psychotherapy as internal relation

(notes for a talk at Confer, 21st Sep 2019)


Describing psychotherapy as a method of treatment is expedient in healthcare settings, but it would take a tin ear to not baulk at it as a characterisation of the therapeutic experience. To unpack our intuition that ‘treatment’ talk fails us, we may use Wittgenstein’s distinction between internal (constitutive) and external (two-part) relations. Psychotherapeutic work embodies, and develops the patient’s capacity to enter trustingly into, mutually implicating (i.e. internal) relations with others – whereas treatment requires only that people are merely externally related. Then again, psychotherapy is also in the business of helping patients separate out from others. This togetherness-in-difference has a name: it’s what we call ‘love’.

Introduction - Grammatical Notes

The question I was invited to address by the organisers of this conference was 'Is psychotherapy a treatment or a relationship?'

Now I take it that this isn't a straightforward question like 'Is animal X a cat or a dog?' For there's as yet too little context, around the mere question, for us to have a sense that the two terms (treatment and relationship) are being used in a sense which excludes one another. Thus perhaps we should want, and do well, to say that psychotherapy is both a relationship and a treatment.

Now there's a trivial sense in which 'psychotherapy' can't easily be taken as the name of a relationship, since psychotherapy is - like 'treatment' but not 'cure' - the name of an activity, and activities are not relationships even when they're done together with someone else. 

And there's also a trivial sense in which nobody shall deny that psychotherapy involves relationship, since even those behaviourists who think of the therapeutic relationship in merely instrumental terms - i.e. merely as a means to an end - as merely a delivery mechanism if you like - will not deny that it involves two or more individuals in cooperative interaction. A pizza (the end, as it were) is delivered by a delivery man (the means, as it were). The delivery man himself, though, is not the dinner (unless you’re a murderous cannibal).

Having got these trivia out the way let's move straight on to examine the key issues.

Instrumental vs Non-Instrumental Understandings, and Versions, of Psychotherapeutic Relationships

Now, what contrasts with a merely instrumental conception of the psychotherapeutic relationship is one in which it is in the relating itself that therapeutic change is taken to occur.

Consider that Judith Beck tried to take a stand against those who accuse cognitive therapists of ignoring the importance of the therapeutic relationship by insisting that cognitive therapy

requires a good therapeutic relationship. Therapists do many things to build a strong alliance. For example, they  work collaboratively with clients . . . ask for feedback . . . and conduct themselves as genuine, warm, empathic, interested, caring human beings.

To this the psychoanalytical psychotherapist Jonathan Shedler responded 

This is the kind of relationship I would expect from my hair stylist or real estate broker. From a psychotherapist, I expect something else. [Beck appears] to have no concept that the therapy relationship provides a special window into the patient’s inner world, or a relationship laboratory and sanctuary in which lifelong patterns can be recognized and understood, and new ones created.

Against this response I only note that Shedler’s optimism regarding his barber or estate agent perhaps warrants some of what a CBT therapist would call 'cognitive restructuring' - although maybe here we meet also with the different cultural expectations and practices of the two sides of the North Atlantic...

Now, working 'in the transference', as they say, provides a good example of what I'm here calling a non-instrumental use of the therapeutic relationship. Let's now spell this out using a comparison of, rather than a remark on the disanalogy between, behaviour therapy and psychoanalytic therapy. (I probably got this analogy from Leigh McCullough's work on 'affect phobia'.)

So consider first that in behaviour therapy it isn't imagined that one shall overcome one's dog phobia, say, merely by talking about dogs. Instead, real encounters with dogs must be suffered, and a 'response prevention' (i.e. evasion prevention) protocol be implemented. In this way one learns - or is re-conditioned - directly from within a transformative experience (of a non-devouring dog).

Similarly when it comes to reconditioning our fundamental, conscious and unconscious, affectively charged expectations of how others will treat us: what's important is not that we (I’m here the therapist, you the patient) simply talk about what happened to you in your childhood, or about what's currently going on at home, but rather that the fearful expectations are directly met with within, because they've been activated bythe intimacy of the therapeutic relationship itself

When the transference is activated and these fearful expectations are alive between us, we now stand a chance of truly learning the difference between the malevolent pit bull (in the negative transference) or the impossibly ideal golden retriever (in the positive transference) the therapist is imagined to be - and the, say, slightly up tight but nevertheless caring and discerning labradoodle he really is (in the real relationship).

