FUNCTIONAL BRAIN SYSTEMS AND PERSONALITY DYNAMICS
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA 1985
I am deeply grateful to Max and Ruth Lindquist for their patience and the opportunities they gave me. My debt to my undergraduate mentor, Dr. Carol Van Hartesveldt, who taught me about science and let me make my own mistakes, is also gratefully acknowledged.
I want to thank my doctoral committee members, Doctors Grater,
Morgan, Schauble, Suchman and Ziller, who were also my most respected graduate teachers. Special thanks are due to David Suchman, who first encouraged me in this project, and to my chairman, Harry Grater, who shepherded me through the crises with grace and allowed me to
keep my dignity.
The help of Dr. Bill Froming in the design of this study, and of Mssrs. Denny Gies, Bill Baxter and Ted Shaw (of the North Florida
Evaluation and Treatment Center) and of Ms. Janet Despard (of Mental Health Services, Inc.) in facilitating its execution, is much appreciated. I am especially grateful to Ms. Cheryl Shaw, who tyDed the manuscript, and without whose friendship and organizational help this project would not have been completed.
Warmest heartfelt gratitude goes to my dear friends Gabriel Rodriguez, Marshall and Laura Knudson, and David Kurtzman, whose love sustained me through these difficult years.
I can never properly express my thanks to Ruth Lindquist, to whom this piece of work is lovingly dedicated.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ................................................. iii
LIST OF FIGURES .................................................. vii
ABSTRACT ........................................................ viii
I INTRODUCTION ......................................... 1
II THE PHYSIOLOGICAL SUBSTRATE OF PERSONALITY:
A SELECTIVE REVIEW OF THE LITERATURE ................. 7
Review of Basic Brain Anatomy and Organization ....... 8
Cortical Mechanisms ............................... 8
Subcortical Systems .............................. 12
The Interpretation of Neurology Literature ....... 17
Human Consciousness ................................. 18
Right versus Left ................................ 18
Language and the Left Hemisphere ................. 20
Aphasia: Anatomy and Syndromes .................. 21
Language, Symbolism, and Meaning ................. 24
Human Consciousness, Self-Awareness,
and Thought ...................................... 26
Discussion ............................. 35
Cortical Mobilization: Attention, Arousal,
and Activation ...................................... 39
The Reticular Activating System and
Tonic Arousal ................................. ... 39
Phasic Control Systems: The Frontal Lobes
and Thalamus ..................................... 41
Discussion ....................................... 44
Motivation: Emotion and Affect ..................... 45
Amygdala Circuits and the Prefrontal Lobes ....... 46 Affective Expression ............................. 48
Discussion ....................................... 50
Memory Functions .................................... 51
Human Amnesia Syndromes .......................... 57
Memory and the Neocortex ......................... 63
Discussion ....................................... 68
The Limbic System, RAS, and Memory ............... 70
Papez Circuit and Memory ......................... 74
Fornix ........................................ 75
Mammillary bodies and mammillothalamic tract. .80 Cingulate cortex .......................... 81
Interfaces and Interactions of the Monitoring,
Motivating, and Mobilization Systems.............. 89
The Biochemistry of Emotion, Motivation,
and Learning................................. 92
Emotion, Amygdala Circuits and Memory........9 Discussion.................................. 103
Biochemical and Electrophysiological Aspects
of Cortical Mobilization Processes............ 104
The Functional Elements of the Personality
Structure ..................................... 108
System ..................................... 109
The Memory System ........................... 110
The Motivating System........................ 112
The Mobilization System...................... 114
Learning and Memory: Animal Studies ...........115
A Functional Meta-system........................ 120
Lateralized Mobilization Processes............ 125
Asymmetrical reaction time to laterally
presented stimuli......................... 125
Altered GSR following unilateral brain
Asymmetrical biochemical and electrophysiological processes ................... 127
Bilateral Interaction in Emotion and Cognition. .128
Mental Health and Psychopatholgy ................ 136
The Psychopathological Correlates of
Unilateral Temporal Lobe Epilepsy............. 139
Discussion: Hyper- and Hypo-dominance
Spectrum Disorders .......................... 141
Schizophrenia and the Affective Disorders .......143
Anxiety Disorders, Obsessive-Compulsive
Illness, and Paranoia........................ 145
Sociopathy and Hysteria...................... 147
III METHOD ........................................ 151
IV RESULTS ....................................... 156
Left versus Right Hemisphere Cognitive
Functioning Between Groups ................... 156
Overall Performance on the Tests Sensitive to
Right versus Left Hemisphere Cognitive
Differences in MMPI Scores ................... 157
V DISCUSSION...................................... 160
A HYPO- AND HYPER-DOMINANCE SPECTRUM DISORDERS BY
DSM III DIAGNOSTIC CLASSIFICATION .................165
B INFORMED CONSENT STATEMENT....................... 166
C REVISED SCORING CRITERIA FOR THE STREET (1931)
GESTALT COMPLETION TEST.......................... 167
BIBLIOGRAPHY ................................................. 168
BIOGRAPHICAL SKETCH........................................... 188
LIST OF FIGURES
1 Cytoarchitectural map of the lateral and medial
surfaces of the human cerebral cortex, with numbers
representing the areas of Brodman .................. 10
2 Partially schematized representation of the
limbic system .................................... 13
3 The position of Papez's circuit within a
larger cortical circuit ........................... 86
4 Schematic representation of the proposed model of
the physiological substrate of personality ..........111
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
FUNCTIONAL BRAIN SYSTEMS AND PERSONALITY DYNAMICS
Chairman: Dr. Harry Grater
Major Department: Psychology
A heuristic model of the physiological substrate of personality
structure was developed from a review of recent neurological and physiological psychology literature. The formulation of the model was based on assumptions that the phylogenetic transition from instinctive responding to self-determined behavior required the evolution of automatic neural mechanisms to insure that the organism would monitor the environment for significant stimuli, be motivated to respond in the presence of those stimuli, and mobilize the appropriate psychological operations to determine the form of that response. Functional elements of these basic mechanisms were identified and a functional meta-system was outlined which would organize the elements and optimize the -utilization of lateralized cognitive processes in the interest of assuring the emission of adaptive behavior.
The proposed model suggests that certain psychodiagnostic entities might be classified as hyper- and hypo-dominance spectrum disorders viii
based on the form of dysfunction within the meta-system. The ability of the model to predict membership in diagnostic categories was tested by assigning 42 adult psychiatric inpatient and outpatient subjects to hyper-dominant and hypo-dominant groups by diagnosis, according to the constructs of the model, and comparing performance on instruments shown to be sensitive to right and left cerebral hemisphere dysfunction. The Street Gestalt Completion Test and the Object Assembly, Similarities, and Information Subtests of the Wechsler Adult Intelligence Scale-Revised were administered to each subject. An abbreviated form of the Minnesota Multiphasic Personality Inventory (MMPI) was used to compare symptomology between groups.
Significant between-group differences (p < 0.001) in the ratios of test scores sensitive to right versus left cognitive functioning were found in the predicted directions, while the groups did not differ in overall performance on the instruments. Significant differences (p< 0.005) in the ratios of selected MMPI clinical scales, in the predicted direction, provided further support for the hypothesized relationship between lateralized cognitive functioning, symptomology, and diagnosis.
It was concluded that the proposed model provides tenable and potentially useful operational definitions of personality functions and psychopathology. Results were discussed in terms of their imrplications for psychotherapeutic interventions and additional methods to test the validity of the model.
The various schools of psychotherapy agree that the practitioner's task is to facilitate change in the client. They agree on little else. There is, as yet, no consensus regarding the two major issues in psychotherapy: what is to be changed, and how that change is to be brought about (Strupp, 1978). Strong opinions about both of these basic problems are available; data based conclusions are not. ProDonents of fundamentally different viewpoints debate acrimoniously (Eysenck, 1974); yet verifiable differences in outcome between theorybased forms of intervention are rare (Bergin & Lambert, 1978).
This unsatisfactory state of affairs in applied psychology has unfortunate consequences for practitioners and their clients alike.
The problem has been ascribed by Watson to the "preparadigmatic"
state of the discipline of psychology. According to Watson "psychology has not experienced anything comparable to what atomic theory has done for biology, what laws of motion have done for physics. Either psychology"s first paradigm has not been discovered yet or it has not been recognized for what it is" (Watson, 1967, p. 53). Hanson stated the problem succinctly, "The issue is not theory using, but theory finding" (1965, p. 3).
The two major forces in psychological thought, psychoanalysis and behaviorism, have encountered significant problems. Theories
based on the former have been criticized as untestable and therefore unscientific (Eysenck, 1970). The latter movement lost impetus with the discovery by Olds and Milner (1954) of "pleasure" centers in the brain. This revelation undermined the basic assumption of the learning theorists that behavior could be explained simply by defining the rules governing stimulus-response relationships.
Successful scientific theories are built on paradigms that
describe the fundamental properties and mechanics of their subject. The fundamental units of personality are networks of neurons in the brain. Sigmund Freud (1948) attempted to relate mental structures to anatomical locations but was forced to abandon his effort because the neurology of the time was not adequate. Instead he and subsequent theorists were forced to base their models on suppositions about the products of the personality processes. As noted above, the results have been less than satisfactory. The science of neurology has made significant progress in the interim and a large amount of useful information has accumulated. These data have been virtually ignored by the discipline of psychology. The integration of neurological data and psychological theory may provide a basis for a useful paradigm for the psychotherapist. The present work is intended as a step toward such an integration.
The purpose of this study is to develop and test a heuristic,
model of personality function, based on an understanding of its physiological substrate, with the ultimate goal of inDroving the effectiveness of psychotherapeutic interventions. Such a model
should identify the basic elements and processes of the personality
structure and describe the ways in which these interact to produce psychological health and psychopathology. Such a model might lead to new operational definitions of psychological phenomena which, in turn, may suggest new intervention points and methods.
The purpose of a theory is to integrate known facts within a
single framework and account for them in terms of a small number of interrelated concepts. Existing theories suffer from a lack of integration. The discipline of psychology has hindered such integration by institutionalizing a tradition of subspecializatiOn: theories of attention, perception, cognition, motivation, emotion, behavior, etc. are developed in relative isolation. The failure of individual theorists to understand and to appreciate the interrelatedness of these various personality operations may account for the inadequacy of psychological theory in general. It is assumed here that the component parts of the personality can only be properly understood in the context of their relationship to the whole; that more will be gained from a gross (but comprehensive and testable) description of the personality infrastructure than from a detailed examination of a single personality operation.
Rather than subdivide the personality structure on the basis of notions of what it should do; it might be more productive to approach the problem on the more basic level of what it must do. The formulation of the model will be guided by a set of assumptions about the evolutionary pressures which shaped the personality structure. These "evolutionary imperatives" are as follows:
1. The physiological substrate of personality evolved to
assure the emission of adaptive behavior by the organism. Survival pressures shaped an organization that was able to respond effectively to significant stimuli and produce behavior that enhanced the organism's well-being.
2. The evolution of this substrate proceeded from an organization based on "hard-wired" stimulus-response instincts to an organization which allowed increased latitude in behavior in order to take advantage of the organism's developing problem-solving abilities.
3. In order to permit self-determination of behavior and,
at the same time, to insure survival, new mechanisms were required to assure that the organism would (a) attend to significant stimuli; (b) be motivated to respond effectively to those stimuli;
(c) accomplish the mobilization of the necessary psychological resources to determine the form of that response (referred to hereafter as the Monitoring, Motivating, and Mobilization systems).
4. Although these systems interact, they must be functionally separated to the extent that they do not interfere with each other's normal operation. Similar functions might be carried out in other brain areas, but the automatic activation of these systems will assure that they dominate responding to stimuli which relate to the survival and well-being of the organism.
5. As these mechanisms are essential for the survival of the individual and species they will form the central organizing processes-of personality; personality dynamics will center on their operation.
Neurological theories of mental function center on clinical
observations of patients with localized brain lesions. The history of neurological thought has revolved around the question of how these data are to be interpreted. For many years higher mental functions were treated as discrete "faculties." The results were inconsistent and of little value. The failures of these "narrow localizers" led to theories which attempted to account for mental functions on the basis of the "mass action" of the brain. Where earlier models were too specific, these theories proved too general to be useful.
In the 1920s Goldstein broke with the tradition of attempting to infer functions directly from deficits and proposed instead an "analysis of basic disturbances." This approach led finally to Luria's conceptualization of mental activities as the product of the interaction of complex functional systems. In Luria's formulation a mental function is the result of contributions from a number of concertedly working zones. Therefore, that function may be destroyed, or disturbed differently, by lesions in different locations. Luria (1973a) described the characteristics of a functional system:
The presence of a constant (invariant) task, performed
by variable (variative) mechanisms, bringing the process
to a constant (invariant) result, is one of the basic features distinguishing the work of every "functional
system." The second distinguishing feature is the
complex composition of the "functional system," which always includes a series of afferent (adjusting) and efferent (effector) impulses. (Luria, 1973a, p. 28)
Luria outlined three principal functional units in the brain: the "units for regulating tone and waking and mental states," centered on the reticular activating system in the brainstem; the
"unit for receiving, analyzing and storing information," operating in the post-central (sensory) areas of the brain; and the "unit for programming, regulation and verification of activity," operating in the frontal lobes (Luria, 1973(i. ch. 2).
Luria's concepts represent a major advance in the understanding of the fundamental operating characteristics of brain systems. Although his formulations are too basic to be of much use to the applied psychologist, he has established a format and a methodology which will be followed here. The present investigation will focus on identifying and describing the interactions of the functional brain systems which satisfy the requirements of the evolutionary imperatives outlined above. Evidence suggests that the substrate for these systems will be found in those anatomical areas for which Luria acknowledged he had inadequate data for his own analyses: the medio-basal zones of the cortex and the right hemisphere of the brain.
THE PHYSIOLOGICAL SUBSTRATE OF PERSONALITY:
A SELECTIVE REVIEW OF THE LITERATURE
In the following sections the neurological and physiological psychology literature pertaining to the functional/anatomical organization of personality will be reviewed. The data will be related to psychological factors and any conclusions will be noted in discussions at the end of each section. The first section is a review of basic brain anatomy and organization. The second section will examine the physiological substrata of consciousness and conclude that human consciousness, characterized by self-awareness, is a manifestation of processes which occur in the left hemisphere of the brain. The third section will trace the systems involved in cortical activation and note the existence of two mechanisms within each hemisphere which have opposite effects on the form of cognitive processes. The fourth section will examine the role of the amygdala and frontal lobes in the subjective experience of emotion and of the right hemisphere in the expression of affect. The fifth section will develop the basis for a theory of memory function and the role of memory systems in organizing affective, arousal, and cognitive processes. The sixth section will focus on the interactions of the emotion, arousal, and memory systems and examine the role of biochemically mediated systems in coordinating these processes. In the seventh section a model of four functional systems which
constitute the basic elements of the personality structure will be proposed. The eighth section will describe the way in which the basic elements are organized into a functional meta-system which
forms the infrastructure of personality and functions to assure the emission of adaptive behavior. This model will be supported with
evidence concerning the psychological correlates of neurological syndromes. In the final section the psychological phenomena associated with various forms of psychopathology will be related to neurological indices which reflect the operation of the lateralized subsystems and their interaction within the functional meta-system. It Oill be concluded that many psychodiagnostic entities may be classified as hypo- or hyper-dominance spectrum disorders which are functionally related to chronic, and maladaptive, under- or over-utilization of the mechanisms in the left hemisphere of the brain.
