Persecutory Alters and Ego States:
Protectors, Friends, and Allies.

Lisa Goodman & Jay Peters
borrowed from


Persecutor alters in Dissociative Identity Disorder are uniformly described in behavioral terms as belligerent, abusive, and violent. While most authors agree that persecutors begin as helpers there is no consensus about their later development or function within the system. This paper presents a theoretical model of the etiology and development of persecutor alters. It elucidates the underlying and continuously protective nature of the alter which becomes masked by the apparently “persecutory” behavior.

Using clinical examples which build on their appreciation of the positive function of persecutor alters the authors present their treatment techniques, which include: engagement, building rapport with the underlying protective function, psychoeducation of the alter, and finally, family therapy style negotiations of roles, expectations, and boundaries.

The paper concludes with an examination of the countertransference issues which commonly arise in working with persecutor alters and their impact on the clinician and the therapeutic task.


“ She should die, she deserves to die. She’s a loser and has been all her life and that’s why I tried to kill her (Christine, speaking of the host).”

Therapists working with either Dissociative Identity Disorder (DID) (American Psychiatric Association, 1994) or Ego State Disorder (Bloch,1991; Watkins & Watkins, 1993, 1992) recognize this angry wail as coming from a “persecutor” or “malevolent” alter or ego state. Such alters are present in somewhere between 50 and 84 percent of Dissociative Identity Disorder cases (Putnam, 1989; Ross, 1989) and can pose considerable risk to the host, frequently disrupt the therapy, and often scare both host and therapist with their vehemence and determination.

Most of the literature on the development of these persecutory alters reports that they usually begin life as protectors and then, for some reason, turn on the host, becoming persecutory. The theoretical reasons given for this change are numerous: it is due to either a “masochistic turning inward of expressions of hostile affect” (Kluft, 1985, p. 183) or an identification with the aggressor (Bloch, 1991) or the "identification with the evil motivations of others" (Bloch, 1991, p. 29).

After reviewing the current literature on persecutory alters we will attempt to formulate a developmental theory which makes clear the persecutors underlying positive role within the system. We will then turn to the treatment implications which result from this perspective. Finally, we will explore some common countertransference responses to persecutor alters and their effect on us as therapists and on the work we do.

Before we begin, however, we are faced with an insurmountable problem: In the literature there is no consistent nomenclature or system of differential diagnosis of persecutor alters. What, exactly, we are talking about is unclear. Many authors such as Ross, Beahrs, and Bloch have established different categories of aggressive alters. Bloch calls them either persecutory or malevolent (1991); Beahrs (1982) differentiates between persecutors and demons; while Ross places them in subgroups of uncooperative alters, angry adolescents, or internal demons who “really want to be contained and loved.” (1989, pp. 255-257). The categorical criteria or descriptive differences each author uses is different from the other authors.

In addition there appear to be aggressive alters who “may be sadistic sex murderers who have committed numerous crimes and be beyond rehabilitation” (Ross, 1989, p. 259-260). While Ross, like Putnam (1989), groups these alters together with the persecutors we are not convinced that they are dynamically and functionally the same. In this paper we are not talking these alters who are truly “beyond rehabilitation,” but rather about persecutors who, as we shall see, are described as malleable, changing from childhood protector to persecutor and then, with proper treatment back into forceful ally. Unfortunately there is no existing nomenclature with which to denote this difference.

Finally, each author proposes quite different treatment approaches for their different persecutor types. This is reminiscent of Ross's critique of the personality disorders field in which one expert's "borderline" is another's "narcissist" who would do poorly if treated like a" borderline" (Ross, 1994).

We believe this confusion and uncertainty both warrants and necessitates further study and the establishment of a system of differential diagnosis of aggressive alters so that in both clinical and forensic settings therapists can accurately, consistently, and safely differentiate the types of aggressive alters.

At present we must simply heed Ross’ advice to “enter negotiations with persecutors cautiously, with eyes and ears open” (1989, p. 260) and, through the techniques we outline here, to assess the function of the behavior. Through this assessment the true nature of the alter will emerge as well as its capacity to form a treatment alliance. In the vast majority of cases this capacity appears to be excellent once understanding has been mutually achieved. It should not. however be assumed without careful assessment.


Description of Persecutor Alters

In the literature on DID and Ego State Disorder (ESD) the description of persecutor alters’ behavioral manifestations is remarkably consistent.

