The limitations of ISTDP. Part 1: Jon Frederickson

What are the limitations of ISTDP? What would a balanced view of ISTDP be like? Just as any approach to psychotherapy, ISTDP is subject to both idealization and devaluation. Over the past few years, we at have published quite a lot of positive stories and news about ISTDP. Now it’s time to do some balancing. We sat down with some prominent ISTDP clinicians to discuss the shortcomings and downsides of ISTDP. Here’s the first part, an interview with Jon Frederickson.


As we’ve talked about before, ISTDP is not a panacea. Which type of problems and patients are not suitable for ISTDP? 

Jon Frederickson discussing a balanced view of ISTDP
Jon Frederickson

Jon Frederickson: Nothing is a panacea in the field of mental health. Types of problems not suitable for ISTDP would include the treatment of traumatic brain injury, neurocognitive deficits, and genuine autism spectrum disorders (not including those mistakenly diagnosed).

Generally, we should offer supportive and not exploratory psychotherapy to patients currently abusing drugs until we have built the affect tolerance that would make exploratory therapy possible. Likewise, some psychotic patients in a severe regression and severely depressed patients may require medication and supportive psychotherapy before a trial of exploratory therapy should be attempted.


Just how difficult is ISTDP to learn? As far as I’ve heard, no one ever graduated from Davanloo’s training. Should learning ISTDP be easier? 

Jon: It’s not just a matter of ISTDP being hard to learn. Learning to be a really good therapist is hard. That is why it is relatively rare. Twenty percent of therapists get eighty percent of the good results. And that is true within each model of therapy. It is really hard to become a highly effective therapist in any model of therapy. You may be under the illusion that you’ve “learned” the model, but the outcome research shows that there is no relationship between our perception of our ability and our actual effectiveness.  

Should learning this be easier? Should learning to be a professional musician be easier? Should learning to be a chess master be easier? No.

It should be hard because it is hard. That is reality. However, in the case of psychotherapy: should our teaching be better? Yes.

Research shows that graduate training has no effect on therapist outcome. What a disaster! Should our supervision be better? Yes, because research shows that 93% of therapy supervision is inadequate and 35% harmful.

At least in music and chess, it is clear what skills need to be learned and there are materials which train students in those skills. We have no agreement on the fundamental skills necessary for effective practice in psychotherapy and no materials for training in those skills. So, in response to your question, yes and no. Learning a complex skill like psychotherapy should be just as hard as becoming a violinist.

Yet, it is currently way too difficult to achieve this skill level as therapists because of the poor quality of supervision generally available. As well as the inadequate, indeed, useless quality of graduate training. The useless seminars offered which do not show effective treatment, and the failure to use videotapes to develop an empirically validatable model of teaching and supervision.

In case you wonder if I am outraged by this state of affairs, you read me accurately.


Unlocking the unconscious is sometimes described as a unique aspect of ISTDP. But other models also facilitate emotional breakthroughs and spontaneous reporting of previously repressed material. Could the jargon mystify the therapy process and put ISTDP at risk of distancing from other models?

Jon: Obviously, any emotionally transformative human experience involves a breakthrough to feelings that were previously out of awareness. It even happens at movies! One danger in any model occurs when we use jargon to “professionalize” our field and to create a sense of mystique such that outsiders “could not possibly understand” what goes on behind closed doors.

Jargon creates another danger: we might accept a piece of jargon, usually a description, and mistake it for an explanation. As a result, steps in logic are skipped, and flaws in an argument remain invisible. In case you wonder what I mean, here are some common vague terms which are ill defined and have come to mean everything: mindfulnessawareness, and superego. Here is a term which doesn’t mean what it claims: diagnosis. In fact, what we call diagnoses are merely a description of symptoms, not a diagnosis of their cause.

Do you think there’s a need for a conceptual “makeover” in ISTDP to facilitate dialogue with other models? 

