When Mediation Goes Awry: Reflections On The Internal Complexities Of Being Human

One observes that in a large number of people, placed in the most diverse conditions, the normal unity of consciousness is disintegrated. Several distinct consciousnesses arise, each of which may have perceptions, a memory, and even a moral character, of its own…. Alfred Binet (1892)

The degree of dynamical interconnectedness of the different parts of the person can be nearly equal within the whole region of the person, or certain regions can separate themselves to an especially high degree from others and develop relatively independently. This can be observed in the normal person….

Kurt Lewin (1936)

Many people lose their battles on the outside because they waste their energy on the inside. Our inner life and our outer coping are linked. One feeds into the other.

Virginia Satir (1978)

We all have times when we are “out of character,” feel outside and beside ourselves, and act in unexpected ways. And we’ve all observed such aberrations in others. Sometimes we are intrigued or even amused by these atypical behaviors; more often, however, we are puzzled, distressed, put off, and embarrassed by them. Naomi Quenk (1993)


I have been learning to practice transformative mediation in the workplace relatively late in my professional life, after having spent twenty years as a clinical psychologist. As a psychologist, two phenomena I soon encountered during mediation sessions seemed especially interesting.

The first observation is that things can be progressing fairly well during the problem-solving phase, and then, suddenly, one participant can become recalcitrant, hostile, or extremely distressed, or inexplicably start to ham-string the process for no apparent reason. The second situation occurs when two individuals seem to be on the brink of signing a mutually advantageous agreement. Then, all of a sudden, one party reneges and refuses to sign for no apparent reason — or for reasons that don’t seem to “make sense” to the mediators or the other participant. A variant of this second phenomenon occurs when an individual thinks about the agreement he or she has signed overnight and then refuses to honor terms of the agreement the next day.

I would assume that more experienced, professional may be starting to wonder about “rookie mistakes” or incomplete recognition of some aspect of “the process” of mediation. Indeed, such technical errors on the part of the mediators may well contribute to these two phenomena. However, I would like to raise another possibility; namely, that our common, everyday, seat-of-the pants personality theories are incomplete and unhelpful when it comes to understanding rapid, emotionally-based changes in human behavior. In short, many participants in mediations may be substantially more complex than we realize.

The Need for Personality Theory

Whether or not they have studied psychology intensively, most mediators are armchair personality theorists. In the words of Naomi Quenk (1993):

Each of us operates with some system for understanding people, whether we use a formal theory or just an implicit set of general guidelines. Most of us use our daily observations of other people and our own reactions to what we observe as a basis for generalizing, explaining, and predicting our own and others’ behavior.

Whether formal or informal personality theories attempt to organize observations of people by providing some kind of underlying framework for classifying and describing behavior. Formal personality theories, such as those put forth by psychologists, tend to use concepts and language that are not easily understood by most people. As a result, most of us devise our own system for making sense of people’s attitudes, motives and behaviors.

On rare occasions, the concepts in a formal personality theory are also meaningful and useful to people who are not specifically trained to understand psychological complexities. (p. 2)

The Limits of Traditional Personality Theories

Most academically respectable personality theories invoke a certain number of key traits that supposedly account for much of the variance in human behavior. Implicit in trait views of personality psychology is the assumption that we all have one basic personality (i.e., just as we each have a single body housing our multiple organ systems). It is further assumed that this basic personality that can be adequately described as having so much of this trait and some much of that trait. Such traits are most often measured by scales of 20-30 items that are grouped together in structured personality inventories. For example, the Myers-Briggs school of personality assessment (frequently used in business settings) holds that the important explanatory traits are: introversion/extraversion, sensing/intuition, thinking/feeling, and judging/perceiving. In statistically based, factor-analytic studies of various personality inventories, the five cardinal traits of human personality are typically: Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness.

Trait approaches to personality assessment provide somewhat helpful predictions about an individual’s behavioral tendencies in the longer term (e.g., over the long haul John is more likely to be a conscientious employee than Jim). However, the predictions derived from the finest trait-based approaches to personality have limited utility for understanding intense, short-term behavioral turnabouts or meltdowns, such as those alluded to above. Because mediations often involve intense emotional conflicts between the participants, mediators often observe dramatic emotional transformations and need some means of understanding and making sense of the behavior of virtual strangers (for whom no personality test data is available) within a mediation session.

When a Mediation Participant SNAPs

For want of a better term, I have taken to referring to such abrupt and intense shifts in affect as “SNAPs” — for Sudden, Negative Alterations of Personality. Often, when a mediation client SNAPs, some crucial aspect of the process gets derailed, leaving the mediators and other participants confused, frustrated, or even frightened. Conventional personality theories have little capacity to predict at the micro-interactional level. That is, trait theories generally cannot tell us who is going to SNAP, at what time, for what reasons, or how a SNAP is likely to influence the subsequent course of the mediation session that is already in progress.

