From Movements to Moments

A paper for Enfield Family Therapy Conference "Power, knowledge and expertise in contemporary family therapy: have we lost the baby with the bath water?" April 30th 1999

Imagine, if you will, this scene from nine centuries ago. The massed armies of Christendom have reached the walls of Jerusalem terrifying its Muslim inhabitants. A deputation of knights rides up within shouting distance of the city and their leader calls out: "We would like to draw your attention to the benefits of Christianity so we've brought you some literature. But no pressure; just think about it for a while. Thanks for listening - we'll go home now."

Of course, it didn't happen like that. Crusades don't work that way. Nothing less would do than to re-claim Jerusalem for the one true faith and if almost the entire population of the city needed to be put to the sword along the way - then so be it. Looking back we might well ask - why did these knights bother - there wasn't much money in it and there was every chance of dying of sickness or wounds thousands of miles from home. Historians tell us we have forgotten the fear of hell that was the constant background to everyday life in those times and crusades offered the ultimate kind of security - membership of a righteous movement - an almost guaranteed front-row seat in heaven.

I'm talking of pre-modern times when you didn't just do what you were told by the religious law or its priests - you were what you were told. But the excitement and zeal of contemporary movements, even postmodern ones, often have this pre-modern side-effect - a sort of callous and destructive neglect of that which existed before.

"Throwing the baby out with the bath water", is of course a wonderfully hackneyed phrase but it does capture something of this neglect - like a scene from a Punch and Judy show; both jolly and chilling. But we must handle this imagery with care. Invoking our wish to protect babies from a callous enemy is an old rhetorical device for stirring communities to take up arms. The best antidote to the side effects of a movement may not be - as I will argue - either a mobilisation into an opposing movement or a defensive retreat into the past but a new kind of thoughtfulness.

What I want to do in the first part of this Conference paper is to offer a deconstruction of some developments in the theory and practice of family therapy. By deconstruction I mean a questioning of some of the thinking which is becoming relatively fixed and nonnegotiable around terms like "expert", "first order", "knowledge", "power" "etiology" and so on. In the second part of the paper I want to offer some other kind of reconstruction or gathering together of old and new ideas. I will talk about these through developments in my thinking and practice. Throughout, I will be welcoming the challenging thinking brought by the cybernetic and postmodern movements while seeking to counter what I fear to be their destructive side-effects.

I am going to define two major changes of bath water or as Thomas Kuhn [1962] in his Structure of Scientific Revolutions might call them - paradigm shifts. First was the emptying out of psychoanalytic oriented approaches in the name of circularity. (I'm talking now of the cybernetic movement beginning in family therapy in the early 60s.) At that time, if an effect was seen as produced by a cause, then that was "linearity" and we didn't want it. The bath water of apparent psychodynamic linearity contained many interactional concepts - projective identification being one - ... but, more than that, it contained a developing language to think about the meeting between people at a deep level. The later unavailability of this language has been, I believe, a substantial loss to mainstream family therapy.

I would define the 2nd change of bath water - one we are still living through - as the draining off, not just of early cybernetics, but of all notions of objective knowing. A driving force for this change is the triple wave of constructivism, social constructionism and postmodernism. I am getting a bit bored with this baby metaphor now but let me squeeze it one last time and say that the biggest baby we are in danger of losing in this second change - a sort of meta-baby if you like - is the search for explanation and understanding.

The first from Alan Parry in 1991 "It may yet be realised that Freud's greatest discovery was not of the unconscious but of the validation that a person receives simply in telling her story to an attentive listener." All Freud had discovered, Parry is saying, was how to listen attentively. Oops .. there goes 100 years of psychoanalysis. We may vary in how much we regret this attack on our neighbour - but Parry is surely saying something to family therapists as well; that the surface narrative needs to be heard and understood - but nothing else.

And Lynn Hoffman in a 1998 paper refers to etiology (the interest in causes) in family therapy as the "blame and change template". She says ".. a who-done-it question has been built into the family therapy movement which has had a chilling effect on families and practitioners alike. During the movements early years it seemed that mothers were at fault. Then whole families came under the gun. With the advent of feminist therapies the onus shifted to husbands and fathers .. But the moving finger moved on to distinctions based on ethnicity, class, and race. Ideas about the etiology of emotional problems have been extended to the larger cultural discourses that entangle us all. p146.

Considerations of causality are then for Hoffman nothing more than pejorative finger-pointing.

And we have Michael White at a UK conference - shortly after a major crime of violence which had hit the headlines - speaking of a "psychopathologists field day", summoning up, for me, the image of a frenzy of the kind of finger-pointing that Hoffman alludes to.

There is nothing in these accounts of a benign role for theory, explanation, hypothesising. In other words there is nothing in it for our clients; it serves only our narcissistic need for prestige, certainty and professional security - an opportunity to stare from our superior expert pedestals at other peoples deficiencies.

Should we then abandon any attempt to theorise, hypothesise - to attempt to do any more than know how to listen, deconstruct, and free family members from the oppressive expert discourses of others? If knowledge is power then are the utterances of therapists always acts of oppression? Are all accounts that we can give about a family or families nothing more than "just our stories"?

