Sage Journals: Discover world-class research

Abstract

Objective:

The aim of this study was to review and discuss the evidence for dimensional classification of personality disorders and the historical and sociological bases of psychiatric nosology and research.

Method:

Categorical and dimensional conceptualisations of personality disorder are reviewed, with a focus on the Diagnostic and Statistical Manual of Mental Disorders–system’s categorisation and the Five-Factor Model of personality. This frames the events leading up to the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition, personality disorder debacle, where the implementation of a hybrid model was blocked in a last-minute intervention by the American Psychiatric Association Board of Trustees. Explanations for these events are discussed, including the existence of invisible colleges of researchers and the fear of risking a ‘scientific revolution’ in psychiatry.

Results:

A failure to recognise extra-scientific factors at work in classification of mental illness can have a profound and long-lasting influence on psychiatric nosology. In the end it was not scientific factors that led to the failure of the hybrid model of personality disorders, but opposing forces within the mental health community in general and the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition, Task Force in particular.

Conclusion:

Substantial evidence has accrued over the past decades in support of a dimensional model of personality disorders. The events surrounding the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition, Personality and Personality Disorders Work Group show the difficulties in reconciling two different worldviews with a hybrid model. They also indicate the future of a psychiatric nosology that will be increasingly concerned with dimensional classification of mental illness. As such, the road is paved for more substantial changes to personality disorder classification in the International Classification of Diseases, 11th Revision, in 2017.

Keywords

Diagnosis philosophy of science methodology history of psychiatry personality disorders

The classification of mental disorders in the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5) remains overwhelmingly categorical, despite substantial evidence that personality disorders (PDs), at least, are better conceptualised from a dimensional perspective. Why is this so? What would it have meant, on a clinical, professional and scientific level, if dimensions – derived from research on normal personality – had been applied to psychiatric disorders? In the following, we review the literature supporting the legitimacy of a dimensional model of PDs, and discuss various explanations for why such a model was ultimately rejected – including the possible fear that it would induce a ‘scientific revolution’ in psychiatry, leading to a fundamental paradigm shift in the field.

Discussion of the fundamentals of psychiatric nosology is as relevant today as a hundred years ago (Kendell, 1975; Spitzer and Wilson, 1975; Zachar and Kendler, 2010). As explicitly acknowledged in the DSM-5, diagnosis should be guided by current knowledge and based on theoretical and empirical research (American Psychiatric Association [APA], 2013). We argue that this has not been the case with regard to PDs, which poses a problem for the validity of these diagnoses.

PD as a dimensional diagnosis, existing on a spectrum from normal to abnormal, is in some ways the most likely case (Clark, 2007). This is because there is (1) a lack of aetiological evidence for a categorical conceptualisation and (2) relative academic agreement on a theory of dimensional trait models of personality. This paper presents the events surrounding the publication of the DSM-5 PDs section and introduces the debate surrounding dimensional and categorical conceptualisations. It attempts to explain the lack of a change to extra-scientific factors at work within the APA, centring on Thomas Kuhn’s philosophy of science to consider why the Work Group’s proposal failed, despite the fact that it incorporated the strengths of several dimensional models while retaining the possibility for categorical diagnoses.

The process leading up to the publication of the DSM-5

During the process leading up to the publication of the DSM-5 in late 2013, the APA considered a fundamental change to the conceptualisation of the PD diagnoses (Skodol and Bender, 2009). In several proposals, the Personality and Personality Disorders Work Group recommended that dimensional personality trait scores should be the foundation for PD diagnoses (Bastiaansen et al., 2013). However, in the end, the APA Board of Trustees vetoed the final proposal, and instead substituted the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision (DSM-IV-TR) criteria almost verbatim (APA, 2013; Black, 2013; Oldham, 2013). The hybrid proposal was subsequently placed in section III of the DSM-5 as an alternative model for PDs. With the removal of the axial system that had acknowledged the special circumstances surrounding the PDs diagnoses (Williams, 1985), the PD section is now a set of categorical diagnoses on par with the others in the DSM-5 – in a sense, making these diagnoses even more categorically based than they had been previously.

The DSM-5 Personality and Personality Disorders Work Group, as the name implies, was specifically created to generate a coherent perspective on personality and personality psychopathology together, rather than focusing solely on psychopathology. Most of its members had been working with dimensional models in research on personality for many years. They had published criticisms of categorical models of psychopathology in general (Skodol et al., 2013a) and argued for a ‘paradigm shift’ in psychiatric nosology away from a neo-Kraepelinian model (Kupfer et al., 2002). At the time, a shift to a dimensional model of PDs was seen as a ‘spearhead’ that could lead to similar revisions of other diagnostic categories in a ‘contagion’ effect (Rounsaville et al., 2002; Widiger and Trull, 2007).

Sometime during the late 2000s, a change away from a focus on dimensionality took place within the DSM-5 Task Force. The APA Board of Trustees established an oversight committee and two peer-review groups, which reported directly to the Board of Trustees (Skodol et al., 2013a; Widiger, 2013). When the Personality and Personality Disorders Work Group published its first proposal in 2010, it was one of the only work groups that had retained dimensional elements, a position supported by interest groups and the influential National Institute of Mental Health (Insel et al., 2010). At this point, the National Institute of Mental Health had signalled to end funding of categorical research because of a limited scientific yield. These events seem to have had a destabilising effect on the Work Group and ‘catalysed tensions’ within the group (Skodol et al., 2013a: 344). It also spawned criticism from psychologists who thought that the initial ‘hybrid’ proposal, which consisted of both categorical and dimensional elements, did not go far enough (Livesley, 2010; Widiger, 2011a). In the end, two members resigned because of this disagreement (Blashfield and Reynolds, 2012). The DSM-5 leadership cut the funding for a separate field trial for the hybrid model proposal, even though earlier trials had demonstrated good reliability. Those in favour of a change felt that opposition was largely an effect of personal investments in previous diagnostic approaches (Skodol et al., 2013b). The conclusion to this ‘debacle’ (Paris, 2013; Widiger, 2013) seems to be that the proposal was largely rejected due to opposition within the APA and from the general field of psychiatrists, even though PDs interest groups and personality researchers were in favour of a change (Lanier et al., 2013; Skodol et al., 2013a; Thomas et al., 2013).

