Harry Stack Sullivan – The Interpersonal Theory of Psychiatry

The Interpersonal Theory of Psychiatry

For 2 years, Harry Stack Sullivan was an honorary member of Karen Horney’s new psychoanalytic institute. However, he was angered by the seemingly arbitrary expulsion of Erich Fromm (Chapter 6). Sullivan therefore resigned from the institute in April 1943, arguing that it is wrong to attack the integrity and judgment of gifted colleagues who prefer the path of innovator and critic.

In the course of defending Fromm, Sullivan also evoked the wrath of the Freudians. He took exception to a scathing review of Fromm’s classic Escape From Freedom by the psychoanalyst Otto Fenichel, which concluded with the claim that only the true faith—Freudian psychoanalysis—was “pure gold.” Sullivan contended that this review was a pro-Freudian diatribe lacking any substantive quality, whose purpose was primarily political. Once again, orthodox psychoanalysis responded to challenge by imposing the penalty of excommunication. Many analysts were asked (in effect) to choose between Freud’s beliefs and associating with Sullivan, and those who selected the latter alternative were subjected to various forms of professional intimidation. This political rivalry grew so intense that even today there are psychoanalysts who have adopted important Sullivanian theories yet who steadfastly refuse to credit him accordingly. (See Perry, 1982, pp. 386–389.) Nevertheless, Sullivan ultimately emerged from these professional difficulties as one of the leading figures in the realm of personality theory.


  • To devise a theory that retains the importance of the unconscious but stresses the social determinants of personality, notably the parents but also schoolmates, friends, and love interests, rather than instincts.

  • To dispense with virtually all Freudian constructs, including libido, and use his own unique terminology.

  • To correct Freud’s pessimistic view of human nature by showing that we have an inherent drive toward mental health.

  • To show that personality development proceeds through a series of seven stages from infancy to adulthood, rather than ending at age six as Freud contended.

  • To show that intense anxiety is an extremely important aspect of psychopathology, and that anxiety in the mother (or whichever adult fulfills this function) is its primary cause.

  • To emphasize that we need other people, and denying this need is also a major cause of psychopathology.

  • To explain the causes and dynamics of schizophrenia.

  • To devise improved methods for treating schizophrenia and other psychological disorders.


Harry Stack Sullivan was born on February 21, 1892, in Norwich, New York. He was the only surviving child of a taciturn father, a farmer and skilled workman, and a mother who “never troubled to notice the characteristics of the child she had brought forth.… ‘Her son’ was so different from me that I felt she had no use for me, except as a clotheshorse on which to hang an elaborate pattern of illusions” (Sullivan, 1942). Partly because the Sullivans were the only Catholic family in a Protestant community, Harry had a lonely childhood. This helped him develop an unusual empathy for the intense isolation of the schizophrenic, together with a rather withdrawn personality of his own.

Sullivan encountered significant personal problems during his freshman year at Cornell University. He became the cat’s-paw for a gang of boys in the dormitory, and shouldered the blame for some illegali- ties engineered by the group. He is also believed to have undergone a schizophrenic breakdown at about this time. (See Perry, 1982, pp. 3, 143–146, 151.)

Sullivan never returned to Cornell after his one-semester suspension. In 1911 he entered the Chicago College of Medicine and Surgery, the medical branch of Valparaiso University of Indiana. His grades were erratic; but he completed his course work in 1915, and ultimately received the M.D. degree from Valparaiso in 1917 (Perry, 1982, pp. 156–159, 165).

Sullivan demonstrated considerable skill as an internist, but he preferred a career in psychiatry. He therefore entered psychoanalytic therapy as a patient during 1916–1917. After serving in World War I as a first lieutenant in the Medical Corps, Sullivan worked at government and private hospitals in Maryland and Washington, D.C. Here he began his intensive studies of schizophrenia; came under the influence of William Alanson White, who later founded a prestigious psychiatric foundation and named Sullivan its president in 1933; and gradually modified Freudian psychoanalysis to suit his own theoretical predilections.

In 1931, Sullivan moved to New York City and pursued further psychoanalytic training. He suffered financial problems that forced him to file for bankruptcy, but ultimately established a lucrative private practice. During World War II, he served as consultant to the newly formed Selective Service System, and subsequently participated in UNESCO and other world health projects.

Sullivan was a lifelong bachelor. In 1927 he began a close relationship with a young man he describes as a former patient, James Inscoe. “Jimmie” lived with Sullivan for some 20 years as a “beloved foster son … friend and ward,” though he was never legally adopted. (See Perry, 1982, pp. 209–210; Sullivan, 1942).

Sullivan’s writings fill seven volumes, only one of which he completed himself (1932–1933/1972). Five were published posthumously, and consist of edited lectures. Harry Stack Sullivan died of a cerebral hemorrhage in Paris on January 14, 1949, while returning home from a mental health conference in Amsterdam.


Like Horney and Fromm, Sullivan emphasizes the interpersonal nature of personality. But whereas Horney concentrates on neurosis, and Fromm stresses the pathogenic role of society, Sullivan is primarily concerned with two other important areas: the development of personality, and the dynamics and treat- ment of schizophrenia. “If we go with almost microscopic care over how everybody comes to be what he is at chronologic adulthood, then perhaps we can learn a good deal of what is highly probable about living and difficulties in living” (Sullivan, 1953/1968, p. 4).

The One-Genus Postulate

Like Fromm, Sullivan makes some allowances for the effects of heredity on personality. We are all influenced by such physiological motives as hunger, thirst, respiration, sexuality, and the maintenance of body temperature. This animalistic aspect of personality accounts for individual differences in physical characteristics, sensory abilities, intelligence, and the rate at which we mature.

However, Sullivan also concludes that the similarities among human personalities far exceed the differences (the One-Genus Postulate). Even the most retarded individual differs far less from the greatest genius than from any member of any other species. Thus Sullivan (unlike Adler) prefers to minimize the importance of individual differences, and to devote his theoretical attention to those phenomena that humans have in common. (See Sullivan, 1947/1953, p. 16; 1953/1968, pp. 32–33.)

The Need for Others

Sullivan shares Fromm’s and Horney’s distaste for Freudian libido theory, arguing that it is “completely preposterous” to assume that our behavior is rigidly determined by instincts. Except for such hereditary disasters as congenital idiocy, human nature is extremely pliable and adaptive. “[Even] the most fantastic social rules and regulations [could] be lived up to, if they were properly inculcated in the young, [and] they would seem very natural and proper ways of life” (Sullivan, 1953/1968, pp. 6, 21).

