‘Reports that say that something hasn't happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns – the ones we don't know we don't know’.

I heard this quote (http://news.bbc.co.uk/1/hi/programmes/bh/1773951.stm) from Donald Rumsfeld in February 2002 whilst he was the US defence secretary. I enjoyed the sound of it, but struggled a little with applying to research. However, I've resolved this through conceptualizing it as a benign factorial thought experiment; the first factor is the things we know (or don't know) – essentially these are data or research facts – which can be known or unknown, and the second factor is the synthesis of these facts into questions we raise or become aware of – things we don't know – the unknowns. There may be some value in taking this approach with reviewing a book on an area of clinical science which is at the core of psychiatry, and the subject of this book – Hallucinations. Thus, one can reflect how well the book covers the data – the first set of knowns or unknowns – reviewing the research and the important areas that require further research; and then evaluate how well it synthesizes this data to address the second set of questions; the opportunity for the authors to raise additional questions, but they can only refer to the knowns; only the reader can develop the unknowns; questions which the authors have not questioned or raised.

The authors cover the first factor very well, there is an excellent review of the literature, enhanced by a comprehensive reference list; and relatively unusually, separate author and subject lists in addition to the references. The book is well laid out, with each chapter terminating in a summary and the key points from the research. The chapters which I particularly enjoyed, included the introductory coverage of the history of hallucinations and the early definitions; from Socrates and Galileo through Freud and the more contemporary Esquirol's definition ‘thorough conviction of the perception of a sensation, when no external object, suited to excite the sensation has impressed the senses’. The phenomenology of hallucinations is also very well described, it was fascinating to consider how cultural factors have influenced the content of hallucinations; in the middle ages hallucinations were almost exclusively religious in content, while in more recent times, the content has more frequent technological and persecutory themes. The attempts to draw out the structure of hallucinations made interesting reading; the classification-based approach, separating hallucinations into three distinct broad dimensions; pertaining to the linguistic complexity of the hallucinated material, whether they were attribute to oneself or others and whether they were experienced in internal mental space or as emanating from external space. In the groups of hallucinators, I was fascinated by the experiences of hallucinations in deaf patients, who described the same characteristic features of hallucinations as non-deaf patients, with hallucinations in the second and third person and inside and outside the head. While hallucinations are accepted as normal in certain circumstances, around sleeping and waking for example, they also occur commonly after bereavement. However, the social acceptability of hallucinations has a significant effect on the reported frequency of hallucinations; intriguingly, in Japan where hallucinations related to bereavement are not considered stigmatizing, up to 90% of recently bereaved widows reported these experiences. In the UK, most bereaved people were reluctant to talk about these experiences, for fear of negative reactions. In a similar vein, there is a comprehensive review of the continuum hypothesis of hallucinations, with increased interest in the view that hallucinations are not pathological per se, but exist in a continuum between healthy and pathological samples, differing only quantitatively. Perhaps, this view suggests, the distinction between pathology and normality may lie in how the subject responds to the hallucinatory experience.

Now, on to the second factor. What questions do the authors synthesize on the basis of the data? This is contained in the chapter entitled ‘Towards a comprehensive model’, where they propose a model to explain hallucinations which operates at three levels. The first is through aberrant hyperexcitation of attentional processes. They suggest that the thalamic reticular system and dopaminergic and cholinergic perturbations may be involved. The second is via disinhibition of perceptual areas, or perceptual release and both of these are commensurate with pharmacological or neurological hallucinations. The third route is a higher order cognitive route which manipulates imagery processes; this is exacerbated by affective states, particularly hyper-arousal secondary to stress, and contributed to by self-monitoring errors and an external attribution bias. They suggest different cortical regions which would be involved in these different pathways, lateral prefrontal cortex for the third higher order route and anterior cingulate gyrus and premotor cortex for the monitoring processes, and the thalamus for attention. This is clearly a complex model and I was less convinced by this synopsis, and a little disappointed; perhaps it was naive to expect an answer, but the proposed model appeared to lack not only parsimony, but I found it difficult to link the pathways together; I would have liked more space devoted to the idea of pulling these strands together. However, perhaps as in so many areas of psychiatric illness – this too is a case of a complex multifactorial aetiology.

The book ends with a chapter on treatment modalities and again had impressive coverage of the use of transcranial magnetic stimulation to treat auditory hallucinations, and the potential for linking this more systematically with imaging studies in the future. One very minor gripe was the list format for describing some of these studies, e.g. X described this, and in the next paragraph, Y described that, and Z described that, suggestive of less rigorous editing than the in the rest of the book.

So in summary, an excellent book which covers the known and unknowns of the basic data in delightful detail, and the synthesis of the known knowns and unknowns reasonably well; as for the unknown unknowns, well that's for you to read and decide. I will certainly be using this copy on a regular basis and would recommend it to anyone interested in psychosis; it has something for almost every level of reader from interested members of the public to cognitive neuroscientists interested in phenomenology.