I want now to look at these difficult terms: "Power", "knowledge", "expert". What do they mean?

The first thing I want to say about these terms is that I think they have almost no generally useful descriptive value. For example, if parents were offered the choice of meeting with a therapist with these attributes or without them my guess is that many would vote for the former. "Yes please, I think I would like a therapist who is an expert and who has power and knowledge." Most therapists, I suspect would be much more equivocal. Many clients may use these words to denote a resource, many therapists might use them to denote oppression. These same words would be used - in Wittgenstein's later philosophy - in completely different language games and would have no shared meaning. So if a client and therapist discussed these issues from within those respective definitions they would not be able to understand each other.

Rather than considering these terms as descriptive we should instead notice how they perform in family therapy discourse - especially when left undefined. At best, they are warnings to us to be alert and to tread carefully; they are signs that say "Minefield ahead". At worst, in my view they become badges of shame - to control what is acceptable behaviour if we want to call ourselves a family therapist or systemic practitioner or narrative therapist. This operates through the disciplinary power of discourse in which we become cooperative "docile bodies", as Foucault puts it. They are part of the means by which we become subjects of family therapy culture through internalising its implicit rules of acceptability.

If we want to do more than have these words as shameful controlling introjects then we need to give ourselves some chance of having a meaningful dialogue. This, I think, will only be possible if we can signal the specific way we are using them, otherwise we may simply talk past each other. But we don't have to insist on a single definition that lasts for ever. It is enough to define these terms both locally and temporarily because we may want to use each one in several ways today and to switch from one usage to another. We just need to try to alert each other when we do this and to watch out for the disputes that arise when, as protagonists, we fail to notice our differing definitions of the same word.

The first notion I want to consider is that of the "expert therapist". In one common usage (or language game) among family therapists, "expert" stands for a person who believes she can objectively describe what is going on, can compare her observations to a store of specialised objective knowledge (preferably scientifically validated) and can confidently define both the problem and the solution without much reference to the subjective view of the client. Notice that the idea of a collaborative therapist is usually defined implicitly against this definition of expert. The term used this way is, mostly, a caricature; a bogeyman evoked in contemporary family therapy rhetoric, but we might come close in a positivist who believes that diagnostic categories (like conduct disorder or personality disorder) have an existence completely independent either of her subjectivity as an observer or of the community of ideas in which her subjectivity is constructed.

I doubt anyone here aspires to be this kind of bogeyman expert. We might all agree that this would be a bad thing. But before we get too comfortable about this it is worth considering these two arguments:

First, an ethical argument . Given only a choice of two is it better to operate as an expert even perhaps a manipulative expert - and provide the help wanted by family members or is it better to work openly, collaboratively and avoid manipulation at the price of providing more limited help? Andrew Collier [1987] in discussing the strategic use of paradoxical injunctions in the book Ethical Issues in Family Therapy suggests that if a paradoxical injunction binding the family into a change that they come to welcome was more productive of this change than working collaboratively, then there is an ethical argument in its favour. A collaborative therapy could be claimed to be the unethical choice under those circumstances. I find this to be a disturbingly unfamiliar argument in contemporary debates - that a helpful therapy is more ethical than an unhelpful one.

Secondly, we should reconsider Maturana's dictum that - according to his version of constructivism - there can be "no instructive interaction." This was widely used in the late 1980s to argue against the role of the powerful expert in family therapy. But, I think this misses the main point. If we accept what Maturana says then we no longer need to worry about this reason for acting like an expert. Being an expert is not unethical - merely impossible, since people don't change through being told what to think. If change happens in the so called first-order therapies (structural, strategic, early Milan) then it occurs through some more subtle process of collaboration - which Maturana [1988] calls structural coupling - and not through the authoritarian power of the expert at all. A similar case is made by Efran and Clarfield [1992] in Therapy as Social Construction. They suggest that encouraging therapists to be collaborative or social constructionist would be as logical as encouraging porcupines and salamanders to observe the principles of natural selection. These ideas are not prescriptions of how we should behave but descriptions of how things are - of how language and culture produce us.

Should we then stop worrying about acting like an expert? I'm working my way towards saying that we should still worry (it is always worth worrying) - but we should worry about it differently. The expert's opinion will be right only if the family can collaboratively make something useful out of it. But our bogeyman expert won't see things that way since her diagnosis is made on "objective" grounds. And, since there is no place for subjectivity, the fact that her opinion doesn't fit the subjectivity of the family would be insufficient reason for the expert to change her mind. The danger is, therefore, not so much the authoritarian power of the expert position as its rigidity. So I would argue that what is helpful or unhelpful in therapy is much more usefully related to a notion of rigid vs. flexible than it is to the more familiar notion of expert vs. collaborative.

