Two letters in response to my brief editorial seeking to argue why borderline personality disorder (BPD) is not a mood disorder. One (by Dr Stephen Marwaha) gently offers a ‘supplementary interpretation’ that seemingly positions BPD in an ineffable zone. The other (by Professor Peter Tyrer) is also gracious, but accuses me of ‘tilting at a windmill’. What is it about BPD that promotes intense professional interchanges and risks splitting?

Let me proceed less impulsively. Peter states that he did not claim any connection between ‘bipolar, unipolar and borderline’, but I did not ascribe any such claim to him. Further, he suggests that, in my arguing against BPD as a mood disorder, I was ‘tilting against a windmill’ in whose construction he had ‘absolutely no part’. Here I risk being more coyote-like than Don Quixote-like in recording that Peter did lead me up the path to the windmill. As then Editor of the British Journal of Psychiatry, he invited me (personal communication, 16 July 2013) to write an editorial on the ‘status and relationship of those disorders’, and there observed that he ‘wrote a paper in 2009 saying that BPD was a mood disorder so I have already nailed my placard on the diagnostic wall’.

It would be fairer to Peter, however, to report his actual referenced 2009 paper.^{Reference Tyrer1} Its abstract stated that BPD ‘is better placed with the mood disorders than in odd isolation as a personality disorder’. In the body of the paper, he judged that BPD does not fit within the general descriptors of a personality disorder and, later, in considering whether it were to be ‘accepted as a mood disorder’, he suggested that it might warrant a ‘term such as fluxithymia, or “rapidly changing” mood disorder’. In relation to other points made in his letter, I agree with all but one of his interpretations. Peter states that BPD characteristics are linked closely to affective instability, not normal personality variation. But is not an affective stability/instability dimension a personality dimension? Even if such instability in those with BPD remits with time (as the data show), this does not of necessity eject BPD from a personality home base. If attenuation with age were to discount a personality base, where would we place sociopathy nosologically when it also commonly ‘burns out’?

Marwaha effectively argues that ‘BPD is in large part a mood disorder but is not necessarily a bipolar or unipolar mood variant.’ While there is one public figure who has described his mood disorder as a ‘tripolar one’, and there are usually three options in life, our current classificatory systems in psychiatry allow only for unipolar and bipolar disorders. If BPD is a mood disorder, it has to be positioned as one or other: there is no mood condition beyond their borders.

Marwaha proceeds to argue that ‘affect’ nuances allow BPD to be positioned as a mood disorder. Let us consider paranoid personality disorder (PPD) as an analogy for his points. Do not individuals with a PPD temperament experience changes in affect that are ‘intense, frequent, rapid to occur, slow to dissipate’ and with ‘valence’ components weighting anger and emotional dysregulation? If so, as I would argue, would we not, logically, similarly risk positioning PPD (and several other personality disorders) akin to BPD if we accept Marwaha’s arguments? I doubt whether we would feel comfortable with such a suggestion. In essence, if we concentrate on features shared across conditions (especially emotional dysregulation) we will predictably find convergence and risk diagnostic confusion. My editorial adopted the opposite approach: focusing on points of distinction between BPD and mood disorders, and which seek to advance diagnostic differentiation, especially in a clinical context.