The British Journal of Psychiatry
Personality disorder
Peter Tyrer


It is now respectable to read about personality disorder. It was not always so. Despite the impossibility of practising psychiatry without being aware of the term and the subject matter it describes, it was not appropriate in good psychiatric circles to mention the subject unless presaged by a pause and pronounced with a mocking inflexion that indicated that the words were in parentheses: signposts to somewhere undesirable, usually somewhere in the jungle of forensic psychiatry (a subject about which I write very little in this piece, in an attempt to redress the balance). I think the reason for this was that personality disorder had such a strong flavour of criticism that, even in a discipline in which stigma confronts us on every corner, its words were the ultimate derogatory label that, once attached, became virtually indelible. Or, as my Lancashire grandmother would say about all unsavoury topics, it was “not very nice and no one really wants to know”. So research and writing on the subject became almost a samizdat topic, written about in code, discussed in quiet corners between professionals when they could not be overheard, or in proxy phrases such as ‘relationship difficulties’ or ‘patients who are difficult to place’ (Coid, 1991).


One of my fellow researchers in personality disorder, Per Vaglum of Oslo, has commented that Rolv Gjessing, the famed Norwegian researcher into periodic catatonia, had an inscription on his door, “In the long run it all depends on personality”, a statement I would echo over and over again. But did Gjessing write and expand this view into empirical research that confirmed its importance? I think not. Even the first UK (Scottish) researcher on personality disorder, David Henderson, rarely used the term. The first book I bought on psychiatry before I started my clinical undergraduate career was his renowned textbook of psychiatry (Henderson & Batchelor, 1962), which had just achieved its ninth edition. In it, despite Henderson's eminence in the subject, personality disorder is only subsumed under ‘psychopathic states’ and described as a symptomatic clinical state in the same way as the other mental disorders.

Henderson was, however, a major figure in developing the subject and his work stimulated me enormously. Although his separation of psychopathy into aggressive, inadequate and creative subtypes (Henderson, 1939) has not achieved support from further data, the description of the creative group emphasises that even abnormal personalities can have positive attributes that may have short-term gains obscuring some of the negative aspects and that do not necessarily lead to major adversity, except in the long term. It is only in retrospect that ‘brought-forward social dysfunction’ as a consequence of personality features can be identified, and it is often difficult to know during life what are the fundamental features of a personality and whether they are positive or negative in effect. We should not be too keen to attribute the diagnosis of personality disorder only to those who demonstrate universal counter-transference (Lewis & Appleby, 1988), although it is true that those with the disorder do have less close contact with other people (Tyrer et al, 1994), and it is when both positive and negative aspects of personality are present that diagnosis is most difficult.

Henderson's work stood me in good stead when I started my clinical training at St Thomas' Hospital, London, where the characteristics of our consultant teachers were fertile ground for the detection of a wide range of interesting personality abnormalities and led me to create my first personality questionnaire. I cannot pretend that this was accurate as a diagnostic tool (Tyrer, 1980) but it caused much merriment among the students and at least characterised the working style of our teachers. Their elevation to teaching-hospital consultant status seemed to depend, at least in part, on their ability to present at least one odious personality feature repeatedly in order for us to admire them as ‘ characters’. Ever since this time I have felt that the diagnosis of personality in all its aspects is one of the most exciting and challenging parts of psychiatry, particularly when I realised that some of the features I had so assiduously identified were two-dimensional stereotypes created mainly for interactions with medical students.


Assessment of personality and its disorders is genuinely exciting because it really gets to the heart of the problem of classification in psychiatry, the ability to summarise in one or two words the common features of a disorder. Personality status, unlike other disorders in psychiatry, is much better described by playwrights and novelists than by psychiatrists, and my first advice to those who aspire to an understanding of personality disorder is to read the character formulations of Shakespeare, Thomas Hardy, Jane Austen, Anthony Trollope, George Eliot and J. D. Salinger before turning to the scientific literature. Good writers have the empathic ability to synthesise understanding of the internal and intimate elements of personality with the impact their behaviours have on others. By comparison, most attempts to diagnose personality disorder in a few words seem laughingly inadequate, a two-dimensional parody of something we all know to be exceedingly complex. Nevertheless, it must not be regarded as impossible; one axiom that I have learnt to repeat is “All personalities are unique but disordered personalities offend in similar ways”. In other words, we do not necessarily need to know everything about someone's personality to recognise the elements that make it disordered.

