contemporary cognitive neuropsychologists continue to hold that study


     of neurologically impaired patients contributes to cognitive theories


     of mind (Caramazza, 1992).  Clinical anthropology has tended not to


     adopt cognitivist models (Stein, 1992c), but cognitively oriented work


     in social psychology has been particularly relevant to clinical


     science, and has tackled such important concerns as self and other


     representations, and their role in mood disorders (Markus, 1977;


     Cantor and Kihlstrom, 1981; Sorrentino and Higgins, 1986; Westen,


     1988).  Finally, it may be argued that philosophy, too, constitutes an


     important foundational discipline for clinical cognitive research


     (Mahoney, 1991; Lyddon, 1988; Stein, 1992d).




     Developmental psychologists have long been interested in developmental


     arrests and abnormalities.  Since Piaget, cognitive concepts have


     increasingly been used in this area.  Developmental research that


     draws on cognitive science constructs and focuses on clinical


     implications has been particularly influential (Greenspan, 1980;


     Stern, 1985; Emde, 1983).








     The work of cognitive scientists interested in clinical phenomena


     provides the basis for employing cognitive constructs to conceptualize


     psychiatric disorders, assessment, and treatment.  I will discuss each


     of these areas in turn.  While a variety of psychiatric disorders have


     been studied from a cognitive science perspective, I will focus on


     only a few in order to highlight some important methods and issues in


     clinical cognitive science.




     a) Major Depression




     Since their introduction by Ellis, Beck and others, cognitive


     therapies have become increasingly accepted as an effective treatment


     of depression (Mahoney and Freeman, 1985).  While these therapies are


     not necessarily formal extensions of cognitive science, much of this


     work has been influenced by the cognitive revolution.  The growing


     dialogue between cognitive science and cognitive models of depression


     (Ingram and Reed, 1986; Ingram and Wisnicki, 1991) provides the most


     comprehensive exemplar of the use of symbolic architectures from


     cognitive science to understand psychopathology.


     Symbolic architectures specify the mind in terms of cognitive


     structures, operations, and processes.  Cognitive structures that have


     been hypothesized to be central in depression include associative


     networks (Ingram, 1984; Teasdale, 1983) and self-schemas (Beck, 1967;


     Kuiper, Derry, and MacDonald, 1982).  Associative network theories


     characterize depression as having increased availability and/or


     accessiblity of negative constructs about the self.  Self-schema


     theory characterizes depression as activation of a self-structure that


     is negative in content.  While evidence for negative self-schemas in


     depression is good, there is less evidence that such schemas are


     present in the absence of current depression (Ingram and Reed, 1986;


     Ingram and Wisnicki, 1991).




     Research also suggests that cognitive processes in depression are


     characterized by a negative selectivity bias (Ingram and Reed, 1986;


     Ingram and Wisnicki, 1991).  Studies of encoding may be divided into


     studies of personal semantic information (i.e.  information related to


     concepts of self) and episodic information (i.e.  information related


     to autobiographical events).  Studies of semantic encoding indicate


     that enhanced negative encoding is a key factor in information


     processing in depression, while studies of episodic encoding indicate


     that less accurate encoding of positive information is critical.


     Studies of retrieval, on the other hand, suggest that negative


     information is more efficiently and positive information is less


     efficiently retrieved in depression.  In addition, depressed patients


     selectively monitor negative information with a decrease in attention


     to positive stimuli.




     Finally, cognitive products in depression include attributions and


     automatic thoughts (Ingram and Reed, 1986; Ingram and Wisnicki, 1991).


     Depressed patients are more likely to make dysfunctional attributions


     as to the cause of events.  Negative automatic thoughts appear to be


     characteristic of subclinical and clinical depression, and they remit


     with treatment.




     Cognitive science models of depression influence our conceptualization


     of psychotherapy, and direct further empirical research.  The finding,


     for example, that depressed schemas are not present in the absence of


     depression suggests that cognitive therapy works not by structural


     changes in underlying schemas, but by changes in the activation level


     of negative self-schemas.  While cognitive structures such as schemas


     are difficult to measure, methodological progress has been made


     (Segal, 1988).  Considerable theoretical and empirical research on


     cognitive science models of depression nevertheless remains to be


     done.  In particular, cognitive science models of depression have,


     with few exceptions (Klein, 1976), failed to incorporate


     neurobiological knowledge.




     b) Dissociative and Conversion Disorders




     Symbolic architectures, including schema theory, have not only been


     adopted by the cognitive therapy tradition, but have also been


     employed in psychoanalytically informed research (Wachtel, 1982; Slap


     and Saykin, 1983; Segal, 1988; Horowitz, 1988a; Horowitz, 1991).  The


     area in which there is perhaps most clearly an intersection between


     the constructs of symbolic architecture and psychodynamic theory is


     the study of dissociative and conversion disorders.