As well as this transference work we also have the fundamental work of therapy to install a 'good internal object' and challenge self-criticism, which process we may think of as an internalisation of the therapeutic relationship. Here too the good therapeutic relationship is not merely instrumental – it’s not merely a precondition for the effective deployment of this or that technique, not merely a way to make someone receptive to the actual treatment - but it is itself the substance of the therapeutic process.

We might sum up these reflections by saying that, far from treatment and relationship being antithetical, meaningful therapeutic treatment will itself take the form of a relationship.


I now wish to turn to ways in which talk of a 'psychotherapeutic treatment' may be considered oxymoronic - or, in other words, to think about what it is in our concept of 'treatment' that can strike a bum note when we're thinking of applying it to the case of psychodynamic and person-centred psychotherapy. And I think the bum note rings out most clearly when we consider such uses of the word 'treatment' as 'Dr Gipps was treating Megan for her depression' - i.e. when the treatment is something done by one person (the therapist) to the other (the patient). This is the sort of thing which may be useful to write on an insurance report or in an NHS case file, but the 'treatment' concept does, I think, risk doing an justice to our sense of the therapeutic process.

To get at this I will borrow a concept from Wittgenstein's philosophy - that of an 'internal relation'. We might define this abstractly - as a conceptual or a constitutive relation - rather than a relationship between discrete, separable, entities - but I think that what it is will come out more readily with an example.

So, imagine first a child's toy in which various shapes (stars, squares, circles, etc) are to be posted through variously shaped holes. Here the holes and the shapes are two separate phenomena. We can say what they are without referring to the other. We can tell a meaningful story about what allowed the one to fit so well through the other (it was made to the same specifications, the craftsman was very skilled, etc). Here we meet with phenomena that are externally related one to the other: they don’t each take their shape from the other.

And now imagine that we meet instead only with a circle drawn on a piece of paper. Someone may come along and ask 'How is it that the white disc in the middle fits so well into the black circle around it?' But now we'd have to explain to the person that they’ve made a mistake. These are not really two separate phenomena, but just one phenomenon. The outside of the white disc is defined by its black perimeter, so there can be no question of one thing fitting another. The white disc and the black circle are internally related to one another: they take their shape from one another.


Now what I propose is that talk of 'treatment' is best understood as talk of something that's done by one individual to another separate individual. As such it is a paradigm of an external, causal relationship between discrete beings. 

The individuals in question may of course be very intimate, but the idea of a treatment going on does not in itself presuppose this intimacy. And it is because treatment talk has these external connotations that we should reject it for an apt description of the work of psychotherapy if we want to understand what psychotherapy essentially is. (You and the pizza delivery man may be lovers, but the pizza he brings is (presumably) extrinsic to your lovemaking.)

The question remains, though, as to what it means to describe what are after all the two separately identifiable participants in a therapeutic relationship as 'internally related'? And clearly the answer will be that, to the extent that the relationship stands a chance of being therapeutic, there is some sense in which the participants cannot be considered entirely separate. And yet clearly the last thing we want from a therapeutic relationship is that the participants become somehow merged with one another. Indeed it may be because of problems with her ego boundaries, or because she suffers a want of self-possession, that the patient comes to therapy in the first place. 

In what follows I shall look first at the internal relatedness required in a truly therapeutic relationship and then consider the significance of our human recognition of separateness, finally tying both of these points to the bum note struck by 'treatment' talk.

Internal Relations

The predicament we're often in when we go to therapy may I think be put like this. (I'm now going to tell you some things that I hope you'll agree we both already know.) So (you’re the patient again): you've suffered a partial developmental failure. You're unclear or doubtful in your heart about whether you're truly lovable as you are and about whether other people are truly to be trusted. To manage these concerns you erect a set of defences and develop a pseudo-adult carapace or persona through which you now engage with the world.

Certain others you do 'let in'. Even here however this goes smoother if first you idealise, have a 'love-in' with, identify with, these others, doing your best to overlook or 'forgive' (which is not real forgiveness) their all-too-human sides, all this happening under the aegis of a positive transference. All the other others, however, you 'keep out'; you engage in politeness, decorum, and all the panoply of those forms of latent human shunning or failures of openness that go to make up 'civilised life’. But whilst these defences help manage your anxiety they also create what we call 'symptoms' and thwart your development. Why is this?

