Chapter 4

ANXIETY INTERPRETED PSYCHOLOGICALLY

Anxiety is the fundamental phenomenon and the central problem of neurosis.

—Sigmund Freud, THE PROBLEM OF ANXIETY

DO ANIMALS HAVE ANXIETY?

The investigations of anxiety-like reactions in animals have shed important illumination on the problem of anxiety in human beings. We use the term “anxiety-like reactions” because there is considerable difference of opinion as to whether animals experience anxiety or not. Goldstein believed animals do have anxiety, but he was using the term to refer to undifferentiated fright reactions, parallel to the “normal” anxiety which may be seen in the two-week-old human infant. Harry Stack Sullivan held animals do not have anxiety. O. Hobart Mowrer, in his early studies of the “anxiety” of rats (which will be reviewed later in this chapter), used the terms “fear” and “anxiety” interchangeably. But later on, he concluded that the apprehension his animals were experiencing was fear, and that animals did not have anxiety except as they were placed in a special psychological relationship with human beings, such as the experimenters. In contrast to Goldstein, however, Mowrer was using the term “anxiety” to refer to neurotic anxiety, which by definition presupposes the capacities for self-consciousness, repression, and so forth which are uniquely the possessions of the human mammal.

It is Howard Liddell, in my judgment, who cuts the Gordian knot of this controversy. In a paper which is remarkably pertinent to our study of anxiety, based on his extensive work in experimental neurosis with sheep and goats, Liddell held that animals do not have anxiety in the meaning of that term as applied to human beings, but they do have a primitive, simple counterpart—namely vigilance.1 When an animal is in a situation that involves a possible threat—such as the sheep in the laboratory expecting an electric shock, or the seal sleeping in its natural habitat having to awake every ten seconds to survey the landscape lest Eskimo hunters sneak up on it—the animal exhibits an alertness and a general expectancy of danger. It is as though the animal were asking, “What is it?” This vigilance is characterized by generalized suspiciousness (indicating that the animal does not know whence the danger may arise), with tendencies to act but without any clear-cut direction for acting. Such behavior, as will readily be seen, is the parallel on the animal level to the vague and generalized apprehensive behavior of the human being in anxiety.

Liddell holds that Goldstein was describing this vigilance in his concept of the “catastrophic reaction,” but he adds that the fact that Goldstein pegged the reactions at a high level of intensity has kept other investigators from identifying the catastrophic reaction. This seems very true. In conditioning experiments, vigilance may be shown not only at a high intensity—as in experimental neurosis, when the animal gives a very clear picture of Goldstein’s catastrophic condition. But it is also shown through all gradations down to such a low intensity as a “small movement of the eyes or a slight acceleration of the heart.”

It is this vigilance, states Liddell, which supplies the power for the conditioned reflex. While Pavlov was astonishingly accurate in his description of the neurophysiological mechanics of conditioning, Liddell believes he was inaccurate when he contended that the motive power for the conditioning came from instinctual sources—e.g., the dog’s instinctual desire to get food or to avoid pain and discomfort. Liddell writes that “the conditioning machinery is not powered, as Pavlov believed, by a leakage of energy along a newly formed pathway or channel from a highly energized unconditioned reflex center to a sensory center feebly energized by sensory impulses set up by the conditioned stimulus.” Rather, it is powered by the animal’s capacity for vigilance or, in other words, by the animal’s capacity as a behaving organism to be alert to, and suspicious of, its environment. Liddell’s distinction here, which places the problem on a psychobiological rather than a neurophysiological level, is the point we have emphasized in the previous chapter—namely that the neurophysiological media by which behavior occurs must not be confused with the causes of the behavior. If the animal is to be conditioned—i.e., if it is to learn to behave in an orderly fashion—it must be able to get some reliable answer to its question, “What is it?” Thus in conditioning experiments consistency is all-important.

The reader will be aware that we here step, with Liddell, from the realm of physiology—i.e., instincts—to the organismic realm. It is so easy to think and talk in terms of “energy” released by the instincts as though we were dealing with a species of electricity, and it were some specific power we could measure and hopefully harness. Liddell makes it very clear that the reality is quite different: we are concerned with a defensive reaction of the total organism as seen in Liddell’s dogs and sheep, including their senses of sight, hearing, smelling, touching, etc., as well as the neurological and physiological media for transmitting the signals. These capacities add up to vigilance on the part of the animals, the precursor of human anxiety.

The capacity to experience anxiety, Liddell states, and the capacity to plan are two sides of the same coin. He holds that “anxiety accompanies intellectual activity as its shadow, and that the more we know of the nature of anxiety, the more we will know of intellect.” Thus Liddell states one aspect of the problem, which was attacked by Kierkegaard and Goldstein, and with which we shall wrestle again and again in this book, of the relation between creative potentialities and the potentiality for experiencing anxiety. The human being’s capacity for imaginative reality-testing, for dealing with symbols and meanings, and for changing behavior on the basis of these processes—all are processes which are intertwined with our capacity to experience anxiety.2

It remains only to note that Liddell, like myself and so many other investigators cited in the present study, see the social nature of the human being as the source of our uniquely creative intellectual capacities as well as our capacity for anxiety. (“Social” is defined here as interpersonal and intrapersonal.) Liddell asserts “that both intellect and its shadow, anxiety, are products of man’s social intercourse.”3 This social intercourse would not be possible, I must also emphasize, without the inner potentialities which we associate with individuality.

THE STUDY OF CHILDREN’S FEARS

If we think of children’s fears as responses to specific threats, on the apparently sound assumption that the child will fear things that have actually been of danger to him or her in their past experience, we are in for some surprises. Fear of apes, polar bears, tigers—animals the child will never have encountered except in rare visits to the zoo—will appear most in the reports of children’s fears. Dangers from mysterious agents like ghosts, witches, and the occult, which again the child has never encountered, also bulk large in the reports of fears. Why should the things feared be so largely imaginary? These and similar questions which bear on the relation of fears to anxiety encourage us to probe more deeply into the origin of childhood fears and anxiety.

