Psychological Review

Vol. 64, No. S, 1957

ANXIETY AND BEHAVIORAL AROUSAL1

ROBERT B. MALMO 2
Allan Memorial Institute of Psychiatry, McGill University

During the past two decades there in terms of direction, meaningful dis-
has been a growing interest in objective tinctions may be made between motiva-
physiological studies of psychiatric pa- tion and emotion, and indeed between
tients. In this work, one of the most different emotions. However, for pres-
prominent psychological concepts has ent purposes it is essential to focus on
been that of anxiety. Although there is the question of what these phenomena
general agreement that the areas de- have in common rather than to consider
noted by the term "anxiety" are impor- their differences; in this paper, there-
tant ones for study, there is nonethe- fore, we shall be primarily concerned
less considerable disagreement concern- with the intensity dimension.
ing what the term means. In large The main purpose of the present pa-
measure, this semantic difficulty is part per is to consider recent experimental
of a larger problem facing psychology data in an attempt to find a way out of
today, and that is to find a way out of the present confusion. I shall begin
the confusion surrounding the concepts with a summary of two lines of investi-
of motivation and emotion. Duffy has gation in our laboratory, dealing first
cogently argued that these concepts are with our discovery that certain physio-
second-order ones which reduce to pri- logical measures may serve as indicants
mary factors of intensity and direction, of intensity or "behavioral arousal."
and that along the intensity dimension, These experiments were performed with
at least, the distinction between motiva- nonpatient subjects. Second, in sum-
tion and emotion is unnecessary (9, 10, marizing our investigations of patho-
II). 3 logical anxiety in psychiatric patients,
This is not to say that the directional I shall attempt to use the concept of
aspect is not important or to deny that, behavioral arousal in an integrative way.
1
Third, I shall draw on data from recent
This paper reviews work which was sup- neurophysiological investigations to in-
ported by the Medical Research and Develop-
ment Division, Office of the Surgeon General, dicate possible mechanisms involved in
Department of the U. S. Army, under Con- the pathology and etiology of anxiety.
tract Number DA 49-007-MD-626, by De- Finally, on the basis of these theoretical
fence Research Board Grant Number 9425-04 considerations, I suggest problems re-
(Canada), and by Grant Number A.P. 29 quiring further experimental study.
from the National Research Council of Canada.
2
The author is indebted to Drs. A. K.
Bartoshuk, D. Bindra, F. R. Brush, D. E. PHYSIOLOGICAL INDICANTS OF
Cameron, D. O. Hebb, and R. G. Stennett for BEHAVIORAL INTENSITY
criticizing earlier drafts of this paper. In 1951 we (31) reported finding a
8
I do not wish to imply that this has been
Duffy's only theoretical contribution. Her gradient phenomenon from electromyo-
writings contain prior reference to a dimen- graphic (EMG) recording during mirror
sion of behavioral intensity (conceived as a tracing. Since that time the phenome-
continuum of "arousal," or "activation") ; and non has been observed under various
she has previously cited evidence to support
the argument that physiological measures may conditions in our laboratory. Figure 1
serve as the chief means of quantifying such presents mirror-drawing data from a
a dimension or continuum. study by Bartoshuk (1). Note that the
276
 ANXIETY AND BEHAVIORAL AROUSAL 277

chin lead (which taps the speech mus- been confirmed by three subsequent
cles) also shows a gradient—that is, pro- studies, employing tracking tasks. Sur-
gressively rising muscle potentials from willo (39) demonstrated that raising in-
the beginning to the end of the task. centive had the effect of increasing the
Belanger (3) found similar gradients steepness of EMG gradients in a visual
from the arm in a size-discrimination tracking experiment. Figure 2 presents
task. Wallerstein (42) found gradients confirmatory data from a more recent
in the frontalis muscle in a task about experiment by Stennett (37) who em-
as completely devoid of motor com- ployed auditory tracking under four con-
ponents as one could possibly design. ditions, with increasing degrees of incen-
The subject, reclining on a comfort- tive. Note that the muscle potentials
able bed, listened to verbal material were recorded from the nonactive, left
(short detective story or essay) pre- arm. His "exertion" condition merely
sented to him by a tape recorder. In involved the subject's holding the track-
Wallerstein's experiment, the gradients ing knob over at a fixed point in order
extended over ten minutes and their to control for sheer physical work. Un-
steepness was related to the subject's der the "calibration" condition the sub-
reported degree of interest in listening ject believed that he was just assisting
(2, p. 228 f.). with calibration of the apparatus, and
Bartoshuk (2) was the first to show that his tracking scores were not being
that the fastest and most accurate sub- recorded. The "optimal" condition was
jects (i.e., superior performers on mirror designed to motivate the subject suffi-
tracing) produced the steepest muscle- ciently to elicit his most efficient per-
potential gradients. Such a relationship formance, whereas the "incentive" con-
of EMG gradients to motivation has dition was designed to "overmotivate"
the subject by offering large bonuses for
|2 90 high-level performance and threatening
o with strong electric shock if performance
i 80
did not reach this high level. The dif-
O ferences shown in the figure were sta-
I
tistically significant. In brief, Stennett's
Z 70 findings indicated that the most effi-
Q
cient tracking performance was associ-
15
h-
60 ated with intermediate physiological lev-
els (i.e., intermediate steepness of EMG
gradients and intermediate levels of pal-
50
mar skin conductance). With lower
levels of physiological functioning (less
LU steep gradients, lower levels of pal-
fe
0. mar skin conductance), performance on
30
LU tracking was inferior. However, going
_J
O now to the other extreme, performance
20 on tracking associated with extremely
I 2 3 4 5 6
PRE SIXTHS OF TASK POST high EMG gradients and extremely high
FIG. 1. Graphs showing mean EMG gradi- palmar skin conductance was also in-
ents in Bartoshuk's experiment (1). Note ferior to tracking performance associ-
that gradient was also obtained from chin
lead which records from muscles of speech. ated with moderately high levels of
N=11. physiological functioning.
278 ROBERT B. MALMO

