LEARNING OBJECTIVES

After reading this article, you will be able to:

• • express an informed view on a general psychological theory of how individuals relate to one another

• • understand how interrelating becomes dysfunctional and the importance of this for understanding personality disorder

• • identify the process whereby a therapeutic encounter can resolve conflictual relationships.

Interpersonal theory focuses on how humans relate to one another and has its genesis in the work of Harry Stack Sullivan, an American neo-Freudian and psychiatrist. He emphasised the interpersonal nature of personality, writing ‘personality can never be isolated from the complex of interpersonal relationships in which [a] person lives’ (Sullivan Reference Sullivan1947: p. 5). His work has subsequently been developed by several authors, including Timothy Leary, Jerry Wiggins, Donald Kiesler, Jeremy Safran and others. It is impossible to do justice to its rich developmental history here, but I would refer the interested reader to papers by Wiggins’ (Reference Wiggins1996) and Wagner & Safran (Reference Wagner, Safran, Castonguay, Muran and Angus2010) for helpful summaries of its elaboration. As this is a review bound by history, an update on the theory is portrayed by Hopwood et al (Reference Hopwood, Wright and Ansell2013). In this review, I will sketch only its main elements, but particularly focus on the contributions of the late Donald Kiesler, whose work, I believe, ought to be better known outside of North America. But why should it be relevant to personality disorder?

The interpersonal nature of personality disorder

First, one might argue that interpersonal difficulties are the crucial defining feature of personality disorder, as both DSM-5 and ICD-11 in their general definitions of the condition make clear. ICD-11, for instance, describes personality disorder as comprising ‘impairments in functioning of aspects of the self (e.g., identity, self-worth, capacity for self-direction) and/or problems in interpersonal functioning (e.g., developing and maintaining close and mutually satisfying relationships, understanding others’ perspectives, managing conflict in relationships’ (World Health Organization 2024: p. 553; my italics). DSM-5 is more succinct but similar in defining personality disorder as having ‘impairments in personality (self and interpersonal) functioning and the presence of pathological personality traits’ (American Psychiatric Association 2013: p. 761; my italics).

Here, there is a definitional problem that needs to be addressed. One might query, for instance, whether there is a real distinction between impairments of the self and in interpersonal functioning, as the demarcation lacks clarity. For example, in a narcissistic individual, there is clearly a disordered sense of self encompassing self-entitlement, grandiosity, etc. which inevitably manifests itself in disordered interpersonal relationships by exploiting others, etc. Hence, the demarcation between self and interpersonal functioning is unclear.

Perhaps a more crucial distinction is where there is fragmentation of self (‘identity, self-worth and capacity for self-direction’) as compared with a dysfunction of interpersonal functioning in the presence of a coherent sense of self. Although both ICD and DSM appear to give equivalent weight to the dysfunction of self and interpersonal functioning in their definitions of personality disorder, clinically, it is an impairment of the latter that characterises the majority of those with personality disorder. For instance, in someone with narcissistic personality disorder, one could argue that their ‘identity, self-worth and capacity for self-direction’ are not in any way impaired in the sense of being fragmented; rather, these features coalesce around a dysfunctional identity that is entirely congruent but one that is neither helpful to self nor to broader society, and hence the disorder (Edershile Reference Edershile and Wright2025). Indeed, it is this very congruence of self – rather than its fragmentation – that creates such a resistance to change.

There is one exception and this applies to borderline personality disorder, which, it could be argued, is unique among those with personality disorder in showing this fragmentation of the self. It is here (and only here) that one finds mercurial changes in ‘identity, self-worth and capacity for self-direction’. As ever, it all depends on how the self is defined, but there is no doubt that interpersonal dysfunction is a crucial component of personality disorder and hence it ought to be the main focus of treatment.

Second, one needs to recognise that attempts at exploring and attempting to change this interpersonal component of a person’s personality (or its disorder) can be seen as an attack on the ‘person’ and therefore why it might provoke a negative response. Hence, extreme sensitivity is required in such an endeavour. This sensitivity as to the nature of the disorder would not apply to the same degree to a clinician treating someone with, say, other mental health conditions such as depression or schizophrenia or indeed physical disorders. These are, after all, ailments visited upon, rather than part of, the self, as is the case with personality disorder.