Defining Ego-states

I would like to call attention to another method of thinking about personality functioning that is based largely on the work of Watkins and Watkins (1997), two psychology professors from the University of Montana. Basically, the Watkins have theorized that virtually all human beings in Western societies have different aspects to their personality and behavioral functioning. They term these recognizable and distinct aspects “ego-states.” The Watkins’ formal definition of an ego-state is as follows:

An ego-state may be defined as an organized system of behavior and experience whose elements are bound together by some common principle, and which is separated from other such states by a boundary that is more or less permeable. (p. 25)

Admittedly, this formal definition of an ego-state is a bit abstruse for non-psychologists. The following description by Phillips and Frederick (1995) is somewhat more accessible:

Ego states are energies within the greater personality. They are not real people who are simply smaller or younger than the greater personality but rather aspects or energies of the individual. Ego states are adaptational [sic]. They always come to help. Ego states can be thought of as existing on a spectrum from the least to the most differentiated. People with little ego-state distinction appear to be much the same in all situations and may lack color or complexity in their personalities. Another normal situation is one in which the personality is rich with an integrated complexity of ego states that are in communication with one another and which act cooperatively….

Ego state pathology occurs when one or more parts are not in harmony with the others, act on their own, and produce symptoms. Such ego states can be thought of as being walled off from the other, having thicker membranes, or simply not being in cooperative communication with other ego states. The thicker walls are viewed as protective and are frequently associated with [psychological] trauma. (p. 63-64)

Other experts, such as Rowan (1990), have used the term “subpersonality” to express essentially the same notion. According to Rowan, “…a subpersonality is a semi-permanent and semi-autonomous region of the personality capable of acting as a person (p. 8).” For the purposes of the present discussion, the terms ego-state, subpersonality, internal “program,” “part,” “aspect,” and persona are regarded as more or less equivalent. (An “alter” personality, as in the case of true Multiple Personality Disorder, is somewhat different. Alters often have memory barriers “separating” them, so that “alters” often report not knowing what many of the other alters are doing or thinking. An isolated alter would thus exemplify the most extreme end of the ego-state continuum. Normally, ego-states or “parts” are generally more able to share thoughts and feelings with one another than are true alters.)

The Analogy to a Home Computer

To get a better feel for the notion of an ego-state or subpersonality, it may help to make an analogy with a home computer. When you purchase a home computer, you are buying the “hardware.” Once you get it home, you can load different types of “software” or programs onto the hard disk to adapt the computer to your specific needs. Because your computer has multiple programs loaded and ready to run, its functions (that is, what the computer is capable of doing) can change radically with the simple click of a mouse. The same computer that was used for word processing only moments ago may now be presenting a multimedia presentation, searching the World Wide Web, or flashing up a computer game, among many other functions. Each computer must have an overall operating system. This is analogous with the capacity to develop a personality. However, once the operating system is in place and other software has been loaded, a dazzling or confusing array of options becomes available to the computer operator.

If the computer operator inadvertently presses the wrong key or clicks the mouse several times in a row, the display on the computer screen may change unpredictably to a one that is unfamiliar or not user-friendly (e.g., the “blue screen of death”). With a home computer, the user determines which keys get pressed when and which mouse clicks are likely to lead to success and failure. Under emotionally charged circumstances in mediation, however, something unforeseen and unrecognized may abruptly trigger a different “internal program” to load and run in one of the participants, without the conscious awareness or planning on the part of the mediators. When this happens and a participant SNAPs into an essentially different internal program, there may be some overt clues such as a marked change in emotional attitude (e.g., from conciliatory to confrontational, or from euthymic to dysphoric). There may also be noticeable changes in tone of voice and/or facial features. You may observe abrupt insistence on discussing new demands or issues that were not previously “on the table”, a sudden attack on the character of the other participant, or distorted accusations of unfairness and other types of acting out towards the mediators. A sudden tearful apology may be proffered, or an utter capitulation offered up “just to put an end to things.” Instead of a mouse-click, someone’s behavior in session has most likely acted as a trigger for the given participant to shift into a distinctly different state of mind and emotion. When such a sudden shift occurs, the mediation process then inevitably undergoes a shift toward greater complexity as well, often for the worse.

Think Quilts, not Blankets

If you are not taken with the computer analogy, consider quilts. Rather than being cut from a single homogenous piece of cloth (the way most blankets are), quilts are typically composed of various squares. Each quilt square may have different fabrics and patterns selected and sewn during the different epochs of a person’s life. (Often some of the fabrics used to make a quilt square are leftovers from sewing projects that were finished months or even years before.) Some quilts will have intricate hand-stitched elements that can be fully appreciated only when one stands up close. Other quilts may be comprised of a single design that makes up the entire quilt. Such quilts are best appreciated at a distance. Some quilts may have some lovely individual blocks juxtaposed with some hideous ones that just don’t seem to fit together. When we view a complex quilt with many individual elements, our attention may be drawn to just one square at a time; then a different quilt block may leap out at us and temporarily command our attention.

Though variable in composition, each quilt displayed at an exhibit is distinctly recognizable in its own right, as are its component quilt squares. Unless one tugs on a given quilt’s structure with one’s own hands, it is impossible to tell how strong and securely assembled it is. Furthermore, unless one is a virtuoso right-brained quilt-maker, it is extremely difficult to look at the front of a quilt and envision an accurate image of the quilting pattern on its hidden underside. Nor can one possibly be certain, just by looking, how warm or functional a particular quilt is — or with what material it was stuffed long ago.

In short, ego-state theory suggests the personalities of individual people more often tend to be complex, like ornate quilts, rather than uniform, like common blankets. Such complexity, whether recognized or not, is very likely to influence the process of mediation, especially when feelings run high.

