Critique of CBT and CAT by

Dr Anthony Ryle
The central claim made for CAT (cognitive analytic therapy) is that it
seeks to offer a comprehensive understanding of human psychology
and involves therapists in forming real, clearly defined and therapeutically powerful relationships with their patients. In this respect it is a
psychodynamic theory and is clearly differentiated from CBT (cognitive behavioural therapy).
Understanding the immense complexity of human psychological
processes needs to be based on an understanding of their development.
CBT provides a model of learning but takes little account of early development and its effects upon psychological structures. CAT, in contrast,
revised object relations theories in ways that sought to eliminate unverifiable assertions about the unconscious and were consistent with
observational studies of early development.
The current CAT model of personality emphasizes that human
infants are genetically predisposed to be socially formed to a far greater
degree than any other animal. They show active emotional engagement
with others from birth, communicating through behaviour, expressive
sounds, gestures, rhythms and imitation (Trevarthen, 2001). Over the
first few years they increasingly demonstrate the highly developed
human capacity to make and use symbols.
A childs unique inherited temperamental characteristics will influence its interaction with others, notably with those in its particular family. The repetition over time of early patterns of interaction are seen to
initiate a unique repertoire of reciprocal role procedures (RRPs). RRPs are
reiterated sequences of perception, thinking, affects and action linked to
the expected or elicited responses of others. Their enactments are accompanied by communication which increasingly involves pre-linguistic
signs and, eventually, by language. Self processes and the self-conscious

Change for the Better (4th edn) by Elizabeth Wilde McCormick (2012, SAGE).

CRITIQUE OF CBT AND CAT

I are derived from the internalization of these reciprocal interactions

and communications. Conscious thought is essentially internal dialogue
derived from sign-mediated exchanges with others. The intrapersonal
procedures involved in self care, self control and self judgment are
derived from the internalization of the interpersonal patterns characterizing early relationships and the two remain permeable to each other.
In this way CAT offers an adequately complex but accessible model of self
processes whereas in CBT the self is more often referred to than defined.

The stability of role procedures

The stability of role procedures is maintained by eliciting reinforcing
reciprocations from others, and this is true of both normal, functional
reciprocal role patterns and of those that are harmful or restrictive.
Revision occurs when others withhold the expected reciprocations or
offer more functional ones or enlarge the persons awareness of them,
so that conscious control may become possible. This understanding
underlines the need for therapists to provide both non-reciprocation and
clear descriptions of their patients dysfunctional RRPs.

Personality structure
The most significant patterns acquired in early life are concerned with
issues of care or neglect in relation to need and over-control or cruelty in
relation to submission. The self is normally multiple as individuals
acquire a repertoire of RRPs, different ones being mobilised in ways
appropriate to the context. Normal multiplicity may include the manifestation at different times of contradictory patterns but in general links
between patterns and awareness of the range is established. However,
this is not the case where adversity and predisposition result in a structural dissociation of the individuals repertoire of role procedures. In
such cases the sense of self is fragmented and discontinuous. In borderline personality disorder, which is the most frequently encountered type
in clinical practice, patients commonly show abrupt switches between
states and may have little recollection between them. This is confusing
to the patient and to those around them, including clinicians, who as a
result may feel de-skilled and may become rejecting.
In these patients, therapists and other clinical staff need to support
integration of the dissociated reciprocal patterns. This can be aided by
verbal and especially by diagrammatic descriptions of the repertoire of
RRPs, which demonstrate their dissociation into separate self states

CRITIQUE OF CBT AND CAT

and trace the switches and links between them. Self State Sequential
Diagrams support the consistent, non-collusive attentive engagement of
clinicians. They also have a direct therapeutic role in helping patients
recognize their states and state switches and so gain more control over
them. These are the essential elements of the Multiple Self States
Model (MSSM) of BPD as described in Ryle (1997).
Many borderline patients are prone to switch into states of uncontrolled anger. Rather than relying on anger management the CAT
response would be to trace the dysfunctional RRPs that precede the
switches into anger with the aim of establishing more adaptive modes.
These prior dysfunctional modes usually represent long-term strategies
evolved in response to deprivation and are attempts to avoid anger.
They typically involve patterns of resentful compliance, emotional distancing or the avoidance of vulnerable need, all of which maintain a
sense of deprivation and pain from which switches to rage states may
be triggered. These states, whether expressed in hurting self or others,
are liable to provoke rejection and hence perpetuate deprivation. CAT
would seek to modify these preceding patterns as well as developing
recognition and control of the switches.

