A TRIUNE CONCEPT OF THE BRAIN AND BEHAVIOUR

THE CLARENCE M. HINCKS MEMORIAL LECTURES, 1969

EDITED BY

T.J. BoagMo
D. Campbell PHO
A triune concept of
the brain and behaviour
BY Paul D. MacLean MD

Including
PSYCHOLOGY OF MEMORY, and
SLEEP AND DREAMING,
papers presented at Queen's University,
Kingston, Ontario, February 1969, by
V.A.KRAL
M. D. SUBOSKI
A. MCGH1E
J. INGLIS
R.BROUGHTON
S.G. LAVERTY
J.B. KNOWLES, E. J. BEAUMASTER, A. W. MACLEAN
D. CAMPBELL, J. RAEBURN

Published for the Ontario Mental Health Foundation by University of

Toronto Press
© University of Toronto Press 1973
Toronto and Buffalo
Printed in Canada
Reprinted in 2018
ISBN 0-8020-3299-0
ISBN 978-1-4875-7739-1 (paper)
LC 72-90742
 Contents

Foreword vii

Clarence Meredith Hincks, 1885-1964 ix

Preface xi

PART O"NE

Clarence M Hincks Memorial Lectures, 1969

DR PAUL D. MACLEAN , Hincks Memorial Lecturer, 1969 2

Introduction 4
1 Man's Reptilian and Limbic Inheritance 6
2 Man's Limbic Brain and the Psychoses 23
3 New Trends in Man's Evolution 42
Bibliography 61

PART TWO
Contributors 68

Psychology of Memory
4 The Organic Amnesias/ v .A. KRAL 69
5 Recent Developments in Memory Consolidation
Theory/ M.D. SUBOSKI 81
6 Input Dysfunction in Schizophren ia/ A. MCGHIE 88
7 Similarities in the Side-Effects of ECT and Temporal
Lobectomy in Man/ J . INGLIS 99

Sleep and Dreaming

8 Confusional Sleep Disorders: Interrelationship with Memory
Consolidation and Retrieval in Sleep / R. BROUGHTON 115
 9 Sleep Disorders and Delirium Associated with the
Use of Ethanol/ s .G. LAVERTY 128
10 The Function of Rapid Eye Movement Sleep and of
Dreaming in the Adult/ J.B. KNOWLES, E.J . BEAUMASTER,
A.W. MACLEAN 146
11 Patterns of Sleep in the Newborn/ D. CAMPBELL, J. RAEBURN 156
 Foreword

In 1965 the Ontario Mental Health Foundation established a lectureship,

named in honour of the late Clarence Meredith Hincks, to stimulate profes-
sional workers in the field of mental health. It was the aim of the Foundation
to make it possible for those engaged in training and research to hear, and
more particularly to associate informally with, persons of outstanding pro-
fessional stature from other centres. In furtherance of this purpose the locale
of the lectures was not to be fixed but was to move among the faculties of
medicine of the universities of Ontario.
The Foundation is pleased to present in this volume the second Clarence
Meredith Hincks Memorial Lectures, held at Queen's University in February
of 1969. The Lectures 'The Brain and Behaviour' were delivered by Dr Paul
D. Maclean, Chief, Laboratory of Brain Evolution and Behavior, the National
Institute of Mental Health, Bethesda, Maryland, USA. Papers presented by
various other experts in the mental health field have been included.
A. POSLUNS, SR
Chairman, The Ontario Mental Health Foundation
 Clarence Meredith Hincks BA, MD, DSC, LLD
1885-1964

The development of psychiatry as a major medical speciality and the growth

of public interest in the mental health movement in Canada are clearly associ-
ated with the career of Dr Hincks.
After graduating from the University of Toronto Medical School in 1907
he entered general practice in Toronto and became a part-time medical officer
in Toronto schools. This led to the development of his interest in the mental
hygiene of children and to a voluntary post in the juvenile court. In I 9 I 4 he
joined the newly-organized psychiatric out-patient clinic at the Toronto
General Hospital. Here he came under the influence of Dr C.K. Clarke, at that
time Professor of Psychiatry, Dean of Medicine, and Superintendent of the
Hospital. With his blessing and the assistance of Clifford Beers in New York,
Dr Hincks organized the Canadian National Committee for Mental Hygiene in
1918 (now the Canadian Mental Health Association).
As its first Secretary and later its Medical Director, he planned and carried
out a varied programme of activities which included surveys of mental hos-
pitals, the development of special classes in schools and better institutional
facilities for retarded children, and the psychiatric screening of immigrants
coming to Canada. He was very successful in soliciting financial support from
the great American foundations to assist in the development of psychiatric
and psychological teaching and research in Canadian universities, and in many
other ways he stimulated the establishment of high professional standards in
the mental health field.
In 1930 he accepted the post of Medical Director of the National Com-
mittee in New York and for the next eight years, throughout the depression,
he managed to keep both the Canadian and the American organizations sol-
vent and usefully busy. During this time, he was successful in helping to
launch the Schizophrenia Research Program of the Scottish Rite Masons,
Northern Jurisdiction (I 935), the National Committee on Psychiatric Educa-
tion (1930), and the American Board of Neurology and Psychiatry (1934).
All of these organizations promoted improvements in undergraduate and
graduate instruction in the field in both the United States and Canada.
X Clarence Meredith Hincks

During the second world war, Hincks and his colleagues developed psycho-
logical and psychiatric services in the Canadian armed forces, and for children
evacuated from bombed-out areas of Britain. Later he organized a Mental
Health Consultation Service for Toronto which after his death became the
Hincks Treatment Centre - a small but exemplary comprehensive psychiatric
hospital for children and young people.
Hincks was a public crusader in the interests of the mentally ill. As such,
compassion, empathy, and sensitivity were essential. He had these qualities,
and many more, including courage, conviction, aggressiveness, and an in-
dignant impatience especially for obsolete government policies, professional
irresponsibility, and public complacency. His greatest enemy was 'man's
inhumanity to man.'
 Preface

At the end of the last century Sigmund Freud abandoned his projected con-
struction of a neurophysiologically based model of the mind and turned to
the development of a psychological model, having recognized that the neuro-
physiology of his day was unequal to the task. Nonetheless, he predicted that
a day would come when advances in knowledge of the functioning of the
central nervous system would make possible the formulation of such a model.
Over seventy years later we are still unable to propose a comprehensive
theory, but there are many segments of our knowledge of the nervous system
which can be related to behaviour, and now many investigators work at the
interface between neurophysiology and the behavioural sciences.
When Queen's University was invited to sponsor the second series of
Hincks Memorial Lectures in 1969, the time seemed ripe to review some
aspect of research and theory development in this frontier area between
neurophysiology and behavioural science, preferably one of demonstrable
relevance to the field of mental health.
The choice of a Lecturer was easy. Dr Paul MacLean's investigations and
development of the concept of the limbic system have been seminal in the
field of psychiatry and in related disciplines, and he has combined rigour of
investigation with a willingness to look, if necessary speculatively, at the
broad implications of these findings. We were fortunate to be able to per-
suade him to accept the task of preparing and presenting these lectures, in
spite of his many other commitments.
The three lectures, which constitute the major part of this book, were
delivered, and were the main theme, at a conference of the Faculty of
Medicine on 'Brain and Behaviour' held on February 11-12, 1969.
The conference was attended by representatives from the fields of mental
health, medicine, and the behavioural sciences. To complement the lectures a
number of panel discussions reviewing current work at the interface of neuro-
physiology and behavioural science were convened. Two panels dealt with
work on 'Memory' and 'Sleep and Dreams.' The papers presented as a basis
for these discussions are included in this volume.
xii Preface

It is not possible to acknowledge individually all those whose work in

organization and participation made the conference a success. Two people,
however, worked behind the scenes and must be mentioned here. Mrs
Jacqualine Norman, Administrative Assistant in the Department of Psy-
chiatry, carried the administrative load throughout. Mrs Pamela Jones served
as editorial assistant in the preparation of the manuscripts; without her help
this book would never have come into being. Our thanks to both of them.
T.J. BOAG
D. CAMPBELL
 DR PAUL D. MACLEAN

Hincks Memorial Lecturer, 1969

Dr MacLean is a native of Phelps, New York. He prepared at the Taft School

and received his BA degree from Yale College in 193 5. His original intention
had been to study philosophy, but he turned to medicine as a more direct
means of gaining an understanding of man. After a year of pre-medical work
in Edinburgh he returned to Yale, where he graduated in medicine in 1940.
Following an internship on the medical wards at Johns Hopkins, he continued
his training at Yale. With the outbreak of the second world war he joined the
Army of the United States as a medical officer, serving from 1942-1946.
While in New Zealand as a member of the 39th General Hospital, he collab-
orated with Dr Averill Liebow in showing that the diphtheria bacillus was a
cause of tropical ulcers, a finding that paved the way for prophylaxis and
treatment. He was also in charge of medical and psychiatric wards. After
separation from the Army in 1946, he practised medicine in Seattle and held
a clinical appointment at the new medical school. In 1947 he was awarded a
USPHS Fellowship for study with Dr Stanley Cobb at the Massachusetts
General Hospital. He conducted research on psychomotor epilepsy and pub-
lished his paper on the 'visceral brain,' for which he later introduced (in
1952) the term 'limbic system.' In 1949 he returned to Yale with a joint
appointment in physiology and psychiatry, and in Dr John Fulton's labora-
tory investigated brain mechanisms of emotion. He became Associate Pro-
fessor of Physiology in 1956 and was awarded a National Science Foundation
Senior Postdoctoral Fellowship for study at the Institute of Physiology in
Zurich. In 1957 he joined the Laboratory of Neurophysiology, National In-
stitute of Mental Health, heading a new section on the limbic system. In
addition to investigations on the anatomic and functional organization of the
limbic system, his work dealt extensively with the cerebral representation of
sexual functions.
Dr MacLean has served as an editor of both physiological and neuro-
psychiatric journals. His writings and lectures have often bridged the two
disciplines, as in his chapter on psychosomatics for the Handbook of
Dr Paul D. Maclean 3

Physiology. In 1964, he received the Distinguished Research Award of the

Association for Research in Nervous and Mental Disease. In 1966 he gave the
Thomas William Salmon Lectures under the auspices of the New York
Academy of Medicine and received the Salmon Medal for Distinguished
Service to Psychiatry. He was recently (1972) the recipient of the Karl
Spencer Lashley Award from the American Philosophical Society and the
G. Burroughs Mider Lectureship Award from the NIH .
In 1971, Dr MacLean became Chief of the Laboratory of Brain Evolution
and Behavior, a new facility of the National Institute of Mental Health,
designed for research on animals living under seminatural conditions.
 Introduction*

The air
Breathes invitation ; easy is the walk
To the Lake's margin, where a boat lies moored
Wordsworth, The Excursion

Kingston , the home of Queen's University , enjoys an enviable location at the

northeast end of Lake Ontario. The Indian name Ontario, according to some
authorities, means 'beautiful lake .' Having summered most of my life among
the Islands cooled by the waters of Lake Ontario, and having explored a
number of its tributaries (including the incomparable canal system of the
Rideau) I find it natural to favour this meaning of the name. Nurtured by
such sentiments, I regarded it as a special honour to receive the invitation
from Queen's University on behalf of the Ontario Mental Health Foundation
to give the second series of the Clarence Hincks Memorial Lectures.
In the letter of invitation, Dr Thomas J. Boag explained that the lectures
are 'intended to provide an opportunity for a significant statement on a
theme related to mental health.' In these words one could almost see the
Lady of the Lake holding forth Excalibur for the hand of someone powerful
enough to strike a blow against the nightmare of mental suffering. As would
be true for anyone trying to understand the brain, the trouble in my case was
that the child had not yet become the 'father of the man .' Full of curiosity to
see the sword, but realizing my inability to grasp it, I requested that the
lectures be scheduled in February when the Lake would be frozen over!
Following the lectures I became aware of another lack of sophistication,
and I will comment upon it in explaining the title of this published version.
When I went to Kingston, I did not realize that if you speak publicly in
Canada, there may be a period when the sun will never set on what you say!

* Copyright does not apply to the introduction, which was contributed by an employee
of the US Government
Introduction 5

Through the communication ties of the British Commonwealth, the

underlying theme of my lectures - the three-brain concept - gained rather
wide circulation. In the abbreviated reports it was easy to get the impression
that I conceived of man as though he behaved under the direction of three
autonomous brains. To hedge against this implication I have changed the
original title of my lectures ('The Brain and Behaviour') to read 'A Triune
Concept of the Brain and Behaviour.' In its evolution the human brain ex-
pands along the lines of three prototypes for which I have used the terms
reptilian, paleomammalian, and neomammalian. If the three cerebrotypes are
pictured as intermeshing and functioning together as a triune brain, it makes
it evident that they cannot be completely autonomous but does not deny
their capacity for operating somewhat independently. Moreover, the word
triune has the advantage of implying that the 'whole' is greater than the sum
of its parts, because the exchange of information among the three brain types
means that each derives a greater amount of information than if it were
operating alone.
Some might argue that my underlying theme is simplistic, pointing out, for
example, that one might also speak of prototypical brains in regard to each of
the sensory systems. But what I have wished to emphasize in these lectures
are the three major evolutionary developments, realizing that one cannot say
anything about such a complicated structure as the brain without indulging in
oversimplification.
I wish to acknowledge my indebtedness to the late Wade Hampton
Marshall, first Chief of the Laboratory of Neurophysiology of the National
Institute of Mental Health (I 953-1970) for his unswerving support; and to
the former members of the Section on Limbic Integration and Behavior (now
the Section on Comparative Neurophysiology and Behavior, the Laboratory
of Brain Evolution and Behavior) for their help and collaboration. Finally, in
acknowledging my indebtedness to the Ontario Mental Health Foundation
and Queen's University for the opportunity to present these lectures, I wish
to express my personal appreciation to Professor T.J. Boag and Dean E.H.
Botterell for their gracious letters of invitation and their warm hospitality to
me while I was in Kingston.
PAUL D. MACLEAN
 1 Man's Reptilian and Limbic Inheritance*

The cursed crocodile became to me the object of more horror than all the
rest. I was compelled to live with him; and (as was always the case in my
dreams) for centuries.

Thomas DeQuincey, The Confessions of an English Opium-Eater

The population explosion is an alanning sign that the problems of modem

man are soaring at an unprecedented rate. Calhoun (I 971) points out that
the von Foerster curve for population growth since the time of Christ shows
that each successive doubling of the population required only half the time of
the previous doubling. At the present rate, von Foerster and his colleagues
(1960) calculate that the human population will squeeze itself to death by
the year 2026, a date which they term 'Doomsday.'
In his recent book The Ghost in the Machine, Arthur Koestler, who was
honoured last year by Queen's University, has a final chapter called 'The Age
of Climax' (1967). In it he poignantly reminds us that the same 'brakes-off
situation pertaining to the population explosion also applies to the power of
weapons, the speed of missiles, communications, and scientific infonnation.
He quotes, for example, Morris' figures that whereas there were only 10
scientific journals in 1700, there were l00 in 1800; 1000 in 1850; 10,000 in
1900; l00,000 after the first world war; with an expected total of one million
by the year 2000. Apparently, as many scientists were produced in the last
fifteen years as existed during the entire previous period of science.
I have friends teaching in the humanities who find these figures on the
growth of science almost as terrifying as the prospect of an atomic holocaust!
One must certainly agree with them that we would not be in our present
predicament if it were not for the spiralling achievements of science. But

* Copyright does not apply to this chapter, which was contributed by an employee
of the US Government
Man's Reptilian and Limbic Inheritance 7

man's decisions as to how he will utilize his scientific knowledge and plot his
future course are essentially a matter of politics. This emphasizes the urgency
for a simultaneous effort on the part of all nations to work for world-wide
enlightenment. I am referring now to the enlightenment of self-knowledge,
and not to the much advertised kind of enlightenment of our western bull-
dozer culture. It has long been an abiding faith of psychiatry that self-
knowledge, more than anything else, holds the promise of reducing those
inner tensions of man that otherwise have the potentiality of exploding with
catastrophic consequences. Such was the abiding faith of Clarence Hincks in
whose memory these lectures are given.
It is my own faith, based on the study of the brain, that a wide dis-
semination of available knowledge about basic brain mechanisms and be-
haviour would do much to help man live in greater contentment with
himself and his society. I say this despite the emphasis I give in this first
lecture to psychological difficulties emanating from lopsided differences
between evolutionary old and new systems of man's brain, and for which I
have only a few homely prescriptions. For the real tenet of my faith you will
have to wait for the third and last lecture on 'New Trends in Man's Evolu-
tion .' Perhaps then you will see why I harbour some hope for Calhoun's
optimistic prediction that just about the time of 'Doomsday' there will in fact
be 'Dawnsday' (I 97 I).
In order to appreciate where we are going, we must first look back in
time to see where we have been. Rocketed by our imagination, which
exceeds the speed of light, we will go back 200 million years to the age of
reptiles, when animals which never learned to talk began to work their way
into the brain of man (MacLean 1964a). Man puts so much emphasis on
himself as a unique creature possessing a spoken and written language that,
like the rich man denying his poor relatives, he is loath to acknowledge his
animal ancestry. In the last century it almost killed him to admit his resem-
blance to apes, but t.he time is approaching when he must say 'uncle' and
admit to having far poorer relatives! Intellectually he has been aware of this
for a long time, but emotionally he cannot bring himself to recognize it. It is a
little bit like the 'denial of illness' which is much more familiar to us.
Perhaps the most revealing thing about the study of man's brain is that he
has inherited the structure and organization of three basic types which, for
simplification, I refer to as reptilian, old mammalian, and new mammalian
(MacLean 1962, 1964a, 1967a, 1968a, 1968b, 1969b, forthcoming). It
cannot be over-emphasized that these three basic brains show great dif-
ferences in structure and chemistry. Yet all three must intermesh and func-
tion together as a triune brain. The wonder is that nature was able to hook
them up and establish any kind of communication among them.
 8 The Triune Concept of the Brain and Behaviour

In evolution one might imagine that the brain has developed like a building
to which wings and superstructure have been added. The hierarchy of these
three brains is shown in Figure 1. Man's oldest brain is basically reptilian, 1
forming the matrix of the upper brainstem and comprising much of the reticu-
lar system, midbrain, and basal ganglia.
The reptilian forebrain is characterized by greatly enlarged basal ganglia
which resemble the striatopallidal complex of mammals.2 But, in contrast to
mammals, there is only a rudimentary cortex. The old mammalian (paleo-
mammalian) brain is distinctive because of the marked development of a
primitive cortex which, as will be explained later, corresponds to the limbic
cortex. Finally, there appears late in evolution a more complicated type of
cortex called neocortex, which is the cerebral hallmark of higher mammals
and which culminates in man to become the brain of reading, writing, and
arithmetic.
In the popular language of today, these three brains might be thought of as
biological computers, each with its own peculiar form of subjectivity and its
own intelligence, its own sense of time and space and its own memory,
motor, and other functions (MacLean 1968a, 1968b, I 969b, forthcoming).
On the basis of behavioural observations of ethologists, there are indications
that the reptilian brain programmes stereotyped behaviours according to
instructions based on ancestral learning and ancestral memories. At the new
NIMH field laboratory, The Laboratory of Brain Evolution and Behavior, we·
plan to test the hypothesis that the counterpart of the reptilian brain in
mammals is fundamental for such genetically constituted forms of behaviour
as selecting homesites, establishing territory, engaging in various types of

Here it is worth a reminder that there are only four surviving orders of reptiles, and none
is considered representative of the forerunners of mammals. The four existing orders are:
(1) Chelonia, comprising tortoises and turtles, and so named because of the box-like
shell in which they are encased; (2) Squamata, or scaly reptiles, consisting of two sub-
orders called Lacerti/ia (lizards) and Ophidia (snakes); (3) Crocodilia (crocodiles and
alligators) which share with birds (Aves) an origin from the Archosauria (ruling reptiles);
and, finally, (4) Rhynchocephalia (snout-head) represented by the single species
Sphenodon Punctatum or so-called Tuatara of New Zealand. Some authorities believe
that of the existing reptiles the turtle possesses the type of brain resembling most closely
the one antecedent to that of mammals. Perhaps, however, the behaviour of lizards
would be most similar to that of the mammal-like reptiles
2 Neuroanatomical inferences about the evolutionary homogeneity of the striatopallidal
complex have recently been strengthened by comparative histochemical studies. With the
Koelle stain for cholinesterase, the corpus striatum (caudate + putamen) acquires a vivid
orange-brown color. What are regarded as comparable structures in reptiles and birds
show a similar intense staining (Parent and Olivier 1970). Using the fluorescent tech-
nique of Falck and Hillarp, Juorio and Vogt (1967) have found in avian and lower mam-
malian forms that these same structures glow brightly because of the presence of
dopamine. David Jacobowitz and I have observed a similar fluorescent picture in the
monkey (unpublished). The presence of dopamine in the striatum is attributable to a
projection of dopamine-containing cells in the substantia nigra (Anden et al, 1965)
Man's Reptilian and Limbic Inheritance 9

~~oMAMMAl/
~I\I

Figure I Diagram of hierarchic organization of three basic brain types, which, in the
evolution of the mammalian forebrain, become part of man's inheritance. Each type has
distinctive structural and chemical features. The triune brain, with its extensive inter-
connections, would, in Koestler's (1967) terminology, represent a 'holonarchy.' Man's
counterpart of the paleomammalian brain comprises the so-called limbic system
(Maclean 1952), which has been found to play an important role in emotional
behaviour. (After Maclean 1967a)
IO A Triune Concept of the Brain and Behaviour

display ,3 hunting, horning, mating, breeding, imprinting, forming social hier-

archies, and selecting leaders. The cardinal importance of territory seems to
have been emphasized first by Eliot Howard (1929), the English naturalist,
who observed that the establishment of territory was a necessary preliminary
to mating and breeding.
The reptilian brain seems to be hidebound by precedent. Behaviourally,
this is illustrated by the reptile's tendency to follow roundabout, but proven,
pathways, or operating according to some rigid schedule. Customs of this
kind appear to have some survival value and raise the question as to what
extent the reptilian counterpart of man's brain may determine his obeisance
to precedent in ceremonial rituals, religious convictions, legal actions, and
political persuasions (Maclean 1968a, 1969b ).
In his essay Beyond the Pleasure Principle, Freud (I 922) refers again and
again to man's compulsion to repetition, or as he otherwise calls it, the
repetition-compulsion. Obeisance to precedent is the first step to obsessive-
compulsive behaviour, and this is well illustrated by the sea turtle's returning
to the same place year after year to lay her eggs. It has been shown in recent
studies on mammals that they are also like the reptile in their tendency to
return to home grounds. This has been observed, for example, in the case of
seals, sheep, and goats (Harper 1970).
Man, too, may harbour a continuous yearning to return home after moving
to a distant land. I remember a surgeon's widow in Auckland, New Zealand,
who, during the second world war, was wonderfully hospitable to members of
our 39th General Hospital. A native of England, she had lived in New Zealand
many years and had reared her children there. Once she said to me, 'You
know, I have lived here most of my life, but I have never settled here.' Most
of us have experienced from time to time the compulsion to return to our
childhood home. I remember once smelling hay in Switzerland and feeling
overwhelmed by yearning to return home to our summer place not far from
here. One might go so far as to generalize with Freud (1922, p. 47) that all
repetition-compulsions represent a form of homing. Certainly there seems to
be a recognized tendency among animals, after exploring and reaching out for
food, for a mate, or whatever else, to return to a recognized frame of
reference . Freud saw a biological model for this in the 'circuitous paths' of

As described in a preliminary report (Mac Lean 197 2), I have since found that bilateral
lesions of the striatopallidal complex abolish the innate display behaviour of squirrel
monkeys described in the final lecture. Because of the traditional clinical view that the
striatopallidal complex subserves purely motor functions, it should be emphasized that
large lesions of the complex may result in no apparent motor incapacity if there is no
injury to the internal capsule. These findings, as discussed in the above-mentioned report,
suggest that the striatopallidal complex is fundamentally involved in species-specific and
imitative forms of behaviour
 Man's Reptilian and Limbic Inheritance 11

recapitulation in ontogenetic development, and in line with such thinking was

led to the conclusion that 'the goal of all life is death,' with the animate
returning to the inanimate.
There are several other aspects of reptilian behaviour that invite spec-
ulation about the functions of man's counterpart of the reptilian brain.
Obviously, it will take many lifetimes, if it is ever possible, to pin a num-
ber of these things down. But before leaving this subject, I will mention a
few more examples. One is the capacity for an appropriate dummy, or even
part of a dummy, to precipitate in reptilian forms a sequential acting out
of genetically constituted forms of behaviour. Some of the best illustrations
are provided by birds which, in the branching of the evolutionary tree, are
closely related to reptiles. 4 In Tom turkeys on our farm, for example, I
have many times seen the copulatory act triggered by a mere phantom, and
performed in entirety without a partner. A similar tendency in mammals is
illustrated by the use of partial dummies for collecting semen for artificial
insemination. The mind readily associates to human fetishism. Other
dummy-induced behaviours are aggressive display, fighting, and flight and
following reactions. One wonders how often the caricature of a leader such
as a Hitler is sufficient to deceive people into thinking that they are following
a true leader!
The question of imprinting also arises in connection with human be-
haviour. It is recognized that there are certain critical times in the brain's
development when it is particularly receptive to forming attachments to
things in the environment. Spalding (I 954 ), in the last century, was probably
the first to describe what ethologists call imprinting (Lorenz 1937), when he
reported his observation that, in the absence of a hen, baby chicks would
follow the first moving object to which they were exposed. I am among those
who speculate that in human society the age of adolescence may be a critical
time for imprinting to occur with respect to the same or opposite sex. At this
age boys and girls have many features in common, and there is the possibility
that , in schools of one sex, imprinting at these times may be conducive to a
life of continued homosexuality . This raises the question of the advisability
of providing co-education whenever possible. It is doubtful, however, that
such reasoning lies behind the co-educational trend in our country where, in
some universities, as you know, boys and girls are now sharing the same
dormitories.
Reptilian behaviour raises other intriguing questions. How, for example,
do ancestral memories detennine man's love of hunting and horse racing, his
choice of spouse, and even perhaps his profession? Symbolically, in affairs of

4 See n. l above
 12 A Triune Concept of the Brain and Behaviour

sexual attraction, I am reminded of DeQuincey's relentless search for Ann,

whose face had become idealized and imprinted in his mind. 5
There is also the question as to how reptilian proneness to imitation is
relevant, in human affairs, to mass hysteria, mob violence, and now, thanks to
television, a world-wide adoption of fads and fashions. In summary, the
reptilian brain behaves as though it were neurosis-bound by an ancestral
superego (Macl..ean 1964a), lacking the adequate neural machinery for
learning to cope with new situations.
The evolutionary development, in lower mammals, of a respectable cortex
appears to represent nature's attempt to provide the reptilian brain with a
'thinking cap' and to emancipate it from the ancestral superego (Maclean
1968a). The primitive cortex might be likened to a primitive television screen,
giving the mammal a better picture for adapting to its internal and external
environment. Evidence for its reception of signals from both internal and ex-
ternal sources will be given in the next two lectures. In all mammals, most of
the primitive cortex is found in a large convolution which Broca (I 878) called
the 'great limbic lobe' because it surrounds the brain stem. Limbic means
'forming a border around.' In dealing with the question of function, it
deserves emphasis that this lobe, as illustrated in Figure 2, is found as a
common denominator of the mammalian brain. Its relative stability through-
out mammalian phylogeny contrasts with the mushrooming neocortex which
culminates in man and gives him a large screen on which a picture can be
portrayed by a written and spoken language.
Because of its close relationship to olfactory structures, the old
mammalian brain was formerly believed to subserve purely olfactory func-
tions and was accordingly referred to in many texts as the rhinencephalon
(Schafer 1900, p. 765, n. 1). Papez's famous paper of 1937 struck a mortal
blow to this line of thinking. Since then, extensive investigation has revealed
that in addition to olfactory functions, this brain plays an important role in
elaborating emotional feelings that guide behaviour with respect to the two
basic life principles of self-preservation and the preservation of the species
(Maclean 1958b, 1959). In 1952 I suggested the term limbic system as a
suitable descriptive designation for the limbic cortex and structures of the
brain stem with which it has direct connections.
It should be emphasized that the limbic cortex has similar features in all
mammals and is structurally more simple than the new cortex. From this it can
be inferred that it continues to function at an animalistic level in man as in ani-
mals. Also, in marked contrast to the new cortex, it has strong connections with
the hypothalamus which plays a basic role in integrating emotional expression.

5 Thomas DeQuincey, Confessions of an English Opium Eater

Man's Reptilian and Limbic Inheritance 13

RAIUIT CAT MONKEY

Figure 2 Most of the cortex of the limbic system (old mammalian brain) is contained
in the limbic lobe which surrounds the brain stem. This figure (depicting the relative
sizes of the brains of the rabbit, cat, and monkey) illustrates that the limbic lobe is found
as a common denominator in the brains of all mammals. The rapidly evolving neocortex,
comparable to an expanding numerator, is tinted grey. Upper and lower drawings, res-
pectively, show lateral and medial views of the cerebral hemispheres. (After MacLean,
in Wittkower and Cleghorn (eds. ), Recent Developments in Psychosomatic Medicine,
London: Pitman, 1954)

In proceeding now to consider some elementary functions of the limbic

system, I shall refer to a simplified anatomical diagram, recognizing that it
is impossible to say anything about such a complicated organ as the brain
without being guilty of oversimplification. The diagram in Figure 3 focuses
on three main subdivisions of the limbic system . The three main cortical
subdivisions in the limbic ring are shaded by the small numerals l, 2, and
3, and their principle connecting links with the hypothalamus and other
parts of the brain stem are correspondingly labelled by the large numerals.
You will note that the two upper branches of the medial forebrain bundle
(MFB) meet with the descending fibres from the olfactory bulb and feed
into the lower and upper halves of the ring through the amygdala and the
septum at the points marked no . l and no. 2. In the rest of this lecture we
will focus attention on these two limbic subdivisions, but in anticipation of
the third and final lec,ure, and to point out an important contrast, it
should be emphasized that the third large pathway branching from the
hypothalamus bypasses the olfactory apparatus. The third subdivision con-
nected with this pathway reaches maximum development in man and will
14 A Triune Concept of the Brain and Behaviour

OLF.
BULB

Figure 3 The functions of the limbic system are discussed with respect to three main
subdivisions shown in this diagram. The three main cortical regions in the limbic lobe
are indicated by the small numerals 1, 2, and 3. (The smaller numerals overlie archi-
cortex and the larger, mesocortex [i.e., transitional cortex).) The principle pathways
linking the three cortical regions with the brain stem are correspondingly labelled by the
large numerals. Elementary functions of the two subdivisions (no. 1 and no. 2), closely
related to the olfactory apparatus, are described in Lecture 1. Discussion of the third
subdivision is deferred until Lecture 3. Abbreviations : AT, anterior thalamic nuclei;
HYP, hypothalamus; MFB, medial forebrain bundle; OLF, olfactory. (Adapted from
MacLean 1958a)

be discussed in the concluding lecture on new trends in the evolution of

the brain.
Clinical and experimental findings indicate that the lower part of the lim-
bic ring fed by the amygdala is primarily concerned with emotional feelings
and behaviour that insure self-preservation (MacLean I 958a, 1958b, I 959).
In other words, there is evidence that its circuits are kept busy with the
selfish demands of feeding, fighting, and self-protection. The most convincing
evidence for this is derived from observations on patients with limbic epi-
lepsy. At the beginning of an epileptic discharge in this part of the brain,
patients experience feelings that come into play in the struggle for survival.
These include elementary feelings of hunger, thirst, nausea, suffocation,
Man's Reptilian and Limbic Inheritance 15

choking, racing heart, or the urge to defecate and urinate, which may be con-
joined with a variety of intense emotional feelings such as terror, fear, anger,
sadness, foreboding, strangeness, and paranoid feelings. The automatic be-
haviour that follows these feelings often appears to be an acting out of the
subjective states, as typified by eating, drinking , vomiting, showing anger, or
running and screaming as if afraid.
In stimulating the corresponding region in cats and monkeys, Delgado and
I (1953) elicited similar forms of behaviour. Such findings, therefore, indicate
that this subdivision of the limbic system is primarily concerned with self-
preservation both as it pertains to obtaining the requisites of life and avoiding
the claws of injury and destruction.
The classical studies of Kluver and Bucy revealed that if this part of the
brain was surgically excised in wild monkeys, they (1) lost their sense of fear;
(2) became tame; (3) would eat all manner of objects such as nuts, bolts, and
faeces; and (4) developed bizarre sexual behaviour and other changes that
would be prejudicial to their survival in a natural environment (Kluver and
Bucy 1939).
The changes in sexual behaviour were indicative of a release of other parts
of the brain concerned with sexual functions. This turns our attention next to
the second subdivision of the limbic system connected by the septum (no. 2).
Several years ago we observed that following electrical or chemical stimula-
tion of the septum and related hippocampus, male cats developed enhanced
pleasure and grooming reactions and sometimes penile erection - aspects of
behaviour seen in feline courtship (Maclean 1957a, 1957b ). These observa-
tions suggested that this part of the limbic system was involved in expressive
and feeling states that are conducive to sociability and other preliminaries of
copulation and reproduction. They were of heuristic value because, curiously
enough, there had existed little but indirect evidence from ablation studies
that the forebrain was concerned in sexual behaviour. Penfield, for example,
who had stimulated the greater part of the cerebral cortex in man apparently
never elicited penile erection or erotic sensations (Penfield and Jasper 1954).
The sum total of negative findings seemed paradoxical in view of the highly
organized behaviour required for procreation .
The positive findings in the cat led me to undertake an exploration of the
limbic system and other parts of the brain in the squirrel monkey in search
for sexual responses to electrical stimulation. For these experiments a stereo-
taxic platform with electrode guides was chronically fixed above the scalp on
four screws previously cemented in the skull (1967b ). This device avoids open
surgery and provides a closed system for millimetre by millimetre exploration
of the brain with stimulating and recording electrodes while the monkey sits
in a special chair. Animals become readily adapted to this procedure and are
16 A Triune Concept of the Brain and Behaviour

provided with their favourite forms of fluid and nourishment. After each
experiment the monkey is returned to its home cage.
In the brain stem above the level of the hypothalamus there are two main
locations where stimulation results in penile erection. One is the medial septo-
preoptic region and the other, as will be emphasized in the final lecture, is in
the anterior and midline thalamic region (Maclean and Ploog 1962).
Stimulation at positive loci in the medial septo-preoptic region commonly
results in electrical discharges in the hippocampus, and during these dis-
charges the erection may become throbbing in character and increase in size.
Afterwards, the monkey may become placid and drowsy for prolonged
periods. The general demeanour recalls one of Heath's patients, who, after
stimulation in the septa! region said, 'I have a glowing feeling, I feel good'
(1954). These observations were originally reported at meetings in 1952. Two
years later Olds and Milner (I 954) described their remarkable finding that
rats will repeatedly press a bar for the apparent satisfaction of receiving
electrical stimulation of this part of the brain.
Figure 4 utilizes the brain diagram in Figure 3 for presenting a graphic
summary of the effects of stimulation at points within the two subdivisions
of the limbic system related to the amygdala and septum. The shield of Mars
(0) is used as a symbol for facial, oral, and alimentary responses, and his
sword (1) for genital responses. The symbols for shield are seen to cluster in
the amygdala region, while those for sword are concentrated in the medial
septo-preoptic region . Followed downstream into the brain stem, sword and
shield unite in the anterior hypothalamus, portraying a reconstitution of the
warrior Mars in a region of converging nerve fibres involved in angry and de-
fensive behaviour. Since fighting is frequently a preliminary to mating as well
as feeding, these findings suggest that nature uses the same neural mechanisms
for combat in both situations.
In the dorsal hypothalamic area just above the focal region in the hypo-
thalamus involved in agonistic behaviour, stimulation elicits full erection
usually accompanied by vocalization. Then, as the electrode is lowered a little
further, signs of angry or fearful behaviour begin to appear, as indicated by
the quality of vocalization, struggling, biting, and showing of the fangs
(Maclean, Denniston, and Dua 1963a, p. 280). Afterwards there is charac-
teristically a rebound erection. At the point where the pallidohypothalamic
tract loops over the medial aspect of the fornix, only agonistic signs are
elicited. Finally, as the electrode leaves this focal area, stimulation primarily
evokes biting or chewing.
These findings would seem to throw some light on the neural basis for
aggressive and violent expressions of sexual behaviour. In his three essays on
sex, Freud (1948) noted that 'the sexually exciting influence of some painful
Man's Reptilian and Limbic Inheritance 17

Figure 4 Diagram summarizing effects of electrical stimulation at points within the

two subdivisions of the limbic system interconnected respectively by amygdala and
septum (compare with Figure 3). The shield of Mars (0) is used as a symbol for facial,
oral, and alimentary responses and his sword (J') for genital responses. Symbols for shield
cluster in amygdala region, and those for sword in septa! region. Followed along descend-
ing pathways, sword and shield unite in the anterior hypothalamus, portraying a recon-
stitution of the warrior Mars at a locus where electrical stimulation elicits angry and
defensive behaviour. See text for implications. (Adapted from MacLean 1964a)

effects such as fear, shuddering, and horror is felt by a great many people
throughout life and readily explains why so many seek opportunities to ex-
perience such sensations' (p. 62). He also remarked, ' ... a number of persons
report that they experienced the first signs of excitement in their genitals
during fighting or wrestling with playmates ... The infantile connection
between fighting and sexual excitement acts in many persons as a deter-
minant for the future preferred course of their sexual impulse' (p. 62).
Returning upstream to the amygdala and septum, we find further implica-
tions of the physiological findings. Slow frequency stimulation in parts of the
amygdala elicit movements of the face, chewing, salivation, and swallowing,
followed several seconds later by the appearance of penile erection (Maclean
I 962, p. 295). Such findings help to explain penile erection observed in
18 A Triune Concept of the Brain and Behaviour

animals and human infants when being fed . They also help to illuminate oro-
genital behaviour to which there have been so many allusions in the current
literature and screenplays, such as James Joyce's Ulysses. An archetypal form
of such behaviour is found in the lemur, a primitive primate which has a
greeting display (see Lecture 3, Figure 5) in which the male and female
mutually lick the anogenital region . The close connection of oral and genital
functions is apparently due to the olfactory sense which, dating far back in
evolution, plays a major role in both feeding and mating.
In the neocortex the representation of the body is such that the head and
tail stand at opposite poles like north and south. This is what one would
expect of a structure with the nice discrimination of the neocortex. But in
the limbic lobe head and tail are brought into proximity by the olfactory
sense. Plate 1 shows a cat licking its erect penis following a hippocampal
after-discharge. Civilized man long suspected that the world was round before
Columbus sailed to America, but how could he have imagined that the limbic
lobe was a closed ring and that in voyaging in one direction the head would
be reached by way of the tail and vice versa (Maclean 1968).
I have been told by some psychiatrists that these recent physiological find-
ings have helped to relieve guilt feelings of a number of their patients about
oral-sexual fantasies and related behaviour. But in other respects the physiol-
ogy and anatomy that have been outlined in discussing the three-brain con-
cept only serve to emphasize the special difficulties besetting the patient and
therapist. In therapy the psychiatrist commonly proceeds on the assumption
that since his patient is an articulate being, his psychological processes readily
lend themselves to translation into words. The goals are to clear the way of
resistances and through free association bring out unconscious and repressed
material that will help to give insight and understanding and a relief of
symptoms.
One shortcoming of this approach may be that it fails to take into account
those two ever-present animals which are conscious and wide awake, but
hopelessly inarticulate. What it amounts to is that the neural machinery does
not exist for the reptilian and limbic brains to communicate in verbal terms.
Too often it seems the brain gets verbal insight, only to be more troubled by
failing to see any improvement of the basic disturbance. Hopefully, future
research will suggest other methods that will perhaps be more effective in
communicating with our animal brains.
But as implied in the introduction, the mounting problems of our times
are such that psychiatry may be called upon to deal not only with the sick-
ness of individuals, but also with a generalized world sickness. One present
difficulty seems to be that our neocortex is all out of step with our animal
Man's Reptilian and Limbic Inheritance 19

Plate 1 Following electrically or chemically induced seizure discharges in the hippo-

campus, male cats may develop enhanced pleasure and grooming reactions and penile
erection. This excerpt from a motion picture film shows a cat with penile erection
grooming its genital region following an electrically induced hippocampal afterdischarge.
(From Maclean 1957b)

brains. As opposed to the old cortex, the neocortex receives its information
predominantly from the external environment through signals conducted
from the eyes, ears, and somatic receptors. (Parenthetically, it is of interest
that the energies giving rise to these signals, unlike those for smell and taste,
lend themselves to electronic amplification and broadcasting.) It is evident,
therefore, that the neocortex is externally oriented. Moreover, it seems to
thrive on change, presumably because nature designed it to come up with new
ideas and new solutions. So dramatic and all-engrossing have been the recent
accomplishments of science, that our educational, political, and business
leaders seem to be planning our existence as though we had to satisfy only
our neocortex.
With its imaamation that travels in excess of the speed of light, man's new
brain may be able to keep up with the present accelerated tempo of life
through speedreading, the help of computers, and otMJ conCrivaoces, but his
20 A Triune Concept of the Brain and Behaviour

two animal brains, which forever tag along , must be presumed to move at
their own slow pace. They seem to have their own biological clocks and their
own sequential, ritualistic way of doing things which cannot be hurried
(MacLean 1967a).
Nature , despite all her progressiveness, is also a staunch conservative and is
more tenacious than the curator of a museum in holding on to her antiques.
The reptilian and limbic brains have survived millions of years of evolution,
and it is evident that we can expect no overnight permutation that will
remove them from the brain of man. Indeed, it is questionable whether or not
the human race could survive without limbic emotions because, whatever else
they do, they assure conflict and argument which in tum insure the mixing of
the gene pool of ideas!
Although we are already anticipating public transportation at rocket
speeds, we will still have to move at a horse and buggy pace with our animal
brains. Once this is realized , it is possible that we can learn to live in greater
contentment than at the present time . Perhaps one of the things we need to
do is to spend more time cultivating those simple domestic pleasures for
which the Europeans are famous and which in the last century were so
colourfully prominent in the poetry and paintings of the pre-Raphaelites -
creating pictures and other things with our own hands , making bread, garden-
ing, tending indoor plants and flowers, caring for pets, birdwatching, and
keeping diaries . In symbolically trying to satisfy our hunting instinct, as well
as our delight in newly laid eggs , let it be hoped that society will insist upon
some alternative to the ugly, macadamized sprawl of our shopping centres.
I myself have never recovered from the homesickness of leaving an idyllic
country town at the age of five. It did a lot to relieve this feeling when we
moved to a small farm in Potomac twelve years ago. I discovered that a little
smell of horse manure once a week was more effective than a cocktail for
quieting something deep down inside of me .· In reading about current trends
one gains the impression that we are becoming a nomadic people, moving on
the average of every three years. Under these disruptive conditions it will per-
haps provide a comforting sense of home base for the reptilian brain if
parents teach their children the stars and constellations.
Most urgent at the present time is the need to devise some way of con-
trolling our soaring population and thereby remove pressures that promote
man's reptilian intolerance and reptilian struggle for territory. There is now
an accumulation of evidence with respect to several animal species that ag-
gressiveness increases with increasing density of population (e .g., Calhoun
1962; Ardrey 1966), often leading to mortal combat . There is no reason to
Man's Reptilian and Limbic Inheritance 21

assume that the animal represented in each one of us would not be similarly
affected (MacLean 1967a).
Already the world has become so small that there is almost no place to
retreat except into the far reaches of our own minds. Some have looked to
the 'mind-expanding' drugs as a means of extending the horizons for such
retreat. I will touch on this subject in the next lecture, but only with respect
to mind-distortion, not retreat. More than retreat, the undrugged, unfettered
mind is capable of ascent, and it is the evolutionary direction of this ascent
that I will consider in the last lecture.

SUMMARY

As evidenced by the population explosion, the problems of modern man are

soaring at an unprecedented rate. There is hope that a solution to these prob-
lems can be helped through a better understanding of the brain. In its evolu-
tion, man's brain retains the hierarchical organization of three basic types,
which for purposes of discussion, are referred to in ascending order as rep-
tilian, paleomammalian, and neomammalian. Despite great differences in
structure and chemistry, all three brains must intermesh and function
together as a triune brain.
In popular terminology, the three sub-brains might be regarded as bio-
logical computers, each with its own special form of subjectivity, intelligence,
time measuring, memory, motor, and other functions.
This lecture focuses attention on the reptilian and paleomammalian brains.
There are indications that the reptilian brain 'programs' behaviour according
to instructions based on ancestral memories and ancestral learning, playing a
primary role in what are commonly referred to as instinctual forms of
behaviour.
The paleomammalian brain represents. an inheritance from lower mammals
and corresponds to the so-called limbic system or limbic brain. It is of special
psychiatric interest because of clinical and experimental evidence of its im-
portant role in emotional behaviour. A common denominator in the brains of
all mammals, it stands in a Janus-like position between the reptilian and the
new mammalian brains. The functions of the limbic brain are considered in
reference to three main subdivisions, two of which are closely related to the
olfactory apparatus and are respectively involved in oral and sexual functions
22 A Triune Concept of the Brain and Behaviour

required for self-preservation and the preservation of the species. Psychiatric

implications of the close organization of the two subdivisions concerned with
oral and genital functions are discussed. Finally, it is emphasized that no
matter how fast man may eventually travel with his neomammalian brain, he
will need to acquire self-knowledge that will allow him to accommodate to
the horse and buggy pace of his reptilian and limbic brains.
 2 Man's Limbic Brain and the Psychoses*

For this second lecture I thought it would be timely to take a sidetrip to

explore some volcanic terrain in one of the frontier regions of the brain. The
purpose in going there is to sound out the possibility that rumblings in the
hippocampal portion of the limbic brain may cause certain types of upheaval
in the endogenous and toxic psychoses.
In particular we shall explore the possibility that eruptions in this part of the
brain may give rise to (I) disturbances of emotion and mood;(2) feelings of de-
personalization; (3) distortions of perception; and (4) paranoid symptoms. The
timeliness of this expedition will be apparent when we stop by the way to ex-
amine some of the chemical properties of the limbic brain in the light of specula-
tions about disorders of catechol (e.g., Bunney and Davis I 965; Schildkraut
1965) and serotonin (Gaddum 1953; Woolley and Shaw 1954) metabolism in
the genesis of the psychoses. The timeliness is further brought home by the
alarming, popular use of psychedelic drugs and the realization that the hallu-
cinogenic agents may precipitate psychoses. These drugs, among their other
possible actions, interfere with the metabolism of cerebral amines, and single
doses of LSD have been said to result in recurring psychotic symptoms in
individuals showing no evidence of a pre-psychotic personality.
Our point of departure for this expedition is from ground covered in the pre-
ceding lecture. For the sake of continuity let us briefly retrace some of the steps
taken up to this point. It was emphasized that man and higher mammals have in-
herited essentially three types of brains (see Lecture I, Figure I). The brain of
oldest heritage is basically reptilian. This reptilian brain appears to be genetical-
ly constituted for guiding behaviour on the basis of ancestral learning and
ancestral memories. In carrying out stereotyped instinctual functions it is
neurosis-bound, as it were, by an ancestral superego.
* Copyright does not apply to this chapter, which was contributed by an employee of the
US Government.
An earlier version of this lecture which was given June 14, 1968 on the occasion of
the opening of the Maryland State Psychiatric Research Center, Spring Grove Hospital,
Catonsville, Maryland, has since been published in P. Black (ed.), Physiological Correlates
of Emotion, New York: Academic Press, 1970, pp. 129-46
24 A Triune Concept of the Brain and Behaviour

h2
\

fO.
/

Plate 1 Two examples of distinctive chemical properties of the archicortex (hippo-

campus). Photomicrograph on left shows complete loss of neurons in area CA3 of mouse
following intraperitoneal injection of 3-acetylpyridine, an antimetabolite of nicotina-
mide. (From Coggeshall and MacLean 1958)
Picture of guinea pig hippocampus on right reveals that the mossy fibres innervating
this same area (McLardy 1955; von Euler 1962) are stained by dithizone, a chelating
agent which gives a red colour when combined with zinc. (From Fleischhauer and
Horstmann 1957)

Superimposed on the reptilian brain is the primitive cortex of the old

mammalian brain that presumably represents nature's attempt to provide a
'thinking cap' for the reptilian brain and emancipate it from the ancestral
superego. The old mammalian brain, which is found as a common de-
nominator in the cerebrum of all mammals (see Lecture 1, Figure 2), is other-
wise known as the limbic system or limbic brain. Appearing late in evolution
is a new type of brain which reaches its greatest development in man to
become the brain of reading, writing, and arithmetic. In the present ex-
pedition we will find additional evidence for believing that intercommuni-
cation among these three brain types must present special difficulties because
of their differences in chemistry and functional anatomy.
As a framework for discussing elementary limbic functions I showed a
simplified diagram of the anatomy of the limbic brain (Lecture 1, Figure 2)
depicting three main subdivisions. I summarized evidence indicating that
the subdivision in the lower part of the limbic ring connected with the
amygdala is largely concerned with emotional feelings and behaviour that
assure self-preservation, whereas, in contrast, the subdivision related to the
septum is implicated in feeling states that are conducive to sociability, pro-
creation, and the preservation of the species. In the final lecture I shall
consider functions of the subdivision connected by the third pathway that
undergoes a great expansion in man and becomes functionally associated
with the frontal lobes.
Limbic Brain and Psychoses 25

Plate 2 Radioautogram on right illustrates high uptake of S 35 labelled L-methionine

in hippocampus (white arrow) of rat. Comparable section on left shows failure of uptake
in a rat that had received a convulsive dose of insulin. (From Flanigan, Gabrieli, and
Maclean 1957)

That the limbic brain is a functionally, as well as an anatomically, inte-

grated system is dramatically demonstrated by mapping the propagation of
hippocampal seizure discharges. The hippocampus, perhaps, has the lowest
seizure threshold of any structure in the brain (Green 1964). If a seizure
discharge is induced in the hippocampus by electrical stimulation it has a
tendency to spread throughout and be confined to the limbic system (e .g.,
Maclean 1957a). The impulses of the discharging neurons might be imagined
as stampeding bulls which do not jump the fence and leave the corral of the
limbic system (Maclean 1958a). Nothing drives home so convincingly the
functional dichotomy - or schizophysiology as I have called it (1954,
1958a) - of the limbic and neocortical systems. As regards this dichotomy of
function it is significant, as I will point out , that patients with smouldering
limbic epilepsy may manifest the various symptoms of schizophrenia .
Before moving on into the psychotic area, however, we will interrupt our
metaphorical expedition to look at a number of chemical differences, which
like anatomical and physiological distinctions already mentioned, serve to
distinguish the limbic brain from the reptilian and new mammalian brains.
26 A Triune Concept of the Brain and Behaviour

I
(
\ 78
\ A

9
D )19
,r)
--- E
I "
0 '-

Figure 1 Certain cortical and subcortical structures of limbic system contain relatively
large amounts of 5-hydroxytryptamine (serotonin). B to I give values in micromilligrams
for respective areas of superficial limbic cortex of dog. The high value of 408 was found
in the periamygdaloid cortex (G). Other abbreviations : A, olfactory bulb ; P and Q,
sensory-motor cortex; R, auditory cortex; S, visual cortex. (From Paasonen, Maclean,
and Giarman 1957)

Many years ago the Vogts (I 953) in documenting their topistic theory,
pointed out several conditions suggestive of a distinctive chemistry of dif-
ferent parts of the hippocampus. In the past fifteen years, several additional
studies have served to emphasize the distinctive chemical properties of the
hippocampal formation . As illustrated in Plate 1, dithizone , a chelating agent
for zinc, strongly stains the mossy fibre system (von Euler 1962; McLardy
Limbic Brain and Psychoses 27

Plate 3 The pronounced fluorescence in the region labelled A indicates the presence
of noradrenalin in the radiate layer of the hippocampus. C identifies the pyramidal layer
and B the stratum oriens. (From Fuxe 1965a)

1962), giving a deep red colour to the part of the hippocampus corresponding
to areas cA4 and CA3 (Maske 1955; Fleishhauer and Horstmann 1957). In the
mouse (Plate l) 3-acetylpyridine, an antimetabolite of the antipellagra vita-
min niacin, selectively destroys the neurons of this same egion (Coggeshall
and Maclean 1958). These areas are also among those parts of the brain that
are damaged by methoxypyridoxine (Purpura and Gonzalez-Monteagudo
1960), an antimetabolite of vitamin B6, which induces seizures possibly
28 A Triune Concept of the Brain and Behaviour

BEFORE RESERPINE
L. POST HIPPOCAMPUS

7 HOURS

12 HOURS

21 HOURS

27 HOURS

33 HOURS

70 HOURS

Figure 2 Samples showing development of changes in bioelectrical activity of pos-

terior hippocampus of cat following single dose of reserpine (lmg/kg). Records are
characterized by persisting theta activity, but note also the slower potentials occurring
at 27 and 33 hours. Horizontal scale, 1 sec; vertical scale, l00µV. (From Kim and
Maclean, unpublished observations 1955)

because of interference with glutamic decarboxylase activity and the resulting

reduction in the neural inhibitor, r-aminobutyric acid (Glaser and Pincus
1969; Tower 1958).
Radioautographic studies (Flanigan, Gabrieli, and MacLean 1957) have
revealed an unusually high uptake of methionine in the hippocampal forma-
tion and other limbic areas (Plate 2), suggesting that the limbic cortex has a
greater turnover of protein than the neocortex. Radioautographic techniques
have also indicated that testosterone has an affinity for several limbic struc-
tures, including the hippocampus (Altman and Das 1965; Pfaff 1968).
Limbic Brain and Psychoses 29

HOYr"7(CH1CH1NH1
(1)
v~N)
H

OH

HN
H0~6HCHaNHR
( 2)

(3) +
McaNCHaCHaOCOCHa
Figure 3 Three classes of drugs noted for their hallucinogenic action are ( 1) sub-
stituted indole alkylamines, and (2) phenylalkylamines and those affecting the acetyl-
choline (no. 3) mechanisms. (Adapted from Downing 1964)

There next crops up the interesting matter of the biogenic amines . For-
merly, when pooled samples were used, it was believed that the cerebral
cortex and other parts of the forebrain contained negligible amounts of
5-hydroxytryptamine (serotonin). When we assayed specific areas, however,
relatively large amounts of this amine were found in the amygdala and
overlying pyriform cortex, the septum, and cellular parts of the hippocampus
(Paasonen, Maclean, and Giarman 1957). The amount in the amygdala region
was comparable to that of the hypothalamus. Figure l shows the values
found for the superficial limbic cortex and the neocortex. As regards cate-
cholamines, radioautographic (Csillik and Erulkar 1964; Reivich and
Glowinski 1967) and fluorescent (Fuxe 1965a) studies have detected a
notable concentration of noradrenalin in the radiate layer of the hippo-
campus (see Plate 3). These findings are of special interest in view of evidence
that tranquilizing and psychotropic drugs known to affect the metabolism of
serotonin and catechols induce distinctive electroencephalographic changes in
the hippocampus. Figure 2 illustrates the sequence of changes occurring in
the cat following the administration of reserpine .
Using the Koelle stain for acetylcholinesterase, Lewis and Shute (I 967)
have provided an assortment of evidence that in the rat the main pathway to
 30 A Triune Concept of the Brain and Behaviour

the hippocampus from the septum is 'cholinergic' in type. In the coypu rat
Girgis (1967) has found a high content of cholinesterase material in both the
hippocampus and amygdala. The combined presence of cholinergic and
adrenergic systems in the limbic brain suggests a certain parallel with the
peripheral autonomic nervous system .
The above considerations recall that the three major classes of psychotomi-
metic drugs (Figure 3) are (I) indole alkylamines; (2) phenylalkylamines; and
(3) agents interfering with cholinergic mechanisms (Cohen 1967). In LSD, DMT,
bufotenine, psilocybin, and harmine, one sees the indole nucleus of serotonin.
In mescaline, one sees the phenyl nucleus of catechols such as noradrenalin.
Denckla has made the interesting observation that for every tranquilizer there
is a compound with a similar structure that has a hallucinogenic action. 1
Recently Dewhurst (1968) has criticized the hypothesis (e.g., Bunney and
Davis 1965 ; Schildkraut 1965) that the manic phase of manic-depressive psy-
choses is attributable to an excess of cerebral catecholamines and the depressive
phase to a deficiency. He emphasizes that catecholamines invariably exert a de-
pressant action if administered so as to circumvent the blood brain barrier. He
offers a substitute hypothesis that the fault lies in tryptamine metabolism,
pointing out among other evidence, that the antidepressant monamine oxidase
inhibitors primarily induce an increase in intracerebral tryptamine and that this
amine has an excitatory effect on the brain. Ernst, van Andel, and Charbon
(I 961 ), however, have reported that the in tracistemal administration of trypta-
mine produces catatonia in cats and that this effect can be antagonized by 5-
hydroxytryptamine which never produces catatonia. Tryptemine readily
crosses the blood brain barrier. Denckla and I found that when given intraperi-
toneally to squirrel monkeys, this amine rapidly produces an apparent state of
drowsiness and catatonia which lasts for about thirty minutes ( unpublished ob-
servations 1969). 5-Hydroxytryptophan exerts a similar but more prolonged ac-
tion, with the effects persisting for five to six hours (Gelhard,Perez-Cruet, and
Gessa 1971 ). A highly purified form of tryptamine is now available , and we shall
attempt to learn whether its intracerebral administration produces catatonia or
excitement. 2

l W.D. Denckla, personal communication, 1968

2 It was found, as might have been expected, that the intracerebral administration of
purified tryptamine in the squirrel monkey resulted in the same kind of symptoms (in-
cluding pupillary dilatation) seen with intraperitoneal injections. Intracerebral nora-
drenalin, in doses ranging from 50-S00µg, failed to elicit signs of excitation; the larger
doses appeared to produce a state of quietude.
Saavedra and Axelrod have reported in a current issue of Science (1972) that
tryptamine occurs normally in the rat brain, and that an enzyme that converts it to
dimethyltryptamine (a potent psychedelic compound) is present in the rat and
human brain
Limbic Brain and Psychoses 31

Van Andel and Ernst (1961) have observed that catatonia caused by
tryptarnine can be prevented by eserine. They suggest tryptamine acts as a
competitive inhibitor of 5-hydroxytryptarnine at certain receptor sites, elim-
inating a function which 'tends to strengthen certain acetylcholine activities
in the CNS.'
On the basis of my own experience with chemical stimulation of the brain,
I would not be inclined to favour the hypothesis that noradrenalin acts as an
excitant. In experiments on cats in which I deposited noradrenalin (in solid
form) in the midbrain reticular formation and other structures, I never ob-
served any notable electroencephalographic or behavioural changes (Maclean
1957b, unpublished observations 195 7). On the contrary, the deposit of
cholinergic agents in the third ventricle or surrounding gray matter close to
the aqueduct provoked a state of profound excitement and angry 'hallu-
cinatory' behaviour (Maclean 1957b ). It is of interest that Fuxe (1965b)
finds a high density of catecholamine terminals in the dorsomedial hypo-
thalamus and medial preoptic region, structures in which electrical stimula-
tion induces such parasympathetic effects as penile erection and cardiac
slowing. Is it possible that noradrenalin in these structures may play a role in
the regulation of cholinergic mechanisms?
The symptoms of patients with psychomotor epilepsy provide the most
convincing evidence that the limbic cortex is implicated in the generation of
emotional states, as well as symptoms of a psychotic nature. Structures in the
lower part of the ring are particularly susceptible to ischemia at the time of
birth, as well as to infection and head injury. The herpes simplex virus, for
example, has a predilection for the hippocampus and other limbic areas
(Drachman and Adams 1962; Glaser, Solitare, and Manuelidis 1964),
providing another kind of evidence that the limbic system has a distinctive
chemistry. Irritative lesions in or near the limbic cortex in this part of the ring
give rise to epileptic discharges accompanied by emotional feelings that under
ordinary conditions are important for survival. As mentioned in the preceding
lecture, these feelings include terror, fear, foreboding, familiarity, strangeness,
unreality, sadness, and feelings of a paranoid nature.
Discharges in or near the basal limbic cortex may also result in feelings of
depersonalization or what Hughlings Jackson (Jackson and Stewart 1899)
called mental diplopia, in which one feels as if one is viewing oneself and
what is going on from a distance, a symptom so common in individuals taking
psychedelic drugs. To paraphrase Penfield (I 952), the patient feels as if he
were in the act of a familiar play in which he himself is both actor and
audience .
There may also be distortions of perception of the type reported by pa-
tients with endogenous or toxic psychoses. Knowledge about the correlation
32 A Triune Concept of the Brain and Behaviour

of the site of the lesion and the symptomatology has been gained particularly
from the operative studies of Penfield and Jasper (l 954). They have observed
that electrical stimulation in the involved region may elicit the same kind of
symptoms as those occurring in a spontaneous seizure. Objects may appear
large or small, near or far; sounds may seem loud or faint; one's tongue, lips,
and extremities may seem swollen to large proportions. Time may appear to
speed up or slow down. Persons intoxicated by psychedelic drugs or suffering
from 'return trips' after LSD may experience such disorders of perception.
The interseizure symptomatology of some patients with limbic epilepsy
may be indistinguishable from that of paranoid schizophrenia. I remember
one patient, for example, who was continually obsessed by the feeling that
God was punishing her for overeating. While we were recording her electro-
encephalogram during the expression of these thoughts, random spiking was
prominent in the lead from the tympanic electrode underneath her left
temporal lobe.
Malamud (I 966) has emphasized the high incidence of medial temporal
sclerosis in cases of psychomotor epilepsy, and has stated that sclerosis of the
hippocampus is a common denominator of this condition. On the basis of
clinical and experimental findings, it is probable that in limbic epilepsy the
hippocampus is nearly always involved in the seizure discharge, with the
discharge either originating within the hippocampus or spreading to it
secondarily from related structures.
Before taking up some experimental observations I want to emphasize that
I am not implying that schizophrenia or other types of psychosis represent a
form of epilepsy. Rather, I am placing emphasis on the study of limbic
epilepsy as a means of learning what parts of the brain may be responsible for
some of the symptomatology seen in the psychoses. No other clinical entity
promises to shed more light on mechanisms underlying psychic functions.
It is obvious that in dealing with such questions as depersonalization, one
must depend solely on subjective reports of patients, but there are some
psychotic manifestations concerning which animal experimentation can be
helpful in clarifying the underlying neural mechanisms.
First, let us consider the question of disturbances of emotion and mood.
Sometimes in the wake of after-discharges induced by stimulation of the
amygdalo-hippocampal region, a cat may be in an agitated state for several
minutes, meowing, running around the room, and trying to climb the walls
(Maclean I 959, p. 47). In contrast to this agitated behaviour, after-discharges
induced by stimulating a few millimetres further caudally in the hippocampus
may be followed by grooming, pleasure, and sexual reactions that persist for
several minutes (Maclean I 9 57b ). After stimulations of limbic nuclei that
induce penile erection and hippocampal after-discharges, aggressive monkeys
Plate 4 Medial view of squirrel monkey's brain showing approximate location of limbic cortical areas in which single nerve cells were
activated by visual stimuli-namely, the posterior hippocampal gyrus (H), and the parahippocampal portion of the lingual gyrus (L), and the
retrosplenial cortex (R). The responsive area in the fusiform gyrus (F) falls outside the limbic cortex. The curved black lines schematize the
circuitous path of part of the optic radiations traced anatomically to these areas. This component of the radiations would correspond
to part of Meyer's loop in man. Arrow points to caudal extremity of rhinal fissure. (From MacLean and Creswell, 1970)
34 A Triune Concept of the Brain and Behaviour

may become placid and tame, and these apparent changes in mood sometimes
seem to linger for several hours (Maclean and Ploog 1962). The septum is
one of the main sources of afferent connections to the hippocampus (Daitz
and Powell 1954; McLardy 1955). As mentioned in the preceding lecture,
Heath and his group (I 954) reported that patients experienced pleasurable
feelings and persisting changes in mood following stimulation through
electrodes presumably in the region of the septum.
We turn next to the alterations of perception occurring in the psychoses.
As classical anatomy provides no evidence of inputs to the limbic cortex
from the auditory, somatic, and visual systems, it has always been puzzling
that, with epileptic discharges arising in or near the limbic cortex of the
insula and hippocampal formation, patients may experience alterations of
perception and hallucinations involving any one of these sensory systems
(Penfield and Jasper 1954). Malamud (1966) recently reported a case of a
26-year-old man with a small ganglioma in the uncus-amygdaloid-
hippocampal region, who, during the epileptic aura variously experienced
gustatory, olfactory, auditory, visual, and somatic illusions or hallucina-
tions without loss of consciousness. As not uncommon, this patient sub-
sequently developed mental symptoms diagnostic of a schizophrenic
reaction .
In the 'visceral brain' paper of 1949, I showed a diagram depicting the
convergence of all the sensory systems in the hippocampal gyrus which lies
next to and projects to the hippocampus. It was then known that the
olfactory system had indirect connections with the hippocampal formation,
but there was no experimental evidence of a representation of the other
senses. It has since been shown that the septum relays visceral information
from the hypothalamus to the hippocampus (Green and Adey 1956).
Because of the fundamental role of vision in man and higher primates, I
have been particularly interested in the question of a visual input to the
limbic cortex {MacLean 1966 ). In order to be certain about the location of
neural responses, we have used microelectrodes for recording the activity of
single nerve cells. The stereotaxic technique described in the preceding
lecture provides a closed system for intracerebral exploration with micro-
electrodes in chronically prepared, awake, sitting monkeys . The letters H, L,
and R in Plate 4 overlie limbic areas in which single nerve cells were ac-
tivated by photic stimulation (MacLean, Yokota, and Kinnard 1968). A large
percentage of responsive cells in the posterior hippocampal gyrus (H) were
distinctive from those in all other areas because they gave a sustained on-
response during ocular illumination. The responses of three such cells are
illustrated in Figure 4, and their cortical location is shown in Plate 5. It is
possible that 'tonic' units of this kind may signal changes in background
Limbic Brain and Psychoses 35

A
I II I hi i1) ijl 1~,~1lm~~Uil tUll l~llill~lf lillll~UiUil~I 11 11~11 Il Ifi rt

J100,.v

11 t ~,
'----'
200m sec

B
__I.. . ,,--.._ ___iw IIOOpV
.•···1·1·
tt i 1111!,'
.7,Mf .·, 1111 ' I ,.,• ii Iii i .~;I~~;~ thi·~

. , •.,.,...,. ,.~
200msec

!100,.v
U11 t t ~ llH lu a •
____r---_

Figure 4 Sustained on-responses to ocular illumination recorded from cells in pos-

terior hippocampal gyrus (see Hin preceding Plate 4) of three squirrel monkeys. Such
'tonic' units were not found in other limbic areas or in the visual cortex. Photocell res-
ponse in accompanying records show duration of stimulation, being 2.5 sec in A and
0.4 sec in B and C. (From MacLean, Yokota, and Kinnard 1968)

illumination and thereby play some role in wakefulness, alerting, and/or light-
dependent neuroendocrine changes.
How do visual impulses reach this cortex? Using modification of the Nauta
stain for showing fine degenerating cortical fibres, Creswell and I (1970) have
examined the brains of twenty-four squirrel monkeys with lesions in the
various parts of the geniculopulvinar complex. As illustrated in Plate 6 it was
found that a lesion in the ventrolateral part of the lateral geniculate body (the
main nucleus for transmitting visual impulses) results in a continuous band of
degeneration into the core of the hippocampal gyrus (Plate 5), and that some
fibres enter the cortex here and in neighbouring areas. The inferior part of
this band corresponds to Meyer's temporal loop in man . It has always been a
mystery why this part of the optic radiations takes this temporal detour, but
it now appears, on the basis of the anatomic and microelectrode findings, that
 !
I

Plate 5 Arrows point to the loci of the tips of the microelectrodes recording the unit responses shown in previous figure. The
recording sites appear to have been within or just undernearth the granular layer. In the case illustrated on the far left the electrode
was fixed in place; in the two other cases a small electrolytic mark was made. (From Maclean, Yokota, and Kinnard 1968)
 -
; t:
·, "~,
~~<-
ii..~ ......~ -

· , ~ ""\ . :
..
.
~,
,~

:,
..... 4

'
"

"
Plate 6 Photomicrograph of lesion in ventrolateral part of lateral geniculate body (large arrow) with a continuous band of
degeneration extending into the core of the posterior hippocampal gyrus (small arrow). Some degenerating fibres can be traced into
the posterior hippocampal gyrus and adjacent areas labelled in Plate 4. Section on right shows corresponding section on unoperated
side where only normal fibres are stained. Abbreviations: C, nucleus caudatus; CGL, corpus geniculatum laterale; GH, gyrus
hippocampi, H, hippocampus. (From MacLean and Creswell 1970)
38 A Triune Concept of the Brain and Behaviour

it travels the roundabout course diagrammed in Plate 4 in order to distribute

fibres to the posterior limbic cortex. The inferior pulvinar, which is regarded
as a visual 'association' nucleus, also contributes fibres to this same cortical
region via a band lying just lateral to the optic radiations.
From the standpoint of paranoid symptoms it is possibly significant that
some cells in the limbic retrosplenial cortex (R in Plate 4) are photically
activated only by the contralateral eye, suggesting that the impulses originate
in the primitive temporal monocular crescent. As is well known from personal
experience, objects entering this part of our peripheral field commonly cause
emotional startle and alarm. I remember a young man who at the beginning
of a limbic seizure had the feeling of fear that someone was standing behind
him. If he turned to see who it was, the feeling of fear became intensified.
During interseizure periods patients may have persistent paranoid feelings.
The parahippocampal cortex transmits impulses to the hippocampus which
has large projections to the hypothalamus, anteromedial thalamus, and other
structures of the brain stem involved in emotional, endocrine, and somato-
visceral functions . As will be further discussed in the next lecture, through
such connections there is one possible mechanism by which 'the brain trans-
forms the cold light with which we see into the warm light with which we
feel' (Maclean forthcoming). Such connections, as will also be noted, may be
involved in dreaming. Schizophrenia has often been likened to a waking
dream.
Finally, there is the question with respect to limbic inputs from gustatory,
somatic, and auditory systems. In the search for an answer we have recently
explored the limbic cortex of the insula overlying the claustrum. We have
found in awake, sitting squirrel monkeys that gustatory, auditory, and soma-
tic stimulation evokes responses of units in respective parts of the claustral
insula (Reeves, Sudakov, and Maclean 1968; Sudakov, Maclean, Reeves, and
Marino 1971). No unit has responded to more than one modality. Figure 5
schematizes the distribution of auditory and somatic units. There appear to
be two main types of auditory units, one of which responds with latencies as
short as IO msec. The somatic units have been found to be activated by
pressure alone or by pressure and light touch. The receptive fields are usually
large and bilateral.
There is neuronographic evidence that the claustral insula 'fires' into the
hippocampal formation (Pribram and Maclean 1953). Accordingly, there is a
potential corticofugal pathway by which auditory and somatic impulses could
interplay in the archicortex with those of internal origin, and influence
vegetative and emotional functions of the hypothalamus.
Heinrich Kluver (1951) has emphasized that the visual cortex surpasses all
other sensory structures in guaranteeing the constancy of the external
 Limbic Brain and Psychoses 39

,,,,
, . •
,,,
,,
,,,,
/
----··'
I
,,
,,,,
,,,, i
•
/
:
I
•
••
•
•
=
••
*
= *f
I
I
I
I
I • • •
I
=
I
I
I
•
I
\ =
\
',. ,,
• * VISUAL
e AUDllORY
■ SOMATIC

Figure 5 Diagram of the lateral view of the insular cortex overlying the claustrum
(bounded by broken line), showing loci of units responding to auditory, somatic, and
visual stimulation. This cortex is by definition limbic because it comprises part of the
phylogenetically old cortex bordering the brain stem. The non-claustral part of insula
lies caudally to the right of broken line

environment. It would therefore seem reasonable to look elsewhere than

the primary visual cortex for the neural disturbance underlying visual
hallucinations and delusions. The same reasoning would apply to the other
primary sensory areas of the neocortex. On the basis of the experimental
material we have reviewed, it is possible that distortions of perception
could arise from dysfunction within the limbic structures themselves, or
they might also result from the effects of limbic perturbations on either
the primary sensory areas or the so-called association areas. Such pertur-
bations might be transmitted through cortical association fibres (Pribram and
40 A Triune Concept of the Brain and Behaviour

Maclean 1953) or the diffuse thalamic projection system (Parmeggiani

1967).
A smouldering volcano does not amount to much in the daily lives of a
people living under its shadow. It is only when it erupts that it throws them
into panic. Similarly, a structure such as the hippocampus may have no
disruptive effect on the life of an individual unless it is thrown into perturba-
tion by injury, toxic agents , or stressful reactions. Because of its peculiar
blood supply and its location in the cranium, the hippocampus has long been
recognized to be especially vulnerable to mechanical injury, vascular insuffi-
ciency, and infection . As was particularly emphasized, it has its own peculiar
chemistry and because of its low seizure threshold probably ranks as the most
unstable of structures in the brain . Experimentally, it is striking to observe
the tendency of the hippocampal formation to show prolonged spiking activity
after intense afferent bombardment by electrical stimulation (MacLean and
Ploog 1962). This abnormality is presumably due to the peculiarity of its chem-
istry and synapses. It is conceivable that in constitutionally disposed individ-
uals, day to day bombardment of this unstable 'primitive' cortex by impulses
initiated by stressful situations could lead to a functional disturbance resulting
in persistent paranoid or other abnormal feeling states, with attendant
delusional thinking.
This brings us to the end of our side-trip to explore the volcanic terrain of the
hippocampus. I have presented several lines of evidence suggesting that rum-
blings in this part of the brain may give rise to (I) disturbances of emotion and
mood; (2) feelings of depersonalization; (3) distortions of perception; and (4)
paranoid symptoms - all of which are manifestations that may crop up in endo-
genous and toxic psychoses. Admittedly, I have been able only to scratch the
surface. To probe further and really find out what is going on underneath in the
brain presents a challenge far surpassing that of landing men on the moon. This
obviously requires no explanation because the constellation of neurons which
each one of us carries in our cranium represents the most complicated
mechanism of the known universe.

SUMMARY

In this lecture we take a side excursion to consider how disorders of function

in the hippocampal portion of the limbic brain may give rise to symptoms oc-
curring in the endogenous and toxic psychoses, including (1) disturbances of
emotion and mood; (2) feelings of depersonalization ; (3) distortions of
perception; and (4) paranoid symptoms.
Limbic Brain and Psychoses 41

The timeliness of this topic is emphasized by the alarming, popular use of

psychedelic drugs which may induce the types of symptoms mentioned above
and in some cases precipitate an enduring psychosis. These drugs, among their
other possible actions, interfere with the metabolism of cerebral amines.
There is current speculation that disorders of catechol and serotonin metab-
olism may be involved in the genesis of psychoses. In recent years it has
been shown that parts of the limbic brain contain relatively large amounts of
serotonin and noradrenalin, and that the electrical activity of certain limbic
structures is differentially altered by some psychotomimetic and tranquilizing
drugs. Other biochemical findings are described which serve to distinguish the
limbic from the reptilian and neomammalian brains.
Electrophysiological studies have shown a functional dichotomy ('schizo.
physiology') of limbic and neocortical systems. In this respect it is significant
that patients with smouldering limbic epilepsy may manifest the various
symptoms of paranoid schizophrenia. This does not imply that schizophrenia
is a form of epilepsy. Case histories are cited to illustrate that the symptoms
listed above may result from spontaneous or electrically induced epileptic
discharges in the limbic brain. In conclusion, some relevant experimen-
tal findings in animals are described, along with a discussion of their
implications.
 3 New Trends in Man's Evolution*

In this final lecture I want to present for your consideration what appear to be
some new trends in the evolution of man's brain. As will be explained, this
will require an examination of the connection between sexual and visual
functions.
Twelve years ago when I joined the National Institute of Mental Health
and opened a new laboratory, I had two primary goals. The first was to look
for a representation of sexual functions above the primitive level of the
hypothalamus. Curiously enough, there existed little but indirect evidence
from ablation studies that the forebrain was concerned with sexual behaviour.
Although the brain had been extensively explored by electrical stimulation,
there were only isolated instances of sexual responses. This seemed para-
doxical in view of the highly organized behaviour required for procreation
and the preservation of the species.
My second goal was to investigate the possibility of connections between
the visual system and the limbic cortex. An answer to this question was
crucial for evaluating the importance of the limbic brain in emotional be-
haviour of higher primates and man, in which vision has become the dom-
inant guiding sense. Looked at in another way the question was this: how
does the brain transform the cold light with which we see into the warm light
with which we feel?
Primates, above all other animals, have developed a social sense which in
man becomes conspicuous for its altruistic manifestations. As evidence that a
charitable social sense is still in evolution we need only recall that the word
altruism was coined as lately as 1853 by the philosopher Comte (1955), and
that the word empathy was introduced into our language by Lipps (1914-20)
about 1900.
Altruism depends not only on feeling one's way into another person in the
sense of empathy. It also involves the capacity to see with feeling into

* Copyright does not apply to this chapter, which was contributed by an employee of the
US Government
New Trends in Man's Evolution 43

another person's situation. To accomplish this with vision - our coldest, most
objective, and analytic of senses - nature has had to accomplish a neuro-
logical tour deforce.
A further consideration of altruism shows that the two questions of sexual
and visual representation are closely related. In its highest expression altruism
requires not only insight, but also foresight, in planning for the welfare and
the preservation of the species. Psychologically, this means that libido in its
early Freudian sense must be translated into unselfish concern in the sense of
such humanists as Clarence Hincks. He is said to have acquired from his
parents both a concern for people in trouble and the habit of service. In the
evolutionary process a concern for the welfare and the preservation of the
species is based on sexuality. It is a concern that leads to courtship and the
rearing of a family. It is a concern that fosters education and the building of
schools, libraries, and museums. It is a concern that promotes 'cultural
grooming' in the form of art, music, and architecture . It is a concern that
inspires medical research to prevent suffering and dying of patients who have
not yet become sick or old. It is a concern that has led man to think in terms
of rockets, travel in outer space, and the possibility of immortal life with the
colonization of other worlds (Maclean 1962).
It is evident from mammalian evolution that it requires but a small brain
to assure self-survival and survival of a species at a primitive level. The op-
posum has fared remarkably well for over 130 million years with little more
than its reptilian and limbic brains, and largely guided by its olfactory sense.
In the first lecture I pointed out how the olfactory sense tends to keep the
animal wrapped up in itself by bringing neural mechanisms for oral and
genital functions into close association. Moreover 1 in on-going behaviour the
olfactory sense depends on persisting chemical tracks laid down in the external
environment. In these and other respects olfaction might be regarded as
operating as a narcissistic, existential sense.
The problem before us now is to inquire how the brain gets out from
under the domination of the narcissistic, existential olfactory sense to be
altruistically guided by the futuristic, visual sense. In considering how the
weight of cerebral functions shifts in this respect, we will refer again to the
simplified anatomical diagram described in the first lecture (Figure 3) depict-
ing three main subdivisions of the limbic system. I reviewed evidence in-
dicating that the subdivision related to the amygdala is primarily concerned
with affects and emotional behaviour that ensure self-preservation. In other
words, its circuits are kept busy with the selfish demands of feeding, fighting,
and self-protection. In contrast, the subdivision comprising the septum and
related cortex appears to be involved in expressive and feeling states that are
conducive to sociability and the preservation of the species. It was pointed
44 A Triune Concept of the Brain and Behaviour

Figure 1 Diagram showing how anterior and midline thalamic structures involved in
primal sexual functions might articulate with evolutionary new parts of the medial dorsal
nucleus which project to the prefrontal cortex. A, anterior thalamic nuclei; M, mammil-
lary bodies; MD, medial dorsal nucleus. (From MacLean 196 7a)

out that the reciprocal relationship of the amygdala and septum helps to
explain the interplay of oral and genital functions, and that this situation
presumably evolved because of the primitive use of the snout and olfaction in
feeding and mating. As was brought out in the discussion of psychiatric
implications, it is of additional significance that pathways involved in oral and
genital functions converge in that part of the hypothalamus in which elec-
trical stimulation results in angry and defensive behaviour.
Looking at the anatomy in Figure 3, Lecture I, we obtain a clue as to why
consideration of the third subdivision was deferred until taking up the topic
of the present lecture. It will be noted that its main connecting pathway (no.
3) bypasses the olfactory apparatus. This pathway, which has no true
counterpart in the reptilian brain (LeGros Clark 1936), connects the mam-
millary bodies in the hypothalamus with the anterior thalamic nuclei, which
in turn project to the limbic cortex in the upper half of the ring. As schema-
tized in Figure l, it also brings this subdivision into association with the
medial dorsal nucleus which projects to part of the orbitofrontal limbic
cortex and to the prefrontal neocortex.
New Trends in Man's Evolution

A 16 A 13.5

Figure 2 Frontal sections of squirrel monkey brain, showing loci at which electrical
stimulation elicited penile erection. The open, half-filled, and solid diamonds and squares
denote gradations from partial to full erection; squares give the additional information
that stimulation induced hippocampal after-discharges. Small filled circles indicate that
stimulation was followed by 'rebound' erection ; large filled circles indicate 'rebound'
erection associated with hippocampal after-discharges. Vertical dashes identify negative
points. Key abbreviations: av, anterior ventral nucleus of thalamus; f, fornix; m, mam-
millary bodies; md, medial dorsal nucleus; mt, mammillothalamic tract; po, medial
preoptic area; s, septum; tt, thalamic tubercule. (From Maclean and Ploog 1962)
46 A Triune Concept of the Brain and Behaviour

25

.... o ••••••
••• • ◊ ••••••

.. . .. . • ·•• ♦ ◊
••••• -~ - •• ~ ~ 0
000 15

••••• ◊ ~~• • • ~ 0
• 0 • 0 ~

25 20 15 10

Figure 3 Diagram of medial aspect of frontal lobe of squirrel monkey with diamonds
and squares representing loci in pregenual cingulate and subcallosal cortex at which
electrical stimulation elicited penile erection. See legend to Figure 2 for key to these
symbols. Small open circles indicate negative points. Roman numerals surrounded by
circles give gradation of erection developing during a generalized seizure. (From Dua and
Maclean 1964)

In the evolution of the primate brain it deserves special emphasis that the
septal region remains relatively undeveloped, whereas structures comprising
the third subdivision become progressively larger and achieve their greatest
prominence in man. Our experimental observations on the squirrel monkey
suggest that this transformation reflects a shifting of emphasis from olfactory
to visual influences in sociosexual behaviour, Fundamental in this respect is
the finding that electrical stimulation in parts of the third subdivision, as in
the septa! circuit, elicits primal sexual responses (Maclean and Ploog 1962).
The squares and diamonds in Figure 2 (A I0.5-A6) show the positive loci for
penile erection in the anterior thalamic nuclei and core of the medial dorsal
nucleus. The squares give the added information that stimulation also elicited
hippocampal after-discharges. The solid circles indicate points at which stimu-
lation was followed by rebound erection. Figure 3 shows the distribution of
positive loci in areas of the rostral limbic cortex that are related respectively
to the anterior and medial dorsal nuclei of the thalamus (Dua and Maclean
I 964). Finally, we have found that stimulation at points along the spino-
thalamic pathway (Figure 4) and in caudal intralaminar structures adjoining
New Trends in Man's Evolution 47

Figure 4 The numerals in this diagram of squirrel monkey's brain identify three limbic
circuits in which electrical stimulation has elicited penile erection. The diagram also
schematizes the spinothalamic pathway and the termination of its medial division in
caudal intralaminar nuclei and paralaminar parts of the medial dorsal nucleus (MD).
Electrical stimulation in these thalamic structures and atipoints along the pathway
itself has resulted in scratching of the genitals and/or in ejaculation. The findings suggest
that thalamic structures involved in genital sensation and ejaculation articulate with the
part of the medial dorsal nucleus that is nodal with respect to penile erection. A major
effector pathway for penile erection appears to follow inferior thalamic peduncle (ITP)
to join the medial forebrain bundle (MFB) in the lateral hypothalamic area. Other
abbreviations: AC, anterior commissure; AT, anterior thalamus; M, mammillary bodies;
sept, septum. (From Maclean 196 2)

the medial dorsal nucleus results in scratching of the genitalia and ejaculation
(Maclean, Dua, and Denniston 1963b). The latter may occur even if the
monkey is prevented from scratching itself. Thus we are beginning to see for
the first time the course of sensory and motor pathways involved in the most
primitive aspects of procreational behaviour.
I shall now describe some comparative behavioural observations and results
of cerebral ablations which suggest that, in evolution, the medial thalamus
outgrows the septum partly because of a shifting from olfactory to visual
48 A Triune Concept of the Brain and Behaviour

Figure 5 Drawing of photograph in paper by Andrew 1964, showing greeting display

of male and female lemur (julvus). The lemur ranks as a primitive primate

influences in sexual behaviour. In such a prosimian form as the lemur fulvus

of Madagascar one finds a well developed septum characteristic of animals
with a large olfactory apparatus. As illustrated in Figure 5, the male and
female have a greeting display in which the snout and olfactory sense come
actively into play in mutual licking of the anogenital region (Andrew 1964).
Such a display provides an interesting contrast with that of a monkey
which is intermediate on the phylogenetic scale of primates. Our observations.
on the small South American squirrel monkey have revealed that this animal
has a genital display which depends on visual rather than olfactory communi-
cation. This is confirmed by our finding that one variety of these monkeys
will display equally well to its reflection in a mirror as to another monkey
(Maclean 1964b ).
In the social situation, display is characteristically performed as illustrated
in Figure 6. The displaying animal vocalizes and spreads its thighs and directs
the erect phallus towards the head of the other animal (Maclean 1962; Ploog
and Maclean 1963). The displaying animal may also grind its teeth. Recalling
what ethologists have described in reptilian and lower forms, the display in
this monkey is the same in courtship as in the show of aggression . In adults
the display occurs in its most dramatic form when a new male is introduced
New Trends in Man's Evolution 49

Figure 6 As opposed to the lemur's display shown in Figure 5, the genital display of
the squirrel monkey (Saimiri sciureus) depends primarily on visual communication. The
communal form of the display illustrated above may be used either in courtship or as a
show of aggression and dominance by one male to another male. In each case the dis-
playing monkey spreads its thighs and thrusts the erect penis towards the head of the
other animal. (From Ploog and MacLean 1963)

into the territory of an established colony of squirrel monkeys (Maclean,

unpublished observations). Within seconds all males begin to display to the
new monkey while grinding their teeth, and if the new monkey does not
remain quiet and submissive with its head bowed, it is viciously attacked. In a
50 A Triune Concept of the Brain and Behaviour

Plate I Two varieties of squirrel monkeys that are informally identified as gothic and
roman types because of the pointed and rounded appearance of the ocular patch where
it arches over the eye. Both types perform a characteristic genital display in communal
situations. But only the gothic type shown on left will consistently display to its reflec•
tion in a mirror. The author has used the mirror display test as a means of discovering
what parts of the brain are involved in its performance. See text. (From Maclean 1964a)

special study we found that the incidence of display among males provides a
better measure of aggression and dominance than the outcome of rivalry for
food (Ploog and MacLean 1963). Females will occasionally display with
clitoral tumescence. Ploog, Hopf, and Winter (196 7) have since observed that
without prior exposure to any other animal than its own mother, the infant
squirrel monkey will display as early as the second day of life to another
monkey, clearly indicating that it is an unlearned form of behaviour.
Elsewhere I have discussed the comparative implications of these findings
and have cited examples of remnants of genital display in man (Maclean
1962, forthcoming). In primitive cultures in different parts of the world the
territorial aggressive implications of the display are illustrated by house-
guards - stone monuments showing an erect phallus - used to mark
territorial boundaries. It was as though a visual urogenital symbol was used as
a substitute for the olfactory urinary markings of lower mammals. In chapter
24 of Genesis, Abraham says to his eldest servant, 'Put I pray thee thy hand
New Trends in Man's Evolution 51

under my thigh; and ... swear that thou shall not take a wife unto my son of
the daughters of the Canaanites.' Thigh, it is said, is used here euphemistically
for the genitals. In his Three Contributions to the Theory of Sex Freud
(1948, p. 52) commented that 'the child is above all shameless and during its
early years it evinces definite pleasure in displaying its body and especially its
sexual organs.' In the squirrel monkey, display may be triggered by the re-
flection of a single eye (MacLeari 1964b ), suggesting that 'looking into the
eye' has primitive origins as an aggressive act. In Italy less than 200 years ago
amulets showing an erect phallus are said to have been worn as a protection
against the evil eye (Knight 1865). I have suggested that primitive man may
have learned that by covering himself, he reduced unpleasant social tensions
arising from this archaic impulse to display, and that this, rather than
modesty, has led to the civilizing influence of clothing (Maclean 1962).
The display is also used as a form of social greeting, and as mentioned, we
found that one variety of squirrel monkey will regularly display to its own
reflection in a mirror (Maclean 1964b ). Plate 1 shows the two races of
monkeys we have worked with. We informally refer to the mirror displaying
animal on the left as the gothic type because the ocular patch comes to a
peak over the eye like a gothic arch. We call the other variety romans because
the patch is round in the form of a Roman arch.
For investigative purposes, the mirror display test has the special advantage
that it excludes olfactory and other cues. With daily testing most gothic-type
monkeys living in visual isolation will consistently display to their reflection
in a mirror. This provides a reliable means of obtaining pre- and post-
operative scores. Using this test, I have studied the effects of lesions in various
parts of the brain on the display behaviour of more than 60 monkeys
(Maclean , unpublished observations 1969).
With respect to the medial dorsal nucleus it is significant that the larger the
lesion in this structure, the greater is the decline in the incidence of the
display. Plate 2 shows a histological section from an animal in which almost
the entire nucleus was destroyed. Before operation this animal scored a 100
per cent performance in 30 trials , but during two and a half months of
post-operative testing there was almost a total absence of display (p<0.0001 ).
Yet when introduced into our testing colony this animal was able to defend
itself and to overpower the dominant animal.
These lesion studies, therefore, suggest that the display comes under limbic
and neocortical influences through the evolutionary expansion of the medial
thalamus and its connections with the limbic and prefrontal cortex. Such a
conjecture, however, leaves unanswered how visual influences would be
brought to bear on the medial thalamus. As regards the neomammalian brain,
the possibility exists that visual connections are established by intrathalamic
52 A Triune Concept of the Brain and Behaviour

Plate 2 Photomicrograph of a lesion produced by electro-coagulation which destroyed

the medial dorsal nucleus in a gothic-type squirrel monkey. See text for description of
the effects of the lesion on the mirror display test

'association' fibres passing from the pulvinar to the medial dorsal nucleus. But
there has existed no basis for visual limbic connections. Classical anatomy
would lead one to believe that, in evolution, visual cortical projections by-
passed the limbic cortex and established connections only with a highly
specialized part of the neocortex known as the striate area (Maclean 1966).
The traditional emphasis on the close relationship of the hippocampal
formation to the rhinal fissure and olfactory cortex appears to have diverted
attention from its posterior relationship to the calcarine sulcus and visual
cortex (Maclean 1966). In both primates and subprimates there is an abrupt
transition between the striate and the limbic cortex. Figure 7 shows sche-
matically this relationship in the brain of man (see Maclean 1966, Fig. 1 for
the histological picture in monkey). As described in Lecture 2, we found by
New Trends in Man's Evolution 53

Figure 7 Medial view of human brain, illu strating the close relationship of posterior
limbic and visual areas. At the depth of the rostral end of posterior calcarine sulcu s, the
striate cortex (stipple) meets the limbic cortex. In the monkey photically responsive
units have been found in the posterior hippocampal gyrus and adjacent areas. (Modified
after Sanides and Vitzthum 1964)

microelectrode recording in awake, sitting monkeys that cells in the posterior

parahippocampal cortex are specifically activated by photic stimulation (see
Lecture 2, Plates 4 and 5, and Figure 4). In addition , I described our neuro-
anatomical findings that degenerating fibres from the ventrolateral part of the
lateral geniculate body can be traced into the same cortex (Lecture 2, Plate 6).
The parahippocampal cortex is a major source of afferents to the hippo-
campus, which in turn projects to the anterior thalamic nuclei, hypothalamus,
and other structures of the brain stem involved in emotional, endocrine, and
somatovisceral functions. As diagrammed in Figure 1, visual information from
the limbic cortex might reach the medial dorsal nucleus through the articula-
tion of this nucleus with the anterior thalamic nuclei.
In the evolution of primates there is a great expansion of the perilimbic
and perivisual convolutions in the medial occipitotemporal region. Indeed,
the convolution called the fusiform gyrus which lies parallel to the hippo-
campal gyrus (Figure 7) looms large as a new convolution in the brains of
higher primates (Ariens Kappers, Huber, and Crosby 1936). The work of
Penfield and others (Penfield 1952 ; Penfield and Jasper 1954) makes these
convolutions loom large also from a psychological and psychiatric point of
54 A Triune Concept of the Brain and Behaviour

view. Irritative lesions or electrical stimulation deep in the temporal lobe

result in visual illusions, evocation of memories, and many forms of intense
emotional experience. There may be a feeling of familiarity or deja vu carrying
with it a sense of conviction that one has already seen and experienced what
is occurring (Jackson 1889;Jackson and Colman 1898), or things may appear
strange, altered, or unreal. In the light of the introductory remarks about
altruism it should be mentioned that among the reported effects is a feeling
of sadness and wanting to cry. In speaking to different groups I have asked
how many have experienced a tearful feeling upon seeing an altruistic act, and
invariably almost every hand goes up.
In leading up to the next point I want to make, it is necessary to ask: what
is it psychologically that distinguishes each one of us as individuals? Intro-
spection reveals that the condition most crucial to our feeling of individuality
is our two-fold source of information from the external public world and our
own private world. Signals to the brain from the world within are entirely
private, being self-contained, whereas those from the outside world can be
publicly experienced and lend themselves to comparison among individuals.
Our sense of individuality, therefore, as well as our personal identification
with happenings in the outside world, would seem to depend on a bond of
internal and external experience (Maclean 1969a). As mentioned in Lecture
1, there are indications that the neocortex performs many of its nice dis-
criminations without interference from signals and noise generated in the
internal world . It is probable, however, on anatomical grounds alone , that the
limbic cortex is constantly bombarded by impulses from the internal world.
Recently, we have achieved success in recording from inside single nerve cells
of the hippocampus (archicortex) of awake, sitting monkeys while testing the
effects of electrical stimuli to representative internal and external inputs
(Yokota, Reeves, and Maclean 1970). Both forms of stimulation induced
excitatory postsynaptic potentials, but only the internal input was effective
in discharging cells. Here is evidence that part of the limbic cortex provides a
functional bridge between the internal and external worlds, and that the
tollgate regulating the flow of neural traffic is located on the internal side of
the bridge. The newly discovered visual connections that have been described
would provide one mechanism by which a union could occur between visual
and visceral experience.
Kubie (1953) would claim that memory likewise depends on a combina-
tion of internal and external experience . One might say that the union of
internal and external experience was as important for memory as the com-
bination of antigen and antibody for an enduring immunity. Indeed, there is
evidence from clinical studies of limbic epilepsy that with epileptic disruption
of somatovisceral integrative functions of the limbic system, experiences
New Trends in Man's Evolution 55

/
_., -------
18 16
-
...
~----...:..=...

Figure 8 Figure from paper by Penfield and Perot (I 963 ). See text for verbal report
of patient when point E was stimulated
56 A Triune Concept of the Brain and Behaviour

Figure 9 Figure from paper by Penfield and Perot (1963). See text for results of
stimulating at point no. 18

cannot be remembered. A growing body of clinical evidence indicates that

damage to the hippocampus or structures linking it to the anterior thalamus
(Figure l) may interfere with the storage of memories (see Maclean, 1968b
for some key references). Electrical stimulation of the epileptic brain, on the
contrary, has been able to evoke memories. Figure 8 shows the locus near the
hippocampal formation in one of Penfield's (Penfield and Perot 1963) cases
(point E, case no . 36) where stimulation caused the patient to say, 'A familiar
memory - the place where I hang my coat up, where I go to work' (p . 65 I).
Perhaps the reason that so relatively little attention has been given to
internal experience in mechanisms of learning and memory, is that we are so
often unaware of the subtle respiratory, cardiovascular, alimentary , and other
changes that take place during our preoccupation with happenings in the
external world (Maclean 1969a). How many of us, for example, are usually
aware of the 'breath of recognition' when we see an old friend or arrive at the
solution of a problem?
New Trends in Man's Evolution 57

STAGE A

REMP

II EEG

,~:o,j
1.5
.1Penile
clrcum. (cm~

Figure 10 Graph from publication by Fisher, Gross, and Zuch (1965, p.36), illus-
trating the presence of penile erection (lowest record) during rapid eye movement
periods (REMP). Penile erection was recorded by a delicate mercury strain gauge (SG).
(Copyright 1965, American Medical Association)

As Penfield's (Penfield and Perot 1963) work has also emphasized, the
temporal lobe appears to play an important role in dreaming. Figure 9 shows
a locus in the hippocampal gyrus, in another of his cases (case no. 22), at
which stimulation caused the young woman to say, 'I keep having dreams - I
keep seeing things. I keep dreaming of things' (p. 635).
The work of Aserinsky and Kleitman (I 953), and Dement and Kleitman
(19 57) has led to the recognition that there are cyclic periods of rapid eye
movement (REM) during sleep that are accompanied by dreaming. There are
usually five or six of these periods lasting about thirty minutes. They are
characterized by a waking type of brain wave, irregularities of respiration and
heart rate, muscle twitchings, grinding of the teeth (bruxism), and as il-
lustrated by Figure l O from a paper by Fisher, Gross, and Zuch (I 965), there
is commonly penile erection lasting throughout the REM period. Kara9an,
Marans, Barnet, and Lodge (I 968) have macie similar observations in infants.
In macrosomatic animals the hippocampus shows rhythmic theta activity
during the REM phase. Finally, there is now considerable evidence suggesting
that the hippocampus exerts a modulating effect on autonomic functions (see
Maclean 1968b for brief review). It is of interest to consider these various
findings in the light of the visually responsive areas I described in the para-
hippocampal cortex, because it is conceivable that the archicortex may be
implicated in the visual aspects of dreaming and in autonomic manifestations
of REM sleep. It is impossible, of course, to mention the term archicortex in
this context without being reminded of the archetypal forms of dreaming
described by Jung (1959).
58 A Triune Concept of the Brain and Behaviour

But how can one explain the functional connection between grinding of
the teeth and penile erection during dreaming? Our brain and behavioural
studies offer a clue. Recently we found pontine reticular cells near the motor
nucleus for the jaw that are photically activated (MacLean, Yokota, and
Kinnard 1969). Significantly, these cells belong to a nuclear group (the lateral
tegmental process) in which electrical stimulation results in a penile erection
(MacLean, Denniston, and Dua I 963a ). This will recall that the aggressive
territorial display of the squirrel monkey which I described earlier not only
involves penile erection, but also grinding of the teeth. It is possible, there-
fore, that penile erection of dreaming represents part of an archetypal
primate pattern of display which in our far ancestral past was variously used
in a show of aggression, in courtship, or in greeting.
We have now, in a sense, travelled full circle, and like the earth rotating
from shadow into sunlight, returned to where we started in the introduction,
with the question arising about the continuing evolution of altruism and its
dependence on insight and foresight. Figure I shows diagrammatically how
the visually responsive limbic cortex looks inward, as it were, upon interocep-
tive parts of the diencephalon that in turn looks outwards upon the recently
evolved prefrontal neocortex . Clinical analysis of cases with frontal lobe
injury suggests that the prefrontal cortex provides foresight in planning for
ourselves and others. There is also evidence that it functions in helping us to
gain insight into the feelings of others. The diagram suggests how it might
receive part of its insight - the capacity 'to see with feeling' - through its
connections with the limbic brain (MacLean 1967a). It is also evident from
the diagram that the prefrontal cortex is geared in with the expanding third
subdivision of the limbic system which, on the basis of recent evidence
appears to be involved in sexual functions, and hence the procreation and
preservation of the species. In the complex organization of these phylo-
genetically old and new structures ' ... we presumably have a neural ladder, a
visionary ladder, for ascending from the most primitive sexual feeling to the
highest level of altruistic sentiments' (Maclean I 962, p . 300). It is possible
that these large evolving territories of the brain are incapable of being brought
into full operation until the hormonal changes of adolescence occur. If so, it
would weigh heavily against the claims of those who contend that the per-
sonality is fully developed and rigid by adolescence, if indeed not by the age
of five or six.
In conclusion, I will quote from the final paragraph of an article that I
contributed to a recent book entitled, Alternatives to Violence:

In the dawn's early light of brain research we gain the intimation that we have
in the new brain a mechanism that is fully capable of dealing with the
New Trends in Man's Evolution 59

difficult medical and social problems of our time. The most explosive issue,
of course, is the problem of controlling man's reptilian intolerance and rep-
tilian struggle for territory, while at the same time finding a means of regu-
lating our soaring population. Language barriers among nations present great
difficulties in arriving at solutions. But the greatest language barrier lies
between man and his animal brains; the neural machinery simply isn't there
for communication in verbal terms. Nevertheless, in the last 2000 years the
layman all on his own has made great strides in domesticating his emotions.
With the increasing insights that are being obtained from the brain and be-
havioral sciences, man should be able to continue to harness his emotions for
progressively constructive purposes. If, through education, we could only
apply what the brain already knows, the year 2000 might see the beginning of
a truly golden age. (1968a, pp. 33-4)

SUMMARY

This lecture considers new trends in the evolution of man's brain, turning
attention to the third subdivision of the limbic system which shows pro-
gressive development in higher primates and culminates in man. Unlike the
two subdivisions discussed in the first lecture, it shows a relative inde-
pendence of the olfactory apparatus.
Experimental evidence suggests that the progressive growth of the third
subdivision reflects a shifting of emphasis from olfactory to visual influences
in socio-sexual behaviour. Fundamental in this respect is the finding that
electrical stimulation in parts of this subdivision elicits primal sexual
responses. The inference regarding the role of vision derives from observations
on alterations of a visually induced genital display of the squirrel monkey
following damage to midline thalamic structures. Moreover, we have recently
shown by microelectrode and neuroanatomical studies that part of the limbic
cortex receives visual connections. This cortex is strongly connected to the
hippocampus which in turn projects to the anterior thalamic nuclei, hypo-
thalamus, and other structures of the brain stem involved in emotional, endo-
crine and somatovisceral functions . Such connections may account for a bond
between visual and visceral experience which appears to be essential for effec-
tive identification with what is visually experienced, as well as for what is
visually remembered. Clinical evidence suggests that the limbic cortex in
question is involved in memory functions and dreaming.
Through articulations in the anterior and medial thalamus, the third sub-
division establishes a relationship with the newly evolved prefrontal cortex,
60 A Triune Concept of the Brain and Behaviour

which, on the basis of clinical evidence, is believed to be involved in insight

and foresight in planning for ourselves and others. It would therefore seem
that in the complex organization of the old and new structures under con-
sideration we possess 'a neural ladder, a visionary ladder, for ascending from
the most primitive sexual feeling to the highest level of altruistic sentiments'
(MacLean 1962).
Despite the animalistic problems of communication within the triune
brain, there are grounds for hope that man's neocortical intelligence, if com-
bined with self-knowledge, will be able to cope with the mounting crises of
our time.
 BIBLIOGRAPHY

Altman, J. and G.D. Das, Autoradiographic and histological evidence of post-

natal hippocampal neurogenesis in rats. J. Comp. Neurol., 1965,
124:319-35
van Andel, H. and A.M. Ernst, Tryptamin-Katatonie, eine cholinergische
Hypofunktion im zentralen Nervensystem. Psychopharmacologia, 1961,
2:461-6
Anden, N.-E., A. Dahlstrom, K. Fuxe, and K. Larsson, Further evidence for
the presence of Nigro-neostriatal dopamine neurons in the rat. Am. J.
Anat., 1965, 116:329-34
Andrew, R.J., The displays of the primates. In J. Buettner-Janusch (ed.), Evo-
lutionary and Genetic Biology of Primates, vol. 2. New York: Academic
Press, 1964, Pp . 227-309
Ardrey, R., The Territorial Imperative. New York: Atheneum, 1966
Ariens Kappers, C.U., G.C. Huber, and E.C. Crosby, The Comparative
Anatomy of the Nervous System of Vertebrates, Including Man. 2 vols.
New York: Macmillan, 1936. Pp. 864-1845
Aserinsky, E. and N. Kleitman, Regularly occurring periods of eye motility,
and concomitant phenomena, during sleep. Science, l 953, l 18: 273-4
Broca, P., Anatomie comparee des circonvolutions cerebrales. Le grand lobe
limbique et la scissure limbique dans la serie des mammiferes. Rev.
Anthrop., 1878, 1 :385-498
Bunney, W.E., Jr and J.M. Davis, Norepinephrine in depressive reactions: a
review.Arch . Gen. Psychiat., 1965, 13:483-94
Calhoun, J.B ., Population density and social pathology. Sci. Am., I 962,
206:139-46
Calhoun, J.B., Space and the strategy of life. In A.M. Esser (ed.), Behavior
and Environment. New York: Plenum Press, 1971. Pp. 329-87
Coggeshall, R.E. and P.D. MacLean, Hippocampal lesions following ad-
ministration of 3-acetylpyridine. Pro. Soc. Exp. Biol. Med., 1958,
98:687-9
Cohen, S., Psychotomimetic agents. In H.W. Elliott, W.C. Cutting, and
R.H. Dreisbach (eds.), Annual Review of Pharmacology, vol. 7. Palo Alto,
Calif. : Annual Reviews, Inc., 1967. Pp. 301-18
Comte, A., Altruism. In C.T. Onions (ed.), The Oxford Universal Dictionary
on Historical Principles . London: Oxford University Press, I 955
62 A Triune Concept of the Brain and Behaviour

Csillik, B. and S.D. Erulkar, Labile stores of monoamines in the central ner-
vous system : a histochemical study. J. Pharmacol. Exp. Therap., 1964,
146 : 186-93
Daitz, H.M. and T.P.S. Powell, Studies of the connexions of the fornix
system. J. Neural. Neurosurg. Psychiat., 1954, I 7: 7 5-82
Dement, W. and N. Kleitman, The relation of the eye movements during sleep
to dream activity : an objective method for the study of dreaming. J. Exp.
Psycho/., 1957, 53:339-46
Dewhurst, W.G., New theory of cerebral amine function and its clinical ap-
plication. Nature, I 968, 218 : 1130-3
Downing, D.F., Psychotomimetic compounds. In M. Gordon (ed.), Psycho·
pharmacological Agents, vol. 1. New York and London: Academic Press,
1964. Pp . 5 55-6 I 8
Drachman, D.A. and R.D. Adams, Herpes simplex and acute inclusion-body
encephalitis. Arch. Neurol. , 1962 , 7:45-63
Dua, S. and P.D. Maclean, Localization for penile erection in medial frontal
lobe. Am. J. Physiol., 1964, 207: 1425-34
Ernst, A.M., H. van Andel, and G.A. Charbon, Beruht die experimentelle
Katatonie durch Tryptamin auf einer Verdriingung des 5-Hydroxytrypta-
min? Psychopharmacologia, 1961, 2:425-35
von Euler, C., On the significance of the high zinc content in the hippocampal
formation . In Physiologie de l'hippocampe. Paris: Centre National de la
Recherche Scientifique, 1962. Pp . 135-45
Fisher, C., J . Gross, and J. Zuch, Cycle of penile erection synchronous with
dreaming (REM) sleep; preliminary report. Arch. Gen. Psychiat., 1965 ,
12:29-45
Flanigan, S., E.R. Gabriell, and P.D. Maclean , Cerebral changes revealed by
radioautography with s 35-labelled L-methionine. AMA Arch. Neurol.
Psychiat., 1957 , 77 :588-94
Fleischhauer, K. and E. Horstmann, Intravitale Dithizonfiirbung homologer
Felder der Ammonsformation von Siiugern. Z. Zel/forsch . Mikroskop.
Anat. , 1957 , 46 :598-609
von Foerster, H. , P.M. Mora, and L.W. Amiot, Doomsday : Friday, 13 Novem-
ber, AD 2026. Science, 1960, 132: 1291-95
Freud , S., Beyond the Pleasure Principle. Translated by C.J.M. Hubback .
London and Vienna: The International Psycho-Analytical Press, 1922
Freud, S., Three Contributions to the Theory of Sex. Translated by A.A .
Brill. (Nervous and Mental Disease Monograph, no. 7.) New York: Journal
of Nervous and Mental Disease Pub. Co., 1948
Fuxe, K. , Evidence for the existence of monoamine neurons in the central
nervous system. III. The monoamine nerve terminal. Z. Zell[ors ch.
Mikroskop. Anat., 1965a, 65 :573-96
Fuxe, K., Evidence for the existence of monoamine neurons in the central
nervous system. IV. Distribution of monoamine nerve terminals in the
central nervous system. Acta Physiol. Scand. , 1965b, 64 : 37-84
Gaddum, J.H., Antagonism between lysergic acid diethylamide and 5-
hydroxytryptamine. J. Physiol., 1953 , 121 : 15P
Gelhard, R., J. Perez-Cruet , and G.L. Gessa, Catalepsy induced in squirrel
monkeys by l-5-hydroxyhyptophan (5-HTP). Fed. Proc., 1971, 30 :671
Bibliography 63

Girgis, M., Distribution of cholinesterase in the basal rhinencephalic struc-

tures of the coypu (Myocastor coypus). J. Comp. Neurol., 1967,
129:85-95
Glaser, G.H. and J.H. Pincus, Limbic encephalitis. J. Nervous Mental Disease,
1969, 149:59-67
Glaser , G.H. , G.B. Solitare, and E.E. Manuelidis, Acute and subacute in-
clusion encephalitis. Res. Pub/. Assoc. Res. Nervous Mental Disease, 1964,
44 : 178-215
Green,J .D., Thehippocampus.Physio/. Rev., 1964, 44:561-608
Green , J.D. and W.R. Adey., Electrophysiological studies of hippocampal
connections and excitability. Electroencephalog. Clin. Neurophysiol.,
I 956, 8: 245-62
Harper, L .V. , Ontogenetic and phylogenetic functions of the parent-offspring
relationship in mammals. In D.S. Lehrman , R.A . Hinde, and E. Shaw
(eds.) , Advances in the Study of Behavior, vol. 3. New York : Academic
Press, 1970. Pp. 75-117
Heath, R.G. (chairman), Studies in Schizophrenia. A Multidisciplinary Ap-
proach to Mind-Brain Relationships. Cambridge: Harvard University Press,
1954
Howard, H.E., An Introduction to the Study of Bird Behavior. Cambridge:
The University Press, 1929
Jackson, J. Hughlings, On a particular variety of epilepsy ('intellectual aura'),
one case with symptoms of organic brain disease. Brain, 1889, 11 : 179-207
Jackson, J. Hughlings, and W.S . Colman, Case of epilepsy with tasting move-
ments and 'dreamy state' - very small patch of softening in the left
uncinate gyrus. Brain, 1898 , 21:580-90
Jackson, J. Hughlings and P. Stewart, Epileptic attacks with a warning of a
crude sensation of smell and with the intellectual aura (dreamy state) in a
patient who had symptoms pointing to gross organic disease of the right
temporo-sphenoidal lobe. Brain, 1899, 22 :534-49
Jung, C.G ., The Archetypes and the Collective Unconscious. Translated by
R.F .C. Hull. Collected works, vol. 9, part I. New York: Pantheon, 1959
Juorio , A.V. and M. Vogt , Monoamines and their metabolites in the avian
brain. I . Physiol., 1967, 189:498-518
Karayan, I., A. Marans, A. Barnet, and A. Lodge, Ontogeny of penile erection
during sleep in infants. Psychophysiology, 1968, 4: 363-4
Kluver, H., Functional differences between the occipital and temporal lobes,
with special reference to the interrelations of behavior and extracerebral
mechanisms. In L.A. Jeffress (ed.), Cerebral Mechanisms in Behavior. New
York : J. Wiley, 1951. Pp. 147-82
Kluver, H. and P.C. Bucy, Preliminary analysis of functions of the temporal
lobes in monkeys. Arch. Neurol. Psychiat., 1939, 42:979-1000
Knight, R.P., A Discourse on the Worsh ip of Priapus and its Connexion with
Mystic Theology of the Ancients. London: privately printed , 1865
Koestler, A., The Ghost in the Machine. London : Hutchinson, 1967
Kubie, L.S., Some implications for psychoanalysis of modern concepts of the
organization of the brain. Psychoanal. Quart., 1953, 22 :21-68
LeGros Clark, W.E., Functional localization in the thalamus and hypo-
thalamus. J. Mental Sci., 1936, 82:99-118
64 A Triune Concept of the Brain and Behaviour

Lewis, P.R. and C.C.D. Shute, The cholinergic limbic system: projections to
hippocampal formation, medial cortex, nuclei of the ascending cholinergic
reticular system, and the subfornical organ and supra-optic crest. Brain,
1967, 90:521-40
Lipps, T .,Asthetik. 2 vols. Leipzig: Voss, 1914-20
Lorenz, K .Z ., The companion in the bird's world. Auk, 1937, 54:245-73
McLardy, T., Observations on the fornix of the monkey. I. Cell studies. J.
Comp . Neurol., 1955, 103:305-25
McLardy, T., Zinc enzymes and the hippocampal mossy fibre system. Nature,
1962, 194:300-2
Maclean, P.D., Psychosomatic disease and the 'visceral brain.' Recent
developments bearing on the Papez theory of emotion. Psychosomat.
Med., 1949, 11:338-53
MacLean, P.D., Some psychiatric implications of physiological studies on
frontotemporal portion of limbic system (visceral brain). Electro-
encephalog. Clin. Neurophysiol., 1952, 4:407-18
Maclean, P.D. , The limbic system and its hippocampal formation. Studies in
animals and their possible application to man. J. Neurosurg., 1954,
11:29-44
MacLean, P.D., Chemical and electrical stimulation of hippocampus in un-
restrained animals. I. Methods and electroencephalographic findings. AMA
Arch. Neurol. Psychiat., 1957a, 78: 113-27
MacLean, P.D., Chemical and electrical stimulation of hippocampus in un-
restrained animals. 11. Behavioral findings. Ibid., 1957b, 78: 128-42
MacLean, P.D., Contrasting functions of limbic and neocortical systems of
the brain and their relevance to psychophysiological aspects of medicine.
Am. J. Med., 1958a, 25:611-26
MacLean, P.D., The limbic system with respect to self-preservation and the
preservation of the species. J. Nervous Mental Disease, 19 58b, 1 27 : 1- I 1
MacLean, P.D., The limbic system with respect to two basic life principles. In
M.A.B . Brazier (ed.), The Central Nervous System and Behavior. New
York: Josiah Macy, Jr. Foundation, 1959. Pp. 31-118
MacLean, P.D., New findings relevant to the evolution of psychosexual func-
tions of the brain.J. Nervous Mental Disease, 1962, 135:289-301
MacLean, P.D. , Man and his animal brains. Mod. Med., 1964a, 32:95-106
MacLean, P.D., Mirror display in the squirrel monkey, Saimiri sciureus.
Science, 1964b, 146: 950-2
MacLean, P.D., The limbic and visual cortex in phylogeny: further insights
from anatomic and microelectrode studies. In R. Hassler and H. Stephan
(eds.), Evolution of the Forebrain . Stuttgart: G. Thieme, 1966. Pp. 443-53
MacLean, P.D., The brain in relation to empathy and medical education. J.
Nervous Mental Disease, 1967a, 144:374-82
MacLean, P.D., A chronically fixed stereotaxic device for intracerebral ex-
ploration with macro- and micro-electrodes. Electroencephalog. Clin.
Neurophysiol., l 967b, 22: 180-2
MacLean, P.D., Alternative neural pathways to violence. In L. Ng (ed .), Alter-
natives to Violence. New York: Time-Life Books, 1968a. Pp. 24-34
MacLean, P.D., Ammon 's Horn : a continuing dilemma. Foreword in S.
Ramon y Cajal, The Structure of Ammon 's Horn. Translated by L. Kraft .
Springfield: Charles C. Thomas, 1968b
Bibliography 65

MacLean, P.O., The internal-external bonds of the memory process. J.