How and Why Ego-States Develop

According to the Watkins, ego-states commonly develop for several reasons:

1. As an adaptive response to the requirements of a particular society or culture;

2. Through internalization of parental values and those of other significant figures in a person’s life;

3. To adapt to overwhelming psychological trauma, particularly in early childhood.

Thus, all ego-states are adaptations or unconscious inventions that evolve to cope with significant environmental demands upon the developing individual. Ego-states are described as “well-integrated” when different internal programs work together harmoniously and cooperatively to promote the welfare of the individual person. When the child is confronted with severe, emotional trauma, such as rejection, abuse, and extreme neglect, boundaries between ego-states can become hardened. Later on in life, internal conflicts, personal unpredictability, emotional suffering, psychiatric symptoms, substance abuse, and the failure to develop adequate skills for resolving conflicts with other people can interfere with adaptive functioning in social and vocational domains. An internal solution that was once adaptive for coping with past trauma may become increasingly maladaptive as an individual matures and the challenges and contingencies that govern adult life change. It may be true that “ego-states always come to help,” but the “help” may no longer be appropriate in the adult workplace.

The Mediator Ego-State

Consider the behavioral expectations for professional mediators. During work, mediators are supposed to maintain strict neutrality, yet demonstrate emotional compassion. They are supposed to facilitate creative solutions and yet use a limited number of verbal techniques for doing so. They need to split their focus fairly evenly between what is going on with the parties involved, what the mediator is processing internally, and what is going on with their co-mediator. They typically expend a great deal of emotional and intellectual energy in facilitating a positive outcome for the participants, but are supposed to be just as accepting if no agreement can be reached.

I suggest that the socialization and formal training needed to assume the role of mediator produces, for most of us, a manner of thinking, behaving, and interacting that is noticeably different from the ways in which we are around our spouses and children. Acting in the capacity of a mediator also differs from how we may come across in other aspects of our professional lives (e.g., the practice of courtroom law). As a friend of mine likes to say: “It’s time for me to put on my mediator cap and think like a mediator.” In short, many of us have internalized a “mediation program” or mediator ego-state that we move into consciously for the purpose of promoting a specific outcome — facilitating the resolution of conflict by other people. We are still recognizable as ourselves, but we behave somewhat differently towards others in the mediation sphere of life. In short, being a mediator is something more engaging than “playing a role” but decidedly less severe than having an alter “personality.”

What Sets the Stage for Many People to SNAP

It would seem important for professional mediators to keep in mind that recent data suggest that as many as 30-40 percent of women and 15 percent of men have been sexually abused in childhood, often by members of their own families or so-called “friends of the family.” Media coverage of the recent widespread scandal involving priests of the Catholic Church has helped to bring the issue of child sexual abuse to the forefront of our collective national attention. Even if they have escaped sexual abuse, many children have been emotionally or physically abused and horribly neglected in some way during their formative years.

Childhood trauma can predispose adults to be avoidant of conflict in later life, lest they experience re-traumatization. Alternatively (but less frequently), it can harden children into an in-your-face confrontational attitude during teenage years that persists into adulthood. And sometimes, a chaotic and abusive home environment can lead to alternating between subservience and aggressiveness within the same individual (e.g., “I may be a timid mouse in the workplace, but don’t you dare harm a single hair on my child’s head!”). Any one of these three patterns — lack of appropriate social skills, the presence of overly aggressive social skills, or alternation between avoidance and provocation of conflict — can contribute to a worker being at increased risk for becoming embroiled in on-the-job conflicts. Although I cannot present any data to support my suspicions, I would bet that abused and neglected children grown up are statistically over-represented among the populations of workers who suffer with significant workplace conflicts. This would seem to be even more probable within the last decade, as vocational and educational settings are among the few venues left in daily American life with fixed rules and firm boundaries. As described above, one of the most important factors that leads to the formation of relatively powerful and discordant internal programs or ego-states is having to cope with overwhelming trauma. Through an unconscious mental process called “dissociation” (a type of mental separation of some aspects of experience from others), sub-programs can be formed specifically to deal with the trauma associated with abuse and neglect in childhood. Under the most extreme circumstances, a child may emerge with encapsulated internal states that appear to all the world to be separate alter personalities in case of Multiple Personality Disorder (or Dissociative Identity Disorder, as it now known). Short of having full-blown alters however, most of us develop more or less distinct ego-states to better adapt to specific life circumstances. In the same way, many victims of severe abuse develop highly encapsulated and highly specific internal programs to help them cope with extremely difficult past life circumstances without having full-blown MPD or DID. It is not surprising then that the emotionally charged arena of mediation (with its emphases on “trust,” “truth-telling,” and face-to-face confrontation) may bring one or more sub-personalities to the forefront in unexpected ways.

Who’s Really “At the Table?”

One of the things that mediation trainers emphasize is the importance of bringing all relevant parties “to the table” prior to starting a formal mediation session. In my limited experience with workplace mediations, this usually translates into making sure that a union representative, some subject matter expert, or a support person is present before starting a mediation session. Requesting the involvement of a higher level supervisor (who may have shoe-horned the participants into electing mediation, but who also has a vested interest in the outcome of the process) is another common scenario.

But how does one control (or even know) which relevant subpersonality of an individual participant is going to be present at the table when that participant has distinct ego-states inside himself or herself? The answer is simple: one can’t.

Just as a computer screen generally shows a display generated by the program that is currently running, only one ego-state or part of the personality will be onstage with the mediators at any given time. However, others may be watching from inside (so called “co-consciousness”) or even fighting to come onstage and take charge of the process. It would seem therefore that accepting the internal make-up of the individuals with whom we mediate is part and parcel of the ambiguity we are called upon to tolerate when we practice this curious profession. That having been said, there are some helpful ways of thinking about the sorts of internal “programs” or ego-states that may suddenly become manifest during a mediation session.

