Depression in Couples: Literature Review and a Metaframeworks Approach

[First published on Nathen’s Miraculous Escape, January 7, 2011.]

Nathen B. Lester

University of Oregon


This paper a brief overview of the evidence and discourse about depression, focusing on its occurrence, dynamics, and treatment in couple relationships.  This is followed by an introduction to Breunlin, Schwartz, and Mac Kune-Karrer’s metaframeworks model for therapy and its possible application in the treatment of depression in couples.

Keywords: depression, couples, couples therapy, metaframeworks

Depression in Couples:

Literature Review and a Metaframeworks Approach

In the last few decades, depression has become an obsession in the US.  We are flooded with ideas about depression from television shows, magazine articles, self-help books, celebrity therapists, theoretical articles, feminist treatises, pharmaceutical advertisements and press releases, research results from biological, clinical, epidemiological, psychological, and sociological perspectives, plus critiques and meta-analyses of that research.  This obsession is not surprising, considering the medical, mortality, and employment costs of the condition are well above $80 billion per year and climbing (Greenburg, Kessler, Birnbaum, Leong, Berglund, & Corey-Lisle, 2003), considering the evidence that prevalence rates are increasing and age of onset is decreasing with each generational cohort (Craighead, 2007), and considering what an unpleasant and dangerous condition it is.  The World Health Organization has predicted that in the next 15 years depression will move from the fourth most disabling disease in the world to the second, behind only obesity/diabetes (Sapolsky, 2009).

In my practice as a couples and family therapist in training, depression is just as common a complaint as the ubiquitous “communication problems,” meaning at least one partner in every couple I have seen so far has considered themselves clinically depressed.  This paper represents my attempt to delve into the data and conversation about depression, how it is treated in couples, and to approach it using the perspective of metaframeworks, as put forth in Breunlin, Schwartz, and Mac Kune-Karrer’s (1997) Metaframeworks: Transcending the Models of Family Therapy.


What Do We Know About Depression?

We know how we define it. In the mental health profession, if someone “has depression,” we mean that they meet the criteria in the current edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) for Major Depressive Disorder.  That means the daily presence of a depressed mood, anhedonia, and at least three of the following six symptoms for at least two weeks: (a) weight loss or gain of at least 5%, (b) insomnia or hypersomnia, (c) observable agitation or slowed movement, (d) fatigue, (e) feelings of guilt or worthlessness, (f) subjectively or objectively decreased cognitive ability, and (g) suicidality. Additionally, these symptoms must  impair functioning in some significant way, must not be the result of an illness or drug, and must not be the result of bereavement—that is, must not occur only within two months of the loss of a loved one (American Psychiatric Association [DSM-IV-TR], 2000).

Additionally, we define depression in terms of the number of depression-indicating responses to surveys, like the Hamilton Rating Scale for Depression (HRSD) (Hamilton, 1980) and the Beck Depression Inventory (BDI-II) (Beck, Steer, Ball, & Ranieri, 1996).  The HRSD and BDI have participants rate symptoms, mostly DSM-IV-TR criteria, on Likert scales between none and severe.  They produce numbers which are compared to predefined cutoffs for no, mild, moderate, severe, and (for the HRSD) very severe depression.

We know quite a few correlates. Depression is associated with intense affect, especially sadness and anxiety, as well as physical pain, addiction, suicide, and social problems such as divorce, unemployment, and truancy (DSM-IV-TR, 2000).  It is also correlated with internal attributions of negative events and external attributions of positive events (Pinto & Francis, 1993).  Of the 17% of US residents who qualify as depressed at some point in their lives, between 10.5 and 18 million people at any one time, two-thirds are female (Craighead, Sheets, Brosse, & Ilardi, 2007), though the DSM-IV-TR reports that rates are even between boys and girls before puberty. The DSM-IV-TR also claims that there are no difference between those in different ethnic, socioeconomic, educational, or marital categories (DSM-IV-TR, 2000), though there is some evidence for a correlation with poverty (Kim, 2008; Vilhjalmsson, 1993) and menial labor (Bonde, 2008).

Major childhood stressors such as physical, emotional, and sexual abuse, witnessing domestic violence, separation, or divorce, or drug abuse, criminal activity, mental illness, or suicide in the household significantly increase rates of depression in adults (Anda, Whitfield, Felitti, Chapman, Edwards, Dube, & Williamson, 2002; Chapman, Anda, Felitti, Dube, Edwards, & Whitfield, 2004; Edwards, Holden, Anda, & Felitti, 2003).  Major stressors in adulthood, especially those related to social rejection, can also precipitate depression (Kendler, 2003; Kessler, 1997; Slavich, Thornton, Torres, Monroe, & Gotlib, 2009).  The highly religious appear to be more prone to depression, though more resilient against it once it occurs (Dein, 2006; McCullough & Larson, 1999).  Depression is more common among  people with major medical conditions (DSM-IV-TR, 2000).

There are no biological tests for depression (DSM-IV-TR, 2000).  There do appear to be a few biological correlates, however, including atrophy of the hypothalamus (Patterson, Albala, McCahill, & Edwards, 2010), increased stress-hormone levels, decreased growth-hormone levels (DSM-IV-TR, 2000), and a short allele of the  5-HTT gene (Caspi, Sugden, Moffitt, Taylor, Craig, Harrington, McClay, Mill, Martin, Braithwaite, & Poulton, 2003).  Being related to a person with depression increases your chances of becoming depressed as well, which is suggestive of a biological correlate (DSM-IV-TR, 2000).  Each of these factors correlate to a significant degree with incidence of depression, but none to the extent that they are candidates for a cause.

By far the strongest biological correlation with depression we know of so far are abnormalities in electroencephalogram (EEG) readings during sleep.  The DSM-IV-TR reports a correlation of 40-60% in outpatient groups and up to 90% for inpatients (2000).  Also, complete sleep deprivation has been shown to entirely alleviate symptoms of depression in over half of participants (Wu & Bunney, 1990).

Another category of correlates is treatment outcomes.  These fall into three general categories–therapeutic, somatic, and placebo–most of which are basically equivalent in their efficacy, which is to say they seem to have a lasting positive effect somewhere around half to two-thirds of the time (Patterson et al., 2010).  The evidence that we have about therapeutic treatment comes largely from clinical tests of variants of cognitive and behavioral therapies, and finds these therapies to be effective to the same degree (Wampold, 2001).  In somatic treatments depressed people take medication, have their brains shocked, have their sleep interfered with, or, in the most extreme cases, their brains operated on.  In a placebo treatment depressed people eat sugar pills which they believe might be medication.  There is evidence that some medications work significantly better than placebos for those with severe or very severe depression (Fournier, DeRubeis, Hollon, Dimidjian, Shelton, & Fawcett, 2010).  There is evidence that reducing REM sleep over time can improve symptoms of depression to the same or somewhat better than a placebo (Rieman, Berger, & Voderholzer, 2001), and there is evidence that electroconvulsive therapy (ECT) is the most effective treatment of all, at approximately 80% response rate (Patterson et al., 2010).  There is also evidence that stereotaxic anterior cingulotomy reduces the symptoms of depression (Ballentine, Bouckoms, Thomas, & Giriunas, 1987; Sapolsky, 2009; though see Cohen, Paul, Zawacki, Moser, Sweet, & Wilkinson, 2001).

What Do We Think About Depression and How Do We Treat It?

How we think about and treat depression is shaped greatly by our theoretical perspective; cognitive therapists emphasize the role of thinking  and psychiatrists emphasize the role of neurobiology.  In the face of the evidence, however, very few theorists or clinicians still believe there is a single, causative factor for depression.  Most, instead, believe some version of the diathesis-stress model, which holds that depression is the result of an interaction between biological factors, usually genetic and/or epigenetic, and some form of environmental stress, such as trauma or loss (Monroe & Simons, 1991; Sapolsky, 2009).

That said, each school of thought does think about and treat depression differently.  Psychiatrists and other medical doctors, for example, tend to emphasize the diathesis side, and see depression as primarily a more-or-less mysterious biological disease, caused by a disregulation of neurotransmitters or other hormones, or some other wiring or firing malfunction of our neural networks.  These clinicians prescribe somatic therapies based on their theoretical training.  The vast majority of patients receive medication that is thought to increase the effects of seratonin on the brain (SSRIs), and most of the rest receive one of a couple older classes of medications, thought to act on norepinepherin or on all monoamine neurotransmitters (Olfson, Marcus, Druss, Elinson, Tanielian, & Pincus, 2002; Unutzer, Katon, Callahan, Willians, Hunkeler, Harpole, Hoffing, Penna, Noel, Lin, Tang, & Oishi, 2003).  Somewhere between 8 and 60% receive psychotherapy, depending on their age (older patients get less therapy) (Olfson et al, 2002; Unutzer et al., 2003).  By comparison, other somatic treatments, such as electroconvulsive therapy, vagal-nerve stimulation, REM-deprivation therapy, light therapy, and changes in diet or exercise, are prescribed for only a tiny fraction of patients.

Those operating from the standpoint of a psychotherapeutic modality, on the other hand, tend to focus on the stress side of the diathesis-stress model, and each model has its favorite stressor.  Cognitive therapists, for example, see depression as the result of extreme, inaccurate, and pessimistic thoughts and beliefs, and treat it by helping clients recognize and counteract these distortions (e.g. Beck, 1976; Ellis, 1998).  Behavioral therapists see depression as essentially as a bad habit resulting from environmental rewards for depressive behavior and punishments for non-depressive behaviors, which are best treated by reversing that reinforcement system (e.g. Kanter, Cautilli, Busch, & Baruch, 2005).  The many, many varieties of psychodynamic therapy all have in common the beliefs that problems come from clients’ internal, unconscious conflicts, and that they are best worked out in a conversation with a therapist that leads to insight.  Different schools of psychodynamic therapy each have different emphasis, however, that  affect how they see and treat depression.  A psychoanalyst might see depression as the result of early, formative experiences that caused the client to turn their anger inward into guilt, and ask the client to free-associate and talk about dreams over a period of years, offering periodic interpretations.  Humanistic psychodynamic models such as Rogers’ client-centered therapy, Perls’ Gestalt therapy, and Yalom’s existential therapy focus more on how clients heal, in the present moment, in the presence of a therapist who is behaving authentically, than on how they become troubled.  A client-centered therapist believes that depression is best ameliorated by the client coming to trust the unconditional positive regard of the therapist (e.g. Rogers, 2003).  A Gestalt therapist believes that they can help by challenging the client to behave with complete authenticity (e.g. Perls, 1973).  An existential therapist believes that genuinely coming to grips with the reality and inevitability of suffering and death are the most helpful (e.g. Yalom, 2009).  One psychodynamic approach–psychodynamic-interpersonal therapy–was designed to treat depression.  Like other psychodynamic approaches, it focuses on emotions, interpersonal experiences, and the therapeutic relationship–but it also incorporates Bowlby’s attachment theory and typically uses few enough sessions to be called “brief therapy” (e.g. Klerman, Weissman, Rounsaville, Chevron, & Rounsaville, 1984).