Several decades ago the chief problem of the psychology of fear was to discover the original, unlearned stimuli which give rise to fears, and to account for these fears by instinctive processes. The child was supposed to have instinctive fears of darkness, animals, large bodies of water, slimy things, and so forth—many of these assumed by Stanley Hall to have been inherited from the animal ancestors of man. Then it became the task of many psychologists to disprove these “inherited fears” one after another, until in John B. Watson’s behaviorism the field was reduced to two. Says Watson of the infant, “there are just two things which will call out a fear response, namely, a loud sound, and loss of support.”⁴ All subsequent fears, this hypothesis contended, are “built in”—i.e., established by conditioning.

But later students of children’s fears have pointed out that Watson’s view was a gross oversimplification. Various investigators have been unable to find these two “original fears” with any consistency in infants. As Jersild writes, “the fear stimulus cannot be described as consisting of an isolated stimulus. . . . The circumstances that may give rise to so-called ‘unlearned’ fears in the infant include not simply noises and loss of support, but any intense, sudden, unexpected, or novel stimulus for which the organism appears to be unprepared.”5 That is to say, situations to which the organism cannot react adequately constitute a threat and are reacted to with anxiety or fear.

I submit that the debate on “original fears,” represented by the instinctivists on one hand and the behaviorists on the other, was a tilting at windmills. The endeavor to determine what specific fears the infant is born with causes us to become lost in a labyrinth of misleading questions. The fruitful question, rather, is: What capacities (neurological and psychological) does the organism possess for meeting threatening situations? With respect to inheritance, we need only assume that the organism reacts to situations for which its capacities are inadequate with anxiety or fear, and did so in the days of our ancestors as well as today. The problem of the acquisition of anxiety and fears after birth boils down to the two questions of maturation and learning. I also question whether these reactions of the infants described by Watson ought accurately to be called “fears” at all. Are they not, rather, undifferentiated defensive reactions properly to be termed anxiety? Such a hypothesis would account for the unspecific character of the reactions—i.e., that the “fear” is not found consistently even in the same child in response to a specific stimulus.

Maturation in Anxiety and Fears

Another defect in the Watsonian approach to children’s fears was its neglect of the factor of maturation. As Jersild observed in this connection, “If a child at a certain stage of development exhibits behavior that was not shown at an earlier time, it does not follow that the change in behavior is due primarily to learning.”⁶

It will be recalled from the discussion of the startle pattern above that in the early weeks of life the infant exhibited the startle response and little or nothing which could be called the emotion of fear. But as the infant developed, more and more secondary behavior (anxiety and fear) appeared. Jersild found in his studies that at about five or six months the child showed occasional signs of fear at the approach of a stranger, while before that level of development the child showed no such reaction.

Gesell’s account of the reaction of the infant at different ages to confinement in a small pen is illuminating. At ten weeks the child was complaisant; at twenty weeks he showed mild apprehension, one sign of which was persistent head-turning. (To me, this “persistent head-turning” is a significant picture of vigilance and mild anxiety; the infant is disquieted, but it cannot spatially locate the object of its apprehension.) At thirty weeks his response to the same situation “may be so vigorously expressed by crying that we describe his reaction as fear or fright.”7 As Jersild expresses it, “the tendency to respond to an event as actually or potentially dangerous is relative to the child’s level of development.”⁸

It seems clear that the level of maturation is one determinant in the infant’s or child’s response to danger situations. The data suggest that the earliest reactions are of the reflexive variety (i.e., startle) and of diffuse, undifferentiated apprehension (anxiety). Though this diffuse apprehension may be elecited in the early weeks of the infant’s life in response to certain stimuli (e.g., falling), it is more frequent as the infant develops greater capacity to perceive danger situations. As regards specific fears, are not these the latest to appear in the scale of maturation of the infant? As Goldstein has indicated, responding with a specific fear presupposes the capacity to objectivate—i.e., to discriminate between specific objects in the environment; and this capacity requires greater neurological and psychological maturation than to respond in a diffuse, undifferentiated way.

René Spitz has coined the phrase “eighth-month anxiety” for the apprehension shown by a baby between eight and twelve months when it is confronted by a stranger. The baby may respond with a look of bewilderment, may cry, or may turn and try to crawl to its mother. Spitz explains that this is due to the fact that the child, through his gradual maturation, has learned to put together his observations so that he recognizes mother and familiar objects. But this perception is not yet solidified and is easily upset by the appearance of another person in the field where mother ought to be. Hence the anxiety aroused when his perceptions are disturbed by a stranger.9

As the child develops after the infant period, Jersild points out, new and significant changes occur in the kinds of stimuli which give rise to fears. “With the development of the child’s imaginative abilities, his fears become increasingly concerned with imaginary dangers; with the development of understanding of the meaning of competition and of awareness of one’s own status as compared with others, there frequently come fears of loss of prestige, ridicule, and failure.”¹⁰

It seems clear that in the latter of these apprehensions—those related to competition—the child is engaging in a more or less complex interpretation of environmental situations. Such interpretive processes presuppose a certain amount of maturation, but they also patently involve the experience and conditioning which occur as the culture makes impact upon the child. Not only has it been found that fears related to competitive status increase as the child matures, but also, interestingly enough, that adults reporting on their remembered childhood fears gave a much higher percentage of apprehensions related to competitive status than any of the groups of children studied. This is rightly explained as due to the tendency of the adults to “read back” into their childhood the fulcra of fear and anxiety which had become increasingly important to them in adulthood.

It is unnecessary to review in detail the results of Jersild’s comprehensive studies of children’s fears. But two problems arise as one contemplates these results. We cite these problems because of the illumination they shed upon the relation of fears to underlying anxiety.