Finally, how one understands and manages this interaction between patient and therapist is a critical component in developing effective treatments for those with personality disorder. Although a wide array of treatments – both physical and psychological – have been assessed for efficacy in borderline personality disorder, few of these have been tested for other personality disorders, and although some have shown some modest effects (Storebø Reference Storebø, Stoffers-Winterling and Völlm2020), these gains have often failed to be replicated when compared with high-quality generic clinical care (Chanen Reference Chanen, Betts and Jackson2022). Moreover, even when there is symptomatic improvement in borderline personality disorder, a sense of social isolation and feeling of emptiness persist (Alvarez-Tomás Reference Alvarez-Tomás, Soler and Bados2017). Here, there is an important caveat in that these evaluations consider evidence from groups rather than from the individual, and extrapolating such group findings to the individual is not straightforward. To paraphrase an argument from Kiesler’s (Reference Kiesler1966) critique of psychotherapy research: given personality disorder’s considerable heterogeneity, the more pertinent question is not ‘what works for personality disorder?’ but rather ‘what works for this individual with this personality disorder?’. What is required then is a theory that has sufficient generality that it can be applied to groups, but also is sufficiently nuanced that it can be applied to the individual. Interpersonal theory has the capacity to fulfil both of these requirements.

Interpersonal theory

Interpersonal theory is premised on how humans (and other animal forms) face the challenge of relating to one another by achieving a balance between (a) affiliation and (b) dominance (i.e. a wish to relate to others but also to control them). This produces a dimensional representation of personality organised around these two major axes, orthogonal to one another, with the extremes (the poles) of its ‘affiliation’ axis represented by ‘cold (hostile)’ versus ‘warm (friendly)’ and those of its ‘dominance’ axis by ‘dominant’ versus ‘submissive’ in a circumplex. Combinations of these two dimensions can then be pooled to produce blended variants around the centre (Fig. 1).

FIG 1 The interpersonal circumplex, with the axes of affiliation and dominance and their derivatives.

There have been various elaborations of this basic portrayal of the theory, but it has been criticised as being too value laden, so alternatives have been proposed. John Birtchnell’s octagon theory, for instance, posits the alternative dimensions of ‘close’ versus ‘distant’ and ‘upper’ versus ‘lower’, as these are generic and value free (Birtchnell Reference Birtchnell2010). Although I believe that this nomenclature is superior, here I will use the more conventional terms, such as cold (hostile)/warm (friendly) rather than close/distant, and dominant/submissive rather than upper/lower.

Consequences of interpersonal theory

Normality versus abnormality

How relating takes place around this circumplex has the following implications. First, it implies that normality in relating is close to the centre, as it is here that one finds someone who has a repertoire of responses to interpersonal encounters that is varied and appropriate, but not extreme. Thus, ‘normal’ individuals can move flexibly around the centre, with an array of wide-ranging and adaptive interpersonal responses to the situation in which they find themselves, resulting in harmonious and stable relationships. Second, those who occupy a position at the edge of the circumplex have responses that are not only more extreme, but also more limited, resulting in a constrained behaviour repertoire that remains the same irrespective of the situation. For instance, someone who occupies a position at the ‘dominant’ pole of the axis is likely to behave in a dictatorial and controlling manner irrespective of the circumstances in which they find themselves, so that they appear overbearing and autocratic as they are intent on eliciting a submission from their respondent. So, the first consequence of the theory is that normality is associated with interpersonal flexibility, whereas abnormality is associated with both extreme and constrained interpersonal behaviour. The relevance of this observation is that rigidity and extremity of interpersonal behaviour, irrespective of the circumstances, is one of the core defining features of personality disorder.

DSM-IV shows the position of some of the personality disorders on the circumplex. It is worth noting that borderline personality disorder does not appear, and there is good reason for this omission. As referenced above, borderline personality disorder is unique among personality disorders in that it has a fragmented rather than a coherent sense of self and this is reflected in such a disordered interpersonal style that it cannot be represented within the circumplex. This also has implications for how the patient–therapist interaction is managed therapeutically (see the ‘Borderline personality disorder’ section below).

Complementarity

The theory is further developed by considering the interpersonal exchanges of the interactants within the system, or what has been termed their complementarity. Sullivan (Reference Sullivan1953) introduced this in his ‘theorem of reciprocal emotion’, in which he stated ‘integration in an interpersonal situation is a process in which (1) complementary needs are resolved, or aggravated; (2) reciprocal patterns of activity are developed, or disintegrated; and (3) foresight of satisfaction, or rebuff, of similar needs are facilitated’ (Sullivan Reference Sullivan1953, p. 198). He also emphasised that both parties contribute to the interpersonal exchange; Kiesler (Reference Kiesler, Anchin and Kiesler1982: p. 198) quotes Sullivan: ‘neither Person A’s nor Person B’s needs alone determine the outcome. Interpersonal needs always seek conjoint expression and resolution.’ Finally, Sullivan’s other important contribution is his observation that the interpersonal exchange constrains the response of the interactant. This constraint, in turn, reinforces the first person’s ensuing actions.

