Tracey L. Stulberg


This paper provides a perspective on spouse abuse from an ecosystemic approach suggested by second order cybernetics. An attempt is made to demonstrate the interactional basis for an understanding of spouse abuse and to suggest that the failure to use an ecosystemic perspective may result in therapists and other helpers inadvertently exacerbating the problem. Borrowing from the work of Keeney and Bateson, an analysis of the relational aspects of the problem of spouse abuse is provided which illustrates the similar nature of typical intervention strategies to the problem.

Men who hit their wives, or more rarely, women who are physically violent with their husbands, have attracted wide attention in our society. Individuals in the legal segment, feminists, moral advocates, theorists, therapists, and politicians each have particular understandings about the causes of this phenomenon and how best to intervene. These understandings and related interventions, alone and in combination, are used to help women in both clinics and women's shelters. However, many of these interventions are designed to treat a particular individual while ignoring the social ecology that surrounds the problem. Some researchers believe that treating only part 0f the system may make the problem worse or create new problems for the system (Bateson, 1972; Bobele, 1987; Conran, 1987; Fisch, Weakland, & Sega), 1982; Keeney, 1983; Minuchin, 1974; Minuchin& Fishman, 1981; Watzlawick, Weakland, &Fisch, 1974). Many therapy groups for men who hit their wives report low rates of effectiveness and high rates of recidivism (Conran, 1987).

Individually oriented treatments are frequently used in working with spouse abuse in both women's shelters and other therapeutic settings. However, because the treatments that are used neglect to include the social ecology in which the problem is embedded, they may fall short of their goals. As Watzlawick, Weakland, and Fisch (1974) put it: dealing with a couple without taking into account the social ecology "is doomed to lead to nonsense and confusion" (p. 6). This failure to take the social ecology into account can be likely to be partly responsible for the exacerbation of domestic violence.

This paper proposes a cybernetic understanding, one which includes the social ecology, the treatment agencies, and therapists in understanding the phenomenon of spouse abuse. This paper takes the position that being mindful of the social ecology in which the problem of spouse abuse is embedded will assist therapists as well as the staffs of women's shelters in working with this problem. An alternative understanding will be presented to suggest how individually oriented approaches may fall short, and will point toward the development of alternative methods which would help increase successful treatment strategies in both clinics and women's shelters.

When spouse abuse is seen as a metaphor for the entire system the treatment options available are expanded and the cybernetic ethic of "acting always as to increase the number of choices" (Von Foerster, 1982, p. 308) is observed. Working with the social ecology, not just the individual increases the number of treatment options, thus helping to increase the potential for clients to reach their goals more quickly. This paper will offer an explanation for the inclination to neglect the social ecology, and provide a formulation of the problem of spouse abuse within the larger system. Such a formulation may be useful in the development of more effective clinical programs to treat spouse abuse.

Logical Mistyping

Logical typing and logical mistyping. Logical typing involves assigning meaning to perceptions. In the course of assignment of meaning perceptions are also accorded a place in a logical hierarchy of types, or orders of abstraction. In this manner a chair belongs to a class of chairs, which in turn belongs to a class referred to as furniture, which in turn may belong to a class called household items. This action is unavoidable as "the very process of perception is an act of logical typing" (Bateson, 1979, p. 210). However, when one assigns meaning to perceptions, there is always a chance that errors called "logical mistyping" (Bateson, 1979) will occur.