Personality science: PDs as a dimensional construct

While the history of PD research is largely a psychiatric one, the history of personality research is overwhelmingly psychological. Two approaches, the lexical and the statistical, have been especially influential. Originating with Galton, and most rigorously followed by Allport and Odbert (see McCrae and Costa, 2003), the lexical approach assumes that the most important personality characteristics will become a part of the concepts of the natural language, and that studying trait terms in the dictionary can yield a typology of human personality. The other approach, the statistical, is based on factor analysis of empirical results of surveys, and is the most widely used of the two approaches today (McCrae and Costa, 2003). In Eysenck’s hierarchical model of personality, personality is split into the three traits of extraversion, neuroticism and psychoticism (Eysenck, 1992). Eysenck’s model has influenced contemporary models of personality and personality traits based on factor analysis. The most influential of these is Robert McCrae and Paul Costa’s Five Factor Theory of Personality (McCrae and Costa, 1997).

PDs are defined in terms of experiences and behaviour showing a marked deviation from cultural norms. Whether they are best understood as pathological and maladaptive expressions of normative personality variations, or categorically distinct from normal personality functioning is largely an empirical question. If PDs are dimensional extremes of normal personality traits, specific constellations of traits should meaningfully mirror each disorder, which would pose a problem for a categorical conceptualisation.

Empirical support for a dimensional conceptualisation

Using a dimensional model of personality traits to understand PDs is a well established approach (Samuel, 2011; Watson et al., 2008). The DSM-5 section III alternate model includes a dimensional measurement of self and interpersonal functioning and reduces the number of PDs from 10 to 6. Most importantly, it requires a diagnosis to include a score on each of five dimensional personality traits based on a total of 25 facets (APA, 2013; Lanier et al., 2013; Miller et al., 2015; Morey and Skodol, 2013). These five traits are ‘maladaptive variants’ of the five domains of the Five-Factor Model (FFM) (APA, 2013: 773). The critical question is whether a dimensional model meaningfully expresses PDs. A literature search of the academic databases PsychINFO and MEDLINE yields four meta-analyses on the applicability of a dimensional model of PDs, with a total of 52 independent samples and 13,640 unique respondents distributed over clinical and non-clinical populations (O’Connor, 2005; O’Connor and Dyce, 1998; Samuel and Widiger, 2008; Saulsman and Page, 2004). These meta-analyses conclude that personality pathology can be adequately understood as constellations of extreme scores on the Big Five factors.

Since the last meta-analysis in 2008, there has been a virtual explosion in the number of empirical studies concerned with the relationship of dimensional traits to PDs symptomatology (Widiger et al., 2013). The essence of this research is the finding that a dimensional conceptualisation can be used to predict the response to pharmacological treatment (Saunders and Silk, 2009) and psychotherapy (Barnicot et al., 2012), and that a dimensional model of psychopathology appears valid cross-culturally (Madsen et al., 2006; Rossier and Rigozzi, 2008; Thiry, 2012). For example, Morey et al. (2012) found support for the consistency of dysfunctional and normal personality traits in a 10-year longitudinal study, and Saunders and Silk (2009) found trait neuroticism and impulsivity-aggression to be predictive of response to antipsychotic medications in samples of patients with borderline PD. In general, these newer studies show continued empirical support for the theoretical and practical applicability of a dimensional model (Bastiaansen et al., 2011; De Fruyt et al., 2013; Le Corff and Toupin, 2010; Miller et al., 2015; Morey and Skodol, 2013; Samuel et al., 2013; Thomas et al., 2013; Watson et al., 2013; Widiger, 2011b). Even so, such an approach has not been universally accepted.

Criticism of a dimensional conceptualisation

The most often voiced criticism of a dimensional conceptualisation is the dimensional model’s lack of utility as a practical and prognostic tool (Clark, 1993; Skodol et al., 2005). Critics argue that psychiatrists working with PDs do not have time for the extra work that would go into a multidimensional diagnosis, that mental health professionals, patients and relatives are all used to the categorical model, and that a new conceptualisation would require lengthy training (Black, 2013; Paris, 2013). Furthermore, it is argued that some of the existing research into the treatment of PDs would become obsolete with a new conceptualisation. Psychologists, however, appear to have a different view. Mullins-Sweatt and Widiger (2011), in a survey of almost 1800 American psychologists, found that most considered an FFM PDs model more clinically useful than the DSM model in terms of global personality description, client communication, comprehensiveness and treatment planning. For those patients not fitting the criteria for 1 of the 10 PDs, who are usually given a diagnosis of Personality Disorder Not Otherwise Specified (PD-NOS), a dimensional model also had significantly greater ease of application and was judged to be of higher value to professional communication. Other studies have shown a dimensional model to have moderate to high inter-rater reliability (Few et al., 2010) and convergent and discriminant validity (Miller et al., 2012). As it stands now, the categorical model’s biggest weaknessesare overlapping diagnoses and the overuse of the PD-NOS category (Lowe and Widiger, 2009; Widiger, 2012) – in other words, arguments against the utility of a categorical model.