Jon: I don’t think ISTDP needs a makeover as you suggest. I think all therapists in all models need to abandon vague concepts, acronyms, and made up words for plain English, or whatever your native language is. If you cannot explain what you are doing so it could be understood by an adolescent, either your language is a barrier, or you do not fully understand what you are trying to say.

We work with humans, speaking a human language of the heart. Any theory we describe should be able to be put in these terms. If we dropped jargon, we could even talk to other clinicians. As it is, today much clinical dialogue at conferences becomes useless because the exchange of abstractions takes the place of examining the actual data. And the narcissistic display of mysterious language becomes a way to avoid the humbling act of revealing one’s actual work.


Even though there’s more and more research showing the efficacy of ISTDP as a whole, there’s still not so much high-quality research on the different ingredients of the therapy. A notable contribution is the recent Iranian study showing that ISTDP without challenge was just as effective as standard ISTDP. Given the lack of studies, ISTDP is largely an “oral tradition” where the experience of specific prominent therapists (be that Davanloo or yourself, for example) is very influential. What are some of the challenges with the specific ingredients of ISTDP?

Jon: The Iranian study was important, but like all studies, it’s easy to forget the context. In fact, challenge is appropriate only with about 25-30% of patients, the ones who primarily regulate feelings with isolation of affect. Challenge is not appropriate for the other seventy percent of patients who are in repression and fragility. So it should be no surprise that ISTDP without challenge would be effective, because that is the effective form of ISTDP for 70% of patients!

When students try something I suggest, sometimes it works, sometimes it doesn’t. Why? Sometimes they aren’t doing what I suggest. Sometimes I was wrong; I misread the patient, and the patient’s response gives a clearer idea of how to proceed. Sometimes, the therapist is initially helpful without realizing it, but is unable to understand and categorize the patient’s subsequent responses. I don’t think the issue is the individual clinician per se, although the effect of the therapist is powerful. I see repeatedly that there are certain patterns of response across patients and across cultures. When we address these patterns – feelings, anxiety, defenses, and transference resistance – we find patterns of response to intervention.

Now we get to the interesting question: the relationship between principles and rules. For instance, when a patient is struggling to bear mixed feelings, the principle is to help the patient bear mixed feelings without anxiety shifting out of the striated muscles. Sometimes, to make things simple, people make up a rule: “Thou shalt pressure to feelings in this way. Repeat after me!” The student, alas, learns to become a clone who follows rules rather than a person who operates according to principles. There are many interventions that could embody the principle of building affect tolerance. And those interventions could be in response to specific words or dynamics the patient has used. They could arise from the therapist’s experience, feelings, and intuition. They could arise from their mutual co-created responsiveness.

In music, the voice leading (how voices related to each other, for instance, in a fugue) was not supposed to have parallel fifths. That was a rule. Suddenly Debussy comes along and he uses all kinds of parallel voice leading to create effects of great beauty. What had been a rule was revealed to be subject to a higher principle. Thus, it could be broken.

Alas, the early phase of ISTDP training often involved people following rules without understanding the overarching principles, to which those rules are subject. If we ritualistically follow rules, therapy is very easy to learn, though robotic. If we follow principles, then we understand the purpose of our interventions, and that allows for creativity in the therapist and responsiveness to the patient.

Good therapy is like jazz. A jazz musician knows the key, the melody, the harmonies, the underlying principles and he improvises based on that underlying structure. He appears to be breaking rules, yet he is guided by underlying principles. A good teacher orients you to principles whether he is teaching you chess, music, or therapy.


Historically, the ISTDP community has unfortunately been subject to sect-like behavior such as a strong idealization of charismatic figures (such as Davanloo) along with exclusion and devaluation of critical voices. Is there something in particular that makes ISTDP vulnerable to this? What can we do to safeguard against this in the present and future? 

Freud - a balanced view
Sigmund Freud

Jon: As we know from the work of Bion and other group theorists, when humans form groups, groups become irrational. 

Friedrich Nietzche said that earth is the insane asylum of the universe. Every day we see plenty of evidence for this. Idealization of teachers happens in all models to greater and lesser degrees. Think of Freud, Klein, Davanloo, Rogers, or Beck. Every one of them has been idealized, and each of them has been devalued.