Logical mistyping occurs when a speaker confuses different orders of abstraction in thinking or speaking, but acts as if that confusion has not occurred. As a result, higher-order explanations may be offered for lower-order behaviors as if the explanations were the behaviors. For example, a police officer, social worker, or other helper arrives at a scene where a husband has raised his fist to his wife and in a loud voice has called her a "bitch." In the course of explaining to themselves, each other or the public what is happening the helpers may mistype the order of abstraction to which the action belongs. They may refer to the situation as the husband's "lack of control" or "aggressiveness" when the situation or the context of action is qualitatively different. In describing the scene they may report observing anger, spouse abuse, family violence, or some other abstraction. Leaping from the observation of the cited simple actions to reports of the higher order abstraction is an example of logical mistyping. The helpers have jumped from observing simple behaviors (raised fist and loud voice) to attributing meaning to those behaviors (anger), but act as if that jump has not occurred. While they may be able to see a complexity of interactions, they cannot see "anger." "Anger" is a punctuation of the observer (or helper), not an observable behavior. In an attempt at explanation "there is a jump here from particular to general, from member to class" (Bateson, 1979, p. 133). The helpers have used the member of a class of behaviors as if it were that class. This jump is not necessarily problematic when it is for simple descriptive or colloquial purposes. The difficulty arises from attempts to use these higher-order descriptions as treatment targets.

The use of explanations that result from logical mistyping will influence the type of assistance that helpers decide to provide. For example, if helpers who have walked into the scene described above have utilized an inborn, individual explanation and typed the husband's behavior as "anger," they have failed to account for the interactional context of the actions. These helpers may conclude that the problem rests within the man, so that if they "fix" the man, the problem will cease to exist. They may take the woman to a women's shelter for protection and support. They may arrest the man, or put the man in an "anger control" group. These helpers have used a simple, incomplete, individual description to explain an interactional phenomena. From a cybernetic viewpoint, then, spouse abuse is interactional, involving all parts of the system in its origin and perpetuation. Separating parts of a system and treating each part as if it were a distinct entity while ignoring the reciprocal processes of the family results from logical mistyping.

It is possible that interventions that follow from this logical mistyping, may, in the long run, increase the incidence of spouse abuse because such interventions ignore the response the other aspects of the system may make following the interventions. This ironic increase in violence may be understood as stemming from the direct and indirect messages that the treatment milieu may provide. For example, spouse abuse may be defined by helpers as a problem that results from a batterer's uncontrolled anger. Offering this explanation to the couple while labeling the man as a batterer, who has uncontrollable anger, may function like a hypnotic command (Keeney, 1983) to the client and his wife. That is, he may understand his "condition" as one, which is uncontrollable, in spite of attempts to "control" it, and thus not be seen by himself or his wife as responsible. Since the "anger" is responsible neither of the pair may be motivated to respond to efforts aimed at changing them.

Helpers may not stop spouse abuse by treating the man's "anger." Anger is not a simple, observable behavior but a category of actions. "The things which are categories of action do not obey the reinforcement rules the way action obeys the reinforcement rules" (Keeney, 1983, p. 34). By treating anger with methods that are based on teaching control, the helpers may teach the man how to be better at controlling. At one level then, he learns by example that controlling people is desirable, and at another level he learns more sophisticated control techniques that may become as abusive, or more so, than previously seen physical aggression (Conran, 1987).

A Dialectical Ladder of Family Violence

In order to illustrate the ease with which one can make jumps in logical typing a dialectical ladder (Keeney, 1983) of spouse abuse has been constructed (See Figure 1.). A dialectical ladder is a device that helps one analyze interactions and sort out orders of logical typing. This ladder can be seen as an "alternation between classification (form) and the description of process" (Bateson, 1979, p. 214) under different levels of abstraction or orders of recursion. Description of process refers to the "raw" data or operational descriptions of observed behavior. Classification of form refers to punctuations of patterns or symbolic descriptions by the observer for helper). Hence, this multi-leveled exchange can be conceived as a dialectical ladder between form and process (Keeney, 1983).

