Gentlemen!

Certain basic features of hallucinations emerge directly from our hypothesis. These include, first, the ‘incorrigibility’ [Ed] of hallucinations: The reality of a sensory deception is maintained against the testimony of all other senses, and most fantastic attempts are made to explain it, leaving no room for doubt, or the possibility of their sense being deceptive. As you already know, comparison with the other senses is the only possible means of correction; but once attention has been captured by morbid activation, which rides the ‘crest of the wave’ [Ed] of psychophysical motion, then constraints imposed on consciousness make instantaneous correction impossible. Formation of associations with any normal ideas, which such an image excites, thereby raising to prominence contradictory images, is made difficult or quite impossible by sejunction.

Exactly the same arguments explain the familiar incontestable nature of those hallucinations that either command or prohibit behaviour. Again, countervailing signals become inaccessible by the very fact of sejunction, so that the nerve current, confined to a narrow pre-formed channel, discharges with its elemental force upon motor projection fields. However I have to add that the compelling nature of such hallucinations is usually overestimated, and you often find patients who can resist unreasonable demands, and who even complain about them. Thus, for example, a patient may have a hallucinatory urge to hit his doctor, which may not reach fruition, when faced with that same patient’s attachment to the doctor; but even this fact may be explained, by the different degrees of the sejunction process, as is readily apparent.

A second striking fact is the predominance of auditory verbal hallucinations, usually identified by patients themselves, as a ‘voice’ [W], which, by virtue of their special clinical status, deserve a special name—a ‘phoneme’ [W]. Occurrence of other auditory hallucinations is no more frequent than in other sensory domains. However, the fact that hallucinations of many patients are exclusively speech sounds, and that, for all hallucinations—with few exceptions—hallucinated speech sounds predominate, must be taken as one of the most fundamental characteristics of sensory deceptions, to be traced back in the final analysis entirely to the manner of their formation [1]. Let us remember that we developed our concept of secondary identification using the specific example of the sensory projection field of language. The sound of a word is not sufficient for us to understand it, without, first of all, memory images, which make up the corresponding concept being activated, so that secondary identification can take place, and the sense of the spoken word can be grasped. Although we can generalize this, we cannot fail to recognize that it is just the sensory speech centre which has such close links with the simplest patterns of association: These are the terms for solid objects, if you do go as far as to equate a sound image with its concept. For a mechanical conception of processes taking place during a hallucination, you must now realize that in no other sensory area does a more intimate connection of a concrete concept exist than in the sensory speech centre. Experiments show that the name of a specific object—which, we can assume has five different sensory qualities—can immediately be found when just one of them is activated, excluding all the others. Let us assume that similarly deep-rooted associations exist between s, the auditory speech centre, and each of the five sensory projection fields: Then we will realize that during central excitation of the concept resulting from the ‘damming up’ [Ed] of nervous energy, excitation spreads to the sensory speech field, multiplied fivefold, and in this way will achieve the abnormal level of activation needed to generate a hallucination. Equally favourable conditions are not to be found in any other sensory projection field.

When introducing psychophysiology I commented that you should assume that there are individual variations in this relationship: Sometimes a more conceptual side prevails, sometimes more ‘thinking in words’ [Ed], that is, in speech sounds. Before I made that suggestion, I had felt it necessary to warn you that in some quarters it is claimed that thinking takes place only through speech sounds. However, if we accept that individual variation in thought mechanisms does exist, then we have found a key to understand an important clinical fact, which is probably quite universal: For a single form of illness (such as an acute ‘anxiety psychosis’ [Ed]), which entails an essential mental content, one individual might portray the content itself, while another produces phonemes representing the same content. In this example, in one case only ideas full of anxiety or ‘anxiety ideas’ [W] as I call them, in the other, frightening and threatening phonemes, that is the same ideas, but put into words. We can assess such experiences not only among acutely ill psychotic patients, but in exactly the same way amongst those psychoses arising slowly, and progressing chronically. To some extent, pent-up nervous energy amongst individuals predisposed to hearing ‘voices’ [W] (and who, by the way, are in the overwhelming majority) [Ed] can find greater excitability of the sensory speech field and its converging incident pathways.