There will always be some people who want to idealize their leader and devalue the rest. We have to understand this as not a problem of a given model, but a problem of the human condition. To avoid the anxiety that our knowledge is partial, our theory will be changed and surpassed is the way of all scientific knowledge, and that whatever we create today will be forgotten in the mists of time, we seek magic.

We idealize a model and view it as the final, complete answer. We idealize some figure. Then we devalue other models and teachers. Then we imagine we are part of some secret society of superior therapists in contrast to all those “others.” This pattern has been described in cults, and, sadly, this kind of cult formation is common in the therapy field. All we can do is make ourselves aware of this temptation to idealize and devalue.

Melanie Klein

And we can also step back and realize what makes us anxious: 1) our knowledge is always partial; 2) we will never have all the answers; 3) we will always be flawed and fail with some people; 4) we will never have the final, complete understanding of the human condition in our lifetime; and 5) whatever we achieve, whatever we build is transient and will disappear. This is reality.

When we cannot bear this death anxiety, we engage in the denial of death through the magical claim that we have found the eternal answer, the eternal group, and the theory that has somehow transcended time. Due to death anxiety, this pattern will probably always recur in humanity, including groups of therapists.


Do you see other major limitations or weaknesses in ISTDP? 

Jon: My major concern here does not have to do with ISTDP but with the psychotherapy field as a whole. Our understandings all too often are not linked to other areas of knowledge such as sociology, group theory, family studies, and economics. These different fields appear as silos. Take for instance the study of patients who suffer from borderline personality structure or psychotic patients. There is so much good research on the relationship between their psychological difficulties and predictable patterns of family dysfunction.

Yet this research keeps getting forgotten, only to be done again by the next generation. These patients are often examined only from the individual perspective, and we forget the family system that generates these patterns. We look at psychological issues, yet we seem to have forgotten the role of social class and capitalism in character development. Fromm wrote much on that, yet today in the US it is a taboo to recognize the role of class.

Or look at racism in the US or the caste system in India as examples of the transgenerational transmission of trauma. And then there is the tendency to underestimate the role of neurocognitive deficits and brain injury in borderline and psychotic patients. The psychotherapy field has become so focused on the individual, that we easily lose sight of the group and family context, the class context, and the biological context. Then we end up with these different research silos: each reducing the patient to one of these categories, when we need to open up to the interrelationships between them.

Do you find there are aspects of ISTDP that we have to address and change in order for the method to thrive? 

Jon: It depends on how you define ISTDP. Some describe it as the method. If so, that is ritualism, and, yes, that should be changed. Some describe it as what some teachers do. If so, that is idol worship, and that should be changed. For some, it is a set of rules, and that should be changed.

For me, ISTDP is a set of meta-theoretical principles which allow us to integrate any of a number of techniques. The most important principle is to assess each patient response to intervention to find out if you met the patient’s need in the moment. And these principles are based on a psychoanalytic theory of childhood development and attachment theory. The techniques of cognitive-behavioral therapy, somatic experiencing, gestalt therapy, or internal family systems, you name it, can be incorporated because the key issue, no matter what technique you use in the moment, is: am I meeting the patient’s need in this moment as revealed in her last response to intervention?

In this sense, I am suggesting that we need to move beyond the idea of a model toward an integrative way of thinking and responding. Models can only point toward that. Replication of models does not lead to good outcome. We have to foster a kind of integrative emotional feeling and responsiveness in our work that models and theories can only point toward.

The best therapists in each model look surprisingly alike according to research. This suggests to me that the key factor is not just their model, but a quality of thinking, feeling, responsiveness, and self-reflectiveness that is filtered through their model.

It’s like driving. It doesn’t matter what kind of car we see. It’s the nut behind the wheel.

Jon Frederickson’s latest book Co-Creating Safety: Healing the Fragile Patient came out a couple of weeks ago.