A Brief Taxonomy of Ego-States

In an intriguing book called The Mosaic Mind, attorney Regina Goulding and family therapist Richard Schwartz (1995) have discussed at length how traumatized individuals often have a number of “parts” to their personalities. Each part learned or was forced into a particular role to ensure that the child would survive in the abuse environment. We find it useful to organize these roles into three groups – exiles, firefighters, and managers. The three categories of parts are apparent in all people, but the roles become more extreme in abuse survivors. (p. 109)

Exiled, Child-like Parts

According to these theorists, the first category of parts, the “exiles,” is comprised of developmentally younger programs that are psychologically needy and often quite frightened. The exiles feel sadness, as well as fear, more acutely than other parts and thus are “banished” by the other two types of ego-states from overt participation in much of adult life, particularly in the workplace. In an unpredictable fashion, an exiled child-like part may suddenly come to the forefront during a mediation session and tearfully beg the mediators for understanding or protection from the other disputant. It is as if the exile is seeking parental guarantees of safety and protection from any additional perceived abuse at the hands of the more powerful adversary or authority figure (e.g., supervisor).

Remember, most child abuse occurs that the hands of a child’s parents, other family members, or known authority figures (e.g., youth group leaders, religious leaders, family “friends,” etc.). Such betrayals can lead to a generalized, mistrustful attitude which, whether or not it is consciously articulated, runs something like this: If my own people treated me this badly, what can I expect at the hands of other authority figures or strangers? Thus, although some other part, such as a “manager,” may have given formal consent to participate in a mediation by actually signing the Agreement to Mediate form, internal exiled parts may have grave reservations that “trust” in the process is unwarranted based upon past experiences. An internal exile is likely to view the mediators as “authority figures” no matter how gently the mediators come across during the opening phase of the process. An exiled part may quickly come to the fore and seek to derail the process by behaving in infantile fashion or using childish, passive-aggressive manipulations. Because other types of parts tend to cope with distressing adult emotions by “dumping” their strong feelings onto already emotionally-overloaded exiled parts, an outpouring of tears and distress (prompted by issues of re-victimization in the workplace or powerlessness in the face of bullying) may also complicate the process. At one mediation session I observed, a very tough-minded woman suddenly grew wide-eyed in a quite innocent and naïve way and asked the mediators: “Are you saying that I am bad?” She seemed shocked at learning from her manager that many of her co-workers had actually been afraid of her during outbursts in the workplace. Despite sincere reassurances that no such judgment (i.e., You are bad) had been intended or made, this internal proposition, once triggered inside the participant, was so threatening that wave after wave of sobbing ensued, necessitating caucuses and breaks before productive problem-solving could resume.


According to Gould and Schwartz, so-called “firefighter” parts are active, energetic, and assertive on behalf of the individual. For persons with military or para-military careers (e.g., police), the “warrior” label may be more apt than the firefighter moniker. In general, this type of part tends to be dormant, listening intently for alarms or threats to self-esteem. Then they jump out “to put out the fire” (and sometimes even to threaten the person who was perceived as setting the fire). In extreme instances, firefighters may act like “tyrants or bullies,” concealing fear by attempting to squash psychological threats to the individual. Whereas child “exiles” are too scared to act in the face of conflict, and “managers” are reluctant to act, firefighters do not hesitate to take action and may even relish an opportunity to “strut their stuff.”

Firefighters share in common with “managers” that both types of parts wish to avoid exposing the individual’s emotional vulnerability in the workplace. However, managers will at least superficially comply with basic workplace expectations. While a managerial aspect of the person may sign the consent form, a firefighter will monitor the course of the mediation session for signs of a conflagration that could overwhelm an “exiled’ part and cause the “manager” to lose focus, at least temporarily. If such a threat to the participant’s self-esteem is detected, the firefighter may emerge to forestall the session turning into a four-alarm blaze. Putting the fire out one way or the other (i.e., stopping the perceived attack upon the individual), thus becomes the new number one priority rather than trying to understand the other participant in order to reach a mutually acceptable solution.

An angry firefighter may vigorously deny any sign of weakness, may counter-attack the other disputant verbally, may simply get up and leave without saying a word, or may aggressively berate the mediators in paranoid fashion for favoring the other disputant. If the mediators intervene to curtail one obstructive behavior, then the firefighter part may quickly find another destructive and distracting behavior as a substitute. An enraged firefighter may exercise what I call the Power of Negative Control (i.e., there’s no situation I cannot attempt to control by making it worse) and drive the other party (or the mediators) to terminate the session as unproductive by making other parties extremely uncomfortable emotionally.

Occasionally, a firefighter part may be held in check by a “manager” during the mediation session long enough for a Mediation Agreement to be reached. Then the firefighter may emerge and balk at signing the agreement to which the manager had given his or her consent a few minutes earlier. Or after an agreement has been signed, the firefighter may panic that the manager has been “had” and resist complying with one or more provisions of the agreement. Alternatively, such a bitter firefighter part may go back to the workgroup and disclose or distort the details of the process or the agreement in an attempt to make the other disputant (or the mediators) look bad in the eyes of some larger group.