Sullivan concludes that personality is shaped primarily by social forces, with the child’s lengthy period of dependence making it particularly vulnerable to influence by others. He posits a powerful human need for interpersonal relationships, so that “it is a rare person who can cut himself off from … relations with others for long spaces of time without undergoing a deterioration of personality.” And he insists that personality exists, and can be studied, only through its interpersonal manifestations:

Personality is the relatively enduring pattern of recurrent interpersonal situations which characterize a human life.… A personality can never be isolated from the complex of interpersonal relationships in which the person lives …. [Therefore,] psychiatry is the study of the phenomena that occur in interpersonal situations, in configurations made up of two or more people, all but one of whom may be more or less com- pletely illusory. (Sullivan, 1947/1953, p. 10; 1953/1968, pp. 32, 110–111; 1964/1971, p. 33. See also Sullivan, 1953/1968, pp. 18–20, 367–368.)

Since Sullivan’s definition of interpersonal relationships includes those that are illusory, even a recluse or psychotic does not lack a personality. Such individuals have memories and/or fantasies of relationships with real or fictitious others, so they are strongly influenced by interpersonal situations.

Tension Reduction

In accordance with Horney and Fromm, Sullivan concludes that human beings have a tendency or drive toward mental health. He also shares Freud’s belief that we are motivated by the desire to reduce inner tensions. The ideal human condition is that of total equilibrium (absolute euphoria), a state of utter well-being characterized by the absence of any internal deficiencies or noxious external stimuli. The opposite extreme, absolute tension, is reciprocally related to euphoria and is similar to a state of terror. Like mathematical limits, however, absolute euphoria and absolute tension can only be approached and do not exist in nature. (See Sullivan, 1947/1953, p. 97; 1953/1968, p. 35; 1954/1970, p. 106; 1956/1973, p. 265.)

According to Sullivan, there are four major causes of tension: the physicochemical needs, the need for sleep, anxiety, and the need to express tenderness.

Physicochemical Needs and Sleep. A state of inner disequilibrium is created by such important physicochemical needs as sexual desire, the necessity of eliminating bodily wastes, and deficiencies in food, water, oxygen, or body heat. This is accompanied by tension that is often (but not always) conscious, thereby motivating us to expend energy and satisfy the need. Tension also arises from the need to sleep, which Sullivan regards as significantly different from the physicochemical needs.

AnxietyPerhaps the most important cause of tension is anxiety. This unpleasant emotion varies considerably in intensity. At its most extreme, anxiety resembles the uncanny emotions of dread, horror, and loathing:

Uncanny emotions have a sort of shuddery, not-of-this-earth component . . . [somewhat like one’s] first glimpse into the Grand Canyon.… If you try to analyze the experience, you may talk about your skin crawling … [and] if there were a great deal more of such emotion, you would be very far from a going concern as long as you had it. That is the nearest I can come to hinting at what I surmise infants undergo when they are severely anxious. (Sullivan, 1953/1968, p. 10; see also pp. 4, 8–11, 41–45, 59.)

Anxiety can be provoked by disturbances in the environment, such as a sudden loud noise or threat. But its major source concerns the child’s relationship to its mother, or whichever person fulfills the mothering function. “The basic vulnerabilities to anxiety [are] in interpersonal relations.… The tension of anxiety, when present in the mothering one, induces anxiety in the infant” (Sullivan, 1953/1968, pp. 11, 41; see also pp. 9, 113–117, 144, 204). Anxiety differs from the tension of fear, which occurs when the satisfaction of a need is substantially delayed.

The means by which anxiety is communicated from the mothering one to the baby is uncertain, per- haps including some sort of empathy on the part of the latter. In any case, its effects are extremely trouble- some. Unlike the physicochemical needs, which are readily satisfied through such actions as eating or drinking, anxiety is best alleviated by safe relationships with nonanxious others (interpersonal security). No specific action by the infant is involved, which makes the relief of anxiety significantly different from all other needs—and more difficult to achieve.

To make matters worse, anxiety opposes the satisfaction of other needs. It can interfere with the ability to swallow when hungry or thirsty, or to fall asleep when tired. It can disrupt the capacity for rational thought, much like a severe blow on the head. And it can sabotage potentially gratifying interpersonal relationships, as when the hungry but anxious infant rejects the proffered nipple. For these reasons, we devote much of our lifetimes (and a great deal of energy) to reducing or avoiding the wholly unwanted tension of anxiety—as by ignoring that with which we cannot cope (selective inattention), or by unconsciously converting anxiety into the more palatable emotion of anger. “It is anxiety which is responsible for a great part of the inadequate, inefficient, unduly rigid, or otherwise unfortunate performances of people … [and] for a great deal of what comes to a psychiatrist for attention” (Sullivan, 1953/1968, p. 160; see also Sullivan, 1953/1968, pp. 42, 53, 92–96, 152, 319, 353; 1954/1970, pp. 100, 135; 1956/1973, pp. 38–76).

TendernessTension also occurs when the mothering one observes activity by the infant that indicates the existence of a need, such as crying. This tension “is experienced as tenderness, and as an impulsion to activities toward the relief of the infant’s needs” (Sullivan, 1953/1968, p. 39). Thus the infant’s need for the tender cooperation of the mothering one stimulates her need to give it, and produces the baby’s first significant interpersonal relationship. If the mothering one should respond to the infant’s distress with anxiety, however, her capacity for tenderness will be inhibited. Therefore, “there isn’t any right thing to do with infantile anxiety, except for the mother to cease to be anxious” (Sullivan, 1953/1968, p. 54).


Although Sullivan espouses a tension-reduction model of personality, he rejects the concept of psychic energy. “Energy, when I mention it, is energy as conceived in physics .… There is no need to add adjec- tives such as ‘mental.’ … Physical energy … is the only kind of energy I know” (Sullivan, 1953/1968, pp. 35–36, 97, 101–102). We transform physical energy, rather than libido, into behaviors designed to satisfy our needs.

To emphasize that personality is a dynamic process, and is constantly in a state of flux, Sullivan refers to such energy transformations as dynamisms. A dynamism may take various forms: overt moving or talking, covert reveries and fantasies, or partly or wholly unconscious processes.

