The importance of separating observation and model is a subject often canvassed amongst SPR members. It is not unknown for a worthy phenomenon to fail to attract credible investigation because of assumptions implicit in the terms in which it is described. This is especially the case with gifted individuals, e.g. healers or mediums, whose control of their ability rests on assumed mechanisms which an investigator may find illogical or internally inconsistent.

It occurs to me, however, that the use of such 'inappropriate' models is not unusual in fields requiring individual talent, and I write to describe an analogy which members may find interesting. As a keen chorister I have some experience of the techniques by which singers control the quality of the sound which they produce. These techniques focus largely on mental imagery, most of which is anatomically invalid. We are taught, amongst other things, to imagine a column of air in our bodies, to control our breathing with our pelvic muscles, to imagine our jaw hinging at the temples and to project the sound from our foreheads. While nobody believes these models to represent reality, by using them one can trick one's body into producing the desired result.

It became fashionable a few decades ago to teach voice students anatomy and acoustics so that they could understand the science behind what they were doing. It soon became clear, however, that, far from helping, a correct understanding of the mechanisms involved was more of a hindrance to good sound production. In some cases students even developed chronic and debilitating tension in their throats. The technique rapidly lost favour, and teachers now tend to concentrate on the evocative imagery.

I find it particularly interesting that the desired phenomenon, a well controlled voice, is better produced by the application of a model which is known to be wrong, than by an understanding of the real processes involved. This places the imagery used by e.g. psychic healers in an alternative context, and argues for tolerance on the part of investigators. After all, scientists who study voice and acoustics are well able to discover the real mechanisms involved, irrespective of the mental model of the singer they investigate, provided that they are not led astray by it. It is also significant that a singer's performance can be inhibited by a more correct understanding of the science involved. One might wonder whether pointing out illogicalities in psychic practitioners' models may reduce or even damage their abilities. The analogy also suggests to me that these mental models should not be considered mere distractions. Although they may not represent the real mechanism, they are clearly important as the means of accessing that mechanism, and so deserve investigation in their own right. A certain mental model may be as indispensable for an individual's psi functioning as it is irrelevant to the actual mechanism involved.