PCC DUPING P05T

•—• INCENTIVE
0—o OPTIMAL
•—• CALIBCATION
o—o EXCQTION
60

PA55IVC
PDONATOD TCBCS
>40
o
a

Z 20
tu
2 10
o

40
</} PA55IVC
!m PQONATOP. QUADCATU5
O
&30
o
I
20
Z
LJ
5 10

-30 30 60 90 150
TIME IN SECONDS

FIG. 2. Graphs from Stennett's experiment (37), showing mean EMG

gradients obtained under conditions varying in degree of incentive. Steep-
ness of gradient varies directly with degree of incentive. N = 3l.

If we consider our physiological meas- as palmar skin conductance, EEG 4 and

ures as indicants of arousal level, we gradients in skeletal muscle tension,
may say that performance suffered in heart rate, blood pressure and respira-
the first instance because of under- tion (26) should provide reliable meas-
arousal (or poor motivation), while in ures of the arousal variable. The ob-
the second instance it suffered from jective nature of the physiological meas-
overarousal (or emotional interference). ures is a highly desirable feature which
In short, as Stennett has previously frees the worker from dependence upon
stated (37), we believe that the concept merely manipulating situations in the
of arousal leads us in the direction of 4
Stennett (38) has found that the relation-
working out (empirically) a continuum ship of alpha amplitude to arousal level is
of behavioral intensity which promises nonlinear. On the lower end of the arousal
to have the very desirable feature of in- continuum the relationship is positive, such
that raising arousal leads to increasing alpha
tegrating the concepts of motivation and amplitude; but past the middle range of
emotion. From available data it ap- arousal the relationship becomes inverse. This
pears that physiological measures, such latter function is the better known one.
 ANXIETY AND BEHAVIORAL AROUSAL 279
hope that he is producing intended tensity dimension, in which level of
changes in the arousal level of his sub- physiological activity, arousal, and in-
jects. Moreover, the physiological in- tensive level are employed as roughly
dicants have the further advantage that synonymous terms.
they may be applied to work with ani- In short, the physiological measures
mals as well as with human subjects, appear to be useful tools in establishing
and may thus serve usefully to bridge and precisely quantifying a dimension of
the gap, in the field of motivation, be- behavioral intensity. Indeed, I regard
tween work on human and on infrahu- such objective measures as nearly indis-
man subjects. pensable to the achievement of a really
A word should be said concerning the satisfactory operational definition of be-
different physiological measures which havioral intensity. In the absence of
have served as indicants of behavioral such objective measures, it is difficult to
arousal. Although gradient steepness has see how circularity can be avoided. Con-
proved a very useful measure, level of siderable work is required, of course, in
palmar skin conductance seems equally working out the intensity dimension, and
promising. As a matter of fact, even while present results are indeed encour-
with EMG, the correlation between av- aging, many further data are required.
erage EMG level and gradient steepness It may be helpful just here to relate the
is usually so high that it is meaningless arousal continuum to the intensity di-
to ask which is a better indicant. We mension which Boring described (4).
still have much to learn concerning the While Boring's main concern was with
application of physiological techniques sensation, I believe that it is appropriate
to our problems. It may be that, as to consider that operations of measure-
Lacey's work suggests (23), for most ment comparable in precision to those
accurate assessment of arousal, special of psychophysics may be possible in the
consideration should be given to indi- field of action.
vidual differences in relative reactivity
of different physiological systems. EXPERIMENTAL STUDIES WITH
Following the usage of Freeman (15) PSYCHIATRIC PATIENTS
and Hebb (18), the term "arousal" is Having elucidated the concept of
used to refer to the intensive dimension. arousal with these reference experi-
I am aware that the term "arousal" is ments, we are now in a favorable posi-
used by some EEC workers to denote tion to take a fresh look at the data
flattening of an EEC tracing (e.g., 8, p. comparing patients and nonpatients with
132). When I use the term, I use it in respect to level of physiological reaction
a much broader sense, as a dimension of under controlled stimulating conditions.
behavior, and I am not using this term At the outset, we may say that the chief
to refer to the EEG phenomenon called impression which one gets in going over
"arousal" or "activation." It is for this all of these data is that, under "stress,"
reason that I specify behavioral arousal psychoneurotic patients appeared to
in the title of this paper. As investiga- show a higher level of physiological re-
tive work proceeds, it may become heu- action than controls, and that level of
ristic to make a definite distinction be- reaction seemed particularly high in pa-
tween physiological arousal and behav- tients suffering mainly from pathologi-
ioral intensity. Granting this possibility, cal anxiety. By pathological anxiety, I
I believe that for present purposes it mean a state of such severity that work
may be preferable to accept a rather efficiency is seriously affected over long
broad operational definition of the in- periods of time, and a state which is
280 ROBERT B. MALMO