Review of Basic Brain Anatomy and Organization
Although certain phylogenetically new areas of the cerebral cortex are of special interest when discussing "higher mental functions," these areas must be considered in their physiological and evolutionary context. A brief review of basic brain systems and anatomy will establish this perspective.
In man, as in all mammals, large portions of the cerebral
cortex are devoted to the more elementary functions of processing sensory stimuli and the initiation and control of movements. The brain structures subserving these basic functions are divided at the
central fissure (Rolando). Incoming somato-sensory, visual and auditory nerve impulses are relayed, via the thalamus, from contralateral receptor surfaces to primary projection areas located in the parietal, occipital and temporal lobes (see Fig. 1, areas 1, 2, 3; 41; 17), respectively (Noback & Demerest, 1972). In each case the modally specific, somatotopic organization of nerve impulses in the projection areas is transformed into functional information (i.e., acquires meaning) in an adjacent secondary association area (areas 5, 7, 18, 19; 42). With each new level of processing there is increasingly complex synthesis of information and decreased modal specificity (Luria, 1973a). Damage to a primary projection area results in a loss of sensation (e.g., blindness) while lesions of a secondary association area are likely to produce the inability to recognize a stimulus in that modality (agnosia). Conversely,
artificial stimulation of a projection area produces a discrete sensory experience while stimulation of a secondary association area elicits a more elaborate sensory hallucination whose complexity is related to the level within the hierarchy that is activated (see Mullan & Penfield, 1959).
Progression within the hierarchy is reversed in the motor
systems. Specificity of control increases as the secondary (premotor) areas (areas 6 and 8) coordinate and fine tune their influence on the pyramidal cells of the primary motor cortex (area 4) with the assistance of continuous feedback from the sensory modalities. Lesions of the primary motor cortex produce contralateral paresis while stimulation elicits flexion of individual muscle groups.
/,- 64 7
Figure 1. Cytoarchitectural map of the lateral and medial surfaces of the human cerebral cortex, with numbers representing the areas of Brodman. (Redrawn from Noback & Demerest,
10 23 7 9
Figure 1. Cytoarchitectural map of the lateral and medial surfaces of the human 'Cerebral cortex, with numbers representing the areas of Brodman.' (Redrawrn from Noback & Demerest, 1972).
Stimulation of secondary motor areas produces smooth, coordinated movements and ablation of these areas may result in the loss of ability to perform skilled motor acts (apraxia) (Noback & Demerest, 1972).
The systems described thus far are typical of all mammals and become progressively more elaborate and efficient in the higher primates. Evolution proceeds generally by modifying and elaborating existing hardware, allocating new functions to tissues which are in some way pre-adapted to assume the new tasks (Campbell, 1974). The development of higher mental functions in man is correlated with bilateral anatomical expansion of two cortical areas which are adjacent to the secondary association areas described above. These regions subserve the highest level of organization in the hierarchies and are called teritary association areas (Luria, 1973a). Both areas are involved in what Penfield (1975) referred to as "transactions of the mind."
The inferior parietal lobule (IPL) (areas 23; 39, 40) lies at the anatomical confluence of the secondary association areas in the post-central cortex. This area is the "association cortex of association cortexes" (Geshwind, 1979). Here processed information from the surrounding sensory association areas is further integrated and synthesized. The area is called "supramodal" because its
individual units can only be excited by the simultaneous stimulation of two or more sensory modes (Luria, 1973a). Information processing at this level is "abstract" in that it is independent of a particular sensory modality (cf. Osgood, 1953). This ability allows the
simultaneous synthesis of information which permits the mental manipulation of the relationship between information units and as such is a prerequisite for the high level mental functions that are characteristic of human beings (Luria, 1973a).
At the anterior pole of the brain, contiguous with the secondary motor association areas, the prefrontal lobes (areas 9-12) are dramatically enlarged and now represent up to one-fourth of the entire cortical mass. The prefrontal lobes have extensive two-way connections with all other parts of the cerebral cortex (Luria, 1973a). The
coordinating and control operations carried out by the lower level (motor) systems in this hierarchy are evident in the functions of the tertiary integration areas. The frontal lobes have been called the "executive of the brain" (Pribram, 1973). They are the seat of Luria's "unit for programming, regulation and verification of activity" (Luria, 1.973a). Lesions of the frontal lobes lead to a defect in the patient's "capacity for planned initiative" (Penfield & Evans, 1940), and to disturbances on impulse control (Pribram, 1973).
The limbic system consists of a group of interconnected structures, situated between the midbrain and the neocortical mantle, including the hypothalamus, amygdala, hippocanpus, septal area, and cingulate cortex (see Fin. 2). Because of their relationship to the olfactory bulbs, these structures were originally thought to be concerned with that function and so this area of the brain was designated the rhinecephalon ("nose-brain"). Although
*0 to (M~ C UC( (-J M, C O 4-W C- M Ica- 0
- (n CO uVE a C
U -) E3 -0 a- -0 >1 4. -. .. '
ro J W LLU- M: C to L.) =.
-- U 0 *E~U
(A 4- X F-.
co~>- 4-.) :3 o4
ai m 0i. 4-) /
lower animals depend on the sense of smell for such survivalrelated activites as food-getting, the detection of enemies, and mating (MacLean, 1949), the role of these subcortical structures in motivational and emotional processes was not appreciated until the 1930s. Since Kluver and Bucy (1938) published their description of altered emotional behavior following the bilateral removal of the temporal lobes, including the amygdalae and hippocampi, limbic system components have been the subject of intense scrutiny. The massive literature which has accumulated is filled with confusing inconsistencies and conflicting findings. This may be due in part to the fact that the function of these structures appears to vary according to enviromental circumstances (Olds, 1958). Although a number of tentative models of limbic system function have been offered (e.g., Papez, 1937; Gloor, 1956; Olds, 1958) none has found general acceptance.
MacLean (1970) distinguishes three major types of systems in the mammalian brain which correspond to stages in its evolutionary development. He describes a protoreptilian core-brain, a paleomammalian brain (the limbic system) and a neomanmnalian brain (the neocortical areas). This triune brain conceptualization provides a useful perspective on the hierarchical arrangement of anatomical and functional systems and their relationship to behavior (Isaacson, 1974),
The protoreptilian brain represents the fundamental core of the nervous system consisting of areas homologous to parts of the upper brainstem and midbrain, hypothalamus, and basal ganglia.
In primitive organisms this brain produces a repertoire of instinctive,
stereotyped behaviors which are sufficient to insure the survival of the individual and species. These unlearned, species-specific
behavior patterns relate to the elementary functions of obtaining food and shelter, establishing and defending a home territory, breeding, maternal behavior, etc. (Isaacson, 1974). Programs for these behavior patterns are apparently stored in the brainstem and triggered by the hypothalamus: complex behavior sequences such as eating, drinking, sexual acts, aggression and many other types of unlearned behavior can be elicited by electrical stimulation of the hypothalamus at levels below those needed to activate motivational systems (Isaacson, 1974). In protoreptilian animals the performance of these acts would be coordinated by the basal ganglia. In these
forms the striatum is the highest center for sensory and motor processing, these functions being subserved by structures corresponding to the :audate/putamin and globus palladus, respectively (Schade' & Ford, 1973).
Stimulus and response are yoked in the protoreptilian brain. Since its repertoire of behaviors is unlearned (i.e., instinctive) this neural system "reacts to changes in the environment by increasing or decreasing the intensity of the predominant response sequence. . Suppression of response on the basis of non-reward or punishment is difficult" (Isaacson, 1974, p. 273).
The development of the paleomammalian brain allows the suppression of stereotyped ways of responding. The addition of the limbic
system circuitry permits the organism to adjust its behavior based
on new sources of information about the internal and external environments and to utilize new and more efficient forms of learning in organizing its responding. It is important to appreciate that the final product of limbic system operations in the paleomammalian brain is the inhibition of lower centers. However, these same mechanisms later came to exert important influences on the neocortical systems.
The cerebral cortex of the neomammalian brain developed in close association with the limbic system and basal ganglia. New nuclei were added to existing sensory and motor systems, culminating finally in the arrangement now found in primates. More efficient information processing in the neocortical additions permits more precise sensory discriminations and rapid, fine-grain movements of the extremities (Isaacson, 1974) but these new systems still operate in close conjunction with the older subcortical mechanisms (Schade' & Ford,
The importance of the limbic system in human experience can hardly be overstated. These structures are implicated in most, if not all, forms of psychopathology. They are the probable substrate for the therapeutic effect of most psychotropic medications (Broekkamp & Lloyd, 1981) and the targets for all forms of "psychiatric surgery." Unfortunately, existing theory regarding the functional significance of the limbic system is primarily descriptive in nature or consists of generalizations so broad as to be of little use to the applied psychologist.
The Interpretation of Neurology Literature
The neurological data which form the basis of this review were gathered over many decades, from a variety of populations, by scientists with widely divergent theoretical beliefs. The psychologist who is unfamiliar with this literature must become aware of its inherent problems before venturing interpretations based upon it.
Those who come under the neurologist's care have, almost invariably, suffered a catastrophic change in their lives and personalities. Much of the data is drawn from individuals who have incurred brain damage. as a result of a head wound, cerebrovascular accident (stroke), or cerebral tumor (neoplasm). While offering an otherwise unavailable opportunity to study higher brain functions, these devastating "natural experiments" invariably preclude the rigorous application of scientific method to that study.
Although techniques continue to improve, it is unusually
impossible to define the parameters of a lesion accurately. Thus, crucial independent variables cannot be precisely isolated and controlled. Cause and effect interpretations are further hampered by the fact that a patient's pre-morbid level of functioning may be impossible to ascertain. Generalizability is also compromised when the subject has a chronic brain disease, such as epilepsy, or has undergone a surgical intervention to control that condition, such as temporal lobectomy or cerebral commissurotomy. In these cases, as with patients who have been subjected to "psychiatric surgery," it is reasonable to suggest that the "pre-treatment" psychological organization may have been grossly abnormal.
The human brain has remarkable ability to compensate for
injury by utilizing undamaged tissue. This recovery of function in the time elapsed between injury and testing is a further
Much of this literature is in the form of case reports and the problem of individual variability is especially difficult to manage. Research designs utilizing groups and statistical analyses are becoming more frequent, but these studies invariably suffer from the difficulties noted above.
The most difficult problems to be encountered here have to do
with the specification of dependent variables: "higher brain functions" tend to resist definition and to defy quantification. Generally, only gross operational definitions have been available. The physician/experimentor was often forced to resort to a judgement as to whether a given (and hypothetical) function was "intact' or "lost," with the attendant risks of experimentor bias and error. Related to this is the understanding, which emerged slowly in the development of this literature, that the function of a circumscribed anatomical area cannot be inferred directly from a deficit which follows the ablation of that area: one can only specify how the brain functions in the absence of that tissue.
Right versus Left
There have been three basic approaches to the study of lateralized cortical function: the comparison of patients with unilateral brain damage; the lateralized presentation of stimuli or task to the
sensory receptor fields of normal subjects; and the study of patients whose cerebral hemispheres have been surgically separated (Nebes, 1977). The early studies of split-brain subjects prompted a great deal of speculation about the abilities of the right hemisphere and a listing of the published works concerned with asymmetrical hemispheric functioning is beyond the scope of this paper. Evidence pertaining to the differences between the cognitive operations performed by the left and right hemispheres has been ably reviewed by a number of writers (e.g., Bogan 1969a, 1969b; Galin, 1974, 1977; Gazzaniga, 1970; Gazzaniga & Ledoux, 1977; Nebes, 1974, 1977; Sperry, 1968). Some of the more important differences will be noted very briefly here.
The orimary deficits seen after right hemisphere lesions involve difficulties in perceiving, manipulating, and remembering spatial relationships and in perceiving and remembering sensory stimuli which resist verbal description (Nebes, 1977). The right hemipshere is superior to the left at generating a percept of the whole from fragmentary information and seems predisposed to notice complex gestalts or patterns rather than the parts (Nebes, 1971, 1974, 1977). The right hemisphere tends to process input simultaneously (in parallel) as opposed to the left hemisphere's preference for sequential (serial) processing, Cohen, 1973). In spite of many claims to the contrary, the evidence indicates that the right
hemisphere considers only the immediately perceived context and performs its tasks in a reflexive, automatic fashion.
Language and the Left Hemisphe
The prominent Soviet geneticist, Theodosius Dobzhansky, observed that ". . while all other organisms become masters of their environments by changing their genes, man does so mostly by changing his culture, which he acquires by learning and transmits by teaching" (Dobzhansky, 1964, p. 145). Cultural evolution was made possible by the development of language, but language is more than a special form of social communication by which culture is transmitted: it is the mechanism which produces the adaptive behavior patterns that released the species from the constraints of biological evolution. Speech is the fundamental tool of intellectual activity. Language processes are essential in the operations of abstraction and generalization, the basis of categorical thinking, and the vehicle for organizing and regulating mental processes and behavior (Luria, 1973).
The clearest example of functional brain asymmetry is the lateralization of language to the left hemisphere, a fact long established by two observations: deficits in language functions (aphasia) following damage to the left hemisphere and the retention of language following right hemisphere injury (right handedness is assumed throughout the text unless otherwise noted), The patterns of language deficit following circumscribed lesions provide the most concrete evidence that is available about the mechanics of consciousness and its product, cognition. The focus of the following review is not on language itself, but on the evidence which indicates that the active operations of consciousness which produce
adaptive behavior are part of a functional system that is lateralized in the left hemisphere of the brain. Aphasia: Anatomy and Syndromes
Three general types of aphasia have been defined and traced
to different anatomical structures in the left hemisphere. Through an analysis of these syndromes it is possible to deduce the outline of a functional system in which these separate areas work together to produce the complex function of language. The following summary is based on reviews by Gardner (1975) and Zaidel (1978).
Broca's aphasia, resulting from damage to the inferior, postfrontal zones of the left hemisphere, is an expressive disorder. Utterances are difficult to initiate and speech is painfully slow and labored. Patients are usually able to name objects and to repeat, having less trouble with words that are familiar and concrete. Although they manage to convey their meanings in a peculiar, telegraphicc" style, they are unable to produce a fully formed sentence. Their productions consist almost entirely of substantives; grammatical parts and forms are absent or impoverished, Nouns are usually delivered in the singular and verbs appear in their simplest,
noninflected form. Parts of speech that are purely gramatical in function (conjunctions, prepositions, articles, adverbs) are exceedingly uncommon. The same deficit pattern is evident in the patient's reading and writing.
