Editors' Introduction

This issue's Target Article, by Sean Spence and colleagues at the University of Sheffield, is an empirical research paper that explores the nature of free association. Such research is difficult, and little work has so far been done on the topic. One reason for this is that it is only recently that high-level mental functions have been researchable using imaging technology. Another is that the concept of free association remains unclear in its meaning and function. If you ask several psychoanalysts about it, you will get opinions ranging from “There is no such thing,” to “Most analysands can't free-associate, or do so rarely,” to “My patients do free-associate, and it is the engine of analysis.” In addition, there has been an evolution of the concept as psychoanalysis has itself evolved. Freud at first viewed associations as free if they were not influenced by the analyst, by way of hypnosis or the forehead-touch technique. In the topographic model of Freud's early work, if the patient were told to speak freely, it was expected that thoughts would follow each other under the influence of the inner drives and unconscious wishes, thereby bypassing the censor and letting the unconscious emerge. Later, with the structural model, things became more complicated. Many of the defenses against free speech were themselves unconscious, so one could not simply command them to cease functioning. Whatever emerged from the patient's free association was now a compromise between conscious and unconscious aspects of id, ego, and superego. Even more recently, in the relationist model, if the term is used at all, it refers to what emerges from the particular dyad: in other words, it emerges not from an inner unconscious template as conceived in more classical models, but is unique and emerges unpredictably from this particular analysand and this particular analyst. With such a changing concept, it is not surprising that the research has been halting and sparse, as the re