We have already reviewed how the DSM and ICD have different perspectives with regard to DID, and specifically its etiological origins. The DSM points towards trauma as a factor which is typically involved (posttraumatic model), whereas ICD indicates that it could be mainly iatrogenic (that is, induced by the therapist) or culturally related (both being part of the sociocognitive model). While the posttraumatic model has been dominant for over 100 years, the alternative view is supported by solid evidence and should not be dismissed out of hand. Some examples that support these alternative views include: (1) the fact that the number of alter personalities increased by the end of the twentieth century, which was in line with media depictions at the time; (2) an increase in the number of diagnoses during this same period; (3) a greater prevalence of the disorder among patients of therapists that use suggestion-prone techniques such as hypnosis; and (4) the appearance of the alters being typical during therapy [9–11]. In response to these critiques, other authors indicate that empirical studies support the existence of DID as a non-iatrogenic, “real”, and valid diagnosis, since it has content, criterion, and construct validity [5]. Moreover, experts have also defended the usefulness of trauma-focused psychotherapy aiming to integrate identity fragmentation and to decrease dissociative amnesia [12]. The fact that the diagnosis of DID is highly controversial could be related to a broader problem concerning the lack of a specific definition of dissociation.

In a broad sense, dissociation means that mental processes that are normally integrated within an individual are abnormally detached. Hence, dissociation includes the domains of conscious awareness, memory, or personality [13]. If we accept the least restrictive definition of dissociation, it would include even the perception of a stimulus under the threshold of full consciousness [13]. Prototypical examples would be the attentional neglect of over-learnt sequences such as driving, or the cocktail party effect (the ability to focus on a particular conversation neglecting the noisy background) [14]. Nevertheless, this conceptualization does not take into account whether individuals can actually change their awareness status and shift attention towards the momentary “dissociated” perception [13]. An alternative definition subordinates the diagnosis of dissociation to the unexpectedness of the disrupted integration. Therefore, “dissociation applies to mental processes, such as sensations, thoughts, emotions, volition, memories, and identities, that we would ordinarily expect to be integrated within the individual’s stream of consciousness and the historically extended self, but which are not” [13]. Typical examples of this definition of integration are post-traumatic amnesia or memory lapses in DID patients. It is important to note that any mental process—such as memory, volition, or emotion—could be dissociated. The current diagnostic criteria use this latter conceptualization [13]. In the particular case of DID, discontinuity of memory (e.g., stream of consciousness) leads to a deficit of self-integration and hence to the existence of multiple identities within the same individual, which are at least partially independent [13]. If this definition of dissociation is accepted, there is another issue to deal with: the degree of compartmentalization between the dissociated processes. In the case of DID, both the ICD and DSM-III proposed that it was almost complete, although very early historical accounts indicated that this was not the case [8]. As previously discussed, latest versions of the DSM are less strict with respect to the degree of independence of the disrupted mental processes.

However, we would like to note that neither of the two previous approaches takes into account the subjective experience of the dissociated individual. To the best of our understanding, both assume that subjective experience is mostly unchanged, in as much as patients are not aware of the incoherence of their inner processes. Moreover, this assumption was explicit in earlier versions of the DSM and still remains in the ICD, but it does not fit well with clinical data. Taking into account the subjective state of the patient, a different interpretation of the concept of dissociation arises: it is understood as a subjective sensation of disconnection between oneself and one’s environment, which is a byproduct of a lack of integration between mental processes [13]. The latest versions of DSM, together with recent scientific reports, have proposed a distinction between two types of dissociative processes or symptoms in the case of the clinical domain, namely detachment and compartmentalization [15]. Compartmentalization refers to the concept of dissociation that we presented first, i.e., the lack of conscious integration. Volitional processes are a key issue when speaking of compartmentalization, since it has been defined as an inability to control processes and take actions that non-pathological individuals would be able to do. On the other hand, detachment refers to the sensation of alienation, and thus to the subjective experience of an altered state of consciousness [15].