First, Jersild’s results show the “irrational” qualities of children’s fears. There was a sharp discrepancy between the fears the children reported and what they described, later in the interview, as the “worst happenings” in their lives.11 The worst happenings were cited as illnesses, bodily injuries, misfortunes, and other experiences that had actually befallen the children. Their fears, on the other hand, were “predominantly described in terms of somewhat vague calamities that might occur.” Actual terrifying experiences with animals constituted less than 2 per cent of the “worst happenings,” whereas fears of animals accounted for 14 per cent. The animals feared were chiefly remote creatures like lions, gorillas, and wolves. Being left alone in the dark and being lost were given as 2 per cent of the actual experiences, while fears of such situations amounted to about 15 per cent. Fears of mysterious agents like ghosts, witches, and the occult accounted for over 19 per cent (the largest group) of the total fears expressed. As Jersild sums it up, a “large proportion of fears that were described have little or no direct relation to misfortunes that actually have befallen the children.”¹²

The above data are puzzling. One would expect the child to fear what actually had given him trouble. Noting that the “imaginary fears” of the child increase with age, Jersild suggests as one explanation the fact that “imaginative capacities” of the child are developing. Such maturing capacities may explain why the child deals in imaginary material. But in my judgment this does not adequately account for the fact that these imaginary things should, in such large measure, be feared.

A second problem raised by Jersild’s data is the unpredictability of fears. Jersild notes that his data emphasize how difficult it is to predict when a child will be afraid:

the same child may face a given situation at a certain time without showing fear but at a later time, with no grossly apparent intervening causal factors, the same situation gives rise to fear. . . . A certain noise causes fear and another does not; a child is taken to one strange place and exhibits no fear, while in another strange situation he does show fear.¹³

It is significant that “fear of strange persons” is most often mentioned as a fear that appears in certain situations and not in other similar situations. Since a child’s fears shift unpredictably, some process is occurring that is vastly more complicated than the usual concept of conditioning. But the question of what this process is remains unanswered.

Fears Masking Anxiety

I submit that these two problems—the irrational and unpredictable qualities of children’s fears—become intelligible if many of these so-called “fears” are regarded not as specific fright reactions (that is, fears as such) but rather as appearances in objectivated form of underlying anxiety. Fear is regularly defined as a specific reaction, and it is apparent that something is occurring in these “fears” which cannot be explained as a specific response inseparably and intrinsically related to a specific stimulus. If the hypothesis is made that these fears are rather expressions of anxiety, the high percentage of “imaginary” fears becomes understandable. It is well known that the anxiety of children (as well as of adults) is often displaced upon ghosts, witches, and other objects which do not have a specific relation to the child’s objective world but do fulfill significant functions for his subjective needs, specifically in relation to his parents. In other words, fears may mask anxiety.

In some cases, for example, the process is to be understood as follows: the child feels anxiety about his relation to his parents. He cannot face this directly, in such terms, for example, as “I am afraid my mother does not love me,” for such a realization would markedly increase his anxiety. Or he is aided in covering up the direct form of the anxiety by reassurances from the parents, which often have little to do with the real fulcrum of anxiety. The anxiety is then displaced upon “imaginary” objects. I have placed the term “imaginary” in quotation marks frequently in this immediate discussion because it would no doubt be found on more profound analysis of the mysterious fears, that the imagined object stands for something all too real in the child’s experience. Adults, of course, likewise engage in a similar pattern of displacement of anxieties, but adults are much more skilled in rationalizing the anxieties so that the objects become more apparently “logical” or “reasonable”.

Our hypothesis, that these fears are the expression of underlying anxiety, would also make clear why the fears of animals were not of animals actually encountered by the child but remote ones like gorillas and lions. Fears of animals are often projections of anxieties the child feels in relations with objects or persons (such as parents) who are by no means remote. Freud’s case of Little Hans is a classic example.14 I here suggest that fears of animals may also be projections of the child’s own hostile feelings toward members of his family, feelings which entail anxiety because of the probability of punishment or disapproval if these hostile tendencies were carried into action.

Our hypothesis likewise throws light upon the unpredictability and shifting quality of children’s fears. If these fears are appearances in objectivated form of underlying anxiety, the anxiety could focus now on this object and now on that. What appears as inconsistency on superficial analysis would then be quite consistent when viewed on a deeper level. Jersild himself notes this shifting character of fears when the fears are an expression of underlying anxiety:

As long as there are underlying difficulties that press upon the child from many sides, the elimination of one particular expression of fear may shortly be followed by other fears of a slightly different cast.¹⁵

Several years after the first edition of this book was published, Jersild remarked to me in personal conversation that he agreed with my conclusion that these fears really expressed anxiety. He was surprised that he had never seen this earlier. I think his not seeing it shows how hard it is to get out of our traditional ways of thinking.

Another indication that anxiety rather than fear is occurring in many children is in the observation that studies indicate the frequent ineffectiveness of verbal reassurance in overcoming (“in contrast to covering up”) children’s fears. As Goldstein held, if the emotion is a specific fear, it can normally be allayed by verbal reassurance. If the child who has a fear that the house is on fire receives adequate demonstration that such is not the case, the fear vanishes. But if this apprehension is actually an objectivated form of anxiety, either the terror will not be allayed, or it will shift to a new object.

There is indirect presumptive support for the hypothesis that anxiety often underlies children’s fears in the close relationship between children’s “fears” and those of their parents. Hagman’s study yields a correlation of .667 between the gross number of children’s fears and the gross number of mothers’ fears.16 Jersild found “a good deal of correspondence between the frequency of fears of children of the same family; the correlations ranged from .65 to .74.”¹⁷ It would seem that something more is occurring than the parents’ fears having an “influence,” as Jersild puts it, upon the fears of the children—i.e., that the child learns to fear certain things because the parents do. It has been so frequently pointed out as to become a platitude that the development of anxiety in children arises centrally out of their relations with their parents.¹⁸

I suggest that the close relationship between children’s fears and those of their parents, and between fears of children of the same family, can be understood more clearly as a carry-over on the anxiety level. In other words, in families in which the parents have a good deal of anxiety, the interpersonal relations with children will be disrupted by this anxiety, and there will be greater anxiety (i.e., increased tendency to have “fears”) on the part of the children.