Although patients with Broca's aphasia may have difficulty in unravelling complex grammatical relationships (e.g., "The lion was eaten by the tiger: which animal is dead?"), their comorehen-
of language is generally intact. The ability to understand and utilize nonverbal symbolism (e.g., gesture, pantomime) is also spared. Intellectual functioning is relatively unimpaired. Patients retain the ability to reason logically, to abstract and to generalize, and to respond to context appropriately. Their associational processes are not loosened, tangential or pressured. They do not produce paraphasias or confabulate. Finally, these patients are acutely aware of their deficits; they may be appropriately depressed and are prone to sudden, transient emotional outbursts.
Exactly the opposite clinical picture is evident when Wernicke's area, on the lateral, convex surfaces of the left temporal lobe, is damaged. Wernicke's aphasia is characterized by impaired language comprehension with fluently articulated but nonsensical speech. Unlike the Broca's aphasic, words are spoken clearly with normal sounding cadence, intonation and melbdy (prosody). However, the speech of the Wernicke's aphasic is lacking in content and may consist almost entirely of semantic jargon which has little conmunication value. These patients appear to have lost control of the language mechanisms at all levels. It is as if the selection thresholds for phonemes, words, and ideas were all lowered. Repetition is poor and marked by paraphasic errors in which the correct sounds may be present but emerge in the wrong order. Patients are able to name only the most familiar objects accurately, although the word produced may come from the same category as the target. Prompting with the initial wordsounds seldom helps.
In spontaneous speech the key substantives are often missing and the remaining parts of speech lack their organizing influence. Grammatical parts and forms are used abundantly but incorrectly. Adjectives,, adverbs, conjunctions and prepositional phrases are strung together haphazardly and the result may resemble a schizophrenic's "word salad" (Luria, 1973). Nonsense syllables and neologisms are frequent. The guiding thought behind a verbal production may be evident, but it becomes obscured by tangential associations and incomprehensible babbling.
The Wernicke's aphasic understands little of what is said and seems to rely on nonverbal cues in order to respond to a situation. They are also unable to comprehend nonverbal symbolism and so their communication is vague and concrete at all levels.
Perhaps the most striking feature of Wernicke''s syndrome is
the patients.' complete lack of awareness of and indifference to their deficits. They appear unconcerned and will vehemently deny any problems.
The same deficits are evident in reading and writing tasks.
The patient can read single words but does not seem to grasp their meaning or relate them to him or herself. They may correctly repeat a simple command written on a card but will make no attempt to comply. Unable to formulate an acceptable sentence on their own, they are able to arrange cards with words printed on them to form a syntactically correct sentence. However, the key substantives are likely to be misplaced (e.g., "The man bit the dog.") confirming again that the patient's facility is with grammar as Opposed to meaning.
There is less agreement about the third major language
disorder, known as anomic or amnesic aphasia, which results from more posterior lesions located in the angular gyrus in the parietooccipital area. Part of the confusion may stem from the fact that its main symptom, the loss of the ability to name objects, is common to all aphasic disorders. However, the anomic aphasia syndrome is distinguished by the fact that the naming disorder is accompanied by relatively intact comprehension of written and spoken language and normal spontaneous speech. The ability to read and to repeat are also spared in anomic aphasia.
The anomic aphasic has no difficulty using words in their appropriate context but cannot find the word in isolation of context; he is unable to divorce himself from the immediate situation. The anomic aphasic cannot produce the name of objects on demand even though he knows what they are. When an object is designated the patient is unable to produce its name, and conversely, given a name, the patient is not certain what it refers to.
Other language difficulties are evident. The patient's
spontaneous speech seems to be either too detailed or too general. Thinking is very concrete; the patient will interpret proverbs literally. The patient is aware of the diabilities and will often develop strategies to compensate for them. Lanugage, Symbolism, and Meanini
The theory of the functional organization of language developed by Wernicke in 1885 is still generally accepted today. According to this model the underlying structure of an utterance arises in
Wernicke's area. It is then transferred via large fibre bundles (the arcuate fasciculus) to Broca's area where it evokes a detailed program for vocalization which is governed by the rules of grammar and syntax. The program developed in Broca's area, the linear scheme of the sentence, is supplied to the adjacent face area of the motor cortex which in turn drives the muscles which produce the vocalization. Thus, the content of speech originates in Wernicke's area and finds its form in Broca's area.
Wernicke's area is also essential for the comprehension of language. Auditory stimuli are relayed from the Organ of Corti to the primary auditory projection areas in Heschell's gyrus in the left temporal lobe. At this point, as with the other sensory projection areas, the information is somatotopically organized and retains its modal specificity; to be understood it must be transferred to the secondary auditory association area (Wernicke's area) where the somatotopical organization is converted into a functional organization (Luria, 1973a). Here, the fundamental phonemic characteristics of language are isolated and identified. Processing by Wernicke's area is essential for both the encoding and decoding of meaning. An intact Wernicke's area is also essential for the expression and comprehension of meaning through symbolic gesture and pantomime (Goodglass & Kaplan, 1973; Gionotti & Lemmo, 1976). It is interesting to note in this regard that "illusions of interpretation" emerge in consciousness after electrical stimulation of the temporal lobe in the right hemisphere but not after stimulation of any other brain area (Mullan & Penfield, 1959).
Human Consciousness, Self-Awareness, and Thought
In the 1960s the general belief in cerebral dominance gave
way to the idea of cerebral specialization due, in large part, to the "split-brain" studies conducted on the cerebral commissurotomy patients of Drs. Vogel and Bogan. In 1969 Bogan took the progression a step further when he revived Wigan's (1844) notion of "the duality of mind." Wigan, (noting the anatomical duality of the brain, autopsy findings of hemispheric atrophy in patients whose personality was apparently intact, and introspective evidence of concurrent, opposing trains of thought) argued that if one hemisphere can sustain a mind, "it necessarily follows" that a man with two hemispheres must have two minds (p. 271). Bogan endorsed this concept in the third of his influential "Other side of the Brain" papers (Bogan. 1969a) and concluded that
Pending further evidence, I believe (with Wigan) that each of us has two minds in one person. . Various
kinds of evidence, especially from hemispherectomy,
have made it clear that one hemisphere is sufficient
to sustain a personality or mind. We may then
conclude that the individual with two intact hemispheres has the capacity for two distinct minds.
This conclusion finds its experimental proof in
the split-brain animals who can be trained to perceive, consider, and act independently. (1969,
The "dual mind" concept implies two relatively equal but
functionally independent entities which act as opposed forces in the process of determining behavior. Bogan contributed the hypothesis that the two hemispheres utilize different "modes of thought" in this process: "propositional" on the left and "appositional" on the right (1969, p. 160). These appealing ideas were enthusiastically
embraced by laymen and professionals alike. They have been cited to support all manner of theories concerning psychological, philosophical, and spiritual dualities (see the critique by Kinsbourne, 1982). However, a closer analysis of the data cited by Bogan suggests that ushc conclusions are misleading.
Michael Gazzaniga, an author of the pioneering animal studies referred to by Bogan as perhaps the most dedicated and prolific of the "split-brain" researchers, complained about the "overpopularization" of basic data produced by himself and his colleagues:
These popular psychological interpretations of "mind left: and "mind right" are not only erronious: they
are inhibitory and blinding to the new students of
behavior who believe classic styles of mental activity
break down along simple hemispheric lines. (1977, p. 416)
There is no doubt that one cerebral hemisphere can, in the absence of its counterpart, support high level intellectual activity if the loss of the other hemisphere occurs early in development. Griffith and Davidson (1966), for example, report that children show relatively good recovery from hemispherectomy for infantile hemiplegia. Smith and Sugar (1975) reported on a 26 year old man who showed superior intelligence (WAIS VIQ:126, PIQ:102, FSIQ:116) 21 years after undergoing left hemispherectomy at age five and one-half. However, it is improper to infer normal functioning directly from grossly abnormal cases such as this, or from animal studies. While it is clear that the right hemisphere may have the capacity to develop higher mental processes, there is no evidence that it does so normally, and considerable evidence to the contrary. Research
efforts in this area have been hampered by the lack of adequate operational definitions (for such phenomena as consciousness, cognition, and thought) which distinguish the hither mental processes of humans from the brain functions of lower forms.
Cerebral dominance for consciousness has been investigated
using the Wada carotid amobarbital test, a procedure developed to localize language functions prior to neurosurgery. The Wada test involves injecting sodium amobarbital into one common carotid artery and results in the anesthetization of only the cerebral hemisphere on the side of the injection. Terzian (1964) reported an absolute and immediate arrest of any communication, both verbal and nonverbal, in the first thirty to sixty seconds after the injection of the drug into the carotid artery of the dominant side which he interpreted as a transient loss of consciousness. Serafetinides and his co-workers reported similar results and noted that the phenomenon rarely occurred following barbiturization of the non-dominant hemisphere (Serafetinides, Driver & Hoare, 1964, 1965a, 1965b). They concluded that unconsciousness, and by
implication consciousness, is in general linked with the function of the hemisphere dominant for speech (Serafetinides et al., 1965a).
Rosadini and Rossi ('1967) attemDted to replicate these findings using more strictly operationalized definitions of consciousness. In one group (48 cases) the criteria consisted of an "analysis of the capacity of the patient to keep in contact with the examiner through verbalizations or movements, to react to noxious stimuli
and to describe at the end of the examination what happened during the examination itself" (p. 103). In a second group the criteria
for consciousness consisted of "a simple stimulus-response test" in which "the patients were instructed to work a switch held in the hand ipsilateral to the intracarotid injection [i.e., the hand controlled by the unanesthetized hemisphere] any time they heard a given sound or saw a flash of light" (p. 103). Behavioral and clinical events indicating unconsciousness were required to last more than one minute in order to "Permit their safe detection." They found that 47 of their 69 cases did not meet their criteria for unconsciousness and the 22 cases which did occurred in roughly the same percentage following left and right injections; aphasia occurred in 16 cases (15 left, one right); only five of the 21 cases (three left, two right) evaluated with the stimulus-response test failed to operate the switch held in the hand controlled by the unanesthetized hemisphere in response to the signals used.
Although their results were complicated by existing neuropathological and cerebrovascular abnormalities, these authors concluded that "the existence of a cerebral dominance for consciousness is not supported" (,p. 111). The usefulness of this study appears to be severely limited by the criteria used to evaluate consciousness in the unanesthetized hemisphere: a large number of animals of various species have demonstrated the ability to "work a switch" in response to visual and auditory stimuli and to respond to noxious stimuli, but these lower forms are not considered conscious in the same sense that humans are. The authors were looking for "the
occurrence of signs revealing the capacity of the subject to keep in contact with the external world" (p. 103). They acknowledged that "the Suppression of expressive and receptive speech functions make such a task quite difficult with the patients receiving barbiturate in the dominant hemisphere" (p. 109). It seems that it is only the appearance of aphasiz (after dominant hemisphere anesthetization) that is specific to human consciousness in this study, and these
findings are consistent with those of Terzian and Serafetinides et al.
Rosadini and Rossi did not report the results of their test of subjects' ability to recall what had occurred during the Wada procedure but this question was addressed directly in an experiment reported by Gazzaniga (1977). This author found that "information encoded while the left hemisphere was anesthetized was uninterpretable by the verbal system when the left hemisphere returned to normal functioning .. when information is encoded by other than the verbal system the person is not consciously aware of the information" (p. 150).
Another approach to the localization of conscious awareness
involved an analysis of the temporal discrimination for simultaneity when two visual stimuli were presented separately to the left and
right visual half-fields, separated by a very brief interval. Efron (1963a; 1963b) found that normal right-handed subjects reported that the two flashes occurred simultaneously only when the light flashed in the left visual half-field was presented several milliseconds earlier than the light flashed in the right visual field.
Efron argued that this was because the "conscious comparison" of the two flashes takes place only in the hemisphere dominant for language, the time lag representing the extra neural steps involved
in relaying sensory information from the right hemisphere over the corpus callosum:
It is only after sensory data have reached the left hemisphere that one is "conscious" of the occurrence of an
event. . To be conscious of something is to be conscious
of something now. It is the thesis of this paper that
the "now" is the moment of arrival of sensory data in
the dominant temporal lobe. (1963b, p. 421)
The most convincing evidence of a correlation between human consciousness and language ability emerged from studies of a unique cerebral commissurotomy patient known as "case P.S." P.S., a right handed boy, developed epilepsy following an injury to his left hemisphere incurred at age two. He subsequently developed language skills in both his right and left hemispheres.
At age 14 he underwent complete surgical section of his corpus callosum to relieve his epilepsy. Following surgery it was found that P.S.'s right hemisphere could spell, comprehend verbal commands, process parts of speech and make conceptual judgements involving verbal information. It was also discovered that his right hemisphere, although unable to speak, could generate answers to printed questions presented tachistocopically to his left visual half-fields. He accomplished this by arranging Scrabble letters with his left
hand. These answers were often different from those given verbally by his isolated left hemisphere. Gazzaniga and Ledoux (1977) argued that P.S.'s right hemisphere possessed qualities deserving of conscious status because.
His right hemisphere has a sense of self, for it knows the name it collectively shares with the left. It has
feelings, for it can describe its mood. It has a sense
of who it likes and what it likes, for it can name its
favorite people and its favorite hobby. The right
hemisphere in P.S. also has a sense of the future, for
it knows what day tomorrow is. Furthermore, it has
goals and aspirations for the future, for it can name
its occupational choice. . The fact that this mute half-brain could generate personal answers to ambiguous and subjective questions demonstrates that in P.S. the
right hemisphere has its own inde *pendent responsepriority determining mechanisms, which is to say, its
own volitional control system. (pp. 143-145)
P.S. is the only split-brain patient with advanced language skills in his right hemisphere and the only patient to demonstrate double consciousness. Ledoux, Wilson and Gazzaniga (1977) stressed the fact that "in all other patients, where linguistic sophistication is lacking in the right hemisphere, so too is the evidence for cons,,_-Iousness" (p. 420).
The capacity for speech and conceptual thought is clearly innate in homo sapiens; only the symbols themselves must be learned (Campbell, 1974). Recent evidence has clarified the anatomical substrate of this genetically transmitted specialization. The development of language capabilities in the human species is correlated with the anatomical expansion and interconnection of the association areas in the left hemisphere (Campbell, 1974). The posterior area of the planum temporale, .%hich forms a part of the secondary auditory cortex (Wernicke's area) is significantly larger on the left side (Geshwind & Levitsky, 1963. The enlargement of this area can be explained in terms of its distinctive cellular organization (Galaburda, LeMay, Kemper & Geshwind, 1978) and the
incomplete development of this cellular architecture has been related to language dysfunction (see Geshwind, 1979).
In her exhaustive study of hundreds of brain injured war veterans, Semmes (1967) discovered that elementary sensory and motor capacities were focally represented in the left hemisphere and diffusely represented in the right. She proposed that this difference indicates the mechanism of hemispheric specialization: focal organization favoring fine control and the integration of similar units (e.g., manual skills and speech) and diffuse organization favoring multimodal coordination (e.g., the various spatial abilities).