Couples and family therapists have many of their own systemic views of the stress side of the diathesis-stress model.  Confronted with a client complaining of depression, an MRI-style brief therapist, for example, would first look for problematic, outdated rules that the family system is acting under, in the form of their attempted solution for the problem.  They would design an intervention that represented as much as possible the opposite of the family’s attempted solution, perhaps by prescribing depressive behavior (e.g. Fisch, Weakland, & Segal, 1982).  A structural family therapist would imagine that depression is the result of inappropriate power alliances between, for example, a parent and child against a second parent.  They would design a behavioral intervention to realign the family hierarchy, such as advising the parents to go out on a secret date (e.g. Minuchin & Fishman, 1981).  A strategic therapist might also see depression as the result of outdated rules or inappropriate power alliances, but also looks for how the depression might be serving a function for the family system—the interpersonal payoff.  If the latter is suspected, a strategic therapist might prescribe an Ericksonian ordeal, intended to make the problematic behavior more difficult to maintain than non-problematic behavior (e.g. Haley, 1984).

An experiential family therapist would find the depressogenic stress in a client’s low self-esteem, and attempt to remedy it by modeling and by leading communication exercises designed to increase authentic communication (e.g. Satir, 1983).  For a Bowenian intergenerational family therapist, stress is the result of a lack of differentiation, meaning anxiety and rigidity of behavior in response to external or internal emotional pressures.  To help increase a client’s differentiation, a Bowenian therapist explores the quality of extended-family relationships and interrupts emotional triangulation, while maintaining their own differentiation (e.g. Nichols & Schwartz, 2008).

A feminist family therapist looks for and tries to resolve stress in the form of gender- or culture-based power differentials (e.g. Leupnitz, 1988).  A solution-focused family therapist believes that stress is the result of a focus on problems, and attempts to alleviate it by focusing on and enhancing exceptions to the problems (e.g. deShazer, 1988).  A narrative family therapist sees stress as the result of clients having internalized a negative discourse about themselves from the dominant culture, and attempts to alleviate it by finding examples in the client’s life that are contradictory to that discourse and re-author a new story.  Narrative therapists also rely on a technique called externalization, where a client’s problem is given a name and the client comes to see it as not part of their self (e.g. White & Epston, 1990).

Emotionally-focused couple therapy (EFT) finds stress is in unmet attachment needs for safety, proximity, and connection.  Therapy consists of an emotional negotiation of what gets in the way of asking for and responding to requests for attachment needs, especially in relationships with a pursue-distance dynamic (e.g. Johnson, 2004).  In Gottman’s couple therapy, the stress resulting in depression is an internal working model of worthlessness and lack of control, and therapy consists primarily of developing a new story about the self, in conversation, creating new and different kinds of positive interactions (e.g. Gottman, 1999).

One other school of thought about depression does not come along with a therapeutic modality, but is worth mentioning.  Evolutionary theorists have recently suggested that while depression is unpleasant and dangerous, it may be an adaptive behavior.  There is evidence to suggest that depression is essentially a physiological state designed to enhance our analytic-rumination process in the face of complex social problems (Andrews & Thomson, 2009).  In other words, depression may help us concentrate on and think about difficult interpersonal situations.

Depression in Couples—Two Literature Reviews

Social psychologists, apparently spurred on by an insatiable desire to understand the minute details of the relationship between depression and marital distress, have amassed a mountain of evidence, from the obvious to the baffling, about depression in couples and how depression affects couples’ and family’s dynamics.  The simple version of the story is that, while complicated by gender and many other factors, there is a relationship, and it is not good.  Depression in couples is not good for the couple, the individuals in the couple, or the children that come from the couple:

Depression correlates with dysfunctional marital interactions and marital distress (Gabriel, Beach, & Bodenmann 2010; Rehman, Ginting, Karimiha, & Goodnight, 2010), but there is also evidence that it does not contribute in any unique way to a couple’s communication (Eldridge, Jones, Sevier, Clements, Markman, Stanley, Sayers, Sher, & Christiansen, 2007).  Depression is negatively correlated with relationship satisfaction (Cramer & Jowett, 2010) and positively correlated to violent and abusive behaviors in relationships (Vaeth, Ramisetty-Mikler, & Caetano, 2010).  Depression reduces dyadic adjustment (Tilden, Gude, Hoffart, & Sexton, 2010), is less likely to remit in the presence of dyadic discord (Denton, Carmody, Rush, Thase, Trivedi, Arnow, Klein, & Keller, 2010), and predicts low relationship quality (Papp, 2010).  Depression is a precursor to divorce (Doohan, Carrere, & Riggs, 2010).  Depression hinders attachment in couples (Mehta, Cowan, & Cowan, 2009).  Depression in one spouse predicts depression in the other (Goodman & Shippy, 2003) (but not in dating couples (Segrin, 2004)).  Depression in either partner predicts marital dissatisfaction (Whisman, Weinstock, & Uebelacker, 2002).  Depression can trigger relationship problems/marital distress, and relationship problems/marital distress can trigger depression (Reich, 2003; Mead, 2002).  Distressed relationships increase depression relapse (Whisman & Schonbrun, 2010).

There is some evidence about what might help. A sense of enduring control in a relationship, for example, may buffer against depression (Keeton, Perry-Jenkins, & Sayer, 2008), and marital satisfaction is higher whether or not there is depression if your spouse likes you (Sacco & Phares, 2001).  And there is some evidence about how the interactions work.  Evidence exists, for example, that marital discord and depression may be mediated by hopelessness and blame (Sayers, Kohn, Fresco, Bellack, & Sarwer, 2001).  Partners’ depressive symptoms tend to become more similar, which may be driven by their coping styles becoming more similar (Holahan, Moos, Moerkbak, Cronkite, Holahan, & Kenney, 2007).  Depression may cause marital dysfunction and psychological distress by reducing the empathic accuracy of the depressed partner—that is, because they imagine incorrectly that the non-depressed partner feels as badly as they do (Papp, Kouros, & Cummings, 2010)–but there is also evidence that depression does not affect empathic accuracy in couples (Cramer & Jowett, 2010).  It is far more common for people seeking help for depression to have a relationship in distress than for those seeking help for relationship distress to have depression (Atkins, Dimidjian, Bedics, & Christensen, 2009).

Many of the findings are gender-specific. The effects of depression on relationship problems and vice versa are greater for women than men (Reich, 2003).  For women, neuroticism increases the effect of marital distress on depression, but for men it decreases it (Davila, Karney, Hall, & Bradbury, 2003).  Negative marital interactions make women’s but not men’s depression worse if they don’t have confidence in the relationship (Whitton, Olmos-Gallo, Stanley, Prado, Kline, St. Peters, Markman, 2007).  Husbands’ depression predicts wives’ depression, but not vice versa (Kouros & Cummings, 2010).

Finally, there is a lot of evidence that depression in parents is not good for their children, correlating with less secure attachment as infants (Martins & Gaffan, 2000), for example.  Parental depression affects  children’s adjustment to school more than parenting ability does (Cummings, Keller, & Davies, 2005). And depressed parents tend to have more depressed and disabled adult children (Timko, Cronkite, Swindle, Robinson, Turrubiartes, & Moos, 2008).

Treating Depression With Couples Therapy: The Evidence

Most of the research on treating depression with couples therapy has involved behavioral marital therapy (BMT).  I found ten empirical studies looking at BMT for depression.  Five of them compared BMT to individual cognitive or cognitive-behavioral therapies (CT or CBT).  In three of those, BMT was equivalent to individual therapy in its amelioration of depression, but more than individual therapy, it reduced marital discord  (Beach & O’Leary, 1986), increased marital adjustment for depressed women (Beach & O’Leary, 1992), and increased marital satisfaction (Emanuels-Zuuveen & Emmelkamp, 1996).  In the fourth, BMT reduced depression as much as individual CT, but only for distressed couples, not non-distressed couples (Jacobson, Dobson, Fruzzetti, & Schmaling, 1991), and in the fifth, a “behavioral spouse-aided therapy” for depression was equivalent to individual CBT but had no boost for marital satisfaction (Emanuels-Zuurveen & Emmelkamp, 1997).  In a sixth study, BMT reduced wives’ depression significantly more than a wait-list control, with 67% of participants improved and nearly half recovered three months later.  Additionally, wives’ marital satisfaction increased and husbands’ psychological distress decreased (Cohen, O’Leary, & Foran, 2010).  In the seventh study, two-thirds of depressed women improved with BMT, and nearly half recovered (Cohen, O’Leary, & Foran, 2009).  In the eighth, two-thirds of depressed spouses of both genders recovered from depression with BMT (Sher, Baucom, & Larus, 1990).  In a final study, depressive symptoms actually predicted positive outcomes for BMT (Jacobson, Follette, & Pagel, 1986).

Three studies looked at cognitive-oriented approaches.  CBT family therapy for depressed parents significantly alleviated parental depression, but the nearly-large effect size (d = .49) at six months fell to a small-to-medium effect size (d = .26) at a 12-month follow up (Compas, Forehand, Keller, Champion, Rakow, Reeslund, McKee, Fear, Colletti, Hardcastle, Merchant, Roberts, Potts, Garai, Coffelt, Roland, Sterba, & Cole, 2009).  Cognitive marital therapy reduced depression as much as individual cognitive therapy for depression and was a little better for severe depression (Teichman, Bar-El, Shor, Sirota, & Elizur, 1995).  Finally, seven sessions of cognitively-oriented family therapy produced substantial and long-term recovery from depression in depressed female parents and small recovery rates for depressed male parents (Beardslee, Wright, Gladstone, & Forbes, 2007).

Five more studies looked at a variety of other couples treatments.  An “integrative approach,” combining systemic, narrative, cognitive-behavioral therapies and psychoeducation, for residential adults with depression showed improvement with significant and large effect sizes (Tilden, Gude, Sexton, Finset, & Hoffart, 2010).  A “coping-oriented couple therapy” matched CBT and individual interpersonal therapy for improving depression (Gabriel, Bodenmann, Widmer, Charvoz, Schramm, & Hautzinger, 2009).  Systemic couples therapy was at least as effective as drug therapy, and more acceptable to participants (Leff, Vearnals, Brewin, Wolff, Alexander, Asen, Dayson, Jones, Chisholm, & Everitt, 2000).  EFT was also as effective as drug therapy for depression (Dessaulles, Johnson, & Denton, 2003).  A trial that included but did not separate the effects of integrative, systemic, psychodynamic, Gestalt and behavioral therapies reduced depression by about half in couples with infidelity and by a quarter in couples without infidelity (Atkins, Marin, Lo, Klann, & Hahlweg, 2010).