How does therapy support change?

In CBT theory the emphasis is placed on the description and challenging of the links between thinking, feeling and acting. When based on a
skilful analysis of the sequences and sensitively carried out this can
clearly be an effective ingredient of therapy but expressed through
simple-minded analyses and schoolteacherly assertions of the patients
faulty thinking it can be experienced as critical or disrespectful. The
therapists positive responses to more effective behaviours may be reinforcing but in the case of persistent negative behaviours, especially
when these undermine therapy, there is little theoretical understanding
of what to do. In such cases CBT therapists rely upon patient compliance with its technical procedures and the personal qualities of the
practitioner to maintain the therapeutic work.
In CAT, creating an effective working relationship with patients is a
main aim, but this must include skilful work with those whose destructive role procedures undermine the therapy. In CAT these are seen to
represent the enactment of some part of the patients dysfunctional
repertoire; the patient brings into the therapy relationship the problems
for which help is sought. Once identified and recognized with the help
of the diagrammatic reformulation these can be challenged rather than
reinforced by reciprocation.

CRITIQUE OF CBT AND CAT

In understanding RRPs in the context of the therapy relationship, sensitivity to non-verbal affective communication are crucially important.
Therapists need to be aware of the feelings induced by their patients.
They may be induced by actions or words but often they can only be recognized by emotional resonance to the patients expression, posture and
other indirect messages. To recognize these, therapists need to be emotionally open and also aware of their own contributions to the patterns.
This aspect of CAT theory is a re-conceptualisation of the psychoanalytic understanding of transferencecountertransference, understood
as a particular example of the general way in which relationships
depend upon the meshing of reciprocal patterns. The CAT concept of
RRPs is found accessible and useful by patients and by non-therapy
staff and provides a more adequate and comprehensive framework for
therapists than CBT.

Objectivity, psychotherapy research and the

nature of evidence
CAT seeks to offer a scientific basis for understanding human nature but
insists that objectivity, in the sense of studying humans as objects, is
inappropriate. We need to observe and record evidence in ways capable
of being replicated by others but full evidence concerning human behaviour and experience can only be obtained by human beings and must
include intersubjective understanding. We know each other through
shared contexts and histories, speech, observation, expressive behaviours
and also through incompletely understood, non-verbal communication:
The heart has its reasons of which reason knows nothing, as Pascal said.
In studies of infant development it is recognized that infants cognitive abilities are evident early and develop in observable ways but their
parallel capacity to convey and respond directly to affective aspects of
experience and behaviour to feel other minds, to know others noncognitively are neglected or denied. The overemphasis on cognition
and the need to explain affects as secondary derivatives of it is dominant in cognitive-behavioural approaches. Therapy practice and
research needs to acknowledge and make use of such subjective knowing and avoid the over-valorisation of cognitive processes, a view which
receives strong support from studies of early development focusing on
how the child develops in relation to others (Reddy, 2008). Mirror cells
and other yet-to-be discovered neurophysiology can throw some light
on the neural pathways involved in our coming to know other minds
feelingly but the fact that we do and how we do can be experienced
and further studied without the neuroscience.