     The first detailed psychological theory of dissociative and conversion


     disorders was developed by Janet (1907).  Like Freud, Janet studied at


     Charcot's clinic at the Salpetriere.  He too was interested in the


     notion of cognitive processes that took place outside of awareness,


     for which he coined the term "subconscious".  Janet described the mind


     in terms of "automatisms", or elementary structures controlling


     experience, thought, and action in various domains, and which together


     make up the flow of consciousness.  At times of stress or exhaustion,


     one or more automatisms separate from the rest, breaking the unity of


     consciousness.  The dissociated automatisms continue to operate, but


     are no longer accessible to phenomenal awareness or amenable to


     voluntary control, resulting in various dissociative or conversion






     Hilgard (1977) used concepts from cognitive psychology to provide a


     neodissociative account of these disorders.  He described the mind in


     terms of subordinate cognitive structures each with a degree of


     autonomy, a hiererarchical control managing the competition between


     these structures, and a central monitoring and controlling structure


     (the executive ego).  Subordinate cognitive structures can operate


     outside of awareness, or they may be processed consciously through the


     executive ego.  Certain physiological (e.g.  anesthesia) or


     psychological (e.g.  hypnosis) conditions lead to disruptions in the


     connections between the subordinate structures.  Information


     processing in the subordinate structures may then take place outside


     of phenomenal awareness or voluntary control.  More recently,


     Kihlstrom (1990) has used an associative-network model of memory to


     further elaborate this view of the dissociative and conversion


     disorders.  He describes how disruption of links between


     representations of memory or of sensory input and representations of


     self may result in lack of awareness of particular memories or


     sensations, with continued processing of related representations.


     While psychoanalytic theory has also emphasized the importance of


     unconscious processes in the genesis of these disorders,


     neodissociative theory has a view of the unconscious that is


     closer to that of cognitive science.  Although there may be


     theoretical difficulties with the psychoanalytic energy based view,


     the success of the cognitivist revival of Janet's ideas ultimately


     depends on its ability to generate empirical work on these relatively


     unresearched disorders.




     c) Obsessive-Compulsive Disorder




     Earlier the question of the relationship between cognitive science and


     neurobiological models was raised.  This becomes particular relevant


     for the clinical cognitive science of disorders about which


     neurobiological knowledge is advancing.  Current research on OCD, for


     example, has demonstrated involvement of the serotonergic system and


     orbitofrontal-basal ganglia-thalamic pathways, and the disorder is


     increasingly seen as having a neurobiological basis (Stein and


     Hollander, 1992).




     Cognitive science constructs may, however, be helpful in thinking


     through the relationship between neurobiological findings and the


     psychology of OCD.  Several authors have, for example, suggested that


     neuroanatomical structures involved in OCD play a role in


     goal-feedback mechanisms (Gray, 1982; Rapoport and Wise, 1988).


     Disruption of cybernetic control in OCD results in psychological


     features such as repetitive symptoms and strategies of seeking control


     (Pitman, 1987).




     Work on neuropsychiatric impairment and serotonergic dysfunction in


     OCD has led Stein and Hollander (1992) to propose that OCD may involve


     a biologically based impairment in determining or in evaluating


     goal-feedback responses.  Some patients, perhaps those with high


     neurological soft signs, visuospatial difficulties and rapid erroneous


     responses on neuropsychological testing, may have difficulty in


     determining goal-feedback responses.  Other patients, perhaps those


     with serotonergic dysfunction and slow correct responses on


     neuropyschological testing, may overevaluate harm associated with


     goal discrepancy.




     This model suggests that both biological approaches (normalizing


     serotonergic function) and psychological approaches (changing harm


     assessment) may be helpful in the treatment of OCD.  On the other


     hand, in patients where goal-feedback response impairment is


     associated with evidence of structural brain damage, treatment may be


     less successful.  The model provides a heuristic for further research


     correlating biological, cognitive, and clinical variables in OCD.




     d) Psychotic Disorders




     Sincle the time of Kraepelin and Jung, psychiatrists have used


     techniques from experimental psychology to explore


     information-processing deficits in psychotic disorders.  More


     recently, cognitive science methods have been used to study these


     disorders.  There is now a large literature on information-processing


     in schizophrenia (Magaro, 1980; Neuchterlein and Dawson, 1984;


     Saccuzzo, 1986), and some work on mania (Grossman and Harrow, 1991).


     In their review, Neuchterlein and Dawson (1984) highlight two types of


     information-processing deficits in schizophrenia - poorer detection of


     single, highly familiar stimuli during vigilance, and increases in the


     time required to recognize an uncomplicated stimulus.


     Information-processing tasks may constitute indicators of


     vulnerability, may have prognostic value, and may be useful measures


     of medication response.  Information-processing deficits in


     schizophrenia may also be conceptualized in schema terms (Magaro,


     1980).  Finally, a number of theorists have attempted to integrate


     information-processing and neurobiological findings (Neuchterlein and


     Dawson, 1984).




     Of considerable current interest, however, is work that employs


     connectionist architectures from cognitive science to study psychotic


     disorders.  This work is of particular interest insofar as


     connectionist models succeed in incorporating neurobiological


     knowledge.  In an early paper Hoffman noted that different


     perturbations of a particular kind of network (the Hopfield network)


     led to different forms of neural dysfunction (Hoffman, 1987).  Memory


     overload led to disturbances similar to those seen in schizophrenia,


     while increased randomness led to disturbances similar to those seen


     in mania.  Neural dysfunction in perturbed Hopfield networks is,


     however, only loosely similar to that seen in psychosis, and the


     neurobiological viability of these perturbations is arguable.