These parts, also labeled “executives” or “internal self-helpers” by some psychologists, generally permit the person to function more or less satisfactorily on the job and in the world at large. In the words of Gould and Schwartz (1995), A managerial part goes to work, eats nutritious meals, drives the children to school on time, gets a haircut, and attends church committee meetings. Some parts acting in a managerial function appear very efficient, competent, smart, and in control. Many people are quite comfortable with their managers. Managerial parts are very likable, quick, and funny. (p.112)

The basic function of a manager is to keep the individual’s life running as normally as possible in the eyes of the external world. They are often driven to maintain a high enough achievement level so that other people will not look deeply into the inner emotional life of the individual. In general, managers believe that it is important to please other people in order to be accepted in the workplace. However, problems arise when others in the workplace become jealous of the individual’s accomplishments or actively begin to bully the individual in ways that are experienced as re-victimization. Manager parts can become exhausted when too much is demanded of them emotionally and their super-rational and achievement-oriented defenses against experiencing emotional pain from the past are breached by the misbehavior of other co-workers. It is as if manager parts insist on being the guardians of social normalcy, even at the risk of trying to avoid or deny all overt conflict.

To promote job security, a manager part is likely to accede reluctantly to a request for mediation from someone else (especially if it comes from a superior). After signing the consent form, a manager part will try to use the individual’s intellect to avoid any disclosure or close examination of the individual’s emotional needs. Although feelings of persecution are unpleasant, they are at least familiar in that they fit the template laid down by prior abuse in childhood. Overtly expressing (or even admitting to) anger may feel even more threatening to a manager part than tolerating prolonged mistreatment at the hands of another. A variety of maneuvers may be employed, both consciously and unconsciously, to avoid letting the mediation session get down to the emotional “heart of the matter.” Denial/minimization, people-pleasing and care-taking of others, avoiding risks, and saving public face at great cost – all of these are self-protective strategies that “manager” parts can use to protect their personality systems and to avoid public displays of intense affect. An internal exiled part may be cringing with apprehension about the workplace conflict; an internal firefighter may be champing at the bit to take action to end the conflict (at least in the short-term). Meanwhile, the manager part may be fighting to stay in control of the entire personality system during the mediation session long enough to comply with the superficial, procedural requirements of the process.

Some Suggested Guidelines for Things You Can Think About Doing

Keeping in mind the potential implications of ego-state theory for the practice of mediation, some practical suggestions are presented for using this psychological model to enhance in-session understanding and facilitation of complex situations.

1. First off, remain aware that most individuals appear to have distinct parts or internal programs within the context of their overall personalities. Accepting that this state of affairs is not at all unusual provides mediators with a better short-term model of real human behavior under stress than do trait models of human behavior. Trait models assume that we all have one basic personality that is more or less consistent across a wide variety of situations. Radical shifts in ongoing behavior are therefore attributed to strong moods or psychopathology. Ego-state or other types of part/whole models of human behavior assume that individuals are capable of running a wide variety of different programs and that the distinct programs (or parts) with very different functions can load themselves very quickly, sometimes in the blink of an eye.

2. The development of rather fixed internal parts or programs is often associated with a history of child abuse or other past trauma. Traumatic memories may condition one or more participants in mediation to mistrust authority and be extremely uncomfortable. A trauma history may also cause a participant to link the current work situation to horrific events in his or her past, either consciously or unconsciously. When this happens, then the resolution of a seemingly simple work dispute may become a “life and death” matter for that participant. Mediation may seem subjectively more like an ordeal that threatens to rip apart a delicately balanced internal personality system, than a benign forum for airing grievances and resolving conflicts.

3. Mediation is not therapy. In general, try to do the minimum exploration of strong feelings necessary to permit graceful transition to the problem-solving phase, if you suspect you are dealing with an individual with a strongly formed internal system. Such an individual may become extremely uncomfortable with in-depth exploration of personal feelings and feel even more vulnerable than a less complex and better- integrated mediation participant, while disclosing personal information in front of total strangers and adversaries. The goal of therapy is to promote understanding to facilitate the mental health of the client. The goal of mediation is to facilitate both parties reaching a mutually acceptable agreement in the workplace. Thus, commonsense dictates that exploration of the personal issues of either one of the disputants should be kept to the bare minimum necessary to reach an agreement. If mediators begin to focus on the internal dynamics of either participant in great detail, this can actually be counter-productive for the process and goals of the mediation session. Most psychologists, let alone non-psychologist mediators, are not prepared to help a stranger draw an accurate “internal map” of himself or herself during a 15 minute individual caucus. Working with a co-mediator can help to ensure that mediation is taking place and not individual therapy. Obviously, if a participant begins to decompensate or discloses suicidal or aggressive intent to the mediators, the mediation should be stopped. It is the responsibility of the mediator the support the participant and the agency or corporation is maintaining everyone’s safety and finding appropriate clinical assistance in the interest of basic human welfare.

4. Relax. You cannot know ahead of time to what degree the individual participants in a mediation session will be well-integrated in terms of their personality functioning (as opposed to having strongly formed and conflicting internal programs). Most people believe that human beings have one basic personality, and that belief includes themselves. Thus, if an individual mediation participant is not consciously aware of “having parts,” how can the mediators possible know ahead of time what to anticipate with regard to that person’s internal personality organization? Nevertheless, there are some behavioral clues that can suggest when a mediation participant’s programs are changing internally or different “parts” are in conflict:

  • The person refers to himself or herself as “We” with no obvious referent (i.e., it is not clearly the bloated, royal “We” that some managers use to represent the company or agency, nor is it clearly the “We” that would clearly refer to a department or sub-unit in the workplace). It may, under certain circumstances, be appropriate or helpful to simply ask who the “We” is that the participant has mentioned. However, if the initial explanation is unclear, then you may be better off letting the matter drop.