Finally, a meta-analysis in 2008, using seven of the studies mentioned above and ruling many others out for various reasons, found that couples therapy and individual therapy are equally effective treatments for depression and that couples therapy has the additional benefit of relieving distress in couples, but that there is so little evidence and it has been produced and described so poorly that it is too early to recommend couples therapy as a treatment for depression 46 (Barbatto & D’Avanzo, 2008).


Metaframeworks is an approach to therapy synthesizing ideas from many different models of family therapy, plus postmodern philosophy, multiculturalism, feminism, and Bateson’s interpretation of Bertalanffy’s general systems theory.  It was developed by family therapists and theorists Breunlin, Schwartz, and Mac Kune-Karrer, and initially presented in their 1992 book,Metaframeworks: Transcending the Models of Family Therapy.  (All information about information presented below about metaframeworks, however, is based on my interpretation of their second, revised, edition of that book, published in 1997.)  Breunlin and colleagues propose that humans suffer as a result of constraints upon their intrinsically resourceful and self-actualizing nature.  These constraints exist on any of several levels of organization: biological, individual, relational, familial, communal, and societal.  Constraints also exist in any of several forms, which give the names to the six metaframeworks: mind or “internal family systems” (IFS), sequences, organization, development, multicultural, and gender.  The level and form that constraints exist on and as interact to form a web of constraints, which can be conceptualized as in Figure 1 (see appendix).  In this conception, clients’ problems can exist as a result of constraints at any combination of points in the web.  The job of the therapist is, in a flexible, collaborative, and intuitive process, identify and release the constraints that are not allowing clients to resolve their problems.

The Role of Systems Theory

Systems theory is a body of thought concerned with “whole” patterns of organization and interaction and the relationships between them.  In metaframeworks, those wholes (calledholons) are characterized as a distinct network of parts which interact recursively by a set of organizing rules that allow for both adaptation and stability.  Holons exist and interact both in parallel and in hierarchies. For example, the organs in a human body are each holons, interacting with the other organs in various ways. At the same time, each organ is also interacting with the holons at higher (the whole body and “up”) and lower (individual cells and “down”) levels of organization.  The nature of all of these interactions is said to be recursive, meaning the action of each part of each whole at each level of organization affects the actions of every other part of every other whole, at every other level or organization.
A metaframeworks-oriented therapist uses the lens of systems theory to work towards a rich, holistic view of their clients, and to keep in mind all of the levels of organization that constraints might exist on, and how those levels might interact.  An individual client might be constrained by a biological disease process, for example, or an oppressive economic system.  Additionally, the disease, the person, and the economic system all interact with each other, each one possibly generating, regulating, or supporting the others in different ways.  Without a systemic view, we miss might these interactions and, therefore, possible points of intervention.

Epistemology and Assumptions

A metaframeworks-oriented therapist works grounded by a view of reality, health, human nature, and change based on four assumptions.  First, they take a middle-of-the-road constructivist stance, called perspectivism.  That is, our senses and thinking do produce maps of a reality that exists independently, but our maps are always limited by our perspectives—our senses, beliefs, and cultures.  Furthermore, some maps are more accurate than others.  Second, it is more accurate and useful to think of human problems in terms of constraints than pathology.  That is, if we assume that a client is suffering or dangerous because we have yet to identify and resolve factors that are holding them back from their potential, we are much more likely to help them than if we assume that they are evil or broken.  Third, healthy systems are characterized by balance, harmony, and leadership.  It is the job of certain parts of a system to provide flexible coordination and regulation for that system, and lack thereof indicate that the leadership of the system is acting under constraints.  Likewise, escalation, rigidity, and failures to adapt or meet the needs of some parts of a system are indications of constraint.  Fourth, metaframeworks-oriented therapists believe that intrapsychic work with the subpersonalities of an individual client is often important to resolve constraints.  The rejection of inner work by behaviorists and early family theorists limited the ability of therapists to identify and resolve constraints.  Furthermore, subpersonalities are organized and interact in the same way that parts of other systems are organized and interact:  They are ideally balanced and harmonious, and have strong leadership in the Self, which is similar to Freud’s concept of ego, but they can become polarized, extreme, or rigid and thus constraining.

Using Metaframeworks

Each of the metaframeworks is a perspective from which to look for and at constraints at any level of organization.  Each has its own conceptual framework, vocabulary, and can suggest different hypotheses and interventions.  The metaframeworks approach uses these perspectives with a blueprint for continually and collaboratively expand and fine-tune therapy for each client.

Internal family systems. IFS is a model of therapy as well as a metaframework.  As a metaframework, it is essentially a psychodynamic theory, positing the existence of a Self plus any number of subpersonalities, each filling a specific role.  We may have a creative part, a cautious part, an adventurous part and so on.  IFS holds that trauma can cause a breakdown in the organization of our subpersonalities.  Instead of balanced and harmonious parts lead by a strong self, some parts can become dissociated and others can come to overshadow the Self.  A typical result of trauma is that the Self becomes weak compared to three categories of parts: exiled parts holding scary, painful memories which can be triggered, manager parts which are vigilant for danger and keep the exiled parts exiled, and firefighter parts which react aggressively if apparent danger or an exile get too close.

As a model of therapy, IFS gently works with the Self of a client to calm and contain the manager and firefighter parts so that the exiled parts can be heard and healed.  When this happens, the Self naturally regains leadership of the intrapsychic system.

Sequences. The sequences metaframework is a way of looking for constraining patterns of behavior at different time increments.  Constraining patterns can exist at the level of individual interactions, such as conversations or arguments, at the level of daily routines, such as work or school schedules, at the level of longer cycles, such as yearly holiday rituals or seasonal production cycles, or at the level of intergenerational patterns, such as inheriting a tendency for abusing people or substances from parents or grandparents.  Breunlin and colleagues also acknowledge the possibility of longer cycles, such as Strauss and Howe’s (1991) four-generation cycle, or even larger societal shifts of values.  The different temporal levels of patterning also interact with each other recursively.  Patterns at any level can exacerbate or ameliorate patterns at any other level.

A metaframeworks-oriented therapist who suspected constraining sequences would choose interventions from models of therapy which specialize in the relevant temporal level of sequence.  Constraints at the level of face-to-face interactions, might best be resolved using techniques from the MRI school of thought, for example, while constraints at the level of intergenerational patterns might be best resolved using Bowenian interventions.  Or if a therapist suspects that one pattern is ameliorating a different, more constraining pattern, solution-focused interventions may be the most appropriate.

Organization. The organization metaframework is used to assess for constraints of balance, harmony, and leadership at various levels of organization.  Balance is a state in which each part of a system has an appropriate amount of resources and power compared to other parts. Harmony is the presence of cooperation and a balance between autonomy and interconnectedness between parts.  Some parts of a system appropriately provide leadership for that system, regulating the flow of resources, and planning for the future with the health of the whole system in mind.  Parents, for example, appropriately fill this role in nuclear families, and elected leaders in democratic societies.  Weak, rigid, or extreme leadership indicates constraints on a system’s organization.

A metaframeworks-oriented therapist who suspected organizational constraints might choose interventions from models of therapy which specialize in that kind of organizational constraint.  Structural, strategic, and Bowenian family therapies, for example, focus in different ways on the quality of boundaries and leadership in a system which can constrain harmony and balance.

Development. The development metaframework reminds us that there are more or less predictable developmental pathways that each level of organization follow.  It is important to keep the milestones of those paths in mind in order to recognize, normalize, and ameliorate constraints that can appear at each stage of each level: biological, individual, relational, and familial, recognizing a recursive relationship between developments at all levels, including societal.  Breunlin and colleagues use an adaptation of Carter and McGoldrick’s (1988) family life cycle theory for family development, an adaptation of Wynne’s (1988) relational development theory for relationships, and posit a theory of individual development in which people in a transition between more stable stages exhibit an oscillation between behaviors appropriate to the old and new stages.  In the absence of constraints, the oscillation dampens towards the new stage.  An oscillation of competence that is maintained is a clue to the presence of significant constraints.

A metaframeworks-oriented therapist should notice and resolve four types of developmental constraints, promoting synchronous development between all levels of organization:  First, if a family is not adapting flexibly to a new life-cycle stage, they are encouraged to notice and adapt to their new circumstances.  Second, if an individual is exhibiting behavioral oscillations, the therapist collaborates with the family to facilitate age-appropriate competence.  Third, in the presence of a biological constraint to development in an individual, the therapist encourages the family to adjust their expectations appropriately.  Fourth, therapists  help enhance any underdeveloped relationship qualities: attraction, liking, nurturing, coordinating meaning, rule setting, and metarules.

Multicultural. The multicultural metaframework is a perspective for therapists to use which takes into account the cultural diversity present in both our clients and ourselves, and the potential interactions between those cultures.  They include immigration status, economic status, education, ethnicity, religion, gender, age, race, minority/majority status, regional background, sexual orientation, disability, and the presence of marginalizing physical characteristics.

A metaframeworks-oriented therapist uses the multicultural metaframework to generate questions, conversations, and knowledge about how the therapist’s and clients’ diversity affect the experiences we have and the meanings we make of them.  The degree of cultural overlap, orfit, between client and therapist, majority and minority factions, or between immigrant and host cultures is a clue to the possibility of constraints; any area of cultural incongruity is likely to produce constraints on the part with less resources and control.

Gender. The gender metaframework is a systemic feminist perspective on therapy, based on the ideals of egalitarianism between genders and ecofeminism, a philosophy of interrelatedness and compassion between all.  Breunlin and colleagues propose a developmental scheme for gender in relationships from traditional, with patriarchal gender roles, to gender aware, where women begin to become angry and men fearful about power and gender roles, to polarized, where those angry and fearful parts are running the show, to balanced, where an egalitarian organization is idealized by both men and women, who work to achieve it.

A metaframeworks-oriented therapist uses the gender metaframework to highlight and attempt to release clients from the constraints of patriarchy.  The therapist identifies the gender stage a client and/or client system is at, and uses questions, statements, and directives designed to move the client or system toward the balanced, egalitarian stage.