The argument that a change of diagnostic system would be negative because it would require clinicians to spend time learning the new model is, for obvious reasons, not in itself a convincing argument against the dimensional model, since this holds true for all changes. Nor is the argument that some research might become obsolete. A direct analogy is that it would have been unthinkable to retain a diagnosis of hypertension as a categorical disease simply because people were used to thinking in this way or because certain strands of research would lose their applicability.

A last argument against a shift is that it would potentially create unanticipated changes in diagnoses and treatment in the future (Samuel et al., 2012). This, however, is true of both changes and lack of changes. The above research, however, indicates that the majority of changes would be positive. While unresolved uncertainties do exist, they are smaller and of a more transient nature than the weaknesses of the existing categorical model. These weaknesses include, as has been noted, extensive co-occurrence of PDs; heterogeneity between patients; little empirical basis for diagnostic thresholds and reification of disorders; over-usage of the PD-NOS category implying poor coverage of personality psychopathology; and a lack of new scientific breakthroughs (Blashfield and Reynolds, 2012; Skodol et al., 2013a; Widiger and Trull, 2007).

Extra-scientific factors at work in classification

Viewing researchers as individuals and members of scientific communities helps explain the chain of events described above. Three complementary perspectives can explain why a categorical model was kept. The first of these deals with cognitive processes and is based on Rosch and colleagues’ work on categorisation. The second is about the jurisdictions of different mental health professions, most notably psychiatry and psychology. The final perspective, which encompasses the preceding ones, sees the events as part of a scientific cycle of paradigm shifts and normal science.

Cognitive psychology and categorisation

Humans intuitively categorise things and create concepts, and from the moment they are born live in a highly categorised world (Rosch, 1978, 1999). Grouping distinguishable instances of something is principal to the creation of any nomenclature. In the history of psychiatry, mental illness sometimes has had a tendency to become reified, to be made thing-like and seen as part of the same system of natural phenomena as processes in the human nervous system and subject to the same laws governing the classification of the natural world of animals and plants. To classify is fundamental (Barsalou, 1983, 2009), and it can be argued that the partition of disorders of personality into discrete categories is the cognitively primary way of looking at the world. The intuitive tendency is to attempt to ‘carve nature at its joints’, grouping mental disorders as diseases categorically different from normal functioning. This is corroborated in the field of medicine and psychiatry through centuries of work to classify and separate diseases. Apparent even in folk taxonomies of mental disorder (Blashfield and Flanaghan, 2008), some of the opposition to dimensional categories can be interpreted as a reaction to the perceived ‘chaos’ it would impose on the classificatory system, as evidenced in the (empirically un-grounded) claims of low clinical utility (Spitzer et al., 2008). Concepts and categories exist not in a vacuum but in social situations. When diseases are conceptualised as discrete things, admitting to the dimensional nature of some things would upset the entire system of classification. But even in this classical conceptualisation of PDs, there are several problematic elements. As we have seen, the borders of each PDs are inherently vague and there is a lack of specific necessary and sufficient conditions for membership. Rather, the categories seem more like prototype classification in a categorical guise. For example, the prototypical borderline personality is an abstraction that the diagnostic criteria of the PD attempt to express (Miller, 2012). It therefore seems that even the current conceptualisation of PD does not conform to an entirely categorical understanding of disease. A related approach is theory theory, in which conceptualisation is analogous to scientific progress, and that changes in conceptualisation are analogous to the gestalt-shift that occurs in a change of paradigm (Gabora et al., 2008; Quinn and Oates, 2004), akin to a Kuhnian framework.

Professional jurisdiction

Another hypothesis is that the failure of the Personality and Personality Disorders Work Group proposal was due to extra-scientific forces struggling for professional jurisdiction and enclaves of researchers with specific agendas. An example of an extra-scientific argument is some of the Work Group members proposing that a number of the psychiatrists’ resistance to change is due to their personal investment in the existing model (Skodol et al., 2013b). The argument here is essentially social-psychological and sociological, and falls into two parts: professional jurisdiction and ‘invisible colleges’. Furthermore, as the ‘debacle’ of PDs in DSM-5 has shown, these factors are also relevant today.

According to the sociology of professions (Freidson, 2001), professions have historically been engaged in conflicts for jurisdiction. In the field of mental health care, several professional groups are present, the most important for this discussion being psychiatrists and clinical psychologists. They often have similar lengths of training and experience and both require licensing to practice, but are schooled in different traditions. Moreover, top positions in psychiatric facilities are usually open only to psychiatrists, and there are differences in average salaries reflecting a difference in status (Starr, 1982). The preparation of the Diagnostic and Statistical Manual of Mental Disorders, 3rd Edition (DSM-III), a psychiatric classification, met resistance from the American Psychological Association, which attempted to create an alternative psychological classificatory system (Blashfield, 1984). This directly reflected fears of a loss of jurisdiction if the DSM were accepted as the authoritative classification. One reason for the ultimate rejection of the PDs proposal could be the fear that it would constitute a spearhead for psychologists to gain a larger influence on classification. Mirroring this, work on the DSM-5 has created a backlash from the British Psychological Society, recommending ‘a revision of the way mental distress is thought about, starting with recognition of the overwhelming evidence that it is on a spectrum with “normal” experience’ (British Psychological Society, 2011), a position they are still advancing today (British Psychological Society, 2014). Early on, people who favoured the dimensional model have been psychologists, and advocates for the categorical model have tended to be psychiatrists (Black, 2013; Blashfield, 1993). A counter argument is that there were both psychologists and psychiatrists present in the Work Group (Blashfield and Reynolds, 2012) and on the DSM-5 Task Force. However, that it was ultimately the APA Board of Trustees who rejected the hybrid proposal points to professional interests playing a role. Their job is to further the interests of the American psychiatric profession.

