ArticlePDF AvailableLiterature Review

Donald O. Hebb and the Organization of Behavior: 17 years in the writing

Authors:

Abstract and Figures

The Organization of Behavior has played a significant part in the development of behavioural neuroscience for the last 70 years. This book introduced the concepts of the "Hebb synapse", the "Hebbian cell assembly" and the "Phase sequence". The most frequently cited of these is the Hebb synapse, but the cell assembly may be Hebb's most important contribution. Even after 70 years, Hebb's theory is still relevant because it is a general framework for relating behavior to synaptic organization through the development of neural networks. The Organization of Behavior was Hebb's 40th publication. His first published papers in 1937 were on the innate organization of the visual system and he first used the phrase "the organization of behavior" in 1938. However, Hebb wrote a number of unpublished papers between 1932 and 1945 in which he developed the ideas published in The Organization of Behavior. Thus, the concept of the neural organization of behavior was central to Hebb's thinking from the beginning of his academic career. But his thinking about the organization of behavior in 1949 was different from what it was between 1932 and 1937. This paper examines Hebb's early ideas on the neural basis of behavior and attempts to trace the rather arduous series of steps through which he developed these ideas into the book that was published as The Organization of Behavior. Using the 1946 typescript and Hebb's correspondence we can see a number of changes made in the book before it was published. Finally, a number of issues arising from the book, and the importance of the book today are discussed.
Reverberatory circuits and the cell assembly. These figures show the evolution of the reverberatory circuits that Hebb developed into his cell assembly. (a) A diagram of the pathways connecting interneurons among themselves and with the ocular motoneurons. V = vestibular nerve; I to 6 = cells in the primary vestibular nuclei; 7, 8, 9 = cells in the reticular formation in the medulla (Med.) and pons (P.); 10, 11, 12 = cells in the reticular nuclei in the midbrain (M.b.); Oc.n. = oculomotor nuclei; Fl, F2 and Col. = positions of the stimulating electrodes. The response of the motoneurons can be recorded with electrodes (R) from the trochlear or oculomotor nerve (III). Delivery of a shock to these nerves outside the brain stem through electrodes (A) causes the arrival of antidromic impulses at the motoneurons. The diagrams below illustrate the two types of chains, M = multiple and C = closed, that are found in the internuncial system. The closed loop system (C) represents the reverberatory circuit. ([125], page 407). (b) Schematic arrangement of neurons to account for conditioning by closed chains. The assumption is made that activation of a nerve cell occurs only when the cell receives excitation from two axon endings simultaneously. Impulses in CS are ineffective with respect to neurons a and b. If impulses in UncS and CS arrive simultaneously, however, they summate to excite b. This sets the closed chain b-c in continuous activity, and impulses in CS now summate with collaterals from fibers in the chain to activate a and RuncS. Simultaneous excitation of b will occur by chance if CS and UncS are stimulated together repeatedly. The chances of simultaneity are increased if the frequency of impulses in CS and UncS is greater; i.e., if the intensity of the stimuli is greater. While the speed of conditioning in any single neuron unit is thus largely a matter of chance according to this scheme, the sum of the changes in many such units would result in a gradual increase in the number of RuncS fibers activated. This scheme is not elaborated here to account for extinction, or other phenomena of conditioning. Inhibitory effects might be introduced by the addition of specific inhibitory collaterals, or by consideration of temporal relations resulting in refractory period decrement. CS = Conditioned stimulus; UncS = Unconditioned stimulus; Rcs = conditioned response; RuncS = Unconditioned response. (From [91], page 331). (c) Diagram of neurons and their synaptic connections illustrating the principle of recurrent (reverberatory) nerve circuits as seen in the IIIrd nerve nucleus (After Lorente de Nó). (From [139], page 64). (d) Hebb's diagram of his cell assembly. Arrows represent a simple "assembly of neural pathways or open multiple chains firing according to the numbers on each (the pathway "1, 4" fires first and fourth, and so on), illustrating the possibility of an "alternating" reverberation which would not extinguish as readily as that in a simple closed circuit. (From Hebb [77], page 73). (e) An illustration of the way in which learning might modify the functioning of cortical circuits and establish a cell-assembly. It is assumed that A-B-C and D-E-F, in association cortex, are excited by the same sensory event (axons from the sensory cortex are not shown, but it is assumed that they excite these cells separately). If A then delivers impulses to B at the moment when B is being fired by axons from sensory cortex, the synapse A-B will be "strengthened", and similarly with the other synapses. As a result of this strengthening the excitation of one cell may become able to set up reverberation in the circuit. Broken lines show possible connections between the two circuits, which would permit them to function as one system. (From [80] page 104). a Reprinted from Lorente de No R. Transmission of impulses through cranial motor nuclei. J Neurophysiol. 1939;2:402-64 [125]. Copyright (1939), with permission from the American Physiological Society. b Reprinted from Hilgard ER, Marquis DG. Conditioning and learning. New York and London: Appleton-Century Company; 1940 [91]. Copyright (1940), with permission from Appleton-Century-Crofts. c Reprinted from Morgan CT. Physiological Psychology. New York and London: McGraw-Hill Book Company Ltd; 1943 [139]. Copyright (1943), with permission from McGraw Hill. d Reprinted from Hebb DO. The organization of behavior; a neuropsychological theory. NY: Wiley; 1949. [reprinted 2002 by Lawrence Erlbaum Associates, Mahwah, New Jersey] [77] Copyright (2002), with permission from Lawrence Erlbaum Associates. e Reprinted from Hebb DO. A textbook of Psychology. Philadelphia and London: W. B. Saunders Company; 1958 [80]. Copyright (1958), with permission from Mary Ellen Hebb
… 
Content may be subject to copyright.
R E V I E W Open Access
Donald O. Hebb and the Organization of
Behavior: 17 years in the writing
Richard E. Brown
Abstract
The Organization of Behavior has played a significant part in the development of behavioural neuroscience for the
last 70 years. This book introduced the concepts of the Hebb synapse, the Hebbian cell assemblyand the Phase
sequence. The most frequently cited of these is the Hebb synapse, but the cell assembly may be Hebbs most
important contribution. Even after 70 years, Hebbs theory is still relevant because it is a general framework for
relating behavior to synaptic organization through the development of neural networks. The Organization of
Behavior was Hebbs 40th publication. His first published papers in 1937 were on the innate organization of the
visual system and he first used the phrase the organization of behaviorin 1938. However, Hebb wrote a number
of unpublished papers between 1932 and 1945 in which he developed the ideas published in The Organization of
Behavior. Thus, the concept of the neural organization of behavior was central to Hebbs thinking from the
beginning of his academic career. But his thinking about the organization of behavior in 1949 was different from
what it was between 1932 and 1937. This paper examines Hebbs early ideas on the neural basis of behavior and
attempts to trace the rather arduous series of steps through which he developed these ideas into the book that
was published as The Organization of Behavior. Using the 1946 typescript and Hebbs correspondence we can see a
number of changes made in the book before it was published. Finally, a number of issues arising from the book,
and the importance of the book today are discussed.
Introduction
This meeting celebrates the 70th anniversary of the publi-
cation of The Organization of Behavior by Donald O. Hebb
[77]. Since its publication, The Organization of Behavior
has become one of the most influential books in Psych-
ology and Neuroscience (over 31,200 Google scholar cita-
tions in January 2020). In 2017 Dick Passingham named
The Organization of Behavior one of the top 5 books in
Cognitive neuroscience
1
because Hebbsideas have turned
out to be incredibly powerful in understanding how the
brain actually works[44]. According to Adams ([2], page
419), HebbsOrganization of Behavior and DarwinsOn the
Origin of Species are two of the most influential books in
the history of biology. Because Hebbs book had been out
of print since 1966, Peter Milner and I had it reprinted by
Lawrence Erlbaum in 2002. In our Foreword to this edition
[22], we included a list of Hebbs publications and a list of
biographies and obituaries of Hebb. This paper will focus
on the work done by Hebb between 1932 and 1949, which
led to the publication of The Organization of Behavior.
The Organization of Behavior introduced the concepts
of the Hebb synapse,theHebbian cell assemblyand
the Phase Sequence. While the Hebb synapse has be-
come the most cited, and better known than Donald
Hebb himself[166], the cell assembly may be his most
lasting legacy [41,97,122,153,163]. Research on the
phase sequence has lagged behind, but multi-electrode re-
cording techniques have enabled researchers to investigate
© The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data.
1
The other four books were The concept of mind by Gilbert Ryle,
Perception and communication by Donald Broadbent, Evolution of the
brain and intelligence by Harry Jerison and Images of mind by Michael
Posner and Marcus Raichle
Correspondence: rebrown@dal.ca
Department of Psychology and Neuroscience, Dalhousie University, Halifax,
Nova Scotia B3H 4R2, Canada
Brown Molecular Brain (2020) 13:55
https://doi.org/10.1186/s13041-020-00567-8
the integration of cell assemblies into larger phase se-
quences [4,129]. The Organization of Behavior remains
influential because it continues to stimulate research in
many areas of neuroscience including studies of learning
and memory; the long-term effects of environment on de-
velopment; aging; computer modeling of the brain, robot-
ics, and artificial intelligence [21]. This meeting is
evidence of Hebbs continued influence on research into
synaptic function in learning and memory. This paper de-
scribes the long and arduous process that Hebb went
through in order to write and publish The Organization of
Behavior. It uses unpublished notes, papers, letters and a
typescript of the first draft to show the development of
the ideas that went into this seminal book.
A brief overview of Hebbs life and early
education (Table 1)
Hebb [81,83] published two autobiographical papers
and details of Hebbs life and work are given in previous
papers [18,19,23], so only a brief overview is given here.
Hebbs father, Arthur Morrison Hebb, and his mother,
Mary Clara Olding, both received their medical degrees
from Dalhousie University and were physicians in the
village of Chester, Nova Scotia. Donald was born on 22
July 1904, the first of four children. His brother Andrew
(19052005) received a law degree from Dalhousie and
went into business. Peter (19091955) was a physician
and Catherine (19121978) received her PhD in Physi-
ology from McGill University, studying the physiology of
the digestive system with Dr. Boris Babkin. Later she
won a scholarship to Edinburgh University and did re-
search on the biosynthesis of acetylcholine at the Insti-
tute of Animal Physiology at Babraham, Cambridge [56].
Hebb went to the Chester School until grade 11 and
then completed high school at the Halifax County Acad-
emy before entering Dalhousie University in Halifax in
1921. He majored in English and Philosophy, with the
intention of becoming a novelist [83]. During the time that
he was an undergraduate, Psychology was taught in the
Philosophy Department and in the 192223 academic year
Hebb took Philosophy 1, Logic and Psychologyfrom
Professor H.L. Stewart (who wrote Questions of the day in
philosophy and psychology, in 1912). According to the
Calendar of Dalhousie University, 192122, the textbooks
for this class were An Introductory Logic by James Edwin
Creighton (1919), and the Textbook of Psychology by Wil-
liam James (1892), with references made to The Principles
of Psychology by William James (1890) and An Introduc-
tion to Social Psychology by William McDougall (1908).
In 192425, Hebb took Philosophy 8 Philosophic ideas
in Literature with Professor H.L. Stewart, a course which
included a study of philosophic ideas in Tolstoy, Hardy,
Anatole France, H.G. Wells, Ibsen, Morley, Fredric Harri-
son, Mrs Humphrey Ward, Rabindranath Tragore, Wilfred
Ward, George Meredith(Calendar of Dalhousie Univer-
sity, 192425). In the same year he took Philosophy 9 Ex-
perimental Psychology, from Professor N.J. Symons, in
which the textbook was An Elementary Laboratory Course
in Psychology by H.S. Langfeld and F. H. Allport (1916).
Thus, Hebb had taken 3 courses that contained topics in
Psychology while completing his B.A., even though he took
no Psychology courses per se.
Hebbs M.A. Thesis [58]
After graduating from Dalhousie in 1925, Hebb obtained a
teaching certificate from the Provincial Normal College in
Truro, Nova Scotia, and became the principal of his old
school in Chester for a year. He then moved to Montreal,
where he was a teacher and a part-time graduate student
in Psychology at McGill University under the supervision
of Professor Chester Kellogg. During this time, he started
an educational experiment in his school as he found that
students of all intellectual abilities were failing. He decided
to change the school procedures to facilitate education,
giving students no homework and no punishment for in-
attention. He persuaded the children that school-work
was a privilege, gave them interesting things to do in class
and sent any who disrupted the class outside to play [57].