The blueprint for therapy and the role of the therapist. A metaframeworks-oriented therapist aims to find and remove constraints that are holding clients back from flexible adaptation, balance, and harmony, keeping in mind that problems can be generated by a recursively interacting constellation of constraints existing at any level of organization, and in the form of any of the metaframeworks’ domains.  This means that therapy for each client will be unique, addressing their particular constellation.  That unique therapy is accomplished with four recursively interacting processes:  (a) Hypothesizing is taking an educated guess at what relevant constraints exist, taking care to remain curious rather than adamant about the truth of hypotheses.  (b) Planning is collaboratively implementing techniques and interventions from any model of family therapy that the current hypothesis suggests, taking care to tailor those interventions to fit the assumptions of metaframeworks.  Planning includes relating, or maintaining the therapeutic relationship, staging, or keeping an eye on the current position in the therapeutic arc, and creating events, which is the actual implementation of interventions.  (c) Conversing means conducting the therapeutic conversation, using questions, statements, and directives, taking care to use language that does not produce defensiveness or confusion in clients.  (d) Reading feedback is the process of recognizing and distinguishing the importance of what clients do and say, taking care to remember that clients are speaking and we are listening from the standpoint of personally and culturally limited perspectives.

A metaframeworks therapist relates to their clients under two balancing principles.  First, it is the therapist’s job to provide leadership for the process of therapy.  Second, for a true collaboration, the therapist must remain honest and clear about what they are doing, thinking, and why.  Under those principles, the therapist moves intuitively between the four components of therapy, hypothesizing, planning, conversing, and reading feedback, and between the six metaframeworks, mind/IFS, sequences, organization, development, culture, and gender.  Doing so, they are most likely to recognize the relevant constraints and deliver effective interventions in a respectful way.

Evidence and a Rationale for the Use of Metaframeworks

There appears to be very little experimental evidence to date supporting Breunlin and colleagues (1997) theory of therapy specifically.  A PsychInfo search on December 3, 2010 for “metaframeworks” produced, for example, only 15 relevant hits.  Two of those were editions of Breunlin and colleagues’ book (1992; 1997), 5 were theoretical chapters (Breunlin & Mac Kune-Karrer, 2002; Breunlin, Rampage, Eovaldi, Mikesell, 1995; Foy & Breunlin, 2001; Lebow, 2003; Rampage, 1998), 3 theoretical doctoral dissertations (Luther, 1995; Nehring, 1998; Schneider, 1998), 2 theoretical journal articles (Breunlin, 1999; Pinsof, 1992), 1 journal article about a metaframeworks training video (Cohen & Abed, 2003).and 1 book review ( Lawson, 1993).  The only hits purporting empirical evidence about metaframeworks (Oulvey, 2000) admits in the abstract that the research design invalidated any results.

A PsychInfo search for “internal family systems” did little better:  The approximately 20 relevant hits were overwhelmingly theoretical presentations or critiques.  The 3 empirical hits consisted only of case reports.  Two were of a successful treatment of a 17-year old female with bulimia (Schwartz, 1987; Schwartz & Grace, 1989), and the other of an African-American family with a young, female sexual abuse survivor (Wilkins, 2007).

Though this may not represent every piece of empirical evidence supporting the metaframeworks theory of therapy, it is clear that neither its tenets nor effectiveness have any strong empirical support.  While we cannot assert with any certainty what forthcoming evidence will suggest about metaframeworks, we can safely assume that the present dearth of evidence in either direction is a result of the newness of the theory rather than the bias of scientific journals against publishing null results.

A rationale for the use of metaframeworks at this point, therefore, must be primarily theoretical.   Theoretically, the efficacy of a therapist using metaframeworks is likely to fall within the same range as other models of therapy, since all models of therapy that have been tested so far appear to function in approximately the same range (Miller, Duncan & Hubble, 1997).  Why should metaframeworks be different?  There are optimistic and pessimistic views:

An optimist about metaframeworks might say that metaframeworks stands to outperform other models for two reasons. First, the stress in the diathesis-stress model is likely to exist in many different forms, and metaframeworks looks systematically at most of the forms we know about, where other models tend to focus on one or two.  Second, to the extent that the efficacy of therapy relies on model-specific techniques and interventions, famously estimated at 15% by Lambert (1994), metaframeworks can benefit from the techniques of every model of family therapy, plus multiculturalism, feminism, and psychodynamics.  Adopting those ideas from other models when appropriate will potentially benefit a metaframeworks-oriented therapist without falling prey to any specific model’s blind spots or other weaknesses.  Metaframeworks’ strength is the sum of the strengths of other models.

A pessimist about metaframeworks might argue that the complexity and sophistication of metaframeworks could stand in the way of success.  We still operate inside a 50-minute-per-week schedule, after all, and there’s only so much one can accomplish in that time.  Perhaps metaframeworks is just an extra-confusing and complicated way to provide the benefits of a therapeutic relationship.  If that is the case, metaframeworks may turn out to be somewhat less effective than other models of therapy.

At this point, there is not enough evidence to say who is right, or where on that optimist-to-pessimist spectrum “right” falls.  For the therapist who is a metaframeworks optimist, like myself, it is clearly important to know the model thoroughly, be able to negotiate its concepts and connections fluidly, and to be fluent with nearly all of the interventions of IFS and the other family therapy models.  Not to do so would be to fall prey to the critique of the pessimist.


Though an optimist about metaframeworks, I believe I see weaknesses in the model. As a family therapist in training, all I can offer is a theoretical critique of the theory, and having just made contact with metaframeworks, my critique may be severely limited by my understanding, but it will show the extent of my understanding.

First, it is not clear to me how the IFS model is a true metaframework the way sequences, organization, and the others are, existing at every level of organization.  IFS seems to be primarily a model for working with individuals’ internal dynamics—almost more of a level than a metaframework.  How does a society-level constraint show up in the IFS framework, for example?  Also, once you admit that working with psychodynamics is useful, you will need to justify using a new model like IFS over a more mature model, of which there are many.  The idea of using systems thinking to approach psychodynamics is intriguing but somewhat less intuitive than systems thinking for families.

I have similar questions about the developmental metaframework.  While Breunlin and colleagues (1997) mention changing values at the level of societies, there is no attempt to put forth a developmental scheme.  Change is not development in the psychological sense.  There are society-level developmental schemes available, such as those put forth by Gebser or Graves (e.g. Graves, 1970; Wilber, 2000) which might be useful in creating a true developmental metaframework.  Development is also ignored on the community level, though I have no ideas on how to improve that.  The developmental scheme put forth for individuals, that of oscillation at developmental nodes, is simplistic compared to the wealth of knowledge developmental psychologists have discovered.  They make no mention even of the foundational work of Piaget, Kholberg, or Gilligan.

The gender metaframework has elements of cultural imperialism:  The therapist knows better than the clients and is duty-bound to change them, if possible.  I believe the problem lies in the gender-relations developmental scheme.  In the description of traditional to gender aware to polarized to balanced there is the clear judgment of very bad to less bad to even less bad to good.     While there is truth to this judgment, it limits the therapist, I think.  A better developmental scheme would, as an option, allow for and encourage a positive expression of traditional value-structure,s rather than a simple, negative judgment of the large majority of the earth’s cultures.

Finally, while metaframeworks seems to do an admirable job of integrating the good ideas from the original systems-oriented therapies, it provides no clear space for the good ideas from other therapy models.  I am specifically missing three elements.  First, Johnson’s twin breakthroughs of incorporating attachment theory and a focus on emotions in couples therapy (e.g. Johnson, 2004).  Second, the tools of cognitive therapy for examining meaning and resolving problematic meaning-making (e.g. Beck, 1979).  Third, mindfulness practices and interventions (e.g. Dimeff, Koerner, & Linehan, 2007).


Metaframeworks and Depression in Couples

As a metaframeworks-oriented therapist, when presented with a couple who complain of depression, I would assume that these are naturally resourceful people who are operating under some constraints right now.  It is likely that some number of factors in the web of possible constraints is providing the stresses that have triggered and maintained the symptoms of depression.  I would enter into a conversation with this couple, listening for clues as to what these constraints might be, at what level of organization, and seen clearest with the perspective of which metaframework.

It may be that one of those constraints is biological and at the individual level of organization—that is, a genetic tendency for depression—and that antidepressant medication or some other somatic intervention would be necessary.  If a thorough assessment for suicidality is negative, however, I would try out other hypotheses first, to see if releasing other constraints is adequate, as antidepressant medications have side effects while the releasing of other constraints does not (Breunlin et al., 1997).

I would take the couple’s lead in our conversation.  What do they think is important to talk about?  Chances are, they have some strong and useful ideas about where their stress is coming from, or at least when it is they feel the most stressed and when and where they feel the least stressed.  During the conversation I would keep the metaframeworks in mind, and the levels of organization, listening for clues about which hold the relevant constraints, which I would follow up on, developing and checking out hypotheses.

Do they complain about communication problems?  That would be a clue that the sequences metaframework may be involved at the interaction level.  If that checked out, it might suggest an interaction-oriented intervention, such as Gottman’s soft-startup psychoeducation, or one of Satir’s congruent-communication exercises.  I would also keep in mind the other levels of sequences.  Is there evidence for constraining daily routines?  Any chance of seasonal affective depression or holiday blues?  How about a history of depression in the family, a possible intergenerational sequence?  I would keep in mind that any sequences I come across can interact recursively with each other, generating, maintaining, regulating, or exacerbating each other.  I would hold my hypotheses lightly, listening for evidence against them, and when I felt like I had a good one, I would try an intervention, watching and listening for, and respecting, any feedback I got from the couple.  Whether or not the intervention brought out evidence for or against my hypothesis, I would use that evidence to generate new directions in the conversation, new questions, more accurate hypotheses.

I may come to suspect that internal family systems are involved in the depression, perhaps in the form of a polarization between angry and disappointed parts of each partner, or between perfectionist and overwhelmed parts of one of the partners.  If that seemed like the case, I could try working with the parts, calming the extreme parts and encouraging the leadership of the Self.  Throughout the process, I would keep in mind that both the levels of organization and the metaframeworks interact recursively.  Are these parts reflected in an intergenerational sequence?  A family-level organization pattern? How about a society-level pattern?  I would follow up on any relevant-seeming clues, getting feedback from the couple, developing a richer understanding of their situation.

I would keep an eye out for organizational and developmental constraints.  Does the couple lead their family appropriately?  Are there parentified children? Cross-generational coalitions?  These might begin to suggest structural or strategic interventions, modified for maximum respectfulness, of course.  Are they at a nodal point in the family life cycle? Having their first child? Launching their last child?  Is there evidence of an oscillation in competence?  To what stage has their relationship developed?

I would investigate and keep in mind the couple’s multicultural and gender aspects, watching for friction from a poor cultural fit, or gender-based power dynamics.  Is either partner an immigrant?  Second- or third-generation American?  Is their socioeconomic class a fit with each other, or suffer from the constraints of poverty?  What is their sexual orientation and what other ways do they identify with majority or minority cultures?  How do they think about religion?  Education?  How about gender?  Have they taken on traditional gender roles or have they begun to chafe under them?  Achieved balance?  How might I support the partner who has less power to stand up for themself?  How might I support the partner with more power to more gently move through the transition to gender balance?