Gazzaniqa and Ledoux (1977) observed that nearly every demonstration of a right hemisphere advantage in split-brain patients has involved mani pul 0-spatial acti vi ties and concluded that
[iThis advantage] exists so long as manipulative activities
are involved in either the stimulus perceptions or the
response production . . The probable neural substrate of
these manipulo-spatial acts involves the inferior parietal
lobule of the right hemisphere in humans. In the left
hemisphere, however, linguistic functions occupy the
inferior parietal lobule .. . The superior performance
of the right hemisphere of split-brain patients on such
tasks does not reflect the evolutionary specialization
of the right hemisphere, but instead represents the
price paid by the left hemisphere in acquiring language,
Our view is not that the right hemisphere is
specialized in some unique way in man. Rather, it
continues to do what it does elsewhere in the phyla.
(pp. 420-421, emphasis added)
Campbell (1974) noted that "spatial relationships involving
depth and distance may appear to be predominately spatial concepts, but they are not of space but about space; of themselves they are spaceless and concerned with pattern rather than place" (p. 337).
It appears, then, that the left hemisphere extracts meaning from the relationship of individual parts to each other, while the right gathers meaning from the pattern of the whole.
Bogan (1969b) proposed that the right "mind" utilized a
different mode of thought which he characterized as aDpositional to denote the ability to appose, or compare, information. He contrasted this with the propositional mode utilized by the left hemisphere. A number of investigators have distinguished similar dichotomies of information processing style. Luria (1973a) spoke of narrative versus relational processes. Galin (1974) suggested that the right hemisphere solved problems through a process Of Multiple converging determinants as opposed to a left hemispheric style which utilized a single causal chain. Sechenov (quoted by Luria, 1973a) postulated that the human brain utilizes two forms of integrative activity: organization into simultaneous and primarily Spatial groups, and into temporally organized successive series. This is consistent with Spearman's conclusion that intelligence comprises two components: the eduction of correlates used in analogical reasoning and the eduction of relations, the basis of abstract reasoning (see McFie & Piercy, 1952). Campbell (1974) noted that "abstraction means escape from the present . what distinguishes man from animals is the length of time through which his consciousness extends (p. 335). Finally, Bogan (1969b) observed that the most important distinction between the left and right hemispheric modes miqht be "the extent to which the linear concept of time participates in the ordering of thought" (p. 160).
There is a clear consensus recognizing two modes of information processing. However, the ability to process information is not necessarily a sufficient condition for consciousness. Although the notion of right brain "thought" has gained wide currency, to date, there has been no conclusive evidence that any cognitive operation occurring in the right hemisphere is directly experienced in consciousness. A possibility not considered by any of the above authors is that one mode might be an ancillary resource utilized by the other. The data reviewed thus far suggests that one must be
very careful to avoid anthropomorphizing when attempting to describe right brain processes. However, some inductive conclusions may be drawn,
It is evident that human consciousness is inexorably linked
with the abstract symbolic processes associated with language. Perhaps the most universally accepted characteristic of human thought is self-awareness. Self-awareness (and its product, the self-concept) requires the abstraction and appreciation of defining features which are consistent over time and situation. Only the left hemisphere, with its temporal acuity, can consider and appreciate changed or conserved relationships in different conditions or contexts. Thus, only the left hemisphere can define itself. The resulting selfawareness provides a reference Point for all of the memories, feelings, intentions and thoughts that are collectively known as themind and which allow the individual, thus defined, to interact intelligently with the environment. Lacking the temporal organizing skills to
construct such a consistent frame of reference the right hemisphere is bound to the immediate context with only the influences of the physiological status of the organism (and the left hemisphere) to guide its processes. Complex motivations, therefore, cannot exist in the right hemisphere. Likewise, so-called "pictorial thinking," if temporally ordered and goal directed, must be organized by the left hemisphere. Galin (1974) suggested that the context bound, egocentric and impulsive nature of right hemisphere condition resembled Freud's notion of primary process thinking. Higher mental processes in the right hemisphere almost certainly qualify as cognitions (i.e., a way of knowing) and may account for the phenomenon of intuition (knowledge without awareness of the process by which it was gained). However, the term "thought" seems misleading and "information processing" might be preferable. As noted above, there is no direct evidence that mental events occurring in the right hemisphere are directly experienced in the conscious left hemisphere; one is left to ponder the question of whether a tree falling in the right brain would make a sound if the left wasn't listening.
The restrictions outlined above are in no way inconsistent with the demonstrated role of the human right hemisphere in the analysis of emotional communications and the modulation of affective expression (e.g., Ross & Mesulam, 1979). The brains of lower forms (and, apparently, the right brain in humans) are primarily concerned with neuronal signals which represent the survival needs of the organism within the immediate environment. Interaction with
the social environment is critically important to survival throughout the phylum. Campbell (1974) noted that animal vocalizations and signals are "emitted only in the presence of the appropriate stimulus" (p. 349) and warned against equating these vocalizations with human speech: "the signals . are generated or motivated by the phenomenon of emotion, and find their neurological origin not in the cortex but in the limbic system of the brain" (1974, p. 348). The cortical organization of these functions in the right brain of humans appears to mirror that of language in the left hemisphere with comprehension and expression utilizing anatomical areas homologous to Wernicke's area and Broca's area, respectively (Ross & Mesulam, 1979). Right hemisphere responses might achieve direct expression in circumstances where control by Broca's area in the left hemisphere is impaired or attenuated, a case in point being the clearly enunciated emotional exclamations of the frustrated Broca's aphasic. Similarly,
poorly defined and undifferentiated emotionally generated behavioral impulses (e.g., approach, avoidance) might also achieve motor expression in the absence of adequate left hemisphere control.
Hughlings Jackson (1864) suggested that if the "faculty of expression" was proven to be lateralized in the left cerebral hemisphere it would then be reasonable to expect that its corresponding opposite, perception, might be lateralized to the right. Although the concept of mental "faculties" has given way to an appreciation of complex functional systems, the role of the right hemisphere within those systems might, in a broad sense, be said to conform to Jackson's prediction. While the information processing
style of the right hemisphere is not suited to solVing problems or making decisionsit is uniquely qualified to perceive the quality of significance (as defined by the individual's experience) in complex environmental stimuli. The evidence suggests that the human right hemisphere attends to the overall pattern of stimuli, searches out ("apposes") associations which are correlated with important stimulus configurations and collates them into percepts that have meaning for the organism. Its associational processes are unencumbered by rules of logic and its perceptions uninfluenced by expectations.
The thrust of Bogan's (1969b)"dual mind" thesis was a reaction against the traditional concept of hemispheric dominance which relegated the right hemisphere to the role of an "automaton" or reserve organ (e.g., Henschen, 1926; Strong & Elwyn, 1943). A basic assumption in the present work, however, is that evolutionary pressures required in the development of an automatic environment monitoring system in order to permit the transition from instinctual to self-determined behavior. It appears that evolution solved this problem by taking advantage of the fact that the human central nervous system contains two relatively autonomous brains which could be yoked together by the limbic system. Within this configuration the left hemisphere may be seen as a problem-solving and responsegenerating system and the right hemisphere might be said to
function as the repository and librarian of the individual's reinforcement history.
Cortical Mobilization: Attention, ArousaT-, and Activation
Consciousness and cognition become possible only when minimum levels of cortical tone are attained. These tonic levels, reflected in the desynchronized EEG pattern, permit sensory discriminations, motor acts and other cognitive operations to take place. Once activated, the cortex has the ability to make phasic modifications of its level of activity and to voluntarily direct its attention. The fundamental systems which govern cortical tone and attention, however, are automatic and capable of overriding voluntary controls. It is clear that these systems, with their ability to control the level and content of consciousness, will exert a significant influence on the personality structure. The Reticular Activating System and Tonic Arousal
The mobilization of the cortex is accomplished by the brainstem reticular formation (RF). This structure is a network of highly interconnected neurons which adjusts its level of activation by integrating input from the sQnsory pathways, limbic system structures, and the neocortex. Impulses from the reticular formation lower the activation thresholds of the neurons it projects to. When the reticular formation is relaxed, cortical tone is lowered and the organism sleeps (Moruzzi & Magoun, 1949).
The reticular activating system (RAS) regulates the state of activation of the brain in two ways: the ascending reticular activating system (ARAS) affects the brain diffusely and sets the generalized (tonic) level of arousal; descending influences direct
RAS impulses to accomplish localized (phasic) arousal of specific areas of the brain. The ascending pathways of the RAS project rostrally from the brainstem reticular formation (via the central tegmental tract/medial forebrain bundle) to the hypothalamus, septal area, and nonspecific intralaminar thalamic nuclei (ILTN). The second path extends from the interpeduncular nucleus to the ILTN
via the habenula. The only direct ascending connections from the RAS to the neocortex are projections from the nonspecific thalamic nuclei (midline and ILTN) to the orbitofrontal cortex (via the ventral anterior nucleus of the thalamus). Descending influences are conveyed from the prefrontal neocortex to lower structures by way of the thalamocortical radiations, corticoreticular fibers,
medial forebrain bundle, and thalamotegmental fibers. Hippocampal output reaches the reticular formation via the fornix, mammillary bodies, and mammillotegmental tract. The septal area, has an additional connection with the reticular formation by way of a stria terminalis--habenula--habenulointerpeduncular tract--interpeduncular nucleus pathway (Noback & Demerest, 1972).
Based on their analysis of some 200 experiments, Pribram
and McGuinness (1975) outlined two major subsystems in the brainstem which control cortical mobilization and identified separate forebrain mechanisms which modulate their functioning. These systems initiate two different types of cortical activity. Diffuse
cortical "arousal," which is associated with the orienting response, is ased on the serotonergic brainstem median raphe" nuclei located in the core of the reticular formation. Arousal is modulated by
a lateral-frontal--amygdala--lateral-hypothalamus facilitory circuit and an inhibitory orbitofrontal--amygdala--medial-hypothalamus circuit. "Activation"is an attention focusing process involved in perceptual expectancies and motor readiness to respond. This system is based on the locus ceruleus, in the periaqueductal gray, which supplies norepinephrine to the forebrain. Activation was thought to be modulated by the ancient motor control system in the basal ganglia. Together, these systems provide for appropriate attending to novel or significant stimuli and prepare the organism to respond cognitively and behaviorally. The hippocampus was seen to integrate the functioning of these systems and to exert ultimate control over cortical mobilization through a mutually inhibitory relationship with the reticular formation.
Phasic Control Systems: The Frontal Lobes and Thalamus
A novel (possibly significant) stimulus elicits an orienting response (OR) from the organism. The psychological phenomena associated with the OR are familiar to all who have had experience with "things that go bump in the night." The complete orienting reaction includes
The suppression of ongoing behavior, the orienting of
the body and receptor towards the new stimulus, changes in the peripheral autonomic nervous system, and, perhaps
less obvious, preparations for associating the new stimulus
with memories from the past and expectancies of the
future. (Issacson, 1974, p. 110)
Pribram (1973) noted that the stimulus sampling aspects of the
orienting response differed from the processes necessary to register a stimulus in awareness and memory (which must be accomplished
before the organism can habituate to a stimulus). In contrast to the indiscriminate arousal associated with orienting, these latter processes require the focusing of attention.
The mobilization of selective attention ("activation") appears
to be reflected in the contingent negative variation (CNV) or expectancyy wave" (Tecce, 1970). The CNV is a special form of
cortical evoked response which consists of a spreading wave of negative potential that appears whenever there is a contingent relationship between two stimuli. Negativity develops when brain tissue is maintaining a readiness for processing (Pribram & McGuinness, 1975). Thus, the CNV appears whenever the organism is expecting to perform a perceptual or motor act. The negativity becomes abruptly positive when that act is executed (Walter, Cooper, Aldridge, McCallum, & Winter, 1964). High amplitude CNVs are related
to greater efficiency of perceptual and motor responses; concentration facilitates the CNV while inattention, boredom, or fatigue decrease it (Cohen, 1974). Elithorn et al. (1958) postulated that frontal lobe injuries somehow damaged the mechanism underlying anticipatory sets. It is interesting, in this light, that the CNV generally appears in the prefrontal lobes and sweeps posteriorally over the post-central cortex.
Patients with frontal lesions are unable to sustain their
attention. While intelligence, as measured by standardized tests, may be unimpaired, these individuals are highly distractable and cannot carry out purposeful activity which is normally directed by intentions (Luria, 1973 i). Luria pointed out that patients with
lesions of the temporal, parietal, or occipital lobes may have
sensory, orientation, or intellectual deficits, but their attention and concentration remain sustained and directed by intentions. Luria and his co-workers suggested that the frontal syndrome reflected the loss of this selectivity (Luria, Homskaya, Blinkov, & Critchley, 1967). An understanding of the functioning of thalamic systems suggests a mechanism by which the frontal lobes might select or
"recruit" psychological operations in the post-central cortex.
The thalami are a pair of egg-shaped masses located beneath the cortex in the center of the cerebral hemispheres, The thalamus
is the final processing point for cortical input and the central integration station of the nervous system. The brief review of thalamic anatomy and functioning presented below is based on reviews by Noback and Demerest (1972) and Chusid (1976).
The ventral half of the thalamus contains the specific relay
of nuclei of the sensory-motor systems. The nuclei of the dorsal tier are association nuclei which have reciprocal connections with the association areas of the post-central cortex and no subcortical connections; the dorsolateral and posterolateral nuclei are interconneected with the parietal lobe, and the pulvinar with the temporal and Parietal lobe.
The dorsomedial and anterior thalamic nuclei are association nuclei involved in emotion and memory, respectively. The dorsomedial nucleus receives input from the amygdala and lateral hypothalamus and has reciprocal connections with the association areas of the prefrontal lobe. The anterior nuclei of the thalamus receive
the output from the hippocampus and have reciprocal connections with the cingulate cortex.
Lying between and separating the major thalamic association nuclei, the nonspecific (intralaminar, midline and reticulate) nuclei have only intrathalamic and subcortical connections. They receive their main input from the brainstem reticular formation and from the rostral end of the ascending reticular activating system (ARAS). The intralaminar nuclei have the "remarkable property of being able to exert a controlling influence upon the rhythmic electrical activity of the entire cortex" (Jasper, 1949, p. 406). Jasper noted that this system is in a position to provide a central
coordinating mechanism for cerebral activities:
A central integrative mechanism with ready access to all
afferent and elaborative systems of both hemispheres,
and closely related to autonomic spring of action; is necessary to explain consciously directed thought and
behavior. It seems that the thalamic reticular system,
with its diffuse cortical projections, relations to
afferent and efferent systems, relations to mesencephalic
hypothalamic and striatal systems, is a good candidate
for this office. (Jasper, 1949, p. 419) Discussion
The brain's mobilization systems with their brainstem, thalamic, and forebrain components, regul ate consciousness, unconsciousness, and the differential consciousness of attention. Sensory signals representing possibly significant stimuli cause the reticular formation to initiate diffuse cortical arousal. When a stimulus has been identified, cortical activation systems facilitate the organization of cerebral activity to deal with the situation appropriately. It appears that the frontal lobes direct this process
by recruiting psychological operations in the post-central cortex. The frontal lobe may accomplish this through its influence on the nonspecific thalamic nuclei which, in turn, control the phasic activation of specific cortical systems.