characterized by one or more of the fol- arousal must be reached in order to

lowing complaints: persistent feelings of demonstrate differences between patients
"tension" or "strain," "irritability," "un- and nonpatients or whether such differ-
remitting worry," "restlessness," "inabil- ences could be obtained under resting,
ity to concentrate," "feelings of panic in "basal" conditions. From reviewing the
everyday-life situations." I should like literature prior to conducting our own
to make it very clear that I do not em- experiments, we were led to suspect that
ploy the term "anxiety" to refer to some stimulation would be necessary be-
transient affective states. When I use cause experiments which had been car-
the term I am talking about a patho- ried out under resting conditions had
logical condition which—as far as we usually yielded negative or inconclusive
can determine—develops slowly, over results.
months or years, and from which re- Our findings did indeed clearly show
covery (when it occurs) is also slow that, in differentiating between patients
and gradual. The experiments which and controls, some form of stimulation
we shall consider in this section em- was definitively superior to merely tak-
ployed patients suffering from "patho- ing records under resting conditions.
logical anxiety," as we have just denned This has been demonstrated for blood
it. For the sake of convenience, these pressure (28, p. 89), for muscle poten-
subjects will be called "anxiety pa- tials in motor tasks (31, p. 54 and pp.
tients." 59 ff.), and again for muscle potentials
In a study with pain as standard in two separate investigations of startle
stimulus (27), the following physiologi- (30, p. 327; 7, p. 181). The only meas-
cal measures showed significantly greater ure which we have found to discriminate
reaction in anxiety patients than in other well between patients and controls under
psychiatric patients: finger movement "resting" conditions was frontalis-muscle
(and number of voluntary pressures to potentials (33). However, we know that
indicate pain), neck-muscle activity, "resting" conditions associated with a
deviation in amplitude and rate of res- testing session are by no means basal,
piration throughout the test, respiratory and that—for example—significantly
irregularities occurring at time of stimu- lower blood-pressure readings may be
lation, and heart-rate variability. In a obtained from patients resting quietly
different study (29) with a perceptual on the ward than in the so-called "rest-
test and a Luria-type recording from ing" condition of our experiments (32).
the left hand, finger movement was sig- "Specific" vs. "nonspecific" stimulat-
nificantly more irregular in anxiety pa- ing situations. In producing higher lev-
tients than in other psychiatric patients. els of arousal in patients, is it necessary
To repeat an earlier statement, these to present material to which patients
findings indicate that under standard are specifically sensitized or is it pos-
conditions of stimulation psychoneurot- sible to demonstrate the difference be-
ics are more reactive than controls, and tween patients and controls by employ-
that patients with anxiety predominat- ing the same standard stimulating situa-
ing in the symptom picture are the most tion for all subjects? Our experiments
responsive of all. clearly show that the latter is true. It
is not necessary to present the patients
Need for "Standard Stress" in Demon- with words or situations which have spe-
strating Differences Between Patients cial meanings for them in order to pro-
and Controls duce more arousal in them than in con-
Another question which we sought to trols.
answer was whether a certain level of As an example of a "specific-corn-
 ANXIETY AND BEHAVIORAL AROUSAL 281

