Within the context of psychiatry and neurology, the term dissociation is extremely difficult to define satisfactorily. Since its introduction in the late nineteenth century, the term has been applied to a wide range of neurologic, psychiatric, and psychological phenomena. As a result, there is considerable confusion over what actually constitutes dissociation, and the concept is frequently misapplied. This is particularly true within the field of epilepsy. Many epileptic phenomena have been labeled as dissociative, including the sensory, affective, and cognitive features of partial seizures, behavioral automatisms, postictal amnesia, and fugue.²⁷ Similarly, certain psychiatric phenomena that mimic epileptic events, so-called “nonepileptic seizures,” have been identified as primarily dissociative in nature. Indeed, many authorities have argued that the dissociative nature of nonepileptic seizures could provide the basis for their conceptual and practical differentiation from “genuine” epileptic events.⁴⁴ If, however, dissociation is experienced both by individuals with epilepsy and by those with pseudo-epileptic seizures, how can its occurrence aid in the differential diagnosis of these conditions?

The term dissociation originates in the work of Pierre Janet, who proposed one of the earliest systematic accounts of the psychological mechanisms underlying hysteria. According to Janet,^36,38 a fundamental weakness in the hysterical individual’s mental character makes the person susceptible to a breakdown in the normally integrated functions of consciousness when faced by environmental stress or trauma. As a result, organized sets of knowledge pertaining to the trauma may become “dissociated” from the main body of consciousness and may serve to take control of behavior and experience if activated by environmental events. The automatic activation of these dissociated memories results in a hysterical reaction (or “somnambulism”) that, in some instances, takes the form of a nonepileptic attack. Following Charcot’s demonstrations at the Salpêtrière in which hysterical symptoms were shown to be both induced and removed by hypnosis, Janet’s dissociation theory assumed that this process of dissociation was driven by an autohypnotic state. According to this view, two aspects of the hysterical individual’s psychophysiologic makeup are responsible for the processes of dissociation and somnambulism. First, the hysterical individual possesses an abnormally high degree of suggestibility that allows ideas from the external environment to develop within him or her in the absence of his or her effort or awareness. Second, the hysterical individual suffers from an attentional dysfunction or “retraction of the field of consciousness” (Janet,³⁸ p. 314), which prevents them from entertaining alternative states of mind, thereby accentuating their responsivity to external suggestion. The resulting process of dissociation leads to “dedoublément,” or double consciousness, whereby two or more discrete but conscious modes of being existed alongside one other, separated by amnesia. In extreme cases, the autonomy of this dissociative consciousness gave rise to one or more alter personalities.⁵⁴

Although it was originally assumed that these processes were triggered by external traumas, subsequent psychoanalytic theory gravitated from external happenings to an inner efficient causation based on traumatic conflict arising from the patient’s psychically unacceptable desires and fantasies. Thus, the etiology was expanded to include subjective traumatization or “vehement emotions” whereby “every memory, every thought, competent to arouse strong and lasting emotions, can play the part of a fixed idea, and may originate hysterical symptoms.”³⁹ In these terms, symptomatology became a composite of fact and fantasy, metaphor and symbolism. This was the case even when there was an association with objective traumatic events because, in classical Freudian theory, before the onset of symptoms, there is a reconstruction in memory overlaid with fantasy (Freud,²³ p. 625). Freud also proposed that falsifications are introduced into memory in order “to interrupt disagreeable and causal connections” (Freud,²⁴ p. 446, footnotes).

These discoveries led Freud to view hysterical etiology as fantastic, so that, as with dreams, the symptoms became the “the royal road” to the unconscious:

This move from fact to fantasy is evident in Freud’s revision of his early seduction theory,³⁵ which suggested that some hysterical patients suffered from unconscious memories of childhood seduction. Later he asserted that such memories were infantile wish fulfillments, although not always: “So often they are not fantasies but real memories…. A fantasy of being seduced when no seduction has occurred is usually employed by a child to screen the autoerotic period of his sexual activity” (Freud,²² p. 417).

Although popular for nearly a century, recent interest in the prevalence of childhood abuse has led to a backlash against Freudian theory, coincidental with a renewed interest in Janet’s work on dissociation,¹⁹ adaptations of his theory in cognitive psychology,^30,31 and a mushrooming of publications linking objective traumatic experiences to dissociative psychopathology. This is particularly relevant for the evaluation of nonepileptic seizures. Indeed, both the Diagnostic and Statistical Manual for Mental Disorders, 4th edition (DSM-IV),² and the International Classification of Diseases (ICD-10)⁶⁵ make an explicit link between traumatic events and the onset of dissociative symptoms. Moreover, a number of studies have found disproportionately high rates of physical, sexual, and emotional abuse in patients with dissociative disorders.^16,34,53 Bowman,⁷ for example, found that 70% and 77% of her sample of 27 nonepileptic seizure patients had experienced physical or sexual abuse, respectively. Similarly, Betts and Boden⁶ obtained positive sexual abuse histories from 54% of 96 patients with nonepileptic seizures.