But this debate goes further and deeper. Recently, other authors have proposed that dissociation only refers to a lack of structural integration of the personality [6]. A critical point of these authors is the proposal that all dissociated personality elements involve a minimum level of sense of self or, in other words, a rudimentary first-person perspective. It is important to note that these authors acknowledge the fact that these “personality fragments” are not completely independent. In fact, they speak of “division” vs. “separation”, and use the “corporation metaphor”, according to which the different departments of the same corporation can share functions or aims, but are somewhat independent. With regard to this first-person perspective, they propose that the dissociated parts fulfill at the very least the minimal requisites for consciousness: “situatedness”, phenomenal now, and transparency [6]. This means that every dissociated part of the individual live in a here, in a now, and experience their own version of the world as real. These are, as we mentioned, the minimum requisites of the dissociated parts or alter for being considered as such. However, alters will usually go beyond this minimum threshold. It has been argued that this phenomenological proposal avoids major philosophical concerns by comparing discontinuities of the self with dissociations of consciousness, especially in uncertain cases of dissociation [16]. Furthermore, this proposal has been criticized for being overly narrow at a clinical level, as it would consider DID as the only dissociative disorder, leaving thus other mental problems such as depersonalization, amnesia, or derealization outside this domain [17]. As a final note, and in relation with the following debate, this proposal has been found problematic as well by those who defend dissociation as a human disposition (i.e., not necessarily pathological [18]). We will explain this debate in some detail in the following subsection.

The answer to this question has provoked a prevailing debate that started with the first descriptions of dissociation [8]. Those who advocate for dissociation as a human disposition propose that it occurs within a continuum ranging from the pathological, non-adaptive clinical syndromes, to inconsequential daily-life dissociations such as day-dreaming [15, 19]. There is statistical evidence indicating that there are some qualitative differences between pathological and non-pathological dissociation: for example, amnesia or identity disturbances are rarely present in the latter. Furthermore, patients diagnosed with a dissociative disorder are more prone to experience “normal” dissociative symptoms, such as absorption in one’s own thoughts. These differences have been termed the dissociation taxon [19, 20].

From a more philosophical perspective, Braude has proposed that dissociation is a human disposition, that is, an ability that can bring both positive and negative consequences: this is termed the “capability assumption”. This interpretation of dissociation involves not only everyday phenomena, but also extreme pathological dissociative states, since in both cases—and the whole spectrum between them—the own mental states of the person are dissociated: this is the “ownership assumption” [7]. If the thesis of dissociation as a human disposition is accepted, there is a deeper debate to be held: should dissociation be treated? If so, at what point within the spectrum should dissociation be considered pathological? These are extremely relevant questions for the clinical practice: if dissociation is considered in an excessively loose way, it could deny the existence of a maladaptive state of the person that entails an extreme psychological suffering. This denial would prevent the person receiving an adequate assistant to recover from this suffering mental state.

Interestingly, dissociation has been used as a therapeutic tool to overcome severe mental conditions. For example, hypnosis benefits from the suggestibility of the person to achieve certain therapeutic goals, although it is not successful for all patients. From this point of view, dissociation could be understood as a defensive mechanism to overcome the original conflict that caused the pathological mental state. However, it should be considered that this “transitional” (dissociative) state is maladaptive itself, since it prevents the person from properly adapting to the environment. An important fact to consider is that DID patients are highly suggestible to hypnosis [21]. Another problematic issue of considering dissociation as non-pathological is allowing the subject to freely going back and forth from his or her dissociative state. Once again, this would entail psychological suffering and a lack of integrity of the self. For those who are able to self-hypnosis, the dissociative state may become a mechanism of self-protection against traumatic memories [21]. However, this protection should not substitute the final goal of the therapist with respect to the patient, who should achieve: (1) a proper adaptation to the environment; (2) an acceptance of their condition; and (3) the integrity of the self.