As a graduate student at McGill University, he took a
course in experimental psychology and seminars in sys-
tematic psychology, with a minor in Education. For his
M.A. thesis, he studied SherringtonsIntegrative activity
of the Nervous System [170] and PavlovsConditioned
Reflexes [149], and wrote a theoretical M.A. thesis enti-
tled Conditioned and Unconditioned Reflexes and Inhib-
ition [58]. In 1980, Hebb wrote that My M.A. thesis,
written in bed [while he was ill with a tubercular hip]
tried to show that skeletal reflexes are a product of
intra-uterine learning. This was nonsense, but no imme-
diate disproof was available at the time([83], page 282
283). However, a reading of Hebbs M.A. thesis shows
that this comment is untrue. As stated in the Introduc-
tion to this thesis:
The purpose of this paper is to present a theory of
the functioning of the synapse based on the experi-
mental work of Sherrington and Pavlov, on reflexes
and inhibitions.
The implications of these things for psychological
theory, in some aspects, has been far from clear. In
looking for a firm basis for psychology in physi-
ology, there are some peculiarities about the results
of both investigators which demand serious consid-
eration and suggest another interpretation of their
work.([58], page 2).
There is nothing about intra-uterine learning in this the-
sis. The four chapter headings are (I) Functioning of the
Brown Molecular Brain (2020) 13:55 Page 2 of 28
synapse in the conditioned reflex, (II) The unconditioned re-
flex, (III) Unconditioned inhibition and (IV) Conditioned in-
hibition: hypnosis, sleep, and the waking state. Chapter 1
contains Hebbs first illustration of his idea of the synaptic
changes associated with conditioning (see Fig. 1a). This fig-
ure shows branches of a stimulus input going to two
unconditioned reflex arcs. Pavlov showed that if one
of the reflex arcs is active when a stimulus is applied
to the input, the branch going to that reflex becomes
more potent; branches going to inactive effectors do
not. Hebb summarized this by stating that: An ex-
cited neuron tends to decrease its discharge to in-
active neurons, and increase this discharge to any
active neuron, and therefore to form a route to it,
whether there are intervening neurons between the
two or not. With repetition this tendency is prepotent
in the formation of neural routes.([58], page 13).
This was the first description of Hebbstheoryofsyn-
aptic function in learning.
At the end of section II of this thesis, Hebb states that:
There are only two types of simple unconditioned reflex,
though these may combine in more complex forms: (1)
the reaction such as to increase stimulation (positive) and
(2) such as to decrease it. This is a feature of reflex activity
which has not been taken account of before as a general
characteristic of all reflexes, and it suggests a genesis of the
reflex as due to environment just as the laboratory condi-
tioned reflex is due to an experimenter.([58], page 28).
One of the important features of this thesis was the dis-
cussion of unconditioned and conditioned inhibition. On
Fig. 1 Hebbs early concepts of synaptic change during learning. (a) Figure 1from Hebb's MA thesis [58] shows his early concept of the synaptic
changes underlying conditioning. (b) Figure 2from [59] shows two possible routes of synaptic activity. An axon A with two terminal branches A
and Aactivates B and C. With greater activity of B, the route A’–B will be strengthened, but if there is an interneuron D, which is activated by A,
the route A’–DC will be strengthened. (c) A possible mechanism of reflex inhibition. This figure was attached to Hebbs 1934 essay [59]. It shows
that his concept of inhibitory synapses was well developed in 1934, but was later deleted from his theory. a,b,c: Reprinted from unpublished
papers of D.O Hebb held in McGill University Archives, Montreal, Quebec, file MG1045 [58,59]. Permission has been obtained from Mary Ellen
Hebb to reprint them. aand bare also in: Brown RE, Milner PM. The legacy of Donald O. Hebb: more than the Hebb synapse. Nat Rev Neurosci.
2003;4:10139 [23]. Ownership of copyright in original research articles remains with the author, Richard E. Brown
Brown Molecular Brain (2020) 13:55 Page 3 of 28
unconditioned inhibition, he says: In reciprocal inhibition it
is noteworthy that it occurs with reflexes which could not
function simultaneously even if one of them were not
inhibited. Flexion inhibits extension, but if this were
not so the two reactions still could not take place at
the same time. This suggests strongly that inhibition
is not established by heredity but by the actual func-
tioning of the organism itself; a result of the oppos-
ition of two reflexes.(page 31). The summary
statement of the thesis (page 6062) is as follows:
Supposing the individual synapse to obey the same
general laws in all neural activity, an analysis of Pav-
lovs work on the conditioned reflex gives the fol-
lowing generalization: An excited neuron tends to
decrease its discharge to inactive neurons and in-
crease this discharge to any active neuron, tending
to form a route to it. With repetition this tendency
is prepotent in the formation of neural routes.
The unconditioned reflex appears then as a reflex
conditioned by the environment. This occurs in two
ways, according to the kind of reaction. The uncon-
ditioned reflex may be grouped into (1) those re-
flexes whose reaction increases the stimulus, as the
exterior thrust, or muscle tone, (2) those whose re-
action decreases the stimulus, which is strong or
persistent, as the salivary reflex to HCl or sand
placed in the mouth, and (3) those which may com-
bine (1) and (2) in varying ways, as the scratch re-
flex. If the generalization of neural action above is
sound, these reflexes may be regarded as the result
of the functioning of the law which established the
conditioned reflex in the laboratory, under the
interaction of environment and organism only.
Similarly the inhibition of the spinal cord may, under
this generalization, be a conditioning by environment,
in the interaction of the spinal reflexes. Activity of the
opposed effectors would facilitate a re-routing of the
impulses from the inhibited receptor so that the
inhibited effector is not excited. The exact adjustment
of this process is effected by the repeated opposition
of the two reflexes, until synaptic block between
inhibited receptor and inhibiting effector is lowered
to the point that impulses may follow this route
whenever the inhibiting effector is more active than
the inhibited effector. The latter thereupon loses tone
because of the change of route of impulses from the
corresponding (inhibited) receptors.
This account of unconditioned inhibition seems to be
considerably more satisfactory than the rather improb-
able refractory-phase hypothesis. [The hypothesis that
inhibition is due to the development of a prolonged re-
fractory phase in the tissues concerned. See Telford
[178]]. Together with a simplified account of the un-
conditioned reflex, this is argument for giving serious
consideration to the generalization above. Applying it
to conditioned inhibition adds further support.
Here also inhibition appears as a re-routing of the
nervous activity. The inhibiting agent is the con-
stantly active reflexes of posture, breathing, and of
visceral activity generally. This means that upon es-
tablishment of an internal inhibition these activities
(postural and visceral) would be re-inforced, but this
in a localized inhibition would be too weak to be dis-
cernable. When inhibition becomes general, however,
as in hypnosis, this re-inforcement becomes very evi-
dent, in the catalepsy of the animal: a very strong
maintenance of the posture. ([58], page 6062).
These detailed quotations show that this thesis had
nothing to do with intra-uterine learning and that Hebb
had seriously considered the importance of inhibition as
well as excitation in the reflex arc. Thus, the ideas for the
Hebb synapse began with Hebbs M.A. thesis and these
ideas developed and changed over the next 17 years.
Hebbs unpublished paper on neural action [59]
After the completion of his M.A. thesis, Hebb started his
PhD research at McGill on classical conditioning with
Drs. Boris Babkin and Leonid Andreyev, both of whom
had worked with Pavlov in St. Petersburg. Hebb, how-
ever, became disillusioned with Pavlovian conditioning
procedures and his graduate studies, and left McGill in
the fall of 1934 to complete his PhD with Karl Lashley at
the University of Chicago [83]. Lashley had just pub-
lished his book Brain Mechanisms of Intelligence [112]
and his American Psychological Association presidential
Table 1 A Brief Time-Line of Hebbs Career
Brown Molecular Brain (2020) 13:55 Page 4 of 28
address outlined the state of the art of physiological
psychology [113]. After he moved to Chicago, Hebb be-
came part of a group of exceptional psychobiology stu-
dents that studied with Lashley. These included Frank
Beach, David Krech, Norman Maier, and Theodore
Schneirla [37]. At Chicago Hebb took classes with L.L.
Thurstone, Wolfgang Köhler, Karl Lashley, C.J. Herrick
and Nathaniel Kleitman [83] and began to think ser-
iously about the neural control of behaviour.
In November 1934, Hebb submitted a paper, entitled
The interpretation of experimental data on neural ac-
tionfor his class in Elementary Neurology (Anatomy
316), taught by C. J. Herrick [59]. In this paper, Hebb
drew two diagrams of possible synaptic activity during
conditioning (see Fig. 1b) and stated that: There is as
yet no understanding theoretically of the process of con-
ditioning, and when this is understood, it may throw
light on the reflex activity of the [spinal] cord, and per-
haps account for the existence of some of the apparently
rigidly inherited patterns. The conception of block and
facilitation at the synapse postulates a process which can
strengthen and perpetuate a route once formed, but
none whatever to account for the establishment of the
route in the first place - a most important weakness in
the whole theory.([59], page 15). Hebb had hoped to
re-write this paper for publication, but this was never
completed. In a letter to Babkin (Hebb to Babkin, 6
January 1935), Hebb explained that Lashley did not favor
the publication of papers which did not include experi-
mental tests of the hypothesis proposed and concluded
that Professor Lashley did say that the paper might be
re-written, so as not to claim too much, but is on the
whole against publication without experimental data.
A closer examination of this paper shows that Hebb
was interested in the neural route of the reflex arc. He
stated that the conception of the reflex route [is based
on] the assumption of relatively unchanging paths along
which an excitation peripherally aroused is conducted
([59] page 2). Much of this essay concerns the concept
of inhibition, and he stated that: Studies of route forma-
tion, of the reflex arc, of inhibition, facilitation or synap-
tic fatigue, depend on the conceptions obtained from the
phenomena of the isolated nerve.([59], page 1). He
stated that In Pavlovs[151] discussion of the reflex he
makes no mention of specific routes; his definition of
the term is essentially that whatever reaction is com-
pletely determined by natural law is a reflex([59], page
3). What Hebb was saying in this paper was that excita-
tion and inhibition of reflexes must occur in specific
neural routes and the patterns of activity (synaptic effi-
cacy) in these routes of nervous activity can be stimu-
lated or inhibited by experience. One can see Hebb
struggling with the concept of how a reflex arc is formed
by the combination of excitatory and inhibitory stimuli.
He suggested that there was an alternative to Pavlovs
conception of the neural route of the reflex:
In his interpretation of his data, of course, Pavlov
[151] represents the most complete assumption of
the linear passage of excitations along specific
routes. Not even in his defence of the reflex concep-
tion and of his interpretations does he make any ef-
fort to justify the point-for-point conception of the
receptor surfaces of the cortex; yet the actual data
suggest, if we have not this preconception, a rather
different process of excitation. The alternative, that
excitation spreads at every synaptic level, is both
more in conformity with the neurological facts of
structure and better able to account in a simple
manner for the generalizationof stimuli. On this
assumption, it is clear, there is not a separate spe-
cific cortical excitation corresponding to each per-
ipheral excitation, but a diffuse excitation of which
a large fraction might be common to two or more
peripheral excitations. This diffusity would be nar-
rowed down to a more specific focus only with the
process of differentiation.([59], page 7).
In preparation for rewriting this paper he added a note
that references would be made to Lorente de Nó (1933
and 1934), Forbes, Herrick (1929), Uexkull, Goldstein,
Graham-Brown ([59], page 17). Appended to this paper
was a drawing of a possible mechanism of reflex action
(see Fig. 1c).