The second approach views researchers as social-psychological in-groups, as ‘invisible colleges’, a term that originates from the psychology of science: they are colleges because they collaborate on research, attend the same conferences, cite each other and share the same social affiliations and professional attitudes; they are invisible, because they are not institutionalised or apparent to people outside their field (Blashfield, 1984; Blashfield and Reynolds, 2012; Price and Beaver, 1966). A basic assumption is that researchers are influenced by the approaches and traditions they have studied under and that dominate the institutions in which they work, and that these ‘non-cognitive’ factors influence the direction of research (Harari, 2001). Blashfield (1984) found one of these invisible colleges surrounding the Feighner paper by looking at the citations leading up to the DSM-III. Analysing the references given in the official Work Group proposal, Blashfield and Reynolds (2012) found signs of an invisible college made up of the university affiliation and citation patterns of the group members.

While invisible colleges have positive effects, in this particular instance, it might explain the failure of the proposed model. For example, the Work Group failed to discuss influential models such as Theodore Millon’s (2011) dimensional model of PDs or the theories of influential neo-analysts such as Kohut and Kernberg. The argument is that a lack of authority in the psychiatric as well as in the psychological community further weakened the Work Group’s proposal. What it does show, however, is how social-psychological and political elements influence research and professional practice. The analogy of a biological taxonomy is therefore not entirely adequate when it comes to the science of psychiatric classification.

An objection to these two elements would be that they are mutually incompatible, since they point to two different divisions, namely between psychiatrists and psychologists on the one hand, and in-groups of researchers on the other. However, neither explanation should be seen as complete, but instead as processes influencing behaviour. They are compatible because they do not claim to be exhaustive.

Insights from Kuhn’s philosophy of science

Thomas Kuhn’s theory of scientific progress provides a meta-framework for these approaches. Although the term ‘paradigm-shift’ has been used to describe the hybrid model (Lanier et al., 2013; Skodol et al., 2013b; Widiger, 2013), no paper has explicitly referred to Thomas Kuhn’s original use of the term in his philosophy of science. ‘Paradigm’ has become so ubiquitous and has been used in so many ways that it is in danger of becoming an empty concept (Andersen and Faye, 2006). However, the theory originally proposed by Kuhn in 1962 and revised until his death holds promise as an integrative framework to describe the evolution of the science of psychiatric nosology as a social-psychological enterprise.

An outline

Kuhn thought that research communities are best understood as science-based professions and research as a kind of social behaviour. Most of the time, scientific research is normal science, defined as ‘a strenuous and devoted attempt to force nature into the conceptual boxes supplied by professional education’ (Kuhn, 1974 [1962]: 5). It is a process of ‘mopping-up’ and assimilation, of solving the puzzles set out by the overarching set of scientific findings and methods that all members of a community of researchers adhere to. Researchers subject themselves to paradigms through their education and membership of a certain community of science, akin to what we saw above concerning professional jurisdiction. Changes of the governing paradigm, scientific revolutions, are caused by recurring, accumulating anomalies that cannot be prescribed to the individual researcher. The following extended period of crisis leads to a gestalt switch, where the old paradigm is gradually superseded by a new paradigm. The new paradigm is said to be incommensurable with the old one, because translating concepts between paradigms is not possible without a loss of meaning (Kuhn, 1970). The paradigm fundamental to a science is a kind of mindset prerequisite to having the shared concepts in a community, and not the concepts in themselves. The analogy here is that of evolution by speciation as opposed to mutation. Science is not cumulative in the same way Popper or the logical positivists thought.

Normal science in psychiatry and psychology

The historical efforts to classify psychopathology mirror Kuhn’s fundamental ideas. The state of disarray in the field before the 20th century, the gradual emergence of distinct approaches, and the eventual dominance of a neo-Kraepelinian approach fit Kuhn’s idea of the move from a period of pre-scientific research to a widely accepted paradigm and normal science. The DSM-III can be said to be the result of a specific paradigm set out in Robins and Guze’s (1970) work on validation. It was based on a textbook, an essential element in the move to normal science (Kraepelin, 1899; Kuhn, 1974 [1962]). It had scientific principles and assumptions about the nature of psychopathology as qualitatively distinct mental disorders. Furthermore, it presented a bounded set of puzzles to solve, as set out in the Feighner paper (Feighner et al., 1972; Kendler et al., 2010). The proponents of this approach specifically felt themselves a part of a ‘paradigm’ (Klerman, 1978). Institutionalised in the APA, normal scientific research in this paradigm consisted of validating distinct syndromes and discovering common aetiologies via laboratory markers of disease. The paradigm is supplied by the psychiatrists’ training in the medical model and the neo-Kraepelinian ‘paradigm’ (Klerman, 1978).

The continuation of a categorical model of PDs fits the framework set out by what might be called the neo-Kraepelinian paradigm. It is important here to acknowledge that normal science is not a lesser form of science than revolutionary science, a view erroneously attributed to Kuhn (Baltas et al., 2000). The theory is largely conservative and places emphasis on the continuity of science, even in the face of anomalies. Scientific progress in the sciences is only possible within a paradigm. Showing the field of psychiatry to be conducting normal science is therefore not in itself a criticism. Psychiatry cannot be understood as mono-paradigmatic, at least not in the same way as, for instance, physics. It does seem clear, however, that there has been a dominant paradigm, which has had both informal support through invisible colleges and formal support through institutionalisation of its approach.