The purpose of this discussion of children’s fears, in addition to throwing light upon the specific problem of fear as such, has been to demonstrate that the study of fears leads inexorably to the study of anxiety. The particular hypothesis suggested here is that many children’s fears are objectivated forms of underlying anxiety.¹⁹

A NOTE ON STRESS AND ANXIETY

Hans Selye’s first book, Stress, was published, interestingly enough, the same year as the first edition of my Meaning of Anxiety, 1950, the exact middle of the century. This marked the beginning of the extensive concern with stress in psychological and medical thinking. In a book six years later, Selye defined biological stress as an “adjustment through the development of an antagonism between an aggressor and the resistance offered to it by the body.” Stress is the response to “the rate of wear and tear in the body.”20

He proposed a General Adaptation Syndrome (called G.A.S.). The G.A.S., through our various internal organs (endocrine glands and nervous system), helps to adjust us to the constant changes which occur in and around us. “The secret of health and happiness lies in successful adjustment to the ever-changing conditions on this globe; the penalties for failure in this great process of adaptation are disease and unhappiness.”²¹ Each person, he believes, is born with a certain amount of adaptation energy.²²

This may well be true physiologically, but I question it psychologically. Is not energy partially a product of the zest and commitment of the person for the task at hand? Do we not now find, in our studies of gerontology, that persons grow senile not wholly as a function of age but also because psychologically they have nothing to interest them? And that the brain depends considerably on commitment to tasks which call forth zest, if it is to retain its energy?

A statement needs to be made about the tendency among psychologists to use the word “stress” as a synonym for anxiety. Books purportedly on anxiety use the word “stress” instead; conferences seemingly on anxiety now are entitled interchangeably with the term “stress.” I will here argue against the identification of stress and anxiety, and will hold that stress is not an adequate substitute for describing the apprehension we ordinarily refer to as anxiety. This is not an argument against the classical work of Selye, whose field and insightful work is in experimental medicine and surgery. In his field the term surely fits; in psychology I do not believe “stress” encompasses the rich meaning of anxiety.

The word stress is borrowed from engineering and physics. It seems to have become popular in psychology because it can be defined readily, handled easily, and generally measured satisfactorily, all of which are difficult with the term “anxiety.” It would seem relatively easy to define the point at which a person will break under stress. Our culture is obviously one which subjects its citizens to great and increasing stress, arising from the radical changes in technology, loss of values, and so on. These are obviously related to the prevalence of stress disorders like cardiac problems, arteriosclerosis, ad infinitum. No cocktail party occurs these days without people talking about stress and the ill effects it causes. “Mental stress” has become an acceptable term even though it appears only eighth on the list of meanings of “stress” in my dictionary.

Psychologically speaking, how the person interprets the threat is crucial. Aaron Beck has pointed out that stressful life situations per se are less important in the production of anxiety than the way in which these situations are perceived by the individual.23 As Bourne, Rose, and Mason also say in their article on anxiety in combat soldiers in Vietnam (in this case, operators of helicopters), flying, or even death, cannot be interpreted as stress without consideration of the manner in which each individual perceives the threat.24 “Perceive” and “interpret” are subjective processes, rightly included in anxiety but not in stress.

If, furthermore, we use stress as a synonym for anxiety, we cannot distinguish between the different emotions. Protracted anger or protracted guilt-feeling causes stress as much as protracted fear; we blur this distinction if we use stress as the catch-all term. We also cannot distinguish between fear and anxiety. In the case of Tom, as we saw above (p. 76–77), when Tom was responding with fear, (i.e., after he had misplaced some valuable papers in the doctor’s laboratory), his gastric readings were very low. The stomach, so to say, shut off. Whereas when he responded with anxiety, as when he lay awake all night worrying about how long his job would continue, his gastric readings were the highest of any time in the study. Opposite to how it functioned in fear, the stomach now worked overtime. If we lump both of these together under “stress,” we will have lost a crucial distinction for our understanding.

No matter how strongly Selye argues against it in his recent book, his earlier statement, “any stress produces damage” is taken, it seems in America, as meaning that all stress should be avoided. Or at least as often as possible. Selye saw this problem when he dedicated one of his books to those “who are not afraid to enjoy the stress of a full life, nor too naïve to think that they can do so without intellectual effort.”²⁵ But let us remind ourselves of the remark attributed to Hudson Hoagland, “Getting up in the morning is a great cause of stress.” Yet we don’t stay in bed.

Furthermore, let us recall that additional stress may also bring a great relief from anxiety. In wartime Great Britain, amid the bombing, painful austerity, and other great stress, there was a clear diminution of neurosis.²⁶ This situation has been demonstrated in many countries. The neurotic problems are allayed in times of stress because the persons have something definite on which to pin their inner turmoil, and they can thus focus on concrete pressures. Indeed, in such cases stress and anxiety work directly opposite to each other. When there is great stress there may be freedom from anxiety.

Finally, we can see the inadequacy of the term “stress” used as a synonym for anxiety in Lidell’s statement that “anxiety accompanies intellect as its shadow, and the more we know of the nature of anxiety, the more we will know of intellect.” If we say “stress accompanies intellect as its shadow,” we do not make sense. The same is true when we consider the basic statement made by Kubie: “Anxiety precedes the development of thought.” To say “Stress precedes the development of thought” does not at all do justice to Kubie’s meaning, i.e., that the gap between stimulus and response, self and object makes thought necessary. “Stress” is, as Selye used it, chiefly a physiological term.