Thus, 100 years of thought about hysteria has brought the traumagenic theory full circle, and yet the difficulties encountered by nineteenth century investigators remain. Trauma is ill conceived and now tends to be exclusively considered only as an objectively measurable event, ranging from natural disasters to childhood abuse. As a result, there is a lack of consideration given to inner efficient causes, such as an individual’s perceptions of events, which would seem important in the process of traumatization. Not only could this explain the obvious lack of psychopathology in some victims of abuse,⁶⁶ but it might also account for the presence of psychopathology following relatively harmless events. For example, in LaBarbera and Dozier’s study of pseudoseizures,⁴⁵ three of the four girls reported no history of sexual abuse but had experienced minor sexual events perceived as traumatic. Equally, in cases in which there have been false allegations of serious trauma,¹⁴ there is often an incidence of minor sexual traumatization. Finally, the determinism of trauma in the development of nonepileptic attacks and psychopathologically related disorders is problematic in terms of the unquestioned and overgeneralized evocation of dissociation as a defense mechanism as well as the lack of explanations for its prolonged use, even years after the event.⁶³

The Dissociative disorders category in the latest edition of the Diagnostic and Statistical Manual² encompasses dissociative amnesia, dissociative fugue, dissociative identity disorder (formally multiple personality disorder), depersonalization disorder and dissociative disorder not otherwise specified (see Table 1). According to DSM-IV,² “the essential feature of the dissociative disorders is a disruption in the usually integrated functions of consciousness, memory, identity, or perception of the environment” (p. 477). A slightly broader definition is offered in the latest edition of the International Classification of Diseases,⁶⁵ which identifies the loss of control over bodily movements as an additional dissociative phenomenon. As such, the ICD-10 Dissociative (conversion) disorders category encompasses dissociative amnesia, dissociative fugue, dissociative motor disorders, dissociative convulsions, dissociative anesthesia and sensory loss, dissociative stupor, mixed dissociative (conversion) disorders, other dissociative (conversion) disorders, and dissociative (conversion) disorders, unspecified (Table 1). The Somatoform and Dissociative disorders are related in that these phenomena are characterized by symptoms that, on the face of it, resemble those that occur in certain physical conditions but are presumed to be psychological in origin. Dissociative and somatoform phenomena differ in that the symptoms of the former resemble those of neurologic illness, whereas the symptoms of the latter are more akin to those encountered in internal medicine. As Kihlstrom⁴⁰ cogently argued, all of the phenomena identified as dissociative disorders within DSM-IV and ICD-10 are linked by the fact that each has a temporary disruption in consciousness or volition as its primary defining feature.

The differences between DSM-IV and ICD-10 in their classification of the Dissociative and Somatoform disorders are readily apparent. First, unlike ICD-10, DSM-IV places nonepileptic attacks in the Somatoform rather than the Dissociative disorders category along with other so-called “conversion” phenomena, such as unexplained motor and sensory symptoms, that are identified as dissociative in ICD-10. This difference is more practical than conceptual, with DSM-IV placing greater emphasis on the importance of excluding physical illness in the differential diagnosis of these phenomena.² Second, unlike DSM-IV, ICD-10 does not identify depersonalization as a dissociative phenomenon, due to the lack of any significant loss of control over sensation, memory, or movement in this condition and its limited affect on personal identity. Third, DSM-IV identifies a distinct category for multiple personality disorder, relabeled dissociative identity disorder in the latest edition of this scheme. In contrast, ICD-10 places multiple personality disorder in the other dissociative (conversion) disorders category, reflecting controversy over whether this syndrome is iatrogenic or culturally bound to North America. Inconsistencies aside, both DSM-IV and ICD-10 explicitly state that physical conditions such as epilepsy should be excluded in the differential diagnosis of the Dissociative and Somatoform disorders.

To some extent, the differences between DSM-IV and ICD-10 demonstrate ongoing controversy about the definition of the term “dissociation.” When the term was originally introduced in the nineteenth century, it was used to refer to a specific mental mechanism thought to be associated with a relatively limited set of psychological symptoms. Over the years, however, the number of phenomena thought to be attributable to dissociation has expanded considerably, and the dissociation label is now applied to an extraordinary range of psychological symptoms, states, and processes. Cardeña¹⁵ described a useful taxonomy that captures the different ways in which the concept of dissociation has been used. According to this scheme, there are three major facets of the dissociation construct: (a) dissociation as nonconscious or nonintegrated mental modules or systems; (b) dissociation as an alteration in consciousness; and (c) dissociation as a defense mechanism.