Within psychology, the last hundred years can best be described as pre-scientific, where different paradigms struggle for dominance (Kuhn, 1974 [1962]). Nevertheless, this might have changed in the last two decades, as least concerning the study of the structure of personality. In the 1990s, Kuhn argued that work in psychology and the social sciences was beginning to resemble normal science (Kuhn, 2000). The relative consensus on a five-factor trait structure and the large body of empirical research support this view, even if any single term describing a scientific field with hundreds of researchers and research communities will inherently be an ideal-type model. The strength of this approach is its ability to explain the events surrounding the DSM-5 PDs.

Paradigm change, revolution and gestalt switch

The evidence of the dimensionality of PDs now appears as a single anomaly amongst many rather than a falsification of a categorical approach. The ‘DSM-5 debacle’ follows a course similar to other crises within a science. The neo-Kraepelinian paradigm has failed to identify specific biological markers of DSM PDs, the diagnostic co-occurrence has not been adequately dealt with, and the diagnoses have failed to provide coverage of personality pathology. The main problem is therefore not only that another model has accrued more empirical support, but that the internal consistency of the existing model is low. The sine qua non of categorical concepts consists in their being mutually exclusive and collectively exhaustive. Discord from within psychiatry and pressure from without is more readily explained from a Kuhnian perspective of crisis rather than from a positivist perspective. Other signs of crisis involve the loss of support from the National Institute of Mental Health and more overt criticism from psychological circles. In this way, this is a crisis stemming both from the failures of the old paradigm, and the emergence of a new one.

This perspective is also better than its falsificationist equivalent in explaining the emergence of an alternate model, which also has its weaknesses and limitations. If a theory had to adhere to the standard of falsification, says Kuhn (1974 [1962]), every theory would be born falsified. The alternate, emerging paradigm is incompatible with the traditional medical model, but perhaps not with the idea of endophenotypes or ‘subclinical markers’ of psychopathology (Widiger, 2012).

Ultimately, adding to the failure of the model are the problems inherent in a hybrid model that tries to bridge two paradigms. According to the incommensurability thesis, even though individuals can be ‘bilinguals’, concepts cannot simply be carried across the chasm between paradigms, which explains why the proposal was met by resistance from both camps. But if paradigms are largely unconscious mental frameworks, would this not mean that change in the conceptualisation of PDs in the future is also unlikely? An argument against this is the possibility of one paradigm superseding another not because of a dying-out of a generation of practitioners, but by individuals making the gestalt switch to the new paradigm. As such, a scientific revolution as something that occurs slowly and over an extended period explains the controversies over the DSM-5 quite well. This also does not rule out progress in conceptualisation: that by changing paradigms we actually come closer to describing ‘something permanent, fixed, and stable’ (Kuhn, 2000: 104).

An overarching Kuhnian framework

The theoretical, empirical and pragmatic evidence pointing towards a dimensional conceptualisation of PDs combines with anomalies in the neo-Kraepelinian paradigm to produce a crisis in the field and disagreement on which road to take in the future. However, the dominant paradigm is still categorical with a conceptualisation of psychopathology as disease, and the momentum of a tradition of normal scientific research is still with us, because reconceptualisation to a dimensional model is neither cognitively primary nor an easy task. Different educational backgrounds make for different perspectives within the field of psychopathology, and issues of professional jurisdiction further hinder a change. Admitting to a dimensional model of PDs risks fundamental change to other diagnoses as well.

The Work Group members failed to take these things into account, largely because they themselves constituted an invisible college. The hybrid model attempted to span two paradigms, and was subsequently criticised by practitioners from both communities. The incommensurability or taxonomic differences created a kind of untranslatability. In other words, the odds were heavily stacked against a hybrid dimensional-categorical model. That it was not rejected entirely, but placed in section III of the DSM-5, tells us something about the future of the field.

Limitations of the approach

Two limitations of the framework as it is used here are particularly relevant. First, it follows from this framework that no entirely objective criteria with which to judge the conceptualisation of PDs exist. As such, one might argue that this defeats the purpose of this paper, namely to discuss the two approaches. If there are no objective criteria, all perspectives are equal. However, this relativistic stance is not necessary. Even if the lack of the ‘perspectiveless perspective’ is accepted, this is not as problematic for a Kuhnian view of science as it might seem. There is a normative aspect to the approach, and rationality to the gestalt switches. Bilingualism means that individuals can change paradigms and weigh the merits of each approach. Furthermore, the Kuhnian framework is stronger than both a falsificationist approach and a post-modernist approach in explaining the events. On the one hand, a falsificationist approach does not seem to be able to explain the events surrounding the continuation of the categorical model. On the other hand, a post-modernist approach, which would assert that no criteria for choosing between paradigms are possible, would also throw away the possibility of progress in the sciences. This seems counter-intuitive in light of the progress actually observed on a continual basis and would in essence mean that any discussion of paradigms would be fruitless.

Second, another criticism of this approach is that it deals largely with ideal types. In reality, the field is more complex than this framework admits to. However, the very concept of a model is a simplification of a complex reality. And even in a complex field, distinct positions emerge, positions that we have argued align with what would be expected from the sociology of professions and the philosophy of science literature. The strength of a model is its ability to explain and predict, and the Kuhnian framework has been able to explain the events described in this paper.

Towards the ICD-11

Empirical studies of the hybrid model of PDs continue to emerge. Over time, clinicians might become more accustomed to this model, which might increase the support for it in psychiatric circles. Selective research endowments from the National Institute of Mental Health might also make the change more likely. But even though the Institute supports a dimensional approach, they also argue for a strict biologically based approach to psychopathology, characterising all mental disorders as brain diseases, which might be problematic for other approaches to psychiatry.

The next big change to psychiatric classification is the International Classification of Diseases, 11th Revision (ICD-11), scheduled to be released in 2017. The diagnoses of PDs in the International Classification of Diseases, 10th Revision (ICD-10) are categorical, but the chair of the ICD-11 Working Group for the Revision of Personality Disorders, Peter Tyrer (2012, 2013), is supporting a radical shift to a dimensional trait model of PDs. The ICD-11 proposal comprises a severity rating in each of four trait domains. Even though it is more comprehensively dimensional than the DSM-5 Personality and Personality Disorders Work Group proposal, this means that it is also mono-paradigmatic. According to the framework outlined above, the ICD-11 proposal therefore stands a better chance of being accepted. As such, it presents a possible test of this framework.

Conclusion

A discussion of scientific progress is ultimately incomplete unless it also acknowledges researchers as psychological beings and the influence of social-psychological processes on research communities. The primary reason why the proposed hybrid dimensional-categorical model failed to gain traction as the superior conceptualisation of PDs was that it tried to bridge two incommensurable worldviews and therefore failed to garner support from either.

The evidence pointing to the superiority of a dimensional model of PDs over a categorical one is substantial. With signs of discontent from influential groups in the debate, the path seems set for a paradigmatic change to the conceptualisation of not just PD, but perhaps psychiatric nosology in general, in the upcoming ICD-11 and the abandonment of research into categorical entities. This might set the stage for a change in a future revision of the DSM. This change, however, is some years away. Until then, the classificatory system of PDs will remain fundamentally flawed, as it does not reflect the best understanding of what constitutes mental illness, largely due to extra-scientific factors and fears of a possible paradigm shift having influenced modern psychiatric classification.

Footnotes

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

American Psychiatric Association (2013) Diagnostic and Statistical Manual of Mental Disorders, 5th Edition. Arlington, VA: American Psychiatric Publishing.

Andersen

Faye

(2006) Arven Efter Kuhn. København: Forlaget Samfundslitteratur.

Baltas

Gavroglu

Kindi

(2000) A discussion with Thomas Kuhn. In: Kuhn

(ed.) The Road Since Structure: Philosophical Essays, 1970–1993 with an Autobiographical Interview. Chicago, IL: The University of Chicago Press, pp. 253–325.

Barnicot

Katsakou

Bhatti

. (2012) Factors predicting the outcome of psychotherapy for borderline personality disorder: A systematic review. Clinical Psychology Review 32: 400–412.

Barsalou

(1983) Ad hoc categories. Memory & Cognition 11: 211–227.

Barsalou

(2009) Simulation, situated conceptualization, and prediction. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 364: 1281–1289.

Bastiaansen

De Fruyt

Rossi

. (2013) Personality disorder dysfunction versus traits: Structural and conceptual issues. Personality Disorders: Theory, Research, and Treatment 4: 293–303.

Bastiaansen

Rossi

Schotte

. (2011) The structure of personality disorders: Comparing the DSM-IV-TR axis II classification with the five-factor model framework using structural equation modelling. Journal of Personality Disorders 25: 378–396.

Black

(2013) DSM-5 is approved, but personality disorders criteria have not changed. Annals of Clinical Psychiatry 25: 1.

10.

Blashfield

(1984) The Classification of Psychopathology. New York: Plenum Press.

11.

Blashfield

(1993) Variants of categorical and dimensional models. Psychological Inquiry 4: 95–98.

12.

Blashfield

Flanaghan

(2008) Clinician’s folk taxonomies of mental disorders. Philosophy, Psychiatry, & Psychology 14: 249–269.

13.

Blashfield

Reynolds

(2012) An invisible college view of the DSM-5 personality disorder classification. Journal of Personality Disorders 26: 821–829.

14.

British Psychological Society (2011) Response to the American Psychiatric Association: DSM-5 development. Available at: http://apps.bps.org.uk/_publicationfiles/consultation-responses/DSM-5%202011%20-%20BPS%20response.pdf (accessed 18 May 2014).

15.

British Psychological Society (2014) Understanding psychosis and schizophrenia. Available at: www.bps.org.uk/system/files/user-files/Division%20of%20Clinical%20Psychology/public/understanding_psychosis_-_final_19th_nov_2014.pdf (accessed 18 January 2015).

16.

Clark

(1993) Personality disorder diagnosis: Limitations of the five-factor model. Psychological Inquiry 4: 100–104.

17.

Clark

(2007) Assessment and diagnosis of personality disorder: Perennial issues and an emerging reconceptualization. Annual Review of Psychology 58: 227–257.

18.

De Fruyt

De Clercq

De Bolle

. (2013) General and maladaptive traits in a five-factor framework for DSM-5 in a university student sample. Assessment 20: 295–307.

19.

Eysenck

(1992) Four ways five factors are not basic. Personality and Individual Differences 13: 667–673.

20.

Feighner

Robins

Guze

. (1972) Diagnostic criteria for use in psychiatric research. Archives of General Psychiatry 26: 57–63.

21.

Few

Miller

Morse

. (2010) Examining the reliability and validity of clinician ratings on the five-factor model score sheet. Assessment 17: 440–453.

22.

Freidson

(2001) Professionalism: The Third Logic. Cambridge: Polity.

23.

Gabora

Rosch

Aerts

(2008) Toward an ecological theory of concepts. Ecological Psychology 20: 84–116.

24.

Harari

(2001) Whose evidence? Lessons from the philosophy of science and the epistemology of medicine. Australian and New Zealand Journal of Psychiatry 35: 724–730.

25.

Insel

Cuthbert

Garvey

. (2010) Research Domain Criteria (RDoC): Toward a new classification framework for research on mental disorders. The American Journal of Psychiatry 167: 748–751.