     A more recent connectionist model is that of Cohen and


     Servan-Schreiber (1992).  Their neural network simulates performance


     of schizophrenic patients on tasks related to the processing of


     context.  Furthermore, their network incorporates the neurobiological


     finding that dopamine has a modulatory effect on prefrontal cortex by


     influencing the responsivity, or gain, of cells in this region.  Thus,


     their model simultaneously posits that schizophrenia is characterized


     by a particular information-processing deficit with a specific


     neurobiological underpinnning, and provides a novel and robust 17


     methodology to explore this deficit.




     e) Other Disorders




     There are several other disorders which can be understood in terms of


     cognitive science constructs.  Space considerations preclude a


     discussion of anxiety disorders (Rapee, 1991; Eysenck, 1991), autism


     (Litrownik and McInnis, 1986), impulsive (Barratt, 1987) and


     antisocial (Gorenstein, 1991) behavior, learning disability (Swerling


     and Sternberg, 1992), memory disorders (Shimamura, 1989), personality


     disorders (Stein and Young, 1992b), and substance abuse (Hull and


     Reilly, 1986).








     Conceptualizing various psychiatric disorders in terms of cognitivist


     constructs immediately raises the question of employing cognitive


     science in clinical assessment.  Cognitive science has assessment


     constructs that differ from those of more traditional frameworks


     (106).  In general, cognitivists attempt to use empirical methods to


     assess the particular psychological structures and processes


     responsible for generating mental events.  Many of these methods may


     be useful in clinical assessment, for example, of neuropsychological


     impairment (Delis, Kramer, Fridlund, Kaplan, 1990), or of maladaptive


     schemas of self and other (Young, 1990; Horowitz, 1991).




     The variety and range of paradigms and techniques developed by


     cognitive scientists holds promise for their application to clinical


     measurement (Merluzzi, Rudy, and Glass, 1981).  It may however be


     necessary to modify cognitive science laboratory methods before


     employing them in the clinic.  For example, early


     information-processing studies in patients with depression did not to


     use stimuli with negative and depressive content, and therefore failed


     to elucidate relevant cognitive structures and processes (Ingram and


     Reed, 1986).




     On the other hand, the clinical setting may allow for more precise


     assessment than is available to most cognitive scientists.  For


     example, measurement of schemas in personality disorder by self-report


     questionnaires is problematic, as patients may not be aware of or may


     avoid their schemas.  However, during psychotherapy schemas are


     explored in various ways and links are drawn between the patient's


     behaviors and these schemas (Young, 1990).  In this way psychotherapy


     techniques allow assessment to move beyond laboratory based cognitive


     science methods.








     How can a cognitive science approach to psychiatric disorders and


     their assessment contribute to psychiatric treatment?  Clearly, the


     reformulation of any particular disorder in cognitivist terms will


     lead to conceptualization of the treatment of that disorder in the


     same terms.  Thus, the paper has already alluded to the treatment of


     negative self-schemas in depression, and the treatment of maladaptive


     person schemas in personality disorder.  Perhaps more important,


     however, is the idea that cognitive science contributes to the theory


     of treatment integration.  In recent years there have been frequent


     calls for psychotherapy integration and for the integration of


     psychotherapy and pharmacotherapy, but there has been little work on


     the development of theoretical frameworks that allow such an


     integration to take place.




     It is notable, however, that within contrasting schools of therapy,


     there have been shifts to an increasingly cognitive orientation.  The


     cognitive-behavioral treatment tradition has moved from behaviorist


     models to linear cognitive models to more complex cognitive models


     (Dobson, 1988).  Psychoanalysis has moved away from energy based


     models to more cognitivist ones, and several psychoanalysts, beginning


     with Blum (1961) and Peterfreund (1971), have pointed to the value of


     cognitive science for psychoanalytic meta-theory (Clippinger, 1977;


     Erdelyi, 1985; Westen, 1988; Horowitza, 1988; Stein, 1992b).  Finally,


     neurobiological findings are increasingly being incorporated into


     clinical cognitive science models.




     Schema theory, for example, has allowed both cognitive-behaviorists


     (Beck, 1967; Young, 1990; Beck, Freeman and Associates, 1991) and


     psychoanalysts (Horowitz, 1988a, 1991; Klein, 1976; Wachtel, 1982;


     Slap and Saykin, 1983) to focus on mental structures, their biological


     basis, their development and change, and on the way in which they


     direct psychological events (Stein, 1992e).  Similarly, neural network


     models may allow an integrative conceptual approach to these issues


     (Caspar, Rothenfluh, and Segal, unpublished manuscript; Forrest,


     1991).  Insofar as cognitive science has conceptual continuities with


     other approaches, but does not have their weakness (e.g., emphasizes


     empirical methods but avoids reduction of mind to stimulus-response


     chains, focuses on mental structures but avoids an energy based


     meta-psychology), it may have theoretical advantages.  Furthermore,


     conceptual reformulation may also foster technical integration.


     During psychotherapy, for example, schema change can then be


     undertaken by means of affective, cognitive, behavioral, and


     interpersonal interventions (Young, 1990).








     A number of advantages of clinical cognitive science are immediately


     apparent.  First, it offers integrative theoretical models of clinical


     disorders, assessment, and treatment.  Whereas so much contemporary


     clinical science may be criticized for being either mindless or


     brainless (Lipowski, 1989), clinical cognitive science offers a


     multifaceted approach that incorporates and extends earlier paradigms.