  • When a history of frequent self-defeating behavior is revealed, it becomes more probable that one may be dealing a personality system with conflicting internal parts. In the words of the Watkins (1997): “Factions within a nation compete for power. Nations war with one another. Corporations engage in economic battles…Within an individual a similar pattern can tale place.” Scarring on the wrists, self-administered tattoos, scars from cigarette burns, problems with substance abuse, self-defeating exploitations by sexual partners, and being a “frequent filer” (i.e., multiple grievances) — these factors increase the likelihood that you are dealing with a person who has a background of significant abuse or neglect. However, you should also remember that having distinct parts to one’s personality make-up does not necessarily mean that a given individual has been abused.
  • Dramatic and puzzling shifts in emotion and overt behavior that seem to “come out of nowhere” may actually be new “parts” or programs coming onstage. Do not assume that what was discussed or agreed upon 5 minutes ago is necessarily the major issue of importance to that part right now. The only thing that you can safely assume is that “all parts or some parts have been listening” to previous storytelling and negotiation. If there appears to be sudden, inexplicable puzzlement on the part of a participant, then the use of summarizing as a technique may be quite valuable in getting that “part” up to speed.
  • If a participant begins to pump an arm or a leg in unilateral fashion or rock in place, this can be an ominous sign that the person may be waiting to explode (i.e., using flash anger to cover over inner fears or conflicts). Don’t be surprised if a “firefighter” persona emerges shortly thereafter.

  • In response to a simple question, if a participant needs to close his or her eyes and “go inside” silently to retrieve information, you may be dealing with a person that has a strongly defined inner system. This may mean that a manager part was not “onstage” when the incident occurred or is somehow lacking the capacity to retrieve crucial information and needs to “dial in” to another part of the personality system. Similarly, if a participant seems to be staring off blankly for a period of time in an oblivious fashion, this may be a sign that the participant is actively dissociating during a portion of the mediation to maintain emotional numbing in the face of stress. Calling a side caucus and gently exploring this may be fruitful. “It seems like it is harder for you to focus at some times rather than others?” is an innocuous way to explore how much of what is actually going on is, in fact, actually being processed in the mind of the participant.

  • Firefighters, exiled children, and managers do not exhaust the list of possible internal parts. For example, psychiatrist Eric Berne, the developer of Transactional Analysis, encouraged therapy clients to consider their Child, Adult, and Parent ego-states. Other methods for examining internal personality systems abound (e.g. Polster, 1995; Rowan, 1995). In my experiences working with military veterans, I routinely find that there is a “warrior program” inside men and women who have spent time in the armed forces. If it is known that a mediation participant has had prior military experience (particularly in combat), there is likely to be a “warrior” program inside the individual instead of, or in addition, to a firefighter part. It is the nature of the warrior that he or she is prepared to “go to war” over certain issues or positions that are viewed as non-negotiable. After all, like it or not, going to war is a means of conflict resolution. Thus, acknowledging the “warrior” for his or her integrity and determination (possibly in a private caucus) may prove necessary before a military veteran can accept some sort of reframe. Only then may a dearly held interest become negotiable. An entrenched warrior will often prefer to inflict pain or fight to the death rather than be made to look weak or lose face. The “warrior program” is a stellar example of a relatively permanent program that did not develop in response to an abusive childhood, but rather as an outcome of a “normal” adult experience, namely, military service.

  • As a theoretically neutral mediator, if you begin to sense that a participant is trying to “make you feel guilty,” responsible, or sorry for their “victim-hood,” that feeling could be one sign that the participant has been victimized in the past. Any history of serious or repeated victimization at the hands of another person, especially early in life, makes it more likely that different adaptive programs have evolved to help the individual cope with traumatic circumstances. Remember to “own the process” — but not the participants’ problems. Trauma survivors are often tougher and more resilient in certain ways than non-traumatized individuals.

  • If, when you look deeply into a participant’s eye and there appears to be “someone else” looking back out at you, there may well be another persona (or other personae) inside.

5. Assume that only one “part” of the individual’s personality system can be onstage at any given time. Other parts are likely to have co-consciousness of what is transpiring within a session. However, generally speaking, only one part can be interacting with you at a time, in much the same way that a computer screen only shows one operative program at a time. All you can assume about the parts that are offstage is that “no one may be listening or everyone may be listening.”

6. Until proven otherwise, assume that you are dealing with a “manager” persona. However, you must be aware that other parts are probably eavesdropping. Thus, censuring one disputant for “acting childishly” could have repercussions for the exiled “inner child” of the other disputant.

7. Accept the fact that you cannot tell what is going to happen next. You can only control your responses to what is going on within the session.

8. Unless someone is being threatened or there are clear safety issues, stay calm.

9. Assume that the individuals who are the participants in the mediation are at least as internally complex as you are and, in all probability more complex. Among mental health professionals (and “professional” people in general), I have often sensed an unspoken and unrecognized bias toward assuming that the “client” to whom services are being vended is psychologically less complex than the professional. However, after spending many hours with previously traumatized individuals, I have been impressed that the opposite is often true. Frequently, people who have attempted to survive and surmount serious problems in living have developed ingenious and adaptive methods for sustaining themselves.

