A BEHAVIORAL VIEW OF THE THERAPEUTIC RELATIONSHIP
(From “The Therapeutic Relationship”, pp. 133-144. Brooks/Cole, 1985)
If learning is defined as those changes in behavior that are not due to native response tendencies, maturation, or temporary states Of the organism (as a result of fatigue or drugs, for example),’9 then counseling or psychotherapy must involve learning. But there are different views of the nature of learning, differing descriptions of the learning process, and different theories of learning. When I referred to perceptual or phenomenological changes leading to behavior changes, I was speaking in terms of one theory of learning. The detailed considerations of the therapy process by Carl Rogers are formulated, essentially, in terms of a perceptual theory of learning and behavior.“ Perceptual theory is not currently popular or influential in psychology, however. The prevailing orientation in the field of learning is toward behavior theory, and discussions of a learning approach to counseling or psychotherapy usually assume this orientation.
To a certain extent at least, behavioral change, or learning, can be described by, or cast into, various theoretical orientations. Each orientation is to some extent a different way of viewing the same phenomena. One orientation may be more restricted orlimited than another, but even a restricted viewpoint can be expanded or extended to encompass, at least in a general way, a wide variety of phenomena. In this respect, reinforcement, a basic concept of the behavioral approach to learning and behavior, can be extended to apply to the therapy process, as can other concepts as well. I shall therefore describe the process of counseling of psychotherapy that has been developed in this and the two preceding chapters in terms of a behavioral approach to learning.2′ The way in which the therapeutic conditions operate can be described-or „explained“-in terms of this approach. Truax and Carkhuff suggested that there are four learning modalities or channels by which the basic facilitative conditions operate in psychotherapy: (1) reinforcement of human relating, (2) reinforcement of self-exploration, (3) elimination of specific anxieties and fears, and (4) reinforcement of positive self-concepts and self-valuations.“2 In this section we consider each of these modalities in terms of behavioral theory. We also consider a fifth important learning modality: modeling.
Reinforcement of Human Relating
Clients have difficulties in interpersonal relations; thus, interpersonal relations are sources of anxiety and fear. Their interpersonal relations have been associated with painful experiences-rejection, lack of understanding, hostility, threat, and so on, from others. As a result, they are hesitant to approach others and are threatened by the prospect of relating to others. Yet they come for therapy because the pain or disturbance is so great that they will do almost anything to seek help. They approach the therapist, though at the same time they have strong avoidance responses. They suffer from the approach-avoidance conflict described by Miller.“
The therapist responds to the client with warmth, respect, concern, and understanding. These constitute, then, rewards for the approach responses of the client. The therapist presents a nonthreatening relationship. It is more than a situation in which the process of extinction operates, where clients find that the therapist does not act toward them in the negative ways in which others have acted or continue to act toward them. They are actually rewarded for efforts to relate to another.
The therapist must begin reinforcement of those behaviors approaching or approximating the desired behaviors. Even slight, tentative, or hesitant relating behaviors must be met with warmth and empathy, with positive responses. Truax and Carkhuff state that, for most effective results:
„The therapist would tend to offer relatively high levels of warmth and empathy to any and all verbalization by the patient (random continuous reinforcement) and also, on a selective basis, provide higher levels of warmth more frequently to the patient’s attempts toward human relating. Thus the effective therapist might respond with empathy and warmth 40 percent of the time to definite attempts at human relating. As in all learning situations, for „shaping“ to be effective, the reinforcement must begin with the responses currently available in the patient. Thus, responses with even the slightest resemblance to attempts at human relating would immediately be followed by a relative increase in empathy and warmth. With some patients this would particularly mean that even negative or hostile reactions to the therapist would be met by heightened expressions of warmth and accurate empathy. Over time, as the patient consequently begins to relate more frequently, more closely and more effectively, the therapist should become more selective in his offering of warmth and empathy. The criterion for reinforcement is gradually raised as the patient is better able to relate. This suggests that the therapist, to be more effective, would offer high levels of warmth and empathy more and more selectively across time in therapy.“ 24
This passage suggests that while the therapist begins by being unconditional (providing continuous reinforcement for any client responses), he or she should move toward discrimination, or conditional reinforcement. The implication is that the therapist should consciously and deliberately deliver reinforcements on a contingency basis. However, attempts to do this in research, either in interview situations or counseling analogue studies, indicate that it is not possible for therapists to discriminate and respond to, or to recognize and reinforce, predetermined classes of behavior with accuracy and consistency. Skinner has noted that „the contingencies of reinforcement which are most efficient in controlling the organism cannot be arranged through … personal mediation…. Mechanical and electrical devices must be used. Personal arrangement and personal observation of results are unthinkable.““ He is referring to teaching in a class of children, but the same can be said for the individual personal situation. Thus, conscious, deliberate attempts to reinforce specific client responses are not very effective, since to be effective every (or almost every) response must be reinforced, and this is not humanly possible. Moreover, the conscious and deliberate dispensing of predetermined rewards or reinforcements leads to an unnaturalness and lack of spontaneity or genuineness that can dilute or destroy their effectiveness as the client becomes aware of their artificial nature.