Brian Cade in an article in 1992 calls himself "an unashamed expert". He is prepared to stand up to a regular tyranny in therapy and the social sciences - that of fashion. We should admire his courage and his willingness to consider that the concept of expert is still worth arguing about. But we should also notice that he is using the word "expert" differently to the way I have used it in the bogeyman version above. In a 1996 article when discussing the topic of persuasion by the therapist he says this:

"Pragmatically, persuading [family members] is limited by the fact that self-generated arguments are far more influential than arguments produced by [therapists], and it appears that the more arguments that are self-generated in favour of a position the more likely it is that the position will persist. It also appears that when considering a series of persuasive messages, [family members] remember with more clarity their own thoughts and arguments than the messages themselves.." p245

Brain Cade may not like me to say so but here he and Maturana are talking about much the same thing. The power of the expert therapist is not so much to be feared as to be debunked as a myth. This kind of power doesn't change people except perhaps through compounding a feeling of hopelessness. It helps or hurts through a process of collaboration.

Yes, I think we should worry about our potential rigidity when we chose to become powerfully persuasive with our clients but I think there is yet a greater danger in that which Foucault [1975] calls disciplinary power. Defining a difference between power-over-others (what we might call repressive power) and disciplinary power helps us to draw a distinction - for example in racism - between, on the one hand the repressive structures of disadvantage, discrimination and persecution which are the constant grinding exterior to the lives of black and Asian people in this country and, on the other hand, the kind of unexamined internalised racism that can create white subjects as indisputably superior and conversely can construct black and Asian subjects as inferior - like the 6 year old child I saw recently who asked his white mother to help scrub his skin so he could be a good pale colour like his sister and friends.

Words in family therapy discourse tend to have a gloss of approval or disapproval depending on fashion. Words like "hierarchy" or "boundary", for example, sound rather tarnished these days. Compare them with "collaboration" which currently glows with our warm mutual acceptance. But there is danger in this. I would like to add another much older gloss to the term - one that would have been familiar to those living in occupied countries of Europe during the 2nd world war when "collaboration" was feared and hated. Disciplinary power in therapy works not through the boot of oppression but through a collaborative taking in and identification with implicit and unexamined assumptions of the therapist. In Michael White's [1991] words: "[T]his collaboration is rarely a conscious phenomenon. The workings of this power are disguised or masked because it operates in relation to certain norms that are assigned a "truth" status."

Which do we think would be more powerful, a) for the therapist to give her expert opinion or b) for the therapist to hold back on her ideas and instead ask a circular question? Current thinking might steer us towards the latter - indeed I have a colleague who trained fairly recently at one of the advanced family therapy courses in London who tells me she was discouraged from ever openly sharing ideas with the family and encouraged, instead, only to ask questions. I think that a Foulcaldian view of power would surely consider the asking of a question to be much more insinuating of the therapists assumptions than a straightforward declaration of those assumptions. (See also Golann, 1988.) A therapist who says, "if you and your partner stop arguing do you think your daughter or your son would feel happiest about that" is inviting the parents to concentrate more on the difference between the son and the daughter and less on the difference between arguing or not arguing. The question assumes that arguing causes unhappiness but without saying so. This, to my ear, is strikingly close to the Eriksonian hypnotherapy technique in which the recipient is provided with an illusion of alternatives.... something like this - ( said to a child) - "Do you want to have green vegetables for lunch or for supper?" Apparently very democratic. This hypnotically avoids (in theory anyway) the more profound and difficult question - "Are you ever going to eat anything which is coloured green?"

Anyway back to the couple. A therapist who says "I'm thinking about your arguments and wondering how big an effect they are having on your children - I know that children can get to feel pretty helpless sometimes" (an expert-sounding statement) is at least putting something up for thinking and discussion between the therapist and the parents. In the circular question the implicit assumption that arguing is causing the children to be unhappy gives no way forward for a dialogue around the issues of arguing in this family. At worst it might lead to a situation in which the couple feel more guilty each time they argue without quite understanding why.

I've said already that if the ideas of the therapist have no fit for the family then an overtly powerful therapist is, at worst, likely to be an irrelevance. However the concept of fit offers no protection in collaboration. For example, a guilty (or self-attacking) family member may find something in the conversation to confirm this self-attack and leave feeling much worse. I recall one parent who told me she lay awake for several nights reproaching herself over the possibility of her daughter being angry with her following something I said which, I thought at the time, had been put quite neutrally and in the context of a supportive relationship. I hadn't, at that point, understood the mother well enough to appreciate her profound sense of having failed her daughter. Here collaboration is working at a level out of sight of the therapist and indeed much collaboration of either the helpful or unhelpful variety operates at this covert or even unconscious level.

To summarise this point, we cannot make people think or behave in ways that don't fit for them. All influence operates via a process of collaboration which, in itself, provides no guarantee of helpfulness.

Those of you familiar with social constructionism may recognise that, at this point, what I am saying is pretty close to some social constructionist critiques. Why bother with explanatory models at all since they are potentially so insinuating of the therapist's agenda and collaboration per se is no protection? Indeed, why not just listen attentively and warmly support everyone's viewpoint? Note that even asking questions is no protection against the malign possibilities of clients colonising themselves with our normative assumptions since our curiosity does not thrive in a vacuum but is connected to our own values, theories and experience. We give ourselves away so easily.