     In addition, it employs a number of methodologies which encourage


     empirical research.  Clinical cognitivists have researched clinical


     phenomena outside the office, and have introduced new scales and


     measures into the clinic.  Conversely, the study of clinical phenomena


     may enrich cognitive science (Stein, 1992a).  Nevertheless, a number


     of criticisms of clinical cognitive science may also be offered.  The


     foundations of cognitive science have been extensively debated in its


     subdiscipline of philosophy; here I will focus on more clinically


     oriented issues.




     A first objection states that clinical cognitive science is overly


     general or that it is "old wine in a new bottle".  The psychoanalytic


     clinician may feel that he or she uses cognitive science concepts and


     methods, but simply employs an older terminology.  Clinicians with a


     cognitive-behavioral orientation may feel that clinical work is


     primarily a pragmatic exercise and that it is unnecessary to introduce


     extraneous theoretical baggage.  Thus cognitive science may be seen as


     only reframing cinical theory, rather than advancing it.  I would


     argue, however, that more accurate reframing constitutes an essential


     part of scientific progress.  Psychoanalysts who reject energy based


     models, for example, are faced with the choice of discarding


     metapsychology or of viewing psychoanalysis as a hermeneutic rather


     than scientific enterprise.  In providing a possible theoretical


     conceptualization of psychodynamic work, cognitive science immediately


     goes beyond mere translation of older terminology (Stein, 1992b).


     Similarly, while cognitive therapists and cognitive scientists could


     conceivably work independently of one another, reframing clinical


     phenomena in terms of cognitive science may have immediate advantages.


     Successful reframing leads to more comprehensive and powerful models,


     and encourages novel methods of treatment and research (Stein, 1992e).




     A converse objection is that cognitivist and computational models are


     overly restricted or insufficiently complex.  The word "cognitivist"


     itself points to a lack of focus on emotion.  Computational machines


     operate along automatic and rigid lines, whereas humans have


     context-dependent and fallible minds.  Clinicians with a strong


     neurobiological or sociocultural orientation may doubt the ability of


     cognitivist models to include biological or sociocultural findings.


     Clinicians with an experiential or phenomenological orientation are


     particularly likely to question the value of a computational model in


     understanding human perceptions, feelings, and experiences.




     I would suggest that this criticism is partially correct and partially


     incorrect.  The criticism is wrong insofar as it suggests that the


     object of cognitive science models is to replicate human minds.  It is


     a truism that there are differences between computers and humans, but


     it does not follow that computational models cannot help us understand


     human minds.  This distinction can be illustrated by the debate


     between Colby (1977), the author of PARRY, and Weizenbaum (1976), who


     developed a program called ELIZA, that simulated a psychotherapy


     session.  Colby is a strong supporter of clinical cognitive science.


     However, when Weizenbaum noted that people took his program as an


     indication that psychotherapists could be replaced by computers, he


     emphasized the limitations of cognitive science, noting that computers


     should not be allowed to take over certain human tasks.  However, the


     issue is not whether computers can or should replace minds, but rather


     whether cognitive science models can and ought to be used to


     conceptualize and investigate minds.  It is noteworthy that whereas


     PARRY was based on a specific model of paranoid processes (and


     therefore directly contributed to the science of psychopathology),


     ELIZA was not a model of psychotherapy per se (and did not make a


     direct contribution to the science of psychotherapy), but rather a


     study of the rules of dialogue.




     What is right about this criticism is that it points to the complexity


     of human minds.  Furthermore, early cognitivist models conceived of


     the mind as a linear information-processor and may have contributed to


     the development of overly simplistic clinical cognitive models.


     However, current cognitive science models and clinical cognitive


     models have become increasingly sophisticated.  Gardner (1) has


     described these developments in terms of the "computational paradox";


     namely, that it took the early formal cognitivist models to describe


     and elaborate the differences between the serial digital computer and


     the human mind.  These advances are increasingly apparent in clinical


     cognitive science as well.  Thus clinical cognitive science is


     beginning to address both the neurobiological and the sociocultural


     limits of cognitive processing.  Furthermore, it may be argued that


     clinical cognitive science models have begun to succeed in


     incorporating both positivist concerns with empirical measurement and


     hermeneutic concerns with meaning and context (72).  The complexity of


     human minds therefore constitutes a challenge rather than an a priori


     objection to clinical cognitive science.




     An important criticism of clinical cognitive science that employs


     symbolic architectures involves questions of theoretical precision.


     Cognitive science models of mental structures and processes vary in


     important ways, and their adoption by clinicians has led to further


     diversification.  For example, whereas Horowitz's (1988a) division of


     schemas into motivational, role, and value schemas is consistent with


     Freud's division of the psyche into id, ego, and superego; Beck and


     colleagues' (1991) division of schemas into cognitive, affective, and


     action schemas is consistent with their view of the linear progression


     of thoughts, feelings, and behaviors.  Furthermore, a variety of


     related concepts such as scripts, self-systems, personal constructs,


     plans, and frames are popular in the cognitive and clinical


     literature.  This diversity may in part reflect the theoretical


     fertility of cognitive science, and may be helpful in providing a


     range of ideas and models.  Nevertheless, the diversity also reflects


     a lack of agreement about the taxonomy of representation, and a lack


     of detailed knowledge about the complexities of human cognitive


     architecture.  Clearly, further work is necessary to clarify and


     consolidate the theoretical apparatus of clinical cognitive science.




     Theoretical diversity contributes to the difficulties that researchers


     have experienced in measuring constructs such as schemas.  Problems in


     measurement are to some extent expectable with research on posited


     entities which underly observable behavior.  Nevertheless, there are


     important unresolved complexities in determining how best to approach


     the processing of information that differs in target, favorability,


     content, congruence, and personal relevance (Ingram and Wisnicki,


     1991).  Detailed clinical observation and advances in assessment


     techniques will be necessary to resolve these methodological


     difficulties (Horowitz, 1991).




     Clinical cognitive science that employs connectionist architectures


     has emerged only recently.  This field has the potential advantages of


     methodological precision and of close links with neuroscience.


     However, current paradigms sometimes appear removed from clinical


     experience and the dialogue between connectionism and neurobiology


     remains rudimentary.  Further work is necessary to narrow the gap


     between behavior of neural nets and clinical phenomenology, and to


     integrate mechanisms of neural net perturbation with knowledge of


     pathogenic mechanisms in biological psychiatry.








     The interdisciplinary nature and rapid development of cognitive


     science is of immediate interest to psychiatry.  In this paper, I have


     reviewed areas in which cognitive science and clinical science have


     established a dialogue.  Clinical cognitive science appears to provide


     an integrative and sophisticated framework for conceptualizing and


     researching psychiatric disorders and their treatment.  Further


     theoretical and empirical work needs to be done to consolidate this


     early work.  Nevertheless, there has been sufficient progress to


     suggest the promotion of cognitive science as an important theoretical


     framework for psychiatry, and to encourage further exploration of this












     Barratt, E.S.  (1987).  Impulsiveness and anxiety:  Information


        processing and electroencephalograph topography.  Journal of


        Research in Personality, 21:453-463.




     Beck, A.T.  (1967).  Depression:  Clinical, Experimental, and


        Theoretical Aspects.  New York:  Harper and Row.


  Beck, A.T., & Freeman, A., and Associates (1991).  Cognitive Therapy



         of Personality Disorders.  New York:  Guilford.



 Blum, G.S.  (1961).  A Model of the Mind.  New York:  Wiley.




 Bowers, K.S., & Meichenbaum, D. (eds) (1984).  The Unconscious


        Reconsidered.  New York:  Wiley.



 Breger, L. (ed) (1969).  Clinical-cognitive Psychology:  Models and


        Integrations.  Englewood Cliffs, NJ:  Prentice-Hall.




     Cantor, N., Kihlstrom, J. (eds) (1981).  Personality, cognition, and


        social interaction.  Hillsdale, NJ:  Erlbaum.




     Caramazza, A. (1992).  Is cognitive neuropsychology possible?  Journal


        of Cognitive Neuroscience, 4:80-94.




     Caspar, F., Rothenfluh, T., & Segal, Z. (unpublished).  The appeal of


        connectionism for clinical psychology.




     Chomsky, N. (1959).  A review of B.F.  Skinner's Verbal Behavior.


        Language, 35:26-58.




     Clippinger, J. (1977).  Meaning and Discourse:  A Computer Model of


        Psychoanalytic Speech and Cognition.  Baltimore:  Johns Hopkins


        University Press.




     Cloitre, M. (1992).  The avoidance of emotional processing:  A


        cognitive science perspective, in Stein, D.J., & Young,


        J.E.  (eds), Cognitive Science and Clinical Disorders.  San Diego:


        Academic Press.




     Cohen, J.D., & Servan-Schreiber, D. (1992), Context, cortex, and


        dopamine:  A connectionist approach to behavior and biology in


        schizophrenia.  Psychol.  Rev.,  99:45-77.




     Colby, K.M.  (1981).  Modeling a paranoid mind.  Behavioral and Brain


        Sciences, 4:515-560.




     Collins, A.M., & Loftus, E.F.  (1975).  A spreading-activation theory


        of semantic processing.  Psychol.  Rev., 82:407-428.




     Craik, F.I.M., & Lockhart, R.S.  (1972).  Levels of processing:  A


        framework for memory research.  J. Verb. Learning Verb. Behav.,






     Craik, K. (1943).  The Nature of Explanation.  Cambridge:  Cambridge


        University Press.




     Delis, D.C., Kramer, J.H., Fridlund, A.J., & Kaplan, E. (1990).  A


        cognitive science approach to neuropsychological assessment, in


        McReynolds, P., Rosen, J.C., & Chelune, G.J.  (eds), Advances in


        Psychological Assessment.  New York:  Plenum Press.




     Dobson, K.S.  (1988).  Historical and philosophical bases of the


        cognitive-behavioral therapies, in Dobson, K.S.  (ed).  Handbook of


        Cognitive-Behavioral Therapies.  New York, Guilford.




     Eagle, M.N.  (1986).  The psychoanalytic and the cognitive


        unconscious, in Stern, R. (ed), Theories of the Unconscious and


        Theories of the Self.  Hillsdale, NJ:  Erlbaum.




     Ellenberger, H.F.  (1970).  The Discovery of the Unconscious:  The


        History and Evolution of Dynamic Psychiatry.  New York:  Basic






     Emde, R.N.  (1983).  The prerepresentational self and its affective


        core.  Psychoanalytic Study of the Child.  New Haven:  Yale


        University Press.