"Descriptions of simple action," or simple behavior, in this spouse abuse ladder include individual units of action such as the husband's raised fist and loud voice. On the same behavioral order, these ac-Figure 1tions can be typed by the observer as belonging to a "category of action" called anger. On a contextual level, "descriptions of interaction," or process, include the relationship between the simple actions. For example, the husband's actions may be followed by the wife raising her fist and speaking loudly, which may be followed by the husband telling the kids to leave, after which he strikes his wife, and his wife strikes him back. This particular "category of interaction," or context, may be classified by the observer as a symmetrical relationship (or more commonly, fighting, battering, wife beating, disagreeing, or spouse abuse). The relationship is symmetrical because it involves behavioral interactions such that, the more the husband displays particular behavior (e.g. striking the wife), the more likely the wife is to display a similar behavior (e.g., striking back) (Bateson, 1979).

On the same contextual level the descriptions of interaction could point to a different relationship classification of the simple actions (See Figure 2). For example, the husband's raised fist and loud voice would be followed by his wife telling him how much she loves him and in a soft voice trying to calm him down. This would be followed by the man telling the kids to leave and then striking his wife, who cries saying that she is sorry and that she loves him. This category of interaction may be classified by an observer as a complementary relationship (or once again, fighting, battering, wife beating, disagreeing, or spouse abuse). However, this relationship is complementary because actions of the husband and wife are different, but mutually complete one another (Bateson, 1979).

There is nothing about the symmetry or complementarity of the relationship itself that leads to spouse abuse. When interactions of either type continue unchecked, however, a runaway escalating process may ensue (Bateson, 1979; Keeney, 1983). In other words, when interactional sequences are solely complementary or symmetrical, there is a potential for the relationship to move toward the extreme. Hence, a symmetrical extreme may result in excessive competition, while a complementary extreme may result in excessive dependency-nurturance. These extremes, rather than balanced expressions of these types of relationship, can lead to runaway escalations or "schismogenesis" (Bateson, 1979). When schismogenesis goes unchecked, "intolerable stress and dissolution of the relationship system" results (Keeney, 1983, p. 40). Abusive relationships, therefore, can be seen as arising from repeated, unchecked symmetrical, or complementary escalation.

Dialectic of Complementary Husband-Wife Interaction Figure 2 It is often at the point of schismogenesis that people become involved with social service agencies or helpers. This involvement may result from the couple's initiative or because the situation has come to the attention of the community. It may be that the couple seeks help for a relationship they fear will dissolve, or a well-intentioned relative or neighbor may call an agency and request help for the couple because of similar fears. These helpers may offer these couples explanations (e.g. uncontrolled anger, poor impulse control, or excessive dependency needs) and advice as part of a therapeutic effort to assist the woman and her husband. In the process of offering explanations for particular behaviors, they often unknowingly confuse different orders of abstraction. When different orders of abstraction are confused in this way, the helpers have engaged in logical mistyping by using a member of a class (hitting someone) as if it were that class (anger). This then leads to influencing clients to commit the same error in logic. In this manner clients as well as helpers focus on changing a category of action such as violence, anger, or impulse control, which does not respond to the same rules of reinforcement as does simple action. Or they run the risk of succumbing to "hypnotic" phenomena to which Keeney referred.

Family Violence in its Recursive Whole

When helpers focus on only one part of the "dysfunctional" system, and the involvement in therapy of themselves and the wider system is not made explicit or even seen by the helper as part of the therapeutic whole, systems thus viewed are called taciturn systems (Pask, 1969). As a taciturn system, helpers "act as though (they are] distinct from the system of interest . . . and overlook any ongoing interaction between operator and machine" (Keeney, 1983, p. 75). As noted, treatment that ignores the larger social context may fall short of treatment goals and may, on occasion, actually exacerbate the problem.

Unfortunately, efforts to help sometimes are multiple and fragmented, involving several uncoordinated entities. For example, a woman was sent to a women's shelter by the commander of a nearby military base because of what was described by the commander as a "spouse abuse" problem. The problem came to the commander's attention through other military personnel who were neighbors of the couple. Based on their own perception, understanding, and explanation of the problem the base commander, the shelter's social worker, volunteers, and others immediately and independently began their own variations of therapy with the woman. The base commander with advice from his staff ordered the husband to remove himself from the home for the time being. The shelter staff began their work in trying to build the women's self-confidence by initiating a program of individual and group counseling with her.

