More Options

Multiple Personality Disorder: Witchcraft Survives in the Twentieth Century


August Piper Jr.

Skeptical Inquirer Volume 22.3, May / June 1998

Since 1980, some psychotherapists have claimed that thousands of Americans are afflicted with multiple personality disorder. Believing such claims requires ignoring their many serious deficiencies.

Any people, given over to the power of contagious passion, may be swept by desolation, and plunged into ruin.

— Charles W. Upham, 1867

An epidemic of psychiatric illness is sweeping through North America. Before 1980, a total of no more than about two hundred cases had ever been found in the entire world, throughout the entire recorded history of psychiatry. Yet today, some proponents of the condition claim that it afflicts at least a tenth of all Americans, and perhaps 30 percent of poor people — more than twenty-six million individuals. An industry involving significant sums of money, many specialty hospitals, and numerous self-described experts, has rapidly grown up around the disorder.

The illness is multiple personality disorder (MPD), a condition that has always attracted a few wisps of controversy. Lately, these wisps have coalesced into clouds that, in drenching rainbursts, pour criticism on the disorder. An examination of the flawed reasoning, unsound claims, and logical inconsistencies of the MPD literature shows that well-founded concerns drive this storm of criticism.

What Is MPD?

MPD is classified as a dissociative disorder. The term dissociation refers to disruption in one or more mental operations that constitute the central idea of “consciousness": forming and holding memories, assimilating sensory impressions and making sense of them, and maintaining a sense of one’s own identity (American Psychiatric Association 1994, 477). The essence of dissociation is that material not in awareness influences behavior, mood, and thought (Spiegel and Schleflin 1994). Thus, the behavioral disturbances prominently manifested in dissociative disorders are considered to be unconscious: that is, resulting from forces beyond the patient’s awareness, beyond voluntary control.

The king of dissociative disorders is MPD,1 also called dissociative identity disorder. Afflicted people episodically fail to recall vital data about themselves, but what distinguishes MPD from all other psychiatric conditions is the putative cause for these memory failures. The condition’s proponents claim the memory failures occur because patients are periodically taken over by one or more “alter personalities” (variously referred to as “identities,” “ego states,” “alters,” or “personality states”). These guest personalities, submerged since being formed during childhood — more on this later — rise to the surface and impose their own memories, thoughts, and behaviors on patients.

The essential feature of MPD, it is said, is that an individual’s behavior is controlled by two or more alters (Putnam et al. 1990); the separate identities are assumed involuntarily (Sarbin 1995; Watkins and Watkins 1984). One personality may feel “carried along in a panicked helpless state” as another endangers it or engages in behavior repugnant to it (Kluft 1983, 75). Patients are said to experience a sense of being made to misbehave or hurt themselves (Putnam 1991). Some theorists even claim the existence of “omnipotent alters,” which can simply compel patients to do their bidding (Lewis and Bard 1991). As an example, C. A. Ross writes of alters that “force [the patient] to jump in front of a truck. [The alters] then go back inside just before impact, leaving the [patient] to experience the pain” (Ross 1989, 115).

The image of all this is of an invading army usurping a government, an operator taking control of a machine, or a parasite attacking another organism. For example, contributors to the MPD literature frequently make statements such as, “If [the patient] drops her guard, the alters take over” (Bliss 1980, 1393). Proponents describe the original personality as the “host” — again recalling notions of a parasite — and describe the change from host to alter, or from one alter to another, as “switching.” Thus, a librarian may one minute be her forty-two-year-old true shy self, but behave in the next like a nine-year-old child, a deep-voiced, foul-mouthed logger, or a promiscuous woman who picks up men in bars (Putnam 1989, 111, 119-120).

These guest personalities, or “alters,” are believed to have many truly remarkable capabilities and qualities. Some have the task of reproducing — of creating new alters. Others, it is claimed, determine which alter will take control of the body at any particular time (Kluft 1995, 364). There are alters of people of the opposite sex, of the treating therapist, of infants, television characters, and demons. Alters of Satan and God, of dogs, cats, lobsters, and stuffed animals — even of people thousands of years old or from another dimension — have been reported by MPD proponents (Fifth Estate 1993; Ganaway 1989; Hendrickson et al. 1990; Kluft 1991b, 166; Kluft 1995, 366; Ross 1989, 112; Ross et al. 1989).

MPD proponents assert that all manner of activities — creating a work of art, driving a car, fighting, doing schoolwork, engaging in prostitution, cleaning a bathtub, or even baking chocolate-chip cookies — are performed by alters (Braun 1988; Putnam 1989, 104; Ross 1989, 112).

Alters are often wily, secretive, and elusive. For instance, R. P. Kluft (1991a) says he has identified guest personalities whose role is to deny that the patient has MPD, thus obscuring the diagnosis. Personalities are also said to try to trick therapists by hiding and impersonating each other (Putnam 1989, 113). They are said to be plastic: “Alter A may be somewhat different when it has been preceded by alter B than when it follows alter C” (Kluft 1988, 49). They are said to multiply: each alter can undergo a cascade of splits, resulting in what is called “polyfragmented” MPD (Frontline 1995; Ross 1994, 60). Or the opposite may occur: during therapy, several alters may coalesce into a kind of “superalter” (Kluft 1988). It is even claimed that they can permanently stop growing at some time, or temporarily stop aging by going into “inner hibernation” and then emerging to resume growing older (Ross 1989, 112). Cases reported in the last few years have shown a median number of two alters at the time of diagnosis; however, during treatment, a further six or twelve usually appear (Putnam et al. 1986; Ross et al. 1989). Sometimes many more are found: as many as one quarter of cases have twenty-six or more alters (Kluft 1988). And the longer patients remain in treatment, the more guest personalities are discovered (Kluft 1988; Kluft 1989): “It is the rule rather than the exception for previously unknown personalities to enter the treatment” (Kluft 1988, 54). Patients with 300 and 4,500 personalities have now been reported (Kluft 1988; Ross 1989, 121; Ross et al. 1989). Kluft has been consulted “several times” on cases where therapists claim — wrongly, Kluft says — to have counted “upward of 10,000 alters” (Kluft 1995, 363).

Why this nearly endless flowering of personalities? According to MPD proponents, it occurs because each trauma or major life change experienced by an MPD patient causes some or all of the alters to be created anew (Kluft 1988).

What Causes MPD?

According to proponents, extraordinary childhood traumas — usually sexual or other abuse by adults — lead to MPD.

The theory is as follows. Because the child cannot physically escape the pain, its only option is to escape mentally: by dissociating. Dissociation is said to defend against pain by allowing the maltreatment to be experienced as if it were happening to someone else (Atchison and McFarlane 1994; Braun 1989; Kluft 1985a; Kluft 1987; Ross 1995). The distress of this childhood maltreatment is also endured by employing repression, a mental mechanism that supposedly allows the child to forget that the abuse happened at all (Lynn and Nash 1994): “Now, not only is the abuse not happening to me, [but] I don't even remember it” (Ross 1995, 67).

Eventually, MPD proponents claim, these defenses begin to be overused — that is, enlisted more and more to cope with commonplace, everyday stressors (Braun 1986, 66; Putnam 1991). The abuse victim’s “dissociated internal structures are slowly crystallized” until they become personalities (Atchison and McFarlane 1994; Putnam 1989, 53-54; Ross 1995a, 67). As mentioned earlier, this alter-building process is supposed to occur almost exclusively in early childhood (Greaves 1980; Vincent and Pickering 1988).

What’s Wrong Here?

So stands the tottering house of MPD theory. Its foundation crumbles and termites gnaw; the storm beats upon it.

The house suffers from at least four serious ailments.

The first: What, exactly, is an “alter personality"?

One might believe that the disorder’s proponents would long ago have taken the elementary step of answering this fundamental question. Such a belief would be mistaken. The MPD literature contains not one single plain, understandable definition that would allow an alter to be recognized if it were encountered on the street, in a person one has known intimately for years, or even in oneself.

The vagueness and imprecision of the alter concept are shown by the frequency with which even MPD experts contradict each other on the fundamental attributes of these entities. As an example, Ross (1990) says patients’ minds are no more host to many distinct personalities than their bodies are to different people; another theorist believes that alter personalities are imaginary constructs (Bliss 1984). But in contradiction, DSM-IV and the writings of several MPD theorists repeatedly stress that alters are well-developed, distinct from one another, complex, and well-integrated (Kluft 1984b, Kluft 1987; Taylor and Martin 1944). Also, MPD-focused practitioners routinely report patients who have dozens or hundreds of personalities — yet Spiegel (1995) has recently claimed that because MPD patients cannot integrate various emotions and memories, such patients actually have less than one personality, not more than one.

Contradictions abound elsewhere, too. On the one hand, Bliss (1984) believes personalities have specific and limited functions, and possess only a narrow range of moods. But on the other, Braun (1984) and other proponents (Putnam 1989, 104; Ross 1989, 81) say that fragments do not have a wide range of mood or affect. One proponent states that fragments “carry out a limited task in the person’s life” (Ross 1989, 81), but then later in the same publication (111-118) argues that personalities may perform only one specific function, represent only a single mood or memory, or exhibit only a narrow range of skills.

This failure to rigorously define the concept of a guest personality leads to all manner of excesses. For example, MPD proponents discover MPD in people whose close relatives, and others who have known those people for years, have never once seen any evidence of alters (Ganaway 1995). Kluft (1985b), for instance, diagnosed the disorder in a series of people — even though he himself acknowledged that almost half of them showed “no overt signs” of MPD. These proponents also find MPD even in people who lack any knowledge whatever of having the condition (Bliss 1980; Bliss 1984; Kluft 1985b), and at least one enthusiast recommends that people be treated for MPD even if they claim not to have the disorder (Putnam 1989, 139, 215).

The imprecision of the alter concept allows MPD adherents to claim that scores of patient behaviors should signal the possible presence of guest personalities. Thus, adherents claim that the following behaviors — and many others — are important diagnostic clues for MPD: glancing around the therapist’s office; frequently blinking one’s eyes; changing posture, or the voice’s pitch or volume; rolling the eyes upward; laughing or showing anger suddenly; covering the mouth; allowing the hair to fall over one’s face; developing a headache; scratching an itch; touching the face, or the chair in which one sits; changing hairstyles between sessions; or wearing a particular color of clothing or item of jewelry (Franklin 1990; Loewenstein 1991; Putnam 1989, 118-123; Ross 1989, 232). In one case known to the author, a leading MPD proponent claimed that the diagnosis was supported by behavior no more remarkable than the fact that the patient changed clothes several times daily and liked to wear sunglasses.

These beliefs about personalities raise some difficult questions that MPD enthusiasts fail to answer. First, how does alter-induced behavior differ from behavior people show every day — say, when they are angry or happy (Piper 1994a)? Do indwelling alters or fragments cause all feelings? If not, how does one determine which emotions result from the activities of alters, which from those of fragments, and which from neither? One proponent acknowledges the difficulty posed by these questions: he says alters may be indistinguishable from the original personality (Kluft 1991b).

Second, how do persons claiming they are overpowered by “irresistible alters” differ from those who attempt to avoid legal sanctions by claiming that, when they committed crimes, they couldn't control their behavior (Piper 1994c)?

Finally, one wonders how seriously to take MPD enthusiasts’ claims that they can accurately keep track of fifteen or thirty invisible alters — or 4,500 — when those alters are deceiving the therapist, growing, splitting, ceasing to age, reproducing, coalescing, going into “inner hibernation,” and changing their characteristics depending on which personality preceded or followed their appearance.

In summary, knowing how to test or prove an assertion that an individual has more than one personality, or how to clinically distinguish between personalities, ego states, identities, fragments, personality states, or the like, is impossible in the absence of agreement about what any of these terms mean (Dinwiddie et al. 1993; Aldridge-Morris 1993, ch. 1). It follows, then, that few limits exist to the number of “personalities” one may unearth. The number is restrained only by the interviewer’s energy and zeal in searching, and by his or her subjective — and perhaps idiosyncratic — sense of what constitutes an alter (Dinwiddie et al. 1993).

Enthusiasts thus expand the concept of personality beyond all bounds. If such a grandly expansive definition is employed, finding thousands of MPD “patients” becomes simple. Without clear behavioral criteria allowing the observer to know when a personality has been encountered, the term personality comes to mean anything and everything patient and clinician want it to. It thus comes to mean nothing.

The second affliction of the house of MPD is laid bare by one startling fact: the disorder’s most dramatic signs appear after, not before, patients begin therapy with MPD proponents.

Those eventually given this diagnosis seek professional help because of many different kinds of psychiatric difficulties. When first presenting for treatment, these patients can exhibit signs or symptoms of each and every psychiatric condition (Coons et al. 1988; Putnam et al. 1986; Bliss 1984). One complaint, however, is conspicuously absent: evidence of separate alter personalities (Brick and Chu 1991; Franklin 1990; Kluft 1984a; Kluft 1985a; Ross 1989, 93).

But when the patients enter MPD-focused therapy, signs of alters’ behaviors skyrocket. For instance, one patient’s guest personalities created apparent grand mal seizures (Kluft 1995); another sold drugs when the host was supposed to be at work (the host would supposedly “come to” miles away) (Putnam 1989, 198). According to proponents, much of the behavior of MPD patients results from alters’ “personified intrapsychic conflicts” (Putnam et al. 1986, 291); the personalities create crises in the patient’s life by attempting to dominate, sabotage, and destroy one another (Kluft 1983; Kluft 1984c). As one example, an alter may lead the patient into compromising circumstances — say, a sexual encounter, an episode of firesetting, or an illegal drug purchase. This personality then vanishes, leaving the patient, who “wakes up” not knowing how he or she got into the situation, to handle the problem (Confer and Ables 1983; Kluft 1991b).

MPD patients often significantly deteriorate during treatment (Kluft 1984c; Ofshe and Watters 1994, ch. 10; Pendergrast 1995, ch. 6). One of the disorder’s leading adherents acknowledges that MPD psychotherapy “causes significant disruption in a patient’s life outside the treatment setting” and that suicide attempts may occur in the weeks following the diagnosis (Putnam 1989, 98, 299). As MPD psychotherapy progresses, patients may become more dissociative, more anxious, or more depressed (Braun 1989); the longer they remain in treatment, the more florid, elaborate, and unlikely their stories about their alleged childhood maltreatment tend to become (Ganaway 1995; Spanos 1996, ch. 20). This worsening contributes to the lengthy hospitalizations — some costing millions of dollars (Frontline 1995; Piper 1994b) — that often occur when MPD patients who are well-insured are treated by the disorder’s enthusiasts. Hospitalizations occur more frequently after the MPD diagnosis is made (Piper 1994b; Ross and Dua 1993).

MPD-focused therapists have struggled mightily to explain these rather embarrassing results of their interventions. Examining these explanations is beyond the scope of this article: see Piper 1995; Piper 1997; Simpson 1995. However, several recent malpractice juries have found the explanations unimpressive. The juries have preferred a simple and logical explanation for the worsening status of these patients: patients worsen after beginning MPD-focused therapy because therapists cause them to do so — by, among other things, encouraging ever-more-dramatic displays of “alters.”

One important way in which therapists encourage such displays is to behave as if alter personalities were real. For example, leading authorities in this field routinely call alters out, hypnotize them, engage in “lengthy monologues” with them, name them, establish treatment alliances with them, talk to their stuffed animals, take them for walks to McDonald’s ("The outside world often seems very big and frightening to child personalities”), engage in playful parody and sarcasm with them, allow them to work on age-appropriate childrens’ projects in occupational therapy ("to show respect for the alter”), and recruit one alter to keep another from hurting still a third (Ross 1989, 227, 252-254; Ross and Gahan 1988). Other MPD adherents encourage alters to solve problems among themselves, to learn the Golden Rule, to participate in “internal group therapy,” and even to decide whether or not the host should enter treatment (Caul 1984; Kluft 1993; Ross 1989, 209).

In 1988, Vincent and Pickering noted that in the published reviews of the literature, exactly one case presenting in childhood was reported in the 135 years prior to 1979. After reviewing the literature published since 1979, they were able to gather a mere twelve cases. (It seems, however, that Vincent and Pickering had to stretch a bit to find even those — four of the twelve were examples not of MPD, but rather of something the authors called “incipient MPD.”) Nine additional cases were found by Peterson (1990).

These minuscule numbers, standing in stark contrast to the thousands of adult cases discovered in recent years, reveal the third weakness: if MPD results from child abuse, then why have so few cases been discovered in children?

The fourth and final weakness of the house is that it is built in a bog, namely, the belief that childhood maltreatment causes MPD. The literature strongly implies that childhood trauma has been unequivocally established as the primary cause of the disorder, and that severe sexual abuse more or less directly leads to MPD (Braun 1989, 311; Ellason and Ross 1997; Putnam 1989, 47; Ross 1989, 101; Ross 1995, 505; Schafer 1986).

Several commentators have recently noted this formulation’s deficiencies. Esman (1994) warns of the dangers of attempting to discover unitary causes of psychiatric disorders; he urges “measured skepticism” about assigning a role for sexual abuse, independently of other aspects of disturbed family function, in the genesis of later adult psychopathology. Numerous investigators, raising similar cautions, state that general family pathology in childhood better predicts adult dysfunction than does childhood sexual abuse alone (Bifulco et al. 1991; Fromuth 1986; Harter et al. 1988; Levitt and Pinnell 1995; Nash et al. 1993). Further, studies repeatedly note the difficulty of separating effects of abuse from the “matrix of disadvantage” giving rise to that abuse (Nash et al. 1993; Bushnell et al. 1992; Hussey and Singer 1993; Mullen et al. 1993). And finally, recent studies warn of the “very real uncertainties that surround evidence” concerning the relationship between childhood sexual abuse and psychiatric disorders (Fergusson et al. 1997), and conclude that available evidence to date does not support sweeping generalizations about childhood sexual abuse as an isolated cause of adult psychopathology (Beichtman et al. 1992; Finkelhor 1990; Levitt and Pinnell 1995).

The evidence for and against a relationship between trauma and dissociative pathology has also been examined. The data should “inspire skepticism, or at least serve to mute the grand conclusions about univariate cause and effect between trauma and dissociation that abound in the professional and lay literatures” (Tillman et al. 1994, 409).

Yet another weakness of this literature is inadequate verification of its child-abuse claims (Frankel 1993; Piper 1994a; Piper 1997). MPD patients very often report bizarre and extremely improbable experiences. For example, in a recent case familiar to the author, one patient claimed to have witnessed a baby being barbecued alive at a family picnic in a city park; another patient alleged repeated sexual assaults by a lion, a baboon, and other zoo animals in her parents’ back yard — in broad daylight. (It should be mentioned that both therapists in these cases are prominent MPD adherents, and neither appeared to have any difficulty believing these allegations). Despite the frequency of claims of this type, “repressed memory patients are seldom referred to medical doctors for examination and possible corroboration of past abuse [though one would assume that] the horrific physical abuse allegedly experienced . . . would require medical care at some point” (Parr 1996). (Space limitations limit discussion of this weakness; see Jones and McGraw 1987; Lindsay and Read 1994; Ofshe and Watters 1994; Pendergrast 1994, chs. 3-5; Spanos 1996, ch. 20; Wakefield and Underwager 1995, ch. 10).

The logic of the claim that childhood trauma causes MPD demonstrates a final serious flaw. If the claim were true, the abuse of millions of children over the years should have caused many cases of MPD. A case in point: children who endured unspeakable maltreatment in the ghettoes, boxcars, and concentration camps of Nazi Germany. However, no evidence exists that any developed MPD (Bower 1994; Des Pres 1976; Eitinger 1980; Krystal 1991; Sofsky 1997) or that any dissociated or repressed their traumatic memories (Eisen 1988; Wagenaar and Groeneweg 1990). Similarly, the same results hold in studies of children who saw a parent murdered (Eth and Pynoos 1994; Malmquist 1986); studies of kidnapped children (Terr 1979; Terr 1983); studies of children known to have been abused (Gold et al. 1994); and in several other investigations (Chodoff 1963; Pynoos and Nader 1989; Strom et al. 1962). Victims neither repressed the traumatic events, forgot about them, nor developed MPD.

Concluding Comments

In the epigraph that begins this article, Upham speaks of the excesses of the seventeenth-century New England witchcraft craze. The story of Sarah Good exemplifies those excesses (Rosenthal 1993). In March of 1692, when thirty-eight years old and pregnant, she heard her husband denounce her to the witchcraft tribunal. He said that either she already was a witch, “or would be one very quickly” (Rosenthal 1993, 89). No one had produced evidence that she had engaged in witchcraft, no one had seen her do anything unusual, no one had come forward to say they had participated in satanic activities with her. But no matter.

On July 19, 1692, Sarah Good died on the gallows.

Three hundred years later, a woman in Chicago consulted a psychiatrist for depression (Frontline 1995). He concluded that she suffered from MPD, that she had abused her own children, and that she had gleefully participated in Satan-worshiping cult orgies where pregnant women were eviscerated and their babies eaten. Her failure to recall these events was attributed to alters that blocked her awareness. No one had produced any evidence for the truth of any of this, no one had seen her do anything unusual, no one had come forward to say they had participated in satanic activities with her. But no matter.

The doctor notified the state that the woman was a child molester. Then, after convincing her that she had killed several adults because she had been told to do so by satanists, he threatened to notify the police about these “criminal activities.”

The woman’s husband believed the doctor’s claims. He divorced her. And, of course, because she was a “child molester,” she lost custody of her children.

Charles Upham recognized the importance of erecting barricades against addlepated ideas blown by gales of illogic. The twentieth-century fad of multiple personality disorder indicates that even after a third of a millennium, such bulwarks have yet to be built.



  1. In the fourth and latest edition of the American Psychiatric Association’s Diagnostic and Statistical Manual, the disorder has been renamed. Although the third edition called the condition MPD, the fourth calls it dissociative identity disorder. The differences between the two disorders’ diagnostic criteria are slight and mainly cosmetic: in the newer criteria, terms such as identities or personality states are employed, rather than the older personalities. Also, the newer definition emphasizes the patient’s inability to recall important personal information.

    Whether the newer term will become popular has yet to be seen; because MPD has the distinct advantage of familiarity, it will be used in this paper.

August Piper Jr.

August Piper Jr., M.D. is a psychiatrist in private practice in Seattle. His book, Hoax and Reality: The Bizarre World of Multiple Personality Disorder, was published in January 1997 by Jason Aronson, Inc.