“On first meeting they will be fearsome, loathsome, demon-like entities totally committed to the malicious harassment and abuse of the patient” (Putnam, 1989, p. 205). Watkins and Watkins (1988) describe them as “loaded with rage, they may be both suicidal and homicidal. They slash the patient, strike at others, initiate bizarre behavior and threaten all, including the therapist” (p.68). They may initiate: "headaches, internal bullying, increased blank spells, interference with function, or imposition of unpleasant states on the host personality" (Ross, 1989, p. 255). In a word they are abusive toward the host and often toward other alters (Bloch, 1991), the therapist, family members, and other people.

In addition to the physical abuse of the host there are other forms of “torment” which are “inflicted” on the patient:

“Self-mutilation by persecutors to punish the host or other alters is common.

“The host may also find threatening notes or even more graphic warnings of future mutilation … for example, … a threatening message written in … blood on [the] bedroom wall. (Putnam, p. 206)

The harassment and abuse also frequently take the form of internal talk by the persecutor. “These voices will berate and belittle the patient, threaten or urge suicide, and sarcastically and gleefully taunt the patient about their total control over him or her” (Putnam, 1988, p. 206). The voices will also often demean and belittle the therapist and urge the host to drop out of therapy.

Finally, the persecutors engage in numerous behaviors which compromise the well being of the host. These include such things as alienating friends and family (who frequently withdraw) and anti-social behavior for which the host is then responsible.

Taken together, “the various forms of harassment and the patient’s reactions to them constitute a major source of torment for an MPD patient” (p. 205) as well as very real threats to the health and well being of the host. From observations of the patterns of "harassment" and from the life histories of the persecutors, theorists have attempted to create explanatory developmental and dynamic models which we shall examine in the next section.

Origin and Development of Persecutory Personalities

It is generally agreed in the literature that persecutors start out as friendly, in fact protective alters. This is supported by Kluft’s findings that in childhood DID "persecutor personalities … are notable for their absence" (Kluft, 1985. p. 183) and Bliss’ observation that “all of the personalities begin as friends and allies…” (quoted in Putnam, 1989, p. 208).

The most common explanation in the literature of the childhood protective function, with several variations, is that the persecutor started life as some kind of repository for various painful experiences and emotions. Watkins and Watkins state that dissociation leads to splitting off the rage into a separate ego state which “lays the basis for forming an unconscious destructive, malevolent ego state divorced from normal super-ego controls" (1988, p. 69). Ross also speaks of the persecutory alter as "carrying all the anger" (1989, p. 256). Kluft describes the persecutors as initially "taking all the suffering for the others" (1985, p. 185). Elsewhere they are described as containing the “affect and energy the depressed and apathetic host cannot sustain” (Putnam, 1989, p. 208), as a "personification of the patient's vital life energy (Beahrs, 1982, p. 141), and finally as serving “as [a] crystallization of the client’s aggressive, destructive impulses” (Bloch, 1991, p. 55).

There is less agreement in the literature about the process by which this initially protective container or repository becomes persecutory towards the host. It has been described as the result of the alter’s “becoming impatient” (Bloch, 1991, p.55) or “resenting suffering for others” (Kluft, 1985, 185). The implication here is that due to the level of distress the alter turns on the host. Alternatively, it has also been suggested that the alter changes from protector to persecutor through “a masochistic turning inward of hostile affect ... identifying with the aggressor" (Kluft, 183-185), and similarly, through a process of identification “with the evil motivations of others” (Bloch, 1991, p. 55). Finally the process is explained as occurring, “when later repression breaks down, this [malevolent] state emerges, takes over executive control of the body and vents it’s rage on the patient or others” (Watkins and Watkins, 1988, p, 69).

Functions of Persecutor Alters

Finally in our review of the literature on persecutor alters we noticed a conspicuous gap. In DID and ESD, alters and ego states are generally seen as having certain roles, functions, or purposes within the system (Bloch, 1991). While we have noted some mention in the literature of initial or childhood function, Putnam and Ross are alone in their discussion of persecutor's later roles (Putnam, 1989; Ross 1989).

Putnam states that persecutors “contain the energy and affects that the depressed and apathetic host cannot sustain” (p. 208). He also states that through their threatening behavior persecutors preserve the secrets of the past abuse. In addition to protecting the facts of the abuse the persecutor “serves to keep noxious reminders of the experience at a distance” from both therapist and host. It accomplishes both goals by “creating such an uproar in therapy that the therapist never has a chance to focus on the past” (p. 208).