Unpublished notes on scientific methods in
psychology [60]
While in Chicago, Hebb wrote five chapters (90 pages;
of which I have found only 69 pages) of a proto-book
entitled Scientific methods in psychology: A theory of
epistemology based on objective methods in psychology
[60]. In this manuscript, Hebb tried to integrate what
he learned about the study of mind as a philosophy
student with the neurophysiology he was learning in
Chicago. In chapter I, Thetheoryofmind, he dis-
cussed the philosophical concept of mind and the
introspective method in Psychology. He rejected
introspection as a method and then attempted to in-
tegrate the theory of mind with the concepts of Be-
haviorism and the physiological basis of mind. In
chapter 2 Idealism, he considered the role of the
brain and the mind in perception. While he did not
get very far with his arguments about the physio-
logical basis of mind, there are some glimpses of his
thinking on mind and brain. First, he states that:
We can say, like primitive man, that the existence
of the mind is to be inferred from human (or
Brown Molecular Brain (2020) 13:55 Page 5 of 28
animal) activity; we can say, strange as it may ap-
pear to the Behaviorist, that mind is itself, and al-
ways has been, simply a theory of behavior.([60],
page 22).
Second, he begins to develop a physiological theory of
mind:
In order to understand human activities in terms
of physiological theory it is not sufficient to sub-
stitute the words conditioned reflexfor associ-
ation,andbrainfor mind, thinking of the brain
as a distinct organ receiving impulses from differ-
ent routes, each route indicating some distinct
sensory perception. The modernising of older
forms of thought has sometimes been rather in-
genious. As soon as the neuron-synapse theory
began to make it possible to think of a com-
pletely neural determination of human activity, it
was assumed frequently that cortical or at any
rate cerebral processes alone correlated with the
process called conscious. There may be no con-
sciousness without the cortex; but there is no
reason to conclude that in the presence of the
cortex the functioning of sensory and sub-
cerebral neurons have nothing to do with con-
sciousprocesses.([60], page 39).
This proto-book includes chapters on The induction
in scientific method(Chapter III) and Mathematical
method, probability, and deductive reasoning(Also la-
beled Chapter III) and ends on page 69 without return-
ing to a discussion of the physiological basis of
behaviour. Parts of this book were to be included as a
final chapter in The Organization of Behavior but
were deleted during the writing process. Some of the
ideas expressed in this manuscript, however, did find
expression in Essay on Mind,Hebbs final book [84].
It is important here as it shows Hebbs early struggle
to develop a physiological theory of psychological
processes.
Hebbs PhD at Harvard [61]
Only a year after Hebb arrived in Chicago, Lashley ac-
cepted a position at Harvard and Hebb completed his
PhD thesis at Harvard on the visual abilities of rats
reared in the dark [62,63]. He also completed the re-
search that he had started in Chicago on field orienta-
tion in rats, and in these papers Hebb [64,65] first used
the term Organization of Behavior. In these early pa-
pers, Hebb tried to develop better behavioral tests for
the study of the effects of cerebral lesions. He stated
that: It is clear that the success of the physiological ana-
lysis depends on the adequacy of the behavioral tests
used. Much of the difficulty found in the evaluation of
the effects of cerebral destruction, both clinically and
with animal experimentation, is due to the fundamental
difficulty of the analysis of behavior and to the unsatis-
factory available accounts of it.([64], page 333).
When Hebb came to write The Organization of Behav-
ior, he re-analyzed the data from his PhD experiments. He
had originally reported that rats reared in total darkness
learned a visual discrimination task in the same way as
normally reared rats [62,63], but when he re-examined
his own data, he found that the dark-reared rats took six
times as many trials as the normally reared rats to learn a
discrimination of horizontal versus vertical stripes, and
twice as many trials for discriminating erect versus
inverted triangles ([77], page 113). In his memoirs ([83]
page 286), he wrote that It was nine years later, when I
was trying to account for some of Wilder Penfieldsbrain-
injury results and developing the ideas that led to The
Organization of Behavior, that I went back to my own
published data and for the first time saw what kind of
beast they were.Of these papers, Hebb notes: The re-
sults were interesting. In fact they were very interesting,
but the fact also is, as far as I can discover, that no one at
the time ever looked at either of my two papers [62,63].
At the Montreal Neurological Institute (MNI) with
Wilder Penfield (193739) and Queens University
(19391942)
After graduating from Harvard, Hebb moved back to
Montreal to study the psychological effects of temporal
lobe and frontal lobe surgery in the patients of Wilder
Penfield, the founder of the Montreal Neurological Insti-
tute [66,67]. His most complete study [88] was of patient
KM, who had a frontal lobotomy. Hebb tested this patient
before and after surgery and found little effect of the sur-
gery on his scores on the standardized tests available at
the time, and concluded that the removal of large
amounts of frontal lobe tissue either had no effect on the
mental abilities of the patient or that the tests used were
not sensitive enough to detect the effects of the surgery.
His experiences in testing patients at the MNI led to a
number of ideas about the nature of intelligence and how
it should be tested [87]. Hebb also observed that lesions of
different brain areas produced different cognitive impair-
ments and that the age at which a brain injury occurred
was important in determining its effects on intelligence
[68]. In this paper, he also observed that intelligence was
composed of two components; a fixed component and a
variable component that could be influenced by environ-
mental experiences. He called these Intelligence A and
intelligence Band, as has been shown elsewhere [20],
Raymond B. Cattell took this idea of two types of
intelligence and renamed them fluid and crystallized
Brown Molecular Brain (2020) 13:55 Page 6 of 28
intelligence [29]. In The Organization of Behavior (page
294) Hebb used the terms Intelligence A and Intelligence
B and did not refer to fluid and crystallized intelligence.
In 1939, Hebb took a position as Lecturer in Experi-
mental Psychology in the Philosophy Department at
Queens University in Kingston, Ontario. It is possible
that he learned a great deal from George Humphrey, the
head of his department, who had written a book entitled
The Nature of Learning (1933), but there is only one ref-
erence to Humphrey [94]inThe Organization of Behav-
ior. It was at Queens University that he designed a
variable path maze with a student, Kenneth Williams,
the so-called Hebb-Williams maze [89], which he later
used to test rats reared in enriched environments [76].
At the Yerkes primate center with Lashley (194247)
When Karl Lashley became the director of the Yale Primate
Laboratories at Orange Park, Florida in 1942, Hebb was
hired as a Research Fellow. He conducted studies on fear
and anger in chimpanzees and related these findings to hu-
man emotionality [73,75]. He also conducted a study on
the behaviour of dolphins [130] and continued his work on
the development of rat intelligence. To determine the ef-
fects of early experience on learning, Hebb reared rats as
petsathomeandshowedthatenrichedexperienceduring
development resulted in improved maze learning in adult-
hood. Only an abstract of this study, which was presented
at an APA meeting, was published [76]. These results
formed the basis of later studies at McGill on the effects of
environmental enrichment on behaviour by Hebbsstu-
dents and established the field of research on how environ-
mental influences shape neural development (see [21]).
During his years at Orange Park (Fig. 2), Hebb com-
pleted the first seven chapters of a manuscript of a book,
eventually published as The Organization of Behavior. The
intellectual climate at the Yerkes laboratories, as described
by Dewsbury [38], stimulated Hebb to return to his earlier
ideas on the physiology of behaviour and to formalize these
in a coherent form. In the preface to The Organization of
Behavior, Hebb states that My greatest debt, perhaps, is to
the weekly colloquium and the persistent theoretical de-
bate at the Yerkes Laboratories of Primate Biology between
1942 and 1947; and to a small group taking part therein
who have also read the entire manuscript and have con-
tributed greatly to it -- Professor Harry W. Nissen, Mr. and
Mrs.Robt.BlumandDr.AustinRiesen.([77], page viii).
Hebbs influences in writing the organization of
behavior
Hebb had a number of influences in writing The
Organization of Behavior. These included some of the
textbooks available at the time, the work of Lorente de
Nó on the synapse, the work of von Senden on visual
development and, as mentioned above, his colleagues at
the Yerkes Primate Centre. Lashley himself was both an
influence and an impediment. Figure 3shows three of
the most important influences on Hebb: Lashley, Köhler
and Lorente de Nó. In the Preface to The Organization
of Behavior, Hebb notes that there were five textbooks
that he would recommend as background reading:
Though I have done my best, it may be chimerical
to hope that my discussion is extensive and clear
enough to stand on its own feet, for the nonpsycho-
logical reader. The reader who needs it will find
more of the details of psychological theory in Mor-
gan [139] on physiological psychology, Hilgard and
Marquis [91] on the theory of learning, Woodworth
[189]onexperimental(normal human adult)
psychology, and Moss [140] or Maier and Schneirla
[127] on animal psychology. Of these, Morgan is
most directly relevant, and in several places I have
assumed a knowledge of fact to the extent provided
by his text.([77], page vii).
In this section, I provide brief outlines of the main in-
fluences on the development of Hebbs theories as pre-
sented in The Organization of Behavior. These include
the textbooks of Hilgard and Marquis [91] and Morgan
[139], the work of Lorente de Nó ([123,124], and von
Senden [186], Wolfgang Köhler and Gestalt Theory, and
Lashley himself.
Hilgard and Marquis [91]
Hebb gave a great deal of credit to Hilgard and Marquis
[91] for stimulating his ideas on synaptic change in
Fig. 2 Hebb and his family in Orange Park, Florida in the 1940s. This
photo shows Hebb, his wife and two daughters and Helen Riesen.
Reprinted from Dewsbury DA. Monkey farm: a history of the Yerkes
Laboratories of Primate Biology, Orange Park, Florida 19301965.
Lewisburg: Bucknell University Press; 2006 [38]. Copyright (2006),
with permission from the author and from Bucknell University Press
Brown Molecular Brain (2020) 13:55 Page 7 of 28
learning. Early in 1944 Hebb learned that Lorente de Nó
[123125] had recently shown that closed circuits were
to be found throughout the brain, and that one neuron
by itself may not be able to excite a second neuron at
the synapse, but can do so if supported by simultaneous
action from another neuron([83], page 295). It was
these ideas of Lorente de Nó that Hebb realized were
just what he needed to develop a neurological theory of
the mind([134], page 127). In his book Essay on Mind,
Hebb ([84], page 8384) said:
The problem of perception remained intractable
for about five years (1939 to 1944) and as a result I
made no progress in my attempt to understand con-
cepts and thought. In fact, by 1944 I had given up
trying to solve the problem. What happened then
was that I became aware of some recent work of
Lorente de Nó in conjuction with some observa-
tions of Hilgard and Marquis [91], which led me to
think of a possible solution for a different problem:
the problem of set and attention. And this brought
me straight back to concepts and the thought
process, but now from a different point of view.
The section of Hilgard and Marquis [91] that Hebb re-
ferred to is entitled The nature of synaptic modifica-
tion(pages 326334) in which the ways that synaptic
change can lead to the permanent neural modifications
responsible for learning are outlined. Five theories were
described: (1) anatomical growth of axons and dendrites;
(2) changes in the physico-chemical properties of the
nerve cell; (3) continued activity in closed neural chains,
(4) the influence of bio-electric fields and (5) Inhibition.
According to Hebb [83] it was the discussion of the
work of Lorente de Nó in these pages that inspired him
to develop the idea of a reverberating circuit in the cell
assembly.
Morgan [139]
Another reference cited by Hebb, Physiological Psych-
ology ([139], pages 6169) also has a section on synaptic
functionswhich contains a diagram of Lorente de Nós
recurrent nervous circuits and Gassers[46] diagram of
reciprocal inhibition. Thus, the ideas of Lorente de Nó
were part of the Zeitgeist as Hebb started to write his
book. Figure 4shows the different representations of
Lorente de Nós recurrent nervous circuits. Morgan
([139], pages 137-141) also discussed spontaneous neural
activity and referred to the data of Weiss [188] on tad-
poles and Jasper [98] on brain waves in humans, exactly
the papers used by Hebb ([77], pages 510). In his dis-
cussion of the neural basis of learning, Morgan ([139],
pages 518525) describes the neurobiotaxis, synaptic re-
sistance, fiber conductance, reverberation and resonance
theories as well as the drainage theory and the irradi-
ation theory. He also includes the Gradient theory and
the Pattern Theory of Lashley. Hebb ([77], pages 1011)
states that both Lashley and Köhler opposed the theory
of neural connections as the basis of learning, but he
notes that their critiques applied to the older theory of
linear, sensori-motor connections, in which a single cell
was supposed to be always capable of exciting a single
cell with which it synapsed. This was the idea of a rigid
synaptic reflex and Hebb notes that the concept of
modifiable synaptic resistance eliminated this critique.