Subsequently, it was left to others, but particularly to Robert Carson, to develop this aspect of the theory further. Here, I shall do no more than mention the main features of the theory of complementarity and I refer interested readers to the work of Carson (Reference Carson1966) and Kiesler (Reference Kiesler1983) for a more complete account.

The ‘evoking message’

As human interpersonal interaction aspires to be as anxiety free and as stable as possible, this is achieved through ‘unconscious evoking messages that function to constrain the others’ reactions to those that are predictable and comfortable for the individual’ (Wagner Reference Wagner, Safran, Castonguay, Muran and Angus2010: p. 218). Thus, one individual is manoeuvred by another to adopt a relationship position that is ‘complementary’ to the one being offered and is mutually rewarding to both participants. If this does not occur, then the relationship is unlikely to endure, or the interaction is altered such that complementarity is re-established.

Correspondence versus reciprocity

Interpersonal theory posits further that in order to maintain stability in relating, correspondence occurs on the affiliation axis (friendliness invites friendliness, and hostility invites hostility), whereas reciprocity tends to occur on the power axis (dominance invites submission, and submission invites dominance). For instance, a complementary response is established when an individual in the Hostile/Dominant quadrant evokes a hostile/submissive response from someone in the Hostile/Submissive quadrant. Similarly, someone in the Hostile/Submissive quadrant would seek a complementary hostile/dominant response from someone in the Hostile/Dominant quadrant. When these trait styles are not concordant, interpersonal messages are sent to pull a complementary response to alter the partner’s behaviour towards one that is reinforcing to their preferred style (Fig. 2).

FIG 2 Complementarity within the interpersonal circumplex. D, dominance; S, submission; H, hostility; F, friendliness. The direction of the arrows shows a complementary response.

Anti-complementarity

If two individuals are in opposite quadrants (e.g. Hostile/Dominant and Friendly/Dominant or Hostile/Submissive and Friendly/Submissive), their evoking messages are unlikely to work, so a state of anti-complementarity results (Fig. 3). In such an interaction, as neither the individual’s affiliation nor their status is supported, this will be experienced as minimally rewarding or aversive. Consequently, the relationship is unlikely to survive. This has relevance to the psychotherapeutic process as here the therapist needs to take an anti-complementarity stance in order to effect change. In so doing, they face the challenge of maintaining the relationship while also seeking to move the patient from their preferred position (see the ‘Therapeutic implications’ section below).

FIG 3 Anti-complementarity within the interpersonal circumplex. D, dominance; S, submission; H, hostility; F, friendliness. The direction of the arrows shows an anti-complementarity response.

The unconscious element of interactions

As regards the interaction, it is important to realise that this interpersonal communication in relationships is more unconscious than overt. Here, Kiesler was heavily influenced by communication theory and figures such as Beier, Haley and Watzlawick and others. To quote Beier (Reference Beier1966) for instance,

‘The evoking message is probably one of the basic tools used by individuals to maintain their consistency of personality. With this message an individual can elicit responses without being aware that he is responsible for doing so. To a certain extent he can create his environment without feeling that he is accountable for the responses which come his way’ (p. 13).

To capture this non-verbal messaging, Kiesler and colleagues developed the Impact Message Inventory (IMI).

Therapeutic implications

Kiester (again) provides a useful summary of the implications of interpersonal theory in psychotherapeutic practice (Kiesler Reference Kiesler1983). At the beginning of therapy, the patient intentionally and unintentionally shapes the therapist to respond to them from a restricted aspect of the therapist’s own internal experience and behavioural repertoire. Here, the patient/client has the whip hand so that ‘the therapist cannot but be hooked or sucked in by the client, because the client is more adept, more expert in his distinctive, rigid, and extreme game of interpersonal encounter’ (Kiesler Reference Kiesler, Anchin and Kiesler1982: p. 281).

The therapist’s task is then to become aware of how they are being manoeuvred by the patient by being attentive as to how their own actions, attributions and fantasies are being altered by the interaction. In interpersonal theory, this refers to an awareness that one is being pulled to give a complementary response by the patient to re-establish harmony and stability in the relationship. Becoming aware of this ‘pull’ and a realisation that this is the unconscious activity of the patient rather than something within the therapist is a crucial component of the therapeutic encounter. Having this self-awareness goes to the heart of any psychotherapeutic encounter in which the therapist has not only the competence, but also the confidence, to decide on how they are being manoeuvred by the patient to respond in a way that is out of character. With personality disorder, this can be especially challenging as it encompasses a judgement on much of what passes for ordinary human interaction. It emphasises the need for extensive training, together with ongoing support and supervision for those engaged in such work.