If you liked this article, you might find our other material interesting. Following this link you can find more material in english. Below you’ll find a list of our recent interviews.

Joel Town: “Teaching the ‘intensive’ is the central challenge”

This is an interview with Joel Town. Joel Town is one of the most important ISTDP researchers out there, being the first author of several empirical studies of ISTDP. Among them is the most rigorous ISTDP study thusfar, the “Halifax Depression Study“. He is an Assistant Professor at Dalhousie University, a lead researcher at the Halifax Centre for Emotions and Health and he runs Dynamic Health Psychological Services. Last year, he visited Göteborg for a two-day workshop on treatment-resistant depression. We had a chat with him about where ISTDP is at and where it should go.

joel town portrait
Joel Town

How did it feel to present to a swedish audience? 
As you know, this was actually my first time both visiting Sweden as well as teaching. I was very happy to see some old friends who I’ve met at past conferences, meet many engaging new colleagues, as well as make some new friends. It was a pleasure to be with you in Sweden and to see how your ISTDP community is growing!

You presented some thought-provoking ideas for modifying the graded format of ISTDP, managing the thresholds differently. Can you explain your perspective on this? And how is it different from the standard graded approach?
With the graded format, I was trained to first think about the use of ‘pressure’ to mobilize complex feelings. Next, we look for a threshold to detect when the patient is struggling to intellectually hold in mind complex emotional states and instead become flooded with anxiety. At this point, the therapist helps to reduce anxiety using different strategies. One observation around the clinical application of this approach that I spoke about during the workshop is how easily we can teach the process as though there are explicit “go” and “stop” signs.

The concept of a threshold can be helpful when initially learning this approach in order to avoid too much anxiety being triggered. But viewing patient tolerance more as a Threshold Window can allow therapists to involve patients more in the process, and help them better learn to self-regulate. I’d describe this more as principle we can be aware of during learning, teaching and supervision that can allow therapists to begin to incorporate what Allan Abbass has called “bracing” interventions. This can be done instead of formal recapping or other anxiety regulating interventions. 

You offered some modifications to the phase of inquiry that were very well received by the swedish audience. How did that come about?
It’s something that I noticed over time when I was reviewing trial therapy tapes. I felt like I was missing something if I wasn’t asking patients about feelings, clarifying anxiety or defences within the first few minutes. However, there were other occasions when I was using these kinds of interventions early in sessions – but I would be left questioning why am I doing this and how helpful is was. This made me think about some of the learning challenging when teaching. And it made me rethink the timing of the transition from ‘inquiry’ to phases of structured ‘pressure’ in ISTDP.

In Gothenburg, I showed a tape in which the patient came into a trial therapy exhibiting a mixed bag of responses that we might consider examples of unconscious anxiety and defence combined. In the tape, I didn’t comment explicitly on these processes for around 10 minutes and instead stuck with a phase of inquiry. The subsequent group discussion raised some good questions about the importance of the pace and timing of therapist interventions early in sessions. The audience appeared to appreciate me saying that it isn’t always entirely clear what is happening moment-to-moment, so collecting more data from sitting with the patient can be helpful.

I think in our effort to provide and teach “intensive short-term” treatments we can easily prioritise the need to intervene. We even sometimes intervene before we understand why we are intervening. During the 2-day seminar it felt like we were able to have a very a constructive discussion about these issues. My thought is that these are likely learning and training challenges in ISTDP as much as they are about technical elements related to the phase of inquiry in treatment.

What are you struggling to learn as a clinician right now?
I have begun seeing a series of patients with chronic symptoms that have an explicit behavioural component such as OCD (e.g., compulsive behaviours), Tourette’s and other tic-based presentations. There hasn’t been a lot written on this topic around the use of ISTDP and in my experience these cases present infrequently to dynamically orientated therapists. It’s been a challenge and learning curve to think about how to adapt and tailor a dynamic approach to specifically target change in symptoms that involve repetitive behavioural patterns.