Culture and gender are also where I would keep a close eye on myself:  How might my culture and gender be constraining my thinking or compassion with this couple?  How might our similarities or differences blind me to possible constraints?  I would quite possibly check in with a colleague or supervisor about this.

Ideally, I would move our conversation with purpose and fluidity through, and back and forth between, the different metaframeworks and the different levels of organization, identifying and releasing constraints by staging and creating events and reading the feedback.  If the depression was the result of just a few constraints, the process might be fairly simple.  If the depression was the result of a dense, heavily embedded constellation of constraints, the process might take longer and involve many iterations of rethinking, shifting hypotheses, trying out different angles and interventions.  Ideally the process would also be intuitive, so that I could just follow my natural curiosity and the natural flow of our conversation.  And, ideally, as a result of our conversation, the other events, like interventions, we orchestrate in our relationship, and the quality of our relationship, the constraints release and the couple finds that they no longer struggle with depression.






















American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (Revised 4th ed.). Washington, DC: Author.

Anda, R. F., Whitfield, C. L., Felitti, V. J., Chapman, D., Edwards, V. J., Dube, S. R., & Williamson, D. F. (2002). Adverse childhood experiences, alcoholic parents, and later risk of alcoholism and depressionPsychiatric Services, 53(8), 1001–1009.

Andrews, P. W., & Thomson, Jr., J. A. (2009). The bright side of being blue: Depression as an adaptation for analyzing complex problems. Psychological Review116(3), 620-654.

Atkins, D. C., Dimidjian, S., Bedics, J. D., Christensen, A. (2009). Couple discord and depression in couples during couple therapy and in depressed individuals during depression treatment. Journal of Consulting and Clinical Psychology, 77(6), 1089-1099. doi:10.1037/a0017119

Atkins, D. C., Marín, R. A., Lo, T. T. Y., Klann, N., & Hahlweg, K. (2010). Outcomes of couples with infidelity in a community​-​based sample of couple therapy. Journal of Family Psychology, 24(2), 212-216. doi:10.1037/a0018789

Ballantine, H. T., Bouckoms, A. J., Thomas, E. K., & Giriunas, I. E. (1987). Treatment of psychiatric illness by stereotactic cingulotomy. Biological Psychiatry, 22(7). 807-819.

Barbato, A. & D’Avanzo, B. (2008). Efficacy of couple therapy as a treatment for depression: A meta-analysis. Psychiatry Quarterly, 79, 121-132.

Baucom, B., Eldridge, K., Jones, J., Sevier, M., Clements, M., Markman, H., Stanley, S., Sayers, S. L., Sher, T., & Christensen, A . (2007). Relative contributions of relationship distress and depression to communication patterns in couples. Journal of Social and Clinical Psychology, 26(6), 689-707. doi:10.1521/jscp.2007.26.6.689

Beach, S. R. & O’Leary, K. D. (1986). The treatment of depressionoccurring in the context of marital discord. Behavior Therapy, 17(1), 43-49. doi:10.1016/S0005-7894(86)80112-5

Beach, S. R. & O’Leary, K. D. (1992). Treating depression in the context of marital discord: Outcome and predictors of response of marital therapy versus cognitive therapy. Behavior Therapy, 23(4), 507-528. doi:10.1016/S0005-7894(05)80219-9

Beardslee, W. R., Wright, E. J., Gladstone, T. R. G.,& Forbes, P. (2007). Long​-​term effects from a randomized trial of two public health preventive interventions for parental depression.Journal of Family Psychology, 21(4), 703-713. doi:10.1037/0893-3200.21.4.703

Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York: Meridian.

Beck, A. T. (1979). Cognitive therapy and the emotional disorders. New York: Meridian.

Beck, A. T., Steer, R. A., Ball, R., & Ranieri, W. (1996). “Comparison of Beck Depression Inventories -IA and -II in psychiatric outpatients”. Journal of Personality Assessment 67(3): 588–97. doi:10.1207/s15327752jpa6703_13

Bonde, J. P. (2008). Psychosocial factors at work and risk of depression: A systematic review of the epidemiological evidence. Journal of Occupational and Environmental Medicine, 65(7), 438–45. doi:10.1136/oem.2007.038430

Breunlin, D. C. & MacKune-Karrer, B. (2002). Metaframeworks. In F. W. Kaslow (Ed.),Comprehensive handbook of psychotherapy: Integrative/eclecticVol. 4, (pp. 367-385). Hoboken, NJ: John Wiley & Sons.

Breunlin, D. C. (1999). Toward a theory of constraints. Journal of Marital and Family Therapy25(3), 365-382. doi:10.1111/j.1752-0606.1999.tb00254.x

Breunlin, D. C., Rampage, C., Eovaldi, M. L., & Mikesell, R. H. (1995). Family therapy supervision: Toward an integrative perspective. In D. D. Lusterman & S. H. McDaniel (Eds.),Integrating family therapy: Handbook of family psychology and systems theory, (pp. 547-560). Washington, DC, US: American Psychological Association. doi:10.1037/10172-031

Breunlin, D. C., Schwartz, R. C., & Mac Kune-Karrer, B. (1992). Metaframeworks: Transcending the models of family therapy. San Francisco: Josey-Bass.

Breunlin, D. C., Schwartz, R. C., & Mac Kune-Karrer, B. (1997). Metaframeworks: Transcending the models of family therapy (2nd ed.). San Francisco: Josey-Bass.

Carter, B. & McGoldrick, M., (1988). The changing family life cycle: A framework for family therapy. (2nd ed.) Boston: Allyn & Bacon.

Caspi, A., Sugden, K., Moffitt, T. E., Taylor, A., Craig, I. W., Harrington, H., McClay, J., Mill, J., Martin, J., Braithwaite, A., & Poulton, R. (2003). Influence of life stress on depression: Moderation by a polymorphism in the 5-HTT Gene. Science, 301, 386-389. doi:10.1126/science.1083968

Chapman D. P., Anda, R. F., Felitti, V. J., Dube, S. R., Edwards, V.J., & Whitfield, C. L. (2004).Epidemiology of adverse childhood experiences and depressive disorders in a large health maintenance organization populationJournal of Affective Disorders 82, 217–225.

Cohen, R. A., Paul, R., Zawacki, T. M., Moser, D. J., Sweet, L., & Wilkinson, H. (2001). Emotion and personality changes following cingulotomy. Emotion, 1(1), 38-50.

Cohen, R. S. & Abed, F. (2003). CD-ROM simulation training in perceptual, conceptual, and executive skills in family therapy: A metaframeworks approach. Journal of Systemic Therapies,22(3), 27-40. doi:10.1521/jsyt.

Cohen, S., O’Leary, K. D., & Foran, H. (2009). A randomized clinical trial of a brief, problem​-​focused couple therapy for depression. Behavior Therapy, 41(4), 433-446. doi:10.1016/j.beth.2009.11.004

Cohen, S., O’Leary, K. D., & Foran, H. (2010). A randomized clinical trial of a brief, problem​-​focused couple therapy for depression. Behavior Therapy, 41(4), 433-446. doi:10.1016/j.beth.2009.11.004

Compas, B. E., Forehand, R., Keller, G., Champion, J. E., Rakow, A., Reeslund, K. L., McKee, L., Fear, J. M., Colletti, C. J. M., Hardcastle, E., Merchant, M. J., Roberts, L., Potts, J. Garai, E., Coffelt, N., Roland, E., Sterba, S. K., & Cole, D. A., (2009). Randomized controlled trial of a family cognitive​-​behavioral preventive intervention for children of depressed parents. Journal of Consulting and Clinical Psychology, 77(6), 1007-1020. doi:10.1037/a0016930

Craighead, W. E., Sheets, E. S., Brosse, A. L., & Ilardi, S. S. (2007). Psychosocial treatments for major depressive disorder. In P. E. Nathan & J. M. Gorman (Eds.), A Guide to Treatments that Work (3rd ed.) (pp. 289-307). Cary, NC: Oxford.

Cramer, D., Jowett, S. (2010). Perceived empathy, accurate empathy and relationship satisfaction in heterosexual couples. Journal of Social and Personal Relationships, 27(3), 327-349. doi:10.1177/0265407509348384

Cummings, E. M., Keller, P. S., & Davies, P. T. (2005). Towards a family process model of maternal and paternal depressive symptoms: Exploring multiple relations with child and family functioning. Journal of Child Psychology and Psychiatry, 46(5), 479-489. doi:10.1111/j.1469-7610.2004.00368.x

Davila, J., Karney, B. R., Hall, T. W. & Bradbury, T. N. (2003). Depressive symptoms and marital satisfaction: Within-subject associations and the moderating effects of gender and neuroticism. Journal of Family Psychology, 17(4), 557-570. doi:10.1037/0893-3200.17.4.557

Dein, S. (2006). Religion, spirituality and depression: Implications for research and treatment.Primary Care and Community Psychiatry, 11(2), 67–72. doi:10.1185/135525706X121110.

Denton, W. H. Carmody, T. J. Rush, A. J. Thase, M. E. Trivedi, M. H. Arnow, B. A. Klein, D. N. & Keller, M. B. (2010). Dyadic discord at baseline is associated with lack of remission in the acute treatment of chronic depression. Psychological Medicine40(3), 415-424. doi:10.1017/S0033291709990535

DeShazer, S. (1988). Clues: Investigating solutions in brief therapy. New York: Norton.

Dessaulles, A., Jonson, S. M., & Denton, W. H. (2003). Emotion focused therapy for couples in treatment of depression: A pilot study. American Journal of Family Therapy 31, 345–353.

Dimeff, L. A., Koerner, K., & Linehan, M. M. (2007). Dialectical behavior therapy in clinical practice: Applications across disorders and settings. New York: Guilford.

Doohan, E. M., Carrère, S. Riggs, M. L. (2010). Using relational stories to predict the trajectory toward marital dissolution: The oral history interview and spousal feelings of flooding, loneliness, and depression. Journal of Family Communication, 10(1), 57-77. doi:10.1080/15267430903396401

Edwards, V. J, Holden, G. W., Anda, R. F., & Felitti, V. J. (2003). Experiencing multiple forms of childhood maltreatment and adult mental health: results from the Adverse Childhood Experiences (ACE) StudyAmerican Journal of Psychiatry, 160(8), 1453–1460.

Ellis, A. (1998). The Albert Ellis reader: A guide to well-being using rational emotive behavior therapy. Secaucus, NJ: Citadel.

Emanuels-Zuurveen, L. & Emmelkamp, P. M. (1996). Individual behavioural-cognitive therapy v. marital therapy for depression in marital distressed couples. British Journal of Psychiatry, 169, 181–188.

