[First published on Nathen’s Miraculous Escape, December 8, 2009.]
I consider the theoretical conversation about intervention at the level of systems versus individuals, including systems holarchies, pathology, the instrument fallacy (“When you have a hammer, everything looks like a nail.”), and the fit of evidence. I include an argument for family-systems interventions as both the underdog and the eventual winner.
Intervention at the Level of Systems versus Individuals
Entering into the field of psychotherapy at this moment in history means coming to grips with an ongoing conversation or debate between those who favor intervening at the level of systems, usually families, and those who favor intervening at the level of individuals. The conversation is mostly one-sided right now because of the history of psychotherapy. Psychotherapy arose under the western European enlightenment, with a modern value structure and philosophy. This means it was formed on the foundation of Christianity, reductionist science, and medicine, and so focuses on the individual using the modern lens of linear causality. It was also formed during a time which had only very primitive understandings of psychology, communication, and statistical modeling. Family therapy and systems-oriented interventions flowed from the more recent scientific and philosophical developments of general systems theory and postmodern philosophy (see Becvar & Becvar, 2006, or Nichols & Schwartz, 2008 for descriptions of this development), with at least two unfortunate results. First, those in the dominant paradigm of individual psychotherapy and especially psychiatry have been able to largely ignore the arguments and breakthroughs of family systems thinkers. Second, the arguments of the family systems paradigm remain a bit reactive and dualistic, along the lines of, “individual interveners have missed the boat, because problems simply do not exist on the level of individuals.” That, of course, is an oversimplification for the purpose of a strong thesis. In fact, the individual and family are convenient points of intervention on a larger continuum from biological to public policy, and it seems to remain to be seen which of these sets of tools provides the most effective intervention for various kinds of human suffering. In part this is because the toolboxes of each intervention-style camp come part and parcel with a set of philosophical information filters that make the tool wielder right about the question at hand, as if the tools in your toolbox actually shape the question into a problem best handled by your system of intervention. It is also in part because the available modes of observation and evaluation, the science and statistics of very complex systems, are still in their infancies, because they are both new and contain very difficult problems. Because of all of this, it can be difficult for the various factions to even agree on adequate outcome variables.
A Holarchy of Systems
One way to conceptualize the nature of systems is as a holarchy, a series of levels of increasing complexity in which each, new, more complex level of systems both transcends and includes all of the previous levels (see e.g. Wilber, 1996). Families, for example, are not just more complex systems than individual humans, they are a more complex kind or level of system. Families are made up of individuals and their interactions. They are systems of individuals. Individuals could (conceivably, at least for a time) exist without families, but families would immediately cease to exist without individuals. Therefore, families represent a higher order of complexity than individuals do, both transcending and including them.
Individuals are systems too, of course, made up of biological systems like organs and tissue, which in turn are made up of chemical systems, each level of which both transcends and includes all of the simpler levels of systems. There are levels of systems more complex than the family, as well; cities, for example, or countries, perhaps even the biosphere (see e.g. Capra, 1997). Even though this is a big oversimplification, it can be a useful way to think about things: systems all the way up and all the way down.
So when we consider where pathology exists, or where best to intervene, the question is not really whether to intervene at the level of individuals or at the level of systems, but rather at what level of systems does pathology exist and at what level of systems will it be most effective to intervene. It is only a matter of convention that by “individual” we tend to mean individual humans, and by “system” we tend to mean families and up. Additionally, it is a convention of family therapists to mean “family system” when we say “system,” as that is our area of primary interest.
What and Where is Pathology?
“Pathology” is a dirty word in holistic, post-modern circles, which is understandable. It smacks of the medical model, with its simplistic, linear causality. It has also been used to justify an incredible amount of oppression and suffering. After all, who gets to say what is normal and who is sick? “The man” does. Family therapists tend to use the word “problem” instead of “pathology” and let themselves off the pathologizing-hook by allowing the client to define the problem: The problem is whatever the client is complaining about, and we are given permission to intervene by their coming to us for help (e.g. Fisch, Weakland, & Segal, 1982). This is a bit disingenuous, however, because outside of the strict behavioral camp, family therapists tend to believe that the real problem exists meta to the complaint. That is, it exists in the level of system above the complaint, in the outmoded rules of the family, for example, or in dysfunctional patterns of communication. This allows us to not pathologize individual humans, but we are still saying that something is not working right, and that is what other disciplines call pathology.