Fortunately, there is a guaranteed way of avoiding this kind of malign collaboration which I can exclusively reveal to you here. Phone in sick then stay in bed and read a novel or watch daytime TV. Under no circumstances allow yourself within the sphere of influence of a client or family. Failing that, there is a less certain method of supporting resistance to disciplinary power. I'll take an analogy from advertising since advertisers understand perfectly well the constitutive powers of discourse. During the time I've been writing this paper there has been a certain small French car being heavily advertised on TV (I'll give you a clue - "Size matters!"). Imagine the effect on sales if no sophisticated sexual imagery seeking to bypass our thinking processes was allowed and instead all that was permitted was a brief notice that read as follows:

Men, are you looking to buy a small French car but are worried that it will make you feel sexually inadequate?

Let us reassure you that buying our small French car is the psychological equivalent of adding 4 cms to the length of your penis.

When a communication is put into words it encourages us to think. And I imagine thought in this case could lead to a bit of detumescence in the sales figures. Critical thinking offers some resistance to the disciplinary power of discourse. So my account so far of the dangers of collaboration has left something out: the capacity of people to reflect upon their situation, to ponder, to play with ideas, to select and discard and to make use of others in this process when there is a safe place for thought.

I hope now I have prepared the ground for one of the major points of my paper. My main critique is of the "persuasion" of the therapist.

I have deliberately chosen an ambiguous term to discuss this issue. I want to use its ambiguity in the way that railway points operate by swiveling from one track to another. The original track is our consideration of whether we should be as frightened as we are of the personal persuasive power of the leadership therapists like Minuchin, Madanes, Haley, Palazolli, Whitaker and many others from that era to which we now refer un-affectionately as "first order family therapy". I have argued that our fear of this kind of persuasion of the therapist has been overdone. And the result of these fears has been, as we know, to fairly successfully remove these models from the therapeutic marketplace.

I now want to switch to a separate use of this term - persuasion of the therapist. I am referring to the definition of "persuasion" that my dictionary calls "a state of being persuaded: a settled opinion: a creed". A therapist of persuasion in this second definition is a therapist who seeks authority or legitimacy for what they do by reference to a particular model or movement.

I think it is generally quite a good idea, though, to be both for and against most things in therapy. To have a working position and simultaneously a critique of ones' position to keep doubt alive. So before I use my critical voice, let me first make the case for movements and models and I will take one aspect of psychoanalytic culture as an example. That culture, as I am coming to appreciate, is very diverse (like all cultures) so any general statement about it would be a misrepresentation. But in my view it does, in some quarters, include strong conservative, authoritarian and elitist tendencies. There is, then, something tremendously refreshing about family therapy's willingness to allow waves of new thinking to wash over it periodically. As Stephen Frosh [1998] recently pointed out (and I agree with him) - family therapy now has quite an impressive record of allowing itself to worry about issues of gender, sexuality, class, racism, ethnicity and expert objectivism.

And I can hardly conceive of a training in family therapy that attempted to do without some period of concentration on models. A too-early eclecticism in which the separate coherence of the models has not been digested gives not so much a rich and flexible choice as a baffling and rather frightening labyrinth of possibilities.

Now I'll use my critical voice to speak against movements or more specifically against their pre-modern crusader side-effects. Movements can richly stimulate our thinking but they can also shut it down. How do they do that?

First, they appeal to our need to belong. If you go to a football match or tennis match and watch the crowd you will see that the committed supporters have far more fun than the neutrals - they feel part of something much bigger and more powerful than themselves. Movements offer an experience of undiluted belonging and this feeling of cohesion is easier to achieve when you have an opponent. A fashionable movement gives an opportunity to belong to something fresh, exciting and quietly superior. No-one wants to be a train-spotter (except perhaps train-spotters) and I guess we would all recognise that early cybernetics have, these days, about as much fashionable appeal as the bobble hat.

Better still, a revolutionary fashionable movement allows us to hide our wish for belonging and superiority behind a commitment to the disadvantaged whom we uniquely know how to help in a way that our authoritarian forebears did not.

Secondly and more simply, movements offer the hope of freedom from too much uncertainty and ambiguity. Being on the right side appears to offer protection from the fear of getting it wrong; of not being able to work out what to think or do for the best. Frosh in a 1997 paper, amply illustrates this point about the defensive retreat into pre-modernism in his discussion of the rise of religious fundamentalism and nationalism as a reaction to the shifting sands of postmodernity.

And third is the point I have been making throughout - the effects of disciplinary power in which we don't so much join a movement as have it reshape our identity. Although I have come to realise that psychoanalytic thinking contains potentially rich resources for family therapists, much of my identity as a therapist has been formed in family therapy discourse, a good deal of which is, at least implicitly, antagonistic to what psychoanalysis is - or to what family therapy thinks it is. That antagonism is now in me too, so every time I mention psychoanalytic ideas within the hearing of family therapists I feel guilty - in the same way that for ten or more years after I knew I didn't believe in god I still felt there was an outside chance of eternal damnation. As Boszormenyi-Nagy [1986] might say, these loyalties reach down deep.

If movements have pre-modern side effects then models have modernist side effects. While movements are driven by the righteous nature of the quest, models are content to be the most objectively rational players on the field. When the question is posed "How best to do therapy?" one's favoured model will have arrived at an answer superior to any other model. It will have the most logically coherent framework, or the best empirical evidence, or the most elegant techniques or the most impeccable ethics. I've separated models from movements here but, more likely, model choice is mixed in with some devotion to the movement and the pre-modern idealising of its priesthood.