plex" technique of producing high-level assume that there is no need to look for
arousal, Luria (25) employed the method pathology in central and motor mecha-
of controlled association in which he nisms, because they believe that amount
compared motor reaction to "critical" of physiological disturbance is com-
words (those which were especially mensurate with the special significance
arousing for the subject because of their which the situation has for the patient.
association with specific life experiences) Implied in this view is the assumption
with reaction to indifferent words. Our that only those stimuli which, through
situations, on the other hand, were learning, have acquired special meaning
chosen for their general arousal value, for the patient have the power to pro-
and we sought to avoid situations which duce an "abnormal" level of arousal.
would have special meaning for particu- It assumes that the patient may partici-
lar individuals. pate in many situations without show-
With this point especially in mind we ing abnormally high levels of physio-
devised our standard situation of pain- logical reaction.
ful stimulation, because of the nearly However, this view may well be ques-
universal avoidance reaction to pain. In tioned because it does not appear to fit
order to permit more generalized con- with clinical observations. Cameron has
clusions, we also employed standard written as follows:
situations other than pain. One study It will be noted that nearly all such patients
is of especial interest because we repro- [with anxiety states] complain that they can-
duced the essential features of Luria's not go into crowded places or into any situa-
procedure, only substituting a series of tion where sustained efforts will be required
size discriminations for the series of of them. Their symptoms are made more
severe by anything which elicits emotional re-
verbal stimulations which Luria em- actions, such as altercations or participating in
ployed (29). Conclusions from these a discussion of illness. Nearly all find, at least
experiments were as follows. All meas- at first, that their symptoms are increased by
ures of motor activity recorded during visiting their former places of employment or
performance of speeded size discrimina- meeting fellow-workers. In other words, their
symptoms are exacerbated by anything which
tion yielded reliable differences between serves to increase tension. Emphasis should
patients and controls. In every instance be placed upon the fact that their symptoms
there was evidence of greater physio- are elicited or intensified, not primarily by the
logical disturbance in the patients. The reactivation of any conflict situation which
may exist, but literally by everything in the
measures employed may be distinguished course of the day which serves to increase
as skeletal-motor (motor control, muscu- tension (5, pp. 56-57. Italics mine).
lar tension) and autonomic (systolic
blood pressure). These differences in In therapy, relaxants of various kinds
motor activity were manifested even are devised to "damp" the "autono-
though psychoneurotics, acute psychot- mous" reaction before proceeding with
ics, and controls were practically identi- psychotherapy (41).
cal with respect to perceptual perform- Strong auditory stimulation. Strong
ance. auditory stimulation served as another
These results led us to question cer- and very different kind of standard
tain views concerning determinants of stimulating situation for comparing pa-
higher arousal levels in psychoneurotics. tients and controls. Two separate stud-
In much current writing there is the un- ies, the first one (30) with induced ten-
derlying assumption that physiological sion (produced by squeezing a rubber
disturbances in the psychoneurotic can bulb), and the second (7) without in-
be accounted for entirely in terms of duced tension and with a less intense
situational explanations. These writers stimulus, agreed in showing that the
282 ROBERT B. MALMO

most reliable difference between anxious While Eccles' work was done on cells
patients and controls was in "after-re- in the spinal cord, it nonetheless seems
sponse" following the period of primary reasonable to suggest that the reticular
reflex-startle reaction. formation could produce widespread in-
hibition in the cortex by hyperpolariz-
NEUROPHYSIOLOGICAL CON- ing cortical cells. Because the study of
SIDERATIONS neuronal discharge in the cortex is a
In the interpretation of our findings new field of research, sufficient data to
in the experiments on strong auditory decide this point are not at hand. But
stimulation (7, 30), we cited the paral- data which are presently available seem
lel between these observations on pa- to be in line with the proposition that
tients and findings in neurophysiologi- some impulse arriving in the cortex may
cal experiments on the reticular forma- have facilitatory effects, while others
tion. In certain animal preparations, may produce opposite results (21, Fig.
after-discharge in the cerebral cortex 19, p. 62).
was abolished by stimulation in the If Eccles' theory is essentially cor-
reticular formation of thalamus (20) rect,5 we may work with inhibition as
and brain stem (35). We believe that an independent process, and seek to un-
it is reasonable to suggest that some derstand the pathology of anxiety in
such inhibitory mechanism (as the one terms of weakened inhibition. To make
which abolished after-discharge) may matters more concrete, we may draw on
be weakened in pathological anxiety. Eccles' hypothesis of a chemical trans-
Having implicated inhibition, we are mitter for inhibition (12, p. 163) and
required to examine this concept criti- on the recent experimental work of El-
cally for a moment. Although there is liott and Florey (13) to suggest that, in
by no means complete agreement on the anxiety, the effectiveness of this sub-
matter of inhibitory mechanisms in the stance has been reduced.
central nervous system, present evidence
appears to point more and more in the THE PROBLEM OF ETIOLOGY
direction of inhibition as a phenomenon The disorder of pathological anxiety
in its own right, independent of excita- may be conceived of almost entirely in
tion (i.e., not merely absence of excita- terms of constitutional factors. It is
tion). logical to consider that certain indi-
Of the current theories of inhibition viduals may inherit a deficient inhibi-
known to me, Eccles' view seems most tory mechanism. Such a person would
reasonable (12). Eccles and his co- consistently suffer from inability to re-
workers developed a technique for plac- lax throughout life, and would be seri-
ing a microelectrode within a single ously limited in the amount of stimula-
spinal motoneurone, and they were thus
B
able to observe the electrical potential Recent findings, although supporting Ec-
between the inside and the outside of cles' main conclusions, suggest that the phe-
nomenon may be somewhat more complex
the cell. They observed that when they than he originally supposed. The observations
stimulated an inhibitory nerve fiber it of Kuffler and Eyzaguirre (22) on inhibition
increased the polarization of the nerve of stretch receptor organs in crustaceans indi-
cell on which it ended. Eccles called cate that the polarity of the "inhibitory po-
this effect "hyperpolarization," which, tential" varies with the state of the cell.
When the cell is depolarized, an inhibitory
electrically is the opposite of what oc- volley causes polarization; when the cell is
curs when a nerve cell is fired (depolar- resting, an inhibitory volley causes depolariza-
ized). tion.
 ANXIETY AND BEHAVIORAL AROUSAL 283