CRITIQUE OF CBT AND CAT

Cost effectiveness and the evidence base of

CBT and CAT
The problems of demonstrating clinically useful effects for a psychotherapy model are formidable. Therapy is not comparable to the
prescribing of a drug; it involves two people and what happens between
them is only partly determined by the model. Trying to overcome this
by delivering manualised therapy seems to me to be incompatible with
a humanly respectful therapy and cannot eliminate the individual
variations in therapists work. The effects of factors common to any
therapy delivered with reasonable tact and attention are difficult to
distinguish from the effects of specific interventions. Change may occur
during active therapy but may not endure, but adequately long followup is costly and rarely done. The aims pursued in different therapeutic
models may differ, as they do between CBT and CAT, and standard
measures are sensitive to only some types of change. In controlled trials
the power of the model being researched may be inflated where the
control intervention often business as usual is clearly ineffective.
These problems have not been overcome in CBT research and the
loud and persistent claims that it (almost alone among therapies) is
evidence-based are increasingly criticised. David Richards (2007), a past
president of the BABCP, attributes the growing criticisms of CBT to its
selective use of evidence and to the naive belief that the randomized
control trial is the only weapon we now need. He observes how we
ourselves write the research questions that now get funded; reviews
have shown that RCTs can both exaggerate and under-estimate the
likely real effect most CBT trials are small and poorly executed; quality thresholds for RCTs in NICE guidelines are notoriously low, allowing
the meta-analyses of small poor quality studies to direct policy; we pay
no attention to qualitative evidence . He points to the unproven contention that it is possible to take the results of experiments conducted
by charismatic product champions in highly controlled environments
and implement them in the widespread manner suggested by Layard .
CAT has different weaknesses and in particular is often criticised for
its small research output; it has indeed been slow to accumulate evidence from large-scale RCTs. However, small-scale descriptive and controlled studies and evaluated case studies accompanied the development
of the model and research into process and into aspects of theory and
practice have continued to take place. This included work on borderline
personality disorder which for a long time had been neglected by psychiatrists and by dynamic and CBT therapists. CAT was shown to have
an effect within a 24-session format (Ryle and Golynkina, 2000) and has
been found to be particularly helpful by those who work with these

CRITIQUE OF CBT AND CAT

patients and with other hard to help (i.e., usually unhelped) patients,
including abuse survivors, the elderly and offenders. This research also
contributed to an understanding of the importance of dissociation, a factor little attended to except in conditions directly attributed to trauma. It
led to the development of the Multiple Self States Model (MSSM) of BPD
which emphasizes the alternating dominance of a range of dissociated self
states, each characterized by contrasting RRPs expressed in subjective
symptoms and behaviours, some of which may be extreme. The MSSM
is of importance in relation to diagnosis, case formulation, management
and treatment. In work of particular relevance to mental health services
it has been shown that brief training and supervision enables staff with
no formal psychotherapy training to use CAT reformulations to plan
interventions and avoid collusive responses(Kerr et al., 2007).

Outcome research
A review of CAT research will be found in Appendix 1 of Ryle and Kerr
(2002). Since then a well designed RCT comparing the effect of the
addition of either CAT or a humanistic cognitive intervention to a comprehensive treatment programme for late adolescent borderline patients
has been reported by Chanen et al. (2008). This showed clinically relevant advantages from CAT. Other published outcome studies and
those in progress are listed on the ACAT website.
In my view the findings of process research and the use of single case
designs are more likely to influence therapy practice than are the largescale RCTs on which CBTs claims to be evidence-based are based.
Other CAT research has been concerned with the development of
instruments for clinical and research use. The eight-item Personality
Structure Questionnaire (PSQ) (Pollock et al., 2001) was developed to
assess the degree of dissociation and provides a reliable measure. High
scores are characteristic of BPD; the mean scores of outpatients
referred for psychological treatment are between those of normal subjects and those with BPD. Bedford, Davies and Tibbles (2009) administered the PSQ to more than 1000 outpatients; they confirmed the
psychometric qualities of the PSQ and showed that scores fell in
patients receiving a range of treatments while remaining stable in those
not treated. The evidence suggests that the level of integration varies
across the spectrum of psychological disorders.
Support for the Multiple Self States Model (MSSM) of BPD has been
provided by the use of the States Description Procedure (Ryle, 2007), a
clinically useful method which contributes to the reformulation of borderline patients.