OK. Let me acknowledge that in playfully highlighting these issues I have probably overstated the problem. I will put it more moderately - that problems of power in therapy are linked to model and movement loyalty. I define a therapist of persuasion as one whose loyalty can lead her into two such problems:

1. She will know in advance what she is going to do and think regardless of the contingencies of the moment. In other words she will, in some areas, be relatively rigid.

2. She may be relatively unwilling to make her working assumptions available for critical reflexive examination. The price of her loyalty is to accept such assumptions as givens. They form the powerful preconscious agenda of the therapist against which clients have less resistance since they permeate the session but are not offered up for thinking.

To be clear, I am drawing no distinction here, for example, between a classical analyst who interprets everything through drive theory, a structural therapist who sees only subsystems, hierarchies and coalitions and a narrative therapist who is preoccupied with dominant and subjugated knowledges.

I find Lyotard's [1979] definition of postmodern - "incredulity towards metanarratives" - to be most helpful in this argument. In plain language, postmodern is here defined as a position of scepticism to the claims of universality that models and movements make. From this postmodern position I doubt that models and movements can be trusted to legitimate the knowledge and activity of therapists. That legitimacy has to be found elsewhere - it has, as I will suggest, to be worked out in a moment to moment dialogical process between therapist and family members: a process that Lois Shawver [1998] - borrowing a term from Lyotard - calls paralogy.

To summarise again: the concern about expert power and control in family therapy has acted to remove many family therapy models and, more importantly, much family therapy thinking from the marketplace. But the most insidious power is not to be found in the charismatic persuasive power of therapists. Instead it slips under the dialogue between therapist and the family. So collaboration, which has broadly emerged as a solution to the problem of potential abuses of power is more centrally defined as their most likely carrier. I have begun to make the case that disciplinary power is more likely to hold sway first when therapists - perhaps through fear of appearing like experts - do not share openly their ideas for playful consideration and, secondly, when therapists seek legitimacy for what they do in movements or models.

What I have been doing so far - deconstruction - or raising doubts about established beliefs and practices is relatively safe. I'm now going to offer some reconstructions of my own (or should that read - "put my head on the block".) I will sketch these ideas out under the loose headings of epistemology, theory and practice but all these issues naturally interweave so I won't be able to keep them apart so neatly.

In a [Pocock] 1997 paper I began with the analogy of the border between two countries and used it to symbolise the relatively flimsy construction that keeps the cultures of family therapy and psychoanalysis safely apart. When I thought about it again recently, I realised how much the metaphor "border" - a single dividing line - had altered my memory. Now I recall that border, like many such borders, does not consist of a single fence or wall but instead the perimeters of each country sandwich a narrow strip of nameless territory where no-one belongs. It is a buffer zone which reduces the intensity of differences between the nations. In this zone nothing is built, nothing much is allowed to grow and it is the kind of place that you can get shot at by either side.

I will use this revised analogy to talk about the dispute between social constructionism and realism. Let me once more tie my use of these terms down with definitions. Social constructionism, while not necessarily denying the existence of a real world, suggests that who we are and how we are in that world are made by the social processes inherent in culture and language. (At the moment I will include much postmodern philosophy within this definition.) I am using the term realism to denote a philosophical position that assumes that while we cannot have a perfect objective knowledge we can come to better know the world as it is. This dispute is not just theoretical but underlies the potential - or in some places actual - split in many disciplines including family therapy and is, perhaps, one of the main reasons for holding this conference. (For our purposes I am defining first-order theories and psychodynamic theories as realist since they are attempting a description of the world. Those quotes from Parry, Hoffman and White, if you remember, are all in opposition to realist ideas of causality.)

In a thoughtful chapter in Ian Parker's [1998] recent book - Social constructionism, discourse and realism - Kenneth Gergen, social constructionism's main protagonist in psychology and a veteran battler with realists, readily acknowledges the high drama of being committed to a movement, and the delightful - albeit impossible - prospect of arguing one's opponents out of existence in academic debate. "Failure to participate in that drama is" he says "to miss one of the greatest pleasures of academic life".

But need this split be as meaningful in practice as it has been in the academic world? I'll give an example. Last year in the Guardian I read a report on a research study in America in which male subjects who described themselves as heterosexual were shown erotic pictures of men and the amount they were sexually aroused physiologically was then measured. The research came up with the rather satisfying finding that the participants who were most aroused were those who had identified themselves most firmly as heterosexual and were, frankly, homophobic. I'm afraid the full significance of the report didn't occur to me at the time so I don't have a reference, but assuming my memory of it is reasonably accurate what are we to make of it?

Consider how a representative from each of three neighbouring but hostile territories might respond. A social constructionist would notice how sexual identity has been constructed. Western culture allows quite a limited choice of sexual identity and it is by no means clear how many homoerotic thoughts a person is allowed to have before they should consider themselves bisexual or gay. These men were taking no chances with their heterosexual identity.