     Erdelyi, M.H.  (1985).  Psychoanalysis:  Freud's Cognitive Psychology.


        New York:  WH Freeman.




     Eysenck, M.W.  (1991).  Anxiety and cognitive functioning:  A


        multifaceted approach, in Lister, R.G., & Weingartner, H.J. (eds),


        Perspectives on Cognitive Neuroscience.  Oxford:  Oxford University






     Forrest, D.V.  (1991).  Mind, brain, and machine:  Object recognition.


        J. Am.  Acad.  Psychoanal.,  19:555-577.




     Foulkes, D. (1985).  Dreaming:  A cognitive-psychological analysis.


        Hillsdale, NJ:  Erlbaum.




     Freud, S. (1894).  The neuro-psychoses of defense.  Standard Edition,


        3.  London:  Hogarth Press, 1962




     Gardner, H. (1985).  The Mind's New Science:  A History of the


        Cognitive Revolution.  New York: Basic Books.




     Globerson, T., & Zelniker, T. (eds) (1989).  Cognitive Style and


        Cognitive Development.  Norwood, NJ:  Ablex.




     Gorenstein, E.E.  (1991).  A cognitive perspective on antisocial


        personality, in Magaro, P.A. (ed), Cognitive Bases of Mental


        Disorders.  Newbury Park, CA:  Sage.




     Gray, J.A.  (1982).  The Neuropsychology of Anxiety.  New York:


        Oxford University Press.




     Greenberg, L.S., & Safran, J.D.  (1990).  Emotion in Psychotherapy.


        New York:  Guilford.




     Greenspan, S.I.  (1980).  Intelligence and Adaptation:  An Integration


        of Psychoanalytic and Piagetian Developmental Psychology.  New


        York:  International Universities Press.




     Grossman, L.S., & Harrow, M. (1991).  Thought disorder and cognitive


        processes in mania, in Magaro, P.A. (ed), Cognitive Bases of Mental


        Disorders.  Newbury Park, CA:  Sage.




     Hebb, D.O.  (1946).  On the nature of fear.  Psychol.  Rev.,






     Hebb, D.O.  (1949).  The Organization of Behavior.  New York:  Wiley.




     Hestenes, D. (1991).  A neural network theory of manic-depressive


        illness, in Levine D.S., Leven, S.J.  (eds), Motivation, Emotion,


        and Goal Direction in Neural Networks.  Hillside, NJ:  Lawrence






     Hilgard, E.R.  (1977).  Divided Consciousness:  Multiple Controls in


        Human Thought and Action.  New York:  Wiley.




     Hoffman, R.E.  (1987).  Computer simulations of neural information


        processing and the schizophrenia-mania dichotomy.  Arch.  Gen.


        Psychiatry, 44:178-188.




     Horowitz, M.J.  (ed) (1988a).  Psychodynamics and Cognition.  Chicago:


        University of Chicago Press.




     Horowitz, M.J.  (1988b).  Introduction to psychodynamics:  A new


        synthesis.  New York:  Basic Books.




     Horowitz, M.J.  (1991).  Person schemas and maladaptive interpersonal


        behavior patterns.  Chicago:  Chicago University Press.




     Hull, J.G, & Reilly, N.P.  (1986).  An information processing approach


        to alcohol use and its consequences, in Ingram, R.E.  (ed),


        Information Processing Approaches to Clinical Psychology.  San


        Diego:  Academic Press.




     Ingram, R.E.  (1984).  Toward an information processing analysis of


        depression.  Cognitive Therapy and Research, 8:443-478.




     Ingram, R.E.  (ed) (1986).  Information Processing Approaches to


        Clinical Psychology.  San Diego:  Academic Press.




     Ingram, R.E., & Reed, M.E.  (1986).  Information encoding and


        retrieval processes in depression:  Findings, issues, and future


        directions, in Ingram, R.E. (ed), Information Processing Approaches


        to Clinical Psychology.  San Diego:  Academic Press.




     Ingram, R.E., & Wisnicki, K. (1991).  Cognition in depression, in


        Magaro, P.A.  (1991).  Cognitive Bases of Mental Disorders.


        Newbury Park, CA:  Sage.




     Janet, P. (1907).  The Major Symptoms of Hysteria.  New York:






     Jeffress, L.A.  (ed) (1951).  Cerebral Mechanisms in Behavior.  The


        Hixon Symposium.  New York:  Wiley.




     Johnson-Laird, P.N.  (1988).  The Computer and the Mind:  An


        Introduction to Cognitive Science.  Cambridge:  Harvard University






     Kihlstrom, J.F.  (1987).  The cognitive unconscious.  Science,






     Kihlstrom, J.F., & Hoyt, I.P.  (1990).  Repression, dissociation, and


        hypnosis, in Singer, J.L. (ed), Repression and Dissociation:


        Implications for Personality Theory, Psychopathology, and Health.


        University of Chicago Press:  Chicago.




     Klein, G.S.  (1976).  Psychoanalytic Theory:  An Exploration of


        Essentials.  New York:  International Universities Press.




     Koukkou, M. (1988).  A psychophysiological information-processing


        model of cognitive dysfunction and cognitive treatment in


        depression, in Perris, C., Blackburn, I.M., & Perris, H.  Cognitive


        Psychotherapy:  Theory and Practice.  Berlin:  Springer Verlag.