Lorente de Nó
Hebb ([84], pages 8388) describes how the work of Lor-
ente de Nó [123,124], as discussed in Hilgard and Marques
[91]andMorgan[139] inspired his ideas on cell assemblies.
UntiltheworkofLorentedeNó,itwasthoughtthata
nerve impulse was activated only by an external stimulus
and lasted only a few milliseconds. Lorente de Nó showed
that nerve fibers could have recurrent loops and Hilgard
and Marquis ([91], page 327) suggested that If two cells
Fig. 3 Three of the most important influences on Hebb. Karl S. Lashley, Wolfgang Köhler and Rafael Lorente de Nó in 1951. Reprinted from
Jeffries, L.A. 1951. Cerebral Mechanisms in Behavior: The Hixon Symposium. New York: John Wiley and Sons. (Frontspiece). Copyright (1951), with
permission from John Wiley and Sons
Brown Molecular Brain (2020) 13:55 Page 8 of 28
ab
c
d
e
Fig. 4 Reverberatory circuits and the cell assembly. These figures show the evolution of the reverberatory circuits that Hebb developed into his cell assembly.
(a) A diagram of the pathways connecting interneurons among themselves and with the ocular motoneurons. V = vestibular nerve; I to 6 = cells in the
primary vestibular nuclei; 7, 8, 9 =cells in the reticular formation in the medulla (Med.) and pons (P.); 10, 11, 12 =cells in the reticular nuclei in the midbrain
(M.b.); Oc.n. = oculomotor nuclei; Fl, F2 and Col. = positions of the stimulating electrodes. The response of the motoneurons can be recorded with electrodes
(R) from the trochlear or oculomotor nerve (III). Delivery of a shock to these nerves outside the brain stem through electrodes (A) causes the arrival of
antidromic impulses at the motoneurons. The diagrams below illustrate the two types of chains, M =multiple and C = closed, that are found in the
internuncial system. The closed loop system (C) represents the reverberatory circuit. ([125], page 407). (b) Schematic arrangement of neurons to account for
conditioning by closed chains. The assumption is made that activation of a nerve cell occurs only when the cell receives excitation from two axon endings
simultaneously. Impulses in CS are ineffective with respect to neurons a and b. If impulses in UncS and CS arrive simultaneously, however, they summateto
excite b. This sets the closed chain b-c in continuous activity, and impulses in CS now summate with collaterals from fibers in the chain to activate a and
RuncS. Simultaneous excitation of b will occur by chance if CS and UncS are stimulated together repeatedly. The chances of simultaneity are increased if the
frequency of impulses in CS and UncS is greater; i.e., if the intensity of the stimuli is greater. While the speed of conditioning in any single neuron unit is thus
largely a matter of chance according to this scheme, the sum of the changes in many such units would result in a gradual increase in the number of RuncS
fibers activated. This scheme is not elaborated here to account for extinction, or other phenomena of conditioning. Inhibitory effects might be introduced by
the addition of specific inhibitory collaterals, or by consideration of temporal relations resulting in refractory period decrement. CS = Conditioned stimulus;
UncS = Unconditioned stimulus; Rcs = conditioned response; RuncS = Unconditioned response. (From [91], page 331). (c) Diagram of neurons and their
synaptic connections illustrating the principle of recurrent (reverberatory) nerve circuits as seen in the IIIrd nerve nucleus (After Lorente de Nó). (From [139],
page 64). (d)Hebbs diagram of his cell assembly. Arrows represent a simple assembly of neural pathways or open multiple chains firing according to the
numbers on each (the pathway 1, 4fires first and fourth, and so on), illustrating the possibility of an alternatingreverberation which would not extinguish
as readily as that in a simple closed circuit. (From Hebb [77], page 73). (e) An illustration of the way in which learning might modify the functioning of cortical
circuits and establish a cell-assembly. It is assumed that A-B-C and D-E-F, in association cortex, areexcitedbythesamesensory event (axons from thesensory
cortex are not shown, but it is assumed that they excite these cells separately). If A then delivers impulses to B at the moment when B is being fired by axons
from sensory cortex, the synapse A-B will be strengthened, and similarly with the other synapses. As a result of this strengthening the excitation of one cell
may become able to set up reverberation in the circuit. Broken lines show possible connections between the two circuits, which would permit them to
function as one system. (From [80] page 104). aReprinted from Lorente de No R. Transmission of impulses through cranial motor nuclei. J Neurophysiol.
1939;2:40264 [125]. Copyright (1939), with permission from the American Physiological Society. bReprinted from Hilgard ER, Marquis DG. Conditioning and
learning. New York and London: Appleton-Century Company; 1940 [91]. Copyright (1940), with permission from Appleton-Century-Crofts. cReprinted from
Morgan CT. Physiological Psychology. New York and London: McGraw-Hill Book Company Ltd; 1943 [139]. Copyright (1943), with permission from McGraw
Hill. dReprinted from Hebb DO. The organization of behavior; a neuropsychological theory. NY: Wiley; 1949. [reprinted 2002 by Lawrence Erlbaum Associates,
Mahwah, New Jersey] [77] Copyright (2002), with permission from Lawrence Erlbaum Associates. eReprinted from Hebb DO. A textbook of Psychology.
Philadelphia and London: W. B. Saunders Company; 1958 [80]. Copyright (1958), with permission from Mary Ellen Hebb
Brown Molecular Brain (2020) 13:55 Page 9 of 28
are simultaneously excited, the resulting ionization is as-
sumed to direct the growth of axons toward the cathode re-
gion, and of dendrites toward the anode, such that a new
synaptic connection is established.HilgardandMarquis
([91], pages 329331) also noted that the work of Lorente
de Nó [124]demonstratedthatspatialsummationofat
least two simultaneous impulses is necessary to excite a
neuron and, in their section on Continued activity in
closed neural chains(page 330), they state that:
Recent observations indicate that the path of neural im-
pulses from sensory to motor nerves is not a single
straight throughpathway [124]. Histological and experi-
mental data indicate that the internuncial neurons are ar-
ranged in two types of chains: multiple chains, in which
several collaterals of a single fiber, after traversing one or
more synapses, converge upon a motor neuron, and closed
chains, in which a collateral excites a circle composed of
several neurons. In the latter case, the chain of neurons
may maintain its activity indefinitely in the absence of per-
ipheral afferent impulses. This arrangement suggests an-
other possible mechanism of learning which would not
necessarily involve any permanent alteration in the
physico-chemical properties of the neurons. Closed chains,
set into activity by the training procedure and continuing
in the absence of any external excitation, would summate
with otherwise inadequate afferent impulses to produce
the conditioned response.(See Fig. 4b).
Hebb cited the work of Lorente de Nó ([123126]) and all
of the papers in the 1939 special issue of the Journal of
Neurophysiology (volume 2, number 5) on the synapse.
Hebbs indebtedness to Lorente de Nó has been discussed by
Martinez and Gil [128]andbyHaider[50], who points out
that Lorente de Nó, Hilgard and Marquis and others whose
work influenced Hebb were all at Yale University in the
1940s. In addition, Robert Yerkes who founded the Primate
Research Centre in Florida did so under the auspices of Yale
University [38]. Only after Lashley became the director did
the Yerkes Primate Centre become a Harvard University fa-
cility, at which Hebb worked as a Harvard employee.
von Senden
In his book Essay on Mind, Hebb ([84], pages 8991) de-
scribed how the work of von Senden [186] and Riesen
[159] on visual development influenced his thinking dur-
ing the period between 1944 and 1945 in which he was
developing his theory:
Nine years earlier in fact I had reared rats in darkness
myself and tested their vision, as the basis of a Ph.D.
thesis. When I recalled that work I recalled too a feel-
ing that there was something peculiar, something re-
markable, about a book I had read at the time. The
book was von Sendens[186] valuable compilation of
all the published case reports of people who were born
blind, with cataract, and who were later made able to
see by removing the cataract. I looked at von Senden
again and was astonished at what I found. His subjects
were in effect blind at first. They could distinguish
and respond to colors but had practically no percep-
tion of shape or pattern. A prolonged learning process
was needed after operation before the patient began to
see the world in the way that a normal person does
(page 90).
Hebb ([84], page 90; See also page 113 of Hebb [77])
discusses the errors in his own PhD thesis:
My failure to see the apparent meaning of von Sen-
dens evidence is a perfect example of how theory,
firmly entrenched, can block ones vision. My thesis
research was done under the influence of holistic and
more or less nativistic ideas--with the result that I did
not see the implication even of my own data. My ex-
periments had shown no difference in the perception
of brightness and size by normal and dark-reared rats.
I went on to look at pattern perception and found
that the perception, once established, is about the
same in normal and dark-reared--patterns that look
alike to one look alike to the other but failed totally
to reflect on the fact that establishing the perception
took six times as long in the dark reared. Only when I
came back to my own published data, nine years later
and with light from another theoretical idea, did I see
what I had done.(page 90).
At about the same time, Austin Riesen [159]testedthe
vision of two chimps reared in darkness and found abso-
lutely no evidence of visual discrimination ability. Until
the research of von Senden and Riesen, many had thought
that there was an innate visual organization of the brain as
Hebb [62,63]sawin his PhD thesis. As Riesen ([159],
page 108) states, The prompt visual learning so charac-
teristic of the normal adult primate is thus not an innate
capacity, independent of visual experience, but requires a
long apprenticeship in the use of the eyes.Given the evi-
dence of von Senden and Riesens research, Hebb ([77],
Chapter 6) proposed that the young chimps reared in the
dark had not formed any visual cell-assemblies ([84]page
91). In his work with fear in monkeys, Hebb [73]proposed
that fear was also a disruption of cell assemblies. Von Sen-
dens book was not translated into English until 1960 and
in this book, Riesen [160] wrote a short commentary dis-
cussing the influence of von Senden on Hebbs ideas.
Köhler and Gestalt Theory
Although Hebb [77] did not specifically acknowledge his
debt to Wolfgang Köhler and Gestalt Theory in the Pref-
ace to The Organization of Behavior, Hebb had taken a
Brown Molecular Brain (2020) 13:55 Page 10 of 28
course from Köhler in Chicago and was well versed in Ge-
stalt Psychology. One might argue that Hebbs thoughts
on the failings of the Gestalt theory of perception stimu-
lated the development of his cell assembly theory (see Sec-
tion The first draft of The Organization of Behavior
below). In his Scientific Method in Psychology, Hebb
([60], page 3637) is critical of Köhler and in his Precis
([71], page 6 and page 20), Hebb criticises the approach
of the Gestalt psychologists to perceptual learning. He said
In general, the Gestalt treatment of the figure-ground re-
lationship has excluded the case in which experience is of
primary importance, although not denying its existence
([71] page 6). Following the introductory chapter, Hebb
([77], page 17) begins The Organization of Behavior with
a revision of perceptual theoryand points out two faults
with Gestalt theory: (1) that perceptions depend on a pat-
tern of excitation whose locus is unimportant; and (2)
that when one perceives a simple figure one perceives it
directly as a distinctive whole, without need of any learn-
ing process and not through prior recognition of the sev-
eral parts of the figure. The focus of the cell assembly
and phase sequence theory in the first chapters of The
Organization of Behavior is on perception, and it is inter-
esting that Hebb ([77], page 17) says in a footnote to
Chapter 2, that This and the following chapter may be
disregarded by the reader who is not particularly inter-
ested in the theory of perception. However, the chapters
that define the cell assembly and phase sequence (Chap-
ters 4 and 5) are entitled the first stage of perception
and perception of a complexand focus on perceptual
learning. It is not until Chapter 6 that Hebb begins to talk
about learning in general.
Although Hebb did not refer the reader to the works of
Köhler [104] or Koffka [103] in the Preface to The
Organization of Behavior, these were important for his PhD
research [62,63]. In his PhD thesis he stated that The Ge-
stalt school ... implies the innateness of certain organizations
of the visual field in stressing the dominance of primitive
Gestalten over learned ones. The kind of evidence cited here
includes the persistence of visual illusions in spite of experi-
ence, and the failure of experimental subjects to pick out fig-
ures which they have been trained to recognize, when these
are part of a more dominant figure([61], page 2). Once
Hebb understood the data of von Senden and reinterpreted
thedatafromhisPhDthesis,heunderstood the flaws in the
Gestalt theory and this stimulated his revision of perceptual
theory, which was to describe (1) the neural pattern of per-
ceptions and (2) how these were learned by experience.