It should be noted that this is again a behavioral description in which the overarching function of or reason for the behavior is hinted at but never stated. In an early work Helen Watkins, while not outright saying what we think is the important discovery she has made, alludes to her increasing understanding that a persecutor is "blindly protecting [the client] according to its view of itself and its originally created purpose for existing" (Watkins, 1978, pp. 368-369). Ross is alone in formulating the concept that "the hostile behavior is actually protective" (1989, p. 259) or more generally that the persecutor "has a good reason for what she is doing and that her behavior makes sense from her point of view" (p. 258). In the following sections we shall follow Ross's lead and attempt to look at persecutor or malevolent alters not so much in terms of what they do but why they act as they do, the "good reason." We shall attempt to distinguish between means and end, behavior and intent. We shall begin by proposing a theory of persecutor's etiology and development.


Protector Initially formed in childhood

We agree entirely with the observation that persecutors evolve out of helper or protector personalities who first come into existence in the host’s childhood or early adolescence. We disagree, however, with the implication in the literature that the persecutors undergo a transformation of identity role to “become” a malevolent entity. This assumed malleability of essence is taken further in the literature when authors talk about therapy with persecutors and note that the persecutor can be "turned into [a] constructive force" (Watkins, 1978, p. 397) and become “one of the therapist’s strongest allies and can play a major role in the healing of the patient” (Putnam, 1991, p. 205).

In our work with these alters we find no such changes. The protectors are still and always protectors. What changes is the form of the protective behavior which no longer looks obviously protective and which may, in actuality be harmful and life threatening to the host.

Though he is writing about the development of aggression, hostility, and hate in childhood, we find Parens' formulations helpful in understanding the development of these feelings in persecutory alters. Hostility is not inborn but is “generated by unpleasure-inducing events … invariably object-related, which are experienced as excessively painful by the child (and ego)” (Parens, 1994, p. 88, italics original). Under the pressure of “excessive unpleasure” aggression is transformed into hostility which “has as it basic aim to act upon, assert oneself over, and control” the object in an effort to stop the excessive unpleasure (pp. 80-81). The hostility is thus instrumentally used and "the inflicting of pain or harm upon, and thus effecting the hostile destruction of, the object" is in the service of control of the object (p. 81).

In our view the most important change in the development of persecutor alters is that in adolescence or adulthood the protector perceives the host or the host’s actions as the source of the threat (the object to be controlled) and consequently acts to protect the system from him or her. To understand the change from childhood protector to adolescent or adult persecutor we need therefore to look not at the alter but at the host. We shall do this in the next section but a preliminary case vignette may be helpful. In an HBO special on MPD (Mierendorf, 1993), a persecutor alter who identifies herself only as "Me, Myself" is questioned about her mutilation of the host:

"I would do anything in the world to destroy anything she did and hurt her any way I could. I used to be one of those inside who would belittle her and called her names and swear at her. I used to cut the shit out of her --- and I'm very good at it. I'm the one who severed the artery and 4 tendons."

The therapist asks: "Why'd you do that?"

"I wanted to kill her."

"What did she do?"

"I stopped growing at 14 because that was when she began becoming interested in boys, and dates, and all that ..........[she trails off], and I hated it and I didn't want any part of it. So I quit, I wasn't going to be any more than 14 'cause nobody was ever gonna touch me ........."

"And whenever that would happen with Gretchen [the host] it would hurt me and I would hate it and I'd hate her, and I'd hate her for letting that happen ... so I'd cut her.

The Change to Harassment: Changes in Host’s Behavior

With the increased agency of adolescence and adulthood the host now starts to engage in behavior which the protector assesses to be dangerous. To protect the host he or she must be controlled. The means of that control are the same aggressive thoughts, feelings, or acts which may have previously been directed toward others (in the protective role) but which are now experienced as "persecutory" as they are directed at the host.

Naturally, given the underlying protective role of the persecutor, this ego state may also feel genuine and intense hostility toward the host for putting the system in danger.

The following discussion of specific host behaviors which elicit this change in the protector’s focus of control is not intended to be exhaustive but rather suggestive of the possible range.

Risk-taking behavior is an obvious trigger of the protector’s efforts to control the host. “Traumatized people relive the moment of trauma not only in their thoughts and dreams but also in their actions. In their attempts to undo the traumatic moment, survivors may even put themselves at risk of further harm” (Herman, 1992, p. 39). Now, in adolescence and early adulthood, the host has much more opportunity than in childhood to put herself at risk. He or she has increased mobility, more unsupervised time, and vastly expanded exposure to potential victimizers.

The host increases her exposure to potential abusers and revictimazation not only through her increased exposure to other people but also through what Kluft has termed the “sitting duck syndrome” which leads to frequent involvement in exploitative and abusive relationships (Kluft, 1990). In a context of such ongoing trauma it is only natural that there will be an increasing load of hostility within the system.

