Beginning with my clinical experience, I have been interested in researching speci®c psychic phenomena where primitive, protomental states of mind predominate and are often inaccessible to analysts. Some patients, even though they communicate predominantly on a neurotic level, hide a part of their personality in which certain impenetrable encapsulated nuclei prevail. By impeding access to these aspects of their minds, they make it dif®cult for real changes to take place in the course of their analyses (S. Klein, 1980; Tustin, 1986). Frequently, in these situations, my experience has been that the relationship established between the analyst and analysand is kept within prescribed `repetitive patterns' of communication that conceal perception of the unknown by the analyst and bring about a crystallization of the analytic process and the acceptance of this situation by both analyst and patient. Speci®c manifestations of this phenomenon elude analysts and are left out of their ®eld of observation. I am referring, in particular, to sensorial manifestations in which protomental phenomena prevail and where there is no discrimination between physical and mental stimuli. The impact of these stimuli on the analyst's mind is considerable and, according to the mental phenomena's degree of primitivism, they are not brought out, thus rendering them incapable of transformation. An analyst's reactions to such stimuli should be observed and included in his ®eld of endeavor as an essential part of his work. Often these reactions take the place of thought. Body language and evasion are expressions of these states.