Karl Lashley
Much has been written about the relationship between
Lashley and Hebb [24,144,146]and,althoughitcouldbe
argued that everything Hebb learned about the brain, he
learned from Lashley, he did not always agree with Lashley.
It was Lashley who proposed Hebbs PhD dissertation topic,
based on his tests of the innate organization of vision [120]
and it was Lashley who stimulated his experiments on the
effects of cortical lesions on spatial learning [64,65]. In-
deed, the majority of Hebbs published papers have refer-
ences to Lashleys work. In chapter 3 of The Organization
of Behavior, Hebb says that Köhler and Lashley are the only
writers who recognize the real problem of the neural
mechanisms of perceptual integration and attempt an ad-
equate solution(page 38), but he was not happy with their
solutions and suggested that thelineofthoughtthatthey
have chosen may be a blind alley.Hebbs theory was a re-
sponse to the theories of Köhler and Lashley, but the devel-
opment of this new theory depended on his being in
Lashleys lab environment in Chicago, at Harvard and in
Florida. The development of Hebbs theory also depended
on other students in Lashleys lab, especially Frank Beach,
whose paper on the neural basis of sexual excitement pro-
vided Hebb with the concept of the central excitatory
mechanism [11] for his study of fear in chimpanzees [73].
This paper introduced the concept of the phase sequence
and the dual process based on the separation of external
stimuli and autonomous central processes, which were dis-
cussed in detail by Beach [11]. As noted below, Hebb felt so
strongly about his debt to Lashley that he asked him to be a
co-author on The Organization of Behavior.
Hebbs notes for The Organization of Behavior
Although he does not mention his unpublished notes in ei-
ther of his autobiographical papers [81,83], these show how
he developed the ideas for The Organization of Behavior.
Hebb [69] first outlined his ideas in five pages of typed
notes, which contain his early thoughts on the term
organization. In this definition, he considered both behav-
ioral and physiological organization and considered the con-
cept of set, which he used to denote the pre-existing state
of the nervous system encountered by an incoming stimu-
lus. A key concept in these notes is the LNC, which was the
LorentedeNócircuit.Hebb([69],page1)takesthisashis
central concept in the organization of behavior:
What I mean by organization behaviorally: ready
identification or discrimination, ease of recall and ac-
curacy of recall or identification, ease of association;
recognition in any circumstances. And by physiologic-
ally: assumption of need of number of parallel LNCs
for stable response or perception, possibly of good
organization by temporary facilitations, possibly of
good organization by innate physiological and anatom-
ical factors, possibly of organization by structural
modifications.
In the spring of 1945 Hebb [70] made four pages of
hand-written notes in which he worked out his theory of
Brown Molecular Brain (2020) 13:55 Page 11 of 28
the ACA = Autonomous Central Action (also ACP = Au-
tonomous Central Process). Hebb ([70], page 1) stated:
That ACA must be considered the direct determinant of
behavior; that ACA events are a joint function of afferent
activity and preceding ACA activity (influence of ACA on
perceptions, also, was however seen only after system began
to be developed).AthirdsetofnotesforThe Organization
of Behavior was typed between March and July 1945. In
these notes, Hebb ([71], page 1) states that:
the attempt is made to show that a schema of the
physiological control of behavior can be developed by
utilizing these preliminary ideas. The schema does
not make specific predictions, and thus suggests only
a reformulation of the problem of the control of be-
havior, instead of providing an explanation. There is
an approach to explanation, however, in the fact that
a variety of psychological problems may possibly be
unified, and how it is possible for example to conceive
of a mechanism of the effect of attention on learning,
in physiological terms; and suggests that if the hurdle
of perceptual generalization can be got over, the
problem of motor generalization and its perceptual
control becomes physiologically explicable (in general
terms, of course, and for simple cases).
These notes include a rough draft of Fig. 3on page 52
of The Organization of Behavior. The second section of
these notes entitled Aschemaofperception[72]con-
tains a draft of Figure 11 in The Organization of Behavior.
These four sets of notes [6972] provide an overview of
Hebbs theories that were presented in the first two chap-
ters of The Organization of Behavior. Many of Hebbs
notes refer to KSL (Lashley) and in June 1944, he noted
that he was prepared to accept - even welcome-innate
organization such as Dr. Lashley has suggested,butby
the Precis of 1945 [71], he has re-discovered von Senden
and wrote that Sendens[186] evidence is that both the
normal associability and the normal generalization of sim-
ple figures such as circle, triangle and square are built up
by experience([71], page 9).
In his paper on the nature of fearHebb [73] outlined
some of the ideas that he was introducing into his book in
progress. These include discussions of central versus sen-
sory factors in determining fear and the inference of a
central excitatory mechanism or central motive state
([73], page 267), which had been presented by Beach [11].
He described temporal patterns of cellular activity as a
phaseand stated that Behavior is directly correlated with
a phase sequence which is temporally organizedand
emotional activity disrupts the phase sequence ([73], page
269). He said that Subjectively, the phase sequence would
be identified with the train of thought and perception.In-
deed, the concept of the phase sequence is spelled out very
clearly in this paper, which is the immediate precursor to
The Organization of Behavior. However, the concept of
the cell assembly had not been developed and Hebb ([73],
page 269) used the term phasefor aspecificpatternof
cellular activity.
The first draft of The Organization of Behavior
A draft of The Organization of Behavior was finished in
1946 and given to Hebbs colleagues at the Yerkes Pri-
mate Centre to read. As noted on the typed manuscript
[74] these included Robert and Josephine Blum, George
Clark, Henry Nissen, Roger Sperry, Austin Riessen, Karl
Lashley, and Karl Pribram. Dewsbury ([38], pages 174
176) gives the details of the staff members of the Yerkes
labs during the years that Lashley was the director. Hebb
also sent copies to Frank Beach, Edwin Boring at Har-
vard, Professor R. B MacLeod at McGill, Dr. J. C. R.
Licklider, Dr. J. G. Beebe-Center and Dr. G. A. Miller, as
well as others. However this draft contained only chap-
ters 1 to 7 of the published version of The Organization
of Behavior. Chapters 8 to 11 were added after Hebb
returned to McGill University in the autumn of 1947.
The draft of The Organization of Behavior that was
submitted in 1946 was different in a number of ways
from that which was finally published in 1949. The pub-
lished book has a preface and an introduction, while the
1946 manuscript begins with chapter 1. Sub-titles of
chapter sections also differ. For example, the section
Rejecting the assumption of a complete sensory con-
trol([77], page 3) was Autonomous central process
versus sensory dominancein the 1946 draft. The last
section of Chapter 1, The mode of attackhas been se-
verely trimmed in the book. In the original manuscript,
Hebb basically says that Lashley was wrong, Köhler was
wrong, Pavlov and Hull were wrong and Skinner was
wrong, and he will correct them. It is worth citing these
sections in full to get the flavour of Hebbs original mis-
sion as they were deleted when the book was published.
If one assumes that Köhler and Lashley have provided
the most nearly satisfactory account of behavior, the
next step obviously is to find out what is not satisfactory
about their theories, and what changes are needed.
The changes lead directly toward the position
taken by Hull and others who have emphasized
slow-increment learning as the fundamental psycho-
logical phenomenon.
But Hulls formal system must also be rejected, just
as Köhlers must. Hulls is the most consistent theory,
internally, of those current today; but its consistence
seems possible only by neglecting certain aspects of
behavior (putting them aside for treatment later,
Brown Molecular Brain (2020) 13:55 Page 12 of 28
presumably); and it makes no real attempt to define
its variables physiologically. I propose to show, at the
end of this monograph, that psychological theory is
essentially concerned with physiological facts and
concepts Skinner [173] to the contrary notwithstand-
ing. Assuming that, and assuming that perceptual
generalization must be accounted for, the work of the
Gestalt school and Lashley is the only solid starting
point for theory.
Now let us see, in the next two chapters, what it is
in Köhlers and Lashleys treatment of perception
that needs revision.([74], page 20).
Another section which was deleted concerns Hebbs
critique of Pavlov. He says:
Here the concepts of irradiation and concentration,
as an explanation of generalization, break down.
When one point on the skin is conditioned, so that
its stimulation is followed by secretion of saliva in
Pavlovs method, neighboring points will also produce
some secretion. This Pavlov accounts for by the no-
tion of irradiationor excitation over the cortex. But
with differences of intensity, instead of locus, the mat-
ter is different: now the two stimuli are projected to
the same cortical point. Irradiation cannot account
for the discrimination of relative intensities as it per-
haps might for discrimination of relative place. Pavlov
[151] failed to answer this criticism and the subse-
quent attempts to modify his ideas to cover it by a
verbally postulated generalization gradient([93,
175]) have provided no intelligible conception of a
physiological mechanism.([74], page 44a.).
The lattice and the schema became the cell assembly and
the phase sequence
In Hebbs 1945 Precis [71] and in the original 1946 type-
script of The Organization of Behavior [74],Hebbdidnot
use the terms cell assemblyor phase sequence.Inthe
typescript ([74], page 69), the cell assembly was called a
lattice, and the phrase the specificity of such a lattice
was changed to the specificity of such an assembly of
cellswhen the book was published ([77], page 72). Figure
5shows the original drawing of the lattice. In the original
typescript there was no phase sequence.Theterm
schemawas used and Chapter 5 (Perception of a com-
plex: the phase sequence) was originally entitled Schema:
the perception of a complex.Thetermphase sequence
was introduced in Hebb's paper on the nature of fear [73].
While the term latticeremains in the published book
in only two places, (pages 72 and 75), the term schema
is used quite often. It appears that the term schema
originally referred to both the cell assembly and the phase
sequence. The idea of a schemawas first used by Head
and Holmes [54] to refer to a cortical representation of
body posture that was built up over time. Schematawere
defined as organized models of ourselvesproduced by
incoming sensory impulses; destruction of such schemata
by a cortical lesion renders impossible all recognition of
posture or of the locality of a stimulated spot in the af-
fected part of the body([54], page 189). Bartlett ([9],
chapter 10) used Heads concept of schemata to develop
his theory of remembering, but other than mentioning
nerves, this was a cognitive, not a neurophysiological the-
ory. He stated (page 214) that his theory merely jumbles
together innumerable traces and calls them schemata
and then it picks out a few and calls them images.Itwas
Hebbs job to turn these schema into a neurophysiological
concept, the phase sequence, but this was a three stage
process: synaptic change leading to a cell assembly and a
series of cell assemblies combining to form a phase se-
quence or schema. In recent studies of memory schema
Hebb is seldom mentioned [47] but his concept of the cell
assembly links Bartletts concept of schemata and Tol-
mans concept of cognitive maps with the later theories of
neural networks and place cell assemblies[7,110].
Why did Hebb change the names of the lattice and
schema?
In a letter to Henry Nissen (12 September 1948) Hebb
wrote that I am now about ready to send off the MS
[manuscript], with one major headache. Licklider at Har-
vard, who read the first half last summer, has decided
strongly that latticehas too many wrong implications. I
agreed too and must now choose another term - have you
any ideas? A structural analogy seems impossible, no good
term available, so I think now of some Greek word meaning
a working together, or calling together, or something of the
kind - if you have any ideas, shell out!Will consult Classics
Dept., try to find something innocuous and euphonious.
On 14 September 1948, Nissen wrote back that As to
lattice, my impression in brief is that (1) it has some
wrong implications, (2) any other term you choose will
have wrong implications, (3) you will find no term that is
less objectionable. Look at it this way: a lattice is an
organization, an organization inferred or postulated to exist
in neural structures. So, since you incline towards the
Greek, neuromorphonor neuro-organon.IfIhaveany
better brainstorms, Ill let you know.On 22 September
1948, Hebb wrote to Nissen that Ihavechangedtheterm
latticeto cell-assembly. This at first glance sounds quite
peculiar; but I think you may feel, after you have seen the
way it works out in the actual context of the discussion,
thatitmaybeaveryhappysolutionwithfewerdrawbacks
than most of the others. Today we should applaud Hebbs
Brown Molecular Brain (2020) 13:55 Page 13 of 28
choice of the cell assembly; it could have been a neuromor-
phon or neuro-organon.
In this context, it is interesting that Lashley [117]dis-
cussed the organization of the nervous system in terms of
schema and lattices. Lashley had read Hebbs manuscript
in 1946 and commented upon it in Feb 1947, but this
paper does not mention that manuscript, nor does it men-
tion Bartlett [9]. The influence of Hebb might be felt in
the statements that anatomic studies suggest that a net-
work or lattice of nerve cells forms the basis of the central
integrative processes(page 36) and the sentence that: I
shall designate these patterns of interacting cells as neural
schemata(page 36). However, unlike Hebb, Lashley in-
cluded the concept of neural inhibition in his diagrams.
Commentaries on the first draft
Many people wrote comments on the early draft of the
manuscript, including Lashley, and Boring.
Boring
Boring [16] appears to have read the first 104 pages (4
chapters) and wrote to Hebb that he should be briefer,
less defensive, more positive, and more kindly, gay and
friendly. Boring sent 6 typed pages of comments to Hebb
with a page-by-page critique. It appears that this critique
was based on an earlier version of the book than the
Hebb 1946 typescript that I have [74]. However, one
comment on Chapter 3 is worth noting. In his com-
ments, Boring says This is a good chapter and import-
ant. These things need to be said, though it is plain that
their chief importance is as an introduction to some-
thing else that I have not come to. I think you [r] criti-
cism of K and L [Köhler and Lashley] is valid and proper
and courteous. You seem to be harder on K than on L,
but then I think K is much wronger than L. So far there
is no reason at all to hesitate about publishing these
comments on Lashley. They do not sound disloyal. They
Fig. 5 A comparison of (a) the original drawing of a cell assembly from Hebbs[74] typescript in which the cell assembly was first called a cage, and
then changed to a lattice and (b) the published figure of a cell assembly from The Organization of Behavior [77], page 73. aReprinted from an
unpublished manuscript by Hebb, D. O. 1946, entitled "Carbon of most of the original MS of my book The Organization of Behavior (while the term
"lattice" was still used instead of "cell assembly" [74]. The original is held in McGill University Archives, Montreal, Quebec, file MG1045. Reprinted with
permission from Mary Ellen Hebb. bReprinted from Hebb DO. The organization of behavior; a neuropsychological theory. NY: Wiley; 1949. [reprinted
2002 by Lawrence Erlbaum Associates, Mahwah, New Jersey]. Copyright (2002), with permission from Lawrence Erlbaum Associates
Brown Molecular Brain (2020) 13:55 Page 14 of 28
seem rather to be supporting him toward what he has al-
ways been after, even though disagreeing in particulars.
In this way Stevens and I keep supporting each other
and referring to each other, while dissenting from each
in minor details.(Boring to Hebb [16], page 5).
Lashley
In February 1947, Lashley sent Hebb two pages of com-
ments on the first 91 pages of the manuscript. A few of
Lashleys comments might be noted. P.13. Statements
in 2 following Pp [paragraphs] are questionable without
amplification. Neither spontaneous activity nor rev. [re-
verberatory] circuits in themselves provide solution of
the problems of the synaptic theory of learning.”“P. 53.
configuration theory does not account for anything - it
sets problems. Köhler.Hebbs notes on these two pages
say These 2 sheets are Lashleys commentary (he didnt
approve) plus all entries in red on the MS itself.
Hebb asks Lashley to be a co-author
As he wrote his book, Hebb realized how indebted he was
to Lashley. In his autobiographical paper, Hebb said that
I went to Lashley and proposed that we work it out [the
concept of a schema] and publish together, since this
would have to be a book-length job, which I was not ready
to face by myself. Also I hoped that he could devise a bet-
ter treatment of perception (I thought this part of the
schema was wrong, but could find no alternative). Be-
tween us, a much better job was possible, in both theory
and presentation; and with him as joint author, the book
would get a hearing that I did not expect I alone would
get. But Lashley was entirely uninterested and remained
skeptical.([83] page 296). In an unpublished autobio-
graphical paper Hebb said When it came to working out
my theory in 1945 I could see that it was a book-length
job which I didntfeelabletotackle.IwenttoLashleyto
ask him to be a joint author. He was quite uninterested.
Together we might have done it in two years. Me it took
five, and I finished it at McGill after leaving Florida([85],
page 13). In a letter to Henry Nissen (Hebb to Nissen, 19
May 1948) Hebb says that Lashley told me you know that
the whole thing was very weak, with no value because it
was so vague, which I think had a great deal to do with
my patting myself on the back and being so aggressive in
discussing other work.
Completing the Organization of Behavior at
McGill (19471949)
In December 1946, Edwin G. Boring, the Chairman of
the Psychology Department at Harvard, invited Hebb to
teach a graduate seminar during the summer session of
1947 and Hebb used a draft of his book in this seminar.
One of the students in this class was Mark Rosenzweig,
who wrote that:
I took a graduate seminar with Donald O. Hebb at
Harvard in the summer of 1947 where the text was
a mimeographed version of Hebbs influential book
The organization of behavior which appeared in
print in 1949. Hebbs creative suggestions revitalized
theorizing and research on learning and memory,
and I benefitted directly from them and from fur-
ther contacts with him[162].
When the summer at Harvard was over, Hebb moved
back to Montreal as Professor of Psychology at McGill
University, where he developed a graduate program in
physiological psychology and completed the final chap-
ters of his book. In the Preface to the Organization of
Behavior, Hebb wrote:It is a pleasure to record my in-
debtedness to the colleagues who have read and im-
proved the contents of this book. I owe much to
students in a seminar at Harvard University in the sum-
mer of 1947, and in another at McGill University in the
following winter. Peter Milner, Brenda Milner and
others had read mimeographed copies of the book at
McGill. Brenda Milner ([132], pages 282283) said that:
during [Hebbs] first seminar, we discussed this
book chapter by chapter and did the relevant back-
ground reading, which covered Lorente de Nó, Mar-
shall and Talbot, Hilgard, Lashley, and Sperry. The
graduate students in this seminar included Morti-
mer Mishkin, Lila Ghent (Braine), Herb Lansdell,
and Woodburn Heron, and discussion after the
seminars often continued late into the night. It was
an exciting time and hastened my decision to do a
PhD at McGill. In 1949, I persuaded him to accept
me as a graduate student. I wrote to Peter [Milner]
enthusiastically about the Hebb seminar, with the
result that he decided on a career change.
Peter Milner ([138], page 36) said that When I read
the chapter on the cell assembly in a manuscript of The
Organization of Behavior, I thought Hebb might be on
the way to an answer. If I studied with him, I might even
help him find it.Peter Milner [133] published a revision
of the cell assembly, and other papers culminating in his
book The Autonomous Brain [135].
Henry Nissen edits the Organization of Behavior
In June 1947, Hebbs book was submitted to Charles
Thomas publishers and it was sent to Henry Nissen at the
Yerkes laboratories in Orange Park, Florida to review. It
was 150 pages long and insured for $300.00 (Nissen to
Hebb, letter of 21 June 1947). Nissen edited this manuscript
from June 1947 to September 1948, sending Hebb numer-
ous comments and suggestions for changes (Nissen to
Hebb 11 May 1948). Nissen sent pages of typed notes on
Brown Molecular Brain (2020) 13:55 Page 15 of 28
each chapter as well as comments in pencil in the mar-
gins. His over-all comments were that I agree whole-
heartedly with your plan to tone down the criticismand to
refrain from patting yourself on the back quite so often. A
number of my criticisms would be met by those two
changes. A less argumentative tonewould be an
improvement.
Nissen turns down co-authorship
On 19 May 1948, Hebb wrote to Nissen and asked is
there any way in which I could induce you to share
authorship with me, change the matter and style where
you think it should be changed, developing your own
ideas concerning perception and learning and motiv-
ation, perhaps writing a chapter on the relation of ani-
mal work to human problems incorporating the ideas
that you probably had to skimp on in writing your hand-
book chapter?(Hebb to Nissen, 19 May 1948).
Nissen replied that Iamquiteoverwhelmedbythesug-
gestion that I share authorship of the book with you. A
more complimentary thing has never happened to me. A
quick analysis indicated several factors which may have con-
tributed in bringing forth this suggestion: (1) An over-
reaction to KSLscriticisms-anover-reactionregardless
of how severe those criticisms may have been. (2) A failure
on your part to realize how good the completed parts of the
book are right now. Perfect? No. But it will never be perfect
no matter how many collaborators you ring in on the job.
[Nissen to Hebb, 23 May 1948, concluded 29 May].
In his letter to Henry Nissen on 19 May 1948, in
which he invited Nissen to be a co-author, Hebb out-
lined his completion of the book. He said that What Id
like to do, therefore, is get your help re-writing, your ad-
ditions and modifications of the theory. I would then
write two chapters, one on emotion [chapters 8 and 9]
and clinical deviations of emotion [chapter 10] as far as
they touch on theory, and one on human intelligence
[Chapter 11]. In Nissens letter to Hebb [23 May 1948,
concluded on 29 May] Nissen gave Hebb a number of
bits of advice about finishing the book within the next
8 to 12 monthsand suggested that something could be
added (preferably near the beginning) which would
summarize the inadequacies of conditioning theory and
the essence of your form of association theory by which
those inadequacies are repaired.
In his reply to Nissen [1 June 1948] Hebb said that I
have now been through your criticisms in detail, and
mean to accept practically all of the changes you sug-
gest.Nissen sent Hebb his critique of Chapter 8 on 13
July 1948 and Hebb sent chapter 9 to Nissen on 19 June
1948. Nissen sent Hebb his comments on the revised
Introduction and on Chapter 9 on 6 August 1948, his
comments on Chapter 10 on 8 September 1948 and
those on Chapter 11 on 15 September 1948.
Rejection and acceptance
On 22 September, 1948, Hebb sent the final draft of the
manuscript to the publishers, Charles C. Thomas. The ori-
ginal title of the book was On Thought and Behavior. How-
ever, in January 1949, Thomas returned the manuscript,
saying that they could not publish it for another year or
more [letter of Hebb to Nissen, 17 Feb 1949). In a letter to
Frank Beach [6 February 1949] Hebb says Thomas has just
sent back my MS, regretting he is unable to publish it on
account of circumstances beyond his control - to wit,
strikes and slow-downs by printers, piling up work for the
past two and a half years.Both Frank Beach and Henry
Nissen wrote to the publishers John Wiley and Sons and
Appleton-Century-Crofts to see if they would publish the
book, and on 18 March 1949, James Helming of John Wiley
and Sons wrote to Hebb agreeing to publish his book with
only perfunctoryediting. In the letter, he noted that his
editor-in-chief had said This is by far the best-written
manuscript that has come my way in some time. The ex-
position is lucid, persuasive, and also lively - the author
need have no misgivings about the propriety of his humer-
ous touches; they are distinctly refreshing. Unlike most fac-
tual manuscripts this one has a definite appeal on literary
merits alone.And so it was published in September 1949
by Wiley under the title The Organization of Behavior.
When published, The Organization of Behavior was
reviewed by Kuhn [107], who said that this book will
probably come to be regarded as a landmark in psycho-
logical theory.Attneave [6]saidthatIbelieveThe
Organization of Behavior to be the most important contri-
bution to psychological theory in recent years.Leeper
[121]statedthatThere are so many respects in which
Hebbs book is so high in quality and is so delightfully
written that it will have an assured status in psychology.
Hebbs book lived up to the reviewers predictions and be-
came one of the most important contributions to psych-
ology in the twentieth century. During the 1950s, Hebbs
ideas found a place in many texts. Allport [3]devotedan
entire chapter (entitled The association approach, cell as-
sembly and phase sequence) to a discussion of Hebbs
ideas on perception. Hilgard [90] added a new section on
Hebbs neuropsychological model to the revised edition of
his book and in 1959, Kochs monumental seven volume
survey of psychology as a science had a chapter by Hebb
[81] and references to Hebbs book in every volume [102].
Hebbs evaluation of the Organization of Behavior
10 years later
In discussing his book ten years later, Hebb ([81], p 638)
said that his idea was to deal with set and attention and
perceptual generalization and learning in one theoretical
framework, not have one approach for thinking, another for
learning, and a third for perception -- the position in which
the members of the Gestalt group found themselves.
Brown Molecular Brain (2020) 13:55 Page 16 of 28
He went on to say that:
My theory is the only one that attempts this, and
in my opinion, to be quite frank, is consequently the
only realistic attempt to deal theoretically with the
problems of behavior. Skinner of course has avoided
theory; Tolman and Guthrie have proposed ap-
proaches to the problem of constructing a theory,
but both have remained, essentially, programmatic.
Hulls is the only real alternative to mine; and the
course of development of his ideas, from 1937 to
1951, has shown a narrowing of the range of phe-
nomena dealt with, an increasingly clear set of diffi-
culties to be encountered even in the narrow range
with which his theory does deal, and an increasing
concern with minor modifications of postulates as
defensive measures to meet the attacks of critics. .....
Mine, in short, is the only attempt to deal with the
thought process and perception in the framework of
a theory of learning. It has serious defects, but no
real competitor. This fact I see as the major evi-
dence for the system, together with the body of re-
search that it has, directly or indirectly, stimulated.
([81] page 638639).
Hebbs theories generated a great deal of research by
his students at McGill University. However, Hebb said
that he discouraged experiments designed as a test of
my theory in a narrow sense, feeling that this would
limit the students research too much.([81], page 637).
But he did describe how the research work in his labora-
tory was related to the idea of the cell assembly. The
connection of my theory to some of the work is fairly
clear: the studies of visual perception of Mishkin and
Forgays, Orbach, Heron, and Hunton; the effects of per-
ceptual isolation by Bexton, Heron, Scott and Doane; the
role of the infant environment in mental development
by Hymovitch, Forgays and Forgays, Clarke et al,
Thompson and Heron, Melzack and Mahut; and the
reexaminatioin of the mass-action and equipotentiality
conceptions by Lansdale and Smith.([81], page 637).
Hebb ([81], page 639) also criticized his theory for its
vagueness and the difficulties in conceiving of how the cell
assembly could be subjected to experimental analysis. He
pointed out that new physiological data in the decade
since his book was published would have greatly affected
the formulations of the theory. In his later papers, Hebb
extended his theory to account for some of these new de-
velopments. His Textbook of Psychology [80] was, in many
ways, a continuation of The Organization of Behavior and
in his Presidential Address to Division 3 of American Psy-
chological Association Hebb [79] examined the relation-
ship of the arousal theories of Moruzzi and Magoun to his
theory of motivation and described the relationship
between arousal level and optimal performance as a the-
ory for understanding motivation. Still later, Hebb ([82],
page 314) suggested that inhibitory circuits were import-
ant for the functioning of the cell assembly because inhibi-
tory circuits serve the purpose of promptly shutting off a
cell assembly once it has performed its function of firing,
or taking part in firing, another.
Hebb ([83], page 301) noted that he was surprised by
the success of his book. He said that: Iexpectedthat
years of research with graduate students would be needed
to get a hearing for the book once it did appear, and its
immediate success was quite astonishing.He pointed out
that the arousal system was discovered by Moruzzi and
Magoun in the year that The Organization of Behavior
was published and a few years later Eccles established the
existence of inhibition which I had been afraid to
postulate. One might consider Hebb's 1980 book Essay
on Mind [84], as a review of his thoughts on the cell
assembly theory and its importance in psychology and
neuroscience.
Eight issues arising from the Organization of
Behavior
The three main ideas in The Organization of Behavior
were (1) the Hebb synapse, (2) the cell assembly and (3)
the phase sequence. An important concept underlying
these ideas was (4) the autonomous central process. Two
important omissions were (5) the concept of inhibition
and (6) any reference to chemical neurotransmitters. Fi-
nally, one of the reviewers of this paper asked about (7)
Hebbs ideas on the relationship between instinct and
learning and (8) the relationship between the cell assembly
and the engram. Each of these issues is discussed here.
The Hebb synapse
Much has been written about the Hebb synapse [32,
105,109,137,144,166,168,174,176]. The idea of the
Hebb synapsewas not new to Hebb but had been
thought of by many others (see [137]). As noted by Hebb
([77] page 70), The general idea is an old one, that any
two cells or systems of cells that are repeatedly active at
the same time will tend to become associatedso that
activity in one facilitates activity in the other.Indeed, in
his 1903 speech in Madrid, Pavlov [150] speculated on
the neural mechanism underlying the conditioned reflex
and presented a theory that sounds very similar to the
familiar Hebb synapse. He stated that when the
psychicalstimuli associated with the food:
become connected with the same nervous centre of
the salivary glands to which the stimulation emanat-
ing from the essential properties of the object is
conducted through a fixed centripetal path. It can
be assumed in this case that the salivary centre acts
Brown Molecular Brain (2020) 13:55 Page 17 of 28
in the central nervous system as a point of attrac-
tion for stimuli coming from other sensory surfaces.
Thus a certain path is opened from the other ex-
cited areas of the body to the salivary centre. But
this connection of the centre with accidental points
is very fragile and tends to disappear of itself. Con-
stant repetition of simultaneous stimulation by
means of the essential and unessential properties of
the object is required to make this connection in-
creasingly durable.(page 163).
Hilgard and Marquis ([91], pages 326335) discussed
the theories of synaptic function then current, as did
Morgan ([139], pages 520525). These theories included
Neurobiotaxis, the formation of new synapses; Synaptic
resistance, the reduction of synaptic resistance during
learning; Fiber conductance, the theory that repeated
passage of an impulse along an axon increased its con-
ductivity; Reverberation, the activation of reverberatory
nerve circuits; and Resonance, during learning, neurons
that are out of tune, become more in tune with each
other. Berlucchi and Buchtel [12] summarize a number
of other theories of synaptic change underlying learning
and they reprinted a letter from Hebb, written on 25
August 1977, in which he stated: In my view, in 1946 or
so when I was drafting O of B [Organization of Behav-
ior], I wasnt proposing anything new. All I considered I
was doing was making a more operational statement of
a widely held idea that synaptic resistancewas reduced
whenever an impulse crossed the synapse.Hebb went
on to say that: The bottom two paragraphs of page 60
(O of B) specifically make the point that it was an old
idea, repeatedly rejected by critics of learning theory, but
now revived in a new physiological context (i.e., the idea
was not new).
One of the main critics of learning theory at this time
was Lashley, who thoroughly rejected the idea that synaptic
change underlied learning. This was clear in the first sen-
tence of his 1924 paper in which he said that: Among the
many unsubstantiated beliefs concerning the physiology of
the learning process none is more widely prevalent than the
doctrine that the passage of a neural impulse through the
synapse somehow reduces synaptic resistance and leads to
the fixation of a new habit([111] page 369). Lashley re-
peated his critique of synapticresistancetheoryinhisbook
on Brain mechanisms in intelligence ([112], pages 125127)
wherehesaidthatit is not clear that the synapse is either
essential or important for learning. Lashley ([115], pages
488494) reviewed five theories of the neurobiological
mechanisms of learning and concluded that If our analysis
of the problems of behavior and of neural mechanism is
correct, it follows that the current theories of neural
organization have started from false premises and offer no
hope of a solution of the problems.
As pointed out earlier, Hebb had drawn out his ideas
of synaptic changes in learning in his 1932 MA thesis
[58] and in his 1934 essay [59]. In this essay, Hebb sug-
gested synaptic mechanisms for both excitation and in-
hibition (see Fig. 1c). It is little wonder then, that
Lashley harshly critiqued Hebbs 1934 essay [59] on the
neural basis of learning, as it proposed synaptic change
as the neurophysiological mechanism underlying learn-
ing. And again, it is no surprise that Lashley did not
want to be the co-author of a book whose fundamental
premise was that synaptic change was the basis of learn-
ing and perception. Hebb was amused that as a result of
his neurophysiological postulate, synaptic changes in
learning had acquired his name, because this postulate
is one of the few aspects of the theory he did not con-
sider completely original. Something like it had been
proposed by many psychologists before him, including
Freud in his early years as a neurobiologist.([134], page
127). It has been suggested that in addition to anticipat-
ing a Hebb-like synapse, Freud also anticipated the dis-
covery of LTP [31,165]. At about the same time as
Hebb, Konorski [106] published his theory of synaptic
plasticity underlying learning [13].
The cell assembly
The concept of the cell assembly was Hebbsgreatestcon-
tribution to psychological theory according to Milner
[134]. Hebb developed the cell assembly from Lorente de
s reverberatory circuits and credits Hilgard and
Marquis [91] for introducing him to the papers of Lorente
de Nó ([77], page 61, [84], page 84). The cell assembly is
thus the extension of Lorente de Nós[124,125], reverber-
atory circuit and the closed chain of neurons. The cell
assembly is a group of neurons arranged as a set of closed
pathways ([80], page 103) that have become connected
with each other by the process of perceptual learning. The
self-exciting closed loops among the neurons in a cell as-
sembly can maintain their activity for some period of time
after the cessation of an external stimulus. Because cell as-
semblies may firein the absence of external stimuli, they
become the basis for thought. Different depictions of the
Cell Assembly are shown in Fig. 4.
Much has been written about the cell-assembly theory,
starting with the revision by Peter Milner [133] and ex-
panded in a number of papers [134,135,138]. Although
Orbach [146] claimed that the idea of the cell assembly
originated with Lashley, Milner [136] refutes this in no
uncertain terms. There are numerous other reviews of
the cell assembly that indicate its importance in theories
of brain function [4,41,97,122,144,147,187].
The phase sequence
A phase sequence is a temporal sequence or series of cell-
assemblies which represents a chain of ideas, from
Brown Molecular Brain (2020) 13:55 Page 18 of 28
sensation to thought, or a train of thought (see [84], page
92). While cell assemblies formed the basis of perceptual
learning, the phase sequence was the basis of perceptual
integration as envisioned by the Gestalt psychologists. The
phase sequence was also the basis of associative learning
(see [133], page 248). Hebb ([71], page 31) had originally
used the term "Schema (A schema of perception)and
stated that: the schema represents some features of the
process of perception.Hebb([71], page 39) states that
The schema implies that the difference between two per-
ceptions derives from the mode of combination of cells,
not from activity in totally different sets of cells.The
phase sequence was seen as the temporal activation of a
series of cells in the schema. In his Precis, Hebb [71]had
yet to articulate the cell assembly and phase sequences
fully. However, as mentioned above, Hebb [73] outlined
his concept of the phase sequence as a temporally orga-
nized pattern of cellular activity which could be identified
with a train of thought and this paper gave a clear descrip-
tion of the phase sequence.
The development of these three concepts in The
Organization of Behavior can be followed through the de-
velopment of Hebbs theory. Starting in 1932, the Hebb
synapsewas an attempt to explain Pavlovs conditioned
reflexes in terms of Sherringtons physiology of the ner-
vous system [58]. The cell assembly took into consider-
ation Lashleys[113] critiques of Pavlov by adding a
mechanism for set or attention [48]andprovidinga
mechanism derived from Lorente de Nó for neural activity
to persist beyond the termination of a stimulus. The cell
assembly provided the basis for perception and learning
and the phase sequence provided a mechanism for linking
cell assemblies to form the basis of memory, thought and
imagination; for linking unrelated concepts into new ideas.
As Hebb ([77], page 79) pointed out what we are aiming
at here is the solution to a psychological problem. To get
psychological theory out of a difficult impasse, one must
find a way of reconciling three things without recourse to
animism: perceptual generalization, the stability of mem-
ory, and the instabilities of attention.
The autonomous central process
Hebb ([77] pages 510) discussed the importance of
postulating neural functions that are independent of sen-
sory stimulation. He reviews the concepts of the central
process[91], central motive state[139]orcentral ex-
citatory state[125,172] and calls this an autonomous
central process. Hebb ([77], page 8) proposed that the
autonomous central process as defined by physiologists
should be integrated with psychological theories of atti-
tude, expectancy and set as reviewed by Sherrington
[172]. Hebb cited evidence from a number of sources in-
cluding Beachs[11] paper on the neural basis of sexual
behaviour, Weisss[188] paper on autonomous versus
reflexive activity of the CNS and Jaspers[98] paper on
spontaneous electrical activity in the brain as recorded
by EEG, for the knowledge of such autonomous activity
in nerve cells. Hebb had known Beach and Weiss at Chi-
cago and Jasper in Montreal and learned from all of
them. Morgan ([139], page 65) pointed out that the con-
cepts of a central excitatory state and a central inhibitory
state came from Sherrington and were conceived as the
accumulation of ions at the synapse which, respectively,
facilitated or inhibited synaptic transmission ([139], page
65). So it is a mystery why Hebb did not include inhib-
ition as part of his cell assembly theory.