10. It may facilitate communicating with the fractionated participant if you occasionally use what I call the “language of parts” — for example, “It seems like part of you feels one way, and part of you may feel differently.” Emphasis on the benefits of teamwork can also be a helpful metaphor. Seemingly bland reassurance, such as “It’s been my experience that most everyone has inner resources which can be called upon to deal with difficult issues/times/duties,” may also be surprisingly effective. Using the “language of parts” should be done subtly and naturally; it is generally a big mistake to ask directly if a person has definite internal parts – this can be distressing and disorganizing for the individual. As stated above, mediation is not therapy.

11. Similarly, the “language of choice” (i.e., “you could do this or that”) versus issuing imperatives (i.e., “you really should consider doing this”) is generally less likely to mobilize defensive maneuvers. Such permissive phrasing, makes it less likely that the mediators will be distorted as fearsome or controlling authority figures.

12. Do not invite an individual participant to disclose information about “parts” in a session or a private caucus. To do so only invites trouble. You do not typically share with other human beings everything that you know or sense that you know about them, so don’t do so in mediation. Think complexly about complex participants in your private conceptualizations, but interact with complex individuals with your customary interpersonal style within the sessions themselves.

13. If there is a “little voice” inside your head that is worried that you may be prematurely rushing to achieve an agreement, trust that intuition. Conducting “reality testing” about the feasibility of implementing the proposed agreement in the real world is time well spent. If the participants deny that there are any impediments toward signing but the mediators still feel uncomfortable, it can be helpful to ask the participants individually in side caucuses to imagine having already signed the agreement; then request that they scan their bodies for any traces of discomfort or tension. The somatic aspects of emotional experience often function much like a lie detector, or, in this case, a conflict detector. If a participant notes distress somewhere in his or her body when imagining the agreement in place, this can be explored individually. A helpful question is simply, “What do you think that means?” Another helpful strategy can be asking “Do you recall some other situation in your life when your body felt the way it does right now? What was going on at that time?” Unless the participant is purposely deceiving you (e.g., just to get the session over as quickly as possible), the absence of distress in the participant’s body scan is a positive sign that there are not huge, inner reservations about the agreement that has been reached. Counter-intuitively, leaving the disputants to “sleep on it” and ponder overnight whether they wish to sign an agreement that has been reached also gives the participants time for reservations or conflicts to surface. Curb your enthusiasm to “close the deal” at any cost. A prematurely signed agreement can sometimes be worse that no agreement at all for the individuals or for the larger work unit. Furthermore, a prematurely signed agreement that later hemorrhages hostility and additional conflict can be devastating for the reputation of a fledgling, in-house mediation program, especially if the mediator pool is internal to the workplace.

14. Rather than giving up on the entire mediation process, it may be helpful to take a break during an ongoing mediation session and, in a private caucus, encourage the acutely troubled participant to avail himself or herself of the agency’s or corporation’s EAP program. Later on, mediation might be resumed at a second session. Also, it may be appropriate to remind a less powerful individual that they are entitled to have a trusted support person present (to help participant stay grounded as stressful topics are discussed) at a later session.

15. Just as two different participants are likely to have divergent underlying interests, different parts of the personality system are likely to have different primary interests and issues. Safety, professional advancement, emotional vulnerability, righteous adherence to personal integrity, and other concerns or values may not be equally shared across all sub-programs of a personality system. Thus, it may be necessary to work on building an “inner consensus” during private caucus time by acknowledging the complexity of certain issues. Teammates do not have to be of one mind to play well together, but everyone’s contribution needs to be acknowledged and respected. Thus, an individual caucus session with a person with a complex inner system may seem a bit like conducting an “inner mediation,” with the various parts one senses needing soothing and support. What threatens one part may be embraced by another part. For example, losing one type of emotional control in public (e.g., weeping) may bring relief to a child-like part, embarrassment to a manager part, and disgust to a firefighter. On the other hand, the overt expression or anger in a mediation session could be relished by a warrior part, scary to an exiled child part, and extremely embarrassing for a manager part. Overtly acknowledging mixed feelings and mixed motivations can be a helpful first step in getting the individual to clarify outcomes that an entire personality system can “live with.” Because mediation is not therapy, you are better off not overtly introduced the notion of “ego-states” or subpersonalities into the arena of mediation, even in a private caucus.

16. When you are dealing with two participants, one of which clearly seems to run different programs and the other who seems much more fully integrated, first caucus with the person you suspect of having the more clearly defined (or actively conflicting) sub-programs. The reason for this is to decrease anxiety of the less –well-integrated party. However, there may well be compelling reasons to ignore this guideline, depending upon how the course of the session is going.

17. Use humor sparingly and judiciously. When in doubt, forego making the joke, particularly at the expense of either participant. A humorless manager may take the intended joke literally, a child-like part may not understand, or a firefighter part may take offense or sense danger when none was intended.