     Kuiper, N.A., Derry, P.A., & MacDonald, M.R.  (1982).  Self-reference


        and person perception depression, in Weary, G., Mirels, H. (eds),


        Integrations of Clinical and Social Psychology.  New York:  Oxford






     Lachman, R., Lachman, J.L., & Butterfield, E.C.  (1979).  Cognitive


         Psychology and Information Processing:  An Introduction.


         Hillsdale, NJ:  Erlbaum.




     Lipowski, Z.J.  (1989).  Psychiatry:  Mindless or brainless, both or


        neither?  Can. J. Psychiatry, 34:249-254.




     Lakoff, G. (1987).  Women, Fire, and Dangerous Things:  What


        Categories Reveal about the Mind.  Chicago:  University of Chicago






     Litrownik, A.J., McInnis, E.T.  (1986).  Information processing and


        autism, in Ingram, R.E. (ed), Information Processing Approaches to


        Clinical Psychology.  San Diego:  Academic Press.




     Lyddon, W.J.  (1988).  Information-processing and constructivist


        models of cognitive therapy:  A philosophical divergence.  Journal


        of Mind and Behavior, 9:137-166.




     Magaro, P.A.  (1980).  Cognition in schizophrenia and paranoia:  The


        integration of cognitive processes.  Hillsdale, NJ:  Erlbaum.




     Magaro, P.A.  (ed) (1991).  Cognitive Bases of Mental Disorders.


        Newbury Park, CA: Sage.




     Mahoney, M.J.  (1991).  Human Change Processes.  New York:  Basic






     Mahoney, M.J., & Freeman, A. (eds) (1985).  Cognition and


        Psychotherapy.  New York:  Plenum.




     Mandler, G. (1975).  Mind and Emotion.  New York:  Wiley.




     Markus, H. (1977).  Self-schemata and processing information about the


        self.  J.  Pers.  Soc.  Psychol., 35:63-78.




     Meltzoff, A.N.  (1990).  Towards a developmental cognitive science:


        The implications of cross-modal matching and imitation for the


        development of representation and memory in infancy.  Ann.  N.Y.


        Acad.  Sci.,  608:1-37.




     Merluzzi, T.V., Rudy, T.E., & Glass, C.R.  (1981).  The information


        processing paradigm:  Implications for cognitive science, in


        Merluzzi, T.V., Glass, C.R., & Genest, M. (eds), Cognitive


        Assessment.  New York:  Guilford Press.




     Morton, J. (1984).  Brain-based and non-brain-based models of


        language, in Caplan, D., Lecours, A.R., Smith, A. (eds), Biological


        Perspectives on Language.  Cambridge:  MIT Press.




     Neisser, U. (1976).  Cognition and Reality.  San Francisco:  WH






     Neuchterlein, K.H., & Dawson, M.E.  (1984).  Information processing


        and functioning in the developmental course of schizophrenic


        disorders.  Schizophr. Bull., 10:160-203.




     Oatley, K. (1991).  Best Laid Plans.  Cambridge:  Cambridge University






     Oatley, K., & Johnson-Laird, P.N.  (1987).  Towards a cognitive theory


        of emotion.  Cognition and Emotion, 1:29-50.




     Peterfreund, E. (1971).  Informations, Systems, and Psychoanalysis:


        An Evolutionary Biological Approach to Psychoanalytic Theory.  New


        York:  International Universities Press.




     Piaget, J. (1952).  The Origins of Intelligence in Children.  New


        York:  International Universities Press.




     Pitman, R. (1987).  A cybernetic model of obsessive-compulsive


        psychopathology.  Compr.  Psychiatry, 28:334-343.




     Posner, M.I.  (ed) (1989).  The Foundations of Cognitive Science.


        Cambridge:  MIT Press.




     Prigatano, G.P., & Schachter, D.L.  (eds) (1991).  Awareness of


        Deficit after Brain Injury:  Clinical and Theoretical Issues.


        Oxford University Press:  Oxford.




     Rapee, R.M.  (1991).  Psychological factors in generalized anxiety, in


        Rapee, R.M., Barlow, D.M.  (eds), Chronic Anxiety:  Generalized


        Anxiety Disorder and Mixed Anxiety-Depression.  New York:






     Rapoport, J., & Wise, S. (1988).  Obsessive-compulsive disorder:  A


        basal ganglia disease?  in Rapoport, J. (ed), Obsessive-Compulsive


        Disorder in Children and Adolescents.  Washington, DC:  APPI Press.




     Ruesch, J., & Bateson, G. (1968).  Communication:  The Social Matrix


        of Psychiatry.  New York:  Norton.




     Rumelhart, D.E., Hinton, G.E., and the PDP Research Group (1986).


        Parallel Distributed Processing:  Explorations in the


        Microstructure of Cognition.  Cambridge:  MIT Press.




     Saccuzzo, D.P.  (1986).  An information processing interpretation of


        theory and research in schizophrenia, in Ingram, R.E., Information


        Processing Approaches to Clinical Psychology.  San Diego:  Academic






     Safran, J.D., & Greenberg, L.S.  (1986).  Affect and the unconscious:


        A cognitive perspective, in Stern, R. (ed), Theories of the


        Unconscious and Theories of the Self.  Hillsdale, NJ:  Erlbaum.