A psychodynamic theorist would notice how something the men were in danger of despising in themselves was denied, projected and unconsciously identified in another group that was overtly attacked. And a critical realist as well as noticing the importance of the heterosexual construction to the identity of these men would also highlight the fact that something real could be empirically measured that could be used to counter any cosy notions of pure heterosexuality.

None of these three representatives are natural colleagues but that makes no difference to the value of all the ideas for someone standing in the uncommitted zone between the three. From this central position wouldn't we agree that these perspectives interact together to give some potentially richer understanding .. a synthesised view that goes something like this - that sexual identity may not be the same as embodied sexuality and that a pure heterosexual identity may require processes of denial and projection leading to homophobia? Or at its simplest - that the maintenance in western society of an identity of pure heterosexuality may require homophobia.

This border analogy is a bit desolate - I don't think we actually have to stay outside of movements to make use of them and to avoid crusader side-effects. We can contain the ideas from all these movements without fully identifying with them. In other words we don't have to succumb to their totalising discourse. What can help us to do this is that we are primarily practitioners and not just thinkers. I'll say more on this in a moment. But I want to fairly recognise Gergen's own moves towards some rapprochement in this debate. In the chapter I mentioned he recognises that realists and constructionists often use each others arguments. Speaking of both sides in the debate he says this:

"I suspect that most all of us would stand fast against neo-Nazis, the Mafia, Islamic terrorism, smuggling heroin, cliterectomy, and Tory policies on higher education. Using a compelling discourse of realism we would point to multiple failings and immoralities. Further, in fine constructionist form we would be happy to demonstrate how such groups could - through the circulation and verification of discourse within their midst - come to find their actions both reasonable and right. By the same token, most of us would fight fiercely against such groups when they used either constructionist or realist discourses to achieve their goals." p152

Realism and constructionism, I would contest, are opposite sides of the same coin. We can't see them simultaneously but we can't have one without the other. As I suggested in a [Pocock] 1995 paper (although using other terms), we need them both, each acting as a critical restraint on the potential excesses of the other.

So in line with this eclectic epistemology we can comfortably consider clinical theories like Minuchin's [1974] ideas on rigid triads, or the communication theories of Watzlawick et al [1967] to be social constructions and this will help us to be cautious about how we use them. But they are at the same time more than "just our stories". They are special kinds of constructions - ones that have survived numerous encounters with the reality of practice. As the critical realist Andrew Collier [1998] - also writing in the Parker book - suggests, it is possible to talk a very great deal about something that we know little about - as anyone who has been an undergraduate student will know - but the limitations of what we know show much more quickly when we practice. Then we can learn a thing or two.

Family therapy is heading towards its 50th year of practice, representing unknown millions of hours of face-to-face encounters. Its ideas are distilled out of attempts to make sense of countless moments of success and failure in therapy and I feel deeply uncomfortable when I see these ideas simply written off as mere linearity or first-order pathologising. Harry Goolishian recommended that therapists should completely change their ideas every five years. But why?

I should say a little about the role of science in helping to decide on the appropriateness of theory. A positivist therapist who will only use clinical ideas that are scientifically verifiable will have to live on concepts of the human condition that are very emaciated indeed. Scientific discourse privileges the measurable and frequently its measuring sticks are just too coarse to pick out the fine and rapid moments when therapeutic change takes place - although process research seems much more promising than research into the comparative effectiveness of models. It is not much surprise that measurements of the effectiveness of different models give very similar results. I think one might do just as well to empirically measure the relationship between therapeutic effectiveness and the form of transport the therapist uses to get to work. Therapy needs some kind of vehicle but "helpfulness" in the open system of the personal encounter is, I believe, related to moments in that encounter - not to models.

Regrettably, the postmodern critique has encouraged a fashionable dislike of science when the scepticism of postmodernism to truth claims and appreciation of the power of discourse could, more appropriately in my view, be additional safeguards against the misuse of scientific evidence. One of the best examples of the use of science is in the work of Byng-Hall [1995]. He uses categories of attachment, which are well established as having predictive value, but at the same time there is recognition that these categories are constructed and reductionist. So his therapy is not bound by them - he is no positivist - but they have become a safe base from which he has improvised his practice. This use of science is still comparatively rare. I can think of very few family therapists, for example, who have made any use of the findings of behavioural genetics which has been one of the most outstanding areas of research in recent years.

But for therapists, science is not the only - or even the most important - encounter with the real. The day to day theories of therapists are not, I think, to be found on their bookshelves but derive from numerous small moments of therapeutic and life experience which accumulate in the mind of each practitioner. The working theories of therapists are shaped by theory and discourse but are also unique stores of the traces of each therapist's survival of all her encounters. This is the kind of theory development that Boscolo and Bertrando [1996] in their book - Systemic Therapy With Individuals - refer to as epigenesis. It is like the patina that covers old wooden sideboards: a waxy accumulation built up from grubby handling, numerous spillages and, on good days - a layer of polish.

I am happy to use the term "expertise" to refer to this but in doing so we should recognise its dangers. Much of this epigenetic development is barely conscious - it is, I think, the source of intuition - and can easily operate without much conscious thought on the part of the therapist. The appropriate safeguards against it are, I think, reflexivity on the part of the therapist and, as I will explain, paralogy; both of which processes bring the ideas of the therapist into an open space for conscious thought and dialogue.