Emanuels-Zuurveen, L., Emmelkamp, P. M. (1997). Spouse-aided therapy with depressed patients. Behavior Modification 21, 62–77.

Fisch, R., Weakland, J. H., & Segal, L. (1982). The tactics of change. San Francisco: Josey-Bass.

Fournier, J. C., DeRubeis, R. J., Hollon, S. D., Dimidjian, S., Shelton, R. C., & Fawcett, J. (2010). Antidepressant drug effects and depression severity: A patient-level meta-analysis.Journal of the American Medical Association 303(1), 47-54.

Foy, C. W. & Breunlin, D. C. (2001). Integrative supervision: A metaframeworks perspective. In S. H. McDaniel, D. D. Lusterman, & C. L. Philpot (Eds.), Casebook for integrating family therapy: An ecosystemic approach, (pp. 387-394). Washington, DC: American Psychological Association. doi:10.1037/10395-032

Gabriel, B., Beach, S. R. H., & Bodenmann, G. (2010). Depression, marital satisfaction and communication in couples: Investigating gender differences. Behavior Therapy41(3), 306-316. doi:10.1016/j.beth.2009.09.001

Gabriel, B., Bodenmann, G., Widmer, K., Charvoz, L., Schramm, E., & Hautzinger, M. (2008). Effects of cognitive behavior therapy, interpersonal psychotherapy, and coping-oriented couple therapy for the treatment of depression on the observed dyadic coping behavior. Zeitschrift für Klinische Psychologie und Psychotherapie: Forschung und Praxis, 37(3),  179-189. doi:10.1026/1616-3443.37.3.179

Goodman, C. R. & Shippy, R. A. (2002). Is it contagious? Affect similarity among spouses.Aging & Mental Health, 6(3), 266-274. doi:10.1080/13607860220142431

Gottman, J. M., (1999). The marriage clinic: A scientifically-based marital therapy. New York: Norton.

Graves, C. W. (1970). Levels of existence: An open system theory of values. Journal of Humanistic Psychology, 10(2), 131-155. doi:10.1177/002216787001000205

Greenberg, P. E., Kessler, R. C., Birnbaum, H. G., Leong, S. A., Lowe, S. W. Berglund, P. A., & Corey-Lisle, P. K. (2003). The economic burden of depression in the United States : How did it change between 1990 and 2000? Journal of Clinical Psychiatry, 64(12), 1465-1475.

Haley, J. (1984). Ordeal therapy. San Francisco: Josey-Bass.

Hamilton, M. (1980) Rating depressive patients. Journal of Clinical Psychiatry, 41, 21-24.

Holahan, C. J., Moos, R. H., Moerkbak, M. L., Cronkite, R. C., Holahan, C. K., & Kenney, B. A. (2007). Spousal similarity in coping and depressive symptoms over 10 years. Journal of Family Psychology, 21(4), 551-559. doi:10.1037/0893-3200.21.4.551

Jacobson, N. S., Dobson, K., Fruzzetti, A. E. , Schmaling, K. B. (1991). Marital therapy as a treatment for depression. Journal of Consulting and Clinical Psychology 59, 547–557.

Jacobson, N. S., Follette, W. C., & Pagel, M. (1986). Predicting who will benefit from behavioral marital therapy. Journal of Consulting and Clinical Psychology, 54(4), 518-522. doi:10.1037/0022-006X.54.4.518

Johnson, S. M. (2004). The practice of emotionally focused couple therapy: Creating connection. New York: Brunner-Routledge.

Johnson, S. M. (2004). The practice of emotionally-focused couple therapy: Creating connection. New York: Brunner-Routledge.

Kanter, J. W., Cautilli, J. D., Busch, A. M., & Baruch, D. E. (2005). Toward a comprehensive functional analysis of depressive behavior: Five environmental factors and a possible sixth and seventh. The Behavior Analyst Today, 6(1), 65-77.

Keeton, C. P., Perry-Jenkins, M., & Sayer, A. G. (2008). Sense of control predicts depressive and anxious symptoms across the transition to parenthood. Journal of Family Psychology, 22(2), 212-221. doi:10.1037/0893-3200.22.2.212

Kendler, K. S. (2003). Life event dimensions of loss, humiliation, entrapment, and danger in the prediction of onsets of major depression and generalized anxiety. Archives of General Psychiatry, 60(8), 789–796. doi:10.1001/archpsyc.60.8.789.

Kessler, R. C. (1997). The effects of stressful life events on depression. Annual revue of Psychology, 48, 191–214. doi:10.1146/annurev.psych.48.1.191.

Kim, D. (2008). Blues from the neighborhood? Neighborhood characteristics and depression.Epidemiological Revue, 30(1): 101-117. doi:10.1093/epirev/mxn009

Klerman, G. L., Weissman, M. M., Rounsaville, B. J., &Chevron, E. S. (1984). Interpersonal psychotherapy of depression. New York: Basic.

Lambert, M. J. (1992) Psychotherapy outcome research: Implications for integrative and eclectic therapists. In J. C. Norcross & M. R. Goldfried (Eds.), Handbook of psychotherapy integration(pp. 94-129). New York: Basic Books.

Lawson, J. M. (1993). Family therapy: Reading the subtitles. PsycCRITIQUES8(4), 409-410. doi:10.1037/033246

Lebow, J. L. (2003). Integrative approaches to couple and family therapy. In T. L. Sexton, G. R. Weeks, & M. S. Robbins, (Eds), Handbook of family therapy: The science and practice of working with families and couples, (pp. 201-225). New York: Brunner-Routledge.

Leff, J., Vearnals, S., Brewin, C. R., Wolff, G., Alexander, B., Asen, E., Dayson, D., Jones, E., Chisholm, E., & Everitt, B. (2000). The London depression intervention trial. Randomised controlled trial of antidepressants v. couple therapy in the treatment and maintenance of people with depression living with the partner: clinical outcome and costs. The British Journal of Psychiatry, 177, 95–100.

Leupnitz, D. (1988). The family interpreted: Feminist theory in clinical practice. New York: Basic.

Luther, C. J. (1995). Utilizing nondirective play therapy interventions within a family therapy model: A theoretical and clinical exploration. Dissertation Abstracts International: Section B: The Sciences and Engineering.

Martins, C. & Gaffan, E. A. (2000). Effects of early maternal depression on patterns of infant–mother attachment: A meta-analytic investigation. Journal of Child Psychology and Psychiatry, 41(6), 737-746. doi:10.1111/1469-7610.00661

McCullough, M., Larson, D. (1999). Religion and depression: A review of the literatureTwin Research, 2(2): 126–136. doi:10.1375/136905299320565997

Mead, D. E. (2002). Marital distress, co-occurring depression and marital therapy: A review.Journal of Marital and Family Therapy, 28(3), 299-314. doi:10.1111/j.1752-0606.2002.tb01188.x

Mehta, N., Cowan, P. A., & Cowan, C. P. (2009). Working models of attachment to parents and partners: Implications for emotional behavior between partners. Journal of Family Psychology, 23(6), 895-899. doi:10.1037/a0016479

Miller, S. D., Duncan, B. L., & Hubble, M. A. (1997). Escape from babel: Toward a unifying language for psychotherapy practice. New York: Norton.

Minuchin, S. & Fishman, H. C. (1982). Family therapy techniques. Boston: Harvard UP.

Monroe, S. M. & Simons, A. D. (1991). Diathesis-stress theories in the context of life stress research: Implications for the depressive disorders. Psychological Bulletin, 110(3), 406-425.

Nehring, K. A. (1998). The metaframeworks approach to treatment of a family in therapy.Dissertation Abstracts International: Section B: The Sciences and Engineering.

Nichols, M. P. & Schwartz, R. C. (2008). Family therapy: Concepts and methods. Boston: Pearson.

Olfson, M., Marcus, S. C., Druss, B., Elinson, L., Tanielian, T., & Pincus, H. A. (2002). National trends in the outpatient treatment of depression. Journal of the American Medical Association, 287(2), 203-210.

Oulvey, E. A. (2000). The relationship of long-term training in a multi-systemic model to changes in epistemic style and the symptoms of burnout. Dissertation Abstracts International: Section B: The Sciences and Engineering.

Papp, L. M. (2010). The course and quality of intimate relationships among psychologically distressed mothers. American Journal of Orthopsychiatry, 80(1), 71-79.

Papp, L. M., Kouros, C. D., & Cummings, E. M. (2010). Emotions in marital conflict interactions: Empathic accuracy, assumed similarity, and the moderating context of depressive symptoms. Journal of Social and Personal Relationships,27(3), 367-387. doi:10.1177/0265407509348810

Patterson, J., Albala, A. A., McCahill, M. E., & Edwards, T. M. (2010). The Therapist’s Guide to Psychopharmacology: Working with Patients, Families, and Physicians to Optimize Care. New York: Guileford.

Perls, F. (1973). The Gestalt approach & eye witness to therapy. Palo Alto, CA: Science & Behavior.

Pinsof, W. M. (1993). Toward a scientific paradigm for family psychology: The integrative process systems perspective. Journal of Family Psychology5(3-4), 432-447. doi:10.1037/0893-3200.5.3-4.432

Pinto, A. &Francis G. (1993). Cognitive correlates of depressive symptoms in hospitalized adolescents. Adolescence28(111), 661–72.

Rampage, C. (1998). Feminist couple therapy: Case studies in couple and family therapy: Systemic and cognitive perspectives. In F. M. Dattilio (Ed.), Case studies in couple and family therapy: Systemic and cognitive perspectives, (pp. 353-370). New York: Guilford.

Rehman, U. S., Ginting, J., Karimiha, G., Goodnight, J. A. (2010). Revisiting the relationship between depressive symptoms and marital communication using an experimental paradigm: The moderating effect of acute sad mood. Behaviour Research and Therapy, 48(2), 97-105. doi:10.1016/j.brat.2009.09.013

Reich, G. (2003). Depression and couples’ relationship. Psychotherapeut, 48(1), 2-14. doi:10.1007/s00278-002-0280-3

Rogers, C. R. (2003). Client-centered therapy: Its current practice, implications, and theory.London: Carlton.

Sacco, W. P. & Phares, V. (2001). Partner appraisal and marital satisfaction: The role of self-esteem and depression. Journal of Marriage and Family, 63(2), 504-513. doi:10.1111/j.1741-3737.2001.00504.x

Sapolsky, R. (2009, November 10). Depression in the U. S. [Video file]. Retrieved from (2010, December 5.)

Satir, V. (1983). Conjoint family therapy. Palo Alto, CA: Science & Behavior.

Sayers, S. L., Kohn, C. S., Fresco, D. M., Bellack, A. S., & Sarwer, D. B. (2001). Marital cognitions and depression in the context of marital discord. Cognitive Therapy and Research, 25(6), 713-732. doi:10.1023/A:1012967222638

Schneider, B. L. (1998). Inpatient treatment of a gang-involved adolescent: A metaframeworks analysis. Dissertation Abstracts International: Section B: The Sciences and Engineering.

Schwartz, R. C. & Grace, P. (1989). The systemic treatment of bulimia. Journal of Psychotherapy & the Family, 6(3-4), 89-105. doi:10.1300/j085V06N01_09

Schwartz, R. C. (1987). Working with “internal and external” families in the treatment of bulimia. Family Relations, 36(3), 242-245. doi:10.2307/583534

Segrin, C. (2004). Concordance on negative emotion in close relationships: Transmission of emotion or assortative mating? Journal of Social and Clinical Psychology, 23(6), 836-856. doi:10.1521/jscp.23.6.836.54802

Sher, T. G., Baucom, D. H., & Larus, J. M. (1990). Communication patterns and response to treatment among depressed and nondepressed maritally distressed couples. Journal of Family Psychology, 4(1), 63-79.

Slavich, G. M., Thornton, T., Torres, L. D., Monroe, S. M., & Gotlib, I. H. (2009). Targeted rejection predicts hastened onset of major depression. Journal of Social and Clinical Psychology, 28, 223–243. doi:10.1521/jscp.2009.28.2.223.

Strauss, W. & Howe, N. (1991). Generations: The history of America’s future, 1584 to 2069.New York: Quill.

Teichman, Y., Bar-El, Z., Shor, H., Sirota, P., & Elizur, A. (1995). A comparison of two modalities of cognitive therapy (individual and marital) in treating depression. Psychiatry 58,136–148.

Tilden, T., Gude, T., Hoffart, A., Sexton, H. (2010). Individual distress and dyadic adjustment over a three​-​year follow​-​up period in couple therapy: A bi​-​directional relationship? Journal of Family Therapy, 32(2), May 2010, 119-141. doi:10.1111/j.1467-6427.2010.00490.x

Tilden, T., Gude, T., Sexton, H., Finset, A., & Hoffart, A. (2010). The associations between intensive residential couple therapy and change in a three​-​year follow​-​up period.Contemporary Family Therapy: An International Journal, 32(1), 69-85. doi:10.1007/s10591-009-9104-8

Timko, C., Cronkite, R. C., Swindle, R., Robinson, R. L., Turrubiartes, P. & Moos, R. H. (2008). Functioning status of adult children of depressed parents: A 23​-​year follow​-​up. Psychological Medicine, 38(3), 343-352. doi:10.1017/S0033291707002073

Unutzer, J., Katon, W., Callahan, C. M., Williams, J. W., Hundeler, E., Harpole, L., Hoffing, M., Della Penna, R. D., Hitchcock Noel, P., Lin, E. H. B., Tang, L., & Oishi, S. (2003). Depression treatment in a sample of 1,801 depressed older adults in primary care. Journal of the American Geriatric Society, 51, 505-514.

Vaeth, P. A. C., Ramisetty-Mikler, S., Caetano, R. (2010). Depression among couples in the United States in the context of intimate partner violence. Journal of Interpersonal Violence, 25(5), 771-790. doi:10.1177/0886260509336957

Vilhjalmsson, R. (1993). Life stress, social support and clinical depression: A reanalysis of the literature. Social Science & Medicine, 37(3), 331–42. doi:10.1016/0277-9536(93)90264-5.

Wampold, B. E. (2001). The great psychotherapy debate: Models, methods, and findings.Mahwah, NJ: Erlbaum.

Whisman, M. A., Schonbrun, Y. C. (2010). Marital distress and relapse prevention for depression. In R. C. Steven & M. G. Perri (Eds.), Relapse prevention for depression (pp. 251-269). Washington, DC: American Psychological Association. doi:10.1037/12082-010

Whisman, M. A., Weinstock, L. M., & Uebelacker, L. A. (2002). Mood reactivity to marital conflict: The influence of marital dissatisfaction and depression. Behavior Therapy, 33(2), 299-314. doi:10.1016/S0005-7894(02)80030-2

White, M. & Epston, D. (1990). Narrative means to therapeutic ends. New York: Norton.

Whitton, S. W., Olmos-Gallo, P. A., Stanley, S. M., Prado, L. M., Kline, G. H., St. Peters, M., & Markman, H. J. (2007). Depressive symptoms in early marriage: Predictions from relationship confidence and negative marital interaction. Journal of Family Psychology, 21(2), 297-306. doi:10.1037/0893-3200.21.2.297

Wilber, K. (2000). Integral psychology: Consciousness, spirit, psychology, therapy. Boston: Shambala.

Wilkins, E. J. (2007). Using an IFS informed intervention to treat African American families surviving sexual abuse: One family’s story. Journal of Feminist Family Therapy, 19(3), 37-53. doi:10.1300/J086v19n03_03

Wu, J. C. & Bunney, W. E. (1990). The biological basis of an antidepressant response to sleep deprivation and relapse: Review and hypothesis. American Journal of Psychiatry, 147(1), 14-21.

Wynne, L. C. (1988). An epigenetic model of family processes. In C. J. Falicov (Ed.), Family transitions: Continuity and change over the life cycle. New York: Guilford.

Yalom, I. (2009). The gift of therapy: An open letter to a new generation of therapists and their patients. New York: Harper Perennial.


Mind Sequence Organization Development Culture Gender

Figure 1. The levels of organization as conceptualized by Breunlin, Schwartz, and Mac Kune-Karrer (1997) are on the vertical axis and the metaframeworks are on the horizontal axis.

Intervention at the Level of Systems versus Individuals

[First published on Nathen’s Miraculous Escape, December 8, 2009.]


I consider the theoretical conversation about intervention at the level of systems versus individuals, including systems holarchies, pathology, the instrument fallacy (“When you have a hammer, everything looks like a nail.”), and the fit of evidence.  I include an argument for family-systems interventions as both the underdog and the eventual winner.

Intervention at the Level of Systems versus Individuals

Entering into the field of psychotherapy at this moment in history means coming to grips with an ongoing conversation or debate between those who favor intervening at the level of systems, usually families, and those who favor intervening at the level of individuals.  The conversation is mostly one-sided right now because of the history of psychotherapy.  Psychotherapy arose under the western European enlightenment, with a modern value structure and philosophy.  This means it was formed on the foundation of Christianity, reductionist science, and medicine, and so focuses on the individual using the modern lens of linear causality.  It was also formed during a time which had only very primitive understandings of psychology, communication, and statistical modeling.  Family therapy and systems-oriented interventions flowed from the more recent scientific and philosophical developments of general systems theory and postmodern philosophy (see Becvar & Becvar, 2006, or Nichols & Schwartz, 2008 for descriptions of this development), with at least two unfortunate results.  First, those in the dominant paradigm of individual psychotherapy and especially psychiatry have been able to largely ignore the arguments and breakthroughs of family systems thinkers.  Second, the arguments of the family systems paradigm remain a bit reactive and dualistic, along the lines of, “individual interveners have missed the boat, because problems simply do not exist on the level of individuals.” That, of course, is an oversimplification for the purpose of a strong thesis.  In fact, the individual and family are convenient points of intervention on a larger continuum from biological to public policy, and it seems to remain to be seen which of these sets of tools provides the most effective intervention for various kinds of human suffering.  In part this is because the toolboxes of each intervention-style camp come part and parcel with a set of philosophical information filters that make the tool wielder right about the question at hand, as if the tools in your toolbox actually shape the question into a problem best handled by your system of intervention.  It is also in part because the available modes of observation and evaluation, the science and statistics of very complex systems, are still in their infancies, because they are both new and contain very difficult problems.  Because of all of this, it can be difficult for the various factions to even agree on adequate outcome variables.

A Holarchy of Systems

One way to conceptualize the nature of systems is as a holarchy, a series of levels of increasing complexity in which each, new, more complex level of systems both transcends and includes all of the previous levels (see e.g. Wilber, 1996).  Families, for example, are not just more complex systems than individual humans, they are a more complex kind or level of system.   Families are made up of individuals and their interactions.  They are systems of individuals.  Individuals could (conceivably, at least for a time) exist without families, but families would immediately cease to exist without individuals.  Therefore, families represent a higher order of complexity than individuals do, both transcending and including them.

Individuals are systems too, of course, made up of biological systems like organs and tissue, which in turn are made up of chemical systems, each level of which both transcends and includes all of the simpler levels of systems.  There are levels of systems more complex than the family, as well; cities, for example, or countries, perhaps even the biosphere (see e.g. Capra, 1997).  Even though this is a big oversimplification, it can be a useful way to think about things: systems all the way up and all the way down.

So when we consider where pathology exists, or where best to intervene, the question is not really whether to intervene at the level of individuals or at the level of systems, but rather at what level of systems does pathology exist and at what level of systems will it be most effective to intervene.  It is only a matter of convention that by “individual” we tend to mean individual humans, and by “system” we tend to mean families and up.  Additionally, it is a convention of family therapists to mean “family system” when we say “system,” as that is our area of primary interest.

What and Where is Pathology?

“Pathology” is a dirty word in holistic, post-modern circles, which is understandable.  It smacks of the medical model, with its simplistic, linear causality.  It has also been used to justify an incredible amount of oppression and suffering.  After all, who gets to say what is normal and who is sick?  “The man” does.  Family therapists tend to use the word “problem” instead of “pathology” and let themselves off the pathologizing-hook by allowing the client to define the problem: The problem is whatever the client is complaining about, and we are given permission to intervene by their coming to us for help (e.g. Fisch, Weakland, & Segal, 1982).  This is a bit disingenuous, however, because outside of the strict behavioral camp, family therapists tend to believe that the real problem exists meta to the complaint.  That is, it exists in the level of system above the complaint, in the outmoded rules of the family, for example, or in dysfunctional patterns of communication.  This allows us to not pathologize individual humans, but we are still saying that something is not working right, and that is what other disciplines call pathology.

The real question about pathology is where it exists, at what level of complexity.  It is possible that a problem could exist at just one level of complexity—in the family system, for example, or in an individual’s neurochemistry.  It is also possible that a problem could exist in multiple levels simultaneously, or even that a problem at one or two levels could cause symptoms at other levels that look like problems.  This gets us into an epistemological morass, because we cannot actually see problems.  We can see patterns and behaviors, and we can hear complaints, and we can notice when those behaviors and complaints change, but that tells us little about what level the actual problem exists at.  For this we have to rely on theory.