RECENT RESEARCH ON ANXIETY27

During the last two decades, thousands of articles, plus a deluge of dissertations, have appeared in the literature on anxiety and stress. Charles Spielberger’s indefatigable efforts to bring together major contributors to the field in several symposia led to the publication of no less than seven volumes of research.²⁸ Although these studies have contributed to our understanding of aspects of anxiety, they create an even greater need for an integrated theory of the meaning of anxiety. I cannot hope to do justice to all of these studies. Instead, I trust the reader will permit me to mention a few of the inquiries that seem most significant to me. I do this, with appropriate anxiety, as an exercise in moving ahead in spite of the fact that one cannot cover everything.

Four areas of current research stand out as leading to new understanding of what contributes to anxiety in human beings. First, cognitive theorists, such as Richard Lazarus and James Averill29 and Seymour Epstein,³⁰ concerned about the person’s perception of reality, believe that an individual’s appraisal of the threat serves as a key to understanding anxiety. The significance of these studies is that cognitive theorists place man-as-perceiver at the center of anxiety theory. Although Lazarus and Averill describe anxiety as an emotion based on cognitive mediators between the situation and the personal response, they emphasize that anxiety does not arise from pathology, but humanity. Much of their work, however, seems to be on the effects of psychological stress rather than anxiety.³¹ Epstein describes expectancy as the basic parameter in determining the level of arousal, defining anxiety as “acutely unpleasant diffuse arousal following the perception of threat.” Anxiety is seen as unresolved fear, resulting in diffusion of the threat. Epstein and Fenz³² studied sport parachutists, finding that experienced parachutists had high focused arousal, which served to expand their awareness prior to the jump. Novices, on the other hand, had a defensiveness against stimulation which was experienced as aversive, causing them to be overwhelmed by jumping. Epstein’s most interesting finding seems to us to be the linkage of anxiety to low self-esteem.”³³ Similar to Goldstein’s view of the “catastrophic situation,” Epstein states that “people have an integrated self theory subject to collapse.”³⁴ Acute psychotic reactions can facilitate the development of an individual’s new, more effective self theory. Epstein continues: “acute anxiety is produced by threats to the integrative capacity of the self system.” A person with low-self-esteem collapses more easily than a person of high self-esteem. Epstein elaborates: “Increases in self esteem produce increases in feelings of happiness, integration, energy, availability, freedom and expansiveness. Decreases in self esteem produce increases in feelings of unhappiness, disorganization, anxiety and constriction.”35

A second area of significant current research is Spielberger’s differentiation between “state” and “trait” anxiety. This has inspired literally hundreds of studies. He sees state anxiety as a transitory emotional condition associated with autonomic-nervous-system activity. Trait anxiety is anxiety proneness, or frequency of the manifestation of anxiety over a long period of time.³⁶ This model has been used by many researchers to differentiate between arousal and underlying anxiety. Those experiences which have most influence on raising the level of trait anxiety probably date back to childhood, Spielberger believes, and involve parent-child relations in which the child is punished. This points in the direction of the thesis of my own research in Chapter 9, that anxiety proneness has its roots in maternal rejection. Norman Endler believes that both trait and state anxiety are multidimensional, posing a Person-Situation-Interaction model for anxiety. He sees anxiety an an interaction between interpersonal or ego threat (situational factor) and level of interpersonal A-Trait (personality factor).³⁷

A third area of current research, that on the relationship between anxiety and fear, is one which has led to many debates among theorists. Conditioning theorists, equating anxiety with fear, have developed various systems of behavior therapy based upon learning theory. It should be noted that their greatest success is with phobics. But a phobia is, by definition, a crystallization of anxiety around some external event, and is by common consent, a neurotic fear covering up anxiety. (See case of Hans in Chapter 5). It is not difficult to change the focus of fear. But to deal with the underlying anxiety seems to be avoided in the strict behaviorist technique. My point of view is close to that of H. D. Kimmel who criticizes the behaviorists who equate anxiety with fear. Kimmel believes that Pavlov’s “experimental neurosis” should have been termed anxiety.³⁸ He maintains that conditioned fear cannot serve as a model for the acquisition of anxiety because it is built upon the principle of certainty, whereas anxiety has as its core uncertainty and lack of controllability.

Another of the great contributions that researchers have made to our understanding of anxiety has come from studies of individuals in actual life situations. Yona Teichman, studying the responses of family members of soldiers missing in action in the 1973 Middle East War, demonstrated that parents, wives and children have different styles of coping with personal loss. Parents typically maintained a highly personal experience of grief, at first refusing to share it with others. Most responses centered around their need for personal courage and their feelings of bitterness. Despite extreme withdrawal for a week or so, no perpetuation into pathological, long-lasting withdrawal was reported. Wives, though similar to the parents in desiring to be strong, maintained a milder form of bitterness. Typically they were preoccupied with practical problems and dependent on supporting figures. Children, on the other hand, reacted to the general stress of the home rather than the specific loss. Since children did not express their grief continuously, they received parental hostility because of their “unfeeling” attitude.39 These role comparisons are interesting to view in the light of Lifton’s Protean Man.⁴⁰ Charles Ford’s descriptions of the anxieties of those involved in the Korean Pueblo incident have demonstrated that men who maintained a faith in their commanding officer, religion, or country coped much more successfully with the anxieties of their incarceration. More than half the men reported significant anxiety from the unpredictability of their treatment. Ford concluded that massive repression was evoked as an acute defense mechanism. More significantly, however, was the finding that the long-range psychological response to severe anxiety may be much greater than the acute response.⁴¹ Richard Lynn’s cross-cultural studies of national differences in anxiety, used increased alcohol consumption, rising suicide rates, and the frequency of accidents as some indicators of anxiety.⁴²

Observations of life event changes and anxiety in mental health demonstrate that any change in one’s familiar life pattern, for better or worse, calls for adaptation, often mobilizing anxiety.⁴³

I hope that these inquiries into the cognitive and multidimensional studies of living people in crisis may help us appreciate how protean are the aspects of anxiety.