Moments of positive change in therapy are usually so elusive. I think I can sometimes pick them up by a shift in my own mood; often they are signaled by an increase in empathy for someone I previously felt out of tune with, or even a little surge of love or admiration. Sometimes there is small moment of mutual achievement - like searching for and finding together the missing piece of a jigsaw which opens up some broader more coherent picture. Occasionally family members will tell me what shifted, but often they can't tie it down either and I tend not to ask very often.

But one moment in therapy stands out clearly for me. A girl of seven had for the previous year given up all spontaneity, laughter and play - unlike her younger brother who was exuberant enough for two. In an initial session with both parents, her, and her brother, she sat quietly and properly in her chair and spoke to me politely and carefully while her brother drew and played with the toys and chatted happily and rather noisily to everyone. In our conversation together her concerned, sensitive parents, while quite anxious about speaking in front of her, eventually talked about the time when "Daddy was very tired" and how that had been a worry for everyone. I asked the daughter, "did you feel you had made Daddy tired?". "Yes.." she said, politely, "I think I wore him out."

This idea, we soon discovered, had dated from the day her father had taken a serious overdose and, while the girl was present, had been removed from the house on a stretcher and taken to hospital. This had followed a long period of serious depression - which was known in the family by the euphemism "tiredness". Both parents and myself were quite stunned by the sudden awareness of the burden of guilt and responsibility the daughter had carried. The silence was broken when one of the parents said to her "That must have been terrible for you." She didn't reply and I began to say something when she interrupted me - "Can I play now?" she asked.

Some days, as I'm sure you will agree, it is good to be a family therapist. I met with the parents for several sessions after that and occasionally they asked their daughter, by then much happier, if she wanted to attend again. "No, I just needed to go for that bit about wearing Daddy out." she said.

A realist therapist would say there is something waiting to be discovered, a constructionist therapist would say there is some new meaning to be made through dialogue. What I want to pick out of the example is that there is something both to be searched for with the family - in this case a traumatic experience that had been storied into self-blame - and something new to be constructed - both an acknowledgment of this event and its meaning for the child and a new empathic story in the parents which contained and transformed the daughter's guilt so that she could be a playful child again.

I am grateful that at some point in my professional life I had been exposed to the idea that children of that age are prone to egocentric, or omnipotent thinking - to making themselves the cause of any difficulties around them that they don't understand. But what awful sounding terms these are and what easy targets for postmodern therapists looking for pathologisers. But in practice, explanation is not the same as understanding [Pocock. 1997]. To confuse the two is like pointing at a skeleton and asking "Have you met my friend?" Explanations are the bare bones from which the healing capacities of human understanding can grow.

But I fear that I have still not done enough in this paper to explain what I mean by the relationship between clinical theory - including working theories - and the unknowable reality beyond our constructions. Reality is not like a magician's rabbit just waiting in the dark to be pulled out of a hat. Our theoretical constructions both select and shape our experience of the real. There is no cross-generational coalition just waiting to be discovered but a family therapist who has been immersed in the discourse of structural family therapy may construct such a shape from the patterns of the family dance.

Look, for example, at this image and imagine yourself in the position of someone who thinks that triangles are, without doubt, the most interesting shapes. Your commitment will mean that the two triangular shapes will become salient. It is this investment of salience by the observer that turns patterns into facts. But these "facts" put out of mind other potential ways of seeing. Now try being fascinated by the four sided figures. You will be constructing some different "facts" out of this image and the triangles will fade. You might, from other orientations, invest salience onto acute or obtuse angles or the length of the lines and so on. Most probably, you are finding it hard, as I am, to move away from constructing this image as something like an aquarium or an upside down transparent container. Our perceptual loyalties are hard to put aside.

Each way of constructing the pre-reflective reality of the image as a "fact" puts out of mind - or defers as differénce in the language of Derrida [1972] - other constructions. Note that there is something real here to be constructed but there is no description that captures it, no way of describing it that does not in some way reshape it in our minds. There is reality but not objective discoverable reality. We are all spin-doctors. When we buy a blue car of a particular make we suddenly see them everywhere. Did hundreds of people suddenly buy the same car? No, they just weren't salient enough to us before.

First, the greater the number of ways of seeing open to the therapist the more potential there is to find and construct with the family something of value. For this reason I would never throw any ideas away and see no reason not to be endlessly curious about the ways of seeing that other therapies use.

Secondly, any way of seeing both reveals and conceals something of the complexity of the problem that has been brought to therapy. So the narrow visions engendered by model and movement loyalties will restrict the flexibility of the therapist reducing, in my view, the possibility of helpfulness.

Thirdly, and perhaps most importantly, the therapist cannot make objective discoveries since there are none. But the therapist can see possibilities for the making of new meaning but only within the dialogical thinking space of the session.