However, the therapist is not unconditional in his or her responses, even in the beginning of therapy. The therapist is more concerned about some of the client’s verbalizations than about others; he or she is more interested in some topics than others. It is well known that what clients talk about is related to the theoretical orientation of the therapist. If the therapist values dreams, clients dream and report their dreams. If the therapist is interested in sex, clients talk about sex. If the therapist deals with phobias or fears, clients present these to the therapist.
The presence of minimal levels of the facilitative conditions provides a nonthreatening atmosphere in which the client can begin to relate to another person. If the client is to remain in the therapy situation, the lack of threat must continue. Thus, it would appear that minimal levels of the core conditions must continue throughout therapy. In fact, contrary to the implication in the quotation from Truax and Carkhuff that the more effective therapist would be highly conditional, Truax later found that the overall level of facilitative conditions in group therapy was independent of the degree to which the conditions were offered differentially as reinforcers. 26
While there are those who maintain that an „optimum“ level of anxiety (which is never defined) is necessary to keep the client working on his or her problems, I suggest that there is sufficient anxiety present within the client, without the therapist adding to it. The therapist
„must discriminate between those anxiety changes evoked by the patient’s self-exploration and those evoked by his own threatening or inept attempts at interpretation or empathic responses. If the therapist fails to make this discrimination it is unlikely that he will prove helpful. It may be best to assume that whenever the patient shows an increase in anxiety and withdrawal, the therapist’s response (or lack of response) is the source of the anxiety arousal.“ 27
Reinforcement of Self-Exploration
The therapist does respond, intuitively, to some activities or verbalizations of the client and not to others. After the relationship begins, and after the opening remarks of the interview sessions, the therapist does not respond to social conversation by the client. As indicated in Chapter 6, such material is considered irrelevant. The therapist does respond to self-talk, since this is assumed to be relevant. At the beginning of therapy, such self-talk is usually of a negative character. The therapist thus is reinforcing negative self-references. But negative self-references are only one subcategory of the self-reference category. So when, as normally happens, positive self-references begin to occur, these also are reinforced. But self-references are themselves only a subcategory of a broader kind of client statements-self-exploratory statements. The therapist, in responding to the several categories of self-exploratory statements, participates in a process known as shaping. In this process, in the absence of the desired behavior (here self-exploration), behaviors that approximate this response (here spontaneous verbalization) are reinforced; when behavior that is closer to the desired behavior occurs (such as self-talk), it is reinforced. Eventually, the client engages in self-exploration, which is then reinforced by the therapist, leading to self-exploration at deeper and deeper levels.
Self-exploration is often a painful or anxiety-arousing process. Thus, even though it is reinforced by specific therapist responses of interest, acceptance, and understanding, clients may become anxious or threatened by the nature of the material they are exploring. The therapist must be careful that his or her responses do not increase threat. Probing questions and interpretations are likely to do this. They cause the client to retreat-to resist, the phenomenon noted by interpretative therapists. It is desirable that the client be helped to explore as deeply as he or she can at the moment. The context of the facilitative conditions maintains a safe nonthreatening atmosphere for this. This atmosphere can be considered „a counterconditioning agent, a pleasant stimulus to be paired with the mildly anxiety-provoking cues associated with the leading edge of the conflict region,“ or the self-exploratory process. 18 in this respect, the nonthreatening therapy relationship is similar to the inducement of relaxation by behavior therapists. In fact, „it seems probable that eliciting warmth and comfort as responses incompatible with anxiety and fear and avoidance produces more potent and adaptive counterconditioning than deep muscle relaxation.“ (29)
The warm, accepting, nonthreatening therapy environment would appear to be a more relevant condition for reducing or extinguishing anxieties related to social or interpersonal processes than physical relaxation. However, as Truax and Carkhuff suggest, clients may have developed conditioned anxiety responses to „human warmth“ (probably phony or manipulative in nature), so that the therapist should be cautious about offering strong, overwhelming warmth at the beginning of therapy.
The client, then, engages in the process of self-exploration in a warm, relaxing, nonthreatening atmosphere. Nevertheless, the process arouses anxiety as it progresses; the client stops, retreats, rests, takes a breather by moving into irrelevant, or less relevant ‚ areas. Self-exploration is a step-by-step process, or a two-step forward and one backward process-an approach-avoidance situation. If the therapist, by his or her specific responses (including concreteness and confrontation), helps clients explore themselves in the presence of the facilitative conditions, clients progress. The therapist’s accepting, warm, empathic responses are rewarding. Clients feel satisfaction in their achievement. Anxiety related to the topic that they are exploring is reduced.
Behavior therapists emphasize that they are systematic in their rewards or reinforcements and, while recognizing that other therapists utilize reinforcement, insist that it is unsystematic.“ But it has already been suggested that it is extremely difficult, if not impossible, to really be systematic in dispensing rewards in the ongoing complex process of psychotherapy. (The process of systematic desensitization is systematic, but this is a process of extinction, not of positive conditioning.) I suggest that the relationship therapist is more systematic in his or her reinforcements intuitively and spontaneously than it is possible to be by taking thought and attempting to provide reinforcements on a conscious, mechanical basis.
Martin describes this intuitive reinforcement process when he says:
„The client is progressing through an „anxiety hierarchy“ as he moves to more and more anxiety-provoking cues, and the nature of the hierarchy continually changes as more and more cues become available to increase his understanding. A consequence of this is that an anxiety hierarchy established at any point in therapy will necessarily be at least partially incorrect. The process of therapy I am describing permits the therapist and client to modify continuously the nature of the hierarchy, as correction of it becomes necessary with new knowledge. Clients do indeed progress in therapy along an anxiety hierarchy, but it is a hierarchy that the client and therapist discover together during the therapy process. The client will attempt to take the next step on his hierarchy. The therapist’s job is to hear this attempt and to facilitate it by responding to the leading edge of what the client is trying to deal with.“ 31
The therapist’s empathic responses facilitate self-exploration. Being understood is rewarding. The therapist’s empathic responses are conditional. But they vary not by the intent or design of the therapist. They are contingent on the client’s success in clarifying his or her self-exploration. While minimal levels of warmth, acceptance, and concern maintain the client in his or her self-exploration, highly empathic responses reward especially successful or clear self-exploratory efforts. The therapist cannot be highly empathic if the client is extremely vague and confused. He can only respond to, or reward, the effort. But when the client engages in clear, specific, highly insightful self-exploration, the therapist is able to respond with high levels of accurate empathy, which is highly reinforcing to the client.
The recognition of the influence of reinforcement provides the counselor with a way of meeting the problem of responding to the silent or highly nonverbal client. Counselor responses of attention, interest, concern, and understanding to the verbalizations of the client are rewarding and encourage the client to continue self-talk and self-exploration. Continuous questioning by the counselor will not accomplish this. In fact, it will teach the client to be responsive to the counselor and to wait for the counselor to continue questioning or initiating the interchange.
Recognizing the principle of reinforcement also suggests that the questions of the counselor should be directed toward eliciting self-talk by the client rather than factual information or personal history material. It is not necessary-or perhaps even desirablethat the questions be of the type frequently recommended by instructors in counseling-that is, questions directed at past or present feelings, such as „How do you feel about that?“ or „How did that make you feel?“ Contrary to the assumption that these questions lead to the expression of real or basic feelings, it is more likely that they lead to the expression of considered or intellectualized feelings. The direct focusing upon feelings by the client, rather than the unconsidered spontaneous expression of feelings, introduces a cognitive element that leads to talk about feelings rather than the expression of feelings.
The objection has been raised that encouraging client self-talk and self-exploration would, by generalization, lead to the client imposing such talk on those he or she interacts with outside therapy. However, a moment’s thought should lead to the recognition that, if such talk were not responded to positively, it would be extinguished, and clients would learn to discriminate regarding with whom they could engage in such talk. In this regard also, it would be anticipated that the generalization of clients‘ other learnings in therapy-understanding, respect, warmth and acceptance, and openness, honesty, and genuineness-would be responded to positively by others in their environment and thus be reinforced. Since such characteristics are responded to by the development of the same characteristics in those who are exposed to them, clients would experience an increase in these conditions in their everyday environment, which would enhance the therapy.
Elimination of Specific Anxieties and Fears
If the client has learned specific anxieties or fears in the area of human relating, these may be removed through desensitization or adaptation, reactive inhibition or internal inhibition, and counterconditioning. These processes occur in the traditional psychotherapies as a result of the nonspecific or general nature of the nonthreatening, accepting therapy relationship. The shaping process for progress in self-exploration would appear to fall under this learning modality of Truax and Carkhuff, at least insofar as specific anxieties and fears are dealt with. Where the client’s problem appears to be an isolated specific fear or anxiety, the use of techniques of behavior modification, such as systematic desensitization, would appear to be appropriate.
Reinforcement of Positive Self-Concepts and Self-Valuations
As clients progress in therapy, their responses become more positive in regard to themselves or their self-concepts. These responses are welcomed by the therapist, who sees them as desirable, as indications of progress. The therapist is thus pleased with the client, and indications of this pleasure and approval are communicated to the client. it is important to emphasize here that such positive client responses come late in therapy, as a result of the process of self-exploration. It is not desirable-or possible-to short-circuit the process by extinguishing negative self-references and reinforcing positive self-references from the beginning of therapy. It might be possible to condition clients to emit positive self-references in a relatively short period of time. However, this would simply be the conditioning of verbal behavior. Efforts to do this with „normal“ subjects in analogue studies have not always been successful. Whether this is related to a social set to be self-critical rather than self-praising, or to an innate resistance to glossing over negative feelings, is not known. The ignoring of negative self-references and the reinforcement of positive self-references prevents or inhibits the process of client self-exploration.
In addition to the specific behavioral learning process operating in counseling or psychotherapy, a more general learning process occurs. This is the process called modeling. Modeling, of course, is not strictly a behavioristic technique. The behaviorists did not invent it or discover it, and it did not arise from laboratory research, as the behaviorists are so fond of claiming for other techniques. Modeling is a basic and pervasive method of learning and, in primitive societies, constitutes the educational process, or the process by which the young acquire the behaviors necessary for survival. Modeling as a method of teaching and learning has the advantage over reinforcement of specific behaviors, in that it makes possible the learning of complex wholes, the acquisition of patterns of behavior, rather than the piecemeal learning of wholes by the prior learning of parts that must then be integrated. It also often leads to rapid learning, often one-trial learning, without practice being necessary.
Modeling involves the observation and imitation by the subject of another person. The process is a highly complex one, and its effects or results involve numerous factors, in the model, the subject, and the relationship between them.“ It is not necessary, for modeling to occur, that the model intend to model; he or she becomes a model for others in the appropriate setting.
Thus, the therapist becomes a model for the client. The attitudes of the client toward the therapist make him or her a model. Descriptions of modeling have focused upon nonverbal behaviors. But most of the behaviors in which we are interested in psychotherapy, at least in relationship therapy, are interpersonal behaviors that are verbal, or verbally mediated. The verbal behavior of the therapist in the interpersonal relationship of psychotherapy is a model of interpersonal relationships. Bandura suggests that since change agents model certain attitudes and social behaviors, these activities should be analyzed in terms of the behavior being modeled or its functional value for clients. He then criticizes „conventional interview therapists“ because they „mainly exemplify silence and interpretive behaviors that have limited functional value for clients. To the extent that clients emulate these behaviors in their everyday behavior they become pests or bores.““ This may be the case in psychoanalysis; I have experienced this in groups of persons undergoing analysis. But if we substitute listening for silence and empathic understanding for interpretation, the criticism is inappropriate. Listening to and understanding others are useful and facilitative conditions for all interpersonal relationships. The relationship therapist is modeling the facilitative conditions for good human relationships.
Relationship therapy is thus not a therapy in which clients only talk about their problems-„verbal substitutes,“ as Bandura calls it–but one in which clients deal directly with problem behavior, with difficulties in interpersonal relationships. Clients are thus actually doing what Bandura says social-learning behavior therapists do“devoting the major portion of their time to altering the conditions governing deviant behaviors rather than conversing about them.“ 34 It would appear that relationship therapy fits at least the first part of Bandura’s description: „A powerful form of treatment is one in which therapeutic agents themselves model the desired behavior and arrange optimal conditions for clients to engage in similar activities until they can perform the behavior skillfully and fearlessly.““
It may be questioned whether, if self-disclosure on the part of the client is desired, the therapist shouldn’t model this at the beginning of therapy. The answer, I think, is that this would be inconsistent with other objectives of the therapist-respecting the client, listening to and understanding the client, and not imposing himself or herself on the client. Therapy is for the client, not for the therapist. In his approach to groups, Mowrer does engage in self-disclosure as a model for the group members.“ Whether this is helpful in individual therapy (it is apparently not necessary) is not known“ (see pp. 80-84).
It may be helpful or facilitative for the client to have some pretherapy orientation or training in preparation for entering a therapy relationship, and this could include modeling of self-disclosure. Truax and Carkhuff found that such „vicarious therapy pretraining“ did facilitate therapy to some extent and resulted in greater improvement.“ Hoehn-Saric et al. developed a Role Induction Interview to influence clients‘ expectations about psychotherapy (including the behavior expected of the client and the therapist) and to prepare them for certain developments, such as resistance and the expectation that improvement would occur in four months. The interview appeared to be effective to some extent, including resulting in greater likelihood of the client continuing in therapy.“
Modeling is a complex process, and although it no doubt incorporates elements of conditioning, it is not explainable simply in conditioning terms. While vicarious conditioning (observation of the model being rewarded) may be an element, modeling is effective without this. Cognitive elements are certainly present. Truax and Carkhuff discuss the principle of reciprocal affect, referring to the fact that people respond with affect similar to that to which they are exposed. This would appear to be modeling. But the principle is descriptive, not explanatory.
In summary, it is possible to describe the therapy process developed in this book in behavioral learning terms. Conditioning, both classical and operant, is involved in the process. Whether this description is explanatory or sufficient to understand the total process perhaps depends on the point of view of the reader, or the breadth of his or her concept or theory of learning. At any rate, the description given here is incomplete. The process of psychotherapy is complex and cannot be reduced to simple stimulus-response components. Behavior therapists no longer attempt to do this; the inclusion of modeling is an expansion of the earlier, restricted behavioristic approach. Learning is a complex process involving more than stimulus response elements or conditioning. „Research has begun to make it increasingly clear that the learning in the behavioral therapies involves complex cognitive, emotional and motivational changes operating within a social context. “ 40
If the facilitative conditions can be described in behavioral terms, behavior therapy techniques can be described in terms of the facilitative conditions. Thus, Truax has indicated that (unpublished dissertation) studies show that „the level of empathy, etc., of the therapist is the crucial factor in traditional desensitization therapy. ,41 This, of course, does not show, or imply, that the process is not a learning process. It is only to say that facilitative conditions are part of the learning or behavior changes that occur in behavior therapy. Certainly, psychotherapy is not a simple conditioning process. The fact that behavior changes follow a reward does not mean that the therapeutic relationship is an automatic sequence of conditioning without awareness on the part of the client. It must be recognized that the extension of the concept of reinforcement beyond the simple conditioning that occurs in the laboratory leads to the introduction of other variables that influence and change the nature of the process. In complex situations what is labeled reinforcement may actually be described in other terms. What is termed a reinforcement may operate on a cognitive basis, such as providing information regarding the consequences of behavior.
Learning, it is being increasingly recognized, is a complex process, and a strict behavioristic interpretation is limited. The extension of the concept of reinforcement as a universal factor in learning appears to be impressive, but it must be remembered that such usage extends the term beyond its origins in laboratory situations. „No one will dispute that reinforcement is important. As Michael Polanyi once pointed out, since it includes anything that can change behavior, it must be.““
But this is only one way of describing behavior change, and it is not explanatory. There are different kinds or levels of learning. Razran describes four evolutionary levels-reactive (nonassociative or reflexive); connective (conditioning); integrative (perceiving); and symbolic (thinking) 43 The importance of the last two, or the cognitive aspects of learning, has been neglected by the behaviorists until recently.
There is a useful way of looking at the relationship between the ore or facilitative conditions and the techniques of behavior modification (such as reinforcement, desensitization, extinction, and mdeling). Rather than viewing them as separate or independent, or different approaches or techniques, they can be incorporated in a igle approach, in which the behavioristic techniques are seen as )des of operation of the conditions. Then it becomes clear that ,re is no necessary or inherent contradiction or inconsistency. The influence of the core conditions is mediated, at least in part, through behavioristic conditions. On the other hand, as is being increasingly recognized, the behavioristic techniques operate in the context relationship, and, in fact, their effectiveness depends to a great extent on the nature and quality of the relationship. Behavior therapists have rejected the medical or disease model pychopathology and psychotherapy.“ Traditional or psycho-dynamic psychotherapies, such as psychoanalysis, client-centered therapy, and existential analysis have been associated with the disease or medical model by many behavior therapists. But it should ar that relationship therapy is not based upon the medical In fact, it could justifiably be claimed that it is based upon the behavioral or environmental model, since the cause of emotional disturbance is seen as developing from the nature of the human environment, and its amelioration requires the modification of the human environment. However, the description in behavioral or learning theory terms is not necessarily explanatory. Although psychotherapy is a learning process, it is not simply stimulus-response learning or conditioning. Learning is much more complex than these simple theories would imply, involving complex cognitive and affective elements and a social context that influences the process. The client in the process is more than a rat in a Skinner box being taught tricks by operant conditioning.
Dr.C. H. Patterson