ANXIETY AND LEARNING THEORY

O. Hobart Mowrer’s work is taken for our chief emphasis here because he spans, in his own metamorphosis, the gamut of schools of psychology. Beginning as a staunch behaviorist, Mowrer contributed some of the best formulations of his time of anxiety in stimulus-response psychology. (He is still quoted by Eysenck, who is apparently unaware of Mowrer’s later changes.) This led Mowrer to learning theory, the area in which many psychologists find his most significant contribution. Learning theory was a bridge for him to clinical psychology via the problem of how and why rats learn delinquent behavior. This, in turn, led him to a concern with the problems of time, symbols and ethics. The last underlie his preoccupation, in the latter part of his writings, with guilt and responsibility and their implications for therapy. Such a radical metamorphosis must indeed have been difficult to assimilate. This is one reason Mowrer’s work is especially revealing.

If we express this in another way, Mowrer’s first stage is behaviorism, the second is anxiety and learning theory, and the third is his concern with guilt feeling and its implications for psychology. The changes in his concerns reflect several significant levels in the broadening approach to anxiety in this country. The material reviewed here is mainly in his second period.

Mowrer’s analyses of anxiety, which interest us here, are based centrally upon his researches in learning theory. In view of the frequently proposed assumption that the ultimate bridge between psychoanalysis on one hand and experimental and academic psychology on the other will probably be learning theory, it is presumable that the learning theory base of Mowrer’s work gives his conceptualizations of anxiety greater cogency.

In his early stimulus-response formulations, written when he was still a behaviorist, Mowrer explicitly characterized anxiety as a “psychological problem to which the habits known as ‘symptoms’ provide solutions.”⁴⁴ Anxiety was defined in his first paper as “the conditioned form of the pain reacton.”⁴⁵ That is to say, the organism perceives the danger signal (the stimulus), and the conditioned response which then follows in anticipation of the danger—a response characterized by tension, organic discomfort, and pain—is anxiety. Any behavior which reduces this anxiety is rewarding, and hence, by the law of effect, such behavior becomes “stamped in,” i.e., learned. This analysis has two important implications. First, anxiety is seen as one of the central motivations of human behavior. And, second, as a corollary, the process by which neurotic symptoms are acquired is placed squarely on a basis of learning theory—symptoms are learned because they are anxiety-reducing.

Mowrer’s next researches in anxiety were experiments with rats and guinea pigs, verifying the above hypothesis that reduction in anxiety serves as a reward and is positively correlated with learning.46 This hypothesis is now widely accepted in the psychology of learning.⁴⁷ It has the practical merit not only of emphasizing how important and pervasive anxiety is as a motivation in education, but it also sheds light on healthy and constructive methods of management of that anxiety in the classroom.⁴⁸

These early approaches of Mowrer to the problem of anxiety have two elements of definition in common. First, no specific distinction is made between fear and anxiety. In the first paper the terms are used synonymously, and in the second the factor of anxiety is defined as the animals’ expectation of the electric shock—a state that could be termed fear as accurately as, if not more accurately than, anxiety.⁴⁹ Second, the threat which cues off anxiety is defined as the threat of organic pain and discomfort. Obviously, during the period when these papers were written, Mowrer was endeavoring to define anxiety in physiological terms.⁵⁰

But radical changes occurred in Mowrer’s conception of anxiety following his further researches in learning theory. The changes came particularly after his inquiry into the question: Why do people learn nonintegrative (“neurotic,” consistently punishing) behavior? Experimenting with animals, he demonstrated that rats exhibited “neurotic” and “criminal” behavior because they were incapable of anticipating future, long-time rewards and punishments and balancing them against the immediate consequences of their behavior.⁵¹

In his stimulating discussion of the findings of his many studies, Mowrer concludes that the essence of integrative behavior is the capacity to bring the future into the psychological present. Human beings have this capacity for integrative learning in a form vastly different from animals because they are able to bring the “time determinant” into learning, to weigh future against immediate consequences. This gives human behavior flexibility and freedom and, by inference, responsibility. Mowrer refers to Goldstein’s observations that the most characteristic loss of patients with cortical injuries was the capacity to “transcend concrete (immediate) experience,” to abstract, to deal with the “possible.” Hence these patients were limited to rigid, inflexible behavior. Since the cortex represents the distinctive neurological difference between man and animals, this capacity which the cortically impaired patients lost may be assumed to be the distinctively human capacity.

Mowrer’s findings imply a new emphasis on the historical nature of man, on the human being as the “time-binding” being.52 As he phrases it,

the capacity, then, to bring the past into the present as a part of the total causal nexus in which living organisms behave (act and react) is the essence of “mind” and “personality” alike.⁵³

To be sure, the importance of the individual’s own genetic past—the fact that he carries into the present the experiences from his childhood, for example—has long been generally accepted in clinical psychology. But there is another implication of this emphasis on man as the “time-binding” being which is relatively new in clinical work: namely, since the human being weighs his behavior in terms of symbols which have been developing through many centuries in the history of his culture, he can be understood only in the context of that history. These findings meant for Mowrer a new interest in history in general, and in particular a new interest in ethics and religion, which are the history of the human endeavor to transcend immediate consequences in terms of long-time universal values.