They thwart development because it's only when we let others in that we can grow. The child needs a trusting intimate connection with others in order to get anywhere in life. Before he can become a separate being with a character which is more than a congeries of defences, he must be able to find his sense of himself in his relations with trusted others. Only out of this togetherness may true individuation proceed. I think that the quality of this experience of the child is best summed up by the Christian use of the word 'parrhesia', which in the Catechism is described as "straightforward simplicity, filial trust, joyous assurance, humble boldness, the certainty of being loved".

So: when I trust you, and let you in, I take your reaction to me as a true indicator of how I really am in myself. I cannot, despite the seductive illusions of omnipotent narcissism, truly achieve this calibration for myself. I can only achievemy self-esteem andcalibratemy moral sensibility (i.e. my conscience) through a close trusting relation with you. I find myself in you.

In such trust I allow myself to be calibrated by you - this is the sense in which a close trusting relationship, especially of a parent-child sort, involves a relationship that is partly internal. My self-becoming and your sense of me are not now two quite separate matters.

Now - not globally but at this specific juncture of our connection - we're related not as two separately-shaped phenomena which happen to coincide, but instead as the edge of the white disc to the inside of the black circle which defines it.

My self-esteem - which is simply the confident untrammelled bodying-forth of myself - can now grow. It grows, in particular, to the extent that I get a sense (now I’m the patient or child, you’re the therapist or adult) that you love me - by which here I mean, you want the best for me- with the stress both on 'the best for'and on the 'me'I can now grow as I step forward confidently into my life. My sense of forgivableness - my sense that I shan't be sent into exile when I make mistakes - and my sense of the valuedness of my own love - is also now increased. Because of all this I can allow myself to form valuable and close relationships with others.

One way we might articulate all this is by distinguishing two different senses of 'recognition'. In the one – let’s call it ‘empirical recognition’, as when I recognise that bird over there as a green woodpecker - we meet with something that may intelligibly be described as correct or incorrect. In the other – let’s call it ethical recognition, as when recognition is what you offer me - you affirm me in my essential character. This is what Rogers called 'unconditional[,] positive regard' - which, incidentally has got nothing to do with a regard that is unconditionally positive (since otherwise how could it begin to cohere with anyone but the sappiest therapist's congruence), but everything to do with being treated as an 'end in oneself', to borrow the Kantian terminology. You welcome me in; you show an openness to me as I am. When I trust in this recognition, then I and you are one, not simply in the coincidence of the content of our beliefs, but in a moment in which who we are is jointly enacted and thereby constituted. Now, just at the meeting point between us, we are one another, just as the white disk and the black circumference are of a piece - although, and of course, there are myriad other junctures of our lives where we are not so mutually implicated.

Now this, I think, is what it is to be internally related to one another in psychotherapy. It is only by presupposing a relationship with such a feature that, say, meaningful work in the transference can ever occur - otherwise why should I – I’m still the patient - ever trust your interpretations? It is this quality of the work, this inner involvement and implication, that - I think - explains why talk of ‘treatment’ just doesn't cut the therapeutic mustard. 

And, for that matter, it is this intimate recognition which is rightly described in terms of love, and which prompts the observation that decent psychotherapy is necessarily informed by love.


Having described our togetherness in a meaningful love-structured therapeutic relationship, I now turn to the equally important matter of our separateness and the recognition of it. The important question here is how to think both of these together. 

In a paper called 'The Sublime and the Good', Iris Murdoch writes:

Love is the perception of individuals. Love is the extremely difficult realisation that something other than oneself is real. Love ... is the discovery of reality. ... the apprehension of something else, something particular, as existing outside us. The enemies... of love...are... social convention and neurosis. ... Freedom is exercised in the confrontation by each other, in the context of an infinitely extensible work of imaginative understanding, of two irreducibly dissimilar individuals. Love is the … respect for, this otherness.

I present this here because with its talk of recognising otherness it can seem to give us a very different picture of what intimate connection is than the conception of self and other as internally related provides. 

And I take it too that a lot of therapy is about helping the patient separate out from, and recognise the independence from him of, his significant others. 

Therapy is partly - and it's an important part - about the growth of self-possession. In psychodynamic psychotherapy this occurs, we believe, partly through the interpretation, and thereby the dismantling of, the transference. A precondition of this is that the therapist can offer that unconditional positive regard to the patient which itself stems from a recognition of his independence from her. But what we were talking about before was the need for patient and therapist to join together in a moment of internal relatedness, mutual implication, or conjoint constitution. And this seems now to put us inquirers at theoretically odds with ourselves, to have us searching for both identity and disunity at the same time.

The answer to this puzzle is that that recognition which makes for an internal relation between self and other isn’t a kind of identity. What you offer recognition to is: me in my distinctive selfhood. In my particular character. This particular character is thereby acknowledged as an intrinsically valid way of being human. This is what it is to offer me positive, unconditional, regard. And I gain my sense of myself and the comfort to body forth as myself under your loving gaze. Despite how different I am from you, you recognise my character as a viable form of human selfhood. By loving me you acknowledge me in my utter otherness to you, as Murdoch states. Yet when I in my self-understanding meet you in your recognition of me, there we are at one, and there I can flourish.

Individuality, Psychotherapy and Treatment

Now it’s sometimes said, by psychoanalytic psychotherapists seeking a quick-and-easy response to the challenges of empirical research pundits, that it's wrong to make use of intrinsically generalist evidence-based treatment research because patients are all individuals.

If what this means is that patients all have different characters, then it's obviously wrong to some degree, and right to some degree. In some ways we are really rather like one another, and psychoanalysis has even contributed to our understanding of this. Whether or not one patient with, say, OCD, is sufficiently similar to another to make a research-derived psychotherapy protocol viable is, we might then think, surely an empirical matter, not something to be pronounced on from an armchair.

But this all makes me wonder whether there might not be another way to take what the psychoanalytic psychotherapist says from her armchair by way of reply to the scientist practitioner. And the way I suggest is what’s already been offered above: we're to treat our patient as an individual not in the sense that he has a unique character profile, but in the sense that he has a way of being human that (underneath the destructive defences he deploys) is to be accepted as is

It’s in his way of being human that we offer him recognition. This is not about our empirically recognising his characterological distinctness, but ethically recognising him in his distinctness. (And so we use his name and not a number; we treat him as an intelligible object of love; if he dies we don't treat him either as a dead animal or as a piece of rubbish, and when we grieve him we grieve not our loss but him. If we wrong him then we feel bad - not for having broken some moral law, but because we've hurt him

So the reason why it may not be apt to apply an evidence-based protocol to our patient, then, is not because he has a rare character profile, but because in a therapeutic relationship we specifically encounter him under the description 'human being' rather than, say, ‘homo sapiens’. This requires recognition - not of the empirical but of the ethical, offered, sort. The point isn’t that he’s not an instance of a type (presumably he is) - but rather that a relationship structured by that idea could never itself be a therapeutic one. 

To suppose that it is would make no more sense than someone saying 'I love my husband because he's 6 foot tall, has a handlebar moustache, is a clever clogs, and has large biceps'. One may certainly 'love' (qua greatly enjoy) those things about him, but truly loving him has no because - or, if you enjoy circular explanations, has no ‘because’ other than that in 'I love him simply 'because he is he''. Our patient has struggled with receiving recognition, has erected defences against being met with as a person, may have been judged in his core being, may have been projected into or neglected or used or smothered or spoiled or indulged by those who were instead supposed to care for him by honouring him, offering him recognition, treating him as a moral subject, one capable of wronging and being wronged. This is what we must undo, as we now express our willingness to meet him as a human being.


I began by getting a clear non-instrumental sense of the therapeutic relationship on the table. I then stressed that we can use the idea of 'internal - mutually implicating - relations' to characterise the kind of relationship we find in a therapeutic relationship. When we are open to one another in such a way as to make ourselves visible to the other's ethical recognition of us, then we are related internally and then the transformative work of therapy may begin. Now we take the other's reactions to us as providing more than information about ourselves. Rather it's in your true pity that I find myself; in your moral challenge to me that I understand my thoughtless selfishness; it's here that I can now become myself. Yet whilst this relationship between us is internal in form, like the black circle to the white disc within, it's not one of identity. Far from it: to be real your loving recognition of me must honour me as truly separate from you - as a being with his own character, own inner life, own needs, own form of experience.

And all this, I suggest, helps us grasp what it is that 'treatment' talk misses by way of an apt characterisation of the essential form of a therapeutic relationship. It tends us toward supposing that we and the patient do well to both direct our attention to something extrinsic to the therapeutic relationship, to something called 'the problem', 'the illness', 'the symptom'. But the very 'symptoms' that bring patients to psychotherapy are a product of a failure of personal being - a failure to receive, or a chronic shying away from receiving, recognition. And it's in providing the requisite recognition that we meet our patient, and its in this recognition that he meets himself and so may now pick up afresh the task of being a person in the world. It is, we might say, precisely in our treating him as a human being that the lie is given to the idea that what he needs is something called treatment.


Popular Posts