While the discussion concerning advantages and disadvantages of the single-symptom approach compared to other approaches, such as the syndrome/diagnosis approach, is far from resolvedFootnote *, I agree with Mojtabai & Reider's (Reference Mojtabai and Rieder1998) conclusion that each approach has different aims and answers different questions and therefore, in my opinion, can be viewed as complementary. In the spirit of methodological pluralism – both approaches deserve to be, as it were, in each researcher's ‘toolbox’.

In particular, hallucinations have frequently been examined within a single-symptom approach (for a review see Aleman & Larøi, Reference Aleman and Larøi2008)Footnote †. There has, furthermore, been a noteworthy recent interest in a ‘phenomenological’ approachFootnote ‡, both within psychopathology research in general (cf. Owen & Harland, Reference Owen and Harland2007) and for hallucinations in particular (cf. Larøi, Reference Larøi2006; Larøi & Woodward, Reference Larøi and Woodward2007). That is, for example, researchers are using the phenomenological characteristics of hallucinations to either test the validity of previous findings and/or models, and/or have employed a refined phenomenological analysis of the experience or symptom in question in order to guide future research. The study by Hoffman and colleagues (Reference Hoffman, Varanko, Gilmore and Mishara2008) is an excellent example of the merits of adopting both a single-symptom and a phenomenological approach.

Hoffman et al. (Reference Hoffman, Varanko, Gilmore and Mishara2008) ask the question: ‘Are schizophrenia patients with auditory-verbal hallucinations (AVHs) able to differentiate between own thoughts and voices and if so, how do they do so?’ The results showed that 80% of patients were able to differentiate AVHs from their usual verbal thoughts either ‘most of the time’ or ‘always’. A similar finding was reported in a recent study (Moritz & Larøi, Reference Moritz and Larøi2008) where a large majority of schizophrenia patients expressed being able to distinguish real from hallucinated voices. Although the results from Hoffman et al. (Reference Hoffman, Varanko, Gilmore and Mishara2008) are not entirely clear-cut and seem to suggest considerable individual variability, the results additionally revealed that three characteristics were particularly important in differentiating voices from thoughts: verbal content, degree of control and the non-self ‘sound’ of the speaking voice.

However, after having read the article, two issues popped into mind. Do the findings have clinical implications? It would have been interesting to match the theoretical reflections with reflections regarding the extent to which the findings might have a clinical impact. Briefly, how do we get patients to gradually come to terms with the fact that their voices are, in reality, internal mental events (i.e. thoughts)? Perhaps it would be fruitful to mention to voice-hearers that thoughts may take a verbal form, are also involuntary or unintended, and may also lead individuals to feel as if they do not belong to oneself. Another issue is that future studies will inevitably need to better elucidate the developmental or dynamic nature of the genesis of hallucinations. To begin with, hallucinations are, ultimately, internal mental events. These internal mental events evolve over time through a complex interaction of cognitive, emotional, biological and socio-cultural mechanisms and factors. They are, as it were, transformed into something else – for some individuals, this transformation results in a voice. Regrettably, we continue to predominantly (although the climate has changed in the past years) study hallucinations when they are fully ‘developed’ (e.g. hallucinations present in chronic schizophrenia patients). It will be an important future approach to consider when and how internal mental events, such as ordinary everyday thoughts, are transformed into voices. This calls for a developmental approach to hallucinations – yet another approach that researchers will need to include in their toolbox.