Through his discussion of integrative learning, Mowrer makes an exceedingly useful distinction between the terms integrative and adjustive. All learned behavior is in some sense adjustive: neuroses are adjustive; defense mechanisms are ways of adjusting to difficult situations. Mowrer’s “neurotic” rats gave up taking food, and his “criminal” rats took it despite the future punishment, each group “adjusting” to a difficult situation. But neuroses and defenses, like the behavior of these rats, are not integrative in the sense of preparing the person for future learning. Neuroses and defenses do not permit the further constructive development of the individual.54

The implications of the above considerations for anxiety theory are profound. The problem of neurotic anxiety is placed squarely in its cultural and historical nexus, and is related specifically to man’s distinctive problems of social responsibility and ethics. This is in radical contrast to Mowrer’s previous definition of anxiety as response to the threat of organic pain or discomfort. For Mowrer now the “social dilemma [as illustrated in the child’s ambivalent relation to his parents] is a precondition for anxiety.”⁵⁵ If animals have neurotic anxiety at all, Mowrer holds, it is only in artificial environments (e.g., the “experimental neurosis”) in which they have become to some extent domesticated, “socialized.” That is, by virtue of their relationship with the experimenters, the animals have become something more than “just” animals. This does not imply, in Mowrer’s writings or in my own attitude any depreciation of the value of animal experiments or laboratory studies of human beings, but it does place such methods of study in perspective. In the study of neurotic anxiety, we find the essence of our problem in precisely those characteristics which distinguish human beings from animals. If we limit ourselves to the areas in human behavior which are identical with the infrahuman or to those elements which can be isolated in the laboratory, or, indeed, if we center our study around the strictly biological and organic impulses and needs of man, the essential significance of anxiety for human beings will elude us.

We shall now turn to Mowrer’s later presentation of his concept of anxiety. He notes that the beginning of the “social dilemma” is in the child’s early relations with his parents. The child cannot avoid anxiety cued off in the family situation by simple flight (like the animal in nature), for the anxious child is dependent upon his parents at the same time that he fears them. Mowrer agrees with the Freudian theory that repression occurs in the child because of real fears—generally the fear of punishment or deprivation (withdrawal of love). Indeed, Mowrer wholly accepts Freud’s description of the mechanism by which anxiety occurs: a real fear→repression of this fear→neurotic anxiety→symptom formation as a solution to the anxiety. But mechanism is a different thing from meaning. Mowrer contends that Freud “never succeeded in fully comprehending the essential nature of anxiety itself”56 because of his endeavor to explain anxiety in terms of instincts and his failure to understand the social context of personality. In the maturing of the human individual, social responsibility normally becomes (or should become) a positive, constructive goal. By and large Mowrer holds, the conflicts which are most likely to cause anxiety are of an ethical nature—a point seen by Kierkegaard but not by Freud. The “ethical accomplishment of untold past generations,” writes Mowrer, “as imbedded in the conscience of modern men and women, is not a stupid, malevolent, archaic incubus, but a challenge and a guide for the individual in his quest for self-fulfillment and harmonious integration.”⁵⁷ The sources of the conflicts are social fear and guilt. What the individual fears is social punishment and withdrawal of love and approval on the part of the significant other persons in his constellation of relationships. It is these fears and the guilt associated with them which become repressed. In their repressed state they become neurotic anxiety.

Anxiety is a product, states Mowrer “not of too little self-indulgence and satisfaction . . . but of irresponsibility, guilt, immaturity.” It arises from “repudiated moral urgings,”58 or, in Freudian terminology, anxiety is caused by “repression of the super-ego,” not, as Freud would have it, the reverse. This viewpoint, of course, has radical implications for dealing with anxiety in therapy. Mowrer points out that the endeavor of many psychoanalysts to dilute and “analyze away” the superego (and concomitantly the individual’s sense of responsibility and guilt) only too often results in a “ ‘deep narcissistic degression’ rather than in the growth in personal maturity, social adequacy, and happiness which one has a right to expect from a really competent therapy.”⁵⁹

One of the significant implications of Mowrer’s viewpoint is that anxiety is seen as playing a constructive, positive role in human development. He writes:

There is a common tendency in our day, both on the part of professional psychologists and laymen, to look upon anxiety as a negative, destructive, “abnormal” experience, one which must be fought and if possible annihilated. . . . Anxiety, as conceived in these pages, is not the cause of personal disorganization; rather is it the outcome or expression of such a state. The element of disorganization enters with the act of dissociation or repression, and anxiety represents not only an attempted return of the repressed but also a striving on the part of the total personality toward a re-establishment of unity, harmony, oneness, “health.”⁶⁰

And, again,

Nothing could be truer in the light of my own clinical, as well as personal, experience than the proposition that psychotherapy must involve acceptance of the essential friendliness and helpfulness of anxiety, which, under such treatment, will eventually again become ordinary guilt and moral fear, to which realistic re-adjustments and new learning can occur.61

PERSONAL COMMENTS

As I have been working on this chapter, some ideas have been running through my mind that I thought would be valuable to share. The first has to do with the problem of experimentally induced anxiety.

In a paper in 1950 Mowrer remarked, “There is at present no experimental psychology of anxiety and one may even doubt whether there ever will be.”⁶² Not only was it true that the problem of human anxiety had been absent from the strictly experimental specialties of psychology, but until the 1950s it had largely been omitted from the other branches of academic and theoretical psychology as well. In searching psychological books written earlier than 1950 (except those concerned with psychoanalysis), it is almost impossible to find the topic of anxiety even listed in the indexes. What Kierkegaard wrote over a hundred years ago remained true in the first half of the twentieth century: “One almost never sees the concept of anxiety dealt with in psychology.”⁶³ True, there had been a plenitude of studies of fears in the experimental and academic psychology of this century, for fears can be made specific and can be counted. But on the threshold where the problem of fears moves into the problem of anxiety, the psychological inquiries had halted.

The practical reasons for this, Mowrer believed, were that the effects of inducing anxiety experimentally in the laboratory would be too damaging. But Mowrer either underestimated the ingenuity of psychologists (and their personal defense mechanisms) or overestimated the human sensitivity of some of them. In any case, of the thousands of researches on anxiety which have sprung up since 1950, many utilize some form of experimentally induced anxiety, generally with students as the subjects.

When my colleague and I went through many of these, we found that some psychologists induce anxiety by setting up an experiment where the student is threatened by shock and others capitalize on the threat of failure. It turned out that threat of failure was definitely more potent in getting the desired reaction from students, so that the majority of later studies induced anxiety by threatening failure. The typical set-up would be roughly as follows: the student comes in with an attitude of respect and trust toward the experimenter, perceiving the latter as a representative of a reputable science. He has heard thousands of times that science will save us, and he is prepared to do his bit. The student would be given several questions to answer. He would then be told, regardless of his responses to the questions, “You have not performed well on this task” or “Your responses are not adequate.” Sometimes the student is enticed in under the guise of getting counseling; he is given the Rorschach and then told, again regardless of his responses, “Your responses are similar to sixty percent of emotionally disturbed people” or “You show by this test that you don’t have the ability to succeed at this university.” The overall aim is to break down the student’s self-esteem and then to record the anxiety.

If one identifies with the students who are the victims of these deceptions, one could imagine several reactions. The first would be the trusting student who has absorbed the lesson of his culture that one should believe everything told to him by a person in authority. His self-esteem would, as expected, take a nose-dive. (And it is the height of navïeté to believe that this can be corrected later by explaining to the student that he was deceived.) Or the student might be the sophisticated type who knows that we all survive by lying to each other anyway. He would be partially protected by his suspiciousness; he would certainly be confirmed in his cynical view of the world, and the general climate of suspiciousness in which he lives would also be confirmed. He would wonder how graduate students and professors can actually think that the subjects believe those lies anyway.

But it totally omits the major ethical issue, which is the matter of lying to the subjects and thinking the “explaining” of it afterward clears everything up. I believe that such experiments take their place with lobotomies and shock treatments as practices any profession worthy of the name should control among its members.

WHATEVER ONE’S ethical stand on the above questions may be, several facts stand out clearly in any survey of psychological studies of human anxiety. One is that the most fruitful researches for illuminating this area have been those which employed clinical procedures along with experimental techniques. For example, the studies of ulcer patients and the case of Tom, reviewed in the preceding chapter.64 This includes taking one’s subjects from those populations which already have anxiety in their life situation. Irving Janis did this in studying anxiety and stress in pre-operative patients in hospitals. Other groups which illustrate life-situation anxiety are combat soldiers, unmarried mothers, parachutists, or test-anxious school children. Obviously, then one can use experimental techniques in assessing the phenomena without themselves inducing the anxiety in the subjects.

Another fact that stands out is that the experimental academic psychologists who have pertinently attacked the problem of anxiety have been those who were led to it via their increasing interest in clinical work and who adopted clinical techniques as their method, like O. Hobart Mowrer, Irving Janis, and John Mason.

The third fact which stands out clearly is that most of the significant data on anxiety comes from the psychotherapists—Freud, Rank, Adler, Sullivan, and others—whose clinical methods permit an intensive study of the subjective dynamics and whose focus of attention is by definition on the individual as a totality confronting crises in his life situation.

My other comments have to do with certain curious anxiety phenomena which have appeared in my own psychotherapeutic practice and are inexplicable on the basis of the classical psychoanalytic doctrine of anxiety. I noticed that some patients did not repress their sexual, aggressive, or “antisocial” urges (in Freud’s sense) in any discernible way. Instead they repressed their needs and desires to have responsible, friendly, and charitable relations with other people. When aggressive, sexual, or other behavior in egocentric form emerged in the analysis, these patients showed no anxiety. But when the opposite needs and desires emerged—i.e. to have responsible and constructive social relations—there appeared much anxiety, accompanied by the typical reactions of patients who feel a crucial psychological strategy to be threatened. Such repression of constructive social urges occurs particularly, and understandably, with defiant, aggressive types of patients. (In Greek terms, this is repression of love in agape, rather than the libido, sense.)

While it is true that many persons have guilt and anxiety because of fear of expressing their own individual capacities and urges, sexual or otherwise, as Freud originally indicated, it is at the same time true that many have guilt and anxiety because they have become “autonomous” without becoming “responsible”65

This is repression of what Alfred Adler termed “social interest.” Adler’s viewpoint has the merit of emphasizing a profoundly important point—namely that the needs of the human being to be a responsible social creature are as fundamental as his needs to express his individualisic, egoistic urgings. It might be argued that urgings for self-gratification are more primal than those for social interest and generosity, since the latter develop at a later stage in the child. But we are learning in recent years the implications of the facts that each human being was born in a dyad, namely he and his mother, and that the fetus was gestated in the womb for nine months. Individualism comes after community ties. There is the fact that the human being is bound by social ties from its fetal, in utero stage onward (as Sullivan points out), regardless of whether awareness of social ties and their meaning emerges into consciousness sooner or later.

This adds up to a general agreement with the idea espoused by Mowrer, that we have overlooked the function of guilt and social responsibility as a cause of anxiety in our culture. A revealing demonstration of this is found in the case of Helen (Chapter 9), who would admit none of her extensive guilt feelings about her extramarital pregnancy because such an admission would conflict with her aim of being a “rationally” emancipated person. As a consequence her strong feelings of anxiety likewise remained repressed and unamenable to therapy. It would seem that the repression of guilt feelings, with its concomitant generation of neurotic anxiety, is a prevalent characteristic of certain groups in our culture and in some ways it pervades our culture as a whole.

It is surely true, also, that many patients carry a heavy burden of irrational guilt and anxiety which is not a product of their own irresponsibility. In my experience, borderline psychotics fit this category most dramatically. This irrational guilt certainly needs to be clarified and relived in any adequate therapy. But there are other patients with whom, when guilt feeling is reduced in therapy by the endeavors of the analyst, it eventuates that the genuine, if confused, insights of the patient into himself have been violated and obscured; and that the most valuable and objectively accurate motivation for change is lost. I have known of cases in analysis which have been unsuccessful precisely because the analyst joined the patient in diluting and depreciating guilt feelings. Temporary allaying of anxiety was achieved, of course, but the problems which underlay the anxiety were unsolved and only became buried under a more complicated system of repression.