26.

Kendell

(1975) The Role of Diagnosis in Psychiatry. London: Blackwell Scientific Publications.

27.

Kendler

Muñoz

Murphy

(2010) The development of the feighner criteria: A historical perspective. The American Journal of Psychiatry 167: 134–142.

28.

Klerman

(1978) The evolution of a scientific nosology. In: Shershow

(ed.) Schizophrenia: Science and Practice. Cambridge, MA: Harvard University Press, pp. 99–121.

29.

Kraepelin

(1899) Psychiatrie: Ein lehrbuch für studirende und aerzte, 6th Edition. Leipzig: Johann Ambrosius Barth.

30.

Kuhn

(1970) Reflections on my critics. In: Lakatos

Musgrave

(eds) Criticism and the Growth of Knowledge. London: Cambridge University Press, pp. 231–279.

31.

Kuhn

(1974 [1962]) The Structure of Scientific Revolutions. Chicago, IL: The University of Chicago Press.

32.

Kuhn

(2000) The Road Since Structure. Philosophical Essays, 1970–1993, with an Autobiographical Interview. London: The University of Chicago Press.

33.

Kupfer

First

Regier

(2002) A Research Agenda for DSM-V. Washington, DC: American Psychiatric Association.

34.

Lanier

Bollinger

Krueger

(2013) Advances in the conceptualization of personality disorders: Issues affecting social work practice and research. Clinical Social Work Journal 41: 155–162.

35.

Le Corff

Toupin

(2010) The five-factor model of personality at the facet level: Association with antisocial personality disorder symptoms and prediction of antisocial behavior. Journal of Psychopathology and Behavioral Assessment 32: 586–594.

36.

Livesley

(2010) Confusion and incoherence in the classification of personality disorder: Commentary on the preliminary proposals for DSM-5. Psychological Injury and Law 3: 304–313.

37.

Lowe

Widiger

(2009) Clinicians’ judgments of clinical utility: A comparison of DSM-IV with dimensional models of general personality. Journal of Personality Disorders 23: 211–229.

38.

McCrae

Costa

Jr (1997) Personality trait structure as a human universal. American Psychologist 52: 509–516.

39.

McCrae

Costa

Jr (2003) Personality in Adulthood – A Five-Factor Theory Perspective. New York: The Guilford Press.

40.

Madsen

Parsons

Grubin

(2006) The relationship between the five-factor model and DSM personality disorder in a sample of child molesters. Personality and Individual Differences 40: 227–236.

41.

Miller

(2012) Five-factor model personality disorder prototypes: A review of their development, validity, and comparison to alternative approaches. Journal of Personality 80: 1565–1591.

42.

Miller

Morse

Nolf

. (2012) Can DSM-IV borderline personality disorder be diagnosed via dimensional personality traits? Implications for the DSM-5 personality disorder proposal. Journal of Abnormal Psychology 121: 944.

43.

Miller

Few

Lynam

. (2015) Pathological personality traits can capture DSM-IV personality disorder types. Personality Disorders: Theory, Research, and Treatment 6: 32–40.

44.

Millon

(2011) Disorders of Personality: Introducing a DSM/ICD Spectrum from Normal to Abnormal, 3rd Edition. Hoboken, NJ: John Wiley and Sons, Inc.

45.

Morey

Skodol

(2013) Convergence between DSM-IV-TR and DSM-5 diagnostic models for personality disorder: Evaluation of strategies for establishing diagnostic thresholds. Journal of Psychiatric Practice 19: 179–193.

46.

Morey

Hopwood

Markowitz

. (2012) Comparison of alternative models for personality disorders, II: 6-, 8-and 10-year follow-up. Psychological Medicine 42: 1705–1713.

47.

Mullins-Sweatt

Widiger

(2011) Clinician’s judgments of the utility of the DSM-IV and five-factor models for personality disordered patients. Journal of Personality Disorders 25: 463–477.

48.

O’Connor

(2005) A search for consensus on the dimensional structure of personality disorders. Journal of Clinical Psychology 61: 323–345.

49.

O’Connor

Dyce

(1998) A test of models of personality disorder configuration. Journal of Abnormal Psychology 107: 3–16.

50.

Oldham

(2013) Personality disorders and DSM-5. Journal of Psychiatric Practice 19: 177.

51.

Paris

(2013) Anatomy of a debacle: Commentary on ‘seeking clarity for future revisions of the personality disorders in DSM-5’. Personality Disorders: Theory, Research, and Treatment 4: 377–378.

52.

Price

Beaver

(1966) Collaboration in an invisible college. American Psychologist 21: 1011–1018.

53.

Quinn

Oates

(2004) Early category representation and concepts. In: Oates

Grayson

(eds) Cognitive and Language Development in Children. Malden, MA: Blackwell Publishing, pp. 21–61.

54.

Robins

Guze

(1970) Establishment of diagnostic validity in psychiatric illness: Its application to schizophrenia. American Journal of Psychiatry 126: 107–111.

55.

Rosch

(1978) Principles of categorization. In: Rosch

Lloyd

(eds) Cognition and Categorization. Hillsdale, NJ: Lawrence Erlbaum, pp. 27–46.

56.

Rosch

(1999) Reclaiming concepts. Journal of Consciousness Studies 6: 61–77.

57.

Rossier

Rigozzi

(2008) Personality disorders and the five-factor model among French speakers in Africa and Europe. The Canadian Journal of Psychiatry [La Revue Canadienne De Psychiatrie] 53: 534–544.

58.

