I am suspicious of the analysis of virtues. There seems to me to be a real danger that all that is best about people will be analysed and summed up as ‘just’ (and the word ‘just’ is important here) a function of genes for this or that, or evolutionary pressures, or social structures, or anything that diminishes individuality, agency and difference. It is as if a picture restorer were to say, ‘Well, you realise that the Mona Lisa is really just paint’; which is true in a sense, but misses the point about looking at pictures generally.

So I was not minded to like this book on first sight. But Professor Paul Gilbert is a serious player in the field of the study of affective experience and regulation, both as a clinician and a researcher; and if he says that compassion is worth empirical study, in terms of understanding behaviour and developing effective treatments, then it probably is. By the end of the book, I found myself in agreement: compassion is a human experience that is of immense relevance to psychiatrists and psychological therapists, especially at a time when psychiatry is under pressure to prevent patients from being violent. My work with patients has taught me that most violence begins with the capacity for cruelty, both to the self and others; and I was fascinated to find comprehensive theoretical accounts of the concept of both compassion and cruelty in this book.

The conceptual part of the book draws on a wide variety of paradigms: evolutionary psychology, attachment theory, social learning and Buddhism. The second half of the book describes compassion as ‘caregiving mentality’, and as an essential for therapy and therapists. Again, to begin with, I was sceptical: surely compassion is something that is essential for everyone, not just for therapists. Further, I think that there is a danger of conflating compassion, empathy and sympathy, and of presenting the therapeutic encounter as being a warm and pleasurable experience. As with any form of psychological development, one hopes that overall the therapy experience will ultimately be seen as being positive and rich, but often the process will involve real pain, anger, hatred and cruelty in the therapeutic space between the participants. It is the negative aspects of the patient that the therapist must have compassion for, which in turn means compassion for his or her own anger, cruelty and pain.

Professor Gilbert and his co-authors describe this issue well, and by the end of the book I was convinced that a lack of compassion for the self is an important aspect of conditions such as chronic depression and post-traumatic stress disorder, and of destructive behaviours such as repeated self-harm. A nice chapter by Lee describes the effectiveness of identifying a cruel aspect of the self, and, interestingly, the lack of effectiveness of standard cognitive therapy for people who lack compassion for themselves. I also enjoyed a chapter by Bates about compassion in group therapy, not least because it rang so true for me and my experience of working with forensic patients in groups.

I came away from the book wanting to recommend it to colleagues, especially those working with violent patients or with patients who self-harm or are cruel to their bodies in some way. It is a compassionate book, written in a warm and accessible style, and I am more convinced now that understanding the good parts of our nature is essential for managing the less good parts. I hope Professor Gilbert and his colleagues will write more about the virtues in future.