For example, the hate dynamism involves the transformation of (physical) energy into behavior that will reduce tension through hostility. The individual may strike or insult someone, have murderous fantasies, and/or form powerful unconscious destructive wishes. The dynamism of lust concerns the use of energy to satisfy sexual needs, as by making love or having erotic daydreams. (See Sullivan, 1953/1968, pp. 102–107; 1964/1971, p. 35 n. 3.)

One particularly important dynamism is the self-system, which serves our need to be free from the tension of anxiety. Since this comes as close as Sullivan ever does to some sort of personality structure, it will be discussed in a subsequent section.

Modes of Experiencing

According to Sullivan, human experience consists entirely of tensions, dynamisms, and need satisfactions. He also argues that this experience occurs in one or more of three modalities: prototaxic, parataxic, and syntaxic.

The Prototaxic ModeThe primitive prototaxic mode is the newborn infant’s only way of apprehending the environment. This limited form of experience is like a succession of momentary discrete states, and is incapable of such distinctions as before and after or self and others. Prototaxic experience cannot be communicated in any kind of symbols, so it is only an educated guess by Sullivan as to the inner processes of the very young infant. (See Sullivan, 1953/1968, pp. xiv, 28–29, 35–36, 100.)

The Parataxic ModeAs the infant develops, it becomes capable of the parataxic mode. This mode is characterized by the use of private or “autistic” symbols (such as nonsense words whose meaning is known only to the user), and by the ability to distinguish differences in time. Parataxic experience is illogical, however, and reflects a lack of understanding about causality. Examples include the superstitious belief that misfortune will result from walking under a ladder, or in front of a black cat; and the psychotic who thinks about rain on a clear day, and concludes that these ruminations caused the downpour that occurred some time thereafter.

The Syntaxic ModeThe most elaborate form of experience is the syntaxic mode, which begins to appear as early as the 12th to 18th month of life. Syntaxic experience can be communicated through the use of symbols that are socially accepted and understood, such as “that fantastic evolutionary development, language.” It also includes an understanding of conventional concepts of cause and effect (Sullivan, 1953/1968, p. 20).


To Sullivan, as to Jung and Adler, human behavior must be understood in terms of both causality and teleology. Our capacity for foresight develops in early infancy, and represents one of the striking characteristics of the human species (Sullivan, 1953/1968, pp. 38–39; see also pp. 51, 64, 82). Unlike Jung, however, Sullivan concludes that psychological phenomena can be translated into physiological events, and his writings include numerous allusions to the anatomical and organic aspects of human functioning.


Sullivan shares Fromm’s and Horney’s distaste for the Freudian structural model. According to Sullivan, we organize our experiences by forming mental conceptions of ourselves and others. These personifications consist of learned feelings and beliefs, which often do not correspond well with reality.

For example, a mother is likely to misperceive her child to some extent. She may form a personification that is more like the way she wants the child to be, or one that is influenced by her experiences with previous offspring. The infant gradually develops a personification of the good mother from her tender and need-satisfying behaviors, and also forms a personification of the bad mother from her frustrating and anxiety-producing behaviors. These personifications are also somewhat inaccurate, partly because the baby’s ability to perceive and interpret the environment is limited. In fact, the infant does not realize at first that the personifications of good and bad mother refer to the same person, although significant portions of each do ultimately fuse into a complicated whole. (See Sullivan, 1953/1968, pp. 110–124, 167, 188–189.)

The irrational aspect of personifications is evident in the case of stereotypes, or beliefs that are applied rigidly and equally to a group of people and obscure the true differences among them. Young children commonly form a stereotype of the opposite sex as undesirable, whereas prejudiced individuals incorrectly personify members of a particular group as having certain negative characteristics in common. (See Sullivan, 1953/1968, pp. 236–238, 302–304.) Such irrationality is also apparent in the personification that we form of ourselves.

The Self-System

The growing infant begins to conceive of itself as a separate and distinct entity at about 6 months of age, and it organizes this information by forming appropriate personifications. These are so important that Sullivan tends to divide personality into two major categories: the self-system (or self-dynamism) and everything else. (See Sullivan, 1947/1953, pp. 19–29; 1953/1968, pp. 135–141, 158–171, 198–202; 1954/1970, pp. 101–112, 138.)

The self-system results partly from experiences with the infant’s own body. For example, thumb sucking helps the infant differentiate between self and others because it produces unique feelings of both sucking and being sucked. For the most part, however, the self-system originates from the appraisals of significant others (such as the parents). The child must learn to get along with other people, so the unconditional maternal tenderness of early infancy must eventually be replaced by a system of rewards and punishments that will prepare the child for its place in society.

During later infancy and childhood, therefore, tenderness is used as an anxiety-reducing reward for desirable behaviors (such as achieving success at toilet-training, or abandoning the cherished but socially unacceptable activity of thumb sucking). Forbidding gestures, maternal anxiety, punishment, and the absence of tenderness serve as anxiety-inducing responses to the infant’s errors and misdeeds. This leads to the development of the two personifications that eventually comprise the self-system: the desirable self or obedient “good me” is associated with experiences that are rewarded by a decrease in anxiety, the undesirable self or rebellious “bad me” results from experiences that are punished by an increase in anxiety.

As with the infant’s conceptions of good and bad mother, the good-me and bad-me personifications ultimately fuse into a single entity. The more intense the early experiences of anxiety, the more rigid and extensive this self-system will be. To Sullivan, therefore, self-centered behavior is a learned response to anxiety, rather than the result of innate narcissism. (See Sullivan, 1947/1953, p. 127 n. 41; 1953/1968, pp. 5, 126–134, 151–202.)

Selective Inattention. The primary goal of the self-system is the desirable one of reducing anxiety, thereby enabling the child to get along with its parents and satisfy its needs. However, it accomplishes this objective through such dubious “security operations” as selective inattentionIf the self-system should encounter information that threatens its stability, it simply ignores or rejects the incongruous data and goes on functioning as before.

Selective inattention may occasionally have beneficial aspects, as when you avoid costly distractions by concentrating on the task at hand. But it is primarily disadvantageous, for it impedes our ability to learn from our failures and weaknesses. These sources of information are threatening to the self-system, so it is likely to pay no attention to them. (See Sullivan, 1953/1968, p. 319.) Selective inattention is so pervasive that most of our mental processes occur outside the realm of consciousness.