The real question about pathology is where it exists, at what level of complexity. It is possible that a problem could exist at just one level of complexity—in the family system, for example, or in an individual’s neurochemistry. It is also possible that a problem could exist in multiple levels simultaneously, or even that a problem at one or two levels could cause symptoms at other levels that look like problems. This gets us into an epistemological morass, because we cannot actually see problems. We can see patterns and behaviors, and we can hear complaints, and we can notice when those behaviors and complaints change, but that tells us little about what level the actual problem exists at. For this we have to rely on theory.
Intervention and the Problem of Having a Hammer
A theory is a story about the underlying nature of reality. It explains things we can see in terms of concepts that we cannot see, in terms of ideas. An ideally rigorous scientist never believes their theory, seeing it instead like a constructivist does: the best map I have made or come across so far for the territory of reality. Real scientists, and other real people, like therapists, who imagine they see their theories be useful over and over, come to some degree to mistake their map for the territory; my map, not yours, shows the real lay of the land, and if we are talking about problems, my map shows where the root of the problem is.
The trouble is, theoretical maps tend to be very subtle and sophisticated representations of only one level of system. They tend to have some idea of the levels just above and just below, but in terms of the language of their level. A therapist who specializes in individuals, for example, is trained to see how problems exist at the level of individual humans. They might know that depression, for example, is correlated with certain kinds of brain chemistry, and maybe even certain kinds of patterns of interaction in the family—they may even know that depression is more common for people in certain economic situations—but they will tend to see these sub- or extra-individual phenomena as symptoms of the real problem, which lies in the individual—perhaps in a traumatic past, psychodynamic factors, or lack of insight into their own thought processes. From this standpoint, giving an individual drugs or intervening with their family seem to be treating the symptom rather than the problem.
If you have a hammer everything looks like a nail, and if you have a theory every problem can look subject to intervention at your favorite level of system. Also, if you know where problems happen, you know where they do not happen. People who specialize in one level of system often have the overt belief that if everything is going well at their level, the other levels can take care of themselves. Family therapists tend to see individual psychological health as a given in the context of a well-functioning family (see e.g. Bateson, Jackson, Haley, & Weakland; Bowen, 1960; Jackson, 1998; Minuchin, 1998; 1963 Satir & Baldwin, 1983 ). For political thinkers, who tend to see problems in higher-level systems, at the level of public policy interventions, making people happier by any means lower than institutional-level interventions is just perpetuating the problem; of course people are depressed! We are living in an unjust political or economic system (famously, e.g., in Marx & Engels, 1967). Giving them happy drugs, intervening in their family systems—these are just ways of maintaining the status quo. It is different, of course, with political thinkers whose preferred system is in place. In that case, the problem is obviously at any other level of system. Medical thinkers have a similar situation, only from a bottom-up perspective, instead of top-down; depression is the result of a chemical imbalance, pathology of the tissue of the brain. Happy drugs are not treating the symptom, they are treating the root of the problem, and the symptoms of discomfort at the level of the individual or of strife at the level of the family will just go away (see e.g. Watters & Ofshe, 1999). Many psychological “diseases” from depression to schizophrenia tend to run in families, too, so medical-thinkers also blame genetics (Carlson, 2005).
It is noteworthy that being a systems thinker does not necessarily ameliorate this problem. Our theories are much too flexible for that. Systems-hip people with bottom-up oriented theories can say that they are changing the system using their individual intervention, and they probably are, to some degree. The nature of systems is one of reciprocal causality; if you change the output of one node in the network, you will theoretically change the output of every other node, and therefore the functioning of the entire system. It is rare even for family therapists to insist on seeing the whole family anymore (Nichols & Schwartz, 2008). While he preferred to see families, Murray Bowen, for example, saw individuals for a large part of his practice, and thought that this could “be the fulcrum for changing an entire family system (Nichols & Schwarts, 2008, p. 145). Some family therapists consider seeing only the complainant as ideal, and in no way contradictory to systems principles (Fisch et al., 1982). In that light it is difficult to argue that medical or individual interventions are not also interventions at the level of systems, as long as they change the client’s behavior.
The Problem With the Evidence
Can we settle this by looking at the evidence? Maybe someday, but not today. The main problem is that our science has not caught up to even Freud’s thinking and style of intervention, much less to the leap into complexity that is systems thinking and intervention. Up until the last couple decades, even if we could have collected the data we needed, we could not have analyzed it with the primitive statistical modeling available. Even now we are left with a serious conceptualization problem and a massive data-collection problem.
In order to study things scientifically, whether it is gravity or psychological health, we need to operationalize our latent construct of interest. That means, since we cannot directly observe gravity, our construct of interest, we have to decide on something we can directly observe to act as a proxy. We could measure how long it takes an apple to fall ten feet, for example, or measure how hard it hits the ground from various heights. From that data, we can make inferences about gravity. For psychological health, it is the same but more difficult. We cannot see it, so we have to choose a proxy. That proxy usually turns out to be some form of asking people how they feel. Sometimes we use the number of rehospitalizations or type of diagnosis, but most often we give the person a survey such as the ubiquitous Beck Depression Inventory, asking them what they have been doing and how they have been feeling lately.
These kinds of measures we have are best suited to test the success of medical-level interventions. If you give someone a drug and they say they feel better, they are better! What more do you need to know?
Those who think of problems existing at the level of the whole individual do not have it quite so good. The reason is that, while asking someone if they feel better or checking whether or not they have been rehospitalized is a decent operationalization of individual health, it cannot distinguish between the “real” health that an individual earns permanently through the work of therapy, and the “fake” health handed to them by externally manipulating their brain chemistry. An individual therapist who looks at data supposedly showing that fluoxetine is in any way comparable to therapy (e.g. TADS, 2004) must chafe. The client has not learned anything! What can this be but a band-aid?
Family therapists have it even worse. Anything short of a family-wide, long-term operationalization of health is inadequate if you believe that symptoms such as depression are just a way that the family system is attempting to maintain its homeostasis, staying within its outmoded parameters. If you “fix” one person in the system without altering the rules of the system, someone else is bound to become symptomatic. You need second-order change, change at the level of the rules, rather than first-order change. But how do you operationalize second-order change? How do you measure the rules of the system, which play out in the complex interactions between body language, spoken language, and emotional tone, between interlocking triangles of relationships? Well, you cannot. Family-systems thinkers are left in the uneasy state of relying on individual outcome measures to judge the success of our work.
In a way, those who prefer intervention at higher levels than families have an even harder time with outcome measures. These systems are almost impossibly complex—whole cities, whole countries—how to operationalize the health of a county? Public-policy interveners tend to be idealogs, too, which makes things more complex still; your version of health will depend on the economic or political system you prefer. Still, these problems may not bother public policy types. Ideological immunity, the tendency or ability to completely disregard information that contradicts your belief system, may limit their interest in evidence, except where it serves their purposes.
But perhaps there is a sort of levels-of-systems-wide equifinality. Perhaps intervening at any level produces the same kinds of results. This is what the small amount of evidence that has been gathered seems to suggest, though it is all based on individual-outcome measures (Sprenkle, 2002). So perhaps there is no such thing as a system-level-specific pathology. Is it even possible to determine the level at which a pathology exists? If so, would there be a way to determine whether it was better to intervene at that level, or a level above or below? It is difficult to imagine the kind of nightmare-monster experiment that could answer those questions to everyone’s satisfaction; a longitudinal study, spanning the entire lifespan of multiple generations, collecting rich data on every complaint, and the timing of every complaint, of every member of hundreds of families of different ethnicities, existing under different political and economic systems, using different levels of intervention, and controlling for nutrition, exercise, values structures, belief systems etc ad nauseum. It would be fairly expensive, and somewhat difficult to analyze the data. Modern research tends to shy away from even relatively simple complexities, such as comorbidity, where an individual is diagnosed with more than one pathology (e.g. TADS, 2004).
There is some anecdotal evidence, but it is, well, anecdotal. Reading biographies of therapists (e.g. Family therapy pioneers, 2008), for example, gives the impression that many people left the ranks of individual intervention to work with families because individual work was too slow, or ineffective, while there are no famous cases of family therapists who became individual therapists because the work was too shallow. This could be evidence that individual therapy is in fact inefficient or ineffective, but it could simply be the result of individual therapy having been formulated first; many, many times the number of people have been trained as individual therapists compared to family therapists—almost infinitely more, back in the day that the pioneers of family therapy were emerging.
Don Jackson, one of the founding fathers of family therapy, claimed to have noticed regularly in his individual therapy practice that when he helped solved a problem for one client, someone else in the family would develop a problem (Jackson, 1954, 1965). That this is a reason to intervene systemically is a commonly held belief among family therapists, but virtually unknown among more bottom-up helping professions, and there appears to be no published empirical evidence on the problem, and none may be on the way. It seems likely that family-systems thinkers will have to continue to rely to a large degree on our intuition about the legitimacy of our theories and the applicability of research based on individual outcome measures and the validity of relatively anecdotal evidence from our founders and mentors.