tion that he could withstand. In such able to avoid such situations, and he
a case the constitutional weakness, rather will, therefore, be more or less con-
than learning, would be the primary stantly operating at physiological levels
factor in etiology. While constitutional which are higher than normal. We may
differences of genetic origin may account conjecture that in such a case in which
for degree of predisposition to the patho- stimulation keeps physiological levels
logical condition of anxiety, clinical evi- constantly very high, over a long pe-
dence stands against a purely genetic riod of time there will be a weakening
etiology. The fact that such a large of inhibitory mechanisms from overuse.
number of patients recover from anx-
iety states (17, 34) argues against a FURTHER CLINICAL-EXPERIMENTAL
purely genetic-constitutional explanation CONSIDERATIONS
of pathological anxiety. Anxiety in combat. If our theory is
Declining the genetic-constitutional ex- correct, anxiety may be considered as a
planation of anxiety implies that learn- "disease of overarousal" (or in Selye's
ing mechanisms are somehow involved [36] terms, a disease of "adaptation").
in the pathology. In order to under- That is, the critical neural change is
stand the full implications of this point thought of as being produced by a proc-
of view, it is helpful to consider that de- ess of attrition from excessive and fix-
gree of arousal is not a "given" in the tended overarousal. It would not mat-
stimulating situation. The same stimu- ter whether this overarousal were pro-
lating situation may produce quite dif- duced in an individual whose previous
ferent levels of physiological reaction in learning made him more prone to over-
different persons, depending upon the arousal, or whether the individual were
effects of past learning.8 We may com- anxiety-resistant from past training, and
pare individuals with respect to their was simply "overexposed" to situations
physiological reactions in a large num- (like battle) that everyone reacts to
ber of different situations. We may find, with extremely high physiological levels.
for example, that a certain person gen- With this view we can readily under-
erally shows significantly higher levels of stand why under battle conditions each
physiological reaction than most other soldier would have his "breaking point,"
individuals. If this person can avoid and why despite resistance to over-
stimulating situations with high-arousal arousal from constitution and previous
values he appears no different from learning, if situations of high-arousal
others. However, in ordinary, every- level are repeated over a long enough
day living, it is unlikely that he will be time period, the critical change will
finally occur. This seems to be the pic-
* The reader will recall that in our physio-
logical studies of psychiatric patients we at-
ture which emerges from studies of anx-
tempted to avoid experimental situations iety in combat (16, pp. 85 ff.).
which had special meanings for particular in-
dividuals. In an earlier section of this paper Inhibitory Deficiency in Anxiety and in
we referred to these situations as "nonspecific." Manic States
We assume that an anxiety-prone individual,
before he actually develops the pathological From the clinical point of view,
state (and after he recovers from it), will not Cameron (6, p. 388) has drawn atten-
show higher arousal levels in such "nonspe- tion to the prominence of overactivity
cific" stimulating situations. The stimulating in the anxiety states. Cameron is in-
situations referred to as producing quite dif-
ferent levels of physiological reaction in dif-
clined to believe, however, that the
ferent persons are, of course, what we called manic state best represents "pure over-
"specific" in the earlier section of this paper. facilitation," in comparison with anx-
284 ROBERT B. MALMO

iety, which he has described as "curbed taken. However, as far as I am aware,

overactivity." In drawing this compari- the study has not been carried out with
son, Cameron was influenced by his anxiety states in the way proposed.
careful observation of body movements.
He found that the typical anxious pa- Anxiety and Learning
tient was restless and in constant move- Physiological measures of arousal
ment, but that he did not have the open, should prove valuable in learning ex-
wide, flung-out movements of the manic. periments in which anxiety has been
In general, the movements of the anx- studied as a variable (14). For ex-
ious patient remained within the body ample, workers have employed question-
silhouette.7 naires and scales (e.g., the Taylor scale
The internally generated manic over- [40]), to select subjects high in "anx-
activity ("pure overfacilitation") could iety." The chief purpose of such ex-
reasonably be accounted for by positing periments has been to compare the
increased activity of facilitatory mecha- learning speed of subjects scoring high
nisms. on such a scale with other subjects scor-
ing lower on the scale. It would appear
PROBLEMS FOR FURTHER STUDY that physiological measures could be
The line of reasoning followed in the applied to such problems with consid-
present paper suggests certain hypothe- erable advantage. Subjects who would
ses which might be put to experimental probably react at high physiological lev-
test. In the first place, longitudinal els could still be selected with the scales
physiological study of patients suffer- as an initial screening device; but
ing severe states of anxiety should re- physiological measurements could then
veal changed physiological reaction un- be applied to provide actual values to
der conditions of standard stimulation. place each subject on a continuum.
That is, during performance of a motor Such methodology would appear promis-
task—for example, palmar skin con- ing in providing a continuous variable
ductance—electromyographic gradients (i.e., physiological intensity, or arousal)
and other physiological indicants of for study in place of the rather dubi-
arousal should show decline when the ous anxiety-nonanxiety dichotomy, and
patient is in remission, and should show would have other advantages. For ex-
increase again with relapse and return ample, a low scorer on the scale might
of the anxiety. This is a straightfor- be temporarily upset, and so be misclas-
ward kind of investigation which one sified in an experiment unless his actual
might suppose had already been under- physiological measures were available
7 on the day of the experiment.
On the surface, this appears incongruous
with the notion of weakened inhibition. How- Research with Reserpine and Chlor-
ever, we may account for this constrained
appearance of inhibition by suggesting the promazine
substitution of less efficient mechanisms of Patients exhibiting anxiety as the pre-
inhibition for the one which has suffered im- dominant symptom have been reported
pairment. It may be, for example, that anx-
iety patients compensate for weakened au- to improve significantly following the
tonomous mechanisms by calling on voluntary administration of reserpine and chlor-
motor mechanisms (i.e., the pyramidal mo- promazine (19). It should prove illumi-
tor system). For example, in the absence of nating to study the effects of such drugs
sufficient control from autonomous inhibitory on physiological reaction of anxiety pa-
mechanisms, the anxiety patient may avoid
loss of motor control through co-contraction tients under controlled stimulating con-
of antagonistic muscles. ditions. For example, with administra-
 ANXIETY AND BEHAVIORAL AROUSAL 285
tion of these drugs, would the elec- The main purpose of such an experi-
tromyographic reaction of patients to ment would be to determine whether
strong auditory stimulation resemble the keeping physiological levels constantly
normal reaction more closely (show less high would finally produce "anxiety" in
after-response) than in the absence of animals (i.e., animals with raised physio-
the drugs? logical levels in standard test situations).
It would likewise be of interest to de- If such experiments did turn out posi-
termine the effect of such drugs on lev- tively, valuable animal "preparations"
els of physiological reaction in anxiety would be available for neurophysiologi-
patients under conditions of moderate cal and pharmacological studies.
stimulation, such as those in our experi- Such a "preparation" might be used,
ments with pain and with performance for example, to determine whether in-
tasks. Would drug administration bring hibitory effects from stimulation in the
levels of physiological reaction down reticular formation are weaker than in
close to normal values under these con- normal animals. We might even con-
ditions? ceive of an experiment paralleling the
As a matter of fact, our experiments ones which we carried out with human
with psychiatric patients were performed subjects. It would seem possible to im-
prior to the full development of the con- plant electrodes in the reticular forma-
cept of an intensity continuum in be- tion to search for areas which fire in-
havior, measured in terms of EMG hibitory impulses to the cerebral cortex
gradients, level of palmar skin conduct- following strong auditory stimulation.
ance, and other such physiological indi- Furthermore, pharmacological investiga-
cators. It would be highly desirable, tion (13) might be directed to the ques-
therefore, to apply these more refined tion whether there is an inhibitory sub-
physiological measures to the study of stance in the brain which becomes dilute
anxiety patients. Do they, in fact, show with long-continued overarousal.
steeper EMG gradients than normals in
tracking, and are these gradients re- SUMMARY
duced in slope with administration of The main purpose of this paper is to
reserpine and chlorpromazine? consider some recent experimental data
which suggest a way out of the present
Proposed Animal Experiments confusion surrounding the concepts of
Certain aspects of these problems may motivation, emotion, and anxiety. Two
be more advantageously studied with lines of investigation, each employing
animal subjects. Studies of "experi- physiological methods, are examined.
mental neuroses," as reviewed by Lid- In one experimental program, measures
dell (24), have shown that it is possible such as steepness of muscle-potential
to produce chronic states characterized gradients and level of palmar skin con-
by physiological deviation. For present ductance were found to be useful indi-
purposes it would be desirable to em- cants of arousal level. The results of
ploy a form of stimulation which effec- several experiments demonstrated signifi-
tively maintains high levels of physio- cant relationships between such physio-
logical reaction over long periods of logical indicants and excellence of per-
time. For our purposes it would not formance on various motor tasks, such
matter particularly how the stimulation as mirror tracing and tracking. In this
was produced; the main requisite is that empirical setting, problems of relation-
high physiological levels be recorded ship between concepts of motivation and
continuously over days and weeks. emotion are reconsidered.
286 ROBERT B. MALMO

The arousal concept is then applied 4. BORING, E. G. The physical dimensions

to the problem of pathological anxiety of consciousness. New York: Century,
1933.
in psychiatric patients. The earlier re- 5. CAMERON, D. E. Autonomy in anxiety.
sults from physiological studies carried Psychiat. Quart., 1944, 18, 53-60.
out with psychiatric patients as subjects 6. CAMERON, D. E. Some relationships be-
are reviewed in the light of the more tween excitement, depression, and anx-
recent work on physiological indicants iety. Amer. J. Psychiat., 194S, 102,
385-394.
of arousal. Considerable confusion has 7. DAVIS, J. F., MALMO, R. B., & SHAGASS, C.
arisen because the term "anxiety" has Electromyographic reaction to strong
been used to denote two quite different auditory stimulation in psychiatric pa-
states of the organism: (a) any increase tients. Canad. J. Psychol, 19S4, 8,
in level of arousal, however brief the 177-186.
8. DELAFRESNAYE, J. F. (Ed.). Brain mecha-
rise (or however selective the stimu- nisms and consciousness. Springfield,
lating condition); and (b) a patho- III.: Thomas, 19S4. (See especially dis-
logical state in which the patient ap- cussion by H. H. Jasper, p. 132.)
pears chronically overreactive (physio- 9. DUFFY, ELIZABETH. The conceptual cate-
gories of psychology: a suggestion for
logically) to every stimulating situation.
revision. Psychol. Rev., 1941, 48, 177-
It seems reasonable to restrict the 203.
term "anxiety" to the chronic patho- 10. DUFFY, ELIZABETH. A systematic frame-
logical condition. Results from physio- work for the description of personality.
logical studies carried out with patients /. abnorm. soc. Psychol, 1949, 44, 175-
190.
suffering this pathological condition in- 11. DUFFY, ELIZABETH. The concept of en-
dicated that standard stimulation (or ergy mobilization. Psychol. Rev., 1951,
"stress") accentuated the differences in 58, 30-40.
arousal between anxiety patients and 12. ECCLES, J. C. The neurophysiological ba-
controls. Under resting conditions such sis of mind. Oxford: Clarendon, 1953.
13. ELLIOTT, K. A. C., & FLOREY, E. Factor
differences were usually insignificant. I—Inhibitory factor from brain. Assay.
On the basis of the data reviewed, cer- Condition in brain. Simulating and an-
tain hypotheses concerning the nature tagonizing substances. /. Neurochem.,
and etiology of pathological anxiety are 1956, 1, 181-192.
14. FARBER, I. E. Anxiety as a drive state.
tentatively advanced. It is suggested In M. R. Jones (Ed.), Nebraska Sym-
that anxiety may be produced in an in- posium on Motivation. Lincoln: Uni-
dividual (in animal as well as in man) ver. of Nebraska Press, 1954.
by keeping level of arousal very high 15. FREEMAN, G. L. The energetics of human
over long periods of time. Finally, re- behavior. Ithaca, N. Y.: Cornell Uni-
ver. Press, 1948.
cent neurophysiological findings are cited 16. GRINKER, R. R., & SPIEGEL, J. P. Men
in stating the hypothesis that such con- under stress. Philadelphia: Blakiston,
tinuous overarousal may result in im- 1945.
pairment of central inhibitory mecha- 17. HARRIS, A. The prognosis of anxiety
nisms. states. Brit. med. J., 1938, 2, 649-664.
18. HEBB, D. O. Drives and the C.N.S. (con-
REFERENCES ceptual nervous system). Psychol. Rev.,
1955, 62, 243-254.
1. BARTOSHTJK, A. K. Electromyographic 19. HOLLISTER, L. E., TRATJB, L., & BECKMAN,
gradients in goal-directed activity. W. G. Psychiatric use of reserpine and