     Shannon, C.E.  (1938).  A symbolic analysis of relay and switching


        circuits.  Master's thesis, Massachusetts Institute of Technology.




     Shimamura, A.P.  (1989).  Disorders of memory:  the cognitive science


        perspective, in Boller, F., & Grafman, J. (eds), Handbook of


        Neuropsychology, 3. Amsterdam:  Elsevier Science Publications.




     Simon, H. (1967).  Motivational and emotional controls of cognition.


        Psychol.  Rev., 74:29-39.




     Singer, J.L.  (ed) (1990).  Repression and Dissociation:  Implications


        for Personality Theory, Psychopathology, and Health.  University of


        Chicago Press:  Chicago.




     Slap, J.W., & Saykin, A.J.  (1983).  The schema:  Basic concept in a


        nonmetapsychological model of the mind.  Psychoanalysis and


        Contemporary Thought, 6:305-325.




     Sloman, A. (1987).  Motives, mechanisms, and emotions.  Cognition and


        Emotion, 1:217-233.




     Sorrentino, R.M., & Higgins, E.T.  (eds) (1986).  Handbook of


        Motivation and Cognition:  Foundations of Social Behavior.  New


        York:  Guilford Press.




     Stein, D.J.  (1992a).  Clinical cognitive science:  Possibilities and


        limitations.  In Stein, D.J., & Young, J.E.  (eds), Cognitive


        Science and Clinical Disorders.  San Diego:  Academic Press, 1992.




     Stein, D.J.  (1992b).  Psychoanalysis and cognitive science:


        Contrasting models of the mind.  J. Am.  Acad.  Psychoanal.,






     Stein, D.J.  (1992c).  Medical anthropology and psychotherapy


        integration:  A cognitivist approach.  Presented at the Annual


        Meeting of the Society for the Exploration of Psychotherapy


        Integration, San Diego.




     Stein, D.J.  (1992d).  Cognitive science and clinical knowledge.


        Integrative Psychiatry.




     Stein, D.J.  (1992e).  Schemas in the cognitive and clinical sciences:


        An integrative construct.  Journal of Psychotherapy Integration






     Stein, D.J., Hollander, E. (1992).  Cognitive science and


        obsessive-compulsive disorder, in Stein D.J., Young, J.E.  (eds),


        Cognitive Science and Clinical Disorders.  San Diego:  Academic






     Stein, D.J., & Young, J.E.  (eds) (1992a).  Cognitive Science and


        Clinical Disorders.  San Diego:  Academic Press.




     Stein, D.J., & Young, J.E.  (1992b).  A schema-focused approach to


        personality disorder, in Stein, D.J., & Young, J.E. (eds),


        Cognitive Science and Clinical Disorders.  San Diego:  Academic






     Stern, D.  (1985).  The Interpersonal World of the Infant.  New York:


        Basic Books.




     Sternberg, R.J.  (ed) (1984).  Mechanisms of Cognitive Development.


        New York:  WH Freeman.




     Stillings, N.A., Feinstein, M.H., Garfield, J.L., Rissland, E.L.,


        Rosenbaum, D.A., Weisler, S.E., & Baker-Ward, L. (1987).  Cognitive


        Science: An Introduction.  Cambridge:  MIT Press.




     Swerling, L.C., & Sternberg, R.J.  (1992).  Information processing,


        experience, and reading disability, in Stein, D.J., & Young, J.E.


        (eds), Cognitive Science and Clinical Disorders.  DJ, Young JE.


        San Diego:  Academic Press.




     Teasdale, J.D.  (1983).  Negative thinking in depression:  Cause,


        effect, or reciprocal relationship?  Advances in Behavioural


        Research and Therapy, 5:3-25.




     Tomkins, S.S., & Messick, S. (eds) (1963).  Computer Simulation of


        Personality:  Frontier of Psychological Theory.  New York:  Wiley.




     Turing, A. (1936).  On computable numbers, with an application to the


        Entscheidungsproblem.  Proceedings of the London Mathematical


        Society, 2nd Series, 42:230-265.




     Turing, A. (1950).  Computing machinery and intelligence.  Mind,






     Uleman, J.S., & Bargh, J.A.  (eds) (1989).  Unintended Thought.  New


        York:  Guilford Press.




     Wachtel, P.L.  (1982).  Resistance:  Psychodynamic and behavioral


        approaches.  New York:  Plenum Press.




     Weizenbaum, J. (1976).  Computer Power and Human Reason.  San


        Francisco:  WH Freeman.




     Westen, D. (1988).  Transference and information-processing.  Clinical


        Psychology Review, 8:161-179.




     Wiener, N. (1947).  Cybernetics.  Cambridge:  MIT Press.




     Williams, J.M.G., Fraser, N.W., MacLeod, C., & Matthews, A. (1988).


        An Information Processing Analysis of the Emotional Disorders.  New


        York:  Wiley.




     Williams, J.M.G, & Oaksford, M. (1992).  Cognitive science, anxiety,


        depression:  From experiments to connectionism, in Stein, D.J., &


        Young, J.E.  (eds), Cognitive Science and Clinical Disorders.


        San Diego:  Academic Press.




     Young, J.E.  (1990).  Cognitive therapy for personality disorders:  A


        schema-focused approach.  Sarasota:  Professional Resource