An Argument for Family-Level Interventions as Both the Underdog and the Eventual Winner
For Judeo-Christian culture, the individual is the obvious focus of attention. The individual is the seat of the soul, the source of the will, and the unit that can be judged innocent or guilty by the courts, and will be judged saved or damned by God when that day comes. It is only the individual who seems to have consciousness and a linear narrative of life. It is the individual who has nerve endings and can complain about them. Who was it that said, “Ouch—I have a pain in my family”? I forget, but I bet it was a family therapist, because it is simply not intuitive to do so. Western ethical and philosophical systems all focus on the individual: What is it right or wrong for you to do? In what fashion can you live the best life? When systems are considered, we jump right up to the level of government and public policy, and even then, they are primarily concerned with the relationship of the state with individuals: What are the individual’s rights? When can the state take an individual’s property?
The development of modern science in the enlightenment began with reductionism, in the form of chemistry, linear causality, in the form of Newtonian physics, and a dualistic model of health, from medicine, which were and remain incredibly powerful and successful forces for gaining knowledge. In part, this is because the questions that chemists and physicists ask are simple ones: What are the basic laws of nature? What is this thing made out of? The kind of science that could grapple with more complex topics such as the nature of communication, or the function of the behaviors of the “mentally ill” were for centuries, and to some extent still are, far beyond our capacity for scientific observation, data collection, and statistical modeling.
Because of all of this, family-systems theory and its foundations, postmodern philosophy, general systems theory, and communication theory, are late to the table and sitting across from respected, entrenched adversaries. Postmodernism thoroughly discredits the idea that an observer can be objective. Systems theory makes taking things apart and using idealized, linear models seem quaint and often misguided. Communication theory holds that the difference between a schizophrenic and a non-schizophrenic may be the relationship between spoken language and body language in that person’s family—a very difficult thing to quantify. Family-systems theory would have us believe that, while we might be able to drug ourselves into happiness, we can’t drug our way to health, because the real problems are in the patterns of communication of the family. These ideas are still all quite counterintuitive from the modern mindset.
Still, I imagine that the newbies will be vindicated eventually. First, they are more sound. Postmodern philosophy is simply more accurate than modern philosophy; that we have no direct access to anyone else’s experience or to any other piece of reality has ramifications that modern philosophy simply ignores. Systems theory is truly a breakthrough in science, and very promising; it takes into account the way complex systems are more than the sum of their parts in a way reductionism is simply not geared to. Communication and family-systems theory are ways of looking at health and pathology from a more accurate perspective; medical thinking is too black-and-white, health-or-pathology, and too reductionistic. Treating the symptom with drugs or surgery is simply not always the answer. Objects in the frame cannot change the frame (Watzlawick, Bavelas, & Jackson, 1967), that is the allopathic fallacy. Second, if psychological problems really do exist in the context of systems, in patterns of communication of individual humans, then we don’t have to kill the patient to look inside it; we can observe the problem in live action without cutting anyone open. The guts of the problem are right there in front of us, in the therapy room, in the interactions of the family. That is a huge advantage over people who think that problems exist in individual brains, which must be either dead, in grave danger, or isolated in a very expensive machine fMRI machine which still cannot see much. Third, even though we have to use inadequate, individual outcome measures to test our work, we are already coming out on a par with individual interventions (Sprenkle, 2002). Imagine how we will do comparatively when outcome measures regularly consider the health of the whole family! Last, we are just getting started. Freud lived 150 years ago, and his ideas took off immediately. They have had time to mature. Family systems theory is barely 50 years old and has gotten a relatively slow start. Who knows what we will be able to accomplish by the time our field is mature?
And Yes, Individual Versus Systems Psychotherapeutic Interventions Look Quite Different
The first psychotherapeutic intervention to take off was Freud’s psychoanalysis. Over the course of years, patients (they still called them “patients”) would explore their past and especially their unconscious mind, increasing their insight, using free association and dreams, under the direction of the expert analyzer, the therapist. The patient would come to think of the analyst as their father and eventually get over that idea, in the process transforming from pathologically neurotic to mildly neurotic.
The second great force in psychotherapy was behaviorism, which was largely a reaction to the heavily subjective, unquantifiable theories and process of the Freudians. Behaviorists ignored what might or might not be going on in people’s minds and concentrated on what they did. One of them noticed that dogs could be conditioned with rewards to salivate to a bell and a series of interventions based on that insight were created. If you want to increase a behavior, you either reward or stop punishing the organism when they perform it. If you want to decrease a behavior, you either punish or stop rewarding the organism when they perform it. Simple!