For instance, in ISTDP I would aim to help a patient see harmful patterns and behaviours so that they become motivated to interrupt them independently. In contrast, a traditional CBT exposure and response prevention approach involves a more directive therapist stance in advising a patient to prevent the ‘response’. In the cases I have treated so far, I am struck by how much emphasis there has needed to be on an explicit therapist stance towards response prevention. I think it is a subtle but significant shift for the ISTDP therapist to focus on interrupting an explicit in-session behaviour like a vocal tic in contrast to purely intrapsychic defences.

What are some of the current challenges for the further development and dissemination of ISTDP globally?
One of the challenges for the dissemination of any psychotherapy is having the means to effectively train others to deliver the treatment. Over time, through these dissemination efforts, if enough clinicians can be trained to become both effective therapists and trainers themselves, there reaches a critical mass at which point the treatment is readily accessible for patients. There are probably only a handful of therapies which can be said to have achieved this globally.

The manualisation of psychotherapies has been a key part of what has made this possible. However, the development of treatment manuals to treating mental health as discreet “disorders” defined by symptom clusters is problematic. I think this paradigm has contributed to the numbers of patients who fail to remit or relapse following psychotherapy generally. In contrast, I think ISTDP is best described as an approach that is fundamentally built to achieve ambitious changes in personality. While my own experience as a researcher and clinician confirm that this is possible, as with other therapies, therapist factors and patient factors contribute significantly to outcomes. 

I think the ISTDP Core Training programs conducted in the last decade indicate that there are many elements to ISTDP that can be taught to a broad group of therapists. The programs teach the delivery of effective treatment that is likely comparable to the outcomes achieved in other treatments. However, my current view is that I think there are other elements of ISTDP that are very difficult to learn, particularly given the training resources typically available to most therapists (e.g., access to and frequency of supervision).

If some of the more difficult-to-learn treatment elements were emphasized less, akin to dropping the ‘intensive’ from ISTDP, I think we would have a treatment that could be more easily disseminated globally. It is arguable that in doing so, we could compromise the nature of the changes possible in treatment by de-emphasizing the elements that promote personality level changes. This is a question that would need addressing empirically.

Do you think we should drop the “intensive” then? Or what do you propose? 
I don’t think the field needs a new treatment with a new acronym. What I am pointing out is that there are different elements to ISTDP that require different competencies to be taught and adequately mastered by a therapist. If attempting to gain competency in multiple domains limits the transferability of the treatment, I am proposing that therapists can be trained and encouraged to utilise specific elements as they are able.

This type of learning environment might help therapists to flourish and grow rather than to become discouraged and drop-out. Perhaps within the field of psychotherapy training there is a risk that in an effort to maintain the presumed integrity of the treatment, it is very possible that the alliance between therapists and their trainers/teachers can be adversely effected. I think this is a central challenge around the dissemination of ISTDP.

Do you have any upcoming research in the pipeline?
I am just preparing a manuscript describing the 12-month post treatment outcomes and a cost effectiveness analysis from the Halifax Depression Study. This a randomized controlled trial that compared the outcomes of time-limited ISTDP against the effects of secondary care community mental health team treatment for treatment resistant depression (TRD). We published the initial findings in the Journal of Affective Disorders in 2017 showing ISTDP is an efficacious treatment for TRD with 36% of patients reaching full-remission at the end of treatment. The follow-up findings are also very encouraging. 

What’s your vision for the future of ISTDP? 
Currently there are very few academic centres around the world in which ISTDP research is being conducted. For the growth of any treatment, research is an important part of dissemination alongside offering the possibility of innovation in methods and technique. In particularly, as a clinical psychologist and researcher having both trained and conducted clinical trials in ISTDP, I think my understanding of some the teaching and learning challenges around ISTDP has been enhanced greatly by this work. Moving forward, I would hope there are increasingly more opportunities for people at all stages of learning to be involved in ISTDP training and research within academic centres of excellence.  

If you enjoyed this Joel Town interview, you might find our other interviews interesting. You can find the whole list here. Below you’ll find a list of five of our most recent interviews: