The Conceptual Unity of Dissociation: A Philosophical Argument

Abstract

Psychologists and psychiatrists have studied dissociative phenomena for a long time. However, they demonstrate surprisingly little agreement about what dissociation is and about which things exemplify it. Of course, many agree that certain l orid phenomena count as dissociative—for example, fugue states and DID. But when mental health professionals tackle the topic of dissociation theoretically and attempt to dei ne it, they do so in ways that often conl ict with one another, and (perhaps most surprising of all) they tend to overlook a large and important class of phenomena. Historically—and contrary to what the recent clinical literature would lead one to believe— most (if not all) hypnotic phenomena have been regarded as dissociative (see, e.g., Gauld, 1992; Van der Hart and Dorahy, 2008). In the late 19th and early 20th centuries, researchers into hypnosis were trying to study systematically the same sorts of subconscious mental divisions they believed occurred spontaneously in hysteria and to some extent in somnambulism. Indeed, some considered hypnotically induced systematized anesthesia or negative hallucination to be paradigm instances of dissociation. Yet when clinicians now try to analyze dissociation, hypnotic phenomena are largely ignored. Despite evidence to the contrary (e.g., Crabtree, 1993; Braude, 1995; Van der Hart & Dorahy, 2008), historians of psychology usually credit Pierre Janet with having originated the concept of dissociation, although he regularly used the term désagrégation instead. Janet focused on a distinctive and relatively limited type of traumainduced psychopathology. He considered dissociation to be a kind of weakness, a failure (in the face of disturbing events) to integrate parts of consciousness and maintain conscious unity. However, the concept has evolved in the hundred years since Janet tackled the subject. Subsequent researchers (e.g., James, Binet, Myers, Liègeois, Sidis) also recognized an apparent causal link between trauma and dissociative pathology. But they tended to agree that the processes Janet was describing from cases of hysteria (which 28 Dissociation and the Dissociative Disorders included conversion disorder and double consciousness) were also at work in a wider variety of phenomena, drawn not just from psychopathology but also from experimental psychology and even everyday life (see, e.g., Binet, 1896; Myers, 1903; Sidis, 1902). And along with that, they tended to view dissociation not as a weakness, but as a kind of capacity (not necessarily maladaptive) to sever familiar links with one’s own mental states. Signii cantly, this evolution of the concept of dissociation happened quite rapidly. Other turn-of-the-century researchers, interested at least as much in hypnosis as in psychopathology, were eager to explore the ways in which hypnotic states seemed to produce a kind of division or doubling of consciousness, or creation of seemingly autonomous sets of mental processes (for a quick history of these developments, see Braude, 1995, and Van der Hart & Dorahy, 2008. For a more detailed account, see Gauld, 1992). As Messerschmidt (1927) eventually made clear, these apparent divisions weren’t as fully autonomous as they seemed. But that didn’t undermine the view that the phenomena in question could arise either experimentally or spontaneously or, for that matter, pathologically or nonpathologically. These nonpathological (including hypnotic) contexts, in which the concept of dissociation has historically played an important role, tend to be neglected by most clinicians. Given their pressing clinical concerns, perhaps that is not surprising. Nevertheless, keeping in mind what pathological and nonpathological dissociative phenomena have in common may bring clarity to other issues, such as the difference (if any) between dissociation and apparently similar or related concepts—in particular, repression. In a fairly recent development, some clinicians have examined the concept of dissociation by using diagnostic surveys (e.g., the Dissociative Experiences Scale (DES) and the Multiscale Dissociation Inventory (MDI)) to consider how dissociative symptoms cluster. These survey instruments were initially designed as screening devices, to assess the presence or absence of phenomena already believed by the test designers to be dissociative. However, subsequent research on thousands of survey results has a more ambitious goal—namely, to determine more precisely what dissociation is. But data of the sort elicited by these surveys can’t tell us what the concept of dissociation is. To reiterate, the surveys look only for symptoms antecedently judged as relevant by their designers, who are limited by their selective grasp of the history of the concept. What they most clearly tend to neglect are the many nonpathological hypnotic phenomena that have been considered dissociative, but simply fall outside the purview of the surveys. In some cases, the studies in question are even more problematical than these remarks might suggest. For example, Briere et al. (2005) apply the MDI to determine whether dissociation is a multidimensional construct, and they conclude that it is, and that “the notion of ‘dissociation’ as a general trait was not supported” (p. 221). Apparently, then, the authors see themselves as trying to settle the issue of what sort of thing dissociation is. Indeed, on the basis of their survey they claim that “the term dissociation may be a misnomer to the extent that it implies a single underlying phenomenon” (p. 230). We’ll consider shortly whether dissociation can in fact be regarded as a single underlying phenomenon. But for now, I want only to observe that Briere et al. can’t possibly have shown that it isn’t (quite apart from concerns about using survey instruments for conceptual analysis). Briere et al. purport to uncover what dissociation is on the basis of a survey that tracks relationships among a handful of factors—of course, factors they antecedently determined to be relevant. Moreover, one of those factors is identity dissociation and, obviously, one can’t analyze the concept of dissociation by appealing to that very concept. So if Briere et al. are (as it seems) trying to analyze the concept of dissociation, their attempt is blatantly circular. So I believe we need to do some conceptual and methodological housecleaning. I agree with Cardeña (1994; Prince, 1905) that when clinicians attempt to characterize dissociation, they tend either to exclude too much or include too much. However (and apparently unlike Cardeña), I think it may be possible to pull together many of the varied intuitions about and approaches to dissociation and come up with a single, general, and useful characterization of dissociation that covers both its pathological and nonpathological forms, including many of those once deemed important but largely ignored today. I shall attempt to dei ne a single inclusive concept of dissociation that rests only on reasonable and recurrent assumptions distilled from more than a century’s literature on the subject. I start by identifying specii c assumptions underlying typical uses of the term dissociation, then see if they can be stated plausibly, and then see whether we can extract from them a dei nition that has both generality and utility.1