Even non-abusive relationships may provoke the protector’s vitriolic reaction if the relationship takes on a meaning which feels threatening. Simply feeling the possibility of closeness to another person may be the trigger because of the protector’s prior learning that trust and dependency lead surely and inevitably to abuse and hurt. The history of relational violation leads to the equation of relationship and violation. Another factor may be the perceived threat of sex. To protectors like "Me, Myself" all sex may be experienced as invasion and abuse.

Another threat to the system which the host may pose in adolescence or adulthood is of breaking the silence about the abuse and/or the multiplicity. This threat is often raised to the level of crisis when the host enters treatment. We then witness the protector’s desperate attempt to control the client, to “save her” from the expected dire consequences of revealing the secrets.

While these perceived threats appear to us to be originating from the external world the protector perceives them as caused by the host, as under the host’s control. In the same way that the victim usually blames herself for the abuse, so to the protector blames her. Therefore the protector's abusive behavior is directed at the host in an effort to control her behavior.

Finally, the host is often perceived as a threat for what he or she is not doing; for not protecting herself, for not getting out of an abusive relationship, for not taking better care of her body, for not sticking up for herself, for not expressing anger, and for a thousand other things which we well know are frequently difficult for survivors. What she is not doing is the fuel for the protector’s charge that she or he is a hopeless “wimp.”

In conclusion, we agree with the common belief that persecutors start out as protectors; we disagree, however, with the idea that their basic identity changes and that they “become” persecutors. We think that in fact these supposed persecutors have not changed at all and that they are still protectors. What has changed is what needs to be protected against. In adolescence or adulthood the host him or herself is perceived as the threat and the protector shifts its aggressive behavior toward the host because of this fact. In the next section we shall focus on treatment of persecutors and how we can help them carry out their underlying protective function more appropriately.


There is a lot of valuable literature on the treatment of DID, ESD, and other varieties of dissociative states (Beahrs, 1982; Bloch, 1991; Kluft, 1990; Putnam, 1989; Ross & Gahn, 1988; Ross, 1989; and Watkins & Watkins 1992, 1993). Much of that literature, especially Ross, speaks directly to working with persecutor alters and is highly recommended. In this section we shall augment that body of work and focus on specific techniques which we find useful.

Watkins and Watkins, in their work with covert ego-states have evolved a treatment model which is distinguished by “the use of group and family therapeutic techniques for the resolution of conflicts between the various ego states that constitute a ‘family of self’ within a single individual” (1992, p. 29 italics original). We use this model extensively and in this section expand on their ideas.

In our view, treatment of the “family” with a persecutor can and must be divided into two broad stages. In the first stage the therapist creates an alliance with the “family” and helps all parties (therapist, host, and alter) come to a more accurate assessment of the problem. In our experience the “multiple family” with a persecutor alter is much like the family who presents with an acting out adolescent child whom they label as the identified patient (IP), and the entire source of the problem. In both “families” there is usually intra-system agreement that everything would be just fine if they could change or extrude the IP. Clearly we must resist the temptation to join the “family’s” view of the problem and, as in family therapy, must therefore challenge the system and shift the focus off the IP and onto the broader issues of system roles, functions, and boundaries (Minuchin, 1974; Napier,1978). This is quite similar to Ross who "tell[s] the persecutor that I assume she has a good reason for what she is doing" and "that I assume she is there for a good reason" (Ross 1989, pp. 258 & 256).

Once a more realistic appraisal of the problem has been reached (often no minor accomplishment), the second stage commences in which conflicting needs are negotiated and harmony among the “family members” is increased (Watkins & Watkins, 1992).

During both of these stages there are two essential requirements for effective treatment of persecutor alters. The therapist must enter the work with an unassailable faith in the underlying positive function of the alter. Then, for those times when that faith inevitably crumbles he or she needs access to a supervisor with the same faith, intact. With these requirements met the therapist will be able to avoid the mistakes which we most frequently see: failing to recognize the positive and protective function of the alter or agreeing with the system that the "IP" really is the problem and that everything would be just fine if they could extrude this troublesome member.

Stage one

With our assumption that the persecutor is not the problem firmly in hand we can begin the engagement phase in which our first goal is to explore the function of the persecutor. At this point we discover there are two types of persecutors. First, there are those who know why they persecute the host (“I punish her so she won’t tell about the abuse cause then he’d kill us”).

CASE A: Susan and Shadow

Susan is a 31 year old woman who came in for treatment when she started recovering memories of incest which had begun in early latency with fondling and proceeded to rape just prior to adolescence. Susan reported a host of dissociative symptoms as well as feeling as if she had a "monster inside me who won't not allow me to heal.”

One day Susan suddenly began talking extensively about the abuse. The therapist (L.G.) successfully slowed her down, but even so, as Susan left the building she started having suicidal thoughts which included throwing herself in front of an oncoming bus.