But what we still have at this point is a therapist and family members in the same room but separated by their differing subjectivities. How do they find each other and how do they create something new and meaningful? The constructivism of Maturana gave us "co-evolution" and social constructionism gives us "co-construction" but I have always found these a little empty as theories of clinical discourse. I find a richer expression of the way forward in Lois Shawver's [1998] version of Lyotard's term "paralogy". This is how she introduces it:

In Lyotard's philosophy, the term "paralogy" means a flood of good ideas that are inspired by conversation. Postmoderns, he tells us, have a quest for "paralogy," a hunger for stimulating conversation and ideas that work in a satisfying way. To get those ideas paralogists often share an irreverent attitude towards well accepted theories, breaking them up and recombining them in revolutionary new ways. The point of paralogy is to help us shake ourselves loose of stultifying traditional frameworks that we have come to take for granted, in order to enhance our spontaneous creativity. [Shawver, (www)]

For Lyotard, paralogy emerges from a battle of ideas but Shawver offers a gentler vision, more suitable for our purposes. Therapists need not hold their ideas as theoretical blueprints that the family should conform to, neither need they be discarded as pathologising finger-pointing. They are, instead, possible starting points for the therapist to enter the paralogical process. The therapist has ideas but she listens generously feeling for the way that words are being used locally and provisionally; seeking to enter the meanings of others in the conversation and noticing how her thoughts are shifting along the way.

And now it seems to be the therapist's turn to speak and she may say: "I don't know for sure ... but this is what I was thinking .." and that note of uncertainty and her previous generosity in listening may have earned some entitlement for the meaning of her words to be felt for by the family members. And they in turn may notice that they have some surprising new thoughts which originated neither in them nor in the therapist but came unbidden out of those peculiar moments when people allow themselves to be affected by others. And, as the family members give voice to the effects of these new ideas, it is the therapist's turn to find that she too has begun to think and feel a little differently. And on it goes ..

And this is my provisional answer to the question posed in the title of this Conference. It is good that family therapy is worried about knowing but if we allow the accumulated ideas of the first 40 years of practice to slip through our fingers what will we have left to offer up into the thinking space of the therapy session?

What is the best kind of therapist to be? My answer is "a helpful one". Movements and models are steps along the way to helpfulness, but these are steps whose cement has yet to dry so we should tread lightly on them in case they suck us down into uncritical loyalty and rigidity and prevent us entering fully into the paralogical process.

Boscolo, L., Bertrando, P. (1996) Systemic Therapy for Individuals. London: Karnac Books.

Boszormenyi-Nagy, I. and Krasner, B. R. (1986) Between Give and Take: A Clinical Guide to Contextual Therapy. New York: Brunner/Mazel.

Byng-Hall, J. (1995) Rewriting Family Scripts: Improvisation and Systems Change. London: Guilford.

Cade, B. (1996) Comment: Towards a one-and-a-half order practice: rehabilitating pragmatics. Journal of Family Therapy, 18: 243-248.

Collier, A. (1987) The language of objectivity and the ethics of reframing. In S. Walrond-Skinner and D. Watson (eds), Ethical Issues in Family Therapy, London: Routledge and Kegan Paul.

Collier, A. (1998) Language, practice and realism. In I. Parker (ed) Social Constructionism, Discourse and Realism. London: Sage.

Efran, J. and Clarfield, L. (1992) Constructionist therapy: sense and nonsense. In S. McNamee and K. Gergen (eds), Therapy as Social Construction. London: Sage.

Foucault, M. (1975) Discipline and Punish: The Birth of the Prison. A. Sheridan (trans). London: Allen Lane.

Frosh, S. (1997) Fundamentalism, gender and family therapy. Journal of Family Therapy, 19: 417-430.

Frosh, S. (1998) Review of: David Pilgrim, Psychotherapy and Society. Journal of Family Therapy, 20: 438-440.

Gergen, K. (1998) Constructionism and realism: how are we to go on? In I. Parker (ed) Social Constructionism, Discourse and Realism. London: Sage.

Golann, S. (1988) On second-order family therapy. Family Process, 27: 51-65.

Hoffman, L. (1998) Setting aside the model in family therapy. Journal of Marital and Family Therapy, 24: 145-156

Kuhn, T. (1962) The Structure of Scientific Revolutions. Chicago: The University of Chicago Press.

Lyotard, J. (1979) The Postmodern Condition: A Report on Knowledge. Manchester: Manchester University Press.

Maturana, H. (1988) Reality: the search for objectivity or the quest for a compelling argument. Irish Journal of Psychology, 9: 1-24.

Minuchin, S. (1974) Families and Family Therapy, Cambridge: Harvard University Press.

Pocock, D. (1995) Searching for a better story: harnessing modern and postmodern positions in family therapy. Journal of Family Therapy, 17: 149-173.

Pocock, D. (1997) Feeling understood in family therapy. Journal of Family Therapy, 19: 283-302.

Shawver, L. (1998) Postmodernizing the unconscious with help of Derrida and Lyotard. The American Journal of Psychoanalysis. 58: 361-390.

Shawver, L. [www] Lyotard's Concept of Paralogy. <http://www.california.com/~rathbone/paralog2.htm>

Watzlawick, P., Beavin Bavelas, J. and Jackson, D. (1967) Pragmatics of Human Communication: A Study of Interactional Patterns, Pathologies and Paradoxes. New York: Norton.

White, M. (1991) Deconstruction and therapy. Dulwich Centre Newsletter, (3) 21-40.