The third great force was humanism. The humanists did not like how the behaviorists treated humans like just another animal. They also did not like how the Freudians dwelled so much on the past, pathology, and intellectual analysis. They had a positive, emotional spin. Humans were naturally healthy, not naturally sick. Mildly neurotic was shooting too low. They shot for self-actualization. Psychological problems came from repressing emotions and not being understood. They also believed in the talking cure, but they talked about what was happening right now, in the moment. Their interventions looked like a series of interviews in which the client and therapist formed and explored an intense emotional relationship, and in which the therapist coached the client to be totally genuine.
The systems approach (the fourth great force?) was a reaction against all individual therapies. It can be seen as a resurgence of behaviorism to some degree, in that it treated individuals’ minds as black boxes, at least in the early days (Watsalwick et al., 1967), but always with a systems, family-oriented view. Later, some models incorporated humanist principles (e.g. Baldwin & Satir, 1987; Whitaker & Keith, 1982). There are a great variety of systems interventions. The only constants among them is that they concentrate on process over content—that is, they pay more attention to the rules of the conversation than the direct messages in the conversation—and they will always at least consider intervening with more than one family member at a time. Therapists in the strategic school tend to use paradoxical homework as interventions—prescribing the symptom, for example. Bowenian therapists also use homework in the form of reconnecting with cutoff family members. Experiential therapists use the relationship with the therapist in the humanist style, plus coaching authentic communication. Structural therapists also use homework to re-establish healthy boundaries between parents and children, going on a secret date, for example. All of these are methods of causing an updating of the rules of a family system.
A Personal Reflection
I am happy to find myself fascinated by this topic—deeply and richly confused, as a friend of mine likes to say. I also feel a little embarrassed to have written so much after reading so little, which is to say that ten weeks is not much time to make oneself an expert in such a deep topic! That, and I feel grateful that there is so much to read—that so many very smart people have thought so much about how best to help people. In the end, I still don’t know the answer to any of the questions I came across here. But I have chosen my theoretical stance anyway. The systems view is my hammer, and I am learning to swing it the best I can.
Baldwin, M. & Satir, V. (Eds.) (1987). The use of self in therapy. New York: Hawthorne.
Bateson, G., Jackson, D. D., Haley, J., & Weakland, J. (1963). Toward a theory of schizophrenia. In N. J. Smelser & W. T. Smelser (Eds.) Personality and social systems (pp. 172-187). Hoboken, NJ: John Wiley & Sons.
Becvar, D. S. & Becvar, R. J. (2006). Family therapy: A systemic integration. Boston, MA: Pearson.
Bowen, M. (1960). A family concept of schizophrenia. In D. D. Jackson (Ed.) The etiology of schizophrenia (pp. 346-372). Oxford, England: Basic.
Capra, F. (1997). The web of life. USA: Anchor
Carlson, N. R. (2005). Foundations of physiological psychology. Boston: Pearson.
Family therapy pioneers. (2008, September/October). Family Therapy Magazine, 7(5), 23-60.
Fisch, R., Weakland, J. H., & Segal, L. (1982). The tactics of change: Doing therapy briefly. San Francisco: Josey-Bass.
Jackson, D. D. (1954). The question of family homeostasis. Psychiatric Quarterly Supplement, 31, 79-90.
Jackson, D. D. (1965). The study of the family. Family Process, 4(1), 1-20.
Marx, K. & Engels, F. (1967). Manifesto of the communist party. New York: Penguin.
Minuchin, S. (1998). Structural family therapy. In D. M. Frank (Ed.) Case studies in couples and family therapy: Systemic and cognitive perspectives (pp. 108-131). New York: Guilford.
Nichols, M. P. & Schwartz, R. C. (2008). Family therapy: Concepts and methods. Boston: Pearson.
Satir, V. & Baldwin, M. (1983). Satir step by step: A guide to creating change in families. Palo Alto, CA: Science and Behavior.
Sprenkle, D. H. (2002). Effectiveness research in marriage and family therapy. Alexandria, VA: AAMFT.
TADS (2004). Fluoxetine, cognitive-behavioral therapy, and their combination for adolescents with depression. JAMA, 292(7), 807-820.
Watters, E. & Ofshe, R. (1999). Therapy’s delusions: and the exploitation of the walking worried. New York: Scribner.
Watzlavick, P., Bavelas, J. B., & Jackson, D. D. (1967). Pragmatics of human communication: A study of interactional patterns, pathologies, and paradoxes.
Whitaker, C. & Keith, D. V. (1982). Symbolic-experiential family therapy. In J. R. Neill & D. P. Kniskern (Eds.), From psyche to system: The evolving Therapy of Carl Whitaker (pp. 330-378). New York: Guilford.
Wilber, K. (1996). A brief history of everything. Boston: Integral.