Rounsaville

Alarcon

Andrews

. (2002) Basic nomenclature issues for DSM-V. In: Kupfer

First

Regier

(eds) A Research Agenda for DSM-V. Washington, DC: American Psychiatric Association, pp. 1–29.

59.

Samuel

(2011) Integrating personality disorder with basic personality science. Acta Psychiatrica Scandinavica 123: 325–326.

60.

Samuel

Widiger

(2008) A meta-analytic review of the relationships between the five-factor model and DSM-IV-TR personality disorders: A facet level analysis. Clinical Psychology Review 28: 1326–1342.

61.

Samuel

Carroll

Rounsaville

. (2013) Personality disorders as maladaptive, extreme variants of normal personality: Borderline personality disorder and neuroticism in a substance using sample. Journal of Personality Disorders 27: 625–635.

62.

Samuel

Miller

Widiger

. (2012) Conceptual changes to the definition of borderline personality disorder proposed for DSM-5. Journal of Abnormal Psychology 121: 467–476.

63.

Saulsman

Page

(2004) The five-factor model and personality disorder empirical literature: A meta-analytic review. Clinical Psychology Review 23: 1055–1085.

64.

Saunders

EFH

Silk

(2009) Personality trait dimensions and the pharmacological treatment of borderline personality disorder. Journal of Clinical Psychopharmacology 29: 461–467.

65.

Skodol

Bender

(2009) The future of personality disorders in DSM-V? The American Journal of Psychiatry 166: 388–391.

66.

Skodol

Morey

Bender

. (2013a) The ironic fate of the personality disorders in DSM-5. Personality Disorders: Theory, Research, and Treatment 4: 342–349.

67.

Skodol

Morey

Bender

. (2013b) When is it time to move on? Rejoinder for ‘The ironic fate of the personality disorders in DSM-5’. Personality Disorders: Theory, Research, and Treatment 4: 354.

68.

Skodol

Oldham

Bender

. (2005) Dimensional representations of DSM-IV personality disorders: Relationships to functional impairment. The American Journal of Psychiatry 162: 1919–1925.

69.

Spitzer

Wilson

(1975) Nosology and the official psychiatric nomenclature. In: Freedman

Kaplan

Sadock

(eds) Comprehensive Textbook of Psychiatry, 2nd Edition. Baltimore, MD: Williams & Wilkins, pp. 826–845.

70.

Spitzer

First

Shedler

. (2008) Clinical utility of five dimensional systems for personality diagnosis: A ‘consumer preference’ study. Journal of Nervous and Mental Disease 196: 356–374.

71.

Starr

(1982) The Social Transformation of American Medicine. New York: Basic Books.

72.

Thiry

(2012) An assessment of personality disorders with the five-factor model among Belgian inmates. International Journal of Law and Psychiatry 35: 327–333.

73.

Thomas

Yalch

Krueger

. (2013) The convergent structure of DSM-5 personality trait facets and five-factor model trait domains. Assessment 20: 308–311.

74.

Tyrer

(2012) Diagnostic and statistical manual of mental disorders: A classification of personality disorders that has had its day. Clinical Psychology & Psychotherapy 19: 372–374.

75.

Tyrer

(2013) The classification of personality disorders in ICD-11: Implications for forensic psychiatry. Criminal Behaviour and Mental Health 23: 1–5.

76.

Watson

Clark

Chmielewski

(2008) Structures of personality and their relevance to psychopathology: II. Further articulation of a comprehensive unified trait structure. Journal of Personality 76: 1545–1586.

77.

Watson

Stasik

. (2013) Integrating normal and pathological personality: Relating the DSM-5 trait-dimensional model to general traits of personality. Assessment 20: 312–326.

78.

Widiger

(2011a) A shaky future for personality disorders. Personality Disorders: Theory, Research, and Treatment 2: 54–67.

79.

Widiger

(2011b) Integrating normal and abnormal personality structure: A proposal for DSM-V. Journal of Personality Disorders 25: 338–363.

80.

Widiger

(2012) The Oxford Handbook of Personality Disorders. New York: Oxford University Press.

81.

Widiger

(2013) Changes in the conceptualization of personality disorder: The DSM-5 debacle. Clinical Social Work Journal 41: 163–167.

82.

Widiger

Costa

Gore

Crego

(2013) Five-factor model personality disorder research. In: Widiger

Costa

(eds) Personality Disorders and the Five-Factor Model of Personality. 3rd ed. Washington, DC: American Psychological Association, pp. 75–100.

83.

Widiger

Trull

(2007) Plate tectonics in the classification of personality disorder: Shifting to a dimensional model. American Psychologist 62: 71–83.

84.

Williams

(1985) The multiaxial system of DSM-III: Where did it come from and where should it go? I. Empirical studies, innovations, and recommendations. Archives of General Psychiatry 42: 181–186.

85.

Zachar

Kendler

(2010) Philosophical issues in the classification of psychopathology. In: Millon

Krueger

Simonsen

(eds) Contemporary Directions in Psychopathology. London: The Guilford Press, pp. 127–148.

Personality disorders and the DSM-5 : Scientific and extra-scientific factors in the maintenance of the status quo

Abstract

Objective:

Method:

Results:

Conclusion:

Keywords

The process leading up to the publication of the DSM-5

Personality science: PDs as a dimensional construct

Empirical support for a dimensional conceptualisation

Criticism of a dimensional conceptualisation

Extra-scientific factors at work in classification

Cognitive psychology and categorisation

Professional jurisdiction

Insights from Kuhn’s philosophy of science

An outline

Normal science in psychiatry and psychology

Paradigm change, revolution and gestalt switch

An overarching Kuhnian framework

Limitations of the approach

Towards the ICD-11

Conclusion

Footnotes

Declaration of interest

Funding

References