Canad. J. Psychol, 19S5, 9, 21-28. chlorpromazine. Results of double-blind
2. BARTOSHUK, A. K. Electromyographic studies. In N. S. Kline (Ed.), Psycho-
gradients as indicants of motivation. pharmacology. Washington, D. C.:
Canad. J. Psychol, 19SS, 9, 215-230. Amer. Assoc. for Advancement of Sci-
3. BELANGER, D. J. "Gradients" musculaires ence, 1956.
et processus mentaux superieurs. Canad. 20. JASPER, H. H. Diffuse projection systems:
J. Psychol. (in press). the integrative action oi the thalamic
 ANXIETY AND BEHAVIORAL AROUSAL 287
reticular system. EEC Clin. Neuro- stress. /. din. exp. Psychopath., 1951,
physiol., 1949, 1, 405-420. 12, 45-66.
21. JUNG, R. Neuronal discharge. EEC Clin. 32. MALMO, R. B., SHAGASS, C., & HESLAM,
Neurophysiol., 19S3, Suppl. No. 4, 57- R. M. Blood pressure response to re-
71. peated brief stress in psychoneurosis: a
22. KUFFLER, S. W., & EYZAGUTRRE, C. Syn- study of adaptation. Canad. J. Psy-
aptic inhibition in an isolated nerve cell. chol., 1951, S, 167-179.
/. gen. Physiol., 1955, 39, 155-184. 33. MALMO, R. B., & SMITH, A. A. Forehead
23. LACEY, J. I. Individual differences in tension and motor irregularities in psy-
somatic response patterns. /. comp. choneurotic patients under stress. /.
physiol. Psychol, 1950, 43, 338-350. Personality, 1955, 23, 391-406.
24. LIDDELL, H. S. Conditioned reflex method 34. MILES, H. H. W., BARRABEE, EDNA L., &
and experimental neurosis. In J. McV. FINESINGER, J. E. Evaluation of psy-
Hunt (Ed.), Personality and the be- chotherapy. Psychosom. Med., 1951,
havior disorders. New York: Ronald, 13, 83-105.
1944. Vol. I, pp. 389-412. 35. MORTJZZI, G., & MAGOUN, H. W. Brain
25. LURIA, A. R. The nature of human con- stem reticular formation and activation
flict. New York: Liveright, 1932. of the EEG. EEC Clin. Neurophysiol.,
26. MALMO, R. B., & DAVIS, J. F. Physiologi- 1949, 1, 455-473.
cal gradients as indicants of "arousal" 36. SELYE, H. Stress. Montreal: Acta, 1950.
in mirror tracing. Canad. J. Psychol., 37. STENNETT, R. G. The arousal continuum.
1956, 10, 231-238. /. exp. Psychol. (in press).
27. MALMO, R. B., & SHAGASS, C. Physiologic 38. STENNETT, R. G. The relationship of
studies of reaction to stress in anxiety alpha amplitude to the level of palmar
and early schizophrenia. Psychosom. conductance. EEG Clin. Neurophysiol.,
Med., 1949, 11, 9-24. 1957, 9, 131-138.
28. MALMO, R. B., & SHAGASS, C. Studies of 39. SURWLLLO, W. W. Psychological factors
blood pressure in psychiatric patients in muscle-action potentials: EMG gradi-
under stress. Psychosom. Med., 1952, ents. J. exp. Psychol., 1956, 52, 263-
14, 82-93. 272.
29. MALMO, R. B., SHAGASS, C., BELANGER, 40. TAYLOR, JANET A. A personality scale of
D. J., & SMITH, A. A. Motor control manifest anxiety. J. abnorm. soc. Psy-
in psychiatric patients under experimen- chol., 1953, 48, 285-290.
tal stress. J. abnorm. soc. Psychol., 41. TYHURST, J. S., & RICHMAN, A. Clini-
1951, 46, 539-547. cal experience with psychiatric patients
on reserpine—preliminary impressions.
30. MALMO, R. B., SHAGASS, C., & DAVIS, J. F. Canad. med. Assoc. J., 1955, 72, 458-
A method for the investigation of so-
459.
matic response mechanisms in psycho- 42. WALLERSTEIN, H. An electromyographic
neurosis. Science, 1950, 112, 325-328.
study of attentive listening. Canad. J.
31. MALMO, R. B., SHAGASS, C., & DAVIS, J. F. Psychol, 1954, 8, 228-238.
Electromyographic studies of muscular
tension in psychiatric patients under (Received December 10, 1956)