Much of Hebbs thinking about the autonomous central
process(ACP)wasworkedoutinhisnotes[69] in which he
wrote: Idea of autonomous central LNCs (Lorente de Nó
circuits), ACP with sensory component of organization, but
presumably subject to their own physiological and anatom-
ical determinants: normally, presumably, subject to sensory
activation, but also capable of autonomous activity(page
1). As Hebb worked out his theory, the LNC became the cell
assembly and the ability of one LNC to influence another
became the phase sequence. The ability of an LNC to show
autonomous activity provided the basis for attention or set
and for thought and imagination, the linking of two inde-
pendent LNCs to create a new LNC combining both; a new
thought by association of two old circuits. Using these ideas,
Hebb [69] worked out a neural explanation for the Law of
Effect (page 6) and the results are presented in Chapter 6 of
The Organization of Behavior (Development of the learning
capacity) which uses the concept of the cell assembly, the
central autonomous process and the combination of cell as-
semblies into phase sequences to develop a neural explan-
ation for associative learning.
Inhibition
Although Hebbs[58] M.A. thesis and his unpublished
essay [59] stressed inhibition, there was no concept of
inhibition in the cell assembly hypothesis. This has al-
ways seemed strange. In his revision of the cell assembly
theory, Milner ([133], page 243) states that: At the time
when Hebb was developing his theory, many physiolo-
gists were strongly opposed to the idea of neural inhib-
ition, largely because it was difficult to fit into the
electrical theory of synaptic transmission.But Hebb had
already written a great deal about inhibition. Lorente de
Nó [125] described inhibition and provided a diagram of
inhibition in a closed loop. Morgan ([139], pages 6667)
described successive inhibition and reciprocal inhibition,
and provided a diagram of neurons and synapses in-
volved in reciprocal inhibition that is almost identical to
that proposed by Milner [133] for the cell assembly
mark II. These are shown in Fig. 6.
So, how did inhibition disappear from Hebbstheory?In
Hebbs[69], notes he says I am deliberately vague about
Brown Molecular Brain (2020) 13:55 Page 19 of 28
when or where one LNC will (A) facilitate another, and
hence make possible structural (B) facilitation; simultaneous
activity not enough, because of (1) anatomic separations, (2)
apparent fact that one system can inhibit or extinguish an-
other, (3) factors of timing utilized by Lashley in his theory
of repetitive circuits(page 5). However, when Hebb [71]
got into his theory of reverberating circuits in his Precis, the
idea of inhibition seems to have been dropped. Whatever
the reason, Milners[133] revision of the cell assembly the-
ory added the concept of inhibition. In fact, Hebb [77]does
consider the problem of inhibition on pages 208215 (chap-
ter 9). But he does not consider inhibitory synapses, even
though he refers to Sherringtons([172], page 532) central
inhibitory state (c.i.s.). For Hebb ([77], page 214), inhibition
meant fatigue of excitatory pathways. He stated that If it
weresafetoassumeatruelong-term inhibition (or equally,
fatigue) in cerebral action it would greatly simplify the task
of the psychologist. As mattersstand,however,thisisnot
justified.
Someone must have convinced him that his earlier
work on unconditioned and conditioned inhibition was
misguided. Hebb ([84], page 101) said that the existence
of a neural inhibition was not definitely known in 1949
and was not incorporated in the theory though it would
have helped. Peter Milner [133] showed how to incorp-
orate it.Hebb ([80] page 92) discussed neural inhibition
as the prevention of a nerve from firing. According to
Fairen ([42], page 436), Lorente de Nó conceived the
existence of inhibition, and wrote that internuncial neu-
rons might be the siege of it, then he disdained this con-
cept absolutely, and did not mention it further.
Conclusive evidence for inhibitory neurons was not
available until the concept of chemical neurotransmitters
was accepted and inhibitory synapses were identified
[40]. From this discussion, it seems that since Lorente
de Nó did not refer to inhibition, neither did Hebb.
Chemical neurotransmitters
Hebb never mentioned chemical neurotransmitters; all of
his ideas about synaptic activity were based on electrical
connections between cells, although he does not explicitly
state this. It is interesting that Hebb [77] cited all of the
Fig. 6 Concepts of inhibition as depicted by Lorente de Nó, Morgan and Milner. (a) A diagram explaining the production of reflex reversal by concurrent
stimulation of two fibers (f
1
and f
2
or f
2
and f
3
) from different peripheral sense organs and its maintenance by the impulses conducted by the closed chain C,
after fiber f
3
, which initiated the response of cell 3, ceases conducting. Each one of the links in the closed chain represents a multiple chain of (see M in Fig.
4Aofthispaper).Collateraldbyloweringthethresholdofcellaandthuscausingtwoimpulsestocrossthroughcell3inquicksuccessionmayproduce
inhibition, for cell 3 will acquire a high subnormal threshold. (From [124], page 231). (b) A diagram of neurons and synaptic connections involved in
reciprocal inhibition (innervation). Excitation at the synapses is widely held to be proportional to the number of active endings. For the sake of argument, the
minimal number of endings which must be active for excitation to occur is arbitrarily taken to be two. If a neuron, b, common to two pathways is switched
out of one pathway when it is taken up by another, the necessary condition for reciprocal innervation would be fulfilled. Let us suppose that rhythmic
stimulation of fiber I is maintaining a flexor reflex. Neurons b and c are excited, and their discharges arriving synchronously at F cause it to respond.Thenlet
us suppose that in the course of this response an extensor reflex is set up through stimulation of fiber II. The latter can excite b in the intervals between the
responses to I, because of the stronger excitation which it is able to deliver through its three endings. No discharges can result therefrom in F, as the
impulses in b are out of time with those in c; and I is no longer able to excite b and c in unison, because of the raised threshold of the former. Neuron b is
dominated by fiber II. Its discharges are caused to be synchronous with those in a, instead of those in c, and activity begins in E. Thus, when innervationof
the extensor muscles starts it must be withdrawn from the flexor muscles. (From Gasser [46] and reprinted in Lorente de Nó [124], page 231, and Morgan
[139], page 67). (c) Types of recurrent inhibitory connections. It is proposed that the short-axon inhibitory cells which receive recurrent collaterals from a long-
axon cell have fewer inhibitory connections to that particular long-axon cell than they do to other long-axon cells in the region. When Cell A is firing it
causes the inhibition of its neighbors, B, C, and D, but is not itself inhibited. In fact, because the surrounding long-axon cells cannot now be fired, there is no
way in which the short-axon cells discharging onto A can be fired. Therefore, as long as A continues to fire, it protects itself from being inhibited. (From [133],
page 246). aand bReprinted from Lorente de No R. Analysis of the activity of the chains of internuncial neurons. J Neurophysiol. 1938;1:20744 [124].
Copyright (1938), with permission from the American Physiological Society. cReprinted from Milner PM. The cell assembly: mark II. Psychol Rev. 1957;64:242
52 [133]. Copyright (1957), with permission from the American Psychological Association
Brown Molecular Brain (2020) 13:55 Page 20 of 28
papers from the 1939 symposium on the synapse without
ever mentioning the ongoing controversy between the elec-
trical and the chemical theories of synaptic transmission
that were discussed in many of these papers. The state of
knowledge about synaptic transmission was summarized by
Forbes ([45], page 471), who stated that we must not think
so loosely as to overlook the fact that the electrical potential
which excites a tissue is the same thing whether produced
by a dynamo or by a galvanic cell, and this is quite distinct
from acetylcholine - a substance which can be put in a bot-
tle, whatever its electrical properties.However Lorente de
Nó ([125], pages 420421) was clearly opposed to the con-
cept of chemical transmission and the reality of chemical
neurotransmitters in the brain was not acknowledged until
1954 [39,40].
Morgan ([139] pages 6263) discussed both the humoral
and electrical theories of synaptic transmission, so Hebb
knew about chemical transmission, but, as noted by Mor-
gan, the experiments to demonstrate chemical neurotrans-
mitters in the central nervous system have not proved
conclusiveand it is most convenient to regard transmis-
sion as electrical, for we are better acquainted with the elec-
trical phenomena of nervous activity than the chemical,
and it is easier, perhaps, to think in electrical rather than
chemical terms.([139],page.63).HebbssisterCatherine
was working on the role of the neurotransmitter acetylcho-
line in the brain, but this work did not begin until 1951 [8,
55]. Hebb did not discuss chemical neurotransmission in
The Organization of Behavior nor in his textbook of Psych-
ology [80]untilthefourthedition[86]. His reference to
humoral mechanisms in A textbook of Psychology ([80],
page 52) refers to estrogen.
The relationship of instinct to learning
Hebbs early papers from his PhD thesis were concerned
with innate behaviour. For example, Hebb [62]concluded
that: In the rat, the figure-ground organization and the
perception of identity in such geometrical patterns as the
solid triangle, outline of triangle, and triangle circumscribed
by a circle are innately determined(page 125). Likewise,
Hebb [63] referred to an innate property of visual percep-
tion(page 298). However, when he came to write The
Organization of Behavior, Hebb (in his Precis of [71]) says
that when he re-examined his thesis data, the dark reared
rats took a mean of 129 trials while controls took only 20
trials (page 23), He has described this on page 113 of The
Organization of Behavior and discussed how he came to
makethiserrorinhisEssayonMind([84] page 90).
Hebb discussed the relation of instinct to learning in
chapter 7 of The Organization of Behavior.Inthischapter,
Hebb considered factors that modulate learning, such as
attention, motivation and emotion. He was also concerned
with the innate processes of instinct that are thought to
take the place of learning(page 140). With respect to
innate or hereditary factors, Hebb ([77], pages 165170)
refers to Lashleys[116] paper on the experimental ana-
lysis of instinct, to Beachs[10] studies of instinct and
learning in rats and to Tinbergens[182] paper on the ob-
jective study of innate behaviour. Hebbs view was that it
was important to distinguish innate from learned behav-
iour, however, he pointed out that there is presumably no
mammalian behavior that is uninfluenced either by learn-
ing or by the constitution that makes some learning easy
or inevitable.He stated that Ultimately, our aim must be
to find out how the same fundamental neural principles
determine all behavior(page 166).
Hebb [78] expanded on these ideas when he said that
there is no behaviour, beyond the level of the reflex, that
is not essentially dependent on learning, and that no be-
haviour can be independent of an animalsheredity.
Hebbs conceptualization of innate and learned behaviour
and their interaction led him to examine the factors of
heredity, maturation and early learning in a series of ex-
periments beginning with rearing rats at home in 1947.
The research done by his students at McGill on environ-
mental enrichment and impoverishment and the effects of
environmental experience on perceptual and cognitive de-
velopment were his attempts to understand the inter-
action between heredity and environment [181].
The cell assembly versus the engram
Since the re-discovery of Richard W. SemonsbookThe
Mneme [167] by Schacter et al. [164], it has become popu-
lar to equate Hebbs cell assembly with Semonsengram
[100,185]. However Hebb did not refer to Semon in The
Organization of Behavior and only mentioned the en-
gramon page 15 of his 1945 Precis [71](This makes the
assumption that there is a dual mechanism of the trace, a
dynamicplus a structural engram.). He used the terms
memory traceand mnemonic trace(page 12) in The
Organization of Behavior.Lashley([118], page 3), on the
other hand, made the term engrampopular as the locus
of memory traces. Lashley [112] did not use the term en-
gramnor refer to Semon, but Lashley and Ball ([119],
page 97) refer to The engram of the maze habitand
Lashley [114]referstomazelearningabilityasthe
ekphore of Semon, so he was familiar with Semonsbook.
Lashley ([115], page 472) discussed the localization of the
engram but dismissed Semons ideas by saying Hering,
Semon, Rignano and others have held that memory is a
basic property of protoplasm; neurologists have been in-
clined to regard it as the product of some peculiar
organization of nerve cells([115], page 457). Hebb did
not refer to Semon nor the engram.
Semons mnemic theory was very complex and early re-
views pointed out that the mnemic principle attempted to
connect memory, habit and heredity[141]. Semonscon-
cept of the engram involved a number of Mnemonic
Brown Molecular Brain (2020) 13:55 Page 21 of 28