Before the next session the therapist received a letter from an ego state named Shadow. The letter detailed how big and fierce Shadow was and how she was supposed to be feared, have control, and all the power.

In the next session Shadow revealed that she had been using fear to prevent Susan from disclosing the abuse. Shadow firmly believed that Susan's father would kill them if the abuse were disclosed. Shadow would therefore use either physical ailments or terrifying suicidal ideation to distract Susan or scare her out of talking

With ego states such as Shadow it is relatively easy to discern their underlying function and to clarify this to the alter and to the system as a whole. Here, simply "pointing out that apparently destructive actions are actually meant to be protective softens the though stance" (Ross, 1989, p. 259).

Unfortunately there are also persecutors who do not have any awareness of the purpose of their behavior (“She’s a wimp who deserves to die”). In this second case both alter and host often firmly believe that the persecutor’s function is simply to torment the host. The persecutor can be seen as an IP who has come to believe that he or she really is the problem and is proud of it. A case example, followed by material from subsequent supervision will make this clear.

Case B: Mary and Victor

Mary is a 39 year old woman who presented with the classic history of physical, emotional, and sexual abuse leading to DID. Soon alter personalities began writing to the therapist (J.P.) and communicating internally with Mary. As each state first communicated it would announce how “bad” and dangerous it was, and that it intended to hurt the host “every day of her life”

After successfully working with three of these protective alters, each more vitriolic and abusive than the last, Victor emerged. Victor announced himself as “totally evil.” He talked with glee and gusto about how he liked to hurt women and how much he enjoyed it when Mary was physically abused by her husband. “I liked it when his fists hit her face.” He also “liked it” when she had been raped as “it was good for her, she got what she deserved.”

When asked about his role and function Victor could only talk about his plans to take over the whole system, to kill off all the “good alters” and, in the end to kill Mary.

Case Supervision: Case B, Mary and Victor

Following his initial meeting with Victor, the therapist came into supervision visibly nervous, shaken, distraught, and angry. He announced: “You know our theory about persecutory alters, well forget it, it doesn’t hold up. The guy I just met has no redeeming positive values. This guy is evil, evil incarnate.” He felt overwhelmed by Victor’s anger and his naked sadism. “There’s no ‘positive intention’ in there, he just loves the pain he causes Mary. It was like sitting with a totally unrepentant rapist in your office while he boasted about his crimes. Conscience, none; repentance, none. This guy has got to go.”

The supervisor (L.G.), in contrast, sat there allowing the therapist to struggle with his feelings, struck by how successful the alter had been in inducing those feelings in her supervisee, but feeling absolutely no animosity toward Victor. In fact she felt a strong liking of Victor, an attraction and kindred recognition that stemmed from subtle cues that the feelings stirred in the therapist were being intentionally engendered and did not reflect the essence of the ego state's relation to the host. Victor was inducing those feelings for a reason.

To untangle those subtle cues the supervisor repeatedly asked not about the behavior but about the effect of Victor's behavior on Mary.

"So what does she do differently because she's so scared that Victor will come out and kill somebody?"

Gradually it emerged that Mary's fear of Victor "forced" her to limit the visits from her abusive ex-husband to situations where he could not physically or sexually abuse her, to set increasingly appropriate limits on her exploitative adolescent child, to be more angry and self-protective with her still abusive parents, and, in order to protect the therapist, to distance herself from the therapy which was seen as threatening the stability of the system.

This supervision highlights a number of quite common themes in working with persecutors. The persecutor frequently protects the host through scare tactics; through fear and intimidation. In order for this scare tactic to work the host must be convinced of the persecutor's capacity and willingness (even desire) to use the utmost in force and destructiveness. The persecutor must, in short, be viewed as Putnam described: "a fearsome, loathsome, deamon-like entity (Putnam 1989, p. 68).

When confronted with threats arising from the host's behavior the persecutor appears to say to the host: "I'll do such and such horrible thing if (or unless) you do thus and so." When the danger posed comes from outside the system as is the case in the clinical example, the persecutor uses the same scare tactics, trying to impress the intruder (in this case the therapist) with his or her ferocity and proclivity to violence.

This case is also one in which the host, throughout adolescence and into adulthood, maintained contact with her abusers. The persecutor therefore becomes increasingly loaded with hostility toward the host, to the point were the original protective function is lost to consciousness.

Finally, it should be obvious that the persecutor who uses threats of violence to protect the host may be quite worried about either the host or therapist uncovering the underlying protective intent for to do so is, in essence, to disarm the power of the threat.

In therapy, as in supervision, we try to get at the intention of this type of persecutor's behaviors by reviewing specific incidents and the results that follow. We look primarily for the effect of the behavior on the system and those who impact the system. Whether conscious or not, it is this effect which the persecutor is after. Thus even as Victor bragged about his total commitment to malice and torture we looked at the consequences of the behavior instead of the behavior itself. Using this approach the therapist listened carefully to Victor’s offhand comment that his viciousness “kept all those men away” including the abusive husband. Understanding the intention, the therapist asked: “Oh, so you keep her from getting abused?” When the alter agrees to this sort of reframing of his or her aggressive acts we move quickly, asking about other times he or she had “protected” the host and then mirroring back the intent, divorced from the means. At the same time we begin moving from the specific intentions to the general formulation of the role: “So, it seems like your job is to protect her.”

While this psychoeducational reframing may, as noted above, be met with some initial resistance by the ego state which is afraid of your very understanding, in our experience the persecutor usually relaxes considerably at this point and so begins what many authors have noted as the rapid conversion from (appearing) enemy to forceful ally. As one alter told her therapist (J.P.): “Since I've been coming in here and talking with you and you’ve been explaining things to me, I’ve been feeling a lot better, I haven’t been wanting to hurt her (the host).”

Our second goal in this stage is often achieved simultaneously with the first and consists of an empathic joining with the underlying emotional state of the persecutor alter. As Ross notes: "The most powerful way to form a treatment alliance with hostile alters is to divine their pain and sadness and comment on it" (Ross, 1989, p. 227). It addition to strengthening the treatment alliance the empathic joining allows us to now mirror back not only the alter's function but also the feelings attached to the role and the internal relationship. Joining these two elements inevitably has a profound cognitive impact on the alter and frequently resolves the alter's initial resistance, facilitating engagement.

There are several emotional themes which reoccur in our experience. Frequently the persecutor is tired to the point of exhaustion. He or she feels overburdened by the task, undervalued for the results achieved, and lacking in adequate resources. The persecutor is also frequently angry, angry with people who are perceived as abusing the host and angry with the host for his or her lack of cooperation assuring the safety of the system. Finally, as Putnam notes, while the alter displays “extreme contempt toward the host,” paradoxically, “the dominant emotion of the persecutor may really be love” (1989, p. 207). In fact, it is this very love which propels the persecutor in the harassment of the client through which they are doing the very best that they can to protect her. At the same time the persecutors "act tough but want to be loved" themselves (Ross, 1989, p. 227).

Empathically joining with these feelings is usually quite easy for us for who among us has not felt frustrated and angry with the host when she fails to protect herself, who among us does not occasionally feel exhausted and undervalued by our ego-state clients, and who among us does not simultaneously feel love for them?

Using these countertransferential feelings one of us (J.P.) was able, at the end of a long session with an initially vitriolic persecutor to say honestly: “You must be exhausted, fighting for her all the time as you do.” The protector sighed, appeared to relax her vigilance, and said simply: “Yeah, I want a rest.”

Getting at the love which the persecutor feels for the host is a crucial step in the successful treatment of these ego states. Often this opportunity arises while we are reframing the meaning of the persecutor’s behavior. After the alter told the therapist that she wanted a rest he continued empathizing with how tired she must be, how hard she was working for the host, how her exhaustion reflected just how hard she worked, and that all this work must mean that she really cared what happened to the host, must really love her. Touched in this, her most vulnerable spot, she burst out crying and was finally able, with all the appropriate affect, to talk about her loving, protective relationship with the host.

The other feeling which the therapist must be able to resonate with is the pleasure available to the persecutor in being aggressive. With an adolescent persecutor who talked about the joy of the aggression L.G. responded: "Yea, it's cool to be powerful." The contradistinction to feelings of helpless vulnerability inherent in the past abuse were unsaid but understood between them. What is being validated here is the pleasure in instrumental use aggression and hostility for a purpose. Thus the therapist empathically joins with both the pleasure and purpose of the behavior.

Beyond building the treatment alliance, empathically joining with the full range of emotions of the persecutor is important in the development of ambivalence. As we use the term , ambivalence is " the experience of coexisting feelings … toward an object" (Parens, 1994, p. 98). The development of ambivalence was previously hindered by the fact that the different feelings (love/hostility, mastery/fear) were experienced by different aspects of the personality so that the simultaneity necessary for the development of ambivalence was not possible (Kernberg, 1994, p. 214). Through our empathic attunement we are therefore encouraging co-consciousness of previously dissociated affective states first within a segment of the personality and then within the personality as a whole.

Stage two: Moving on

Now, with a firm working alliance grounded in a truly positive appreciation of the role and emotional state of the alter, the host, persecutor, and therapist are ready to move on to the next stage of the family therapy which has as its goals “increased harmony, communication, and cooperation” (Bloch 1991, p. 71). Quite often reaching this goal means host and persecutor must negotiate differences, resolve long-standing conflicts, and overcome past hurts (both real and imagined).

During this stage of treatment we find that most of our interventions consist of some version of the questions: “What do you need;” “What does everyone need;” or” How can we make sure everyone's needs are met?" Obviously the persecutor usually needs the host to be safe. The host needs to be able to engage in an increased range of activities without interference or retaliation from the protector. Through the repeated articulation and resolution of the question “What do you need” we note an often rapid reduction of hostility between host and persecutor. It appears that as the host increasingly understands the interrelationship between his or her behavior and the persecutor's response a working alliance develops. The persecutor’s perception of the host as listening and taking seriously the persecutor’s perceptions seems to be the key factor. The host does not have to agree completely with the persecutor’s definitions of risk or danger, he or she must simply listen and act respectfully.

In this stage of treatment we, like many other authors, note the persecutor's ready switch to positive, helpful, and life-promoting behavior. We understand this switch as demonstrating that the previously abusive behavior did not reflect an underlying character structure formed through identification with the aggressor and masochistic turning in ward of hostile affects for such a character structure would not be amenable to such rapid changes. Instead we understand the change as demonstrating an past adaptive instrumental use of violence and all it's trappings in an attempt to preserve the system in the face of perceived threat.


Work with persecutors evokes profound and often profoundly distressing countertransference responses in the therapist. Persecutors force us to confront a multitude of issues. On one hand, despite all the injunctions that we not have favorites among the alters, we often feel intense anger toward this alter for the pain, suffering, and humiliation he or she inflicts on our client. We may also want to save our client from this “abuser” as we were not able to save her from the original abuse, thus expiating our “survivors guilt.” At the same time, and most threatening of all, we may share some of the “persecutor’s” feelings of hate, disgust, anger, frustration, and disappointment toward our client for not protecting herself.

Countertransferential feelings such as these toward any client are difficult for the therapist to deal with. When held toward the victim of chronic and often sadistic abuse they become intolerable. To the degree that we can neither express nor sit with these feelings, we must defend against them. Projection, in which “consciously disowned aspects of the self are rejected or disowned and thrown outward and imputed to others (White & Gilliland 1975) comes to our rescue with a ready target in the “persecutory alter.” It is the alter, not us, who has these intolerable feelings toward the host. The parallel process at work here should be noted. In the same way that the persecutor served originally as a container for the host’s split off, unacceptable thoughts and feelings, it now serves that same role for the therapist.

As a result we are then unable to resonate with the true, protective aspects of the persecutor alter because of our investment in maintaining the persecutor as the container for our own uncomfortable abusive feelings toward the host. We join the host in resistance to integration.

To aid in this process we go a step further and create a theory in which the persecutor is seen as undergoing a transformation from an initially helpful alter into an abuser. Now the persecutor is the bad guy while we, on the other hand, can prove ourselves the “good guy” by rescuing our client from the persecutor’s reign of terror. Theory is thus created and defended in the service of our countertransference resistance. Unfortunately this process assures that we remain blinded to the underlying positive function and unable to truly empathize with this personality.

Overcoming this countertransference resistance and its negative consequences demands that:

"Therapists must be prepared to acknowledge that the capacity to abuse, and be abused, is a recurrent and tragic feature of the human condition, and that the pain of bearing witness to this reality resides in our own vulnerability to aggression. That is to say, that our patient's victimization confronts us not only with the evil in the world around us, but with our own capacity to be intentionally or unintentionally hurtful. (Marcuse, 1994)

Working with persecutors requires that we check our responses carefully, putting our feelings aside until, through looking at the effect of the persecutor's behavior we come to fully understand their protective role. We may well then discover that our feelings of fear and loathing were exactly the effect the persecutor wished, in their endeavor to protect the host from the threat of our understanding, our concern, and our closeness.


In conclusion, we agree with the common belief that persecutors start out as protectors. Then, in adolescence or adulthood, the host's actions are perceived by this protector as a threat to the system and the protector shifts its aggressive behavior toward the host in order to control him or her.

In treatment we look steadfastly at the effect of the persecutor's behavior in order to get at the motivating intention. We then form a bridge between that intention and the concept of a protective role and from that role to the underlying affective state. This bridge allows the alter and host to begin real communication and negotiation about their often conflicting needs. This communication inevitably shifts the previously split off function of the protector back into the host's ego sphere leading to integration of the previously split of protective function.

Some areas in need of further study and understanding have already been mentioned such as the need for a unified system of nomenclature for aggressive alters. Such a system of naming will require that there first be a reliable system of differentiation of types of aggressive alters. The problems of establishing reliability and validity of such a system are enormous but would be of equal importance through helping us identify those alters (such as the persecutors we have discussed in this paper) who are quite amenable to treatment from those for whom treatment is not indicated in this time of scarce clinical resources.

In this paper we have proposed a developmental theory of persecutor alters in which they are seen as protectors who change the techniques used to carry out that role but not the underlying role. While this formulation appears to challenge current psychodynamically oriented developmental theories, we hope that other writers will be able to synthesize our ideas with these theories, resulting in understandings of greater complexity and richness as well as enhanced clinical utility.


American Psychiatric Press (1994). Diagnostic and statistical manual, fourth edition. Washington, DC: American Psychiatric Press

Beahrs, J. O. (1982). Unity and multiplicity: Multilevel consciousness of self in hypnosis, psychiatric disorder and mental health. New York, NY: Brunner/Mazel.

Bloch, J. P. (1991). Assessment and treatment of multiple personality and dissociative disorders. Sarasota, FL: Professional Resource Press.

Herman, J. L. (1992). Trauma and recovery. New York, NY: Basic Books.

Kernberg, O. F. (1994). The psychopathology of hatred. Journal of the American Psychoanalytic Association, 39 (Supplement), 209-238.

Kluft, R. P. (1985). Childhood multiple personality disorder: Predictors, clinical findings, and treatment results. In R. P. Kluft (Ed.) Childhood antecedents of multiple personality disorder. (pp. 167-196). Washington, DC: American Psychiatric Press

Kluft, R. P. (1987). First-rank symptoms as a diagnostic clue to multiple personality disorder. American Journal of Psychiatry 144, 293-298.

Kluft, R. P. (1990). Incest and subsequent revictimization: The case of therapist - patient sexual exploitation, with a description of the sitting duck syndrome. In R. P. Kluft (Ed.). Incest-related syndromes of adult psychopathology. (pp. ). Washington, DC: American Psychiatric Press

Kluft, R. P. (1991). Multiple Personality Disorder. In A. G. Tasman, & S. Goldfinger, Review of psychiatry: Volume 10. Washington, DC: American Psychiatric Press.

Marcuse, J. J. (1994). Dissociation and enactment: Countertransference considerations in work with adult survivors of abuse. Paper presented at William Alanson White Institute, New York, NY.

Minuchin, S. (1974). Families and family therapy. Cambridge, MA: Harvard University Press.

Mierendorf, M. (Producer. Director. & Writer), (1993). Multiple personalities: The search for deadly memories. America Undercover Series. New York, NY: Home Box Office.

Napier, A. with Whitaker, C. (1980). The family crucible. New York: Bantam Books.

Parens, H. (1994). A view of the development of hostility in early life. Journal of the American Psychoanalytic Association, 39 (Supplement), 75-108.

Putnam, F. W. (1989). Diagnosis and treatment of multiple personality disorder. New York, NY: The Guilford Press.

Ross, C. A. and Gahn, P. (1988). Cognitive analysis of multiple personality disorder. American Journal of Psychotherapy, 32, 229-239.

Ross, C.. A. (1989). Multiple personality disorder: Diagnosis, clinical features, and treatment. New York, NY: John Wiley & Sons.

Ross, C. A. (1994). The scientific state of the dissociative disorders field. Treating Abuse Today. 3 (5), pp. 4-9.

Watkins, H. H. (1978). Ego-state therapy by Helen H. Watkins. In J. G. Watkins (Ed.). The therapeutic self: Developing resonance—key to effective relationships.. (pp. 360-397). New York, NY: Human Sciences Press.

Watkins, J. G. & Watkins, H. H. (1993). Ego-state therapy in the treatment of dissociative disorders. In R. P. Kluft, Catherine G. Fine. (Eds). Clinical perspectives on multiple personality disorder. (pp. 277-299). Washington, DC : American Psychiatric Press.

Watkins, J. G. & Watkins, H. H. (1992). Hypnosis and ego-state therapy. Innovations in clinical practice: A source book (Vol. 10). (pp. 23-37). Sarasota, FL: Professional Resource Exchange

Watkins, J. G. & Watkins, H. H. (1988). The management of malevolent ego states in multiple personality disorder. Dissociation 1, 67-72

White, R. B. & Gilliland, R. M. (1975). Elements of psychopathology: the mechanisms of defense. New York, NY: Grune & Stratton.