CRITIQUE OF CBT AND CAT

Notions of happiness
Philosophers and writers have offered many different ways of considering whether happiness is a desirable goal and if it is how a person
might pursue it. But therapists cannot contribute by prescribing the
goal or explaining the meaning of life. For most people, survival rather
than happiness is the aim. I would argue that the belief that one is
entitled to, or worse still should be able to purchase, happiness is a
symptom of our individualistic consumer culture and may well contribute to unhappiness by ignoring the extent to which individuals need to
be in a meaningful relationship with others and with their wider social
context. Faced with the fact that our culture produces many hundreds
of thousands of people needing relief from psychological/emotional
distress and unhappiness, the proliferation of therapists, a response
consistent with the Layard thesis, is like dealing with rising sea levels
by issuing lots of buckets. Economic downturns, poverty, modern forms
of colonialism and vast expenditure on weaponry are not natural phenomena like tsunamis, they are man-made and call for political action.
Prescribing counselling or CBT for all serves to distract us from attending to the values and practices of the unhappy society in which we live.
If more resources are made available for psychotherapy and counselling a first step might be to correct the underfunded and unevenly distributed services in the NHS. But prevention would be a more logical priority,
achieved by increasing what might be called psychological literacy, centrally among those involved in parenting, education and management but
also more generally. In that respect the CAT model is particularly suitable
because of its focus on the relationship between individual psychological
functioning and the social and relationship context.

Bibliography
Bedford, A., Davies, F. and Tibbles, J. (2009) The Personality Structure
Questionnaire (PSQ): a cross-validation with a large clinical sample,
Clinical Psychology and Psychotherapy, 16 (1), 7781.
Bennett, D. and Parry, G. (1998) The accuracy of reformulation in cognitive analytic therapy: a validation study, Psychotherapy Research, 8,
84103.
Bennett, D. and Parry, G. (2004) A measure of psychotherapeutic competence derived from cognitive analytic therapy (CCAT), Psychotherapy
Research,14 (2), 17692.
Chanen, A.M., Jackson, H.J., McCutcheon, L.K. et al. (2008) Early intervention for adolescents with borderline personality disorder using

CRITIQUE OF CBT AND CAT

cognitive analytic therapy: randomised controlled trial. British Journal of

Psychiatry, 193, 47784.
Fosbury, J.A., Bosley, C.M., Ryle, A., Sonksen, P.H. and Judd, S.L. (1997)
A trial of cognitive analytic therapy in poorly controlled Type 1 patients,
Diabetes Care, 20, 95964.
Kellett, S.(2005) The treatment of Dissociative Identity Disorder with cognitive analytic therapy: experimental evidence of sudden gains, Journal
of Trauma and Dissociation, 6 (3), 5580.
Kerr, I.B., Dent-Brown, K. and Parry, G.D. (2007) Psychotherapy and mental health teams, International Review of Psychiatry, 19 (1), 6380.
Leiman, M. (1992) The concept of sign in the work of Vygotsky, Winnicott
and Bakhtin: further integration of object relations theory and activity
theory, British Journal of Medical Psychology, 65, 20921.
Pollock, P.H., Broadbent, M., Clarke, S., Dorrian, A.J. and Ryle, A. (2001)
The Personality Structure Questionnaire (PSQ): a measure of the multiple self-states model of identity disturbance in cognitive analytic therapy, Clinical Psychology and Psychotherapy, 8, 5972.
Reddy, V. (2008) How Infants Know Minds. Cambridge, MA: Harvard
University Press.
Richards, D. (2007) Arrogant, inflexible, remote and imperious: is this
what is wrong with CBT?, BABCP magazine, March 2007.
Ryle, A. (1997). Cognitive Analytic Therapy and Borderline Personality
Disorder: The Model and the Method. Chichester: Wiley.
Ryle, A. (2001) Cognitive Analytic Therapy, Constructivism in the Human
Sciences, 1&2, 518.
Ryle, A. (2007) Investigating the phenomenology of Borderline Personality
Disorder with the States Description Procedure: clinical implications,
Clinical Psychology and Psychotherapy, 14, 32941.
Ryle, A. and Golynkina, K. (2000) Effectiveness of time-limited cognitive
analytic therapy of borderline personality disorder: factors associated
with outcome, British Journal of Medical Psychology, 73, 197210.
Ryle, A. and Kerr, I.B. (2002) Introducing Cognitive Analytic Therapy:
Principles and Practice. Chichester, Wiley.
Trevarthen, C. (2001) Intrinsic motives for companionship in understanding: their origin, development and significance for infant mental health,
Infant Mental Health Journal, 22 (1-2), 95131.