Helpers in the approach described prematurely offered solutions to the woman's "problem" without first considering three very impor-tant points: First, how does the woman define the current situation? If the woman did not come into the shelter on her own because she thought that there was a "spousal abuse" problem, are helpers treating a problem which, for the woman, is not important enough to her to elicit her cooperation? This is not to say that the situation is not problematic for the victim; it is possible, however, that she sees the efforts by the staff and the base commander to be off target, and perhaps meddlesome.

Social problems are what people think they are and if conditions are not defined as social problems by the people involved in them, they are not problems to those people, although they may be problems to outsiders or scientists (Fuller & Myers, 1941, p. 320).

The outsiders, or helpers, have cultural values and perceptions that define certain situations as problems. When presented with a situation, they attempt to help by offering solutions to the woman. However, "cultural values obstruct solutions to conditions defined as social problems because (helpers) are unwilling to endorse programs of amelioration which prejudice or require abandonment of their cherished beliefs" (Fuller & Myers, 1941, p. 320). Thus helpers are often inhibited from effective helping because of strongly held values. The cultural value which supports seeing behavior in non interactional, unidirectional mode is one of many which is at work here.

Perhaps in the context of this particular couple, the behaviors of the couple that helpers would describe as "abuse" are described by this client in other, more positive terms. For example, it is not unusual to hear women who are in shelters say "I know that when he hits me it is because he loves me" or "if he didn't get so angry at me, I wouldn't be sure that he loved me." These alternative realities may not make sense to social service agencies or the reader, but they do make sense to the women who share them. Other commonly heard perspectives are that this woman is likely to define the problem to herself as an inability to please her husband or as a husband who is under a lot of stress at his job and is unable to help himself. Frequently women express the understanding that it is not a problem with her husband but some other factor such as alcohol, drugs, or in-laws.

Although these perceptions may not match the helpers' worldviews, they are valuable clues to the problem for which the woman might want help with in order to stop being abused. It may be that the pejorative remarks made by these helpers about her husband or her situation may discourage her from pursuing further help. Spouse abuse may be a larger sounding problem than the one that she thinks that she has. A dangerous consequence of failing to consider the helpers' ideas and part in this interaction is that the woman may think that the helpers do not understand her situation and may decline their help and return prematurely to the same situation. Also to be considered is the danger that the involvement of the outsiders may enrage the husband further.

Second, when helpers neglect to take into account the woman's perceptions, they run the risk of creating a problem, which for that particular system does not exist, while neglecting the problem that does exist. There is a circular, or recursive, relationship between problems and solutions. It has been said that problems arise out of people's attempts to solve them, while solutions arise from people experiencing the problem (Keeney & Ross, 1985). Both problems and solutions are ways that observers describe their experience. When helpers attempt to solve problems described as spouse abuse, they may indeed create another problematic situation or help to maintain an existing problem. If the possibility exists that well-meaning helpers may create a problem, then therapists need to take into account themselves and these other helpers who become as integral a part of the therapeutic process as they have become part of the problem situation. In fact, they may not only be part of the problem, but, as will be illustrated, their intervening may be the situation that puts the woman in greater danger (Bobele, 1987).

Third, while social service agencies are extremely helpful, their recursively related influence to the problem is often overlooked. To illustrate the relationship between the social service agency and the wife another dialectical ladder has been constructed (See Figure 3.).

On a behavioral level, descriptions of simple action could include the helper saying to the woman "You need to leave your husband because he's a batterer." These simple actions could be described by an observer as belonging to a category of action called helping behavior or counseling. On a contextual level, descriptions of interaction might include the woman responding to the helper by saying "I can't leave him because I love him." The helper may respond with "yes, but if you go back he will continue to beat you and he might kill you." The woman may respond to this by trying to explain that she loves him and that she knows that he loves her, and if she tries real hard, she will not upset him again. This category of interaction may be conceived as a symmetrical relationship; the "interaction can be described in terms of competition" (Bateson, 1979, p. 192). In this case the competition involves each in trying to convince the other of the correctness of their view while denying the other's point of view. As it may occur to the reader, the symmetrical relationship the helper has established with the woman may be similar (isomorphic) to the symmetrical relationship the woman has with her husband. There is a sense of competition for the correctness of a point of view, that if goes unchecked, may lead to a symmetrical escalation. This symmetrical escalation between the husband and wife may have led the woman to the women's shelter. This same symmetrical escalation between the woman and the helper may be what leads the woman away from the shelter and back to her abusive situation.

In a similar manner the descriptions of interaction could include a different set of behaviors, behaviors of a complementary nature (See Figure 4). The helper's statement "you need to leave your husband because he's a batterer," could be followed by the wife saying "you are right, I need to leave him. If I leave him what can you do for me?" The helper may say "we will help you to get legal assistance, help you get an education, get you counseling and provide shelter for you and. . ."

This category of interaction may be classified by the observer as a complementary relationship. The interactional sequences of the helper and the wife are "different but mutually fitted to each other" (Bateson, 1979, P. 193). The complementary relationship that the woman has with the helper may also be of the same type that the woman has with her husband. There is sense of over dependency and submission on the part of the woman, and too much willingness on the part of the helper to manage the woman's life. Such a situation can result in the woman's extreme dependency on the helper, like the husband, that when unchecked, can also lead to schismogenesis. This complementary escalation between the husband and wife may have driven her away from her husband. It sometimes happens that helpers become frustrated or discouraged with women who become overly dependent on them and encourage them covertly or overtly to leave the shelter because they are not seen as motivated to change themselves. This same complementary escalation between the woman and helper may drive her away from the shelter, discouraged about receiving help.

Not only can the symmetrical and complementary relationships between the wife and helper escalate and lead to schismogenesis, but they may also create problems for the woman that were previously nonexistent. In the symmetrical relationship, the wife and helper could continue to compete, and the wife could then go back to her husband. During her next symmetrical escalation with her husband she may call him a "batterer." This is new "ammunition" that she has picked up and is now using in interactions with her husband. The husband may get even more angry at the wife for allowing a stranger to talk about him to his wife in this way.

In the complementary relationship, the woman may indeed leave her husband. In the process the helper may continue to show nurturance and dominance while the woman continues to display excessive dependency and submission. Finally, the helper may become so frustrated with the woman's dependent behavior that she stops helping the woman, and the woman leaves the shelter with no more survival skills than when she arrived. It is possible that the woman will return to the previous relationship. There is obviously a recursive relationship between the social service agency and the wife. I do not mean to imply here that the pattern of interaction in the home is always duplicated in the shelter. As a matter of fact, it is possible that a complementary relationship in the home and a symmetrical relationship with helpers or vice versa can also result in an oscillation of the woman between home and shelter.

Second-order cybernetics, or "cybernetics of cybernetics" (Mead, 1968), indicates that the helper is included as part of the system which contains spouse abuse. By including the role of the helper in this analysis of the situation, it may be that helpers can assist their clients even better in making desirable changes in their situations, while maintaining their organization as a couple (Keeney, 1983). Some might object to the maintenance of the couple's relationship as a desirable goal, and sometimes it may be that one of them may decide to end the relationship. However, if the position of the helper is that the termination of the relationship is the primary goal, it may alienate the client who does not share that goal, as well as other members of the community who may need assistance with cases of domestic violence. Pastors, for example, may be reluctant to refer troubled couples to a shelter where the divorce is perceived to be a frequent solution to such problems (Bobele, Langford, &Todtman, 1987). The couple may not get needed help because a referral is not made.

When working with couples described as abusive, therapists could include in the therapeutic process various parts of the system that are involved in trying to solve the problems such as neighbors, law enforcers, the legal system, the shelter staff, and the other shelter women, for instance. It may not be necessary to literally bring them all into a therapy session, but accounting for their influence in planning and conducting treatment is invaluable. By interacting with the system's wholeness in this way, therapists may come to recognize that the "total self-corrective unit which processes information, or, as one may say thinks and acts and decides, is a system whose boundaries do not at all coincide with the boundaries either of the body or what is popularly called the self or consciousness (Bateson, 1972, p. 319). The unit of treatment may be more properly seen as an ecosystem, rather than a single individual or couple in the ecosystem.


This paper has examined from the vantage point of second-order cybernetics many problems faced by helpers in working with spouse abuse. Inclusion of the wider social system that contains the problem is critical in understanding the dynamics of spouse abuse and its treatment. It is hoped that questions about clinical implications were raised in the reader's mind, although the purpose of this paper was to discuss theoretical issues involved in spouse abuse.

In brief, it is important for helpers and therapists to become more aware of their systemic connection with the couples they strive to assist. Becoming aware of this connection has implications for how helpers conceptualize what to do with a particular part of the system in a particular context in relation to the larger system.

While there is no one "correct" theory that conceptualizes the phenomenon of spouse abuse, just as there is no one correct clinical model from which to work with these couples, there are models which may be more helpful in working with various sorts of problems. This paper suggests a paradigm for the treatment of spouse abuse that may extend the effectiveness of clinicians and other helpers in the treatment of spouse abuse. These helpers may become more effective by avoiding the errors in logical typing that permeate the field of spouse abuse research and treatment. Such errors are seen as potentially dangerous in that they may force helpers into the same schis-mogenic escalation with their clients that the clients have been involved in with their spouses. Therefore, if schismogenic escalations with husbands lead women into shelters, it may be possible to assume that the same schismogenic escalations with the helping agency may exacerbate the problem or create new problems.


Bateson, G. (1972). Steps to an ecology of mind. New York: Ballantine.

Bateson, G. (1979). Mind and nature: A necessary unity. New York: E.P. Dutton. Bobele, M. (1987). Therapeutic interventions in life-threatening situations.Journal of Marital and Family.Therapy, 13, 225-239.

Bobele, M., Langford, D., &Todtman, D. (1987). Fundamentalist pastors as referring persons: An ethnographic study. Unpublished Manuscript, Texas Tech University.

Conran, T. J. (1987). Controlling the controlling and overpowering the powerful: A language of violence intervention. Unpublished Manuscript, San Antonio: Union Graduate School, Union of Experimenting Colleges and Universities.

Fisch, R., Weakland, J. H., &Segal, L. (1982). The tactics of change: Doing therapy briefly. San Francisco: Jossey-Bass.

Fuller, R. C. & Myers, R. R. (1941).The natural history of a social problem.American Sociological Review, 6, 320-329.

Keeney, B. P. (1983). Aesthetics of change. New York: Guilford Press.

Keeney, B. P. & Ross, J. M. (1985).Mind in therapy: Constructing systemic family therapies. New York: Basic Books.

Mead, M. (1968). Cybernetics of cybernetics. In H. Von Foerster, H. Peterson, J. White & J. Russell (Eds.), Purposive systems. New York: Spartan Books.

Minuchin, S. (1974).Families and family therapy.Cambridge: Harvard University Press.

Minuchin, S. & Fishman, H. C. (1981). Family therapy techniques.Cambridge: Harvard University Press.

Pask, G. (1969). The meaning of cybernetics in the behavioral sciences. In H. Rose (Ed.). Progress of cybernetics.New York: Gordon & Breach.

Von Foerster, H. (1982). On constructing a reality. In H. Von Foerster (Ed.), Observing Systems. Seaside: Intersystem Publications. Watzlawick, P.

Weakland, J., &Fisch, R. (1974). Change: Principles of problem formation and resolution. New York: W. W. Norton.