18. If you wish to deliver a message to another “part” that you suspect is inside, the indirect technique of “talking through” can be helpful. In “talking through,” the goal of the communication is to talk through the manifest part to some inner part(s). Superficially, the mediator appears to be addressing the “part” that is currently participating. However, the mediator is also talking through that persona to parts presumed to be inside. For example, this technique can be of value in reassuring a frightened child-like part, tacitly acknowledging the power of a firefighter to protect the individual, or inviting a manager to return to the table if that part has fled the scene. When attempting to “talk through,” it is helpful to make gentle but direct eye contact with the individual participant in a deep way, taking pains not to glare at the person. After you are done speaking, then gradually but casually shift your gaze elsewhere. Just because you have tried to send some sort of “message,” there is no guarantee that the message has been attended to or correctly decoded. (This is true of any verbal message and not just for “talking through.”) Usually, if the message has been apprehended as intended, there will be some extremely brief acknowledgement in “micro-momentary” changes in the facial musculature around the eyes. Another way to gauge the impact of an intended message is to watch for acute, positive changes in the participant’s affect and demeanor right after the message has been delivered. While helpful at crucial junctures, this technique should be used sparingly, and, as always, roughly equal attention should be paid to both sides in a mediation session.

19. Fail creatively and blamelessly. Remind yourself that the only true failures occur when mediators don’t free themselves up to do their best on behalf of their clients. Imagine a very simple mediation scenario with two participants and two co-mediators. Furthermore, assume for the moment that both mediators have extremely well-integrated personality functioning – for all intents and purposes, any parts or programs that are inside the individual mediators fit together so well that the mediators’ personalities appear to be seamless. Now imagine that each of the participants has a manager, exiled child, and firefighter program inside. The number of potential participants at this hypothetical mediation has just increased from four to ten. If many human beings are as complex as I am suggesting is the case, isn’t it extraordinary that mediation generally succeeds about 70-80 % of the time? Now consider group mediations: the likelihood of multiple programs with diverse and competing agendas increases almost exponentially. When mediation does not succeed, an important explanatory variable may be the intra-psychic complexity of one or both participants.

20. Don’t be afraid to think about your own life in terms of inner parts or sub-programs. In the words of renowned family therapist Virginia Satir (1978):

…you probably have many parts that you have not yet discovered. All of these parts, whether you have owned them or not, are present in you. Becoming aware of them enables you to take charge of them rather than to be enslaved by them. Each of your parts is a vital source of energy. Each has many uses, and can harmonize with many other parts in ways to add even more energy. (p. 63)

Very few of us are truly one-dimensional. If you have been neglecting aspects of yourself, try to find ways to revive them, particularly the family-oriented, affiliative, spiritual, creative, recreational, and physical aspects of your being. All these facets of existence can and do refresh you. On their death-beds, few mediators have actually wished that they had spent more time at the office.

Personal Responsibility

The proposition that personality is comprised of a number of “parts” in no way diminishes the principle that each human being is responsible for his or her behavior. Regardless of how one conceptualizes the inner complexities of being human, we all engage in “self-management.” We all need to acknowledge responsibility for what we act, what we think, and how we feel. The excuse that “my rogue ego-state made me do it” is no more valid for medical-legal purposes that the late Flip Wilson’s lament: “the Devil made me do it.” (Actually, it was the “Geraldine” part of Wilson’s routine that said this.)

No matter how complexly they are internally configured, all mediation participants should be held responsible for working in good faith to resolve workplace conflicts and for honoring signed agreements. Similarly, all mediators are responsible for continuing to develop personal awareness of their own inner dynamics and monitoring how these dynamics inevitably influence process and outcome. To paraphrase the Watkins’ (1997), it is important what we say to clients, but is also important how we are with clients. As is true for therapists, good mediators try to “meet their clients were they are at,” prior to helping them move to different and hopefully less contentious ground.

I hope that the reflections and guidelines presented in this article will influence some mediators to think more creatively and to feel more comfortable when some mediation sessions start to go awry.


Binet, A. (1892). Les alterations de al personalite. Paris: Alcan.

Goulding, R.A., & Schwartz, R.C (1995). The mosaic mind. New York: Norton.

Lewin, K. (1936). Topographical psychology. New York: McGraw-Hill.

Phillips, M., & Frederick, C. (1995). Healing the divided self. New York: Norton.

Polster, E. (1995). A population of selves. San Franascisco: Jossey-Bass.

Quenk, N.L. (1993). Beside ourselves: our hidden personality in everyday life. Palo Alto, CA.: Davies-Black,.

Rowan, J. (1990). Subpersonalities: the people inside us. London: Routledge.

Satir, V. (1978). Your many faces: the first step to being loved. Berkeley, CA: Celestial Arts. Watkins, J.G., & Watkins, H.H. (1997). Ego states: theory and therapy. New York: Norton.


Richard J. Roberts

Dick Roberts, Ph.D. is a certified mediator who formerly participated in the Alternative Dispute Resolution program at the Iowa City DVA Medical Center.  He was initially trained as a clinical psychologist, psychophysiologist, and clinical neuropsychologist.  Although he no longer mediates due to the demands of clinical work, he remains actively… MORE >

Featured Mediators

View all

Read these next


Conflict & Consciousness

Who is Determining Your Self-Worth? I’m publishing this newsletter series to discuss the nuances of the four timeless topics for people and organizations: conflict, change, transformation, and consciousness. Please subscribe...

By Ralph Kilmann

Interview with Albie Davis – Views from the Eye of the Storm

This is the complete interview by Robert Benjamin with Albie Davis, a pioneering community mediation leader, filmed as part of Mediate.com's "The Mediators: Views from the Eye of the Storm"...

By Albie Davis

Five Tips to Let Go and Forgive after a Break-Up

Conflict Remedy Blog by Lorraine SegalLearning how to let go and forgive helped me a lot after two “break-ups.” Neither of these was with a spouse or romantic partner, but...

By Lorraine Segal

Find a Mediator