Few colleagues will currently doubt the existence of psychotic experiences in non-clinical populations, as many cohort studies have consistently demonstrated (e.g. Tien, Reference Tien1991; Poulton et al. Reference Poulton, Caspi, Moffitt, Cannon, Murray and Harrington2000; Van Os et al. Reference Van Os, Hanssen, Bijl and Ravelli2000; Verdoux & Van Os, Reference Verdoux and Van Os2002). These findings, together with the discovery of the large genetic overlap between psychiatric disorders, has led to a call for revisiting psychiatric diagnoses in general and the Kreapelian dichotomy in particular (Boks et al. Reference Boks, Hoogendoorn, Jungerius, Bakker, Sommer, Sinke, Ophoff and Kahn2008; Carroll & Owen, Reference Carroll and Owen2009). This call is urgent with the review of the upcoming Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-V) and the need for more homogeneous schizophrenia phenotypes.

However, the continuum theory of psychosis not only predicts the existence of a non-clinical variant of psychotic experiences, it goes further, in presuming the existence of a continuous phenotype, measurable in healthy and diseased individuals. Such a continuous phenotype implies that the same psychotic symptoms that are seen in patients with a psychotic disorder can also be observed in non-clinical populations (Van Os et al. Reference Van Os, Linscott, Myin-Germys, Delespaul and Krabbendam2009).

Non-clinical psychotic phenomena are generally assessed with the aid of standard questionnaires, such as the Peters Delusion Inventory (PDI; Peters et al. Reference Peters, Joseph and Garety1999), the Launay and Slade Hallucination Scale (LSHS; Larøi et al. Reference Larøi, Marczewski and Van der Linden2004) or the Diagnostic Interview Schedule for Children (DISC-C; Costello et al. Reference Costello, Edelbrock, Kalas, Kessler and Klaric1982). When applying these scales, cohort studies observe a majority of the population scoring on none of the items, a relatively large minority scoring on only one item, a smaller group scoring on several items and finally a very small group scoring on many items. For example, in the Dunedin cohort, 85% of the children did not score on any item of the DISC-C, 13% scored on one item, 1% scored on two items and 0.3% scored on three or more items (Poulton et al. Reference Poulton, Caspi, Moffitt, Cannon, Murray and Harrington2000). When the total scores of such questionnaires are evaluated, the impression is that of a perfect continuum of psychotic experiences. However, it remains illusive if the 13% scoring affirmative on one item of this scale is phenotypically similar to the 0.35 scoring on three or more items.

These questionnaires do not explore the phenomenology of psychotic experiences and thus provide only little information about the exact phenotype. For example, an item assessing auditory verbal hallucinations of the DISC-C asks ‘Have you heard voices other people can't hear?’ In a clinical situation, an affirmative response to this item would lead to further questions regarding the perceptual quality of this experience, frequency, associated distress and verbal content of this voice. For cohort studies, assessing hundreds of individuals, this is often not feasible. As a consequence, it is currently not known if psychotic experiences in non-clinical individuals have the same quality as those in psychotic patients. This information can only be obtained by assessing detailed phenomenology of individual cases in a clinical psychiatric interview. At the University Medical Centre Utrecht, we made an effort to do so by studying non-clinical individuals who experienced auditory verbal hallucinations at least once a month during at least 1 year. These individuals were assessed by a psychiatrist, using the Comprehensive Assessment of Symptoms and History (CASH; Andreasen et al. Reference Andreasen, Flaum and Arndt1992) and the Structured Clinical Interview for Personality Disorder (SCID-II; First et al. Reference First, Spitzer and Gibbon1995) interview (for details, see Sommer et al. Reference Sommer, Daalman, Rietkerk, Diederen, Bakker, Wijkstra and Boks2008). Auditory hallucinations of 111 individuals who did not meet diagnostic criteria for axis I or II disorders were described in detail, and compared with hallucinations in 118 psychotic patients. Based on the continuum theory of psychosis, we expected to find auditory verbal hallucinations similar to those in psychotic patients in the non-clinical group, but of lower severity, for example less real, or less loud than in psychosis.

However, while loudness and reality of the voices were rather similar in both groups, we found large qualitative differences, with hardly any overlap in the verbal and emotional content of these hallucinations. Almost all non-clinical individuals had a neutral or pleasant emotional content of hallucinations. The verbal content was frequently an advice about daily hassles or encouragements. Examples of these voices experienced by non-clinical individuals were: ‘don't be late, or you'll miss your train’, ‘try to practise what you've learned’ and ‘don't be intimidated by big words’. Although these phenomena definitely had perceptual qualities, distinguishing them from thoughts, the typical content was more alike to normal inner speech, i.e. one could imagine a person saying these phrases to himself. In contrast, the auditory hallucinations in the psychotic patients had a more classical, stereotyped content, either consisting of insults (i.e. ‘you lazy fat bastard’), commands or running commentary. The overlap between the groups was remarkably low: seven of 111 non-clinical individuals experienced voices with a negative content, while only two of the 118 patients had voices with a predominantly positive content (Daalman et al. in press). Another contrast between the non-clinical voice hearers and the patients with hallucinations was the age of onset. Mean age of onset in the healthy subjects was 12 years, while patients had a mean age of onset of 21 years. Such a large difference in onset of symptoms could imply a difference in aetiology, as the pre-pubertal brain in which the non-clinical hallucinations first arise differs in many aspects from the adolescent brain which gives rise to psychotic hallucinations.

So, while standard questionnaires suggest a continuous phenotype of hallucinations throughout the population, detailed study of the phenomenology of these ‘non-clinical psychotic phenomena’ suggests that they are qualitatively dissimilar from the symptoms encountered in psychosis. While the first requirement for the continuum theory of psychosis – the presence of psychotic experiences in the non-clinical population – is beyond reasonable doubt, the second prerequisite of a similar phenotype in clinical and non-clinical individuals needs further study. David's critique (David, Reference David2010) is important and timely; before we can embrace the continuum theory of psychosis and start to think about revisiting our diagnostic classification, further detailed study of the clinical and especially of the non-clinical phenotype is warranted.