It is important to note that the two biochemically mediated subsystems which control the mobilization processes are duplicated in both halves of the brain. Unilateral prefrontal lobe lesions have been found to produce deficits which resemble those seen after damage to post-central lesions on the same side: right frontal lesions have been associated with disturbances of emotion and spatial abilities while left-sided injuries lead to disorders of speech and thought (Zangwill, 1966; Benton, 1968; Luria, 1973a;l. Processes which disrupt the biochemical balance between the two control mechanisms have far-reaching psychological consequences which will be reviewed in a later section.
i motivation: Emotion and Affect
The greatest risk involved in giving up instinct-based responding in favor of self-determined behavior is the possibility that the individual might fail to respond appropriately in survival-related situations. The evolution of the species could not have occurred if this problem had not been solved. The forces which motivate adaptive behavior must, by definition, be the single most powerful influence on the personality structure. The evidence indicates that the source of these forces lies in the limbic system. It appears that separate cortical mechanisms mediate their internal experience and external expression.
Amygdala Circuits and the Prefrontal Lobes
The role of the amygdala in emotional processes established by Kluver and Bucy in 1938, has been assumed to be affected through this structure's close relationship with the hypothalamus. The amygdala seems to direct behavior toward biological goals (Halgren, 1981) and is implicated in the control of species-sDecific behaviors related to survival needs, including defensive and aggressive behaviors, sexual activity, and feeding (Isaacson, 1974). In lower forms these processes might depend on a simplified (instinctive) form of memory in which stimulus and response are yoked (Pribram & McGuiness, 1975). In addition to mediating emotional states the amygdala is involved in the analysis of reinforcement contingencies. Amygdala lesions have been shown to produce impaired recognition of stimuli associated with rewards (Weiskrantz, 1956; Schwartzbaum, Thompson & Kellicut, 1965; Jones & Mishkin, 1972) and inability to respond appropriately to changes in the magnitude of rewards (Schwartzbaum, 1960).
Strong interconnections with the hypothalamus (via the stria terminalis and ventral amygdalofugal fibers) give the amygdala immediate access to information concerning the internal status of the organism (Price, 1981). The amygdala also receives processed sensory information from all of the secondary sensory association areas (Van Hoesen, 1981). Mishkin and Aggleton (1981) noted that this arrangement places the amygdalae in a position to integrate external events with their internal consequences, which would permit the attachment of emotional and motivational significance
to sensory stimuli. Kessner (1981) reported experimental evidence that demonstrated the essential role of the amygdala in encoding and retrieving the positive and negative attributes of a specific memory. In lower forms the identification of a motivationally significant stimulus might result in the release of species-specific behaviors, but in humans behavior is self-determined. Halgren (.1981) concluded from his amygdala stimultion studies with humans that "the amygdala helps organize the discharge of emotional tension into consciousness" (p. 404) and noted that this would allow the directing of consciousness toward biological goals. The amygdala's input to the neocortex is directed to the entire prefrontal lobe both directly, via the uncinate fasciculus, and indirectly, by way of the dorosomedial thalamus (Noback & Demerest, 1972; Price, 1981).
While damage to the dorsolateral area of the prefrontal lobes has been associated with intellectual disturbances, lesions of the orbito-frontal cortex (and orbital undercutting, which disconnects this area from the amygdala) result in emotional changes (Lewin, 1961). Elithorn, et al. (1958) concluded that this type of damage produced a "generalized impairment of the ability to form appropriate emotional responses" (p. 250),,including the ability to elaborate on the affect appropriate to the concepts present in consciousness.
In contrast to the planning deficits, loss of energy and interest, and affective dullness seen after dorsolateral frontal damage,.orbitofrontal injuries often lead to euphoria, impulsive (disinhibited) behavior, and the appearance of "greediness, selfishness, and
tactlessness" (see Lewin, 1961). Faust (1966) noted that such patients resemble psychopaths in that they are unable to profit from experience and are in constant conflict with their environment and the law. Zangwill (1966) pointed out that the tactlessness
common in frontal lobe patients does not result from a loss of knowledge of social conventions, but from the failure to regulate behavior in accordance with those standards.
Disconnecting the orbito-frontal cortex from the am~ygdala
(orbital undercutting) has been reported to be the most effective psychosurgical operation for relieving the symptoms of anxiety and depression (Elithorn et al., 1958; Lewin, 1961; Levinson & Meyer, 1965). Elithorn et al. (1958) noted that this procedure
increased reactions of "a hysterical type" and is contraindicated for those conditions. It is interesting to note that drugs that reduce anxiety and produce euphoria (e.g., the barbiturates) have been found to exert an uncoupling effect between the frontal lobes and the limbic system (Heath & Galbraith, 1965). Affective Express ion
The right hemisphere plays an essential role in both the
comprehension of emotional communications and the expression of affect. Patients with right hemisphere lesions showed impaired recall of stories with emotional content versus neutral stories (Wechsler, 1973). Hielman, Scholes and Watson (1975) demonstrated that judgements of the emotional mood of a speaker (sad, happy, angry, indifferent) made by patients with right temporoparietal
lesions were significantly impaired relative to patients with leftsided lesions and controls. This finding was replicated by Tucker,
Watson and Hielman (1976) who showed also that right hemisphere patients were impaired in the vocal expression of emotion. The efforts of patients with right temporoparietal damage to impart a sad, happy or angry tone to their voices were rated as incorrect significantly more often than controls. Ross and Mesulam (1979) presented case studies of two well-educated patients who had comnparable damage in the right supra-sylvian area, which is homologous to Broca's area on the left side. Both patients showed flattened affect and had completely lost the ability to laugh, cry, or otherwise express any emotion in their speech. Their ability to experience and comprehend emotions was unchanged. The authors noted that the organization of emotion in the right hemisphere seems to mirror that of language in the left: the area homologous to Wernicke's area being essential for comprehension, and to Broca's
area, for expression.
Gazzaniga and Ledoux (1977) suggested that right hemisphere functioning in humans is distinguished only by contrast to the left; it continues to perform its functions in the same manner as elsewhere in the phylum. Although it should be noted that the human right hemisphere is in possession of tertiary association areas and so would perform those tasks more efficiently, the data reviewed above are not inconsistent with Gazzaniga's interpretation. The brains of lower forms (and, apparently, the right brain in humans) are primarily concerned with neuronal signal !s which represent the survival needs of the organism within the immediate environment. Interaction with the social environment
is critically important to survival throughout the phylum. Campbell (1974) noted that animal vocalizations and signals are "emitted only in the presence of the appropriate stimulus" (p. 349) and warned against equating these vocalizations with human speech: "the signals . are generated or motivated by the phenomenon of emotion, and find their neurological origin not in the cortex but in the limbic system of the brain" (1974, p. 348). Discussion
It appears that basic human emotional experience is an
emergent property of the functioning of mechanisms that originally served to regulate the emission of soecies-specific behaviors which were elicited directly by releasing stimuli in survivalrelated situations. The functional systei which evolved in humans decouples stimulus and response. T1us, in humans, it is t~le emtional experience evoked by a stimulus--rather than the stimulus itself-that is the primary motivating factor (reinforcer) which ultimately determines behavior. Further, this emotional experience might be most properly considered to be a part of the experiencing person's environment, since that experience is involuntary and has the power to condition the person's response. These processes appear to have their functional impact in the left hemisphere, where the formulation of behavioral responses occurs.
The physiological mechanisms which motivate adaptive behavior in humans are centered on the amygdala which integrates information from the internal and external enviroments in order to attach emotional significance to stimuli. This structure is involved in the
encoding and retrieval of this information in memory and forwards its signals to the prefrontal lobes where they are experienced as subjective emotions. The prefrontal lobes appear to utilize these signals in the process of forming intentions to direct adaptive behavior. This amygdala-prefrontal pathway appears to be the substrate of anxiety and depression. When the amygdala-frontal connection is severed surgically, or uncoupled pharmacologically, the neocortex experiences euphoria, but fails to behave in an adaptive manner.
It is evident that the functional brain systems which form the infrastructure of personality include separate cognitive, affective and arousal components. It appears that these subsystems evolved to take full advantage of a form of learning which utilizes reinforcement and emotional experience in determining behavior. The product of any learning experience is memory. The importance of memories (or "associations") in the organization of cognitive operations is obvious and most, if not all, arousal and affective processes must depend on the ability to discriminate personally relevant stimuli. Clearly, memory is fundamental to all aspects of personality function, but the material substrate of memory remains a complete mystery (e.g., Lashley, 1950; Luria, 1975a);-our concepts regarding it are, of necessity, only abstract descriptions, Before reviewing the physiological organization of memory systems it will be necessary to define and delimit, as far as possible, those abstract
concepts of memory phenomena that are pertinent to the interests of the applied psychologist.
Experimental psychologists have traditionally approached,
the study of memory by subdividing it into registration, retention, and recall, attempting to isolate and measure these aspects and the variables which affect them. Rapaport (1961) criticized this methodology as artificial, insisting that these functions are inextricably related and that such experiments merely demonstrate how memory can function under given laboratory conditions, Working from a psychoanalytic perspective, Rapaport preferred to treat memory as an aspect of cognition. He argued that "actual memory phenomena are encountered only in the context of thought processes; at best the classical memory experiments could ignore this fact and make us ignore it, but they could not produce memory phenomena outside this context" (Rapaport, 1961, p. 6). He acknowledged the difficulties in determining the relation of indistinct entities such as emotion and memory and attempted to clarify the psychoanalytic viewpoint by suggesting that "memory is a motivated behavior phenomenon and emotions are motivating factors" (Rapaport,
1961, p. 8). This statement, however, appears to beg the question; if memories are activated by emotions, then what initiates arousal and affective processes?
The interaction of cognition, affect and memory in the etiology and cure of psychopathology were central themes in the work, published in 1893 by Josef Breuer and Sigmund Freud, which gave Psychoanalysis its start. The emphasis on the significance of
memory phenomena in psychoanalytic literature (e.g., slips of the
tongue, forgetting, false remembering, repression) can be traced
to this seminal paper in which the authors concluded that "hysterics
suffer mainly from reminiscences." In this work, Breuer and Freud
made a crucial distinction regarding the memory processes operating
in psychoneurosis which may have sowed the seed from which the
notion of unconscious causation of psychological phenomena germinated:
...the causal relation between the determining psychical
trauma tan experience which calls up distressing affects
such as those of fright, anxiety, shame or physical pain]
and the hysterical phenomenon is not of a kind implying
that the trauma merely acts like an agent provocateur
in releasing the symptom, which thereafte leads an
independent existence. We must presume rather that the
psychical trauma--or more precisely the memory of the trauma--acts like a foreign body which long after its
entry must continue to be regarded as an agent that
is still at work.' (Breuer & Freud, 1974, p. 355)
The authors became aware of this "highly remarkable phenomenon"
and its relation to affective processes in the course of their
experimental treatment of hysterical conversion symptoms:
[we found] that each individual hysterical symptom immediately and permanently disappeared when we had
succeeded in bringing clearly to light the memory of
the event by which it was provoked and in arousing its accompanying affect, and when the patient had
described that event in the greatest possible detail
and had put the affect into words. Recollection
without affect almost invariably produces no result.
(Breuer & Freud, 1974, p. 355)
Hillix and Marx (1974) have suggested that "it was necessary for
Freud to invent the psychic apparatus and much of his psychoanalytic theory just to account for what he and Breuer had already
observed" (p. 352). It is to be hoped that recent evidence will
make a more parsimonious accounting possible.
The implicitly verbal form of memory referred to by Rapaport seems to be qualitatively different from the "foreign body" which Breuer and Freud assumed to be the culprit in hysterical neurosis. They referred to the latter type of memory by the less formal term "idea" and indicated that symptom removal depends.on the transformation of this "idea" into a more formal thought process so that its associated affect can be abreacted:
[the therapeutic procedure] brings to an end the operative force of the idea which was not abreacted in the
first instance, by allowing its strangulated affect to
find a way out through speech; and it subjects it to associative correction by introducing it into normal
consciousness. (p. 356).
This special type of memory would seem to merit a more detailed description. It is evident that we are concerned here with a subset of memories which have significance for the individual. By definition, these are memories that are associated with reinforcement and/or emotional experience. They are experiential (nonverbal) and may be isolated from consciousness. It may be noted that this subset of memories will define the relationship between the individual and his or her environment and might be the organism's most important survival resource. A concept from social learning theory seems to encompass this type of memory comfortably and provides a more
Julian Rotter (1966) theorized that "a reinforcement acts to
strengthen an expectancy that a particular behavior will be followed by that reinforcement in the future" (p. 2). Further, "when an organism perceives two situations as similar, then his expectations
for a particular kind of reinforcement, or class of reinforcements, will generalize from one situation to another" (Rotter, 1975, p. 57). Rotter distinguished two types of "generalized expectancies" (GE). The first has to do with the nature of the reinforcement: expectations for a particular kind of reinforcement in a given situation. The second type deals with other properties of situational stimuli and has to do with the perception of control that one can exercise to change or maintain the situation: the kind of behavior that is likely to produce or terminate reinforcement. The first type is designated with a subscript r for reinforcement (GE r ). The second type is designated a problem-solving generalized expectancy (GEPS). Striking insights into the nature and mechanics of this sort of experiential memory were afforded by Penfield's observations of certain psychical phenomena elicited by direct electrical stimulation of the conscious brain (see Mullan & Penfield, 1959).
Wilder Penfield's data were collected from patients undergoing
radical brain surgery, with local anesthesia, for the relief of intrac-, table epilepsy. His observations consist of spontaneous reports from these patients following applications of a mild electric current to the exposed cortex from the tip of a unipolar electrode. The responses to such stimulation which are of interest here fall into three categories:
1. The emergence in consciousness of vivid and coherent
experiential hallucinations which appeared to be recollections of (or abstractions from) the subject's past experiences.
2. Changes in a patient's subjective experience of his or her relationship with the immediate environment.
3. "Illusory" emotional experiences.
Based on his analysis of the data, Penfield (1975) postulated the existence of two related brain systems: a "mechanism of recall," and a "mechanism of interpretation." The latter involved the temporal cortex (exclusive of the speech areas) and was referred to by Penfield as the "nonverbal concept mechanism." Penfield comDared its function with nonverbal concepts to the operation of the speech cortex with verbal concepts.
Somehow [this mechanism] seems to analyze the components
of sensation, compares them with previous experience,
and by that analysis and comparison, transmits into
consciousness their present and immediate significance
: I [an emotional response] is a signal that rises
into consciousness as a result of an interpretation of
what the present situation may bring the subject in
the immediate future.. (Mullan & Penfield, 1959, p. 283) It appears that the cognitive products of these mechanisms fit the criteria for generalized expectations. Penfield's evidence assists in understanding the different types of memory referred to by Rapaport and by Breuer and Freud and indicates the neural substrate of these processes. (Penfield's data and conclusions will be reviewed and evaluated in detail presently.)
It has been assumed here that at the base of personality there are mechanisms whereby cognitive operations and affective processes are appropriately activated by significant memories. The present task is to describe neurological evidence which accounts for the memory phenomena reviewed above in terms that fit the criteria for "functional systems" as defined by Luria, and satisfy the evolutionary imperatives outlined at the beginning of this chapter.