With this realisation, the therapist then needs to disengage and stop the reinforcement of the individual’s pathological style by interrupting their automatic responses to the patient. With a dependent individual, for instance, the therapist may begin to withhold advice and take control so as to interrupt the pattern of reinforcement of the individual’s pathological dependent style. This can then progress to the therapist engaging in anti-complementary actions that challenge the patients’ bid for further reinforcement based on the elements of the interpersonal circumplex. The contents of this phase focus on the self-defeating consequences for the patient of their extreme and rigid behavioural pattern.

When the therapist changes from a complementary to an anti-complementary role there is an opportunity to link the patient’s immediate response to their more habitual behaviour. Clearly, this would need to be done sensitively and in a non-judgemental way, for example ‘I notice that you seem a little disappointed [uncomfortable/angry etc. depending on the response] that I failed to meet your expectation that I would provide you with an extra session as you were lonely. I wonder whether you see any connection between this and how you have reacted to similar perceived rejections in the past’.

Alliance ruptures

Another important advance in psychotherapeutic thinking relevant to this discussion is that of alliance ruptures and their repair, stemming from the operational and empirical study of the therapeutic relationship by Jeremy Safran and colleagues. ‘Alliance ruptures’ are defined as a tension or breakdown in the collaborative relationship between the patient and the therapist, varying in intensity from brief disengagements to full-blown termination (Safran Reference Safran, Crocker and McMain1990). This thinking has been heavily influenced by interpersonal theory, as the therapeutic relationship is conceived ‘as consisting of intersecting dimensions of affiliation and interdependence on the interpersonal circumplex’ (Safran Reference Safran, Crocker and McMain1990).

Hence, they stress that such ‘ruptures’ (a) are interactive: ‘a rupture is not a phenomenon that is located exclusively within the patient or caused exclusively by the therapist. Rather a rupture is an interactive process that includes these kinds of experiences that play out within the context of each particular therapeutic relationship’ (Samstag Reference Samstag, Muran, Safran and Chaman2019: p. 188); and (b) that they recapitulate long-standing interpersonal interactions: ‘Ruptures often emerge when therapists unwittingly participate in maladaptive interpersonal cycles, that resemble those characteristic of patients’ other interactions, thus confirming their patients’ dysfunctional interpersonal schemas or generalized representations of self-other interactions’ (Safran Reference Safran and Muran1996: p. 447). Being aware of this interpersonal pull and being able to detach oneself sufficiently to collaboratively explore what is taking place in the therapeutic relationship, they refer to therapeutic metacommunication (adopted from Kiesler Reference Kiesler1996).

Complementarity and countertransference

This type of therapist reactivity is a very similar to, but a more generalised aspect of, countertransference in the analytic literature. Countertransference was initially conceived as something that impeded therapeutic work – a psychological blind spot in the analyst that needed to be understood and dealt with. Over time, however, far from being a barrier, attentiveness to one’s countertransference was seen as providing ‘an additional avenue of insight into the patient’s unconscious mental processes’ (Sandler Reference Sandler, Dare and Holder1973, p. 66). The value of countertransference in the treatment of personality disorder was noted by Kernberg (Reference Kernberg1965), who wrote ‘that the full use of the analyst’s emotional response can be considered to be of particular importance in the treatment of patients with profound personality disorders’ (cited in Sandler Reference Sandler, Dare and Holder1973, p. 66).

This suggests that one’s interpersonal response in the therapeutic encounter might provide not only valuable information on the patient’s interpersonal pathology but also an opportunity to ameliorate it. There is no better description of the value of this interactive process between therapist and patient than that in a quote from James Stratchey in the early psychoanalytic literature:

This early psychoanalytic quote shows a remarkable concordance with much of what has been written here, as there is an (a) interpersonal encounter enacted with the analyst which is not only (b) conflictual but also (c) involves an ‘actual and immediate situation […] under our control’ and (d) is a great opportunity to engage in a real life encounter between two parties so that progress is possible. Indeed, these intersubjective negotiations between a patient and therapist in developing and maintaining the therapeutic relationship in spite of the conflict, and the patient’s emerging experience of a new and constructive connection with the therapist, are considered the ‘very essence of the change process’ (Safran Reference Safran and Muran2000).

Case vignettes

The following fictitious case vignettes show how an understanding of a person’s position within the interpersonal circumplex can inform therapy and management of individuals with personality disorders in forensic settings.