Because the self-system uses selective inattention to combat anxiety, it differs from other dyna- misms by being extraordinarily resistant to change. This rigidity represents the principal stumbling block to constructive growth. “We are being perfectly irrational and simply unpleasant if we expect another person to profit quickly from his experience, as long as his self-system is involved—although this is a very reasonable anticipation in all fields in which the self-system is not involved.” The security operations of the self-system also create the impression that we differ more from others than is actually the case (a “delusion of unique individuality”), and may even result in a grandiose self- personification somewhat like the Adlerian superiority complex (Sullivan, 1953/1968, pp. 140, 192; see also Sullivan, 1953/1968, pp. 168–170, 247–248, 319, 346, 374; 1964/1971, pp. 198–226; 1956/1973,

pp. 38–76).

The “Not-Me” Personification

Intense anxiety during childhood (“a very poor method of education”) leads to the development of the “not-me” personification, a shadowy and dreadful aspect of personality that is usually unconscious. The not-me personification involves material so threatening that even the bad-me personification cannot cope with it, so it is unconsciously divorced (dissociated) from the self-personification. Dissociation is an extreme form of security operation that resembles “[flinging] something of you into outer darkness, where it reposes for years, quite peacefully,” and it must be maintained through constant (if unconscious) vigilance (Sullivan, 1953/1968, pp. 163, 318; see also Sullivan, 1947/1953, p. 71; 1953/1968, pp. 201, 314–328; 1964/1971, pp. 248–249).

For example, a young child or schizophrenic may seek to avoid punishment by arguing and believing, “Oh, I didn’t do that, it was my hand.” The emergence into consciousness of the not-me personification produces uncanny emotions and the feeling of not being oneself, a terrifying experience that is common in schizophrenia—and in some nightmares and states of shock that befall more healthy individuals.1

Other Defensive Behaviors

Sullivan differs from Freud by interpreting sublimation as a device for reducing anxiety, wherein one behavior is unconsciously substituted for another that would be more satisfying but also more threatening. This conception implies that sublimation is not always advantageous, for it may cause us to accept a less satisfying substitute on those occasions when anxiety has mistakenly become associated with acceptable activities. Sullivan also argues that fantasy can help us make useful, realistic plans for the future.

Sullivan is highly (and wryly) skeptical about various other Freudian constructs. He regards introjection as “a great magic verbal gesture, the meaning of which cannot be made explicit.” Projection is a “nice [topic] for certain late-evening-alcoholic psychiatric discussions.” And regression is something that happens every 24 hours when a child goes to sleep and complicated, recently acquired patterns of behavior collapse, rather than “some great abstruse whatnot” that “psychiatrists often use … to brush aside mysteries which they do not grasp at all” (Sullivan, 1953/1968, pp. 166, 197, 359; see also Sullivan, 1947/1953, p. 54 n. 20; 1953/1968, pp. 113, 191–196, 348–350; 1964/1971, pp. 209–210; 1956/1973, pp. 14–20, 232).


Sullivan regards developmental psychology as the key to understanding human behavior, and he concludes that significant changes in personality often occur during late childhood and adolescence. He posits seven epochs through which personality may develop, each of which represents an optimal time for certain innate capacities to reach fruition.


The stage of infancy begins a few minutes after birth and continues until the appearance of speech (however meaningless). (See Sullivan, 1953/1968, pp. 49–187.) Nursing provides the infant’s first experience in interpersonal relationships. The infant learns to distinguish among such important external cues as the “good and satisfactory nipple,” which is provided by a tender mother and gives milk when the infant is hungry; and the “evil nipple” of the anxious mother, which is so unpleas- ant that it is rejected even if the infant is hungry. (See Sullivan, 1953/1968, pp. 49–50, 73, 79–81, 88, 122.)

Crying is for some time the infant’s most effective method of satisfying needs and reducing anxiety, and varies according to its intent. Crying-when-hungry represents the infant’s crude, prototaxic way of expressing the idea “come, nipple, into my mouth,” and “has no necessary relatedness, in the infant’s experience, with crying-when-cold, crying-when-pained, or crying-under-any-other-circumstances” (Sullivan, 1953/1968, p. 67; see also pp. 52–53, 62, 66–75). When crying does bring the desired relief, such successes help the infant to develop foresight and an understanding of cause and effect. That is, the infant (in some primitive prototaxic fashion) concludes that “I cry when I suffer a certain distress, and that produces something different which is connected with the relief of the distress” (Sullivan, 1953/1968, p. 72).

The self-system begins to develop during mid-infancy. As we have seen, this is due primarily to two factors: bodily explorations such as thumb sucking, and the mothering one’s shift from unconditional tenderness to rewards and punishments that cause the “good me” and “bad me” personifications to develop. These rewards and punishments involve such essential tasks as learning to use the toilet and giving up thumb sucking.

During the 12th to 18th month of life, the use of language begins with the imitation of sounds in the environment. This represents the appearance of the parataxic mode (or syntaxic mode, if the infant’s utter- ances happen to correspond to actual words), and ushers in the second stage of personality development.


During childhood, parental punishments further the growth of the bad-me aspect of the self-system. (See Sullivan, 1953/1968, pp. 188–226.) So long as the parents also assist the development of the good-me personification by providing sufficient rewards and tenderness, no great harm will result. But if the child’s need for tenderness is consistently rebuffed by parental anxiety, irritability, or hos- tility, the bad-me component will dominate the self-system. As with the neglected child in Adlerian theory, this malevolent transformation results in the misguided belief that other people are hostile and unloving:

[The malevolent transformation] is perhaps the greatest disaster that [could happen] in the childhood phase of personality development.… [Such a child learns] that it is highly disadvantageous to show any need for tender cooperation from the authoritative figures around him. [Instead] he shows … the basic malevolent attitude, the attitude that one really lives among enemies.… This distortion, this malevolence … runs something like this: Once upon a time everything was lovely, but that was before I had to deal with people. (Sullivan, 1953/1968, pp. 214, 216.)

The malevolent child may be mischievous, behave like a bully, or express resentment more passively by stubbornly failing to do whatever is required. This transformation also impairs the sufferer’s relation- ships with others, notably authority figures outside the immediate family.