The neuropathological correlates of human amnesia syndromes and data from related animal studies will be reviewed. It will be hypothesized that human beings possess two autonomous, lateralized memory systems, centered on the hippocampi, each of which is intimately associated with its own emotional and cortical activating system. Human Amnesia Syndromes
A number of terms are used to describe different aspects of
memory function and dysfunction. Immediate memory, usually measured
by digit span, probably reflects the ability to hold information in the primary or secondary sensory cortex as long as voluntary attention (directed by the nonspecific thalamic nuclei) is focused upon it (Smithies, 1966). The terms short-term ("recent") and long-term ("remote") memory indicate recollection over increasingly greater periods of time, but both are almost certainly subserved by the same physiological processes (Brierley, 1977). The term retrograde amnesia refers to a period of time before an accident or illness for which the patient's ability to recall is diminished or lott Ant rde amnesia is an inability to retain in memory events that occur after such an injury or illness.
The combination of a severe retrograde amnesia and a debilitating anterograde amnesia is the hallmark of the Wernicke-Kors~koff syndrome, the most common form of memory disease. This illness is most fr ,quently seen as a result of brain lesions brought on by dietary (thiamine) deficiencies in chronic alcoholics, although the lesions and illness may be produced by a number of toxic or disease processes (see the excellent review by Brierley, 1977). This illness
was described independently in the 1880s by Karl Wernicke (whose work with aphasia was reviewed earlier) and S. S. Korsakoff, a Russian psychiatrist. Wernicke focused on the acute stage of the illness during which the patient is usually depressed, fearful and anxious; often paranoid; and always severely confused and disoriented (Wernicke's encephalopathy). Patients who survive this acute stage become stabilized in the phase known as Korsakoff's psychosis. This chronic state is characterized by severe memory disorders and profound changes in the patients personality, motivation, and affect.
Against a background of retained intellectual skills and intact remote memory, the victim of Korsakoff's psychosis suffers a retrograde amnesia for periods of up to several years before the onset of the illness, and an almost total inability to recall any new information once his or her attention is distracted from it. Consequently, these patients live virtually in the immediate present and are always disoriented as to time, place and situation. These patients are, at best, only vaguely aware of their inability to learn new material. They often produce confabulations to cover gaps in
their recollection and fuse or combine ("reduplicate") experiences from different periods in their lives (Gardner, 1975). In both cases they believe that their statements reflect reality. In most cases the patient's affect is blunted, although some patients have exhibited chronic euphoria (e.g., Remy, 1942). They show reduced spontaneity and initiative and a "lack of desire for alcohol, sex, and other traditional reinforcers"'(Gardner, 1975, p.'188).
(Such indifference to alcohol is especially interesting in light of the fact that the brain damage in most of these patients was caused by years of sustained, heavy drinking.)
Post-mortem examination of the brains of persons who suffered
from Wernicke-Korsakoff disease reveals lesions of certain anatomical structures in the limbic system associated with the well-known circuit of Papez. Such damage almost invariably includes, and may be confined to, injury to the mammillary bodies, the relay for hippocampal output on its way to the anterior thalamic nuclei
A pure form of this memory disorder was the unfortunate consequence of bilateral removal of the hippocampi in humans. In the 1950s Scoville performed a series of experimental operations designed to relieve the symptoms of chronic schizophrenia without the undesirable side-effects of a complete frontal lobotomy. The surgical procedure involved the resection of the medial surface of the temporal lobes from 5.0 to 8.0 cm posterior to the tip of the lobe combined in some cases with orbital undercutting. Thirty severely deteriorated schizophrenics had undergone the operation, with slight improvement in their conditions, when a purely temporal resection was performed on a nonpsychotic epileptic patient whose seizures were unresponsive to medication. When this patient, "case H. M.," recovered from the operation it became apparent that he had developed a severe amnestic disorder which resembled Korsakoff's psychosis and which persisted at 14 years (Milner, Corkin & Teuber, 1969). Scoville
and Milner subsequently examined eight of the psychotic patients
who had undergone the operation and who were able to participate in formal testing. They discovered "some generalized memory disturbance in all patients with removals extending far enough posteriorly to damage portions of the hippocampus and hippocampal gyrus (Milner, 1958, p. 112). The degree of memory impairment was more or less proportional to the amount of these structures removed. Bilateral resection of the uncus and amygdaloid nucleus alone did not result in amnesia (Brierly, 1977), nor did removal of the gyri of the outer aspects of the temporal lobes (Bailey, 1946). It has been concluded, therefore, that "the structures necessary for normal memorizing are the hippocampal formations within the temporal lobes, the mammillary bodies and, possibly, certain thalamic nuclei within the diencephalon" (Brierley, 1977, p. 221); that is, the hippocampi,. their output pathways and related projection sites.
There has been, as yet, no definitive explanation of either the nature of the amnesic deficit described above or of its underlying mechanisms. The fact that remote memory seems to be intact in these patients has led many to believe the the hippocampaldiencephalic structures are not involved in the process of recall, although Brierley (1977) pointed out that such a conclusion is unjustified in the absence of adequate pre-post evaluation of this function. Milner (1966, ch. 5) attempted to account for the pairing of a period of retrograde amnesia with an inability to learn new material by hypothesizing that the establishment of a permanent memory trace requires an extended period of "consolidation," which is somehow disrupted in this syndrome. In Milner's view the deficit
represents a failure to transfer sensory impressions into long-term store. The adequacy of the consolidation hypothesis is called into question, however, by demonstrations that amnesic patients are in fact able to recall new information under certain conditions. The most convincing evidence comes from experiments using the technique of "cued recall" in which a subject is given partial information about a stimulus (e.g., a previously presented word or picture) and asked to identify the whole item. Under these circumstances the performance of amnesic subjects was not significantly different from that of controls (Weiskrantz & Warrington, 1970). This suggested to the authors that the amnesic deficit involved problems with mechanisms of retrieval rather than those of acquisition or retention.
Weiskrantz (1979) reviewed a number of experimental paradigms in which normal learning has been demonstrated in amnesic subjects and underscored the fact that, in each case, the patients themselves persistently failed to acknowledge the fact that their performance was based on specific past experience or that they had been confronted with the task before). Thus, amnesia vi,.ctims do not have access to their memories on a conscious level, nor is such awareness necessary for that memory to be demonstrated objectively. Weiskrantz pointed out that this- "striking dissociation between the subjects' commentaries and their objective performance . suggests a dissociation between levels of processing rather than a failure on any particu-lar level" (p. 385).
Levels of processing in memory are the subject of a theory
(summarized by Gaffen, 1972) which is based on arguments by Talland (1965) and supported by experimental evidence (Peterson, 1967; Kintsch, 1970). Briefly, the theory postulates that the process of recall consists of two separate and autonomous stages: retrieval (or search) and recognition. The retrieval process "proceeds at several levels . each being terminated by an implicit act of recognition" (Talland, 1965, p. 304). The recognition stage is based on a record from which the past cannot be read directly, but which can assign a particular response in a particular context a value of "familiarity--unfamiliarity" (the correct response being the most familiar in that context). Thus "[in the retrieval stage] various responses are generated (but not emitted); when finally the correct response is generated, it is recognized as such by the recognition stage, and is then emitted" (Gaffen, 1972, p. 328). The theory postulates that amnesic subjects (animal and human) lack the faculty of discriminating familiarity. This basic deficit is manifested in the premature termination of search cycles, resulting in an incorrect match. These formulations are not inconsistent with those of Butters and Cermack (1974), who concluded from their experiments that increased sensitivity to proactive interference, subsequent to inappropriate encoding of information, was the critical factor- underlying the amnesic disorder. Finally, it is interesting to note that modern theories regarding amnesia seem to have arrived at the point at which they began: Korsakoff (1889), in keeping with the associationist doctrine of his time, believed
that his patients were deficient in making associations among new ideas and in-connecting past and present experience.
The suggestion of bilevel processes in memory noted above are particularly interesting given that language and conscious awareness have been associated with temporal lobe structures of the left hemisphere, and Penfield's (1975) report that electrically induced "illusions of recognition" were elicited only by stimulation of temporal structures of the right hemisphere. Memory and the Neocortex
As noted earlier, Penfield's experiments with electrical stimulation of the cortex in conscious patients led him to postulate the existence of two separate memory systems: a "mechanism of recall" and a "mechanism of interpretation." The existence of the former was suggested by the fact that, following stimulations of the exposed cortex, some of his patients reported vivid auditory and/or visual experiences ("flashbacks"); it seemed to the patients as if they were reliving prior experiences in their lives, although they retained their awareness of the operating room environment. Since many of these sensory sequences were trivial, yet perceived as familiar, Penfield concluded that his electrode was tapping a "continuous record of conscious experience." Although widespread areas of the cortex were exposed and explored, these "experiential responses come only from the temporal lobe, never from any other part of the brain" (Penfield, 1975, p. 31).
Niesser (1967) presented persuasive arguments refuting Penfield's claim that "nothing is lost . the brain of every man contains an
unchanging ganglionic record of subjective experience" (Penfield, 1954, p. 67). Niesser suggested that "most of the cases described by Penfield seem more like generic and repeated categories of events rather than specific instances" (p. 168). (It will be noted that Niesser's formulation is congruent with the notion of a "generalized expectation," as defined earlier.)
Penfield's second mechanism was suggested by another category of electrically induced phenomena which consisted of the "misrepresentation or altered interpretation of present experience" (M611an & Penfield, 1959, p. 269). Prominent among these were "illusions of recognition" during which "present experience seemed familiar, strange, altered, or unreal" (p. 270).' These "illusions of comparative interpretation" were associated with stimulation of the temporal cortex in the hemisphere that was minor for handedness and speech. The authors believed that "in normal life, these are signals that rise into consciousness, signals tnat depend on subconscious comparison of past experience with the present" (p. 283).
In 1951 Penfield proposed that portions of the temporal lobes be called "memory cortex" in the belief that his electrode had activated a neuronal record which was stored there. He was obliged to revise this theory in 1958 because of a new understanding of the physiology of electrical brain stimulation. When an electrode passes a current into the cerebral cortex, the current completely disrupts the patient's normal use of that gray matter (e.g., stimulation of the speech areas produces momentary aphasia). Therefore, any positive responses are produced by axon-conduction and the functional
activation of a distant, secondary ganglionic station (Penfield, 1975, ch. 7). In his later formulations then, Penfield referred to those temporal structures as the "interpretive cortex" and postulated that his electrode had activated a final common pathway to a secondary center which in turn produced the illusions of comparative interpretation. Since the temporal lobe forms the principal source of input into the hippocampus (which was known to be related to memory), Penfield assumed that this was the secondary center in question. He suggested that
The hippocampi seem to store keys-of-access to the
record of the stream of consciousness. With the
interpretive cortex, they make possible the scanning
and the recall of experiential memory.* (Penfield,
1975, p. 36)
Penfield's finding that illusions of familiarity were associated with activity of the temporal lobe in the right hemisphere is complemented by Kimura's (1963) evidence that the right temporal lobe appears to be more involved in the analysis of unfamiliar stimuli. Kimura presented familiar and unfamiliar visual stimuli to the right and left visual fields of patients with lesions of the right or the left temporal lobe. The right temporal group was impaired in the perception of the unfamiliar stimuli but not the familiar. Kimura interpreted her results in terms of the verbal identifiability of a stimulus:
It seems clear that a frequent (though not a necessary) concomitant of familiarity in a perceptual sense is the
possibility of verbal identification. Where increased
familiarity with a stimulus object, or class of objects,
is associated with the repeated naming of the object,
the ability instantly to attach a name to it represents an important step in the development of a concept. It
seems probable that in such cases a large part of the increase in permanent neural -. epresentation which is assumed to correlate with familiarity will take place
in the language centers, that is, in the dominant
hemisphere. (Kimura, 1963, p. 269)
Thus, unfamiliar stimuli, which are not represented in verbal memory by a permanent neural model (e.g., a name or concept) are more likely to be processed by the right hemisphere, which Kimura suggests is more important, than the left in the establishment of such "cell assemblies." Since all of the material in the memory store are initially unfamiliar it follows that many, if not all, verbal concepts (in the left hemisphere) might be based on neural models, or gestalto7n, which were assembled (and are stored) in the right hemisphere. Such an hypothesis is supported by evidence from split-brain studies that the right hemisphere is far superior to the left in the discrimination of part-whole relationships (Nebes, 1974).
The appearance of material-specific amnesia syndromes following unilateral temporal lobectomies suggests that the isolation of language in the left hemisphere extends to verbal learning also, and is thus virtually complete. Milner (1971) reports that,
A comparison of left and right anterior temporal lobectomy in epileptic patients has revealed certain specific
memory defects that vary with the side of the lesion.
These material-specific disorders are to be distinguished
from the global amnesia that follows bilateral damage
in the hippocampal zone (Milner, 1958). Thus, left
temporal lobectomy, in the dominant hemisphere for
speech, selectively impairs the learning and retention of verbal material (Meyer & Yates, 1955; Milner, 1958).
...Conversely, removal of the right, nondominant
temporal lobe leaves verbal memory intact but impairs
the recognition and recall of visual and auditory patterns
that do not lend themselves easily to verbal encoding.
...Thus within the sphere of learning and memory
there is a double dissociation between the effects of
these two lesions.* (Milner, 1971, p. 274)
Butters and Cermack (1974) focused on the specifically verbal aspects of the amnesic disorder in Korsakoff patients and noted that, during learning tasks, these patients did not react to changes in semantic categories on successive lists. They concluded that the amnesic deficit was due to the patient's
inability to encode verbal information along semantic
or meaning dimensions. . Korsakoff patients do not
spontaneously employ semantic encoding strategies,
but rely on basic acoustic and associative categorizations.
If the Korsakoff patient is instructed to encode semantically. he will do so, but in an impaired manner.
Butters and Cermack assumed that the Korsakoff patients'
deficient utilization of "meaning" in learning tasks was a specifically verbal (i.e., left hemisphere) phenomenon. However, Gazzaniga and his colleagues produced evidence which suggests that the right hemisphere may play an important role in imparting "meaning" to verbal memory processes. These authors tested patients with partial or complete section of the cerebral commissures for recall of two lists of paired-associate nouns. On presentation of the second list each patient was instructed to "form a 'picture in his mind' of the two items interacting in some unusual or amusing way" (Gazzaniga, Risse, Springer, Clark & Wilson, 1975, p. 12). Patients with partial sectioning of the cerebral commissures showed marked improvement with the imagery instructions but none was seen where there was complete section of the hemispheric interconnections.
The evidence reviewed thus far suggests that the central problem in the amnesia syndromes involves mechanisms which normally facilitate access to stored memory traces. The problem might occur at the point of encoding and deposition, or retrieval, or both. The hippocampus appears to be critical to these coding and decoding operations; this structure may normally provide the memory "cues" which must be externally administered in its absence.
The evidence is supportive of Niesser's (1967) proposal that
(rather than tapping a "continuous memory strip") "Penfield's electrode may have touched on the mechanisms of perceptual synthesis" (p. 169). There can be little doubt that Penfield's "final common pathways" in the temporal lobes are related to hippocampal afferents. However, it is now clear that the hippocampus cannot be the only secondary ganglionic station which must be activated before a signal indicating familiarity, originating in the right hemisphere, can "rise into consciousness." Studies of split-brain patients have shown conclusively that the ability to give a verbal account of events occurring in the right hemisphere is dependent on the integrity of the cerebral commissures (e.g., Sperry, 1968). The general layout of the commissures is such that a specific area in one hemisphere is connected via commissural fibers to the homologous area in the contralateral hemisphere. The temporal lobes have their own private interconnection in the anterior commissure, wiiijch also connects the two amygdalae (Gray, 1977). Studies of patients with only partial sectioning of these commissures have indicated
that highly processed information might be even more "transferrable" than elementary sensory information and may be able to utilize any commissural pathway that is available (Gazzaniga- et al., 1975).
It appears that there are two separate and autonomous memory systems in the brain which are specialized as to their function: a verbal system, lateralized in the left hemisphere, and a nonverbal (experiential) system lateralized in the right. The evidence suggests that memory functions might be conceptualized as having both vertical and lateral dimensions: the scanning (or search) within each system may involve a temporal-hippocampal interaction and the phenomenon of recognition may be a function of right-left temporal lobe interaction.
A voluntarily initiated search of memory would, most certainly, begin in the left-hemisphere system, but an environmental stimulus might activate an initial nonverbal (right-hemisphere) scan and analysis. In either case, the result of these processes would seem to be the emergence in conscious awareness of a signal indicating the "familiarity" (or "strangeness") of the stimulus and, finally, the facilitation of generalized expectations and verbal associations related to that stimulus. The implications of such a formulation
for the understanding of psychopathological processes and the practice of psychotherapy will be discussed in later sections, following an evaluation of the limbic system's role in experiential and emotional memory.
The Limbic System, RAS, and Memory
Little distinction is made in the human amnesia literature
between the verbal, experiential, and emotional aspects of memory. The verbal manifestations of the disorder are the most obvious, the most amenable to description and testing, and so have become the primary focus of scientific attention. Although it is seldom emphasized, most case studies of amnesia victims relate anecdotal evidence of emotional dysfunction (e.g., blunted or flattened affect; euphoria). There are also more or less vague, but consistent, references to what might be termed disturbances of tonic arousal ("decreased spotaniety," "lack of initiative," "indifference," "passivity"). These emotional and arousal difficulties appear to be associated with systems centered on the amygdala and RAS, respectively. Both of these will be described below following a consideration of the nonverbal, predominantly unconscious mechanisms of experiential memory.
Before an organism can respond to a stimulus on any level
(emotional, verbal, or behavioral) its meaning must be ascertained. At the most elementary level the organism's survival depends on its ability to make appropriate decisions about whether to invest neural energy in attending to and further analyzing a particular stimulus (orienting) or to ignore it (habituation). Such a decision demands a judgement as to the apparent novelty, possible significance, or lack of these qualities in the stimulus configuration, a process which requires access to a patterned memory trace or "neural model" (Sokolov, 1963). Such a process must begin with sensory
input and end with afferents which are capable of modulating the
activity of the brainstem RAS.
A determination of novelty or significance might be made at the level of the secondary sensory association cortices where raw receptor impulses are converted into functional (i.e., "meaningful") information, although such duplication of effort in all the modalities would be cumbersome and inefficient. Further, the relative significance of a stimulus may depend on the internal state of the organism (e.g., satiation, the presence or absence of certain hormones, etc.). The fact that information related to this added consideration is most readily available in subcortical structures is another argument favoring a centralized location for the mechanisms involved with the decision to orient or habituate. The hippocampus meets all of the criteria specified above.
Luria concluded that "many nuerons in the hippocampus and connected nuclei do not respond to modality-specific stimuli of any sort, but serve to compare present stimuli with traces of past experience; they react to every change in the stimulus and thus play to some extent the role both of 'attention neurons' and of 'memory neurons"' (1973a, p. 289). According to Luria the hippocampus provides for the "elimination of responses to irrelevant
stimuli and enables the organism to behave in a strictly selective manner" (1973a, pp. 271-272). It appears that the hippocampus accomplishes this complex task by coordinating the activities of the cortical and subcortical mechanisms which are directly involved in the processes of attention, memory and learning.
Early investigators were puzzled by the fact that cortical
activation (EEG desynchronization) was accompanied by synchronous slow-wave activity (4-8 Hz theta rhythms) in the hippocampus (e.g., Green & Arduini, 1954). The most constant behavioral correlation of hippocampal theta activity in animals of different species is orienting towards, and attending to, stimuli in the environment (Isaacson, 1974). Cortical activation, such as that seen in the orienting response, is accomplished by the brainstem RAS in conjunction with the nonspecific thalamic nuclei. The hippocampus appears to regulate the process of involuntary attention by performing switching functions through a mutually inhibitory relationship with the reticular formation (Smithies, 1966). The Soviet neuropsycholgist Vinogradova provided important insights into the mechanics of this process.
By observing unit activity with microelectrodes Vinogradova
(1970) determined that all of the neurons in the hippocampus monitor incoming stimuli, habituating to repetition and dishabituating to any change in the stimulus configuration. Such responsiveness requires constant matching of the stimulus with a related neuronal model (Sokolov, 1963). That these models exist in the cortex, and not in the hippocampus itself, is established by evidence that the quality of sensory information is almost completely erased in hippocampal neurons (Gloor, 1961).
Vinogradova distinguished two types of neurons in the hippocampus: A-neurons (30-40%) which are activated by a stimulus and I-neurons (60%) which are inhibited. She went on to propose a mechanism whereby the hippocampus is able to modulate the processes of attention and learning:
The hippocampus exerts a tonic inhibitory influence upon
the reticular formation, blocking activatory processes
through the tonic discharge of its I-neurons when
novelty is absent and registration [a change in the
neuronal model] is not needed. But when a stimulus which
is not registered in the memory system appears, this
inhibitory control is blocked (I-neurons become silent),
arousal occurs, and the process of registration starts.
(1970, p. 114)
Vinogradova's hypothesis is supported by observations of electrical activity in the brains of animals in classical conditioning paradigms. Theta rhythms (associated with activity of Vinogradova's A-neurons) are found in the early stages of learning but disappear when the response has become well established (Isaacson, 1974). During conditioning the time course of the theta rhythm and the orienting response are matched. As the latter is replaced by the stabilized conditioned response, theta dies out (I-neurons become active again) and the hippocampus resumes its inhibitory control-over the RAS, thus ending the orienting response and thereby allowing the fully developed conditioned response to materialize (Smithies, 1966, p. 90).
The hippocampal-cortical interaction was apparent in an
analysis of the characteristics of hippocampal theta rhythms in situations requiring different types of cortical information processing. Bremner (1970) investigated the effects of orienting, simple conditioning, discrimination, and discrimination reversal tasks on various parameters of the theta rhythm using the habituated organism (rat and man) as a baseline. He found that theta*.power (amount of energy) increased in the presence of stimuli which elicited orienting and arousal and decreased in the interval preceding a response in the conditioning situation; the range of energy distribution
around the peak frequency narrowed during discrimination; and the location of the peak shifted in discrimination reversal procedures.
In summary, the hippocampus appears to be able to monitor
incoming stimuli and match them against (cortical) neuronal models which represent the past experience of the organism with related stimuli. In addition to facilitating appropriate access to these memory traces hippocampal activities have been directly associated
with the triggering of cortical processes which permit further analysis of a stimulus, emotional and behavioral responses as warranted, and/or alterations in the neuronal model itself (i.e., learning).
Papez Circuit and Memory
The hippocampus forms part of a continuous pathway within the limbic system which Papez (1937) believed to be the substrate of emotion. Papez was aware that destruction or stimulation of limbic structures produced major alterations in emotional behavior and believed that emotional expression depended on the integrative action of the hypothalamus. He was also convinced that subjective emotional
experience required the participation of the cerebral cortex. Papez outlined an anatomical circuit through which he thought
emotion might arise in either of those two centers. Thus
Incitations of cortical origin would first pass to the hippocampal formation and then down by way of the fornix to the mammillary body. From this they
would pass upward through the mammillothalamic tract
t'eto the anterior nuclei of the thalamus and thence
by the medial thalamocortical radiation (in the cingulum) to the cortex of the gyrus cinguli .
Radiations of the emotive process from the gyrus cinguli to other regions in the cerebral cortex
would add emotional coloring to psychic processes
occurring elsewhere. (Papez, 1937, pp. 304-306)
Papez believed that sensory input to the system originated in the thalamus and was communicated via the subthalamus to the hypothalamus. The circuit is completed with the connection of the cingulate gyrus to the hippocampus by way of the cingulum bundle.
It is now known that other limbic structures are more actively involved in the specifically emotional processes. However, Papez's anatomical concepts might be rehabilitated if their context were changed from emotion to memory. Input to the hippocampal circuits would be seen as processed sensory information and its output as
memory indexing information capable of "cueing" associations and generalized expectations related to the input stimulus. It remains to place the role of the hippocampus in the context of the functional system it subserves and to examine the other components of that system.
Fornix. Hippocampal output makes its way via direct and
indirect pathways to the thalamus (anterior, dorsomedial, and intralaminar nuclei), the hypothalamus, and the midbrain reticular formation. Its main efferent fiber system, the fornix, is composed of axons from pyramidal cells in the body of the hippocampus. These fibers converge in the fimbria, traveling backwards within the temiporal lobe, and then arch forward under the corpus callosum as the crura (posterior pillars) of the fornix. Here a number of fibers cross to the other side, forming the hippocampal commissure. The two crura then join to form the body of the fornix which
continues to arch forward, following the course of the lateral ventricle to the rostral edge of the thalamus. Here the bundles separate again to form the anterior columns of the fornix which curve downward in front of the intraventricular foramen and above the anterior commissure (AC). Approximately half of the fibers descend behind the AC as the postcommissural fornix; the other half, in a less compact bundle, pass in front of the AC as the precommissural fornix. Postcommissural fibers pass through the hypothalamus to
the mammillary bodies, giving off fibers to the thalamus on their way. Some of the precommissural fibers distribute to the septal area and others join with septal fibers and continue into the same areas as the postcommissural fornix. Approximately one-third of the fornix fibers reach the mammillary bodies (Daitz, 1953); the anterior nuclei of the thalamus receive as many direct fibers from the fornix as they do from the mammillothalamic tract (Truax & Carpenter, 1969, ch. 21).
Liss (1968), working with the rat, found that hippocampal
and fornix lesions had analogous effects on learning and behavior in passive and active-avoidance tasks. In the monkey, fornix lesions led to impaired learning of a spatial reversal task that was "functionally similar" to the deficit seen after hippocampal removal (Mahut, 1972; Mahut & Zola, 1973). Gaffen (1972) reported a series of six experiments in which rats with fornix lesions were shown to have a defect in "recognition memory" that the author argued was equivalent to anterograde amnesia in humans. In Russel's (1971) survey of brain wound cases, the eight patients who showed
a Korsakoff-type of memory disorder were thought to have had their fornices damaged by metal fragments. Four of the five cases with typical amnesic syndromes in Jarho's (1973) study of brain-injured war veterans were considered to have bilateral interruption of the fornix or mammillothalamic tract.
Sweet, Talland, and Ervin (1959) reported the case of a woman
in whom the anterior columns of the fornix were sectioned to facilitate the removal of a colloid cyst of the third ventricle. This patient showed a rapid recovery of old skills, but developed a "severe loss of memory for recent events which persisted at two years] . a retrograde amnesia of at least several weeks and a subjective
complaint of amnesia for specific events of four or five years past" (p. 76), coupled with intact remote memory. In the discussion following the presentation of this patient, Brenda Milner reported on a similar case operated on by Welsh in 1954. This patient showed a gross initial memory disorder but, at one year, was able (with effort) to effect some compensation for his defect: "Whenever he deliberately sought associative links he was able to improve his performance considerable" (Sweet. et al., 1959, p. 79). Milner concluded that.
I think that one can only account for the paradoxical
diversity of data from the fornix cases, as contrasted
with the consistent and severe memory loss in the hippocampal cases (and maybe in the mammillary body cases
also), by supposing that you are only interfering
with a part of the system by fornix section. Thus,
you are most apt to see a temporary disruption of disturbance of memory with minimal residual loss. '(p. 79)
Data on the fornix is relatively scarce, and there have been reports of negative findings. In his influential review, Brierly (1977) took a very conservative stance on this subject:
The most discrete link between the hippocampal and
diencepahlic regions is the fornix. It is surprising,
therefore, that with the exception of the case reported
by Sweet, Talland and Ervin (1959), bilateral division
of the fornix (usually in the region of the intraventricular foramen) has not resulted in a disorder of memorizing (Dott, 1938; Cairnes & Mosberg, 1951;
Garcia-Bengochea and his colleagues, 1954). This finding suggests that the two groups of structures
linked by the fornix cannot be regarded as a unitary
system subserving the process of memorizing, at least
until major interconnections other than the fornix
have been identified. (pp. 221-222)
Other interconnections are available. Smithies (1966) describes a "massive direct hippocampal-hypothalamic pathway" that runs
diffusely through the subthalamus and which he suggests miaht be "quite able to carry on hippocamoal and limbic circuit function in the absence of [the fornix]" (p. 122). A seond look at the reports
cited by Brierley, however, allows the possibility that his conclusions are premature.
Sweet et al. (1959) emphasized that their patient's "conversations, social amenities, and general demeanor gave little evidence of [her] severe deficits unless they were specifically looked for" (p. 76). They also noted that she lacked spontaniety, made little
effort to converse, and was apparently indifferent to her deficits (cf. Korsakoff's syndrome). The very brief report of Bengochea, De la Torre, Esquival, Vieta and Fernandez (1954), after a "short follow-up" of their patients (whose fornices were severed in an experimental operation to relieve intractable epilepsy) did not
mention any attempt at quantification of behavior. They simply stated that "so far, in none of the 12 surviving cases there has been [sic] any unfavorable neurological or psychiatric sequela" (p. 177). It seems possible that these authors may have missed subtle symptoms in their apparently superficial evaluation.
Wilder Penfield (quoted in Sweet et al, 1959) underscored the fact that "patients who have [bilateral hippocampal lesions] do not forget their skills. Two of them were able to carry out most complicated skills learner previous--glove cutting and engineering drawing" (p. 81). Unfortunately, the only behavioral measure reported by Cairnes and Mosberg (1951) involved a return to work. These authors noted that some of their pateints (who had incurred fornix damage in the course of surgery to remove colloid cysts of the third ventricle) showed initial confusion, loss of memorizing, and amnesia for the period surrounding the operation, but: "after operation all [but one of their nine surving cases] returned to work, and . .showed no disturbance of emotion or intelligence" (p. 564). Thus
Four of the five young women ...are doing normal
housework; three have borne childrn. The other young woman is in regular work as a clerk, and is
free from complaints . . Two older women ...
are also doing their housework [although one has a
'slight impairment of memory'] . . Of the two men,
one is working regularly as a policemen. (p. 568)
The extent of the lesions in these patients is unclear. The authors report only that "each had . partial or complete division of the anterior columns of the fornix" (p. 564). (It seems possible that these surgical fornicotomies spared the precommissural fornix.)
It should be noted that a colloid cyst of the third ventricle tends to produce confusion, dulling of attention and memory, and sometimes a progressive dementia prior to its surgical removal. These factors would make a pre-post evaluation of memory function very difficult. Still, the scantiness of the reported date in the studies reviewed above is unfortunate. It is evident that injuries to different parts of this system result in different expressions of the disorder. It is reasonable to conclude, however, that damage to the fornix has adverse effects on memory function which vary as to the quality and degree, and may leave a greater possibility
of recovery of function.
Mammillary bodies and mammillothalamic tract. The mammillary bodies are a collection of nuclei at the posterior boundary of the hypothalamus. They form a major relay station for hippocampal output on its way to the thalamus (via the mammillothalamic tract) and to the midbrain reticular formation (by way of the mammillotegmental tract).
In humans, damage which is apparently limited to the mammillary bodies has resulted in the full Korsakoff amnesic syndrome (Remy; 1942; Delay & Brion, 1951; Gruner, 1956; Symonds, 1966), although Victor (1964) suggested that additional damage to the thalamus was necessary to produce the disorder. In the rat, lesions of the mammillary bodies or of the mammillothalamic tract impaired the ability to perform a spatial discrimination in a T-maze in order to avoid footshocks (Thompson, Langer & Rich, 1964). Krieckhaus (1962, 1964) found that complete or partial destruction of the
mammillothalamic tract in the cat reduced the retention of a two-way active avoidance task and produced a less striking deficit in the retention of a one-way active avoidance task. These findings were later replicated in the rat (Krieckhaus, 1965). Thomas, Frey, Slotnick and Krieckhaus (1963) studied the post-operative acquisition of the two-way active avoidance learning task and reported mixed results: four of their eleven cats were completely unable to learn the task, while seven of them mastered the problem within the number of trials required by normal animals. (The significance of the distinction between acquisition and retention will be discussed in the following section.)
Cingulate cortex. The cingulate cortex lies above the
corpus callosum on the medial side of the hemisphere, separated from the neocortex above by the cingulate sulcus. It merges with the hippocampal gyrus posteriorly and with the neocortex of the frontal lobe anteriorly. As noted by Papez, the cingulate gyrus receives its main afferent supply from the anterior nuclei of the thalamus and projects to the hippocampus via the cingulum bundle. Stimulation of the cingulate cortex also produces activity in the prefrontal and orbitofrontal regions of the neocortex (Dunsmore & Lennox, 1950). There are reciprocal connections with the anterior and other thalamic nuclei (including the dorsomedial). A strong projection to the interior parietal lobule (IPL) in the post-central neocortex has been demonstrated in the monkey (Mesulam, Van Hoesen, Pandya & Geshwind, 1977). These authors, using the horseradish peroxidase technique, found that "the cingulate gyrus contained one
of the heaviest concentrations of labeled neurons in most cases" following injection of that substance into the IPL (p. 324).
The literature documenting the efforts of physiological
psychologists to define the role of the cingulate cortex in learning and memory is confused somewhat by varying interpretations of the data by those authors (an objective review is available in Isaacson, 1974). Several experimenters have ascribed the learning deficit which follows cingulate lesions to an enhanced fear response. However, Kimble and Gostnell (1968), using two different behavioral measures, failed to find any support for this hypothesis. Lubar, Perachio, and Kavanagh (1966) suggested that incidental damage to the visual cortex could account for the deficits following cingulate lesions, but adequate control lesions (e.g., Kimble, 1968) and lesions which avoid damage to the visual cortex (Trafton, Fibley & Johnson, 1969) are adequate proof of the cingulate's role in the observed impairments (Isaacson, 1974). The essential findings on the effect of cingulate lesions on memory in the rat and cat are straightforward: such lesions impair the ability of these animals to acquire active avoidance conditioning (Peretz, 1960; McCleary, 1961; Thomas & Slotnick, 1962; Lubar & Perachio, 1965; Kimble & Gostnell, 1968; Trafton et al., 1969).
A number of relatively comnlex deficits have been noted after cingulate damage which resemble symptoms seen in human amnesia syndromes. These include disruption of the orderly-sequencing of behaviors (in nest-building and maternal behavior, Stamm, 1955; Slotnick, 1967), deficits in the temporal ordering of responding
(in runway problems and bar-press alternation: Barker & Thomas, 1965, 1966; Barker, 1967); and the failure to exhibit behaviors which were indicative of opium addiction (Marques, 1971). The last, especially, is supportive of the hypothesis voiced by some authors that cingulatelesioned animals are unable to anticipate the emotional consequences of their behavior for both rewards and punishments (Glass, Ison, & Thomas, 1969; Isaacson, 1974).
Anterior cingulectomy has been termed the psychosurgical
"operation of choice" for the treatment of severe obsessional and anxiety disorders (Lewin, 1961; Whitty, 1966). Whitty noted that one of the long-term effects of cingulectomy was "relative neglect of the impact of external events." Finally, Pechtel, McAvoy, Levitt, Kling & Massermann, (1958) concluded that lesions of the cingulate gyrus in humans resulted in, among other things, "amnesia for previous learning" and "impairment of new learning skills." Discussion
The failure to transfer learning between hemispheres seen in
cerebral commissurotomy preparations (animal and human) indicates that the storage of memories is a neocortical function (Geshwind, 1965). Generalized memory disorders, on the other hand, are only produced by bilateral damage to the diencephalic structures of the hippocampal system. It is evident that these structures in the circuit of Papez form part of a functional system which monitors the environment, matches incoming stimuli with representations of the organisms previous experience with similar stimulus configurations, activates the organism in the presence of potentially significant stimuli,
and finally, causes pertinent information concerning that stimulus to be presented to conscious awareness. It appears that there are two such systems in the human brain, each specialized to deal with a different type of information. The first, lateralized to the right hemisphere, performs the functions listed above with experiential data. The second, operating in the 16ft hemisphere, deals with language and verbal concepts which may be based on gestalten that are assembled and stored in the contralateral system. All of the functions noted above are impaired, to a greater or lesser extent, in various manifestations of the amnesic syndrome.
The evidence suggests that some learning does not take place in amnesia victims, that is, memory traces are stored. However, the amnesic subject has difficulty gaining access to those memory traces when they are needed. More specifically, they are unable to recognize and select the appropriate memory trace in a given situation from the set of available traces. Access to the proper trace is facilitated with adequate cueing. It appears that the function of the hippocampal system is to assure the activation of appropriate memory traces based on the requirements of the situation. In the present formulation, significant (experientially based) memory traces have been designated by the superordinate term "generalized expectations." It appears that the hippocampal system is responsible for the generalizing of these expectations from one situation to another, similar, situation.
In lower form, where survival is dependent on instincts, the
identification of a significant stimulus may culminate in the release
of species-specific behaviors that are organized at a subcortical level. The elicitation of unlearned behavior sequences (related to feeding, fighting, fleeing and reproduction) by hypothalamic stimulation is well established and seems to be related to programs: stored in the midbrain or brainstem (Isaacson, 1974). In humans, however, cortical functions are more important in determining the behaviors that insure survival.
Papez's circuit is an "anatomical identity which connects the temporal and cingulo-frontal cortex of both hemispheres" (Barbizet, 1963). Given the recently identified massive interconnection of the cingulate gyrus with the inferior parietal lobule (IPL), Papez's circle may be seen as forming a part of a larger cortical circuit (see Fig. .). As such, it is in a position to directly modulate the activity of the two areas of the brain that are identified with the highest levels of information processing: the specifically human tertiary association areas of the prefrontal lobe and the IPL.
Information flow within this cortical-hippocampal circuit
might be seen as follows: raw sensory data enter the system in the primary projection areas and are processed in the secondary association areas of the post-central cortex; as this information acquires meaning it is passed on to the temporal cortex and into the hippocampal circuit; here the meaningful bits are translated into a code that emerges in the cingulate cortex as memory indexing information; this, in turn, is referred to the IPL where it facilitates associations that fine-tune the continuing input into the system. This process continues until an adequate match is made and recognized at the
u c 0
(3) S- 4-) 4-) L) (o 00)
C: S- r- -0
< -- Su >. M
(Z 3: E
A CL- M 4-)
=3 N CL
x a) CD F .......... .
CII S- F= S- 0
LL. 4-) 4-)
level of the temporal cortex. This match would take the form of a
finished "gestalt" in the experiential system and a formal "concept" in the verbal system. With the appearance of such a match the temporal component would terminate the search process and relay the product to the c~ntralateral system. Thus, the recognition of a gestalt might initiate the search for a related verbal concept or, conversely, a verbal formulation may trigger a scan for pertinent experiential associations. The description of these two complementary systems seems to provide adequate explanation for the qualitatively different types of memory which were described by Breuer~and Freud and by Rapaport, respectively as noted at the beginning of this section.
Substantial support for the model described above may be found in a situation where hippocampal activity is induced from within the limbic system (rather than from the neocortex, as was the case in Penfield's experiments). Such is the case when amygdala stimulation produces after-discharges in the hippocampus. Halgren (1981) reviewed the effects of amygdala stimulation in conscious humans and noted that such stimulations sometimes produces "complex formed hallucinations, sometimes complete scenes as in a dream or vivid recollection and sometimes more vague, apparently similar to an intruding thought .. and illusions of familiarity (deja vu)" (p. 345). Halgren suggested that it is the activation of distant normal tissue subsequent to amygdala stimulation which produces the resulting mental phenomena. He cites evidence that
Amygdala stimulation seldom evokes a mental phenomenon
unless it also evokes an after-discharge . most amygdala stimulations, even if they evoke an afterdischarge (AD), do not evoke any reported mental phenomenon . thus, there is no direct connection between amygdala activity and hallucinatory experiences ....
Simultaneous recordings from multiple brain areas
indicate that amygdala ADs seldom remain localized.
Initial spread is to the ipsilaterla hippocampus and
hippocampal gyrus. AD may remain confined to these
structures, in which case no mental phenomenon is
necessarily evoked. Further spread is to the ipsi
lateral limbic cortex (orbital, insular and cingular) and diencephalon (especially the anterior nucleus of
the thalamus, but also to the centre median, pulvinar
and dorsomedial nuclei). ADs seldom spread to the
neocortex, which may however desynchronize." (pp. 396397, emphasis added)
These findings are congruent with the present formulation, which would interpret the appearance of "complex formed sensory hallucinations" following the stimulation described above as the result of hippocampal output evoking experiential memory traces in the posterior association cortex by way of the anterior thalamus and cingulate cortex. The appearance of identical "experiential hallucinations" following temporal lobe stimulation and limbic system ADs supports Penfield's contention that final common pathways in the temporal lobe may produce activity in the hippocampal system which utlimately results in the appearance of mental phenomena.
The hippocampal system is directly involved in the mechanics
of learning. Hippocampal activity seems to assure that the organism attends to novel stimuli and sets the conditions which permit changes to be encoded into the neural models which form the centeral representations of those stimuli. Pribram and McGuinness (1975) suggested that such changes in neural represenations "may be conceived as
changes of state, set, or 'attitude'" (p. 132). Such experientially based alterations in the disposition of the organism relative to a stimulus object, situation, or event provide an operational definition for the generic term "generalized expectation" as used in the present formulations. This type of memory appears to be the "idea" which Breuer and Freud believed to be of central importance in the etiology of psychoneurosis. (It will be suggested that it is precisely these generalized expectations, in this memory system, which will provide the answer to the question posed in the introduction: "What is to be changed in the process of psychotherapy?")
In addition to their mnemonic defects, amnesia victims with damage to the hippocampal systems also present anomalies in their affect and arousal. The latter may be traced to the failure of the hippocampus to trigger cortical activation (via the RAS) as it normally does in conditions of uncertainty or significance. The former seems to reflect the disruption of hippocampal modulation
of the emotion/motivation processes by way of its interactions with other limbic structures and with the prefrontal areas (via the dorsomedial thalamus).
Interfaces and Interactions of the Monitoring
Motivating, and Mobilization System
Pharmacological research with animals and humans has implicated the two classes of biogenic amines in the modulation of fundamental brain processes. The catecriolamines dopamine (DA) and its metabolite norepinephrine (NE) have been associated With cognitive function and dysfunction (e.g., schizophrenia) while affective disorders seem to
involve an interaction between NE and the indolamine serotonin (5-HT). Histological researchers have identified six major mono amine systems in the rat brain (Fuxe, 1965; Fuxe, Hamberger &
The nigro-striatal DA system originates in cell bodies in the
substancia nigra whose axons extend through the lateral hypothalamus to terminate on cells in the caudate nucleus. This tract is known to regulate the activity of the ancient extrapyramidal motor
The meso-limbic DA system originates in the ventral tegmental area of the pons and projects mainly to the septal area and related nuclei in the limbic forebrain. Dysfunction in this system is
assumed by many to be the source of schizophrenic disorders.
The meso-cortical DA system is less well defined but appears to include projections from the ventrotegmental to the frontal cortex and from the substancia nigra to the anterior cingulate cortex (see Meltzer, 1979).
The ventral NE system originates in the reticular formation (medulla and pons) and ascends through the median forebrain bundle (MFB) to terminate on cells throughout the hypothalamus and amygdala. Stein, Wise and Berger (1972) suggested that this system primarily regulates motivational activities.
The dorsal NE system arises in the locus ceruleus in the pons and may supply up to 70% of the NE in primate brains (Redmond, 1979). The axons in this system ascend through the MFB and give off branches to the hypothalamus, the hippocampus and amygdala,
the septal area (which receives the bulk of these terminals), the anterior-ventral thalamus, the cingulate gyrus and the neocortex. Stein and Wise (1971) suggest that this dorsal NE system is involved in regulating cognitive activities.
The ascending serotonin (5-HT) system originates in the median raphe' nucleus which is situated in the core of the reticular formation and receives collateral branches from the sensory nerves. 5-HT cells in the raphd; have terminations throughout the central gray of the midbrain. Axons from the raphe' also rise in the MEB and distribute to the same areas as the dorsal NE system, with the addition of terminals in the basal ganglia and more widespread distribution in the neocortex. This system seems to be most directly involved in regulating tonic arousal.
Brodie and Shore (1957) suggested that NE and 5-HT exert
opposing effects that modulate a variety of CNS functions (e.g., sleep, appetite, sexual drive). It appears that these parallel systems control behavior by reciprocal action in balanced systems.
While functional specificity is determined by the neural circuitry involved at a given anatomical level, the cumulative effects of specific functional outcomes produced a preponderance of one or
the other transmitter seem to be consistent and to result in coordination across levels. It appears that a preponderance of the indolamine 5-HT, for example, results in the release of tonic mobilizing energy at the brainstem level, the suppression of ongoing behavior at the hypothalamic level, the identification of negativee reinforcement" at the limbic system level, and