But we have not yet learnt to separate the wheat from the chaff. The DSM-III revolution was a victory for diagnostic operational criteria and these have continued to sweep all before them, so that they now dominate the landscape of classification. Unfortunately, the valiant attempts to operationalise the characteristics of personality disorder have to date failed to impress and, although it may be regarded as a thankless task to examine why, it is necessary before one can create a better system. We can begin by blaming Kretschmer (1918) and Schneider (1923). Their beautiful descriptions of a range of abnormal personality types, even if they were all listed under the label of psychopathy, had the same impact as Keats' description of his first reading of the works of Homer, when all psychiatrists “ look'd at each other with a wild surmise” and marvelled at the descriptions of the ‘sensitive Bezichungswahn’ and other classical personalities that we have never been able to forget.

Thus, adjectival descriptions such as schizoid, cyclothymic, hysterical (histrionic) and obsessional have become securely attached to personality disorder and are now extremely well known through “long clinical traditions” (Dowson & Grounds, 1995). Our mistake, as pointed out by that precise medical writer, Richard Asher, is to assume that what is well known is necessarily true: those involved in research in the subject have persistently found that these labels actually hinder description and classification. Some, such as Michael Rutter (1987), have abandoned them altogether and there is now a strong lobby to exclude them from formal classification as they have simply not come up to scratch as descriptions that are useful in either clinical or research practice. “Dimensions, not categories” has become the war cry of the reformists and their case is a powerful one, expressed most cogently in the writings of Thomas Widiger, John Livesley and their colleagues over nearly two decades (Widiger & Frances, 1985; Livesley, 1986, 1991; Widiger, 1991; Livesley et al, 1994), which are worth reading because they adumbrate the current debate on the subject by many years.


The main reason for abandoning the present classification is summed up in one word, comorbidity. Comorbidity is the nosologist's nightmare; it shouts, “ you have failed” every time it is mentioned. Comorbidity demonstrates that the boundaries between the disorders that are found to be comorbid are unclear or may not even be identifiable and that the clinical value of the disorder is compromised as a consequence. Thus, when a study demonstrates that a personality disorder such as the borderline condition is found in pure form in fewer than one in 10 of all (relatively) unselected patients with the condition (Fyer et al, 1988) it is reasonable to question the use of the diagnostic description.

However, this does not necessarily mean that all comorbidity is to be decried. The DSM classification made the bold step of identifying a separate axis for personality disorders (and ‘mental retardation’) because it recognised that personality was in a different domain from other clinical disorders (Axis I) and gave recognition to the fact that it is only when personality disorder tends to be extremely striking (e.g. in forensic settings) that it achieves status as a primary diagnosis. The DSM's justification for this new axis is delicately phrased by the wording “ placing on a separate axis ensures that consideration will be given to the possible presence of personality disorders and mental retardation that might otherwise be overlooked when attention is directed to the usually more florid Axis I disorders” (American Psychiatric Association, 2000; p. 28). A more honest explanation may be that both personality disorder and mental retardation are stigmatic terms that psychiatrists like to avoid just as much as do patients, and they have to be forced to consider them. This stigmatisation is likely to be made even worse by a new addition to the personality disorder lexicon, described sometimes as JSS, the Jack Straw Syndrome, or dangerous and severe personality disorder (D&SPD; Home Office and Department of Health, 1999).


There is no doubt that the task force behind the original DSM-III personality disorder classification achieved a great deal when they produced their 1980 description (American Psychiatric Association, 1980), but this has been largely through the original aim of DSM, stimulating research, rather than advancing knowledge. Although we in the UK tend to regard ourselves as the last ‘ DSM-free zone’ left in the world, we should be aware that some of the strongest criticism of the classification in areas where it has demonstrably failed have come from research workers in the USA. It is fair to add that the DSM classification has repeatedly stated that its “ diagnostic criteria are based on clinical judgement and have not been fully validated” (American Psychiatric Association, 1987: p. xxiv) and it is the success of the classification among rank and file psychiatrists and mental health workers who have no real interest in research that has created many of the current problems. With complex subjects such as personality disorder it is very easy to have a set of prototypical behaviours that allegedly characterise each personality disorder even though every clinician will tell you that this form of stereotyping is quite inappropriate in such a complicated field. John Livesley, with his roots soundly anchored in the UK but now working in Canada, has exposed the inconsistency of the individual criteria for the DSM personality disorders over and over again in his series of elegant papers and two books (Livesley, 1995, 2000), any one of which should have been sufficient to sink USS Personality Disorders without trace. Unfortunately, or fortunately, depending on your point of view, the DSM classification is far too robust to be sunk by mere evidence. Although I used to reclassify DSM as “diagnosis for simple minds” it has been better described to me (admittedly by an ICD enthusiast) as “diagnosis and a source of money”, and there is absolutely no doubt that it is a roaring commercial success. Because of this the alternative classifications for personality disorder tend to get less of a hearing but there are clear options, and for those who wish to know about these the writings of Lee Anna Clark (Clark, 1995; Clark et al, 1997) are particularly well argued. There has also been a welcome move away from the slavish adherence to structured interviews carried out with subjects in the early years of enthusiasm after DSM-III, towards a more rounded view in which the opinions of key informants are acknowledged as a major resource. The early work of Anthony Mann (Mann et al, 1981) and, more recently, Drew Westen (Westen, 1997; Westen & Shedler, 1999) are notable examples of this.