If the parents are too distant and aloof, the child may try to resolve its feelings of loneliness through excessive daydreaming. This will make it difficult for the child to abandon its private, parataxic thoughts and images in favor of the syntaxic mode and effective communication with other people. But the parents must also avoid the opposite extreme of catering to the child’s every whim, which will produce a spoiled and self-centered individual (similar to Adler’s concept of pampering).

The personifications of good and bad mother begin to fuse into a single entity during childhood. The father now joins the mother as an authority to be reckoned with, leading to the child’s formation of a father personification. A knowledge of gender also begins to develop, with the boy or girl wishing to be like the parent of the same sex. To Sullivan, however, such identifications are not due to some sort of Oedipus complex. They occur because most parents are more comfortable with the child of the same sex, and reward behavior typical of that sex with approval and tenderness (Sullivan, 1953/1968, pp. 218–219).

The Juvenile Era

The juvenile era originates with the appearance of the need for playmates, which occurs at about the time of entry into school. (See Sullivan, 1953/1968, pp. 227–244.) Like Adler, Sullivan concludes that the educational system can remedy serious parental errors that occurred during infancy and childhood. Such favorable alterations in personality are possible because the normally rigid self-system is more amenable to change at the inception of each developmental stage, when newly maturing abilities increase the prob- ability of significant changes in behavior.

The juvenile learns to adjust to the demands, rewards, and punishments of such new authority figures as teachers. He or she observes how other juveniles are treated by these authorities (and by each other), learns to deal with peers (including the malevolent bully), and is introduced to the social processes of competition and compromise. School also involves the painful possibility of ostracism by one’s peers. The juvenile era is the time when the world begins to be complicated by the presence of other people, and is typified by inexperienced attempts at interpersonal relationships that often reflect a shocking insensitivity to other people’s feelings. (See Sullivan, 1947/1953, pp. 38–41; 1953/1968, pp. 227–232.)

The syntaxic mode becomes prominent during the juvenile era. In addition, the parents begin to lose their godlike attributes and take on more human, fallible personifications. “[If one] comes out of

the juvenile era with [the feeling that the parents] still have to be sacrosanct, the most perfect people on earth, then one of the most striking and important of the juvenile contributions to socialization has sadly miscarried” (Sullivan, 1953/1968, p. 231). Another potential source of pathology involves parents who constantly disparage other people, which causes the juvenile to feel incapable of knowing what is good:

If you have to maintain self-esteem by pulling down the standing of others, you are extraordinarily unfortunate.… The doctrine that if you are a molehill then, by God, there shall be no mountains … is probably the most vicious of the inadequate, inappropriate, and ineffectual performances of parents with juveniles.… (Sullivan, 1953/1968, pp. 242–243, 309.)


The preadolescent stage is highlighted by the need for an intimate relationship with a particular individual of the same sex, or chum. (See Sullivan, 1953/1968, pp. 245–262.) This relatively brief period tends to occur between the ages of 8 ½ and 10, though it may be delayed by as much as a few years if maturation is relatively slow.

The preadolescent chumship is crucial because it represents a sincere interest in the welfare of another person. The influence of this important individual may be sufficient to modify the otherwise rigid self- system and correct any warps in personality carried over from preceding stages. “Because one draws so close to another, because one is newly capable of seeing oneself through the other’s eyes, the preadolescent phase of personality development is especially significant in correcting autistic, fantastic ideas about oneself or others” (Sullivan, 1953/1968, p. 248; see also Sullivan, 1947/1953, pp. 41–44). An effective chumship may alter such misguided views as arrogance, dependence, or the belief that one should be liked by everyone. It may even reverse or cure a malevolent transformation. Conversely, difficulties in dealing with others of the same sex are invariably due (at least in the case of males) to the failure to develop this essential preadolescent relationship, and to the resulting feelings of intense loneliness.

Early Adolescence

The period of early adolescence begins with puberty and the appearance of the powerful lust dynamism, which leads to the desire for a close relationship with a member of the opposite sex. (See Sullivan, 1953/1968, pp. 263–296.) In contrast to the intimacy of the preadolescent stage, which is not necessarily sexual, lust is expressed primarily through sexual behavior and culminates in the experience of orgasm.

Sullivan warns that early adolescence is rife with possibilities for serious maladjustment, because our culture confronts us with singular handicaps in our pursuit of lustful activity. Essential information and guidance may be totally lacking at this important time, and the parents may add to the problem by providing ridicule and sarcasm instead of emotional support. Thus the adolescent’s fledgling attempts at heterosexuality may lead to such embarrassing outcomes as impotence, frigidity, or premature ejaculation, causing a sharp decrease in self-esteem; and “customarily low self-esteem makes it difficult indeed for the carrier person … to manifest good feeling toward another person” (Sullivan, 1953/1968, p. 351; see also Sullivan, 1947/1953, p. 63; 1954/1970, p. 9).

The adolescent who experiences such difficulties may rush into marriage with the first member of the opposite sex who inspires any feelings akin to love, a relationship that is usually far from satisfying. Or the adolescent may develop a fear of the opposite sex, possibly resulting in celibacy, excessive fantasizing, or homosexuality. Or the adolescent may conduct an endless quest for the ideal member of the opposite sex, and blame the inevitable failures on apparent defects in every candidate rather than on the unconscious fear of heterosexuality. An adolescent who has not emotionally outgrown the juvenile era may form

numerous superficial sexual liaisons (“Don Juanism”). Whereas occasional masturbation is not harmful (and is virtually universal), an anxious adolescent may rely so heavily on self-stimulation that healthy heterosexuality becomes impossible.

Despite the seriousness of such problems, Sullivan does not regard sexual dysfunction as the most important aspect of psychiatry. He prefers to emphasize the inability to form satisfying interpersonal relationships, which underlies the more manifest sexual difficulties (Sullivan, 1953/1968, pp. 295–296; 1954/1970, p. 13).

Late Adolescence

The latter part of adolescence originates with the achievement of satisfying sexual activity. (See Sullivan, 1953/1968, pp. 287–310). The adolescent must now contend with increasing social responsibilities, such as working and paying income tax. Socioeconomic status also affects this stage of personality development, for those who are able to attend college have several years of extraordinary opportunity for observation and learning that others do not. Faulty personality development may now be evidenced by a pronounced tendency to avoid others, or by such pseudosocial rituals as impersonal card games that provide only the most superficial of contacts.


Sullivan has relatively little to say about the stage of adulthood, which represents maturity and the completion of personality development, because psychiatrists do not get many opportunities to observe well-adjusted behavior. Adulthood is denoted by a mature repertory of interpersonal behaviors and the capacity for genuine love, a state wherein “the other person is as significant, or nearly as significant, as one’s self (Sullivan, 1953/1968, p. 34; see also Sullivan, 1947/1953, p. 42; 1953/1968, pp. 297, 309–310).

This final epoch is somewhat similar to the Freudian genital stage (without the sexual implications) and to Fromm’s productive orientation. However, Sullivan is not optimistic about our chances to attain it. “I believe that for a great majority of our people, preadolescence is the nearest that they come to untrou- bled human life—that from then on, the stresses of life distort them to inferior caricatures of what they might have been” (Sullivan, 1947/1953, p. 56).



In accordance with all of the theorists discussed thus far, Sullivan regards psychopathology as a difference in degree rather than in kind. Every patient “is mostly a person like the psychiatrist,” and even the bizarre behavior of the psychotic is related to processes that occur in relatively normal individuals (Sullivan, 1947/1953, p. 96; see also Sullivan, 1953/1968, pp. 208, 223; 1954/1970, pp. 18, 183).

Causes of Psychopathology. Neurosis and schizophrenia are caused by pathogenic interpersonal relationships. Too much maternal anxiety during infancy, not enough parental tenderness during childhood, the failure to find a satisfactory peer group during the juvenile era or a preadolescent chum, or problems in early adolescence with heterosexual relationships and the lust dynamism damage the child’s self-esteem and cause the self-system to become extremely rigid. As a result, the sufferer is unable to develop a repertory of behaviors for dealing effectively with other people.

For example, malevolent individuals cannot apply judicious doses of empathy and compassion because they are constantly preoccupied with the hate dynamism and attempts to dominate others. To Sullivan, then, all mental disorders are to be understood as patterns of inadequate and inappropri- ate behavior in interpersonal relations. (See Sullivan, 1953/1968, pp. 313–328, 344–363; 1954/1970, pp. 183–208; 1956/1973, pp. 200–202.)

Sullivan shares Fromm’s concern about pathogenic societal forces, partly because of the devastation he observed during World War II. “The Western world is a profoundly sick society in which each denizen, each person, is sick to the extent that he is of it” (Sullivan, 1964/1971, p. 155; see also pp. 76–84, 100–107).

Varieties of Psychopathology. Sullivan is critical of the standard psychiatric nomenclature, which he regards as a source of potential confusion. “These trick words, so far as I can discover, merely make one a member of a somewhat esoteric union made up of people who certainly can’t talk to anybody outside the union and who only have the illusion that they are talking to one another” (Sullivan, 1953/1968, p. 7). Nevertheless, his major clinical interests concern two of the standard classifications: obsessive-compulsive neurosis and schizophrenia.

Obsessive-compulsive neurosis reflects an extreme vulnerability to anxiety and a profound loss of self-esteem, caused by never having had outstanding success in one’s interpersonal relations. Repeated thoughts or actions are (unconsciously) substituted for behaviors that would evoke intense anxiety, as when an adolescent who has suffered sexual embarrassments stays at home and obsesses about romantic triumphs instead of risking going out on a date. Secondary gains also play a significant role in this dis- order, as with the patient of Sullivan’s who could not leave the second floor of his home because of an obsession about committing suicide by jumping from a flight of stairs. He not only achieved some security against a threatening external world, but also enjoyed the constant sympathy and attention of his wife—at least until she grew weary of his neurotic demands, and divorced him a few years later. (See Sullivan, 1953/1968, pp. 210–211, 318–319; 1964/1971, pp. 231–232; 1956/1973, pp. 229–283.)

Schizophrenia is caused by the occurrence of uncanny emotions early in life, notably extreme anxiety, or by disastrous blows to one’s self-esteem during the latter stages of development (particularly adoles- cence). If a parent has irrational fears about the infant’s sexuality and becomes horrified when the baby toys with its genitals, the resulting extreme anxiety is likely to prove as numbing and incomprehensible as a severe blow on the head. Rather than associating the genitals with the bad-me personification, as would be the case with less traumatic punishment, the child may instead dissociate this highly threatening issue from its self-system. Thus sexual impulses and behaviors become associated with the unconscious not-me personification and are attributed to external sources, producing a gap in this area of personality that will create serious difficulties during early adolescence. The schizophrenic’s quest for security also involves a regression to parataxic speech that has meaning to the patient, but appears incomprehensible and bizarre to others. To Sullivan, therefore, schizophrenia represents a return to an early form of mental functioning in an attempt to ward off intense anxiety and restore a shattered sense of self-esteem. (See for example Arieti, 1974, pp. 25–29; Sullivan, 1953/1968, pp. 313–328, 360–361; 1956/1973; 1962/1974.)


Theoretical Foundation. To Sullivan, psychotherapy is first and foremost a learning process. “There is no essential difference between psychotherapeutic achievement and achievements in other forms of education … [all of which are] in the end reducible to the common denominator of experience incorporated into the self.… [Thus] I am avoiding the term ‘cure,’ since I do not think it applies in the realm of personality” (Sullivan, 1956/1973, p. 228; 1962/1974, p. 281; see also Sullivan, 1954/1970, p. 238).

Ideally, therapy enables the patient to gain valuable insights into issues that were selectively inat- tended, reintegrate dissociated aspects of personality, and establish a proper balance between the good-me and bad-me personifications. This expansion of the self-system facilitates the development of a wider,

more effective repertory of interpersonal behaviors. Thus therapy may help a patient suffering from dis- sociated sexual impulses to accept the existence of lustful drives, recognize and eliminate the accompany- ing shame and guilt, and develop appropriate behaviors for satisfying this need. Or a malevolent patient may learn to reduce an exaggerated bad-me personification, establish some love of self, develop more accurate interpersonal perceptions, and (ultimately) express tenderness and love to others.

Therapeutic Procedures. Sullivanian psychotherapy focuses on the interpersonal relationship between the patient and therapist. The therapist is an active participant as well as an observer, focusing on what the patient is saying “with me and to me” and preventing lengthy forays into inconsequential territory. “The expert [therapist] does not permit people to tell him things so beside the point that only God could guess how they happened to get into the account” (Sullivan, 1954/1970, pp. 34, 58; see also Sullivan, 1953/1968, pp. 13–14; 1954/1970, pp. 3–6, 19–25, 82–85, 113).