Nervous Mental Disease, 1969a, 149 :40-7
MacLean, P.O., The paranoid streak in man. In A. Koestler and J.R. Smythies
(eds.), Beyond Reductionism. London : Hutchinson, 1969b
MacLean, P.O., Brain and vision in the evolution of emotional and sexual
behavior. Thomas William Salmon Lectures, New York Academy of Medi-
cine, December 7, I 966 (forthcoming)
MacLean , P.O., Cerebral evolution and emotional processes: new findings on
the stria ta! complex. Ann. NY A cad. Sci., I 972, I 93: 137-49
MacLean, P.O. and G. Creswell, Anatomical connections of visual system with
limbic cortex ofmonkey.J. Comp. Neural., 1970, 138 :265-78
MacLean, P.O. and J.M.R. Delgado, Electrical and chemical stimulation of
frontotemporal portion of limbic system in the waking animal. Electro-
encephalog. Clin. Neurophysiol., 1953, 5:91-100
MacLean, P.O., R.H. Denniston, and S. Dua, Further studies on cerebral
representation of penile erection: Caudal thalamus, midbrain, and pons. J.
Neurophysiol., I 963a, 26:273-93
MacLean, P.O., S. Dua, and R.H. Denniston, Cerebral localization for scratch-
ing and seminal discharge. Arch. Neural., 1963b, 9:485-97
MacLean, P.O., S. Flanigan, J.P. Flynn, C. Kim, and J.R. Stevens, Hippo-
campal function: tentative correlations of conditioning, EEG, drug, and
radioautographic studies. Yale J. Biol. Med., I 95 5-6, 28: 380-95
MacLean, P.O. arid D.W. Ploog, Cerebral representation of penile erection. J.
Neurophysiol., 1962, 25:29-55
MacLean, P.O., T. Yokota, and M.A. Kinnard, Photically sustained on-
responses of units in posterior hippocampal gyrus of awake monkey. J.
Neurophysiol., 1968, 31: 870-83
Malamud, N., The epileptogenic focus in temporal lobe epilepsy from a
pathological standpoint. Arch. Neural., I 966, I 4: 190-5
Maske, H., Uber den topochemischen Nachweis von Zink im Ammonshorn
verschiedener Siiugetiere. Naturwissenschaften, 1955, 42:424
Olds, J. and P. Milner, Positive reinforcement produced by electrical stimula-
tion of septa! areas and other regions of rat brain. J. Comp. Physiol.
Psycho/., 1954, 47:419-27
Paasonen, M.K., P.O . MacLean, and N.J. Giarman, 5-Hydroxytryptamine
(serotonin, enteramine) content of structures of the limbic system. J.
Neurochem., 1957, I :326-33
Papez, J.W. A proposed mechanism of emotion. Arch. Neural. Psychiat.,
1937, 38 :725-43
Parent, A. and A. Olivier, Comparative histochemical study of the corpus
striatum. J. Fur Hirnforsch., 1970, 12 :73-81
Parmeggiani, P.L., On the functional significance of the hippocampal 0-
rhythm . In W.R. Adey and T. Tokizane (eds.), Progress in Brain Research,
vol. 27. Amsterdam: Elsevier, 1967. Pp. 413-41
Penfield, W., Memory mechanisms. AMA Arch. Neural. Psychiat. , 1952,
67 : 178-91
Penfield, W. and H . Jasper, Epilepsy and the Functional Anatomy of the
Human Brain. Boston : Little, Brown, 1954
Penfield, W. and P. Perot, The brain's record of auditory and visual experi-
ence. A final summary and discussion. Brain, 1963, 86: 596-696
66 A Triune Concept of the Brain and Behaviour

Pfaff, D.W., Autoradiographic localization of radioactivity in rat brain after

injection of tritiated sex hormones. Science, 1968, 161: 135 5-6
Ploog, D., S. Hopf, and P. Winter, Ontogenese des Verhaltens von Totenkopf-
Affen (Sairniri sciureus). Psycho/. Forsch., 1967, 31: 1-41
Ploog, D. and P.D. Maclean, Display of penile erection in squirrel monkey
(Sairniri sciureus). Animal Behav., 1963, 11 :32-9
Pribram, K.H. and P.D. Maclean, Neuronographic analysis of medial and
basal cerebral cortex. II. Monkey. J. Neurophysiol., 1953, 16:324-40
Purpura, D.P. and 0 . Gonzalez-Monteagudo, Acute effects of methoxy-
pyridoxine on hippocampal end-blade neurons; an experimental study of
'special pathoclisis' in the cerebral cortex. J. Neuropathol. Exp. Neural.,
1960, 19:421-32
Reeves, A.G., K. Sudakov, and P.D. Maclean, Exploratory unit analysis of
exteroceptive inputs to the insular cortex in awake, sitting, squirrel
monkeys. Fed. Proc., 1968, 27:388
Reivich1 M. and J. Glowinski, An autoradiographic study of the distribution
of C 4-norepinephrine in the brain of the rat. Brain, 196 7, 90: 633-46
Saavedra, J.M. and J. Axelrod, Psychotomimetic N-methylated tryptamines:
formation in brain in vivo and in vitro. Science, I 972, 175: I 365-6
Sanides, F. and H. Vitzthum, Zur Architektonik der menschlichen Sehrinde
und den Prinzipien ihrer Entwiklung. Deut. Z. Nervenheilk., 1965,
187:680-707
Schafer, E.A. , Text Book of Physiology. 2 vols. Edinburgh and London : Y .J.
Pentland, 1898-1900
Schildkraut, J.J., The catecholamine hypothesis of affective disorders: a
review of supporting evidence. Am. J. Psychiat., 1965, 122: 509-22
Shute, C.C.D. and P.R. Lewis, The use of cholinesterase techniques combined
with the operative procedures to follow nervous pathways in the brain.
Bibi. Anat., 1961, 2:34-49
Spalding, D.A., Instinct, with original observations on young animals. Brit. J.
AnimalBehav., 1954, 2:2-11
Sudakov, K., P.D. Maclean, A. Reeves, and R. Marino, Unit study of extero-
ceptive inputs to claustrocortex in awake, sitting, squirrel monkey.
Brain Res., I 971, 28: 19-34
Tower, D.B., Pyridoxine and cerebral activity . Nutr. Rev., 1958, 16: 161-4
Vogt, C. and 0. Vogt, Gestaltung der topistischen Hirnforschung und ihre
Forderung durch den Hirnbau und seine Anomalien. J. Hirnforsch., 1953,
I: 1-46
Woolley, D.W. and E. Shaw, A biochemical and pharmacological suggestion
about certain mental disorders. Proc. Nat. A cad. Sci. U.S., 1954,
40:228-31
Yokota, T., A.G. Reeves, and P.D. Maclean, Intracellular olfactory response
of hippocampal neurons in awake, sitting, squirrel monkeys. Science,
1967, 157:1072-4
Yokota, T., A.G . Reeves, and P.D. Maclean, Differential effects of septal and
olfactory volleys on intracellular responses of hippocampal neurons in
awake, sitting monkeys. J. Neurophysiol., 1970, 33: 96-107
 CONTRIBUTORS

E. Beaumaster MA, DIP PSYCH, is a Research Assistant, Department of Psy-

chology, Queen's University, Kingston, Ontario
R. Broughton MD, PH D, is Associate Professor of Medicine, and Associate,
Medical Research Council, University of Ottawa, Ottawa, Ontario
D. Campbell PH D, is Professor, Departments of Psychiatry and Psychology,
and Ontario Mental Health Foundation Research Associate, Queen's Uni-
versity, Kingston, Ontario
J. Inglis MA, PH D, is Professor, Department of Psychology, Queen's Univer-
sity, Kingston, Ontario
J. Knowles BSC, PH D, is Associate Professor, Department of Psychology,
Queen's University, Kingston, Ontario
V.A. Kral MD, is Director, Gerontological Unit, Allan Memorial Institute of
Psychiatry, and Associate Professor of Psychiatry, McGill University,
Montreal, PQ
S.G. Laverty MD, is Director, Addiction Studies Unit, Kingston Psychiatric
Hospital, and Professor, Department of Psychiatry, Queen's University,
Kingston, Ontario
A. McGhie MA, PH D, is Professor, Department of Psychology, Queen's Uni-
versity, Kingston, Ontario
A.W. Maclean PH D, is Assistant Professor, Department of Psychology,
Queen's University, Kingston, Ontario
J. Rae burn MA, PH D, is Lecturer, Department of Psychiatry, University of
Auckland, Auckland, New Zealand
M.D . Suhoski MA, PH D, is Associate Professor, Department of Psychology,
Queen's University, Kingston, Ontario
 V. A. KRAL MD

4 The Organic Amnesias

For a long time research into the pathology of remembering and the nature of
the memorial process in general has been carried out mainly by clinicians,
psychiatrists, neurologists, and clinical psychologists. More recently, animal
experiments on memory function and learning are being pursued in psycho-
physiological and pharmacological laboratories in various centres. Although
these studies have led to interesting hypotheses about the nature of the
memorial process (John 1967), they are, at least at the present stage of
knowledge , only to a limited extent applicable to human psychopathology,
mainly because the results of learning experiments achieved by conditioning
techniques in lower mammals cannot easily be equated with the encoding and
retrieval of memory traces of conscious experiences in man. Moreover,
clinical observation brings into focus certain facts which by their very nature
cannot be studied in animal experiments, but which are nevertheless im-
portant for our understanding not only of the pathology of memory in the
human , but also of the process of remembering itself. The clinician has to
agree with Whitty and Zangwill (1966) who state: 'Far greater heed should be
paid to the findings of clinical inquiry in formulating general theories of
memory.' This presentation deals with some 'findings of clinical inquiry'
which seem to be important for the understanding of memory function in the
human.
There is first of all the close relationship of memory function to the level
of consciousness. The unconscious patient is unable to recall, once conscious-
ness is restored, what had happened during the time he had been in coma.
The patient is unable to recall not only external events, but also what had
happened to himself during that time, including the most painful stimuli or
therapeutic procedures. This anterograde amnesia is permanent, does not
shrink and does not clear up, even with the help of cues. This is the more
remarkable as EEG studies have demonstrated that evoked potentials in the
relevant cortical areas can be produced in unconscious patients (Arfel, Albe-
Fessard, and Walter 1968). Moreover, recent experimental studies have shown
70 A Triune Concept of the Brain and Behaviour

that subliminal stimuli not consciously perceived can exert a considerable

influence on the subject's behaviour, at least as far as the association process is
concerned, although they cannot be recalled in memory (Shevrin and Fritzler
1968).
In other words, only those events, perceptions, and stimuli which previous-
ly were experienced consciously, be it only in the vaguest manner and for the
shortest time, can later be recalled. This does not mean that all perceptions
and events which we experience consciously will , or even can, later be re-
called. Only a very small proportion of the perceptions and events of every-
day life are later recalled, partly because there never arises the need to
retrieve them and partly because they did not have any 'personal meaning' for
the subject at the time of encoding (Cameron 1947 ; Kral 1956).
In any case it seems safe to state that the neurophysiological mechanisms
which are involved in arousal and in maintenance of consciousness are of
primary importance in remembering.
This assumption is well supported by clinico-anatomical experience . It is
now generally accepted that lesions involving the reticular activating system
in the brain stem may lead to loss of consciousness. With short-lasting and
reversible interference with brain stem function, as in most cases of brain
concussion (Kral 1935) or with epileptic discharge (Penfield 1938), loss of
consciousness is accompanied by a circumscribed amnesia for the time of
unconsciousness, sometimes preceded by a retrograde amnesia.
In cases where a patient gradually recovers from a more severe but still
reversible lesion such as a severe brain concussion, or a direct brain stem
trauma (Kral 1934) or a strangulation (Kral and Gamper 1933), a phase of
unconsciousness of a few hours' or days' duration is followed by a clouded
state which may last a few days or weeks until full consciousness is restored .
Systematic investigation of memory function in such cases shows that the
period of unconsciousness is covered by complete anterograde amnesia which
does not clear up . The clouded state is, as far as memory is concerned,
characterized by an amnestic syndrome, namely impaired immediate recall,
inability to recall recent events, disorientation, and sometimes confabula-
tions. The amnestic syndrome gradually clears up, usually in parallel to the
clearing of consciousness, until normal memory function is restored, although
in some cases it may outlast the confusional state for some time. There is,
therefore, in such cases a continuous gradual restitution of memory function
from complete amnesia through an amnestic syndrome to normal memory
function.
The retrograde amnesia observed in many cases covers a variable time
period preceding the onset of the lesion . It consists of two parts: a more
extensive part which usually clears up with or without the help of cues, and a
The Organic Amnesias 71

short-lasting part immediately preceding the onset of the lesion which does
not clear up. The former seems to be due to psychological factors such as the
sudden interruption in the course of meaningful events and the lack of cues in
the contents of thought and behaviour, as well as dynamic factors in the
patient's past history (Klein and Kral 1933). The short part immediately
preceding the lesion seems to be due to the sudden interference of the lesion
with the still labile phase of memory trace formation for the events im-
mediately preceding the onset of the lesion.
With progressive lesions impinging on the midbrain and posterior hypo-
thalamus, such as tumours of the third ventricle, the pineal gland, or cranio-
pharyngiomas, the amnestic syndrome is the first psychiatric disorder to
appear to be followed by clouding of consciousness, and eventually coma and
death. Successful removal of the tumour, as in some cases reported by
Foerster and Gagel (1934) , and Williams and Pennybaker (1954), may inter-
rupt this sequence at any stage. The amnestic syndrome may improve or
disappear completely when no irreversible damage has as yet occurred before
operation .
These observations suggest firstly that the memory impairment in cases
with brain stem lesions can be graded along a scale from non-storage, possibly
due to non-encoding, via a transitional stage of labile short-term storage to
the storage of consolidated permanent engrams, and secondly that the bio-
chemical and electrophysiological processes involved are closely related,
perhaps identical, with the ones underlying arousal and maintenance of the
level of consciousness.
In this connection it should be noted that work by Jasper and co-workers
(Jasper, Khan, and Elliott 1965; Jasper and Koyama 1967) seems to indicate
that acetylcholine and glutamic acid are released at a higher rate from the
cortical surface by arousal and reticular activation. Glutamic acid was released
at a reduced level after upper midbrain transection as compared to normal
sleep, and this again was lower than in the aroused animal. On the other hand,
GABA was released at a higher rate after midbrain coagulation . Furthermore,
Krnjevic (I 965) and co-workers (Kmjevic and Schwartz, 1968) drew atten-
tion to the inhibitory effect of GABA and the excitatory effect of acetyl-
choline on cortical neurons.
It should be mentioned, however, that Deutsch (1966), based on his
animal experiments with di-isopropylfluorophosphate, an anticholinesterase,
arrives at the conclusion that acetylcholine is necessary, not for the storage
of memory, but for its retrieval or accessibility. However this may be, it
would appear safe to say that the understanding of the chemistry of
memory presupposes an understanding of the chemistry of coma and
consciousness.
72 A Triune Concept of the Brain and Behaviour

Another important finding which gradually emerged from clinical and

anatomical investigation and which has by now become a well established
part of our knowledge is the fact that an amnestic syndrome, clinically not
distinguishable from the one which is found with brain stem lesions, can also
be found with bilateral lesions of the various parts of Papez' harmonious
mechanism of emotions. Papez proposed in 1937 that the ensemble of the
following structures: hypothalamus, cingulate gyms, hippocampus, and their
interconnections represent the anatomical basis of the emotions. In this en-
semble the mammillary body bears the main hypothalamic relations to the
cortex. Confirmatory evidence has since been accumulated by clinical ob-
servations, as well as by experimental studies by Cobb (I 950), Fulton (I 951 ),
and particularly by MacLean (1949).
In this connection it is interesting to note that Gamper in 1928 published
his findings in the brains of 16 young patients who had suffered from al-
coholic Korsakofrs psychosis and had died shortly after the onset of the
disease. The pathological changes found in these brains were qualitatively
identical with those usually found in Wernicke's disease. They were found in
the brain stem, in an area extending from the dorsal vagal nucleus cephalad to
the anterior commissure and involved the grey matter of the anterior part of
the floor of the fourth ventricle, the nuclei of the posterior colliculi, the
ventromedial part of the central grey matter around the aqueduct, the medial
part of the third nucleus, the interstitial and Darkschewitch nuclei, the
mammillary bodies, the tuber cinereum, the supraoptic nucleus, the medial
periventricular nuclei of the third ventricle and the medial nuclei of the
thalamus .
In all cases the mammilary bodies showed the maximum changes. They
were found severely affected even in cases where the other structures remain-
ed free, and they showed the maximum atrophy in the older cases. The
cerebral cortex and the basal ganglia, on the other hand, were found free of
the characteristic lesions.
Camper's neuropathological findings have been since confirmed by a
number of authors in alcoholic Korsakoffs syndrome, as well as in patients
who had shown clinically an amnestic syndrome on an etiological basis other
than chronic alcoholism, such as carcinoma of the stomach with metastases in
the liver (Neubuerger 1929) or lymphogranulomatosis (Kant 1933). The
lesions found in these cases were qualitatively, and also with respect to
localization, identical with those observed in alcoholic Korsakoff patients,
with the maximum changes in the mammillary bodies.
Furthermore, the same amnestic syndrome has been found with lesions of
other parts of Papez' harmonious mechanism. One of the first observations is
that of Glees and Griffith ( 1952), who described the case of a woman who, at
The Organic Amnesias 73

the age of 58, developed complete lack of emotional drive and apathy
together with a severe amnestic syndrome.
The patient slipped during the following years into an automatic and
vegetative mode of life . At autopsy the medial part and tip of the temporal
lobe were found to be bilaterally missing. The hippocampus, hippocampal
gyrus, the gyrus lingualis and fusiformis were replaced by a cyst, the fornices
atrophic. The mammillary bodies did not appear grossly abnormal. Further-
more, Penfield and Milner (1958) and Scoville and Milner (1957) demon-
strated that bilateral ablation of the hippocampal area produced impairment
of immediate recall and loss of recent memories with its behavioural
consequences.
Other structures of Papez' circuit, lesions of which can lead to a severe
amnestic syndrome usually combined with deep apathy, are the anterior
thalamus, as in the cases reported by Stern (1939) and Benedek and Juba
(I 940), and in the cingulate gyrus, as reported by Nielsen (I 951 ).
The disease entity, where the amnestic syndrome is most frequently found
is, of course, senile dementia, and an extensive literature exists about the
clinical and psychological aspects of the memory impairment in this condi-
tion (Inglis, Ankus, and Sykes 1968; Wigdor and Kral 1961). Recent
anatomical studies by Corsellis (1967) have shown that senile plaques are
most frequent in the hippocampus and hippocampal gyrus, whereas Ishii
(1966) demonstrated that Alzheimer's neurofibrillary changes are to be
found in marked predilection in certain groups of hypothalamic nuclei, the
nucleus mammillo-infundibularis among them .
Based on these clinical observations we arrive at the conclusion that bi-
lateral lesions of the hippocampus and functionally related structures of
Papez' harmonious mechanism which form the limbic system may lead, as far
as memory function is concerned, to the same clinical picture as lesions of the
posterior hypothalamus. Namely, an amnestic syndrome with the same type
of impairment of immediate recall , in other words, the same unstable, labile,
short-term encoding which is not followed by the formation of permanent
engrams.
One explanation of this interesting finding could be sought in the assump-
tion that, in cases where the amnestic syndrome is found with lesions of the
brain stem, it is actually due to the involvement of the hypothalamic part of
the limbic system and particularly of the mammillary bodies.
Another assumption, more in keeping with clinical observation, would be
that two qualitatively different processes are necessary in order to complete
consolidation of the memory trace, one dependent on the integrity of the reti-
cular activating system and another one dependent on the integrity of the limbic
system . Psychological considerations seem to favour the second hypothesis.
74 A Triune Concept of the Brain and Behaviour

While the close relationship of amnesia and the reversible amnestic syn-
drome to loss and impairment of consciousness in the clinical course of many
cases with brain lesions indicate that the memory impairment is the direct
consequence of impaired consciousness, more specifically, of impaired per-
ceptual awareness, clinical observation and psychological studies (Talland
1965) show that the patient suffering from a chronic amnestic syndrome due
to alcoholism or senile dementia is well capable of integrated perceptions,
originating both from the environment and from his own body. He can read
and write when asked to and is able to follow a conversation for a short time.
He is able to perform purposeful and goal-directed movements. In other
words, perceptual awareness and a high degree of integrated behaviour is
present in cases with chronic amnestic syndrome. The same holds true for
those cases where the amnestic syndrome develops out of a condition of un-
consciousness or a clouded state once the acute state has passed.
In a previous analysis of cases with an amnestic syndrome of traumatic ori-
gin it was found that the patients behaved as if events, even if they concerned
them directly, did not affect them (Klein and Kral 1933). They did not be-
come a problem to them and the patients experienced them in a passive and
apersonal manner. The events seemed to have lost any personal meaning for
them. It is these 'meaningless' events which the patients are unable to recall.
There are, however, certain events in the recent past to which the patient
with a chronic amnestic syndrome repeatedly refers, although frequently in a
distorted manner. These events have one feature in common. They have a
'personal meaning' for the patient, of which other events apparently are
deprived. It is this 'personal meaning' which makes an 'experience' out of an
event and it obviously depends on the balance between the severity of the
amnestic syndrome on the one hand, and the 'meaning' of the experience on
the other hand, whether the latter will be remembered or not (Kral 1956).
Impairment or loss of awareness of personal meaning is characteristic of
the passive and impersonal way in which these patients experience events and
is, apparently, the essential psychopathological deficit underlying the
amnestic syndrome in such cases (Kral 1959).
Very few events in adult life derive their personal meaning solely from the
strength of the sensory perceptions of the actual situation and the emotional
response to them. In most instances the personal meaning of an event is based
on the memories of previous similar situations with their varied affective
components which serve as a conceptual scheme for recognition and recall.
The individual's attitude at the time the event takes place is another im-
portant factor as to whether it becomes meaningful for a person or remains
deprived of meaning, like so many events of our daily life. Although in most
instances the individual may not realize why a particular event becomes
The Organic Amnesias 75

meaningful to him, a reaction is initiated which starts with more or less

automatic focusing of attention to the particular event and eventually leads
to its integrated registration as a personal memory of an experience available
for future recall.
From our clinical experiences it would appear that the chronic amnestic
syndrome, as found with lesions of the various parts of the limbic system, is
due to the impairment of the capacity to integrate the sensory perceptions
and the affective components of a given situation with similar previous ones
to a personal, meaningful experience.
Awareness of personal meaning is based on perceptual awareness as one of
its prerequisites. Impaired perceptual awareness is, therefore, accompanied by
impairment of awareness of personal meaning, whereas the latter impairment
may be found without the former.
This assumption is borne out by clinical facts. The essential features of the
amnestic syndrome, impaired immediate recall and loss of recent personal
memories, are always found in conditions of impaired perceptual awareness
like clouded and delirious states. They may be hidden behind the more im-
pressive psychopathological manifestations so often found with impaired per-
ceptual awareness like illusions, hallucinations, and delirious behaviour, but
they can always be demonstrated on psychiatric examination.
On the other hand lesions which disrupt the integration of the affective
components with the perceptual components of an event will interfere with
the awareness of personal meaning and will lead, as far as memory is con-
cerned, to an amnestic syndrome, even if perceptual awareness remains un-
impaired. Integrity of the limbic system in addition to a normally functioning
arousal mechanism seems to be necessary for the ·consolidation of permanent
engrams out of the short-term memory traces.
The types of memory dysfunction so far discussed have considered the
inability to recall experiences of the person as a whole (Kral and Durost
1953), in this paper called personal memories. But clinical observation shows
that there exists still another type of memory impairment independent of,
and often without, any impairment of personal memories, an inability to make
use of previous memories in certain areas of functioning, the partial or limited
amnesias. This type of amnesia is found in, and actually forms the basis of,
agnosia, aphasia, apraxia, and related disorders.
Meyra t (I 944) devoted a special study to what he called special memories.
Based on the assumption that amnestic functions are part of the functions that
they sustain and that amnestic disorders are but a combination of disturbances
of partial memories, he described special test procedures for their investigation.
Kral and Durost (1953) studied a number of patients with amnestic syn-
dromes of various etiology by means of these tests and found that the loss of
76 A Triune Concept of the Brain and Behaviour

special memories was essentially independent of the amnestic syndrome and

that among the disease entities investigated, only a certain group, namely
patients with senile dementia, showed such a loss of special memories in
various functional areas. Since then, these studies have been extended to
include other disease entities with diffuse cortical involvement, for example,
Alzheimer's disease and cerebral arteriosclerosis. These patients also showed a
loss of special memories in certain functional areas.
There was no uniformity between Meyrat's different subtests; for example,
a patient would reach a full score on repeated verbal memory as measured
with Meyrat's test, but would score zero for recall of a written word, and vice
versa. There was, furthermore , no patient in this group who could not recall
at least once, that is after 10 seconds, a painful stimulus and even locate it,
and some of our senile patients achieved a full score on our pain subtest. No
practice effect was recognizable with these different subtests, although they
were repeated after short intervals.
In general, it was found that partial amnesias are to be observed in patients
with lesions of the neocortex, either localized as in the aphasias and related
disorders, or diffuse as in senile dementia or Alzheimer's disease .
An important difference between the partial amnesias and the amnesia for
personal memories, as found in the amnestic syndrome, is that in the former
old memory traces are not available, whereas in the latter recent memories
cannot be recalled. One could even state that, at least from the clinical point
of view, the partial amnesias are the only example of real loss of previously
encoded material . However, the loss of old memory traces in these conditions
also interferes with the assimilation of new material in the relevant functional
areas, as exemplified in the results of Mey rat's test.
Another type of memory dysfunction discovered by clinical inquiry is the
inability to recall already encoded and retained material, as exemplified by
the 'benign type of senescent forgetfulness' (Kral 1962). This is characterized
by the inability of the subject to recall certain parts of an experience like a
name, a date, or a place, whereas the experience as such can be recalled. The
same data which cannot be recalled on one occasion are available for recall at
another time, an important difference from the senile amnestic syndrome.
This seems to indicate that the 'forgotten' parts of an experience were both
registered and retained, and that the functional deficit is essentially one of
recall.
The reasons why a name or a date cannot be recalled on certain occasions are
manifold : the length of time which has elapsed since the original experience or
since the last time the date was recalled; the frequency of previous recalls;
changes which take place in the registered and retained material during the
period of retention; dynamic factors operative at the time of recall.
The Organic Amnesias 77

Extensive longitudinal studies of the senescent population have shown that

the benign type of senescent forgetfulness is not statistically related to
shortened survival time and increased death rate as is the malignant type of
senescent forgetfulness, the senile amnestic syndrome (Kral 1966).
However, the fact that so many more aging than young people have diffi-
culties in recall of registered and retained material seems to indicate that this
type of amnesia in the aged has an organic basis in that the association
process which underlies the recall of names and dates becomes more
vulnerable in this age-group.
On the basis of the observations described above, a hypothesis of
memory function can be developed which does not seem to be contra-
dicted by clinical facts and at the same time allows for further elaboration
by electrophysiological and biochemical studies. It seems safe to assume
that both the reticular activating system in the brain stem and the limbic
system are necessary for memory function, the former to alert the cortex
for the many perceptual components which simultaneously or in short se-
quence are distributed over its various areas according to their functioning,
and the latter for the integration of these components into an 'experience.'
Interference with the functional activity of the cortex as in coma and
clouded states would lead to interference with encoding, because the per-
ceptual components of an experience which may reach the cortex via the
specific projection system arrive at a time when the area is completely or
partially unresponsive to conscious experience and, as clinical observation
shows, to the formation of any, or at least long-term, structural engrams.
Interference with the limbic system would lead to loss of the capacity to
integrate the partial components and, therefore, to form memory traces of
the new experience as a whole, but not of partial memories. These are regis-
tered and retained in the various parts of the cortex, whether integrated or
not. Once integrated, the experience as a whole can be evoked from the
various cortical areas.
This working hypothesis, formed on a basis of clinical experience, is sup-
ported by findings of the neurophysiologists and neurochemists. It would
seem that a certain state of electrophysiological activity is necessary for the
various cortical regions to register perceptual stimuli, probably due to the
activity of certain, now intensively investigated, enzyme systems. Interference
with this activity by lesions of the reticuloactivating system interferes with
perception as well as with encoding. Only the fully active and alert cortex
seems to be able to form permanent engrams which, however, still need to be
integrated by the hippocampus-fornix-mammillary system, to serve as per-
sonal memories of an experience. The details of the interaction of these two
systems have not yet been clarified.
78 A Triune Concept of the Brain and Behaviour

Recall may originate from any part of the cortex according to the nature
and perceptual quality of the stimulus presented and may then evoke the
integrated engram of the experience. It may be interfered with by the loss of
permanent engrams as in the partial amnesias with cortical disease, or by
interference with the process of association, as observed in the benign type of
senescent forgetfulness.
From the previous consideration it would appear that clinical research still
has something to offer in furthering our understanding of the memorial pro-
cess. We are dealing with experiments of nature which, of course, cannot be
so carefully predesigned as the experiments in the laboratory. But the ex-
periments we are dealing with allow for the observation of a very important
part of human functioning, namely conscious memory function, which ex-
perimental and biochemical studies on animals do not permit. In spite of the
progress made in experimental studies, clinicians still have the obligation to
pursue their own observations and to draw their own conclusions which, of
course, have to be correlated with those of the experimentalists.

SUMMARY

Three organic types of memory dysfunction are discussed: the amnestic syn-
drome, the partial amnesias, and the benign type of senescent forgetfulness.
From the clinical point of view, these three types of organic amnesia seem to
represent disturbances of the three phases of the process of remembering. The
amnestic syndrome, which can be due to lesions of either the reticular ac-
tivating system in the brain stem or of different parts of the limbic system, is
an expression of impaired encoding leading only to short-term retention of
the memory trace. The partial amnesias represent loss of stored memories,
while the benign senescent forgetfulness is due to impaired recall, probably
due to a slowing down of the association process in the aged.
A hypothetical formulation of the memorial process, as seen from the
point of view of the clinician, is presented.

REFERENCES

Arfel, G., D. Albe-Fessard, and S. Walter, Evoked potentials in coma. Electroencephalog.

CTin. Neurophysiol., 1968, 25 :93-4
Benedek, L. and A. Juba, Weitere Beitraege zur Frage des anatomischen Substrates des Korsa-
kowschen Symptomenkomplexes. Arch. Psychiat. Nervenkrankh, 1940, 112: 505-16.
The Organic Amnesias 79

Cameron, D.E., Remembering. (Neivous and Mental Disease Monographs, no. 7 2.) New
York: Williams and Wilkins, 194 7
Cobb, S., Emotions and Qinical Medicine. New York: Norton, 1950
Corsellis, J ., The relationship between quantitative measures of dementia and de-
generative changes in the cerebral grey matter of elderly subjects. Proc. Roy. Soc.
Med., 1967, 60:259-60
Deutsch, J.A., Substrates of learning and memory. Diseases Nervous System, 1966, 27,
Sect. 2, Monograph suppl.: 20-4
Foerster, 0 . and 0. Gagel, Ein Fall von Ependymcyste des III. Ventrikels. Z. Ges.
Neurol. Psychiat., 1933, 149:134-75
Fulton, J .F ., Frontal Lobotomy and Affective Behavior; A Neurophysiological Analysis.
New York: Norton, 1951
Gamper, E., Zur Frage der Polioencephalitis haemorrhagica der chronischen Alkoholiker.
Deut. Z. Nervenkrankh., 1928, 102: 122-9
Glees, P. and H.B . Griffith, Bilateral destruction of hippocampus (cornu Ammonis) in
case of dementia. Monatsschr. Psychiat. Neurol., 1952, 123: 193-204
Inglis, J ., M.N. Ankus, and D.H. Sykes, Age-related differences in learning and short-
term memory from childhood to the senium. Human Development, 1968,
11:42-52
Ishii, T., Distribution of Alzheimer's neurofibrillary changes in brain stem and hypo-
thalamus of senile dementia. Acta Neuropathol., 1966, 6: 181-7
Jasper, H.H., R.T. Kahn, and K.A.C. Elliott, Amino-acids released from the cerebral
cortex in relation to its state of activation. Science, 1965, 147: 1448-9
Jasper, H.H. and I. Koyama, Rate of release of acetylcholine and glutamic acid from the
cortex during reticular activation. Fed. Proc., 1967, 26:373
John, E.R ., Mechanisms of Memory. New York: Academic Press, 1967
Kant, F ., Die Pseudoencephalitis Wernicke der Alkoholiker. Arch. Psychiat. Nerven-
krankh., 1933, 98:702-68
Klein, R. and A. Kral, Zur Frage der Pathogenese und Psychopathologie des amnes-
tischen Symptomenkomplexes nach Schaedeltraumen. Z. Ges. Neurol. Psychiat.,
1933, 149:134-75
Kral, A., Ein weiterer Beitrag zur Frage der Pathogenese und Pathophysiologie der Com-
motionspsychose. Arch. Psychiat. Nervenkrankh., 1934, 101 : 729-38
Kral, A., Zur Pathophysiologie der Commotio cerebri. Med. Klin. (Munich) , 1935,
31 :876-910
Kral, V.A., The amnestic syndrome. Monatsschr. Psychiat. Neurol., 1956, 132:65-80
Kral, V.A., Amnesia and the amnestic syndrome. Can. Psychiat. Assoc. J., 1959, 4:61-8
Kral, V.A., Senescent forgetfulness: benign and malignant. Can. Med. Assoc. J., 1962,
86: 257-60
Kral, V.A., Memory loss in the aged. Diseases Nervous System, 1966, 27, Sect. 2, Mono-
graphic suppl. :51-4
Kral, V .A. and H.B. Durost, A comparative study of the amnestic syndrome in various
organic conditions. Amer. J. Psychiat., 1953, 110:41-7
Kral, V.A. and E. Gamper, Koerperdrehungen um die Laengsachse bei einem wieder-
belebten Erhaengten. Monatsschr. Psychiat. Neurol., 1933, 84 :309-23
Krnjevic, K., Cholinergic inneivation of the cerebral cortex. In D.R. Curtis and
A.K . McIntyre (eds.), Studies in Physiology. New York: Springer, 1965. Pp. 144-51
Kmjevic, K. and S. Schwartz, in The Inhibitory Transmitter in the Cerebral Cortex in
Structure and Function of the Inhibitory Neuronal Mechanisms. Oxford and New
York: Pergamon, 1968. Pp. 419-27
Maclean, P.D., Psychosomatic disease and the 'visceral brain'; recent developments bear-
ing on the Papez theory of emotion. Psychosomat. Med., 1949, 11: 338-5 3
Meyrat, G., Recherche sur les troubles de la memoire. Monatsschr. Psychiat. Neurol.,
1944, 109 :36-48
80 A Triune Concept of the Brain and Behaviour

Neubuerger, K. , Ueber Hirnveraenderungen nach Alkoholm issbrauch. Z. Ges. Neurol.

Psychiat., 19·29, 135 :159-209
Nielsen, J.M., Anterior cingulate gyrus and corpus callosum. Bull. Los Angeles Neurol.
Soc., 1951, 16 :235-43
Papez, J.W ., A proposed mechanism of emotion. Arch. Neurol. Psychiat., 1937,
38:725-43
Penfield, W., The cerebral cortex in man. Ibid., 1938, 40:417-42
Penfield, W. and B. Milner, Memory deficit produced by bilateral lesions in the hippo-
campal zone. AMA Arch. Neurol. Psychiat., 1958, 79:4 75-97
Scoville, W.B. and B. Milner, Loss of memory after bilateral h ippocampal lesions.
J. Neurol. Neurosurg. Psychiat., 1957, 20: 11-21
Shevrin, H., and D.E. Fritzler, Visual evoked response correlates of unconscious mental
processes. Science, 1968, 161 :295-8
Stern, K., Severe dementia associated with bilateral symmetrical degeneration of the
thalamus.Brain, 1939,62:157-71
Talland, G.A., Deranged Memory; a Psychonomic Study of the Amnesic Syndrome. New
York: Academic Press, 1965
Whitty , C.W.M . and O.L. Zangwill (eds.), Amnesia. London: Butterworths, 1966
Wigdor, B. and V.A. Kral , Senescent memory function as an indicator of the general
preservation of the aging human organism .World Congress of Psychiatry, 3rd,
Montreal, 1961; proceedings. 1961 , vol. 1:682-6
Williams, M. and J . Pennybacker, Memory disturbance in third ventricle tumors.
J. Neurol. Neurosurg. Psychiat., 1954, 17 : 115-23
 MILTON D. SUBOSKI

5 Recent Developments in Memory

Consolidation Theory 1

According to memory consolidation theory, memory becomes increasingly

resistant to interference with the passage of time following learning, pre-
sumably as memory is gradually transferred from temporary to permanent
storage. Modern interest in consolidation theory was stimulated by Hebb's
(1949) formulation of a neurophysiological theory of the memory trace. In
Hebb's theory, the production of a permanent memory trace was assumed to
involve a two-phase mechanism. Sensory input to the central nervous system
initiated temporary reverberatory electrical activity in a set of neurons, ac-
companied by a gradual and permanent structural change. The massive elec-
trical interference produced by electroconvulsive shock presumably would
disrupt the reverberatory circuit and halt the consolidation process. This
theory, along with the extensive reports of memory disturbance resulting
from electroconvulsive therapy, has led to frequent attempts to subject con-
solidation theory to rigorous experimental test through the use of electro-
convulsive shock (Ecs).
The literature on this problem is marked by considerable controversy gene-
rated primarily by the difficulties involved in the identification of memory in
animals. The controversy does not concern the existence of a memory con-
solidation process. Hardly anyone would seriously dispute that memory stor-
age requires a structural change in the central nervous system and that such a
change requires time. Rather, the controversy concerns the temporal duration
of the consolidation process and whether Ecs affects memory consolidation.
Beginning with Duncan in 1949, many experiments showed that an ECS
given after each trial in a learning task interfered with the acquisition of the
learned response. Furthermore, the longer the time interval between the
learning trial and the ECS, the less the Ecs interfered with learning. The data
from these experiments were interpreted as indicating that ECS disrupted a

The research by the author and his students described in this paper was supported
primarily by the Ontario Mental Health Foundation, Grant 161
82 A Triune Concept of the Brain and Behaviour

memory consolidation process that took an hour to complete. In 1960 it was

shown by Coons and Miller that another interpretation of Duncan's experi-
ment was possible. Coons and Miller showed that with repeated presentations
ECS becomes aversive. In Duncan's experiment, rats were failing to learn the
response not because they forgot the learning trial but rather because they
remembered that performing the learned response was followed by ECS.
This challenge to memory consolidation theory was met by the develop-
ment of the one-trial passive avoidance paradigm. In the passive avoidance
paradigm a highly probable response is followed by a single punishing foot
shock. Upon subsequent test, the rat presumably 'passively' avoids that foot
shock by showing increased latency or decreased probability of making the
punished response. Many experiments have shown that a single ECS interferes
with the acquisition of a one-trial passive avoidance response. Furthermore,
beginning with Heriot and Coleman in 1962, many experiments have shown
that the amount of interference is a function of the interval between learning
and the ECS. An ECS given immediately after foot shock eliminates the effect
of the foot shock. The rat repeats the response as readily as if he had never
been shocked for making it. An ECS given an hour or more after learning has
no effect. The rat shows as much passive avoidance as if he had not received
the ECS. Intermediate learning-Ecs intervals produced intermediate amounts
of learning. The Ecs-produced passive avoidance gradients have constituted
powerful support for the hypothesis that ECS disrupts a memory con-
solidation process requiring an hour or more to complete.
However, within the last few years the adequacy of the passive avoidance
experiments has come into question. The problem, as was mentioned earlier,
is knowing what measures memory in rats. The question is: what is learned in
one-trial passive avoidance paradigms? A rat may fail to repeat a punished
response because he has learned in one trial that the response leads to punish-
ment, or he may learn nothing at all about the connection between the
response and the punishment but nevertheless fail to make the response
because he now crouches and freezes in the apparatus. To phrase it simply,
ECS may affect performance and not memory. The rat may repeat the
previously punished response after ECS not because he does not remember the
foot shock but rather because he no longer freezes in the presence of the
apparatus cues. A related fact that at the very least introduces equivocation
into the interpretation of passive avoidance experiments is the discovery
within the last several years of passive avoidance incubation gradients. Pinel
and Cooper (1966) conducted a one-trial passive avoidance experiment that
included the standard groups that received ECS at different time intervals
following learning. These groups were tested for retention of the passive
avoidance response 24 hours after receiving ECS in the standard way. They
Memory Consolidaton Theory 83

Plate 1 Top view of the open field, two chamber apparatus used to train a one-trial
discriminated avoidance response.

also included groups that were returned to the apparatus and tested for
retention at the same postlearning intervals at which the standard groups
received ECS. Pinel and Cooper found that the passive avoidance response
incubated, that is, changed in strength over time following learning. The
groups tested very quickly after passive avoidance training repeated the just-
punished response quite readily. However, the groups tested after an hour did
not make the punished response. Furthermore, the incubation gradient and
the Ecs-produced gradient were virtually identical. These, and other similar
data, show that at least two interpretations of the retrograde effect of Ecs on
passive avoidance learning are possible, and only one of them is memory
consolidation disruption.
One technique for obtaining a relatively unequivocal indicator of memory
is to use correct choices in a discriminated avoidance learning task. Correct
choices can only be made on the basis of memory as to which choice is
correct. We have developed several one-trial discriminated avoidance learning
tasks and investigated the effects of a single ECS given at different time
intervals following discriminated avoidance learning. The apparatus used in
one such learning task is shown in Plate 1. The triangular open field is lighted
84 A Triune Concept of the Brain and Behaviour

by the flood lamp. The two small chambers are distinctive in colour and the
shape of the holes leading into them. A rat placed in the open field will
quickly enter one of the small chambers and remain there in preference to the
lighted field. After a number of trials in which the rat is familiarized with
both chambers, the grid floor is electrified in whichever small chamber that
the rat enters. When tested 24 hours later, about 90 per cent of the rats will
not enter the chamber where they had been shocked but rather enter the
other chamber. We (Suhoski, Spevack, Litner, and Beaumaster 1969) have
conducted three separate experiments in which the interval between the time
when the rat first encounters the shock and the time when he receives an ECS
was varied from just over one minute to just under one hour. The results can
be summarized quite simply. A smaller proportion of rats make correct
choices if an Ecs follows learning than if ECS is not given, but the pro-
portion of correct choices does not depend upon the interval between learning
and Ecs.
We (Suhoski, Black, Litner, Greenner, and Spevack 1969) used a different
type of one-trial discriminated avoidance training task in another series of
experiments. Rats were trained to bar press for food reward in a conditioning
box in which the bar and the food chamber were located on opposite walls.
The wall containing the food chamber was constructed of translucent plastic.
After the rats were bar pressing and retrieving the food pellets readily, a light
outside the box was turned on when the rat made a preselected bar press.
Then, as the rat retrieved the food pellet, the grid floor of the box was briefly
electrified and the light went out. The. next day the rat was returned to the
box and left there until he began to bar press and retrieve the food pellets.
When the rat was again bar pressing readily, he was tested by having the light
turned on for half of the bar presses. The single exposure to the light-shock
combination was sufficient to produce good discriminated avoidance. The
rats retrieved food pellets quite readily on trials when the light did not come
on but avoided the food chamber when the wall was illuminated. We con-
ducted two experiments in which a single ECS was delivered after the light-
shock combination at intervals varying from 100 sec to about one hour for
different groups of rats. The results were essentially the same as in the earlier
experiments. The single ECS produced poorer discrimination, but the amount
of interference with discrimination did not depend on the learning-Ecs
interval.
In a third experiment in the same apparatus, we examined the effect of an
ECS given at short learning-Ecs intervals. We found that at intervals between 0
and 100 sec, ECS produced temporally graded interference with discriminated
avoidance. If the Ecs were delivered either 0 or 5 sec after foot shock, the Ss
gave no indication of having learned the discrimination. An ECS delivered 20
Memory Consolidation Theory 85

sec after learning did not eliminate discrimination, but did produce poorer
discrimination than an ECS delivered I 00 sec after learning. Pfingst and King
(1968) have reported results for one-trial discriminated avoidance that are in
essential agreement with ours. They found temporally graded interference
with discrimination by ECS delivered at postlearning intervals of less than 30
sec and temporally ungraded effects for longer intervals. Thus, the results of
experiments in which ECS was delivered following one-trial discriminated
avoidance learning indicate that Ecs produces temporally graded interference
with discrimination but only for 20 or 30 sec after learning. These results can
be interpreted as ECS disruption of a memory consolidation process that is
complete in less than a minute.
In contrast to the one-trial discriminated avoidance results are the results
of a number of experiments in which a single ECS was delivered following
multi-trial discriminated avoidance training. A number of such studies can be
found, the earliest of which is by Thompson and Dean (1955), and all showed
a temporally graded effect of ECS on the number of trials or errors in relearning
the discrimination. In all of the experiments the temporally graded effect of
ECS extended well past the 20 or 30 sec postlearning interval during which
ECS produces temporally graded interference with one-trial discriminated
avoidance. Since these studies seem to indicate that the amount of learning
retained is an increasing function of the learning-Ecs interval for up to an
hour after learning, they are consistent with the hypothesis that Ecs disrupts
long-term memory consolidation.
Thus, the question becomes whether to believe the one-trial or the multi-
trial discriminated avoidance results. One way to decide might be to look at
possible involvement of incubation processes in the two types of discrimina-
ted avoidance. In our laboratory, Black (1969) found no evidence in the open
field, two chamber apparatus for changes in discriminated avoidance as a
function of time following learning except at very long intervals where
discrimination drops off slightly. In multi-trial discriminated avoidance,
however, he found incubation gradients that were virtually identical to the
ECS•produced gradients. Why multi-trial discriminated avoidance improves as
a function of time following initial learning is unclear. It appears, however,
that when the rats are tested at the longer interval they tend to freeze in the
apparatus and thus have a longer time in which to make their choice.
One other set of experiments has yielded results worthy of at least brief
mention. A number of researchers have administered ECS at different time
intervals following one-trial appetitively motivated learning. The usual pro-
cedure is to place a thirsty rat in a box containing an alcove with water
available. After the rat has found and begun to drink the water, he is removed
and returned for test at some later time. The single exposure to water results
86 A Triune Concept of the Brain and Behaviour

in a shorter latency and higher frequency of alcove entrances. An experiment

recently completed by Pinel (1969) yielded typical results. A single ECS pro-
duced temporally graded interference with the appetitively motivated
response only if it followed learning by a brief time interval. Incidently, Pinel
included the appropriate control groups and found no evidence that the
response changed as a function of time following learning.
To summarize briefly, ECS appears to have four different effects on
responses learned prior to its administration. First, ECS produces an apparent
loss of memory which is not selective upon the age of the memory trace.
Second, when delivered repetitively ECS becomes aversive and can produce a
delay-of-punishment gradient. Third, ECS produces an apparent loss of
memory which is selective upon the age of the memory trace, but only for a
few seconds following formation of the memory trace. This effect of ECS can
be interpreted as disruption of a memory consolidation process which lasts
for 20 to 30 seconds following learning. Finally, ECS halts the incubation,
that is, change over time, of a punishment-produced conditioned emotional
response which is characterized in the rat primarily by freezing behaviour.
The duration of the incubation process is an hour or more and ECS disruption
of incubation has frequently been misinterpreted as Ecs disruption of a long-
term consolidation process.

REFERENCES

Black, M., Memory consolidation theory and electroconvulsive shock : interference

paradigms. Unpublished master's thesis, Queen's University, 1969
Coons, E.E. and N.E. Miller, Conflict versus consolidation of memory traces to explain
'retrograde amnesia' produced by ECS. J. Comp. Physiol. Psycho/., 1960, 53:524-31
Duncan , C.P., The retroactive effect of electroshock on learning. J. Comp. Physiol.
Psycho/., 1949, 4 2: 32-44
Hebb, D.O., The Organization of Behavior; a Neuropsychological Theory. New York: J.
Wiley, 1949
Heriot, J .T . and P.D. Colman, The effect of electroconvulsive shock on retention of a
modified 'one-trial' conditioned evidence. J. Comp. Physiol. Psycho/., 1962,
55: 1082-4
Pfingst, B.E. and R.A. King, Time course of consolidation as measured by response-
choice behavior. American Psychological Association, 76th Annual convention;
proceedings, 1968, 3: 32 7-8
Pine I, J .P.J ., A short gradient of ECS-produced amnesia in a one-trial appetitive learning
situation. J. Comp. Physiol. Psycho/., 1969, 70:650-5
Pinet, J.P.J. and R.M. Cooper, The relationship between incubation and ECS gradient
effects. Psychonom. Sci., 1966, 6: 125-6
Suhoski, M.D ., M . Black, J. Litner, R.T. Greenner, and A.A. Spevack, Long and short-
term effects of ECS following one-trial discriminated avoidance conditioning. Neur<>
psycho/ogia, 1969, 7:349-56
Memory Consolidation Theory 87

Suhoski, M.D., A.A. Spevack, J. Litner, and E. Beaumaster, Effects of ECS following
one-trial discriminated avoidance conditioning. Neuropsycho/ogia, 1969, 7: 67-78
Thompson, R. and W. Dean, A further study on the retroactive effect of electro-
convulsive shock. J. Comp. Physiol. Psycho/., 1955, 48 :488-91
 A. MCGHIE

6 Input Dysfunction in Schizophrenia

CLINICAL FINDINGS
Sensory response to external stimuli is quite normal. To be sure, the patient
will complain that everything appears to be different ... However, this strange-
ness is usually attributable to a deficit in customary associations and parti-
cularly to an alteration in emotional emphasis, not to disturbances of
sensation (italics added).
Even though uninterested and autistically encapsulated patients appear to
pay little attention to the outside world, they register a remarkable number
of events of no concern to them. The selection which attention exercises over
normal sensory impressions may be reduced to zero, so that almost every-
thing that meets the senses is registered (italics added) .

These two totally contradictory comments on schizophrenia were both made

by Bleuler in his classic 1911 monograph. Until relatively recent years sub-
sequent clinical observations of schizophrenic patients have tended to
substantiate the view expressed in Bleuler's first statement of a clear sen-
sorium in schizophrenia and to add to the stereotype of the schizophrenic as
a withdrawn 'dreamer in a world awake .' More recently there has been
growing support for the observation expressed in the second quotation in
viewing at least one type of the 'group of schizophrenias' as being a result of a
primary input dysfunction involving an impairment of the normal inhibitory
functions of attention. In contrast to previous conceptions the schizophrenic
patient is seen, not as an individual who has effectively shut out external
reality, but as one who is overwhelmed by too much contact with the
environment.
Perhaps the most convincing exponents of this view have been the patients
themselves. In recent years there has been a healthy revival in the once
traditional phenomenological approach to psychotic patients. The patient's
own description of his experiences is once again being considered as a valuable
source of information. Fortunately, this approach has also included some
Schizophrenia 89

attempt to collate the information in a systematic manner. Thus, in our own

attempts to elicit descriptions from early and acute schizophrenic patients,
information was gathered through a standard type of interview. Descriptions
were deemed to be characteristic of schizophrenics only after a follow-up
study had verified the diagnosis of some of these young patients and allowed
their reports to be compared with other patients whose schizophrenic diag-
nosis was not verified in the follow-up period (McGhie and Chapman 1961 ).
The following extracts from the reports of patients are taken from this study
and from similar studies by Chapman (1966) and Bowers and Freedman
(1966):

I can't concentrate ... It's diversion of attention that troubles me ... the sounds
are coming through to me but I feel my mind cannot cope with every-
thing ... It is difficult to concentrate on any one sound .. . it's like trying to do
two or three things at the one time. (McGhie and Chapman 1961)
Everything seems to grip my attention although I am not particularly
interested in anything. I am speaking to you now but I can hear noises going
on next door and in the corridor. I find it difficult to shut these out and it
makes it more difficult for me to concentrate on what I am saying to you.
(McGhie and Chapman 1961)
It has to do with what is going on around me ... taking in too much of my
surroundings ... vital not to miss anything ... I can't shut things out of my
mind and everything closes in on me ... (Chapman 1966)

The involuntary widening of attention implied in such statements is often

accompanied by reports denoting a heightening of sensory vividness:

Have you ever had wax in your ears for a while and then had them syringed?
That's what it's like now, as if I had been deaf before. Everything is much
noisier and it excites me . (McGhie and Chapman 1961)
The colour of things seems much clearer and brighter ... maybe it's be-
cause I notice so much more ... not only the colour of things fascinates me
but all sorts of little things like markings on the surface pick up my attention
too. (McGhie and Chapman 1966)
One night I woke up and started feeling good again ... I felt alive and vital,
full of energy. My senses seemed alive, colours were very bright. They hit me
harder ... I noticed things I had never noticed before. (Bowers and Freedman
1966)
I feel my tactile senses are enhanced as well as my visual ones, to a point
of great power. Patterns and designs begin to distinguish themselves and take
on significance . (Bowers and Freedman 1966)
90 A Triune Concept of the Brain and Behaviour

Thought disorder is perhaps the most distinctive feature of schizophrenia.

The comments of some of the patients in our own study not only provide
vivid descriptions of the form of schizophrenic thought disorder but suggest
its association with a primary disorder of attention : 'My thoughts get all
jumbled up. I start thinking or talking about something but I never get there .
Instead I wander off in the wrong direction and get caught up with all sorts of
different things that may be connected with the things I want to say but in a
way I can't explain' (McGhie and Chapman I 963).
One patient summed up his difficulties in the following telling comment :
'When I'm trying to think of something I'm like a railway engine, running
along a line where someone keeps changing the points' (McGhie and Chapman
1963).
Some patients also report disturbance of their bodily movements and
general motility :

I'm not sure of my own movements any more. It's very hard to describe this
but at times I'm not sure about even simple actions like sitting down ... I
found recently that I was thinking of myself doing things before I would do
them. If I'm going to sit down for example, I've got to think of myself and
almost see myself sitting down before I do it. It's the same with other things
like washing, eating, and even dressing - things that I have done at one time
without even bothering or thinking about at all. (McGhie and Chapman 1961)

The loss of spontaneity in behaviour which schizophrenic patients

describe would seem to be a natural consequence of their conscious attention
being invaded by the volitional impulses and stimuli from the effectors which
normally function autonomously outside the range of awareness. The patient
now has consciously to initiate and control his bodily movements, every one
of which involves a decision. Activities which before were self-regulative are
now experienced as uncertain and require deliberate co-ordination. It is small
wonder that the patient speaks of a split between his mind and body, and
feels that he is in danger of losing control over his own actions.
Let us conclude these selections from patients' descriptions by quoting
from a passage written by a schizophrenic girl who, while in a stage of partial
remission, tried to interpret her own acute psychotic experiences:

Each of us is capable of coping with a large number of stimuli, invading our

being through any of the senses ... It's obvious that we would be incapable of
carrying on any of our daily activities if even one hundredth of all these
available stimuli invaded us at once. So the mind must have a filter which
functions without our conscious thought, sorting stimuli and allowing only
Schizophrenia 91

those which are relevant to the situation in hand to disturb consciousness.

What happened to me was a breakdown of the filter, and a hodge-podge of
unrelated stimuli were distracting me from things that should have held my
undivided attention ... I had very little ability to sort the relevant from the
irrelevant ... Completely unrelated events became intricately connected in my
mind. (McDonald 1960)

EXPERIMENT AL FINDINGS

A statement as clear as the one quoted above might cause us to wonder why
research workers feel compelled to perform time-consuming experimental
studies to arrive at precisely the same conclusions. There are perhaps two
main reasons for further extended study. The first is the inescapable fact that
clinical psychologists have to justify their existence - doctorates must be
obtained, salary cheques must be justified, and we cannot possibly tolerate
patients substituting facts for hypotheses verified at the one per cent level.
The other and perhaps more serious justification for experimental enquiry is,
of course, that it permits us to seek answers to questions which are less
accessible to introspective reports. Many questions are raised by the clinical
data: is the widening of attention reported by schizophrenic patients specific
to this diagnostic group, or may it be found in other patient groups - or for
that matter, in normal subjects under certain conditions? Is this form of input
dysfunction characteristic of all patients diagnosed as schizophrenic, or is it
related to certain types or stages of the illness? Do the effects of experi-
mentally induced distraction on schizophrenic performance vary with the
aspects of behaviour being tapped by the task involved? Are distraction ef-
fects related in any way to the sensory modalities in which the task informa-
tion or the distracting information is transmitted? To what extent may an
impairment in the filtering of irrelevant information play a causal role in
relation to other areas of schizophrenic impairment, e.g., language disorder,
motor impairment, thought disorder, and changes in affect?
Subsequent experimental studies appear to have gone some way to
answering some of these questions, although frequently only replacing them
with new questions. These experimental investigations have a wide variety of
techniques, reflecting the varying theoretical approaches of those involved.
Payne (I 970) has approached the problem by studying the development of
loosely constructed over-inclusive thinking and has relied mainly on verbal
and non-verbal tests of concept formation . Shakow (I 962, 1963) has con-
centrated predominantly on speed of decision-making and has made use of
tests of simple and complex reaction time. Weckowicz (1959) and others
have interpreted their measures of disturbed perceptual consistency in
92 A Triune Concept of the Brain and Behaviour

schizophrenic patients as another aspect of impaired attentional selectivity.

Silverman (1964) has ingeniously adapted the principles of cognitive con-
trol (scanning and field articulation) to the study of schizophrenic input
dysfunction .
In our own investigations the basic technique used was simply to select
tasks which seemed to include specific areas of mental functioning, and to
examine the patients' performance on such tasks, both in the absence and in
the presence of experimentally induced distraction. The form of the distract-
ing stimuli was systematically varied, being either auditory or visual, either
meaningful or meaningless. The subjects studied in this series of investigations
included schizophrenic patients ( differentiated according to subtype and
chronicity), non-schizophrenic psychotic patients (mainly manic-depressive),
non-psychotic patients (mainly character disorders), patients with a clear
diagnosis of organic involvement (mainly epileptic and arteriosclerotic ), and
normal control subjects (a motley group of nursing staff, students, and
psychiatric and psychology residents too junior to object).
Although these studies confirmed the reports of schizophrenic patients of
heightened vulnerability to external distraction, the findings also indicated
that such vulnerability was contingent upon certain situational variables.
Our main findings may be quickly summarized as follows: (I) On tasks of
psychomotor performance the effects of distraction were slight and affected
all groups equally when the task demanded only a motor response to a
predictable stimulus. However, when the task was more complex, requiring
the processing of unpredictable stimuli (e .g., a signal tracking task), the effect
of distraction was pronounced and specific to the schizophrenic group. (2) In
tasks requiring the patient to integrate information in two sensory modalities,
schizophrenic patients had marked difficulties. It was also noticeable that, in
such circumstances, the falling off in performance was most marked for the
processing of visual information. (3) In a number of tasks requiring the pro-
cessing of either visual or auditory information (in the absence of distraction)
schizophrenic patients fared most poorly with visual information. In the
presence of experimentally induced distraction, schizophrenic performance
fell off sharply, this effect being most pronounced on tasks requiring the
processing of auditory information.
In interpreting our findings we leaned heavily on Broadbent's (I 958)
theory of a limited capacity decision channel. Studies of normal information
processing have already made it clear that the limitation of the human com-
munication channel is an informational one, so that the number of stimuli
which can be responded to at any time is determined by the amount of
information they contain. It is thus possible to deal with more than one set of
data at a time only if the informational demands of each task are small. In
order to function effectively, the individual is forced to perform a selective
Schizophrenia 93

filtering operation on the input to ensure that his limited capacity is not
overloaded. We interpreted our findings as indicating that, in schizophrenia,
this nonnal filtering process has broken down so that the patients are less able
to attend selectively and to process only relevant information. This defect
may be expected to have varying effects depending upon the nature and
demands of any task. In dealing with a situation requiring a response to
simple predictable stimuli, overloading will be less likely, and the patient's
deficit less obvious. In tasks demanding the monitoring of a range of stimuli
involving more complex decision-making and fully occupying the limited
decision channel, the failure in selective attention is more likely to lead to
overloading and consequent breakdown in performance . The relatively slow
rate of recoding visual information into the auditory modality increases the
likelihood of overloading. Similar conclusions have been derived from many
other forms of experimental investigations of schizophrenic patients.
Weckowicz and Blewett (I 959) conclude a series of studies of reduced per-
ceptual constancy in schizophrenic patients by stating, 'Their (schizo-
phrenics') perception is more influenced by the here and now factors of the
immediate situation and less by the experience of the past and the anticipa-
tion of the future ... There is some change in the filtering of stimulation ... The
gates are open wide, the cortex is flooded with irrelevant information so that
maintenance of attention becomes difficult.'
On the basis of his own investigations of reaction time performance in
schizophrenics, Shakow (I 962) arrived at a similar conception : 'It is as if, in
the scanning process which takes place before the response to a stimulus is
made, the schizophrenic is unable to select out the material relevant for
optimal response. He apparently cannot free himself from the irrelevant
among the numerous possibilities available for choice. In other words, that
function which is of equal importance as a response to stimuli, namely the
protection against the response to stimuli, is aberrant.' In a similar vein Payne
(I 964) concluded his survey of schizophrenic thought disorder by suggesting
that such disordered functioning might be regarded as secondary elaboration
of a more fundamental breakdown at the input level: 'The mechanism of
attention itself seems to become defective. Whatever filtering mechanism
ensures that only the stimuli that are relevant to the task enter consciousness
and are processed, seems no longer to exclude the irrelevant. This has
numerous repercussions. Thinking becomes distracted by external events [and
also] by irrelevant thoughts and emotions.'
In our own studies it was repeatedly observed that the hypothesized input
dysfunction was not common to all schizophrenic patients. Comparisons of
distractability performance and the clinical picture made it clear that this
attentional deficit was not apparent in patients whose clinical condition was
predominantly paranoid. Indeed, on most of our tests the paranoid patients
94 A Triune Concept of the Brain and Behaviour

were less distractable than normal subjects. This finding is also prominent in
the work of Weckowicz, Shakow, Silverman, and others. Shakow (1963)
describes a total of twenty varied experiments in which the paranoid and
non-paranoid subjects' scores fell on either side of those of the normals. Using
the concepts of scanning control and field articulation developed by the
Menninger group (Gardner, Holzman, Klein, Linton, and Spence 1959),
Silverman (1964) concluded that, 'Things and events either fit the paranoid
schizophrenic's delusioned system or they do not exist. Under these condi-
tions, a state of information overload - a state which appears to characterize
other schizophrenics - is not possible. For there are only two kinds of in-
formation to be processed - that which is congruent with one's delusional
system and that which is not.'

SUMMARY OF FINDINGS ON INPUT DYSFUNCTION

We might pause at this point to summarize briefly the position with regard to
schizophrenia input dysfunction as follows. The performance of some schizo-
phrenic patients is strongly indicative of a breakdown in the normal selective
and inhibitory functions of attention . These patients suffer from a marked
inability to attend selectively to stimuli in such a way that only relevant
information is processed. The effect of this impairment on the performance
of the patient appears to vary with a number of factors, including the amount
of information to be processed, the modality in which information is process-
ed, and the rate of arrival of information. Clinically, these patients tend to
present, with an insidious onset to their condition, marked thought disorder,
loss of affect, and a progressive deterioration into a chronic state. It is
probably of little importance whether we categorize these patients as 'hebe-
phrenia, nuclear,' or 'process,' schizophrenia, but the clinical picture they
represent is almost identical with the original Kraepelinian conception of
dementia praecox. In marked contrast to this group, the paranoid patient
tends to show a highly selective form of attention which enables him to
screen out extraneous stimulation more effectively than even the normal
subject. It seems likely that such workers as Silverman {1964) may be correct
in suggesting that the filter mechanism of the paranoid patient is set to attend
selectively only to events which might interfere with the rationale of his
delusional system.

CLINICAL CONSEQUENCES OF INPUT DYSFUNCTION

As mentioned earlier in our discussion, the picture of the nuclear

schizophrenic which emerges from these studies is in vivid contrast to that
Schizophrenia 95

of the withdrawn and shut-in individual familiar to those who work in the
chronic mad setting.
It would seem possible that the bizzare fonns of withdrawal demonstrated
by some chronic patients represent strategies developed to reduce the in-
coming flow of information to a more bearable level. Thus the patient who
remains isolated in a quiet corner of a ward with his eyes closed and ears
plugged with cotton wool may almost literally be attempting to shut out a
bad and hostile world. Such an argument is supported by the otherwise
puzzling but repeated finding that these patients who are most socially with-
drawn are in a state of high cortical arousal (Venables 1963; Venables and
Wing 1962).
We might also conclude that schizophrenic patients with this fonn of input
dysfunction might fare most poorly in an excessively stimulating environment
and improve their performance in an environment where the amount and rate
of sensory input is carefully controlled. This suggestion has been supported
by the reports of clinicians (Chapman 1966; Chapman and McGhie 1963)
that secondary symptoms such as hallucinations, catatonic behaviour, and
social withdrawal are intensified when the patient finds himself in large noisy
wards where his senses are constantly bombarded by multiple stimuli. These
workers have also sought to demonstrate an improvement in clinical state
when the same patients are observed in a less demanding environment.
Chapman ( 1966) has demonstrated how the communicative ability of the
schizophrenic patient may be greatly enhanced by the interviewer avoiding
conditions which might cause the patient to be 'overloaded.' Similar improve-
ments in the perfonnance of subjects working in a reduced sensory environ-
ment have been experimentally demonstrated. 1

CURRENT INTERESTS

In this final part of my presentation I should like to become even more

openly egocentric than hitherto and describe briefly a few of my own current
research interests in this field . These studies are being carried out by graduate
students under my supervision - which is the academic way of saying that
they will be doing all the work, while I shall share the credit if the findings
are worthwhile, reserving the right to disown any responsibility should the
results prove negative.
A noticeable feature of the reports of some young schizophrenic patients
is that the gradually increasing range of their attention was at first ex-
perienced as pleasant and conducive to a wide appreciation of reality. Indeed,

P. Slade, personal communication, 1968

96 A Triune Concept of the Brain and Behaviour

some patients had the subjective impression of increased efficiency during the
very early stages of their illness. Their emotional reactions change to alarm
only when the involuntary widening of attention threatened to obliterate all
former stable constructs of reality. Lehmann (I 966) makes an interesting
point in this direction when he describes the schizophrenic patient having 'a
primary possibly constitutional susceptibility ... to be subjected to the impact
of a higher number of discrete sensory stimuli per time unit of experience
than most other pathological or non-pathological individuals.' He then goes
on to say, 'If he is capable of coping with this greater than average influx of
discrete sensory stimuli, he might perform at a better than average level, but
when the extraordinary sensitivity of his receptive apparatus is not matched
by an equally extraordinary performance of his central processing apparatus,
then his integration breaks down and he may become psychotic' (italics
added) .
Studies of the effects on normal performance of progressively increasing
the quantity of information handled per time unit demonstrate wide in-
dividual difference in processing capacity . Some individuals appear to be able
to handle an extraordinarily high rate of input while the processing capacities
of others are severely restricted. This raises the possibility that the impaired
filter functioning which is a serious defect to the schizophrenic patient may
act advantageously in the case of some individuals who are able to effectively
process the increased input. With this possibility in mind we are making a
comparative study of schizophrenic patients and normal adults assessed to be
'creative' thinkers. The hypothesis to be tested is simply that creative thinkers
and schizophrenics will both demonstrate a wider sampling of environmental
stimulation. However, whereas the creative thinkers will show themselves
capable of processing this greater input, the schizophrenics should be unable
to handle the input. Another hypothesis being tested is that the performance
of normal and schizophrenic individuals will near equivalence in conditions
where the normal subject's channel capacity is systematically overloaded .
In such studies we are utilizing a technique devised by workers in the field
of ergonomics to measure mental workload. Subjects are required to cope
with a progressively increasing load on a primary task while performing a
fixed secondary task. The secondary task is thought to initially absorb the
spare mental capacity not used by the primary task, so, by increasing the rate
of input in the primary task, the deterioration in performance on the
secondary task may be seen as an index of the amount of 'distraction stress.'
This method has been used by Brown and Poulten (1961) to assess such
factors as individual differences in the spare capacity of car drivers operating
in situations of varying decision-making complexity. Other workers, such as
Schizophrenia 97

Kalsbeek (I 964) in Amsterdam have used a similar technique to measure

the mental load involved in specific industrial tasks. Kalsbeek has found
that individuals vary greatly in their information handling capacities in this
situation and that the deterioration in performance on the secondary task
as a consequence of such overloading is reminiscent to the performance of
brain-damaged patients. In discussing these and other findings Kalsbeek
comments, 'In general we think that reduced information handling capacity
forces the subject to retreat to simple levels of organized behaviour. The
latter levels are thought of as pre-existing strategies for shedding load in
situations which otherwise would lead to overloading the subject's central
nervous system ... Strategies adopted by the subject in order to avoid the
state of overload can be conceived of as symptoms which would enable
the experimenter to diagnose the level of simplification towards which
the subject is forced to retreat' (I 964 ). Although Kalsbeek is thinking
entirely within an industrial context it would seem that both his tech-
nique and his model of distraction stress are of obvious relevance to the
conception of input dysfunction in schizophrenia. If our view of input
dysfunction in nuclear schizophrenia is correct, one might hypothesize
that the performance of normal subjects on secondary tasks under dis-
traction stress conditions would progressively approximate to that of
schizophrenics as the input flow of the primary task was progressively
increased .

CONCLUSION

In concluding, I should like to reiterate a theme introduced much earlier

in this discussion . In its pressing need for scientific respectability, psy-
chological research in the clinical field has become increasingly ex-
perimental. This movement has brought with it the undoubted benefits
of more rigorously controlled investigation yielding more reliable data. A
less fortunate accompaniment of an increasingly experimental approach
has been the denigration of the patient himself as a direct source of in-
formation . The patient's own account of his changing experience has
been regarded as too 'subjective' and unfit to keep company with the
more objectively derived experimental data. My own feeling is that most
productive experimental studies of psychiatric patients are built upon an
initial foundation of clinical observation. Clinical information from patients,
such as that quoted in this paper, is complementary to experimentally derived
data. Indeed, a more receptive attitude towards the patients' self-descriptions
might render many of our experimental investigations unnecessary.
98 A Triune Concept of the Brain and Behaviour

REFERENCES

Bleuler, E., Dementia Praecox; or the Group of Schizophrenias. Translated by J. Zinkin.

New York : International Universities Press, 1950
Bowers, M.B ., Jr., and D. Freedman, 'Psychedelic' experiences in acute psychoses. Arch.
Gen. Psychiat., 1966, 15 :240-8
Broadbent, D.E. , Perception and Communication. New York: Pergamon, 1958
Brown, I.D. and E.C. Poulten, Measuring the spare 'mental capacity' of car drivers by a
subsidiary task. Ergonomics, 1961 , 4: 35-40
Chapman, J., The early symptoms of schizophrenia. Brit. J. Psychiat., 1966, 112: 225-51
Chapman, J. and A. McGhie, An approach to the psychotherapy of cognitive dys-
function in schizophrenia. Brit. J. Med. Psycho/., 1963, 36 :253-60
Gardner, R.W., P.S. Holzman, G.S. Klein, H.B. Linton, and D.P. Spence, Cognitive con-
trol - a study of individual consistencies in cognitive behavior. Psycho[. Issues, 1959,
1, no. 4: 1-186
Kalsbeek, J.W.H ., On the measurement of deterioration in performance caused by
distraction stress. Ergonomics, 1964, 7: 18 7-9 5
Lehmann, N. , Pharmacotherapy of schizophrenia. In P. Hoch and J. Zubin (eds.), Psy-
chopathology of Schizophrenia. New York: Grune and Stratton, 1966
McDonald, N., Living with schizophrenia. Can. Med. Assoc. J. , 1960, 82:218-21
McGhie, A. and J . Chapman, Disorders of attention and perception in early schizo-
phrenia. Brit. J. Med. Psycho/., 1961, 34: 103-16
Payne, R.W., The measurement and significance of overinclusive thinking and retardation
in schizophrenic patients. Paper presented to APA, 1964
Payne, R.W. Disorders of thinking. In C.G. Costello (ed.), Symptoms of Psycopathology.
New York: Wiley, 1970
Shakow, D., Segmental set: a theory of the formal psychological deficit of schizo-
phrenia.Arch. Gen. Psychiat., 1962, 6 : 1-17
Shakow, D., Psychological deficit in schizophrenia. Behavioral Sci., 1963, 8:275-305
Silverman, J., The problem of attention in research and theory of schizophrenia.
Psycho/. Rev. , 1964, 71 :352-64
Venables, P.H ., The relationship between the level of skin potential and fusion of paired
light flashes in schizophrenic and normal subjects. J. Psychiat. Res. , 1961-3, 1: 279-87
Venables, P.H. and J. Wing, Levels of arousal and the subclassification of schizophrenia.
Arch. Gen. Psychiat., 1962, 7: 1_14-9
Weckowicz, T .E. and T.B. Blewett, Size constancy and abstract thinking in schizophrenic
patients. I. Mental Sci., 1959, 105 :909-34
 JAMES INGLIS

7 Similarities in the Side-Effects of ECT

and Temporal Lobectomy in Man 1

The main contention of this paper is that some of the transient side-effects of
electroconvulsive therapy on memory resemble, in kind if not in degree, those
more severe and chronic learning defects that are known to appear as an
incidental result of temporal lobectomy in man. If this claim can plausibly be
supported it would imply a pressing need for the more systematic study of
other modes of therapeutically effective ECT that would yet interfere as little
as possible with the normal activity of those parts of the human brain that are
essential for adequate learning and memory function.
Support for the notion of such a parallel between ECT and temporal
lobectomy comes from at least three sources of evidence. These relate, respec-
tively, to (I) the physical site of action of the two procedures; (2) the likely
physiological mode of action of electroshock on learning; and (3) the psy-
chological effects of the two procedures. Some of the relevant evidence may
be considered under these three headings, as follows.

PHYSICAL EVIDENCE

No very precise studies seem to have been made of the exact regions of the
brain that may underlie even the most common sites of electrode placement
in ECT. In any case it is likely that there exist individual differences in
skull-and-brain relationships of such magnitude as to make it impossible to
define with absolute accuracy the precise correspondence between external
electrode sites and internal nervous structures.
Nevertheless, some idea of the regions most likely to lie under the ECT
electrodes can be got from any anatomical atlas that shows human skull-

Another version of this paper has been published in the British Journal of Psychiatry
(1970). Grateful acknowledgements are hereby made to that journal for permission to
reprint. Part of the preparation of the present version was carried out while the author
was in receipt of a research grant from the Ontario Mental Health Foundation (OMHF
no. 284) ; this aid is also most gratefully acknowledged.
100 A Triune Concept of the Brain and Behaviour

Figure l A representation of skull-and-brain relationships in man. (After Kampmeier,

Cooper, and Jones 1957)
Side-Effects of ECT 101

Figure 2 Two of the possible positions of electrodes in bilateral ECT. A. Bifrontal

placement of electrodes. C. Bitemperoparietal placement of the electrodes.
(Alexander 1953)

. ,
,,.,.··.
~--·
/~o~
..,.... r·,i}~·
Y"";----- -. ~

Figure 3 The position of the electrodes in unilateral ECT. In the unilateral placement
the lower electrode was midway between the lateral angle of the orbit and the external
auditory meatus and 1½ inches above this line. The upper electrode was 3 inches higher
than the lower and at an angle of 70° to the line. (Slight deviations from these points do
not materially affect the results.) (Lancaster, Steinert, and Frost, J. Mental Sci., 1958)
102 A Triune Concept of the Brain and Behaviour

and-brain relationships. An example has been adapted from Kampmeier,

Cooper, and Jones (I 957) and is shown in Figure I .
When the bilateral form of ECT is used, the common method requires that
the electrodes be placed on or near the patient's temples (Kalinowsky 196 7).
There is, however, undoubtedly much variation in electrode placement in
clinical practice. Two of the many possible positions have been illustrated by
Alexander (1953) and are shown in Figure 2.
In the application of unilateral ECT most users seem to follow the example
of Lancaster, Steinert, and Frost (1958) and place their electrodes more or
less as shown in Figure 3.
Both bilateral and unilateral ECT would, therefore, usually seem to be
applied to areas over or near the temporal lobes of the brain.
An illustration of the surgical approach to these areas for the purposes of
temporal lobectomy has been provided by Scoville and Milner (1957) as
shown in Plate I.
The actual physical sites of action involved in both ECT and temporal
lobectomy would, therefore, appear at least to overlap. There are, in addition,
other grounds for the belief that ECT may affect, in particular, critical struc-
tures within the region of the temporal lobes.

PHYSIOLOGICAL EVIDENCE

It has been shown that the effect of electrical brain stimulation necessary for
elevation in the mood of depressed patients is a generalized neurological
discharge of the kind produced by orthodox ECT, while the common amnesic
effect of such shock is not at all related to therapeutic outcome (Ottosson
1960). There is, however, also evidence that there may be a local action of
electroshock on brain areas underlying electrode placement that is in-
dependent of the general therapeutic activity and which may, indeed, be
deleterious. At least three kinds of evidence seem relevant to this point. These
derive from (a) studies of EEG changes after different forms of ECT; (b)
studies of various levels of convulsive threshold in different brain structures;
and (c) studies of the effects of direct stimulation of the temporal lobes.

( a) EEG studies

There are now available over twenty studies of the effects of unilateral ECT,
many having included comparisons with its bilateral form. Of these, a number
have looked at subsequent EEG records (vide Blaurock, Lorimer, Segal, and
Gibbs 1950; Martin, Ford, McDonald, and Towler 1965; Zamora and Kael-
bling I 965; Valentine, Keddie, and Dunne I 968) and most have found
Side-Effects of ECT 103

Plate 1 A surgical approach to the temporal lobe (Scoville and Milner, J. Neural.
Neurosurg. Psychiot., 1957)
104 A Triune Concept of the Brain and Behaviour

evidence of electrical disturbance (e.g. , slow waves) ipsilateral to the side of

electrode placement.
These observations support the notion that the kind of stimulation used in
ECT may have a local as well as a general effect, at least insofar as laterality of
action is concerned.

(b) Convulsive thresholds

A number of ihvestigators, including Liberson and his colleagues, have mea-

sured the different convulsive thresholds of various brain structures both in
man (Liberson , Scoville, and Dunsmore 1951; Scoville, Dunsmore, Liberson,
Henry, and Pepe 1953) and in animals (Liberson and Akert 1953; Liberson
and Cadilhac 1953). Their findings suggest that, of all the areas studied, the
hippocampus may be one of the brain structures with the lowest threshold to
convulsion. Hence Liberson and Cadilhac conclude, 'These investigations
show that both parameters of stimulation and unequal aptitude to after-
discharge of different structures of the brain may explain the possibility of
dissociation of the electroshock patterns. Whatever the part of the hippo-
campus may be in the total picture of electroshock, it must be involved to the
highest degree because of its low epileptogenic threshold' (I 953, p. 284).
Given, then , the physical position of the electrodes over the temporal
lobes and given the differential sensitivity to shock of some associated sub-
structures, it would therefore seem likely that these brain areas are the ones
to bear the main brunt of any local effects of shock under the usual
conditions of bilateral or even unilateral ECT .

( c) Direct stimulation

The most likely consequence of such local shock in man may be inferred
from studies of direct stimulation carried out either at open brain surgery or
through implanted electrodes.
Perhaps the most extensive series of studies of brain stimulation in con-
scious patients at surgery has been carried out at the Montreal Neurological
Institute. Such intervention in certain areas of the temporal lobes has been
shown to produce automatism with associated amnesia. Wilder Penfield
(1958) has summarized these findings as follows: 'Automatism is not asso-
ciated with complete loss of consciousness but with amnesia. The automatic
individual may move about in a more or less confused state . His behavior
shows lack of normal insight. But the major defect is that he seems to be
deprived of the ability of making a memory record' (1958, p. 215). Penfield
goes on to state, furthermore, that it has been found that" ... the area of the
temporal lobe in which epileptic discharge might produce automatism was the
Side-Effects of ECT 105

peri-amygdaloid area and the hippocampal zone, and further that electrical
stimulation in the peri-amygdaloid area produced automatism but only when
stimulation was followed by epileptic after-discharge' (I 958, p. 215).
Further information from this series has been provided by Jasper and
Rasmussen (1958). Deep temporal stimulation in the peri-amygdaloid region,
they found, was most likely to result in what they labelled as 'psychoparetic'
symptoms. This condition, in fact, comprised that association of automatism
and amnesia already desclibed by Penfield.
Stimulation can also be induced in these areas by electrodes implanted
through burr-holes in the closed skull. Bickford, Mulder, Dodge, Svien, and
Rome (I 958) have reported two cases in which amnesia for recent events
followed upon the passage of unilateral current through sub-temporal regions.
A similar result in one case of bilateral stimulation has more recently been
reported by Brazier (I 964) .
Taken together, these various kinds of evidence (from EEG, differential
threshold, and direct stimulation studies) suggest that ECT might indeed have
an incidental effect on limited areas of the human brain. If these areas happen
to include the particularly sensitive structures of the temporal lobes, then the
consequences may well include transient disturbances of learning.
It would therefore seem not unreasonable to expect that analogies might
be found between the behavioural effects of massive electrical stimulation
and the actual physical removal of brain tissue in these areas.

PSYCHOLOGICAL EVIDENCE

Recent reviews, such as those carried out by Brierley (1966a, 1966b) have
strongly suggested that in many human disorders in which learning dys-
function appears as a significant element there is often also to be found
evidence to indict malfunctioning of the temporal lobes and their adjacent
structures, particularly the hippocampal regions. This is not to say that these
areas are in any sense the 'site of storage.' Their integrity, however, seems
vital for the maintenance of the cycle of recording involved in learn-
ing - perhaps most especially for the transferring of what is to be learned
from a temporary to a more pennanent store.
This view is most powerfully supported by studies of the effects of the
actual surgical removal of temporal lobe tissue in man - an operation usually
carried out for the relief of otherwise intractable temporal lobe epilepsy (Hill
1953). Such studies furthermore demonstrate the crucial importance of both
the lobe and the side of the brain operated upon . Account must, in fact, be
taken of whether the dominant or the non-dominant lobe is the site of
surgery. In most of these studies the dominant side has commonly been
defined as the left side of the brain in right-handed patients, the
106 A Triune Concept of the Brain and Behaviour

non-dominant (or minor) lobe then being, of course, on the right side. This
definition, although not completely satisfactory on all counts, is serviceable
enough and will be the one used here.
The analogies between some of the effects of temporal lobe surgery and
ECT will therefore be further pursued in terms of dominant, non-dominant,
and bilateral interference.

(a) Leaming defects and temporal lobe surgery

(i) Dominant temporal lobectomy

In a series of studies carried out at the Guys-Maudsley Neurosurgical Unit in

London it was shown (Meyer and Yates 1955; Meyer 1959) that, apart from
its therapeutic effect on epilepsy, the main result of dominant temporal
lobectomy in human patients is to produce a defect of verbal learning,
especially of verbal material presented through the auditory modality. In
these studies the verbal learning defect was not evident before the operation
but was manifest afterwards. The same disability was shown to be present in
these patients at least one year after the operation.
Contemporaneous, but quite independent studies at the Montreal Neuro-
logical Institute (Milner 1958) found a slight impairment of verbal learning in
dominant cases even before the operation and showed this to be exacerbated
after surgery. Milner (1962) has since presented persuasive evidence that
verbal content alone, quite apart from modality of presentation, is the critical
factor in the production of a dramatic verbal learning impairment after
dominant temporal lobectomy in man. She has, furthermore, also used con-
trol groups of frontal and parietal patients to show that such defects are
relatively specific to dominant temporal lobe damage and are not simply due
to the side of the brain affected. In other words, there is a highly significant
interaction between lobe and laterality.

(ii) Non-dominant temporal lobectomy

In the Montreal studies, defects of non-verbal learning and discrimination, on

the other hand, have consistently been found by Milner (1958, 1968) and her
associates after surgery involving the non-dominant temporal lobe. Thus,
Kimura (1963) has shown that non-dominant cases find great difficulty in the
immediate tachistoscopic recognition, and in the recognition of previously
presented unfamiliar visual designs. Additional defects in non-verbal learning,
as in memory for tonal patterns (Milner 1962), for visually-guided stylus-
maze learning (Milner 1965) and proprioceptively-guided stylus-maze learning
Side-Effects of ECT 107

(Corkin 1965) have also been reported following surgical removal of parts of
the non-dominant lobe in man. Again, the behaviour of appropriate control
patients has shown the crucial importance of damage to the temporal lobe
itself rather than the mere side of the brain on which the lesion is placed; lobe
and laterality interact.

(iii) Bilateral temporal lobectomy

A gross and general form of memory impainnent is often clinically evident

after bilateral temporal lobe damage. Cases so afflicted have been described
by Milner (1959) and by Scoville and Milner (I 957).
In view of what is known about the effects of damage to a single lobe, on
one side or the other, it seems likely that the particularly harmful effects of
bilateral interference could be due to a combination of two factors.
In the first place, if the dominant lobe is removed then the non-dominant,
minor lobe may still to some extent be able to compensate through a process
of vicarious functioning; that is to say, the minor lobe may, by retraining,
take over part of the functions of the contralateral damaged lobe (and, of
course, vice versa).
Secondly, it may further be the case that patients with dominant temporal
lobectomy, who mainly suffer from a verbal defect, can also partly com-
pensate for this handicap by relying more heavily, for example, on non-verbal
cues - as well as on other aids related to the kinds of learning originally
mediated by the non-dominant lobe.
Bilateral damage that extensively affects both the dominant and the non-
dominant lobe would then rob the patient of any possibility of vicarious
function and would furthermore deny him the opportunity of substituting
one kind of learning for another, thus producing the crippling, generalized
form of learning disorder that has been found after bilateral temporal lobe
resection in man.

( b JLearning defects and electroconvulsive therapy

It is contended in this paper that there may be a close resemblance between

the side-effects of various forms of temporal lobectomy and the corres-
ponding kinds of ECT. It is to be expected, however, that the defects pro-
duced by functional electrical disturbance should be much less severe and also
of much briefer duration than the defects produced by structural damage due
to disease or surgery.
There are, as yet, few studies available in which the effects of the two
kinds of unilateral ECT on different learning functions have been examined
l08 A Triune Concept of the Brain and Behaviour

20 Psychiatrist Patient
N=52

rj
15

ll
L16
10
R18
5
B18 (17)

0
LR B LR B
L/R NS NS
L/R NS NS
R/B NS NS

Figure 4 Depression rating scale results on the first occasion of retest. Mean change
(improvement) scores in two kinds of depression rating scales after 4 treatments.
(Halliday, Davison, Browne, and Kreeger, Brit. J. Psychiat., 1968)

and compared with the effects of bilateral ECT. Those results that are now
available, however, seem to be not inconsistent with the main contention of
this paper.
Thus, Halliday, Davison, Browne, and Kreeger (1968) carried out a com-
parison of the effects on mood and memory of bilateral ECT and of unilateral
ECT both to the dominant and to the non-dominant hemispheres. This investi-
gation involved over 50 depressed patients for whom ECT had been pre-
scribed, each being randomly assigned to one of the three treatment groups.
All of these patients were under the age of 65 and all were right-handed,
without history of left-handedness in the family. Each received a standardized
course of four ECT treatments, the shocks being given at half-weekly intervals,
unilaterally to the right or to the left side of the head, or bilaterally with
conventional fronto-temporal electrodes.
In this study depression was rated 'blind' by the psychiatrist and also
self-rated by the patients, before and after four ECTS and again three months
later following variable amounts of further treatment. There proved to be no
significant difference in the effect of the three kinds of ECT on these depres-
sion ratings. There did appear to be a non-significant trend toward a better
response to bilateral ECT initially, as is shown in Figure 4.
Side-Effects of ECT 109

20 Psychiatrist Patient
15 N=44

ll
L11
10
R16
5
817 (16)
0
LR B LR B
L/R NS NS
L/R NS NS
R/8 NS NS

Figure 5 Depression rating scale results of the three month follow up. Mean change
(improvement) scores in two kinds of depression rating scales at 3-month follow up.
(Halliday, Davison, Browne, and Kreeger, Brit. J. Psychiat., 1968)

This trend, however, was not confirmed at the three month follow-up
assessment, at which time the best result, as may be seen from Figure 5,
was obtained from unilateral treatment to the non-dominant, right
hemisphere.
Learning and memory were assessed before and after four ECTS and also
three months later on both verbal and non-verbal learning tasks (Williams
1968). In one of the verbal learning tasks, for example, the patient had to
learn the meanings of five unfamiliar words and was scored in terms of the
number of failures to give the correct meanings over five trials. One of the
non-verbal learning tests used was the Rey-Davis peg-board (Zangwill 1946).
In this test the patient had to learn the different positions of a single fixed
peg on each of a series of boards each also bearing a number of movable
pegs.
The changes in performance after four treatments are shown in Figure 6.
These results show that ECT to the dominant hemisphere selectively impairs
verbal learning, whereas ECT to the non-dominant hemisphere selectively im-
pairs non-verbal learning, results that are, of course, exactly analogous to the
effects of the two corresponding forms of temporal lobectomy . Bilateral ECT,
however, only produced an intermediate pattern of impairment except on the
110 A Triune Concept of the Brain and Behaviour

Non- Delayed Delayed

Digit Verbal verbal Delayed verbal non-verbal
5 span learning learning recall learning learning
4
3 Left~
2 Right
1 Bilateral
0

IT'
1
2
3
4
5
6
7 N=52
L/R <0.005 <0.025 <0.05 NS <0.05 <0.01 L16
L/8 NS <0.01 NS NS <0.01 NS R18
R/8 NS NS NS NS NS NS 818
Figure 6 Learning test results on the first occasion of retest. Mean changes in memory
test scores after 4 treatments. Testing took place just before ECT and again 2 to 3 days
after the fourth ECT. Improvement is plotted upwards from the zero line and deteriora-
tion downwards. (Halliday, Davison, Browne, and Kreeger, Brit. J. Psychiat., 1968)

Non- Delayed Delayed

Digit Verbal verbal Delayed verbal non-verbal
5 span learning learning recall learning learning
4
3 Left
Right
~
2
1 Bilateral

1
0
1
2
3
4
5
6
7 N=44
L/R NS <0.05 NS NS <0.005 NS L11
L/8 NS NS NS NS NS NS R16
R/8 <0.025 NS NS NS NS <0.05 817
Figure 7 Learning test results at three-month follow-up. Mean changes in memory
test scores at three-month follow-up. Improvement is plotted upwards from the zero
line and deterioration downwards. (Halliday, Davison, Browne, and Kreeger, Brit. J.
Psychiat., 1968)
Side-Effects of ECT 11 I

UNILATERAL EST BILATERAL EST

Right

aWords
I Forms

+10 .....-r--,.......--r-....-...-,.----.--.-...-.--,.....,.--r-...----,-,l"""'T""T""-,--,....,l"""'T-
12345678 12345678 12345678
INDIVIDUAL PATIENTS

Figure 8 Performance of three groups on verbal and non-verbal learning tests. Indi-
vidual post-ECT decrements on the verbal and non-verbal tests for the three treatment
groups. Analysis of variance showed a significant effect for treatment group and a sig-
nificant groups by task interaction. (Cohen, Noblin, Silverman, and Penick, Science,
1968; reproduced by permission of the American Association for the Advancement
of Science.)

verbal tests which were, in fact, less severely impaired than was the case for
either of the other two types of ECT. It can be seen that this latter result is
not exactly in accord with expectation.
The results at the time of the three-month follow-up are shown in Figure
7. It can be seen that the verbal learning impairment was still evident in the
patients that had received dominant hemisphere ECT, but impairment on
non-verbal learning tests seemed to be more persistent after bilateral ECT
rather than after unilateral ECT to the non-dominant hemisphere.
In one other study Cohen, Noblin, Silverman, and Penick (I 968) tested 24
depressed, right-handed female patients, randomly sorted into the same three
treatment groups for ECT. Each patient was tested before treatment and then
retested from 5 to 8 hours after the fifth ECT. Two kinds of test item were
used. One was a verbal learning test involving the paired association of un-
related words, and the other was a non-verbal learning test that required the
completion from memory of a series of drawn designs. Performance on these
tests after treatment is shown in Figure 8.
It can be seen from this figure that the results, even for individual patients,
are remarkably clear-cut. The dominant group does much more poorly on the
112 A Triune Concept of the Brain and Behaviour

verbal learning test after shock; the non-dominant group does much more
poorly on the non-verbal test, while the bilateral group does badly on both
kinds of learning after ECT.
The results of shock on mood were not reported in this study.

CONCLUSIONS

It may be concluded that the evidence from three different sources, physical,
physiological, and psychological, makes it seem likely that the actual position
of the electrodes on the head in the different forms of ECT may be of quite
critical importance in the production of amnesic side-effects in man. The
brain structures that bear the main force of the electrical assault under the
standard conditions of bilateral and unilateral ECT are particularly liable to
such interference and are, at the same time, crucial elements in the mechan-
ism mediating human learning and memory.
If this inference is correct, it strongly suggests the need for a systematic
study of other variants of ECT technique. In one such adaptation the elec-
trodes might, for example, be placed as near the upper part of the front of
the head as is consistent with the actual production of convulsion, thus
keeping the local effects of stimulation as far away as may be possible from
the sensitive regions of the temporal lobes .
On the basis of the evidence reviewed, and the admittedly speculative
inferences drawn therefrom, it would be predicted that such frontally pro-
duced convulsions should improve mood but have significantly less effect on
learning and memory as compared to any of the techniques that have been
systematically studied so far.

SUMMARY

It is contended that some of the incidental and transient effects of ECT on

learning and memory closely resemble those more severe and chronic impair-
ments that are known to be the side-effects of temporal lobectomy in man.
This contention may be supported by three kinds of evidence. At the simplest
physical level the two kinds of procedure are carried out at similar sites on
the human head. From the physiological evidence it appears that electrical
stimulation has a local as well as a general action on the brain. The areas most
likely to be affected by ECT lie within the temporal lobes and the most
probable result of their disturbance is some form of amnesic disorder. The
psychological evidence points to close similarities betwen the behavioural
Side-Effects of ECT I 13

effects of shock and surgery. Both kinds of interference on the dominant side
of the brain produce defects of verbal learning; on the non-dominant side
they produce defects of non-verbal learning. These parallels imply a pressing
need for the systematic study of other modes of ECT that would interfere as
little as possible with the normal activity of those parts of the human brain
that are essential for adequate learning and memory function.

REFERENCES

Alexander, L., Treatment of Mental Disorder. Philadelphia : Saunders, 1953

Bickford, R.G ., D.W. Mulder, H.W. Dodge, H.J. Svien, and H.P. Rome, Changes in
memory function produced by electrical stimulation of the temporal lobe in man.
Res. Pub/. Assoc. Res. Nervous Mental Disease, 195 8, 36: 227-43
Blaurock, M.F ., F .M. Lorimer, M.M. Segal, and F.A . Gibbs, Focal electro-
encephalographic changes in unilateral electric convulsion therapy. AMA Arch.
Neurol. Psychiat., 1950, 64:220-6
Brazier, M.A.B., Stimulation of the hippocampus in man using implanted electrodes. In
M.A.B. Brazier (ed.), Brain Function, vol. 2: RNA and Brain Function, Memory and
Learning. Los Angeles: University of California Press, 1964
Brierley, J .B., The neuropathology of amnesic states. In C.W.M . Whitty and O.L.
Zangwill (eds.),Amnesia. London: Butterworths, 1966a
Brierley, J .B. Some aspects of the disorders of memory due to brain damage. In
D. Richter (ed.), Aspects of Learning and Memory. New York : Basic Books, 1966b
Cohen, B.D., C.D. Noblin, A.J. Silverman, and S.B. Penick, Functional asymmetry of the
human brain. Science, 1968, 162:475-7
Corkin, S., Tactually-guided maze learning in man: effects of unilateral cortical excisions
and bilateral hippocampal lesions. Neuropsychologia, 1965, 3:339-52
Halliday, A.M ., K . Davison, M.W. Browne, and L.C. Kreeger, A comparison of the effects
on depression and memory of bilateral ECT and unila.~ral ECT to the dominant and
non-dominant hemispheres. Brit. J. Psychiat., 1968, 114:997-1012
Hill, D. Discussion on the surgery of temporal lobe epilepsy: the clinical study and selec-
tion of patients. Proc. Roy. Soc. Med., 1953, 46 :965-71
Jasper, H.H. and T. Rasmussen, Studies of clinical and electrical responses to deep
temporal stimulation in man with some consideration of functional anatomy. Res.
Publ. Assoc. Res. Nervous Mental Disease, 1958, 36 :316-34
Kalinowsky, L.B ., The convulsive therapies. In A.M . Freedman and HJ. Kaplan (eds.),
Comprehensive Textbook of Psychiatry. Baltimore: Williams and Wilkins, 1967
Kampmeier, O .F., A.R. Cooper, and T .S. Jones, A Frontal Section Anatomy of the Head
and Neck. Urbana, Ill.: University of Illinois Press, 1957
Kimura, D., Right temporal-lobe damage: perception of unfamiliar stimuli after damage.
Arch. Neurol. , 1963, 8: 264-71
Lancaster, N.P., R.R . Steinert, and I. Frost, Unilateral electroconvulsive therapy.
J. Mental Sci., 1958, 104:221-7
Liberson, W.T. and K. Akert, Observations of electrical activity of the hippocampus,
thalamus, striatum and cortex under resting conditions and during experimental
seizure states in guinea pigs. Electroencephalog. Clin. Neurophysiol.,
1953, 5 :320
Liberson, W.T . and J .G. Cadilhac, Electroshock and rhinencephalic seizure states.
Confinia Neurol. , 1953, 13: 278-86
114 A Triune Concept of the Brain and Behaviour

Liberson, W.T ., W.B. Scoville, and R.H . Dunsmore, Stimulation studies of the prefrontal
lobe and uncus in man. Electroencephalog. Clin. Neurophysiol., 1951, 3: 1-8
Martin, W.L., H.F. Ford, E.C. McDonald, and M.L. Towler, Clinical evaluation of uni-
lateral EST. Amer. J. Psychiat., 1965, 121: 1087-90
Meyer, V. Cognitive changes following temporal lobectomy for relief of temporal lobe
epilepsy.AMA Arch. Neurol. Psychiat., 1959, 81 :299-309
Meyer, V. and A.J. Yates, Intellectual changes following temporal lobectomy for psycho-
motor epilepsy; preliminary communication. J. Neural. Neurosurg. Psychiat., 1955,
18:44-52
Milner, B., Psychological defects produced by temporal lobe excision. Res. Pub/. Assoc.
Res. Nervous Mental Disease, 1958, 36: 244-57
Milner, B., The memory defect in bilateral hippocampal lesions. Psychiat. Res. Rept.,
1959, no. 11 :43-58
Milner, B., Laterality effects in audition. In V.B. Mountcastle (ed .), lnterhemispheric
Relations and Cerebral Dominance. Baltimore: Johns Hopkins Press, 196 2
Milner, B., Visually-guided maze learning in man : effects of bilateral hippocampal, bi-
lateral frontal, and unilateral cerebral lesions. Neuropsychologia, 1965, 3:317-38
Milner, B., Visual recognition and recall after right temporal-lobe excision in man.
Neuropsycho/ogia, 1968, 6: 191-209
Ottosson, J-0., Experimental studies of the mode of action of electroconvulsive therapy.
Acta Psychiat. Neurol. Scand., 1960, Suppl. 145
Penfield, W., Functional localization in temporal and deep Sylvian areas. Res. Pub/.
Assoc. Res. Nervous Mental Disease, 1958, 36 :210-26
Scoville, W.B., R.H. Dunsmore, W.T. Liberson, C.E. Henry, and A. Pepe, Observations on
medial temporal lobotomy and uncotomy in treatment of psychotic states: pre-
liminary review of 19 operative cases compared with 60 frontal lobotomy under-
cutting cases. Res. Pub/. Assoc. Res. Nervous Mental Disease, 1953, 31 : 347-69
Scoville, W.B . and B. Milner, Loss of memory after bilateral hippocampal lesions.
J. Neural. Neurosurg. Psychiat., 1957, 20 : 11-21
Valentine, M., K.M.G. Keddie, and D. Dunne, A comparison of techniques in electrer
convulsive therapy. Brit. J. Psychiat., 1968, 114 :989-96
Williams, M., The measurement of memory in clinical practice. Brit. J. Soc. Gin.
Psycho/. , 1968, 7: 19-34
Zamora, E.N. and R. Kaelbling, Memory and electroconvulsive therapy. Amer. J.
Psychiat., 1965, 122:546-54
Zangwill, O.L ., Some clinical applications of the Rey-Davis performance test. J. Mental
Sci., 1946, 92: 19-34
 ROGER BROUGHTON MD

8 Confusional Sleep Disorders:

Interrelationship with Memory
Consolidation and Retrieval in Sleep 1

I am very pleased to have this opportunity to describe a personal scientific

odyssey to members and friends of my alma mater.
During the last eight years, personal research into the mechanisms of
certain common sleep disorders (i.e., nocturnal enuresis, sleepwalking, the
classical nightmare (or incubus attack) of adults and the sleep terror (pavor
nocturnus) of children) has convinced me that they should in fact be con-
sidered as arousal disorders (Broughton 1968). This does not imply that
psychological factors play no important role in their genesis, but rather that
the prevalent belief that the attacks are precipitated by co-existent mental
activity is questionable. First, their symptomatology is largely explicable by a
combination of physiological changes predisposing to the particular attack
type. Second, there is no direct evidence that recurrent mental activity
actually triggers the attacks. This lack of recall, however, necessitates con-
sideration of the relationships between memory mtchanisms and sleep.
In this paper, I review the apparent pathogeneses of the attacks, the
confusional states in which they arise, the possibility that arousal from dif-
ferent sleep stages has relevance to species survival, and some aspects of
memory and sleep.

PATHOGENESIS OF THE ATTACKS

The apparent lack of an immediately preceding sequence of mental activity in

the genesis of the attacks is particularly surprising, because both traditional
and current hypotheses indicate that such activity should be found : enuresis
supposedly is due to aggression, Oedipus fixation or other psychodynamic

I would like to give thanks to the Medic.ii Research Council of Canada for support during
the different stages of this research, initially a Queen Elizabeth II Canadian Research
Fellowship, then a Medical Research Council Fellowship and, most recently , a Medical
Research Council Associateship. Mr Wayne Gibson was involved in a technical capacity
in the evoked potential and 'decision time' studies in Montreal
116 A Triune Concept of the Brain and Behaviour

factors (Hader 1965); sleepwalking, to the acting out of dream activity

(Andre-Balisaux and Gonsette 1956, Sours et al. 1963); and the incubus
attack and night terror, to terrifying dream-like experiences or to 'repressed
sexual impulses' (Freud). Both physicians and public alike generally have
considered these more or less abrupt disorders as an expression of at least
some type of co-existent mental activity, usually dreaming.
But rather than occurring in rapid eye movement (REM) sleep with its
well-documented high (80-95 per cent) incidence of dream recall on awaken-
ing, the four sleep disorders listed in our early studies almost always took
place coincident with the arousal terminating slow wave stages of NREM sleep
(and generally in the first cycle, when this arousal is often most intense).
Detailed descriptions have been previously reported (Gastaut et al. 1964,
Gastaut and Broughton 1965, Broughton 1968).
Of particular interest was the almost complete lack of reports of mental
activity when subjects were interrogated at the end of the different attacks.
The only reports with any detail followed the incubus attacks in adults and
pavor nocturnus in children. They consisted of poorly defined sensations of
'impending doom,' choking or difficulty in breathing - sometimes likened to
paralysis - and intense palpitations. These features sometimes were elab-
orated into descriptions of a particular situation or scene (such as being
locked in a cave, having a pile of rocks on the chest, etc.). Occasionally,
mixed or 'compound' attacks also were observed. A sleep terror, for instance,
might evolve into a sleepwalking attack. All four attacks contained 'dis-
sociative' aspects, either between the clinical state and EEG (incomplete wake-
fulness, with an EEG of sleep, or vice versa) or between individual clinical
features (some functions having reached levels suggestive of wakefulness,
others not).
We also found that physiological changes were present throughout the
night's sleep, independent of the attacks; and that these were of a type which
would predispose to the appropriate attack during any abrupt or intense
arousal. Enuretic children, for instance, had many primary detrusor contrac-
tions in slow wave sleep. These contractions were spontaneous or induced by
even minimal stimuli (e.g., 'clicks'). They were most prevalent during enuretic
nights and were independent of body movement, deep inspiration, or con-
traction of abdominal muscles. In control subjects, they were essentially
absent. In enuretics, the normal slow wave sleep arousal episodes were asso-
ciated with, and sometimes preceded by, increases in bladder pressure to
levels producing reflex micturition without waking the child. In fact, enuretic
children were extremely difficult to arouse and appeared relatively in-
sensitive to the visceral stimuli, as well as to the exteroceptive stimuli of a
wet bed (Broughton and Gastaut 1964; Gastaut and Broughton 1965;
Broughton 1968).
Confusional Sleep Disorders 117

Sleepwalkers showed a marked tendency towards prolonged complex

movements and gestures, often resembling waking behaviour, in slow wave
sleep without producing significant attenuation of the EEG. Quite the con-
trary, the EEG might contain an increase in amplitude of delta activity (Figure
3 in Gastaut and Broughton 1965). The subjects were difficult to arouse
during these complex movements and, on awakening, had no recall of their
behaviour. And when forcefully aroused at any time in slow wave sleep, the
sleepwalker displayed intense and prolonged confusion, usually in the form of
an actual sleepwalking episode (Gastaut et al. 1964, Gastaut and Broughton
1965, Broughton 1968). These findings have been confirmed by Jacobson et
al. (1965) and Jacobson and Kales (1967). The latter authors recently have
emphasized the marked and paradoxical increase in slow wave activity at the
beginning of arousal in NREM sleep, which to them suggested a relative 'func-
tional immaturity' of the cerebral cortex in sleepwalkers.
Finally, subjects with either classical incubus or pavor nocturnus attacks
showed extremely intense autonomic, muscular, and EEG activation during
the episodes. They also exhibited alterations in sleep patterns independent of
the attack, showing responses throughout slow wave sleep consisting of exag-
gerated tachycardia (from a low baseline) and possibly tachypnea and
muscular hypertonia to various stimuli (or even during apparently spon-
taneous incomplete arousals, with or without gross body movement). Fisher
et al. (1968, 1969) have reported similar findings.
Considerable evidence exists that mental activity resembling thinking, or
even of dream-like quality which is difficult to differentiate from that re-
called after REM awakenings, may occur in NREM sleep (Goodenough et al.
1959; Kamiya 1961; Foulkes 1962; Rechtschaffen et al. 1963; Foulkes and
Rechtschaffen 1964; Foulkes 1967; Shapiro 1967). It appears unlikely,
however, that there is any correlation between such activity and the pre-
cipitation of any of the above phenomena.
First, while quite high incidences of mental activity in NREM sleep have
been reported in normal subjects, this is not true for the sleep when these
disorders occur, i.e., stages III and 1v (Kales et al. 1967; Kales and Jacobson
1967). Moreover, recall of mental activity after NREM awakenings increases
with later cycles, whereas most attacks take place in the first or second
cycles. And individuals subject to these arousal disorders ('sleep disorders')
exhibit a particularly low incidence of recall when awakened from slow sleep,
and perhaps even REM sleep (Jacobson and Kales 1967).
Second, if N REM mental activity were the basis of the attacks, these should
be precipitated by a 'peaking' of such activity at any time in NREM sleep, or
at least in slow sleep. Instead, the attacks recur preferentially during the usual
cyclic slow wave arousal and appear to be related to accentuation of some
components of the arousal mechanism.
118 A Triune Concept of the Brain and Behaviour

Third, the occurrence of the attacks in slow sleep ('quiet sleep,' 'sound
sleep of the Yoga,' etc.) is when vigilance and cardiorespiratory functions are
at their minimum. Consequently, the greatest degrees and intensities of
arousal are possible.
Fourth, physiological changes during the attack explain most, if not all, of
the symptoms (phenomenon of somatization). The same changes featured in
an attack are present, to a lesser extent, throughout slow wave sleep and
would predispose to the particular attack pattern during any intense arousal
(i.e., bladder contractions and hyper-reactivity to enuresis; complex auto-
matisms and durable confusion to sleepwalking; and excessive cardiac, and
probably respiratory and muscular, reactivity to the incubus and pavor
nocturnus attacks).
Finally, and perhaps :nost importantly, forced arousal at any time in slow
sleep stages of NREM sleep may induce the attacks. If the arousal itself were
not the necessary substrate, the experimenter would have to postulate that he
has repeatedly decided to awaken subjects at just that moment when their
dreaming or other mental activity, unknown to him, was about to generate an
attack.
There is no reason, however, to doubt the psychodynamic interpretation
that these arousal disturbances are related to underlying psychic conflicts in
the subjects' lives. The actual constellation of pathophysiological changes
present may reflect adjustments in the 'weakest system' to nonspecific stress,
in the sense of Selye (I 956). These psychosomatic changes are sufficient to
explain the nature and mechanisms of the attacks without further pre-
supposing a psychogenic mechanism just before each episode. It must be
stated, however, that the physiological changes during attack and the co-
existent mental activity must interact. The mind-body dichotomy is false.
Moreover, aspects of behaviour during sufficiently prolonged confusion may
have an evident symbolic aspect, as in a young girl who, during her sleep-
walking episodes (presented only when her father was away), always crawled
into bed with her mother.

CONFUSIONAL AWAKENINGS

The common symptoms of the attacks are themselves interesting. They in-
clude: mental confusion and disorientation; automatic behaviour; relative
nonreactivity to external stimuli; difficulty in bringing subjects to a state of
full lucid consciousness; complete or partial retrograde amnesia for inter-
current events; and only fragmentary reports, if any, of mental activity.
Because the four disorders usually occur at the same moment in the waking-
sleep cycles, it seemed probable that the common symptoms were related to
the arousal itself.
Confusional Sleep Disorders 119

To test this hypothesis, subjects, most of them children (both normals and
sleepwalkers), were wakened during different stages of sleep, paying parti-
cular attention to differences between arousals in REM sleep and in slow wave
sleep (stages m and 1v). They were either forced to sit up in bed or lifted to
their feet and were asked to 'reveille-toi! reveille-toi!' Telemetry was used to
monitor the physiological data (Gastaut and Broughton 1965).
Arousal from REM sleep was fairly difficult, even in normals; but once
awake, the subjects were lucid and generally (85 per cent) recalled detailed
dreams. By comparison, arousal from slow wave sleep in most normals and in
all sleepwalkers led to prolonged confusional episodes with automatic be-
haviour, more or less complete retrograde amnesia and rarely (8 per cent in
normals) recall of any mental activity. When present, such activity lacked the
panoramic hallucinatory features and development in time characteristic of
most dreams and consisted of fragmentary sensations or merely 'thinking
about something.' The confusional states elicited from NREM sleep became
less marked during the night, and were essentially absent from awakenings
(usually stage 11) in the second half of sleep.
The slow wave confusional arousals were much more pronounced in the
sleepwalkers, many of whom, like some normals, actually exhibited complete
episodes of somnambulism. When fluids had been forced before retiring, the
automatisms often included a trip to the bathroom to void.
If one defines sleepwalking as an arousal from sleep in which subjects then
walk some distance in a complex sequence of behaviour usually with sub-
sequent return to bed or, if awakened, retrograde amnesia, then I would judge
that virtually all members of homo sapiens have experienced at least one
episode of somnambulism during their lives. This generalization is possible
because such a sequence is seen at some time in almost all young children
when they awaken, either by parental stimulation or spontaneously, to go to
the toilet.
Even in adults, however, confusional awakenings have been recognized for
centuries. They are described in French literature as l'ivresse du sommeil
(Marc 1840) and in German literature as Schlaftrunkenheit (Gudden 1905).
In fact, as early as 1897, de Manaceine reported having induced them experi-
mentally in man. The question of an individual's legal responsibility for his
actions during such a state is appraised in Marc's fascinating book De la Folie
(1840; p. 658) written to instruct his patient, Roi Philippe III of France.
That confusional awakenings have medico-legal pertinence is well il-
lustrated by the famous case of Schimaidzig who awoke from sleep, saw a
ghost at the end of his bed, picked up a hatchet and gravely wounded his
wife. In my experience, however, aggression has not been a feature of
these confusional arousals . And it has been reported recently that fewer
than two dozen such cases are in the forensic literature. Nevertheless, the
120 A Triune Concept of the Brain and Behaviour

psychological literature abounds in bizarre and often spectacular behaviour

following sudden partial arousal from sleep. Probably almost all of these
occur in slow wave sleep.
It is not difficult to appreciate how important such confusional awaken-
ings might be in situations where an immediate high level of performance is
required. For example, anyone who must wake up and fight a fire , look at a
radar screen, fly a combat plane, or adjust a space vehicle, would be at a
considerable disadvantage , if awakened from slow wave sleep in the first
hours after retiring.
Marie Feltin, now a medical student, made a detailed study under my
supervision at the Montreal Neurological Institute of the behavioural effects
of arousal from REM sleep and slow wave sleep (Feltin and Broughton 1968).
Young paid adults, mostly university students, were trained to respond to
stroboscopic flashes (which elicited a visual evoked potential) by closing a
low inertia relay taped to a thumb . This response was to follow only white
flashes (which appeared in random sequence with blue) to give a combination
of differentiation and reaction which might be called a 'decision time,' or
index of 'critical reactivity' (Pithon). Subjects were tested during pre-sleep
wakefulness, following arousal from stage 1v and REM sleep and in a post-
sleep control, always with eyes open . Fifteen studies were performed in seven
subjects giving a total of approximately 4500 measurements.
The subjects responded to flashes either quite quickly (within 250-700
msec) or after l sec (or not at all). The latter two (I sec or more) were
grouped together as 'overshoots' and appeared to reflect inattention, rather
than true prolonged decision times. They were rare during wakefulness (ap-
proximately l per cent), very frequent initially after slow wave sleep arousal
(16 per cent), and significantly less frequent in early post-REM versus slow
wave sleep arousal (8 per cent incidence, p < 0.05), both being much greater
than wakefulness (p < 0.001).
The mean 'decision times' were relatively short in wakefulness (399
msec ± 60 msec) and were not significantly prolonged initially after REM
arousal (443 msec ± 71, p > 0.05). However, after the first two minutes or so
of post-REM wakefulness, performance dropped off. This was possibly be-
cause unrewarded subjects are unmotivated and so tend to go back to sleep
( confirmed by inspection of EEG). By comparison, arousal from slow wave
sleep produced a greater degree of impaired function (average 510 msec ± 75)
and the arousal progressed slowly.
There are interesting correlations between cerebral evoked potentials and
behaviour revealed by arousal studies. The occipital visual evoked potential
following arousal from stages III and IV often shows either a carry-over into
the confusional period of wave forms characteristic of slow wave sleep, or
Confusional Sleep Disorders 121

simply increased latencies and decreased amplitude of later components

(Broughton 1968, Saier et al. 1968). By comparison, arousal from REM sleep
results in more or less immediate return of the occipital wave fonn resembling
that of wakefulness, but sometimes is decreased in amplitude for a variable
period of time. The carry-over of NREM sleep components and the increased
latencies of later components evolving towards a VEP typical of presleep
wakefulness roughly parallelled the degree and evolution of the confusion
(Broughton 1968).

POSSIBLE EVOLUTIONARY IMPLICATIONS

Snyder recently ( 1966) proposed that REM ( or paradoxical) sleep has a sur-
vival function for mammals. His very attractive 'sentinel hypothesis' postu-
lates that paradoxical sleep is a recurrent cyclic state preparing the brain for
optimal environmental scanning. This scanning during the usual brief awaken-
ings following paradoxical sleep, would allow detection of any predators, and
the animal would be ready for fight or flight. Also, paradoxical sleep would
insure that the animal does not experience prolonged periods of very low
levels of consciousness, which could be detrimental to his survival. Rather, in
this sleep he remains relatively quiescent, thus not attracting any predators
and benefits from both a reasonable continuity of sleep and a cyclic, brief,
and fairly optimal ability to assess his environment for danger.
Our own studies, in which subjects following REM sleep awakenings
displayed relatively high performance levels on testing and had a more lucid
mental state, appear to be supporting evidence for this hypothesis.
Mr Wayne Gibson and I recently attempted to verify this theoretical pro-
posal by investigating the effects of the presence of a predator upon the sleep
of a prey animal (Gibson and Broughton 1969). The 'sentinel hypothesis'
might predict any of the following: (i) a selective proportional increase of
paradoxical sleep ( REM sleep); (ii) a selective decrease of total slow sleep
(NREM sleep); (iii) more frequent or longer post-paradoxical awakenings; (iv)
more frequent awakenings interrupting paradoxical sleep; or (v) a shorter
cycle length, producing more frequent periods of paradoxical sleep.
The sleep of male albino rats has been investigated recording cortical and
hippocampal activity, a horizontal oculogram and nuchal EMG from 11.00
a.m. to 4.00 p.m., when rats (nocturnal animals) prefer to sleep. In control
studies, sleep in the rat is quite stable, with slow sleep representing 88. l per
cent± 2.4 and paradoxical sleep 11.9 per cent± 2.4 of total sleep; and wake-
fulness being 34. l per cent ± 10.1 of total recording time.
After control studies had indicated that the well-fed laboratory cat would
not harm the laboratory rat when they were placed together, cats were
122 A Triune Concept of the Brain and Behaviour

introduced free in the same cage with the rats, after the latter had fallen
asleep. This produced no significant changes in the rats' sleep patterns, slow
sleep being 87 .8 per cent and paradoxical sleep 12.2 per cent of sleep, with
wakefulness 36.4 per cent of total recording time. We believed that this
negative result was both because the animals were naive and in-bred for
laboratory purposes, the true prey-predator relationship not being present,
and because the cats were not hungry.
Cats were then chronically implanted with bipolar electrodes in the dorso-
Jateral hypothalamus. Stimulation (60 HZ, 2 msec, 3 ma) of the cat in this
area produced direct attack without rage and would have Jed to kill, if the
current were maintained. Experiments were carried to this conclusion by
Wasman and Flynn (I 962). As the attack behaviour terminates when the
stimulation ends, this is a reliable technique for controlling the 'threat value'
of a cat without endangering the rat.
When cats were placed beside or in the rats' cage and stimulated once prior
to sleep of the latter, paradoxical sleep was completely suppressed for 2-3
hours. Slow sleep became 100 per cent of sleep and recurred fairly cyclically
but with more frequent awakenings. Some cases resulted in even longer sup-
pression of paradoxical sleep. Usually, however, paradoxical sleep reappeared
towards the end of the recording session leading to over-all statistics of 0.9
per cent of total recording time (versus a control value of 8.1 per cent), and
6.6 per cent of total sleep. Wakefulness increased to 86.4 per cent of the total
recording time. Almost all of the prolonged periods of wakefulness began in
slow sleep. When the cat was introduced after the rat had fallen asleep and
then stimulated as nearly as possible to 20-minute intervals during the rat's
slow sleep, paradoxical sleep decreased even more markedly (paradoxical
sleep 3.1 per cent and slow sleep 96.9 per cent of total sleep). Wakefulness
was then 71.7 per cent of total recording time. We were able in some ex-
periments to suppress paradoxical sleep completely for over 7 hours.
Several interesting behavioural phenomena have also been observed. In one
of the studies, a cat caged with a rat nipped the rat's tail. The rat then ran to
the far corner of the recording area and showed signs of stress (3 fecal
pellets). The rat was subsequently insomniac, and reasonably so, until the cat
fell asleep. Once the cat was asleep, the rat went to sleep. At the end of his
first slow wave period, the rat awoke and, apparently confused , ran to and bit
the cat. The cat awoke and hissed, and the rat scurried once again to the far
corner and defecated. This episode seemed analagous to the confusional slow
wave arousals in man, and obviously resembles most closely those rare reports
of arousal with aggressive behaviour.
In sum then , the effects of a threatening cat in the immediate environment
of a rat were an increase in wakefulness by suppression of paradoxical sleep
 Confusional Sleep Disorders 123

and, to a lesser extent, slow sleep; much more frequent awakenings from slow
sleep; and several prolonged periods of insomnia.
I find it difficult to reconcile the experimental data of a more or less
selective suppression of paradoxical sleep (sometimes with continuation of
cyclic slow wave sleep) with the teleological theory that paradoxical sleep is a
preparatory state prior to scanning utilized for species survival. Threatened
rats do not awaken more frequently during, nor do they remain awake longer
after, or increase their proportion of, paradoxical sleep. Instead, paradoxical
sleep is suppressed and almost all periods of wakefulness are from slow sleep.
It seems more reasonable to suppose that what is needed is wakefulness, for
environmental scanning and detection of danger, plus slow sleep, which is
known to be related to recuperation from muscular exercise (Hobson 1968).
It might be possible, however, that under chronic stress paradoxical sleep
would rebound to a higher proportion than baseline levels and then have
survival value. We intend to initiate studies of chronic predatory threat.

MEMORY AND SLEEP

The last aspect to be considered concerns the inter-relationships between

memory and sleep. In both reports of others and introspection of my own
sleep, I have been consistently impressed by the wealth of such rela-
tionships. First, many of our more vivid dreams, especially those con-
cerning childhood events, resemble endogenous playback of 'film strips' of
the past, although they are also manifestly influenced by environmental ef-
fects preceding and during sleep 2 • Second, there is the retrograde amnesia
of recent events with passage into stage u of NREM sleep, especially well
documented for the loss of recall of all previous post-REM sleep reports
(Wolpert and Trosman 1958). And third, there is the profound amnesia of the
confusional disorders with arousal from slow wave sleep, which we have just
considered.
There are, of course, many other interesting inter-relationships such as the
possibility of learning during sleep, which certainly necessitates a capacity for
memory, and the marked effects upon memory and learning of fatigue from
sleep deprivation. Also, the possibility of an important amount of memory
consolidation occurring in sleep is not without experimental evidence (cf.,
Kleitman 1963).

2 The facts that certain brain structures, mainly of the so called limbic system, important
in the physiology of both REM sleep and waking memory mechanisms are essentially
identical, and that the substrates of these two behavioural events appear to have similar
neurochemical bases, are undoubtedly not fortuitous. They point again to the close
inter-relationships between memory and sleep.
124 A Triune Concept of the Brain and Behaviour

When such an obvious correlation of two very fundamental interesting be-

havioural phenomena exists, it is extremely curious that this is given such scant
attention in literature on memory. The evidence for the existence of at least two
types of memory (short-term memory with its consolidation and a long-term
memory, cf., review by Weiskranz 1966 and Deutsch 1962) is impressive ; but it
comes almost entirely from experimental studies in animals and from such
frankly abnormal clinical conditions as head trauma, electroconvulsive therapy,
toxic and metabolic states, alcoholic encephalopathies, deficiency diseases,
brain infections, cerebral tumours, epilepsy, effects of neurosurgery, etc.
It is not my intention to deprecate the importance of such investigations.
But it does seem evident that the same two fundamental components are
equally well demonstrated in a much more natural setting, i.e., during that
third of our existence that we call sleep. A specific example : after a dream
report has been gathered in a post-REM awakening, the ability to recall the
dream, or the verbal description of it, falls off very rapidly with subsequent
NREM sleep. But the subjects, when awakened, can recall their general fund of
knowledge. Short-term and long-term memory are thereby clearly and
physiologically dissociated.
Perhaps the most pertinent relationship to consider in some detail in this pre-
sentation is the amnesia with arousal from slow wave sleep. This is a phenome-
non seen to a greater or lesser extent in all subjects. But it is most marked in the
first and second cycles of sleep, in children and in individuals suffering from the
arousal disorders already discussed. The amnesia is so profound that, for
instance, subjects do not even remember a complete ten-minute episode of
sleepwalking in which the investigators may have repeatedly attempted to
attract the somnambulist's attention or used very bright lights for filming and
brisk shaking for awakening. If the subjects cannot remember such intense
experiences, how could they be expected to remember any preceding dreams
or other mental activity, possibly related to the attack?
This question is legitimate and germane to the general relative paucity of re-
call of all types of mental activity from slow sleep or even NREM sleep in general.
The point has been raised before (Kleitman 1961 ; Oswald 1962; Gastaut and
Broughton 1965; Clemente 1967; Brazier 1967; Broughton 1968) and still has
not been satisfactorily resolved. (Mention should also be made of the fact that
the extent and degree of recall of mental activity in NREM sleep has not been
fully agreed upon. This does not in any way affect the findings that physio-
logical changes predisposing to an attack type take place on a broad time
base, at least throughout slow sleep, and that endogenous or externally in-
duced arousal is generally sufficient to precipitate them).
Confusional amnesia frequently occurs during either excessive arousal or
insufficient arousal in a number of conditions. Examples of excessive arousal
 Confusional Sleep Disorders 125

include the various battle shock states and the conditions of extreme anxiety
sometimes, and perhaps erroneously referred to as 'hysterical amnesia.' There
are also a number of examples of insufficient arousal producing confusional
amnesic states: post-traumatic and drug-induced states, post-epileptic seizure
states, ictal epileptic 'twilight' states of temporal lobe, frontal lobe, or 'petit
mat' origin, the effects of brainstem lesions, certain depressive amnesic
illnesses, and so forth .
The experimental data of these arousal disorders suggest that the incubus
and pavor nocturnus attacks represent disorders of excessive arousal, and
enuresis and sleepwalking disorders of insufficient arousal. Dement (I 967)
has made the interesting suggestion that sleepwalkers appear like thalamic
animals, with the brainstem and diencephalon functioning as in wakefulness
but with cortico-subcortical mechanisms not yet re-established. 3
The transitory confusional state of arousal from slow wave sleep appears
to make retrieval of short-term memory impossible. The reason is still un-
clear. Short-term memory might be absent either because the individual's
perception is very limited or disturbed (there is evidence for both), or because
the mechanisms of consolidation are not functioning. Perhaps memory re-
trieval itself is impaired, and the experiences in the lacunar period may be
available later with different types of access (perhaps light barbiturate nar-
cosis, hypnotic trance during dreams, and so forth) . Or combinations of these
or other mechanisms may be at play . In fact, we simply do not know .
Here, then, is an example of a non-organic memory impairment seen in
normal lives of normal subjects which is of considerable apparent interest to
the biology of memory and which has received virtually no attention.

REFERENCES

Andre-Balisaux, G. and R. Gonsette, L'electroencephalographie dans le somnambulisme

et sa valeur pour l'etablissement d'un diagnostic etiologique. Acta Neural. Psychiat.
Belg., 1956, 56 :270-81
Brazier, M., Absence of dreaming or failure to recall? Exp. Neural., 1967, Suppl. 4 :91-8
Broughton, R.J ., Sleep disorders: Disorders of arousal? Science, 1968, 159: 1070-8
Clemente, C. (ed.), Physiological Correlates of Dreaming. (Exp. Neural. , 1967, Suppl. 4)

3 Of course, the arousal disorders from slow sleep can be contrasted to: (l) dissociative
arousal disorders from REM sleep, that is, hypnopompic sleep paralysis; (2) the in-
appropriate triggering of some aspect of sleep during wakefulness, such as cataplexy,
hypnagogic sleep paralysis or excessive hypnagogic hallucinations; or (3) the induction
of sleep itself, as in the narcoleptic attack. All of these phenomena indicate breakdown
of the normal integrated mechanisms of the two types of sleep and wakefulness.
126 A Triune Concept of the Brain and Behaviour

de Manaceine, M., Sleep, its Physiology, Pathology, Hygiene and Psychology. London :
W. Scott, 1897
Dement, W., in discussion of A. Jacobson and A. Kales, Somnambulism: all-night EEG
and related studies. Res. Pub/. Assoc. Res. Nervous Mental Disease, 1967, 45 :449
Deutsch, J.A., Higher nervous function: The physiological bases of memory . Ann. Rev.
Physiol., 1962, 24 :259-86
Fellin, M. and R.J . Broughton, Differential effects of arousal from slow wave sleep and
REM sleep. Psychophysiology, 1968, 5:231
Fisher, C. , J .V. Byrne, and A. Edwards, REM and NREM nightmares. Psychophysiology,
1968, 5:221
Fisher, C., J.V. Byrne, A. Edwards, And E. Kahn, paper presented at the 9th annual
meeting of the Association for the Psychophysiological Study of Sleep, Boston,
March 1969. Psychophysiology (in press)
Foulkes, W.D., Dream reports from different stages of sleep. J. Abnorm. Soc. Psycho/.,
1962, 65: 14-25
Foulkes, W.D ., Nonrapid eye movement mentation. Exp. Neural., 1967, Suppl. 4:28-38
Foulkes, W.D. and A. Rechtschaffen, Presleep determinants of dream content: effect of
two films. Percept. Motor Skills, 1964, 19:983-1005
Freud, S., The Interpretation of Dreams. Translated and edited by J. Strachey. New
York: Basic Books, 1955
Gastaut, H. and R. Broughton, A clinical and polygraphic study of episodic phenomena
during sleep. In J. Wortis (ed.), Recent Advances in Biological Psychiatry, vol. 7. New
York: Plenum, 1965. Pp . 197-221
Gibson, W. and R. Broughton, Effect of a predator on the sleep of a prey . Presented at
the 9th annual meeting of the Association for the Psychological Study of Sleep,
Boston, March 1969. Psychophysiology (in press)
Goodenough, D.R ., A. Shapiro, M. Holden, and L. Steinshriber, A comparison of
'dreamers' and 'non-dreamers': Eye movements, electroencephalograms, and the
recall of dreams. J. Abnorm. Soc. Psycho/., 1959, 59:295-302
Gudden, H., Die physiologische und pathologische Schlaftrunkenheit. Arch. Psychiat.
Nervenkrankh., 1905, 40: 989-1015
Hader, M., Persistent enuresis. Arch. Gen. Psychiat., 1965, 13: 296-8
Jacobson, A. and A. Kales, Somnambulism: all-night EEG and related studies. Res. Pub/.
Assoc. Res. Nervous Mental Disease, 1967, 45:424-55
Jacobson, A., A. Kales, D. Lehmann, and J.R . Zweizig, Somnambulism: all night electro-
encephalographic studies. Science, 1965, 148:975
Kales, A., F.S. Hoedemaker, A. Jacobson, J.D. Kales, M.J . Paulson, and T.E. Wilson,
Mentation during sleep: REM and NREM recall reports. Percept. Motor Skills, 196 7,
24:555-60
Kales, A. and A. Jacobson, Mental activity during sleep : recall studies, somnambulism,
and effects of rapid eye movement deprivation and drugs. Exp. Neural., 1967, Suppl.
4:81-91
Kamiya, J., Behavioral, subjective, and physiological aspects of drowsiness and sleep. In
D.W. Fiske and S.R . Maddi (eds.), Functions of Varied Experience. Homewood, Ill. :
Dorsey, 1961. Pp . 145-74
Kleitrnan, N., The nature of dreaming. In G.E.W. Wolstenholme and M. O'Connor (eds.),
Ciba Foundation Symposium on the Nature of Sleep, London, 1960. London:
Churchill, 1961. Pp. 349-74
Kleitrnan, N., Sleep and Wakefulness. Rev. and enl. ed. Chicago: University of Chicago
Press, 1963
Marc, C., De la Folie. Paris: Bailliere, 1840
Oswald, I., Sleeping and Waking; Physiology and Psychology. Amsterdam : Elsevier, 1962
Rechtschaffen, A., P. Verdone, and J. Wheaton, Reports of mental activity during sleep.
Can. Psychiat. Assoc. J., 1963, 8:409-14
Confusional Sleep Disorders 127

Selye, H., The Stress of Life. New York: McGraw-Hill, 1956

Shapiro, A., Dreaming and the physiology of sleep. Exp. Neuro/. , 1967, Suppl. 4:56-81
Snyder, F., Toward an evolutionary theory of dreaming. Amer. J. Psychiat. , 1966,
123 : 121-36
Sours, J.A ., P. Frumkin, and R.R . lndennill, Somnambulism: its clinical significance and
dynamic meaning in late adolescence and adulthood. Arch. Gen. Psychiat., 1963,
9:400-13
Wasman, M . and J.P. Flynn, Directed attack elicited from hypothalmus. Arch. Neurol.,
1962, 6:220-7
Weiskrantz, L. , Experimental studies in amnesia. In C.W.M . Whitty and O.L. Zangwill
(eds.),Amnesia. London: Butterworths, 1966. Pp. 1-35
Wolpert, E.A. and H. Trosman, Studies in psychophysiology of dreams: I. Experimental
evocation of sequential dream episodes. AU4 Arch. Neuro/. Psychiat., 1958,
79:603-6
 S.G.LAVERTY

9 Sleep Disorders and Delirium Associated

with the Use of Ethanol'

The purpose of this paper is to examine the effects of ethanol intoxication

and withdrawal on patterns of sleep, with particular reference to delirium
tremens. A general description of sleep disturbances given by drinkers is
followed by consideration of two components of sleep, the REM or dreaming
state, and slow-wave sleep.

GENERAL DESCRIPTION OF SLEEP DISORDERS

IN RELATION TO ETHANOL USE

Reports of sleep disturbances following ethanol use fall into four main cate-
gories: difficulty in going to sleep; interrupted sleep and early waking;
sensory phenomena including dreams, nightmares, illusions, and hallu-
cinations; motor restlessness.
Difficulty in falling asleep is described, often by inexperienced drinkers,
and may be associated with sensations of giddiness (with eyes closed, less
often when open) when the subject lies down, and as drowsiness appears. The
subject may be aroused with impending nausea and a strong sense of rotation.
Such reports may be seen associated with the nystagmoid eye movements
reported by Goldberg and his colleagues (Aschan, Bergstedt, Goldberg, and
Laurell 1956; Goldberg 1966), of which two varieties are described (posi-
tional alcohol nystagmus and alcohol gaze nystagmus), the former appearing
during both intoxication and late withdrawal phases (PAN I, PAN 11). Alcohol
gaze nystagmus is related to dosage of alcohol; higher dosage is associated
with dizziness and vertigo.
In contrast, alcohol is frequently described as a reliable sedative enabling
the subject to secure a rapid and easy transition into sleep.

This study has been supported by the Addiction Research Foundation of Ontario. The
work has been carried out in the Addiction Studies Unit, Queen's University, Kingston
Psychiatric Hospital
 Sleep Disorders and Delirium 129

Interrupted sleep following drinking may be associated with dreams or

nightmares, or with symptoms such as tachycardia, nausea, anxiety, and rest•
lessness. It is tempting to see this as a rebound phenomenon comparable to
the alcohol withdrawal syndrome . Evidence of such 'rebound states' may be
drawn from Goldberg's description of PAN II, and from the compensatory
rebound described for the REM state following deprivation or suppression
(Dement 1965) by moons which include ethanol administration (Gresham,
Webb, and Williams 1963 ; Yules, Freedman, and Chandler 1966; Yules,
Ogden, Gault, and Freedman 1966; Knowles, Laverty, and Kuechler 1968).
This form of sleep disturbance is a well recognized symptom of mental de-
pression (Kraines 1957), a temporary state frequently observed in the later
stages of heavy bouts of drinking.
Sensory phenomena related to sleep include dream experiences, and ab•
normal waking perceptions (illusions, hallucinations). Very intense dreams,
usually unpleasant, are described by heavy drinkers, and their reports are
familiar to the night-staff of an alcohol drying-out centre in the early hours of
the morning. They may precede waking for days on end while a bout is
continuing, and may be related to periods of wakefulness during the nights of
withdrawal. Such phenomena are often associated with motor restlessness and
the trembling sensations that constitute the commonest symptom of alcohol
withdrawal, objectively seen in the form of muscular tremors and twitchings.
The most dramatic form of sleep disorder associated with ethanol in-
toxication and withdrawal is the state of delirium tremens. This usually ap•
pears in subjects who have been drinking heavily for many days or weeks.2
The age of onset is usually in the fourth decade. A consistent picture is drawn
by many authors (Engel 1967; Thompson 1959; Gross, Goodenough, Tobin,
Halpert, Leport, Pearlstein, Sirota, Dibianco, Fuller, and Kishner 1966;
Kraepelin 19 l 2). Symptoms are more pronounced late in the day and at
night , and are made worse by a reduction in external stimuli, for example in a
dark room or when the eyes are closed or covered.
Tremor, restlessness, hallucinations, and intellectual impairment, often
with clouding of consciousness and confusion, are the central symptoms.
Insomnia is a rule and the condition may resolve with a deep terminal sleep.
The history of drinking prior to the onset of delirium is usually that of a
heavy bout with increasing quantities being consumed over a period, but it may
also occur in steady drinkers usually at the point of withdrawal, or within a few
days. The contributory effects of such conditions as pneumonia and anxiety
have been described (Kat and Prick 1940; Dynes 1939; Moore and Gray 1939).

2 However, a few susceptible subjects may show hallucinatory and delusional symptoms
even after one session of drinking
130 A Triune Concept of the Brain and Behaviour

The bout of drinking leading up to delirium presents some specific clinical

features which may throw light on the mechanisms involved. Insomnia may
begin at the onset of drinking. Increasing quantities of alcohol are taken to
procure sleep. Sleep may then be apparently normal except that the con-
tinuation of alcohol is required to go to sleep. Sleep is interrupted earlier,
waking is more dramatic, and dreams or nightmares are described at the point
of waking. These interruptions become increasingly disturbing. Observations
through the night of patients who are permitted to continue after days of
drinking show a characteristic pattern. The subject takes a drink in order to
sleep and may then fall into a light stupor from which he can easily be
aroused. Spontaneous arousal frequently occurs. If the patient continues to
sleep for two or three hours, sudden awakening occurs usually associated with
the appearance of dreaming. On awakening, the patient often demands
further alcohol. Awakening may occur with a frightening experience which
continues into the waking state and may carry with it evidence of perceptual
anomalies such as perceiving movement or misinterpreting objects in the sur-
roundings. The patient may attempt to control these waking experiences by
drinking. Such experiences may suddenly develop the quality of hallu-
cinations and the patient may then emerge into the state of a true waking
hallucination.
The subject may report that drinking is no longer effective in controlling
his anxiety and the associated perceptual changes. A reduction of alcohol
consumption may be attempted usually by 'tapering off.' At this point of
relative withdrawal the hallucinatory experience may suddenly become ac-
centuated. This is a point at which the patient may seek admission to hospital
since, when he drinks, his symptoms are no longer relieved; intoxication may
become acute, and with abstinence a severe withdrawal syndrome develops.
Delirium may develop prior to complete withdrawal {Thomson 1959;
Kraepelin 1912; Kat and Prick 1940). In other instances the patient may
describe 'the onset of hallucinations during this second sleepless phase as
occurring during the night while they are alone, lying in bed with the lights
off and their eyes closed trying to fall asleep. For a time opening their eyes,
turning on the lights, seeking out companionship and/or the end of the night
may help their hallucinations; then after a brief period of only a day or two
these conditions no longer offer refuge' (Gross et al. 1966).
Visual hallucinations are most frequent, and diverse. Auditory hallucina-
tions occur and usually consist of occasional words or phrases but may be
more discursive. Tactile, olfactory, and gestatory hallucinations are less fre-
quent. Hallucinations may be associated with paranoid interpretations and
illusions, and complex hallucinatory sequences have been described.
In addition to the disturbances of sleep and the appearance of hallucina-
tions other symptoms may appear during the drinking bout. Irritability,
 Sleep Disorders and Delirium 131

tremor, general restlessness, and myoclonic jerkings of various muscle groups

may appear. Anorexia is usually present from the onset of drinking; sweating,
perspiration, and severe generalized weakness may be present. Marked effec-
tive changes with fluctuations occur: fear, anxiety, depression, and extreme
motor excitement.
Seizures, usually in the form of grand ma! convulsions occur, and may
herald the onset of delirium. There is usually no previous history of epilepsy
(Rosenbaum, Lewis, Piker, and Goldman 1938).
The withdrawal syndrome may be represented by all or a few of the
foregoing symptoms and it would appear that the condition will be more
severe in proportion to the degree and duration of drinking, and may range in
severity from a mild hangover to the most severe form of withdrawal mani-
festation, namely epileptic seizures and delirium tremens. The experimental
reproduction of these symptoms, including delirium, by Isbell and his asso-
ciates (Isbell, Fraser, Wikler, Belleville, and Eisenman 1955) is the strongest
support for this view. Mendelson and La Dou (1964) have shown that dis-
organization from alcohol withdrawal may begin while heavy drinking is in
progress. This implies that the withdrawal syndrome increases during the
development of tolerance.

CHANGES IN STAGE 1-REM AND SLOW-WAVE SLEEP

ASSOC IA TED WITH ETHANOL

It has recently been proposed that some features of delirium tremens are
related to specific disturbances of sleep, namely the suppression and rebound
of the REM-state or dreaming sleep.
Greenberg and Pearlman (I 967), and Gross et al. (I 966) first presented
evidence that a very high percentage of REM may be associated with the onset
of delirium tremens and argue that, in fact, the hallucinations of this con-
dition represent REM (dreaming activity) breaking forth from the confines of
sleep into wakefulness.
The validity of this proposition is supported by the observation of high
percentage REM sleep just prior to and in delirium. 3
3 It may be opportune to point to the difficulties attending the recording of REM sleep in
the delirious patient, who is restless, often disoriented, and hypersensitive to manipula-
tions such as electrode placements. Further, there is an intermittent degree of increased
muscle tone which makes use of that indicator as a sign of REM onset unreliable. The
first delirious patient assessed in our laboratory showed continuous eye movements
(both REM and nystagmoid) both awake and while drowsing, increased muscle tone and
restlessness, and a highly uncharacteristic EEG throughout the first night, during which
he alternately hallucinated and drowsed. In no part of this record was a typical sample
of normal sleep stages discernible. Yet on the subsequent night he slept well and showed
21 per cent REM sleep and I 8 per cent slow-wave sleep. If REM rebound occurred it
had been missed
132 A Triune Concept of the Brain and Behaviour

Greenberg and Pearlman ( 1967) and Gross et al. (I 966) observed sleep-
onset REM, with a high REM component of sleep (in some instances 100 per
cent of total sleeping time) in patients in delirium tremens. Three further
instances are described here.

Study 1

Sleep records were taken in subjects with delirium tremens, in hospital. Oc-
cipital EEG, lateral eye movement, muscle tone, heart rate, and respiration
were recorded on Grass and Beckman polygraphs. Criteria for REM and slow
wave sleep were derived from the papers of Dement and K.leitman (1957).
Three patients were studied at the onset of delirium precipitated by
alcohol withdrawal.
Subject 1. Male aged 27 years. Drinking heavily for 5 days and nights,
during which he reported virtually no sleep. Early on the morning prior to
admission he began to hallucinate, seeing 'a large cockroach the size of a cat'
beside his foot. When he closed his eyes a 'rush of faces like ghosts,' and a
great variety of other and unpleasant images 'crowded me in.' On the first
night of withdrawal the subject attempted to sleep, and continuous REM
activity appeared during the onset of stage l sleep. After 7 minutes the
subject woke hallucinating and could not sleep during the rest of the night.
On the second night 3 hours and 20 minutes of sleep was recorded (REM 58
per cent). Waking with dreams and subsequent hallucination occurred after 1
hour of sleep. REM onset at 3 minutes from initial sleep onset, and at 9
minutes following first waking . Subsequent nights are shown in Table l.
Subject 2. Male aged 67 years. Subject was seen after 10 days of heavy
drinking. Many previous episodes of withdrawal delirium. Hallucinations con-
sisted of web-like structures and small animals. First night of withdrawal, no
sleep; second night of withdrawal, 40 minutes of sleep with REM-onset co-
inciding with onset of sleep, but continuing evidence of muscle twitching and
no clear muscular relaxation. Third night, slept 6 hours; no slow wave sleep;
21 per cent REM sleep. Hallucinations reported before and after sleep.
Subsequent nights illustrated in Table 1 .
Subject 3. Male aged 43 years. Drinking steadily for 6 weeks. Slept for 3
hours 10 minutes on first night (partially intoxicated) with 30 per cent REM.
On sleeping for l hour during the next day, early REM-onset (within 5
minutes of sleep onset) and 62 per cent REM sleep. On the second night, 4
hours 17 minutes sleep, 48 per cent REM; third night, 5 hours 22 minutes
sleep, REM 26 per cent. Hallucination during first night, on waking from
dreaming, and for periods the next morning, 'faces and moving objects'; brief
hallucinations and illusions and false preceptions persisted for short periods
for the next 4 days (Table l ).
 Sleep Disorders and Delirium 133

TABLE 1
REM and slow-wave sleep in three delirious subjects

Subject 1 Subject 2 Subject 3

REM Slowt Total REM Slowt Total REM Slowt Total

(%) wave sleep (%) wave sleep (%) wave sleep
(%) (mins.) (%) (mins.) (%) (mins.)

Last night
of drinking H H 30H 190
62 60
(day)

Withdrawal
night
1 100" 0 7 0 0 0 48H 28 257
2 58*" 200 100*" 0 40 26H 14 322
3 24 22 313 21 H 0 360 18H 9 238
4 21 7 347 27 5 331 17H 11 240
5 18 4 371 20 7.5 349
.L
I Slow-wave sleep shown is the combined score of stage 3 and stage 4 sleep (usually
slow waves of l second or longer)
H Hallucination reported that night
* No muscle relaxation

These subjects showed high rates of REM sleep, particularly for short
periods during the period of delirium. However, the third subject reported
hallucinations on three nights when REM activity was in the normal range.
Two further studies of ethanol and sleep were made: one (study 2) con-
sisted of recording sleep activity in patients withdrawing from alcohol after
prolonged bouts of drinking. The other study (study 3) was an experimental
investigation of intoxication and withdrawal in volunteer subjects. In both
these studies EEG and eye-movements were used in determining stages of
sleep. Muscle tone was not recorded, the requisite apparatus not being
available.

Study 2

Eight male subjects between the ages of 28 and 51 years were investigated
during the phase of withdrawal from alcoholic bouts ranging from 5 to 37
days. All were alcohol addicts of at least 4 years' standing. The first night in
hospital was the last day of drinking and rated as the first night of withdrawal.
REM and slow-wave sleep (stage 3 + stage 4) were rated. Scores and means are
shown on Table 2. Wide variations between individual subjects were seen in
the first 5 nights of withdrawal both for REM and sws. Over the first four
nights of withdrawal REM showed a progressive increase (Table 3) for the
 TABLE 2
Percentage of total sleep spent in REM (R) and slow wave (S) sleep during ethanol withdrawal

Night : l 2 3 4 5 6 7 8

R(%) S(%)t R(%) S(%) R(%) S(%) R(%) S(%) R(%) S(%) R(%) S(%) R(%) S(%) R(%) S(%)

Subjects
l 17 9.5 23 20 25 26 33 33 27 25 16 10 23 15 33 3
2 0 0 8 2 23 3 19 0 22 4 17 7.5 17 0 24 0
3 0 0 13 0 23 0 29 2 22 0 14 7 28 l 28 8
4 9 7 17 25 18.5 14 28 18 9 7 21.5 4 22 12 23 9
5 5 0 8 l 14 2 28 24 21 20 26 5 24 2 18 3
6 so 2 46 6 21 4 18 18 18 17.5 28 3 15 6 17 11
7 7 0 6 16 26 5 25 5 20 5
8 17 12 27 48 30 30 32 7 21 9

Mean 13.1 3.8 18.6 15.6 22.6 10.5 26.5 10.4 20.1 10.9 18.9 6.0 21.5 6.0 23.8 5.6

Total sleep
time (mean
in mins) 160 323 278 342 351 338 344 366
...I Slow wave sleep shown is the combined score of stage 3 and stage 4 sleep (usually slow waves of 1 second or longer)
Sleep Disorders and Delirium 135

TABLE 3
Change in REM and slow-wave sleep in ethanol withdrawal

Nights : 1-2 2-3 3-4 4-5 5-<i 6-7 7-8

REM
N 8 8 8 8 6 6 6
X 2 1 3 6 3 3 3
y 6 7 5 2 3 3 3
p 0.145 0-035 0.363 0.145 0.675 0.656 0.656

sws
N 8 8 7 7 6 6 6
X 1 5 5 2 4 3 4
y 7 3 2 5 2 3 2
p 0-035 0.363 0.227 0.227 0.344 0.656 0.344

X =no. of subjects whose REM or SWS decreases on the later night

Y =no. of subjects whose REM or SWS increases on the later night
P =probability of number involved in change (Siegel 1956, p. 250)

majority of subjects; sws showed an increase from the first to second nights.
Only one subject (subj. 6) showed a high percentage of REM (50 per cent, 46
per cent) and this was on the first two nights. Stage 4 sleep was identified
only in one subject (subj. 7) and accounted for 6 per cent of total sleep on
night 2, and 0.5 per cent on night 3.
These studies have focused only on the terminal period of prolonged
drinking bouts; experimental studies of the initial effects of ethanol on sleep
(Gresham, Webb, and Williams 1963; Yules, Freedman, and Chandler 1966;
Yules, Ogden, Gault, and Freedman 1966; Knowles, Laverty, and Kuechler
1968) have demonstrated that REM suppression and rebound occur following
discrete quantities of ethanol. To examine the effect of more prolonged
drinking a further experimental study was made, in which both sleep and the
development of tolerance to ethanol were investigated . Only those relevant to
sleep will be discussed here .

Study 3

Four subjects took part: two alcoholics (one a repetitive bout drinker, the
other a steady drinker), and two non-alcoholic subjects (one of whom had a
high N score on the Eysenck Personality Inventory [Eysenck and Eysenck
1963]). All subjects had been abstinent for one month. The subjects were
exposed to the following schedule: (I) a period of control observations; (2) a
period of morning drinking ( consuming a quantity of ethanol equal to 0.4
gm/k,g/b.w.); (3) a period of morning drinking (at the same level) plus
 136 A Triune Concept of the Brain and Behaviour

40 CONTROL
NIGHTS
I! I MORNING
DRINKING ONLY I AM+ PM
DRINKING
I
WITHDRAWAL
Cl.
w ~
w "'
V)
...J
en
~

oo
:,,
w
> ,.z
,, \
<{
~ 30 ,, 0
f.
00

,,,,
0

~ 0
3: 0
•',,I I
I

I :\I/
0 Or
,,
0
...J Il II \/~
en .I l oo
'0

\t
0 0 0 I I
,.
I-
z II I
w 20
u . II 0f . .·
':
I 10
I I
O \ oJ
ct:
w I 10 I 1

f ,?.
ri I I
,,,,
Cl.
11

~
0 #0 ' i
z () l i
,:
jj 0
<{
0 f l t
10 I # I # t 0
~
t I •
b
0 t
w ,, II
ct:
REM
••
0 ti '
I- I o SLEEP
z H
w
u
u 0 SWS
II
ct: · tl
w 0 0 p
Cl.
5 10 15 20 25 30 35 40
SUCCESSIVE NIGHTS OF SLEEP

Figure 1 Effect of daily ethanol drinking on REM and slow-wave sleep (SWS).
Subject A: 31 years

drinking in the evening (intoxication to 0.08 mgrn per cent blood ethanol or
more); (4) a final period of withdrawal. Sleep records for these periods (REM
and sws) are shown in diagram form in Figures I, 2, and 3, for three of the
subjects (subjs. A, B, c). The fourth subject (alcoholic bout drinker) was
unable to complete the second period; he showed the familiar syndrome of
loss of control over drinking two days after exposure to the morning
drinking schedule. He procured and stored quantities of wine and drank
whenever possible; his behaviour became disturbed, he was unable to sleep
and could not be continued in the experiment without disrupting the situa-
tion, and was therefore withdrawn for hospital care and treatment. This
Sleep Disorders and Delirium 137

40
CONTROL!
NIGHTS !I MORNING
DRINKING ONLY
AM + PM
DRINKING
lw1THDRAWAL

a..
w
,,.,.... □

al)\
w ....
(/)

_J
Cf) 0
::0
w □

L~"\
~ 30
~

tv,~,
~
0
_J
Cf)

1-
z 20 □
~ \ o aD D
□ II
1 I
O REM
SLEEP
a:: □ I I o SWS
w \
a..
,,
I
0
□
· Uc I

" '\'
0
z , I I o
<(
::!: 10 JI o o
, .\
Jlof
I t fl 1
I
I
I
I
II tal It
I .l!l ®,•
w o ,0 I ;,
a:: I I I I t t t· l II o I'
t
·o ~ \
I
1-
z 1 ~o I : I I I: , I I 1' I
I I I o
w \ f I I t f fl t I
II I I
<.) ' \ I 11 II lt I I I o
a:: I\
o
1:till•
0000 b o o
I
"
w 0 □ 0 0
a.. .f:. ~
5 10 15 20 25 30 35 40

SUCCESSIVE NIGHTS OF SLEEP

Figure 2 Effect of daily ethanol drinking on REM and slow-wave sleep (SWS).
Subject B: 33 years

subject had been previously exposed to sleep studies during a drinking bout
of seven days in which consumption was uncontrolled. Marked reduction in
percentage of REM sleep (12, 18, 0, 1, 9, 0, 0, on successive nights)
persisted during the bout with a return to normal levels within four days of
withdrawal, with no rebound effect. sws was not scored. Total sleep time
was greatly reduced during the last three days of drinking.
During the morning period of drinking, assessment was made of
each subject's threshold for recognition of a low amperage electric
shock (Figure 4). Subjects were wakened at 7 am so that the morning
schedule could proceed. A routine bed-time of 11 pm was maintained but
138 A Triune Concept of the Brain and Behaviour

40 CONTROL !
NIGHTS !I MORNING I
DRINKING
AM+ PM
DRINKING
IWITHDRAWA L
~ - -~
(l. □ sr~~p
w ....
w "'....
V>
0 s ws
J
en
,.,
0

z MORN ING DR INKING TO

" RELIEVE WITHDRAWAL

!!

□
□

0
II 0
t-- /I II I
z II I I I
w II
'i
I I I
u I I I I
a:: I I 11
w O oo b
(l.

0 5 10 15 20 25 30 35 40 45 50
SUCCES SIV E NIG HTS OF SLEE P

Figure 3 Effect of daily ethanol drinking on REM and slow-wave sleep (SWS).
Subject C: 54 years

broken occasionally when the intoxicated subjects wished to stay up

later.
After the control period a test-drink was given in the evening (ap-
proximately 3.5 oz ethanol in orange juice) to determine individual reactions
to the proposed drinking schedule. Each subject showed a definite sup-
pression of REM sleep on that night. Only subject A showed a definite
rebound on the following night.
The period of morning drinking was curtailed to five days in the case of
the alcoholic subject (subj. c) because at the end of this period he said that he
could not continue without drinking or obtaining medication outside the
schedule. On a morning and evening schedule he continued to drink for a
further I 8 days. During withdrawal he twice drank covertly to relieve
withdrawal effects.
0 CONTROL I MORNING ALCOHOL I AM+ PM
ALCOHOL
I WITHDRAWAL
5!l
~
~
'O

S?.
...J "'
..,0
~ . 25
(f) ..,~
w "'
~
::,
Cl'.: C.
I
..,~
I- . 20
z 2·
3

6/.\t
0

-
1-
z . 15
0

./:v·
0 '6
(.')
0
~~.4
u '' I •
w 0

Cl'.: . 10 6
~
~l~6 I 'o
u 6"/,o, I
0 , ' I
I o' o
(f).05

~JNORMAL
o ALCOHOLIC
0
TESTING EVERY THIRD DAY

Figure 4 Effect of ethanol on shock recognition threshold

-'°
w
140 A Triune Concept of the Brain and Behaviour

TABLE4
Sleep variables in relation to ethanol

Subject A Subject B Subject C

conditions conditions conditions •

2 3 4 2 3 4 2 3 4

REM%
N 13 10 7 16 10 8 5 18 7
X 6 4 0 6 2 0 1 0 2
p 0.500 0.377 0.008 0.227 0.055 0.044 0.188 0.001 0.227
Direction + + +
REM latency
N 12 11 7 15 10 8 5 17 7
X 3 3 1 6 2 0 0 4 2
p 0.073 0.113 0.062 0.304 0.055 0.004 0.031 0.025 0.227
Direction + +
SWS%
N 13 10 7 16 10 8 5 18 7
X 6 4 2 5 0 3 1 2 0
p 0.500 0.377 0.277 0.105 0.001 0.363 0. 188 0.001 0.008
Direction

SWS latency
N 12 9 7 7 3 6 4 9 7
X 5 1 1 2 0 1 I 2 3
p 0.337 0.020 0.062 0.227 - 0.109 0.020 0.500
Direction +

Total sleep
N 13 10 7 16 10 8 5 18 7
X 1 1 1 3 3 0 1 8 1
p 0.003 0.011 0.062 0. 046 0.1 72 0. 004 0.188 0.593 0.062
Direction

N = nights of observation.
X = smaller of observed frequencies.
P = probabilities given by the binomial test (small samples). (Siegel, 1956, p. 250)
Conditions: 2, morning drinking; 3, morning and evening drinking; 4, withdrawal period

Results (Study 3)

The main findings are summarized in Table 4 and discussed individually for
each subject. Changes from control night observation are shown for REM and
sws under conditions 2, 3, and 4. x represents the number of observations
either above or below the median value obtained on the control nights; the
smaller of these two frequencies is shown as x and the sign below (entered
Sleep Disorders and Delirium 141

wherever p < 0.10) gives the direction of the change, e.g. 3, -, means that 3
observations are less than the median obtained for control nights.
Subject A . In spite of an initial suppression of REM this subject showed
very little continuing disturbance of the pattern of REM sleep. In the heavy
drinking period (3) values for REM tended to crowd around 15 per cent for
several days. Alternation between high values of REM and sws characterize
this very stable sleep record despite an over-all shortening of total sleep time.
The alternation of these variables has the balance of a passage of Mozartian
counterpoint, and in the withdrawal period (4) REM rebound alternates with
the redistribution of sws. This subject withstood the effects of drinking well
except for a brief depressive reaction following the first exposure to alcohol.
Withdrawal symptoms were minimal.
Subject B. REM increase with reduced latency was present in the later
stages of drinking and during withdrawal. More pronounced was the absence
of slow-wave sleep during the heavy drinking period (3). Reduction of total
sleep time occurred during withdrawal. This subject was more disturbed
during the period of prolonged drinking; irritability, depressive symptoms,
awakening with dreaming, and severe morning hangover symptoms were
noted, including tremulousness and anorexia. No perceptual symptoms were
described.
Subject C. The subject with an alcoholic history was older than the other
subjects, 54 years. Pronounced suppression of both REM and slow-wave sleep
developed during the drinking periods (2 and 3). REM recovery did not in-
clude obvious 'rebound' scores. Slow-wave sleep following a brief increase
tended to remain low during withdrawal. Two episodes of drinking after a full
week of withdrawal demonstrated that suppression effects could still be
produced.

Effects on Variables

REM state. In the non-alcoholic subjects REM suppression was only re-
markable at the onset of drinking, and REM 'rebound' was associated with the
withdrawal period. The alcoholic subject, despite REM suppression, showed
no rebound on withdrawal.
REM latency. Decreases in REM latency (during withdrawal) were pre-
dictably associated with REM increase but occurred only in the non-alcoholic
subjects. REM suppression in the alcoholic was associated with postponement
of REM.
Slow-wave sleep. In one non-alcoholic (the more stable), no effects were
noted; the other two subjects showed suppression during heavy drinking; this
continued in the alcoholic subject during the withdrawal period.
142 A Triune Concept of the Brain and Behaviour

sws latency. Decreased latencies for sws in subject A (phases 2 and 3)

were associated with a strong need for sleep which subject described as
'needing to flake out.' Longer latencies occurred for subject c when sws
suppression was observed (2).
Total sleep. Total sleeping time was reduced in all periods of drinking and
withdrawal, except that the alcoholic subject slept slightly longer during the
period of heavy drinking. Most of this sleep was stage I (NREM) and stage 2
sleep, and the subject was characteristically easily woken during this period.
Shortening of sleep time in the non-alcoholic subjects was contributed to by
their delaying the time of sleep onset.
Shock threshold. Figure 4 shows the shock-recognition threshold (derived
from small amperage shock given to the hand) for the morning drinking
periods throughout study 3. Decreased thresholds develop during heavy
drinking, continue in withdrawal, and gradually return to normal values.

Discussion

From the small samples represented by these sleep studies only tentative
conclusions can be drawn. An attempt will be made to piece together the
evidence concerning the effects of ethanol on sleep at the beginning, middle,
and end of a period of heavy drinking.
The initial effect of ethanol in suppressing REM described by Gresham,
Webb , and Williams (1963), Yules, Freedman, and Chandler (1966), Knowles,
Laverty, and Kuechler (I 968), is confirmed in study 3.
Continued drinking in the evening sufficient to produce definite intoxica-
tion (0.08 mgm per cent or more) was followed by continued REM suppres-
sion only in the alcoholic subject c. A similar suppression occurred with
subject o in a separate study. Both these subjects drank more heavily than the
non-alcoholic subjects. Knowles, Laverty, and Keuchler (I 968) have shown
that REM suppression and rebound are dependent on dosage. Continued REM
suppression during steady drinking might be expected to follow the large
quantities and has been described (Greenberg and Pearlman 1967) with occa-
sional nights of high percentage REM consumed by addicted drinkers. That
REM suppression is in evidence at the end of the drinking bout is supported
by the low percentages of REM in the first night of study 2 on the last day of
drinking. However this was also the first night in the sleep laboratory when
low values may be expected (Dement 1965). Values such as 0, 0, 7, 9,
however, are unusually low. Evidence concerning continued REM suppression
from ethanol is therefore conflicting; wide individual differences appear in
these samples; definite continued suppression does occur, but so does
adaptation.
Sleep Disorders and Delirium 143

Increase in REM on withdrawal after ethanol administration is reported

(Knowles, Laverty, and Kuechler 1968; Gross et al. 1966; Greenberg and
Pearlman 1967) and confirmed in the non-alcoholic subjects (subjs. A and B)
in study 3. The subjects in withdrawal (study 2) also showed an increase in
REM during the first four nights, and although this period might be explained
as a result of adaptation to the first nights in the sleep laboratory, the pattern
follows that seen in REM deprivation studies (Dement 1965) where REM is
made up over a period of several nights. Only two subjects in this study
(subjs. 6 and 8) showed levels (50 per cent, 46 per cent, 48 per cent) high
enough to be described as 'rebound' effects. REM rebound is therefore not a
necessary or conspicuous result of the alcohol withdrawal state, even after
days of heavy drinking, and its relative absence in alcoholics may be an
adaptation to repeated bouts of drinking.
Subjects in withdrawal delirium , on the other hand, do show, at least for
short periods, very high proportions of REM sleep (Gross et al. 1966; Green-
berg and Pearlman 1967). That hallucinations in delirium are a direct result
of, or a continuation of, REM-sleep increase appears open to doubt, for the
following reasons. (1) Hallucinations may continue during periods when the
REM component of sleep is in the normal range, as was observed in subject 3
(study 1). (2) REM sleep in delirious states may be misidentified. In parti-
cular, the persistence of muscular activity, generalized restlessness in sleep,
and the presence of a variety of eye-movements complicates such records.
Vogel (1968) raised this objection in relation to other studies. However, in
the current study two good examples of continuous REM sleep were seen in
association with waking into a hallucinatory state. (3) REM deprivation as a
cause of hallucinatory psychosis has not been confirmed. An exhaustive dis-
cussion of REM deprivation studies by Vogel (1968) suggests that total sleep
deprivation might be a more important contributing factor in cases cited.
REM deprivation is conceded to increase general drive and irritability (Vogel
1968; Agnew, Webb, and Williams 1964), changes commonly seen in al-
coholic subjects during drinking and in withdrawal. (4) The disturbances seen
in delirious alcoholics, and in the severe withdrawal syndrome , are much
more general than those attributable to REM deprivation and rebound, and
include the general reduction to sensory thresholds illustrated in Figure 4,
increased motor activity, and even major convulsive seizures. It would seem
appropriate at this point to include the sleep disorders and perceptual distur-
bances as symptoms of the unstable state of the central nervous system rather
than to explain one symptom (hallucinations) in terms of the other (REM
rebound) .
The most striking observation common to the sleep records in the al-
coholic subjects studied here, has been the low proportion of slow-wave sleep,
144 A Triune Concept of the Brain and Behaviour

and particularly the absence of stage 4. Slow-wave sleep was significantly

reduced in experimental subjects B and c (study 3) during the period of
evening drinking, while subject A showed a persistent periodic rebound.
Records from alcoholics with persisting insomnia have also shown low levels
of slow-wave sleep, after abstinence for as long as three months. Smith,
Johnson, and Burdick (in press) report a near-absence of slow-wave sleep in
14 alcoholics studied for 12 days of withdrawal, and low values for stage 3.
Deprivation of stage 4 sleep (Agnew, Webb, and Williams 1964) has been
related to depressive symptoms, and these are common in alcoholics after
periods of drinking.
A possible synergism appears to link REM and stage 4 sleep. Agnew, Webb ,
and Williams (1964), for example, report that stage 4 sleep rebound (follow-
ing stage 4 deprivation) occurs in the succeeding night and may be followed
by increases in REM over the next two nights, although no REM deprivation
has occurred. Deprivation of stage 4 during drinking may precede (or pro-
voke) additional REM rebound, such as may occur in delirium tremens. Cer-
tainly, the finding of changes in both these components of sleep related to
ethanol use deserve further investigation.
Many other drugs produce REM suppression but few of these are associated
with an acute withdrawal psychosis. An exception are the short and medium
acting barbiturates (Kales 1969; Oswald, Berger, Jaramillo, Olley, and
Plunkett 1963; Oswald and Priest 1965; Kales and Jacobson 1967). Like
alcohol, their action and elimination is relatively rapid, and the action of both
is associated with the development of tolerance and changes in sleep pattern.
Drugs with a slower withdrawal effect appear to be less potent and return to a
normal sleep pattern less disruptive.

REFERENCES

Agnew, H.W., W.B. Webb, and R.L. Williams, The effects of stage four sleep deprivation.
Electroencephalog. Clin. Neurophysiol., 1964, 17 :68-70
Aschan, G., M. Bergstedt, L. Goldberg, and L. Laureti, Positional nystagmus in man
during and after alcohol intoxication. Quart. J. Studies Ale., 1956, 17:381-405
Dement, W.C., An essay on dreams: the role of physiology in understanding their nature.
In New Directions in Psychology, 2. New York: Holt, Rinehart & Winston, 1965
Dement, W.C. and N. Kleitman, The relation of eye-movements during sleep to dream
activity : an objective method for study of dreaming. J. Exp. Psycho/., 1957,
53 :339-46
Dynes, J.B., Survey of alcoholic patients admitted to Boston Psychopathic Hospital in
193 7. New Engl. J. Med., 1939, 220: 195-8
Engel, G.L., Delirium . In A.M. Freedman and HJ. Kaplan (eds.), Comprehensive Text-
book of Psychiatry. Baltimore: Williams and Wilkins, 1967
Sleep Disorders and Delirium 145

Eysenck, H.J. and S.B.G. Eysenck Manual of the Eysenck Personality Inventory. San
Diego, Calif.: Educational and Industrial Testing Service, 1963
Goldberg. L., Behavioral and physiological effects of alcohol on man. Psychosomat.
Med., 1966, 28 :570-95
Greenberg, R. and C. Pearlman, Delirium tremens and dreaming. Amer. J. Psychiat.,
196 7, 124 : 133-42
Gresham, S.C., W.B. Webb, and R.L. Williams, Alcohol and caffeine: effect on inferred
visual dreaming.Science, 1963, 140: 1226-7
Gross, M.M., K. Goodenough, M. Tobin, E. Halpert, D. Leport, A. Pearlstein, M. Sirota,
J. Dibianco, R. Fuller, and I. Kishner, Sleep disturbances and hallucinations in the
!'cute alcoholic psychoses. J. Nervous Mental Disease, 1966, 142:493-514
Isbell, H., H.F. Fraser, A. Wikler, R.E. Belleville, and A.J. Eisenman, An experimental
study of the etiology of 'rum fits' and delirium tremens. Quart, J. Studies Ale., 1955,
16: 1-33
Kales, A. Sedatives and sleep-dream alterations. In A. Kales (ed.), Psychology and Path-
ology of Sleep. Philadelphia: J .B. Lippincott, 1969
Kales, A. and A. Jacobson, Mental activity during sleep: recall studies, somnambulism,
and effects of rapid eye movement deprivation and drugs. Exp. Neurol., 1967, Suppl.
4 :81-91
Kat, W. and J.G. Prick, Uber Pathogenese und Klinik des Delirium tremens. Schweiz.
Arch. Neurol. Psychiat., 1940, 45: 303-40
Knowles, J .B., S.G. Laverty, and H.A. Kuechler, Effects of alcohol on REM sleep. Quart.
J. Studies Ale., 1968, 29 : 34 2-9
Kraepelin, E., CTinical Psychiatry. Adapted from the 7th Gennan ed. and translated by
A.R. Defendorf. New York: Macmillan, 1907
Kraines, S.H., Mental Depressions and their Treatment. New York: Macmillan, 195 7
Mendelson, J.H. and J. LaDou, Experimentally induced chronic intoxication and with-
drawal in alcoholics. Quart. J. Studies Ale., 1964, Suppl. no. 2: 1-126
Moore, M. and M.G. Gray, Delirium tremens: study of cases at Boston City Hospital,
1915-1936. New Engl. J. Med., 1939, 220:953-6
Oswald, I., R.J . Berger, R.A . Jaramillo, K.M .G. Keddie, P.C. Olley, and G.B. Plunkett,
Melancholia and barbiturates: a controlled EEG, body and eye movement study of
sleep. Brit. J. Psychiat. , 1963, 109:66-78
Oswald, I. and R.G. Priest, Five weeks to escape the sleeping-pill habit. Brit. Med. J.,
1965, v.2 (5470) : 1093-5
Rosenbaum, M., M. Lewis, P. Piker, and D. Goldman, Convulsive seizures in delirium
tremens. Arch. Neurol. Psychiat., l 93 8, 40: 922-7
Siegel, S., Nonparametric statistics. New York: McGraw-Hill, 1956
Smith, J.W., L.C. Johnson, and J .A. Burdick, Sleep ; psychological and clinical changes
during alcohol withdrawal in NAO-treated chronic alcoholics. (In press)
Thompson, G.N., Acute and chronic alcoholic conditions. In S. Arieti (ed.), American
Handbook of Psychiatry. New York: Basic Books, 1959
Vogel, G., REM deprivation : Ill. Dreaming and psychosis. Arch. Gen. Psychiat., 1968,
18: 320-8
Yules, R.B., D.X. Freedman, and K .A . Chandler, The effect of ethyl alcohol on man's
electroencephalographic sleep cycle. Electroencephalog. Clin. Neurophysiol., 1966,
20: 109-11
Yules, R.B., J.A. Ogden, F.P. Gault, and D.X. Freedman, The effect of ethyl alcohol on
electroencephalographic sleep cycles in cats. Psychonom. Sci., 1966, 5:97-8
 JOHN B. KNOWLES, EUGENE J. BEAUMASTER,
ALISTAIR W. MACLEAN

10 The Function of Rapid Eye Movement Sleep

and of Dreaming in the Adult'

Although REM sleep never regains the prominence it enjoys during neonatal
life, changes in sleep that occur with age suggest that REM sleep in the adult
may be more than a vestige of its neonatal self. Studies conducted in a
number of laboratories agree in showing that with increasing age the total
period of sleep declines, but that within this span of time the proportion
spent in REM sleep remains remarkably constant - approximately 20 to 25
per cent. In contrast, as Table I shows, with increasing years the proportion
of stage 4 sleep declines precipitously from approximately 15 per cent in the
twenties and thirties to about I to 5 per cent in subjects sixty and over. In
other words, while some sleep characteristics change quite dramatically with
increasing age, this is not true of the REM state; it persists, and its persistence
demands some explanation.
What function then does REM sleep perform in the adult? Presumably its
function can derive either from the physiological changes that serve to define
REM sleep or from the psychological events - namely dreaming - that
typically occur during REM sleep. In other words we can distinguish hypo-
theses about REM sleep from those that refer more specifically to dreaming.
One view that clearly falls within the first category is that the cerebral
activation characteristically associated with REM sleep serves to maintain
cortical tonus or arousal (Ephron and Carrington 1966). From this point of
view periods of REM sleep act homeostatically to offset the decline in arousal
which occurs during non-REM sleep . We examined this hypothesis by arguing
that if it were true then we would expect that the prolongation of non-REM
sleep prior to the first REM period would lead to an augmentation of REM
activity in the sleep period remaining to the individual; in a sense the late
starter would have to make up for lost time. Accordingly, we calculated the
regression of REM time on latency to the first REM period in two sets of data

Some of the data cited are derived from studies supported by grant # 14 7 from the
Ontario Mental Health Foundation and by a grant from Merck, Sharpe, and Dohm
 Rapid Eye Movement Sleep 147

TABLE 1
Duration of sleep stages {as percentage of total sleep) in persons of differing age

Author Subjects % stages after sleep onset

N Sex Age Awake 2 1 2 3 4 1 REM

Williams, Agnew, and 16 M 21-31 0.9 5.4 48.7 7.7 13.2 24.l
Webb (1964)
Williams, Agnew, and 16 F 19-28 1.1 5.9 48.0 6.9 16.2 21.9
Webb (1966)
Kales et al. {1967a) 1 10 SM 20-29 - 4.9 49.6 10.3 11.2 24.0
SF
Feinberg, Koresko, and 15 9M 19-36 - 9.6 13.0 22.7
Heller (1967) 6F
Agnew and Webb {1968a) 12 M 30-39 2.4 7.5 53.0 5.5 9.6 21.9
Agnew, Webb, and 16 M 50-60 4.1 10.9 51.1 8.4 2.7 22.8
Williams (196 7a)
Agnew and Webb {1968b) 16 llM 60-69 9.9 11.9 50.6 4.5 2.7 20.4
SF
Kales et al. (196 7b ) 3 10 SM 64-87 - 6.5 56 .9 14.0 1.3 22.3
SF
Kahn and Fisher (1969) 16 16M 71-95 - 53.9 17.2 4.5 20.1
Feinberg, Koresko and 15 9M 65-96 - 16.5 7.3 21.1
Heller (1967) 6F

1 Data for third of three nights of recording

2 In five studies periods ~f wakefulness subtracted from total sleep time
3 Data for fourth of four nights of recording

collected in our laboratory. One refers to a group of 40 subjects for whom

REM sleep was calculated from records obtained on the second of two nights.
The second set of data refers to a single subject seen for 31 virtually con-
secutive nights. Within the sample of 40 subjects we found that though the
slope of the regression line did differ significantly from zero (see Figure 1),
only 17 per cent of the variance in REM time was accounted for by variation
in REM latency. The findings on the single subject are even more striking. As
Figure 2 shows, the regression line does not differ significantly from zero; if
anything, the correlation is negative. Essentially similar results were obtained
when we calculated the regression of REM time on latency to the second REM
period. Therefore we conclude that our data do not support the homeostatic
hypothesis.
The behavioural effects of experimentally induced REM deprivation are
also relevant to a discussion of REM sleep and the function it might perform.
As is well known, Dement (I 960) selectively deprived subjects of REM sleep
for four or more consecutive nights by waking them at the onset of each REM
period. He reported that 'psychological disturbances such as anxiety,
irritability, and difficulty in concentrating developed during the period of
148 A Triune Concept of the Brain and Behaviour

c:,
z
i
0
..J
..J
~ ...
~t
50

40
0

0
O
0
0
0

l&I ·0 .
2 Ill 30 0
l&I z
a: 0 0
20
1&12
c:, l&I
~a: 10
z
l&I
(,)
a: 30 60 90 120 150 180 210 240 270 300 330
l&I
G.
LATENCY OF REM I ( Mina.)

Figure 1 Percentage of REM during sleep following REM 1 onset as a function of the
latency of REM 1 onset (N=40)

- ~-
40
0
30 -0 '11 ;-"' 0 -'"""'-"--70;--------------~-
0 0
20

10

o-~----~-.. . . . --~-.. . . . --~-.. . . . --~-.. . . . --.. . .-

40 60 80 100 120 140 160 180 200

LATENCY OF REM I ( Mina.I

Figure 2 Percentage REM during sleep following REM 1 onset in single subject seen
for 31 nights

dream deprivation ...' (p. 1707) though these disturbances were not 'catastro-
phic.' However, of the seven similar investigations published since that time
(Snyder 1963; Kales, Hoedemaker, Jacobson, and Lichtenstein 1964;
Sampson 1966; Clemes and Dement 1967; Agnew, Webb, and Williams
1967b; Vogel and Traub 1968; Vogel, Traub, Ben-Horin, and Meyers 1968),
only two (Sampson 1966; Agnew, Webb, and Williams 1967) have found
adverse mood or personality changes in the REM deprived subjects, and in one
of these (Sampson 1966) the clinical observations were not substantiated by
psychological test data . In a particularly striking study, Vogel and others
(1968) were even able to demonstrate that in a small series of depressed
 Rapid Eye Movement Sleep 149

patients, deprived of REM sleep for between 7 and 14 nights, those who
reacted by showing progressively diminishing inter-REM latencies and by
showing the characteristic 'rebound' of REM sleep during the subsequent
recovery period were sufficiently improved clinically to be discharged from
hospital without further treatment. On the other hand, patients who did not
respond to the REM deprivation in these two ways failed to improve clinical-
ly. With respect to mood and personality changes it seems reasonable to
conclude, as does Vogel (1968), that the early findings 'were probably a
result of uncontrolled factors such as the expectations of the experimenters
and the subjects' (p. 319).
However, it remains possible that REM sleep plays a more critical role in
maintaining what we may broadly refer to as cognitive functions. Dement and
Fisher (I 963) noted impairment of memory and concentration in their sub-
jects, as did Sampson (I 966), but, in common with most other investigators,
Sampson was unable to demonstrate any significant change in scores on ob-
jective psychological tests. The fact that test performance has not been ad-
versely influenced by REM deprivation procedures might be taken as lending
additional support to the contention that experimenter and subject expecta-
tions contributed to the earlier clinical findings, but we have also to consider
the possibility that the negative findings are attributable to limitations in the
psychological tests employed. Studies of the psychological effects of total
sleep loss are highly informative in this connection. From the results of these
studies it is apparent that to be sensitive to total sleep loss - and, thus,
presumably to loss of REM sleep - performance tests should be long, at least
thirty minutes in duration, lack incentive value, and be of moderate com-
plexity. The duration of the test is particularly important (Wilkinson 1968).
However, performance tests used in studies of REM deprivation have rarely, if
ever, met these criteria. Particularly striking is the fact that for the most part
the tests have been relatively brief in duration.
Thus, it may be premature to conclude that REM deprivation has no detri-
mental effects on behaviour. Even so, cognitive and memory impairment is
more likely to result from stage 4 deprivation than from REM deprivation, if
only because such impairment is most obvious in the elderly in whom, as we
have seen, there is a relative absence of stage 4, but not of REM sleep. 2
The existence of reliable individual differences in REM sleep has permitted
an alternative approach to the problem of REM function. In a relatively early

2 Subsequent to the presentation of this paper, we were able to analyse the results of a
study (Maclean 1969) in which reaction time, dichotic listening, and signal detection
tasks were given to subjects during the day following a night in which REM sleep was
significantly reduced by the administration of Amitriptyline (Elavil) prior to sleep. None
of the tasks were adversely influenced by this experience
 150 A Triune Concept of the Brain and Behaviour

study Rechtschaffen and Verdone (1964) reported that within a sample of

individuals, distinguished only by their willingness to participate and by
stable sleep habits, the proportion of REM sleep was associated with MMPI
indices of introversion and anxiety. Working with a similar group of 40 sub-
jects in our own laboratory, Beaumaster (I 968) found little or no difference
between introverts and extraverts in this respect, but he was able to confirm
that REM sleep was significantly associated with neuroticism as measured by
the Eysenck Personality Inventory (Eysenck and Eysenck 1963). These data ,
which refer to the amount of REM sleep recorded during the first six hours of
sleep on the second of two nights of recording, are shown in Figure 3. During
this period the 'neurotic' and 'stable' 3 subjects spent respectively 23 per cent
and 18 per cent of their time in REM sleep (F = 6.67, p < 0.05). Further
analysis of these data showed that this difference was associated with signi-
ficant differences between the mean latencies of the first REM period (I 36
mins and 89 mins for the stable and neurotic subjects respectively, F = 4.38,
p < 0.05). The differences in REM latency, shown in Figure 4, probably ac-
count for the differing amounts of total REM sleep in that Beaumaster also
found that the neurotic subjects experienced more, rather than longer, REM
periods during the remainder of the six-hour sleep period, and that latencies
between the remaining REM periods did not distinguish the neurotic from the
stable subjects.
One other difference between the two groups is noteworthy. The neurotic
subjects fell asleep significantly later than did the stable. This was due to the
fact that the neurotic subjects went to bed later than did the stable ; both
groups took similar times to fall asleep once in bed. Since subjects were asked
to come to the laboratory one hour before their normal bed time, it is
reasonable to infer that these subjects characteristically retire at a relatively
late hour. Though it is probable that personality types differ in their pre-
ferred scheduling of diurnal activity, occupational demands are likely to place
constraints upon their freedom of choice. Consequently, it is possible that
those who retire late suffer some curtailment of their sleep, particularly of
the latter part of their sleep period, in which REM sleep predominates. This
partial sleep deprivation, if continued for some time, could have two effects.
First there may be an adjustment within the sleep cycle such that REM sleep
occurs earlier (Webb 1969), secondly there could be periodic REM ' rebounds'
occasioning a transitory elevation in REM time.

3 'Neurotic' and 'stable' refer to subjects selected because their N scores were respectively
above and below the mean score of 10.08 for a sample of 233 male undergraduates at
Queen's University. Stable subjects had scores ranging from I through 8 (mean= 4.90);
the neurotic had scores ranging from 11 through 20 (mean= 14.35). There were twenty
subjects in each group
Rapid Eye Movement Sleep 151

11

10

6
f 'STABLE' (N=20)
5

3
'NEUROTIC' (N=20)
f
2

-10 -20-30 -40 -50 -60 -70 -80 -90-100 101+

Minutes

Figure 3 Total REM time (mins) in first six hours of sleep for 'neurotic' and 'stable '
subjects
152 A Triune Concept of the Brain and Behaviour

Either or both processes described above could account for the diminished
REM latency and increased REM times in the neurotic subjects, but the shorter
REM latencies in the neurotic subjects appear to parallel the clinical ob-
servation of Hartmann, Verdone, and Snyder (1966) that in one of their
cases, a male depressive continuously observed for three months, there was a
significant inverse relationship between severity of depression and REM
latency. Moreover, in a study of individual differences in response to REM
deprivation, Cartwright, Monroe, and Palmer (1967) found that when de-
prived of REM sleep by being awakened at the onset of each REM period,
subjects with high levels of anxiety (as measured by the Taylor Manifest
Anxiety Scale) showed the shortest latencies between successive REM periods.
In other words, it was the more anxious subjects who had to be awakened the
most frequently to prevent the onset of REM sleep.
Taken together, these data on personality differences suggest an alternative
view that REM sleep derives significance from its psychological con-
comitants - namely dreaming. In particular, on the basis of their findings,
Hartmann, Verdone, and Snyder (1966) speculated that REM sleep may not
be related to psychopathology in a general sense, but to be sensitive to
periods of what they term psychic disequilibrium or psychic pain. Stated
alternatively, it may be that dreaming permits the expression of material that
is avoided during the waking state, because this avoidance facilitates coping
with stress. If so, we would expect to find increased REM time during periods
of stress and, in addition, to find that REM sleep is particularly evident in
individuals with relatively high levels of neuroticism. Moreover, if this were
true, then one might also expect that dreaming would be particularly
prominent in individuals who characteristically cope with stress by repression
or denial. Pivik and Foulkes' (1966) findings, that 'repressors' as judged by
Byrne's Repression-Sensitization Scale have significantly longer REM periods
than do 'sensitizers,' and that when deprived of REM early in the night they
react by intensifying the content of subsequent dreams to a greater extent
than do sensitizers, are, we believe, consistent with this idea.
However, whether this is a defensible proposition rests not only upon the
validity of this particular hypothesis about dream function, but also upon the
association between REM sleep and dreaming. This remains a contentious
issue. Foulkes (1966), in particular, argues that dream-like mentation occurs
in non-REM sleep, and we have also to consider the possibility that under
unusual conditions, such as repeated awakenings required to deprive someone
of REM sleep, or in some individuals, dreaming may be displaced into or occur
during sleep stages other than stage 1 REM. The results of a recent experiment
by Cartwright and Monroe (1968) suggest that this might be so. In their study
subjects were awakened at the onset of each REM period that occurred during
Rapid Eye Movement Sleep 153

7
6

5 'STABLE' (N=20)

4 - -
3

I I

f 5 'NEUROTIC' (N=20)

0
-20 -40 -60 -80 -100-120 -140-160-180-200 201+
Minutes

Figure 4 Latency of first REM period (mins) for 'neurotic' and 'stable' subjects
154 A Triune Concept of the Brain and Behaviour

the first 200 minutes of sleep. On awakening, subjects were asked either to
recall and elaborate upon the dream material briefly experienced before
waking or they were asked to carry out a digit span task which occupied the
same length of time and required a similar degree of arousal from sleep. They
found that if the subjects were permitted to recount their dream material, to
have, as it were, a waking dream, there was no elevation of REM time in the
remaining 200 minutes of sleep during which they were allowed to sleep
undisturbed, but if they had previously carried out the digit span task, a task
designed to block the expression of dream content, there was a significant
increase in REM time during the following period of uninterrupted sleep.
Also there are a number of reports that dreaming may take place at sleep
onset even though laboratory recording of sleep indicates that in normal
subjects, at least, REM sleep does not occur until the subject has been asleep
some sixty to ninety minutes. For the most part these observations are anec-
dotal, but Singer (I 966) has reported that he himself experienced a vivid
dream shortly after sleep onset, but that rapid eye movements were not
evident in the recording made concurrently.
Thus, although dreaming usually occurs during REM sleep, there is reason
to believe that dreaming is not inevitably linked with the occurrence of REM
sleep. This would seem to be a crucial point when evaluating the results of the
REM deprivation studies - for if dreaming is, as it were, a moveable feast, the
fact that the majority of REM deprivation studies have failed to demonstrate
that REM deprivation eventuates in psychological disturbance need not
denude dreaming of psychological significance.

REFERENCES

Agnew, H.W. and W.B. Webb, Sleep patterns of 30-39 year-old male subjects. Psycho-
physiology, 1968a, 5:228
Agnew, H.W. and W.B. Webb, Sleep patterns of the healthy elderly. /bid., 1968b, 5:229
Agnew, H.W., W.B . Webb, and R.L. Williams, Sleep patterns in late middle age males: an
EEG study. Electroencephalog. Clin. Neurophysiol., 1967a, 23: 168-71
Agnew, H.W., W.B. Webb, and R.L. Williams, Comparison of stage four and 1-REM sleep
deprivation. Percept. Motor Skills, 1967b, 24:851-8
Beaumaster, E.J.B ., Individual differences in rapid eye movement (REM) sleep. Unpub-
lished MA thesis, Queen's University, 1968
Cartwright, R.D. and L.J . Monroe, Relation of dreaming and REM sleep : the effects of
REM deprivation under two conditions. J. Personality Soc. Psycho/., 1968, 10: 69-74
Cartwright, R.D., L.J. Monroe, and C. Palmer, Individual differences in response to REM
deprivation.Arch. Gen. Psychiat., 1967, 16: 297-303
Clemes, S.R. and W.C. Dement, Effect of REM sleep deprivation on psychological func-
tioning. J. Nervous Mental Disease, 1967, 144:485-91
Rapid Eye Movement Sleep 155

Dement, W.C. , The effect of dream deprivation. Science, 1960, 131 : 1705-7
Dement, W.C. and C. Fisher, Experimental interference with the sleep cycle. Can.
Psychiat. Assoc. J., 1963, 8:400-5
Ephron, H.S. and P. Carrington, Rapid eye movement sleep and cortical homeostasis.
Psycho[. Rev., 1966, 73 :500-26
Eysenck, H.J. and S.B.G. Eysenck, Manual for the Eysenck Personality Inventory, San
Diego, Calif: Educational and Industrial Testing Service, 1963
Feinberg, I., R.L. Koresko, and N. Heller, EEG sleep patterns as a function of normal
and pathological aging in man. J. Psychiat. Res., 1967, 5: 107-44
Foulkes, W.D., The Psychology of Sleep. New York: Scribners, 1966
Hartmann, E., P. Verdone, and F. Snyder, Longitudinal studies of sleep and dreaming
patterns in psychiatric patients. J. Nervous Mental Disease, 1966, 142: 117-26
Kahn, E. and C. Fisher, The sleep characteristics of the normal aged male. J. Nervous
Mental Disease, 1969, 148:477-94
Kales, A., F.S . Hoedemaker, A. Jacobson, and E.L. Lichtenstein, Dream deprivation: an
experimental reappraisal. Nature, 1964, 204: 1337-8
Kales, A. , A. Jacobson, J.D . Kales, T. Kun, and R. Weissbuch, All-night EEG sleep
measurements in young adults. Psychonom. Sci., 1967a, 7:67-8
Kales, A., T. Wilson, J.D. Kales, A. Jacobson, M.J . Paulson, E. Kollar, and R.D. Walters,
Measurements of all-night sleep in normal elderly persons: effects of aging. J. Am.
Geriat. Soc., 1967b, 15:405-14
MacLean, A.W., The behavioural effects of Elavil induced deprivation of the rapid eye
movement phase of sleep. Unpublished PhD thesis, Queen 's University, 1969
Pivik, T . and D. Foulkes, Dream deprivation: effects on dream content. Science, 1966 ,
153 : 1282-4
Rechtschaffen, A. and P. Verdone, Amount of dreaming: effect of incentive, adaptation
to laboratory, and individual differences, Percept. Motor Skills, 1964, 19:94 7-5 8
Sampson, H., Psychological effects of deprivation of dreaming sleep. J. Nervous Mental
Disease, 1966, 143 :305-17
Singer, J.L., Daydreaming: an Introduction to the Experimental Study of Inner Ex-
perience. New York : Random House, 1966
Snyder, F., The new biology of dreaming. Arch. Gen. Psychiat., 1963, 8:381-91
Vogel, G.W., REM deprivation : III. Dreaming and psychosis. Ibid., 1968, 18 :312-29
Vogel, G.W. and A.C. Traub, REM deprivation: I. The effect on schizophrenic patients.
Ibid., I 968, 18: 287-300
Vogel, G.W ., A.C. Traub, P. Ben-Horin, and G.M. Meyers, REM deprivation : II . The
effects on depressed patients. Ibid., 1968, 18:301-11
Webb, W.B., Partial and differential sleep deprivation. In A . Kales (ed.), Sleep, Physi-
ology and Pathology; a Symposium. Philadelphia: Lippincott, 1969
Wilkinson, R.T., Sleep deprivation : performance tests for partial and selective sleep de-
privation. In Progress in CTinical Psychology, vol. 8. New York : Grune and Stratton,
1968
Williams, R.L., H.W. Agnew, and W.B. Webb, Sleep patterns in young adults: an EEG
study. Electroencepha/og. Clin. Neurophysiol., 1964, 17:376-81
Williams, R.L., H.W. Agnew, and W.B. Webb, Sleep patterns in the young adult female:
an EEG study.Ibid., 1966, 20: 264-6
 DUGAL CAMPBELL, JOHN RAEBURN

11 Patterns of Sleep in the Newborn'

In 1955 two papers on sleep appeared side by side in the Journal of Applied
Physiology (Aserinsky and Kleitman 1955a, 1955b). Both papers described
sleep in terms of a cyclical alternation of two phases. The first of the two
papers described sleep in adults and drew attention to the fact that during the
periods in which low voltage and fast EEG waves were to be seen, there were
also rapid eye movements. The second paper described sleep in infants in
whom periods marked by body movements coincided with the rapid eye
movements. The two reports included estimates of the duration of the two
major phases of the sleep cycle in both adults and infants; in infants the rapid
eye movement period apparently lasted for about an hour and the companion
quiet phase about 23 minutes; in adults the two phases apparently lasted
about the same time : 20 minutes active sleep and 24 minutes quiet sleep.
This last result suggested that REM sleep in children differed quantitatively
from REM sleep in adults but this aspect of the subject did not at first attract
a great deal of investigation. Perhaps because the REM period was described as
'an objective method for the study of dreaming' it attracted a great deal of
work (Dement and Kleitman 1957). Dreams and the effects of REM depriva-
tion were the major topics examined in the first few years of the new rush of
sleep studies.
More recently three kinds of finding have changed the emphasis of interest
in sleep studies: studies of sleep in animals, neurophysiological studies of
sleep, and studies of sleep made at a variety of times during the human life
span. Sleep of a cyclical nature with two phases resembling those in man has
been found in all mammals so far examined including species with a long
evolutionary history such as the opossum (Snyder 1966). Neurophysiological
studies with animals have shown that control of the active phase is due to an
area in the brain stem (Jouvet 1967) which again suggests that active sleep
may have a considerable history in evolution.

This work was supported by the Ontario Mental Health Foundation (grant no. 150)
Sleep in the Newborn 157

TABLE I
Estimates of total sleep time and proportion of REM sleep
in children (Hartmann 1967)

Total sleep %REM

time (hours)

Prematures ? 50-80
Term infants 16 45-65
0-2 years 12 25-40
2-5 years 11 20-30
5-13 years 10 15-20

The studies of sleep at different ages in human subjects have shown that
the pattern of total sleep and active sleep changes during life : the total time
spent in sleep falls as individuals get older and the proportion of sleeping time
spent in REM sleep also falls (e.g., Feinberg and Carlson 1968). This last result
has led to a renewal of interest in sleep in the newborn. Further studies (e .g.,
Delange, Cadilhac, and Passouant 1962; Monod and Pajot 1965; Parmelee,
Wenner, Akiyama, Stern, and Flescher 1967) have confirmed Aserinsky's
original observation that infants have a greater proportion of REM sleep than
adults. It also appears that the rate of change in the sleep pattern is greater
during early infancy than at any other time during life. Similar results have
been found in kittens and newborn rats in both of whom quiet sleep is
seldom to be found (Jouvet 1967).
The results for children are illustrated in the first table which shows
estimates of total sleep time and estimates of the proportion of REM sleep
during the first decade of life . It can be seen that the decline in the propor-
tion of REM sleep is greater in the first months of life than it is later. More-
over, if one extends the tendency to the period before birth one must assume
that the foetus in utero spends a substantial portion of time in active sleep
and the few observations available confirm this extrapolation (e.g., Parmelee
et al. 1967).
You will notice that not only is the estimated proportion of REM sleep
greatest about term but also that the variation between the upper and lower
estimates is greatest. A starting point of the investigations to be described in
this paper was the discrepancies in the proportion of REM sleep said to be
typical of newborn infants.
An analysis of the original reports showed that a wide variety of criteria
have been used in the classification of infants' sleep: some studies have relied
upon measures which were made by an observer; others depended upon addi-
tional measures for which a permanent record and subsequent analysis were
necessary. Both of the techniques provide data which confirm the idea that
158 A Triune Concept of the Brain and Behaviour

sleep in infants can be divided into phases recurring in cycles but the time
spent in the two phases can be estimated very differently . It is difficult to
compare reports from different laboratories when the criteria upon which the
estimates are based differ.
One may suppose that the question of criteria to define sleep phases is a
question of relatively minor importance. There are two reasons for supposing
the point deserves some attention. First of all, if it is assumed that changes in
the proportion of the various sleep phases during the early life of the or-
ganism can tell us something about the function of sleep and the development
of the central nervous system, then it becomes important to measure these
changes in a precise way. It has been argued, for example, that active sleep in
utero has the function of 'exercising' the CNS (Roffwarg, Muzio, and Dement
1966) by providing inputs which are the equivalent of the stimulation de-
livered via the external senses after delivery. On this argument one might
expect to see a drop in the proportion of active sleep at term when the
internally generated stimulation would no longer be necessary. To establish
such a proposition exact estimates of REM time are required .
A second reason for treating the description of the differing phases of
sleep as a point of importance is that studies of some abnormal groups have
been made, for example mongol infants, with the idea of demonstrating that
their pattern of sleep differs from normal (Goldie, Curtis, Svendsen, and
Robertson 1968). Evidently in such studies it is necessary to have a clear
description of what is normal in order to demonstrate what is abnormal.
The study which I shall report was undertaken with two aims in view. The
first was to compare estimates of infants' sleep phases based upon poly-
graphic measures used both singly and jointly; the second aim was to compare
the polygraphic measures with the results obtained by a human observer. Our
object was to prepare a scheme for scoring sleep states suitable to be used by
a human observer working alone in the infant's home. We therefore required a
scheme which had been validated by systematic comparison with measures
obtained by the objective polygraphic method of recording and scoring.
A group of 17 newborn, term infants was examined . Each child was
selected from a full-term nursery in the Kingston General Hospital after an
examination of the history of pregnancy and delivery showed an uneventful
clinical career. Each baby was taken to the sleep laboratory before a feed and
the electrodes necessary to make a polygraphic recording of sleep were ap-
plied before and during the feed. The following recordings were made : EEG ;
EKG ; EMG; bodily activity; respiration ; horizontal and vertical eye move-
ments. The record began as soon as the baby was put down after the feed and
lasted for 120-180 min. While the baby slept he was watched by three ob-
servers. Each of the observers was required to make a record of the child's
Sleep in the Newborn 159

Horizontal EM

Vertical EM

EMG

Respiration

Heart rate

Figure l Record of a quiet sleep period in a newborn infant

Horizontal E-M

Vertical E-M

EMG

EEG (F3-F4)

~ ----
Heart rate

Figure 2 Record of an active sleep period in a newborn infant

160 A Triune Concept of the Brain and Behaviour

Active

Transitional active

Transitional quiet

Figure 3 Exemplary passages from infant's EEG records

overt behaviour. One observer, who controlled the polygraph, made a note of
the number and kind of body movements; another made use of a sleep rating
scale which had been used in other similar studies; and the third made a
record of the occurrence of visible eye movements.
Scoring of the records began with an examination of EEG and respiration.
In the records one can see the infant pass through a series of graded changes
which describe a continuum ranging between quiet sleep and active sleep.
Figures I and 2 show typical examples of quiet and active sleep records.
A substantial proportion of most records contained passages of sleep
which could not readily be classified as either of the quiet or active sleep
patterns shown in these figures. Rather, they were an admixture of both
types. Nevertheless, it was possible, by using definite criteria, to classify sleep
of this nature as being either predominantly quiet or predominantly active.
To such sleep we have given the names Transitional Quiet and Transitional
Active, respectively.
The two varieties of transitional sleep are of key importance. First, it is a
form .of sleep which is peculiar to the infant - although the phases may later
be subsumed into the four-stage classification of NREM sleep in the adult.
Second, the estimation of the proportion of REM sleep, and the description of
the whole active/quiet cycle must be affected by the treatment given this
Sleep in the Newborn 161

CLASSIFICATION OF EEG DURING SLEEP

Cycle 1 10min

Periods 1-I____
A_c_ti_ve_________~,-Q_u_ie_t----1l1--1I

Cycle2

Periods t-----Ac....;..:_ct-'-'iv_e.;...__ _ _ _ _ _ _ _-+-~---Q..;;.;.;;.u_ie_t_ _ _-H

Cycle3 l■~-■1111
t:JTransitional quiet fl Transitional active

Cycle4 Oauiet ■ Active

Figure 4 Sleep cycles in terms of four categories of sleep

stage. If the transitional phases are assimilated to active sleep then one in-
creases the proportion of REM sleep; if they are assimilated to quiet sleep
then one diminishes the proportion of REM sleep. The differences can be
considerable; for example, in one subject the proportion of the sleep cycle
spent in transitional phases was 25 per cent. If this is said to be active sleep
then 85 per cent of the sleep time is REM sleep but, if it is said to be quiet
sleep, then 60 per cent is REM sleep.
The four categories of sleep we have defined - active, transitional ac-
tive, transitional quiet, and quiet - form the basis of the scoring system
devised for EEG and respiration. Eye movements inevitably accompany
active sleep, but may also be observed in all other stages - even occa-
sionally in quiet sleep. Every minute of the EEG and respiration records
was allotted to one of the four categories: quiet, active, transitional
quiet, and transitional active. The results of this division applied to one
EEG record will serve as an example. Figure 3 shows exemplary passages
from an EEG record. The top line shows an active sleep EEG pattern, the
162 A Triune Concept of the Brain and Behaviour

SLEEP CLASSIFIED FOR EEG RESPIRATION AND EYE MOVEMENTS

EEG~

RESP~

EM~

10min ~Active Dauiet

L....----l

Figure 5 Sleep cycles as defined by independent analyses of EEG, respiration, and

eye movements

bottom line shows a quiet sleep EEG pattern, and the two lines in
between show transitional sleep.
Using these four categories one can analyse an infant's EEG record in the
manner shown in Figure 4 . As may be seen, successive minutes which can be
scored as one category tended to occur in clusters. A passage which is of one
predominant type can be called a period, and such periods correspond to the
usual way of describing phases of infant sleep. However, while rules may be
laid down for defining a period, in practice they are not always easy to apply,
especially where there are frequent minute to minute changes of sleep type.
Moreover, periods often contain passages of sleep which are not of the pre-
dominant type. An alternative way of describing the proportions of record
occupied by different sleep stages is simply to use minute by minute scores,
regardless of the general character of the period in which they occur. We have
calculated sleep stages by both methods but the results cited here are
estimates based upon minute by minute tallies.
One period which can be unambiguously delineated by our criteria is the
active or REM period. We have made use of this fact to determine the length
of the sleep cycle. The cycle is defined as a passage of sleep extending from
the beginning of one active period, through the other sleep stages, to the
beginning of the next active period. This definition is illustrated by cycles 2
and 3 in the fourth figure. The cycle, rather than the whole sleep record, pro-
vides a denominator for statements regarding proportions of sleep occupied by
different stages. It also permits an analysis of the changes in these proportions
as sleep progresses because two or three cycles are usually obtained in a
recording session.
Sleep in the Newborn 163

TABLE 2
Quantitative analyses of sleep cycles

1st cycle 2nd cycle

x SD % x SD %

EEG Q 11.37 8.56 21.39 14.25 7.64 23.60

A 30.87 7.38 57.63 32.00 12.23 53.23
TA 5.50 3.82 10.22 4.25 2.91 7.57
TQ 5.87 5.44 10.76 7.37 4.78 15.60
Total 53.61 57.87

Respiration Q 9.91 6.85 20.99 7.92 4.68 13.98

A 28.33 13.30 52.91 31.17 15.95 51.15
TA 7.67 4. 12 15.34 10.17 5.95 20.60
TQ 5.58 3. 17 10.76 7.58 4.34 14.27
Total 51.49 56.84

Eye movements Q 19.41 5.71 35.89 19.50 4.94 37.54

A 35.50 9.87 64.11 36.75 14.72 62.46

Observer scale Q 20.58 5.21 37.85 21.17 4.80 40.30

A 34.75 9.84 62.15 34.83 13.95 59.69

In our experience the first passage of sleep may fall into any one of the
four defined categories. It has been said (Roffwarg et al. 1966) that infants
pass through an initial REM period but we did not see this in every case. There
follows a relatively unpatterned interchange between different categories
until the emergence of the first substantial active period, usually around the
fortieth minute. Regular cycles then begin. Figure 5 shows, for one baby, the
cyclical pattern arrived at independently for EEG, respiration, and eye move-
ments . It can be seen that there is agreement, but not a perfect agreement,
between these independent measures, and this is the case for many but not all
babies. In the group of 17 babies, the mean length of the first cycle, using the
EEG as the criterion measure, was 53 minutes, of which 58 per cent was spent
in active sleep, 21 per cent in quiet sleep, and 21 per cent in transitional sleep
(equally divided between transitional quiet and transitional active).
Table 2 shows means and standard deviations of the time spent in the four
sleep categories for EEG and respiration, of the time with eye movements
present or absent, and of the time spent in the two major classifications used
by the observers. It may be seen that second cycles tend to be longer than
first and to have a lower proportion of active sleep. However, we may also say
that, in the second cycle, respiration has a tendency to move toward the
active end of the continuum, and EEG has a tendency to move toward the
164 A Triune Concept of the Brain and Behaviour

quiet end of the continuum. Eye movements, it may be seen, are present for
almost exactly the same time as the sum of EEG active and transitional active
minutes in both cycles.
From these results one may deduce that there are at least two reasons why
estimates of the proportion of REM sleep have varied so much: different
criteria provide different figures, and different cycles provide different
figures . But one should note that the different estimates provided by the
different measures used in this experiment are not large enough to account
for the differing estimates found in the literature.
Another source of the variance of these estimates may lie in the clinical
history of the infant. We have observed sleep, using an observer equipped
with a rating scale in another group of 20 infants. Of these 20, seven were
subsequently found to have had at least one feature in their history which
one can regard as a clinically adverse factor (without any necessary long-term
implications) such as a low Apgar score at 5 minutes after birth. In these
seven infants the amount of R EM sleep was significantly greater than it was in
the other 13 infants. When their sleep records were examined in detail it was
found that this increase was due to a change in the pattern of sleep. The
suspect group had an earlier onset of REM sleep so that, by the time the
period of observation had ended, they had completed their second cycle
when the non-suspect group had yet to do so.
Taken together, these results suggest that, in the newborn, a feature of
sleep to which interest must be directed is the amount of transitional sleep in
relation to the criteria used to define transitional sleep and the form of the
sleep pattern. Parmelee (Parmelee et al. 196 7) has recently shown that as the
infant matures the amount of transitional sleep declines, and this decline can
be thought of as a reflexion of the increasing maturity of the brain. The
amount of transitional sleep, and changes in the amount of transitional sleep,
can be interpreted as a result of differential rates of maturation in the brain
stem reticular centres which control the phases of sleep.
One can also ask what light these results throw on the function of active
sleep. The most attractive hypothesis in our view is that the REM phase has a
combined alerting and exercising function which, in the immature system,
provides one element necessary for proper development and growth. In the
mature system other functions may be acquired , and dreaming is perhaps one
of these . But in the infant, as in the many small mammals who spend the
greater part of their time asleep, it seems reasonable to suppose that active
mechanisms are necessary to do for the brain what manoeuvres do for a
standing army in peace time.
Sleep in the Newborn 165

REFERENCES

Aserinsky, E. and N. Kleitman, Two types of ocular motility occurring in sleep. J. Appl.
Physiol., 1955a, 8: 1-10
Aserinsky, E. and N. Kleitman, A motility cycle in sleeping infants as manifested by
ocular and gross bodily activity. /bid., 1955b, 8: 11-18
Delange, P ., J. Cadilhac, and P. Passouant, Les divers stades du sommeil chez le nouveau-
ne et le nourisson. Rev. Neurol., 1962, 107: 271-6
Dement, W. and N. Kleitman, The relation of eye movement during sleep to dream ac-
tivity: an objective method for the study of dreaming. J. Exp. Psycho/., 1957,
53: 339-46
Feinberg, I. and V.R . Carlson, Sleep variables as a function of age in man Arch. Gen.
Psychiat., 1968, 18:239-50
Goldie, L., J.A.H . Curtis, U. Svendsen, and N.R.C. Robertson, Abnormal sleep rhythms
in mongol babies. lancet, 1968, 1: 229-30
Hartmann, E., The Biology of Dreaming. Springfield, Ill.: C.C. Thomas, 196 7
Jouvet, M., Neurophysiology of the states of sleep. Physiol. Rev., 1967, 47: 117-77
Monod, N. and N. Pajot, Le sommeil du nouveau-ne et du premature. I. Analyse des
etudes polygraphiques (mouvements oculaires, respiration et EEG) chez le nouveau-
a
ne terme. Biol. Neonatale, 1965, 8:281-307
Parmelee, A.H., W.H . Wenner, Y.Akiyama, E. Stern, and J . Flesch er, Electroencephalo-
graphy and brain maturation. In A. Minkowski (ed.), Regional Development of the
Brain in Early Life. Oxford: Blackwell, 1967
Roffwarg, H.P., J.N. Muzio, and W. Dement, Ontogenetic development of the human
sleep-dream cycle. Science, 1966, 152 :604-19
Snyder, F ., The new biology of dreaming. Arch. Gen. Psychiat., 1963, 8: 381-91