Whatever misgivings one may have of the descriptions of individual personality disorders, it must be acknowledged that the subject has been stimulated enormously by debate and research enqiury in the past 20 years. The relationship between personality and mental state disorders has proved a valuable sounding board for a range of theoretical constructs, empirical studies and valuable insight. Examples of these include: the psychobiological models of Robert Cloninger and Larry Siever (Cloninger, 1987; Siever & Davis, 1991; Cloninger et al, 1993); the ways in which personality disorder and mental states interact (Lyons et al, 1997); and the relative contributions of personality and symptoms to diagnoses such as social phobia (Herbert et al, 1992), somatoform disorders (Bass & Murphy, 1995) and schizophrenia (Peralta et al, 1991).


Most importantly, this interest has focused attention on treatment and outcome of personality disorders. Fifty years ago the only treatment claiming success in the management of personality disorders was the therapeutic community, and although this remains live and kicking (Campling & Haigh, 1999) we now have at least seven other treatment types: drug therapy (which includes some interesting studies demonstrating clear efficacy (e.g. Coccaro & Kavoussi, 1997)), psychodynamic therapy, day hospital intervention (including group therapy and partial hospitalisation), cognitive—behavioural therapy (Davidson, 2000), cognitive analytic therapy (Ryle, 1997), interpersonal therapy (Benjamin, 1993) and dialectical behaviour therapy (Linehan, 1992). Interpretation of the findings of therapy is difficult because improvement often occurs in spite of treatment rather than because of it (Stone, 1993), and also because most studies have been carried out in patients with borderline personality disorder, a condition that differs fundamentally from the rest of the personality disorder spectrum (Tyrer, 1999). Of all the treatment trials carried out to date the most impressive has been the study of partial hospitalisation by Bateman & Fonagy (1999), and this is worth a close read for those who wish to replicate it in practice. Success is likely to involve extremely good integration of all those involved in treatment.


Finally, it is worth reminding ourselves that the study of personality disorder has humour, not always black, as a companion. One of my first engagements was with someone who asked me early in our contact why she had no friends. I informed her dutifully that, as in my view she had a personality disorder (with primary narcissistic and histrionic features), this was not at all surprising because the possession of this condition tended to alienate others. In attempting to tackle this I suggested she started a self-help group for personality disorders. She returned 2 months later and I politely asked how many members had joined her group. “I'm not a group yet”, she said. “When they heard it was called ‘personality disorder’ no one wanted to join”. I reassured her that this was normal at first and, in any case, her personality structure was used to informing her that she was a special case. After I got up from the floor we had a discussion about the negative implications of narcissism.

When the frustrations, and not a little pomposity, of writings on personality disorder begin to wear away your enthusiasm for the subject it does no harm to read Glenn Ellenbogen (1980), who has described the exploits of his alter ego, Ernst von Krankman, who carried out important work on the phenomenon of vegetarianism and its significance in personality terms.

Von Krankman argues that vegetarians abhor flesh because they are fixed at the oral—sadistic stage of development. Searching fearlessly for a common explanation he alights upon personality as the cause, and illuminates our dim knowledge base with a full description of the vegetarian personality, which is characterised by “ruthless acts of an oral—sadistic nature directed against vegetables”. Oddly enough, he died shortly afterwards (after ingesting poisonous mushrooms) but has left the DSM diagnosis of Vegetarian Personality Disorder (301.85) for posterity in which the diagnostic criteria are unequivocally precise, including “an overconcern for the feelings and physical well-being of animals”, “ a conspicuous lack of empathy for the feelings of murdered vegetables” and “a paranoid and hypervigilant preoccupation with the oral zone and food consumption”, which in the advanced stages of the disorder makes the individual “suspect that there are minuscule animal by-products mixed in with his food”. Garnering such insights helps us to realise the sophistication and precision of current thinking on personality disorder and makes us pause in wonderment at how we must be viewed by the rest of the world.