Sullivanians eschew the use of a couch and sit at a 90-degree angle to the patient, enabling them to detect sudden changes of posture without being distracted by facial expressions. They also reject the use of free association with schizophrenics as too anxiety-provoking, and limit its use with other disorders to times when the patient is blocked and ready to recognize its value. Sullivan dislikes taking written notes during the therapeutic session, arguing that this method is distracting and cannot register subtle nuances in behavior, but does advocate the use of tape recordings. He also favors relatively brief and simple interpre- tations, so as to avoid causing excessive anxiety and intensifying the defenses of the patient’s self-system. (See Singer, 1970, pp. 196–199; Sullivan, 1953/1968, p. 302; 1954/1970, p. 90; White, 1952, pp. 132–133.)

Psychotherapy begins with the stage of “formal inception,” during which the patient first meets the psychiatrist and provides some explanation for entering therapy. Sullivan warns that the therapist’s ini- tial behavior is of considerable importance, for even such apparently minor errors as an overly limp handshake, an excessively warm or cool greeting, or somewhat too much arrogance or diffidence can significantly affect the patient’s perceptions and distort the subsequent course of treatment. (See Sullivan, 1954/1970.)

The second stage of therapy, or “reconnaissance,” occurs when the psychiatrist has formed a fairly good idea as to why the patient is in need of professional assistance. This period generally takes from 7 ½ to 15 hours and consists of an unstructured inquiry into the patient’s life history. The reconnaissance concludes with a summary statement of what the therapist has learned about the patient, whereupon the patient usually agrees that some significant problems have emerged that are worthy of further study.

The third stage, or “detailed inquiry,” represents the “long haul” of psychotherapy. No matter how skilled the therapist may be, the preceding brief stages are unlikely to provide a wholly accurate picture. Many patients try to reduce anxiety by making statements designed to please or impress the therapist, by rationalizing or ignoring their failures and embarrassments, or by exaggerating their successes. Or a patient’s communications may be deceptively difficult to understand because they include many parataxic symbols. During the detailed inquiry, therefore, the impressions gained from the formal inception and reconnaissance are checked against more substantial data provided by the patient. The therapist probes into important aspects of the patient’s developmental history, including such issues as toilet-training, school experiences, the preadolescent chum, sexual relationships, and security operations for avoiding anxiety.

The final stage of therapy (“termination”) includes four major steps: a succinct formal statement of what the therapist has learned during the course of treatment, a prescription for actions that the patient should take or avoid, a formal assessment of the patient’s probable future course in life, and a clear-cut leave-taking that is neither too indecisive nor too abrupt. Pessimistic prognoses are avoided, however, since they may become self-fulfilling prophecies. “I try never to close all doors to a person; the person should go away with hope and with an improved grasp on what has been the trouble” (Sullivan, 1954/1970, p. 211).

Unlike Freud, Sullivan devoted the majority of his attention to the treatment of schizophrenia. He was sharply critical of the inferior methods and conditions of most mental hospitals, and any ward under his supervision was conducted according to his own unique regulations. For example, he prohibited female nurses from appearing in all-male wards because the patients were likely to regard them as threatening

symbols of authoritarianism. Instead he trained his own (male) assistants, and emphasized on them his belief that the patients’ daily life and social contacts on the ward were even more important than the hourly sessions with the psychiatrist. (See Arieti, 1974, p. 541; Sullivan, 1954/1970, pp. xx, 50; 1962/1974, pp. xvi–xix.) Although capable of pronounced sarcasm with colleagues, Sullivan was unfailingly kind and gentle with schizophrenics. Even when an upset patient would slap him in the face, he would strictly prohibit any reprisals and only ask quietly, “Well, do you feel better now?” Nor was he afraid to be somewhat unorthodox, and would compensate for the lack of modern drug therapy by using alcoholic beverages to relax a rigid self-system and make the patient more amenable to change (Sullivan, 1947/1953, p. 219).

Dream InterpretationSullivan does not regard dreams as particularly rich sources of information about personality. He argues that our recall upon awakening is hopelessly distorted by the resurgent self-system, and he criticizes interpretations designed to unearth some sort of latent content as futile efforts to translate the dreamer’s private (parataxic) symbology into communicable (syntaxic) experience.

Sullivan limits dream interpretation to reflecting back important aspects, with the goal of stimulating the patient’s train of thought. A patient of Sullivan’s once dreamed of approaching a highly attractive Dutch windmill, only to find upon entering that it was ruined and inches deep in dust. Sullivan’s reply was, “that is, beautiful, active on the outside—utterly dead and decayed within. Does it provoke anything?” Whereupon the patient responded, “my God, my mother,” recognizing with astonishment that he actually regarded her as a “sort of zombie … [or] weary phonograph offering cultural platitudes” (Sullivan, 1953/1968, pp. 338–339; see also Sullivan, 1947/1953, pp. 69–72; 1953/1968, pp. 329–337; 1956/1973, pp. 19–20).‌

Resistance and Transference. Sullivan concludes that the self-system actively opposes the goals of therapy. But he interprets this as an attempt to reduce anxiety, rather than as an effort to preserve illicit impulses. Sullivan is opposed to the use of transference, which he regards as another erroneous interpersonal perception that the patient must learn to abandon. (See Sullivan, 1953/1968, pp. 237–238; 1954/1970, pp. 104, 139, 219, 231.)

Above all, Sullivan emphasizes the difficulty of doing effective psychotherapy. “There is no fun in psychiatry.… It is work—work the like of which I do not know” (Sullivan, 1954/1970, p. 10).

Psychotherapy and Social ReformLike Fromm, Sullivan (1964/1971) discusses such social applications of personality theory as world tensions, national defense, and propaganda and censorship. However, his untimely death prevented him from devoting more than a few articles to this area.


Criticisms and Controversies

For one who claims to dislike psychiatric jargon, Sullivan is not averse to introducing some formidable terminology of his own. His writings are probably more difficult than any of the theorists discussed thus far except Jung (and perhaps some parts of Freudian theory), and his language poses a considerable barrier to the prospective reader.

Although based on a different premise, Sullivan’s tension-reduction model of human motivation is as subject to criticism as Freud’s. The concept of dynamism has an all-pervasive quality similar to Freudian sexuality and Adlerian inferiority, and has not enjoyed much popularity among modern psychologists. Also, Sullivan often fails to acknowledge his intellectual debts to his predecessors. He ignores obvious

similarities between such concepts as actively maintained dissociations and actively maintained repressions, security operations and defense mechanisms, the malevolent transformation and Adler’s theory of the neglected child, personifications and Adler’s “fictions,” and the self-system and the style of life.

Sullivan has also been criticized for attributing the formation of the self-system primarily to the appraisals of significant others, and for ignoring those distortions introduced by the child’s own misperceptions and faulty cognitions. (See Arieti, 1974, p. 78; Ellenberger, 1970, p. 639.) Sullivan’s theory has not generated much empirical research. In comparison to Fromm, Sullivan’s approach to dreams seems shallow and unconvincing. And Sullivan has been taken to task for overemphasizing the interpersonal aspects of personality.


Perhaps Sullivan’s greatest contributions concern the understanding and treatment of schizophrenia. He has been credited as the first to offer a convincing psychodynamic interpretation of this disorder (Arieti, 1974, p. 25). As with Horney and Fromm, Sullivan’s rejection of libido theory would seem to be a step in the right direction. Sullivan devotes far more attention to the development of personality than either Horney or Fromm, and his emphasis on adolescence appears preferable to Freud’s and Adler’s relative lack of concern with this stage of life.

Sullivan’s conception of anxiety is superior to that of Adler, and reflects a better understanding of the pain and suffering that are involved. As Sullivan puts it, “Under no conceivable circumstances … has anyone sought and valued as desirable the experience of anxiety.… People who ride on roller coasters pay money for being afraid. But no one will ever pay money for anxiety in its own right. No one wants to experience it. Only one other experience—that of loneliness—is in this special class of being totally unwanted” (1954/1970, p. 100).

Sullivan’s extensive neologisms are indeed troublesome, so much so that some students of human behavior prefer to avoid his ideas. This is unfortunate, for a careful study of his major writings will yield quite a few pearls of genuine wisdom.

Suggested Reading

The most complete discussion of Sullivanian theory is presented in The Interpersonal Theory of Psychiatry (1953/1968). Also well-regarded is his work on psychotherapeutic procedures, The Psychiatric Interview (1954/1970). For a biography of Sullivan, see Perry (1982).


  1. The basic nature of human beings. Sullivan is perhaps less optimistic about human nature than Horney and Fromm, but he does posit an inherent drive toward mental health. The One-Genus Postulate: To Sullivan, human personalities more closely resemble each other than anything else in the world. Therefore, unlike Adler, he prefers to emphasize the similari- ties among human beings rather than the differences. The Need for Others: Relationships with other people are essential to proper personality development. In fact, Sullivan defines person- ality in terms of interpersonal factors. Tension Reduction: We are motivated to reduce various tensions, the most notable of which is anxiety. Other tensions arise from the physicochemi- cal needs, the need for sleep, and the arousal of maternal tenderness. Dynamisms: Sullivanrejects the construct of psychic energy, concluding instead that behavior can be explained in terms of transformations of physical energy (dynamisms). Other Factors: Human experience occurs in one or more of three modalities: prototaxic, parataxic, and syntaxic. Like Jung and Adler, Sullivan concludes that behavior must be understood in terms of both causality and teleology.
  2. The structure of personality. The growing child organizes its experience by form- ing mental conceptions (personifications) of other people and, most importantly, of itself. The Self-System: Personifications of oneself result from experiences with one’s own body, and from the appraisals of significant others (particularly the parents). This self-system consists of the good-me and bad-me personifications, has the goal of reducing anxiety, and is remarkably resistant to change. The Not-Me Personification: Aspects of personality that cause intense anxiety are dissociated from the self-system, and comprise the shadowy and dreadful not-me personification. This personification plays a significant role in schizophrenia. Other Factors: Sullivan regards much of personality as unconscious, but he attributes this to selective inatten- tion rather than to repression. He interprets sublimation as an attempt to reduce anxiety, rather than as the diversion of illicit impulses.
  3. The development of personality. Sullivan regards developmental psychology as the key to understanding human behavior. He discusses in detail seven developmental epochs, each of which represents an optimal time for certain innate capacities to reach fruition. Infancy is highlighted by maternal tenderness and anxiety, crying and nursing, and the beginnings of the self-system. Childhood is a time for dealing with parental rewards and punishments, and may give rise to the malevolent transformation or loneliness. The juvenile era originates with the appearance of the need for playmates, and is when the world begins to be compli- cated by the presence of schoolmates and social groups. Preadolescence is highlighted by the need for an intimate relationship with a particular individual of the same sex, or chum. Early adolescence begins with puberty, and includes the desire for a close relationship with a member of the opposite sex. Late adolescence originates with the achievement of satisfying sexual activity, and involves increasing social responsibilities. Finally, adulthood—a stage of true maturity that is probably not attainable by most people—is denoted by the capacity for genuine love.

  4. Further applicationsPsychopathology: Sullivan is particularly concerned with two forms of psychopathology. Obsessive-compulsive neurosis reflects an extreme vulnerability to anxiety and a profound loss of self-esteem, with ritualistic thoughts and actions used to reduce anxiety. Schizophrenia is caused by uncanny emotions that occur early in life, notably extreme anxiety, or by disastrous blows to one’s self-esteem during such later stages as adolescence. It involves the dissociation of highly threatening aspects of one’s personality. Psychotherapy: Sullivan regards psychotherapy primarily as a form of education, rather than as a medically oriented “cure.” The goals are to reintegrate dissociated aspects of personality and expand the self-system, thereby leading to a wider and more effective repertory of interpersonal behavior. Sullivan has devoted considerable attention to the technique of psychiatric interviewing, and has published several articles on social change and international problems.
  5. Evaluation. Sullivan has been criticized for excessive neologisms, the use of a tension-reduction model of motivation, a shallow approach to dream interpretation, failing to acknowledge similarities between his constructs and those of Freud and Adler, and overemphasizing the importance of the interpersonal aspects of personality. Yet he has also been credited as the first to offer a convincing psychodynamic interpretation of schizophrenia, and he was a pioneer in advocating and using more humane treatment methods with such patients. He has furthered our understanding of personality development (including such stages as adolescence, which Freud ignored), the important phenomenon of anxiety, and psychiatric interviewing and psychotherapy. And he played a significant role in numerous international projects seeking the elusive goal of world peace.