Intervention and the Problem of Having a Hammer

A theory is a story about the underlying nature of reality.  It explains things we can see in terms of concepts that we cannot see, in terms of ideas.  An ideally rigorous scientist never believes their theory, seeing it instead like a constructivist does: the best map I have made or come across so far for the territory of reality.  Real scientists, and other real people, like therapists, who imagine they see their theories be useful over and over, come to some degree to mistake their map for the territory; my map, not yours, shows the real lay of the land, and if we are talking about problems, my map shows where the root of the problem is.

The trouble is, theoretical maps tend to be very subtle and sophisticated representations of only one level of system.  They tend to have some idea of the levels just above and just below, but in terms of the language of their level.  A therapist who specializes in individuals, for example, is trained to see how problems exist at the level of individual humans.  They might know that depression, for example, is correlated with certain kinds of brain chemistry, and maybe even certain kinds of patterns of interaction in the family—they may even know that depression is more common for people in certain economic situations—but they will tend to see these sub- or extra-individual phenomena as symptoms of the real problem, which lies in the individual—perhaps in a traumatic past, psychodynamic factors, or lack of insight into their own thought processes.  From this standpoint, giving an individual drugs or intervening with their family seem to be treating the symptom rather than the problem.

If you have a hammer everything looks like a nail, and if you have a theory every problem can look subject to intervention at your favorite level of system.  Also, if you know where problems happen, you know where they do not happen.  People who specialize in one level of system often have the overt belief that if everything is going well at their level, the other levels can take care of themselves.  Family therapists tend to see individual psychological health as a given in the context of a well-functioning family (see e.g. Bateson, Jackson, Haley, & Weakland; Bowen, 1960; Jackson, 1998; Minuchin, 1998; 1963 Satir & Baldwin, 1983 ).  For political thinkers, who tend to see problems in higher-level systems, at the level of public policy interventions, making people happier by any means lower than institutional-level interventions is just perpetuating the problem; of course people are depressed! We are living in an unjust political or economic system (famously, e.g., in Marx & Engels, 1967).  Giving them happy drugs, intervening in their family systems—these are just ways of maintaining the status quo.  It is different, of course, with political thinkers whose preferred system is in place.  In that case, the problem is obviously at any other level of system.  Medical thinkers have a similar situation, only from a bottom-up perspective, instead of top-down; depression is the result of a chemical imbalance, pathology of the tissue of the brain.  Happy drugs are not treating the symptom, they are treating the root of the problem, and the symptoms of discomfort at the level of the individual or of strife at the level of the family will just go away (see e.g. Watters & Ofshe, 1999).  Many psychological “diseases” from depression to schizophrenia tend to run in families, too, so medical-thinkers also blame genetics (Carlson, 2005).

It is noteworthy that being a systems thinker does not necessarily ameliorate this problem.  Our theories are much too flexible for that.  Systems-hip people with bottom-up oriented theories can say that they are changing the system using their individual intervention, and they probably are, to some degree.  The nature of systems is one of reciprocal causality; if you change the output of one node in the network, you will theoretically change the output of every other node, and therefore the functioning of the entire system.  It is rare even for family therapists to insist on seeing the whole family anymore (Nichols & Schwartz, 2008).  While he preferred to see families, Murray Bowen, for example, saw individuals for a large part of his practice, and thought that this could “be the fulcrum for changing an entire family system (Nichols & Schwarts, 2008, p. 145).  Some family therapists consider seeing only the complainant as ideal, and in no way contradictory to systems principles (Fisch et al., 1982).  In that light it is difficult to argue that medical or individual interventions are not also interventions at the level of systems, as long as they change the client’s behavior.

The Problem With the Evidence

Can we settle this by looking at the evidence?  Maybe someday, but not today.  The main problem is that our science has not caught up to even Freud’s thinking and style of intervention, much less to the leap into complexity that is systems thinking and intervention.  Up until the last couple decades, even if we could have collected the data we needed, we could not have analyzed it with the primitive statistical modeling available.  Even now we are left with a serious conceptualization problem and a massive data-collection problem.

In order to study things scientifically, whether it is gravity or psychological health, we need to operationalize our latent construct of interest.  That means, since we cannot directly observe gravity, our construct of interest, we have to decide on something we can directly observe to act as a proxy.  We could measure how long it takes an apple to fall ten feet, for example, or measure how hard it hits the ground from various heights.  From that data, we can make inferences about gravity.  For psychological health, it is the same but more difficult.  We cannot see it, so we have to choose a proxy.  That proxy usually turns out to be some form of asking people how they feel.  Sometimes we use the number of rehospitalizations or type of diagnosis, but most often we give the person a survey such as the ubiquitous Beck Depression Inventory, asking them what they have been doing and how they have been feeling lately.

These kinds of measures we have are best suited to test the success of medical-level interventions.  If you give someone a drug and they say they feel better, they are better! What more do you need to know?

Those who think of problems existing at the level of the whole individual do not have it quite so good.  The reason is that, while asking someone if they feel better or checking whether or not they have been rehospitalized is a decent operationalization of individual health, it cannot distinguish between the “real” health that an individual earns permanently through the work of therapy, and the “fake” health handed to them by externally manipulating their brain chemistry.   An individual therapist who looks at data supposedly showing that fluoxetine is in any way comparable to therapy (e.g. TADS, 2004) must chafe.  The client has not learned anything!  What can this be but a band-aid?

Family therapists have it even worse.  Anything short of a family-wide, long-term operationalization of health is inadequate if you believe that symptoms such as depression are just a way that the family system is attempting to maintain its homeostasis, staying within its outmoded parameters.  If you “fix” one person in the system without altering the rules of the system, someone else is bound to become symptomatic.  You need second-order change, change at the level of the rules, rather than first-order change.  But how do you operationalize second-order change?  How do you measure the rules of the system, which play out in the complex interactions between body language, spoken language, and emotional tone, between  interlocking triangles of relationships?  Well, you cannot.  Family-systems thinkers are left in the uneasy state of relying on individual outcome measures to judge the success of our work.

In a way, those who prefer intervention at higher levels than families have an even harder time with outcome measures.  These systems are almost impossibly complex—whole cities, whole countries—how to operationalize the health of a county?  Public-policy interveners tend to be idealogs, too, which makes things more complex still; your version of health will depend on the economic or political system you prefer.  Still, these problems may not bother public policy types.  Ideological immunity, the tendency or ability to completely disregard information that contradicts your belief system, may limit their interest in evidence, except where it serves their purposes.

But perhaps there is a sort of levels-of-systems-wide equifinality.  Perhaps intervening at any level produces the same kinds of results.  This is what the small amount of evidence that has been gathered seems to suggest, though it is all based on individual-outcome measures (Sprenkle, 2002).  So perhaps there is no such thing as a system-level-specific pathology.  Is it even possible to determine the level at which a pathology exists?  If so, would there be a way to determine whether it was better to intervene at that level, or a level above or below?  It is difficult to imagine the kind of nightmare-monster experiment that could answer those questions to everyone’s satisfaction; a longitudinal study, spanning the entire lifespan of multiple generations, collecting rich data on every complaint, and the timing of every complaint, of every member of hundreds of families of different ethnicities, existing under different political and economic systems, using different levels of intervention, and controlling for nutrition, exercise, values structures, belief systems etc ad nauseum.  It would be fairly expensive, and somewhat difficult to analyze the data.  Modern research tends to shy away from even relatively simple complexities, such as comorbidity, where an individual is diagnosed with more than one pathology (e.g. TADS, 2004).

There is some anecdotal evidence, but it is, well, anecdotal.   Reading biographies of therapists (e.g. Family therapy pioneers, 2008), for example, gives the impression that many people left the ranks of individual intervention to work with families because individual work was too slow, or ineffective, while there are no famous cases of family therapists who became individual therapists because the work was too shallow.  This could be evidence that individual therapy is in fact inefficient or ineffective, but it could simply be the result of individual therapy having been formulated first; many, many times the number of people have been trained as individual therapists compared to family therapists—almost infinitely more, back in the day that the pioneers of family therapy were emerging.

Don Jackson, one of the founding fathers of family therapy, claimed to have noticed regularly in his individual therapy practice that when he helped solved a problem for one client, someone else in the family would develop a problem (Jackson, 1954, 1965).  That this is a reason to intervene systemically is a commonly held belief among family therapists, but virtually unknown among more bottom-up helping professions, and there appears to be no published empirical evidence on the problem, and none may be on the way.  It seems likely that family-systems thinkers will have to continue to rely to a large degree on our intuition about the legitimacy of our theories and the applicability of research based on individual outcome measures and the validity of relatively anecdotal evidence from our founders and mentors.

An Argument for Family-Level Interventions as Both the Underdog and the Eventual Winner

For Judeo-Christian culture, the individual is the obvious focus of attention.  The individual is the seat of the soul, the source of the will, and the unit that can be judged innocent or guilty by the courts, and will be judged saved or damned by God when that day comes.  It is only the individual who seems to have consciousness and a linear narrative of life.  It is the individual who has nerve endings and can complain about them.  Who was it that said, “Ouch—I have a pain in my family”?  I forget, but I bet it was a family therapist, because it is simply not intuitive to do so.  Western ethical and philosophical systems all focus on the individual: What is it right or wrong for you to do?  In what fashion can you live the best life? When systems are considered, we jump right up to the level of government and public policy, and even then, they are primarily concerned with the relationship of the state with individuals: What are the individual’s rights? When can the state take an individual’s property?

The development of modern science in the enlightenment began with reductionism, in the form of chemistry, linear causality, in the form of Newtonian physics, and a dualistic model of health, from medicine, which were and remain incredibly powerful and successful forces for gaining knowledge.  In part, this is because the questions that chemists and physicists ask are simple ones: What are the basic laws of nature? What is this thing made out of?  The kind of science that could grapple with more complex topics such as the nature of communication, or the function of the behaviors of the “mentally ill” were for centuries, and to some extent still are, far beyond our capacity for scientific observation, data collection, and statistical modeling.

Because of all of this, family-systems theory and its foundations, postmodern philosophy, general systems theory, and communication theory, are late to the table and sitting across from respected, entrenched adversaries.  Postmodernism thoroughly discredits the idea that an observer can be objective.  Systems theory makes taking things apart and using idealized, linear models seem quaint and often misguided.  Communication theory holds that the difference between a schizophrenic and a non-schizophrenic may be the relationship between spoken language and body language in that person’s family—a very difficult thing to quantify.  Family-systems theory would have us believe that, while we might be able to drug ourselves into happiness, we can’t drug our way to health, because the real problems are in the patterns of communication of the family.  These ideas are still all quite counterintuitive from the modern mindset.

Still, I imagine that the newbies will be vindicated eventually.  First, they are more sound.  Postmodern philosophy is simply more accurate than modern philosophy; that we have no direct access to anyone else’s experience or to any other piece of reality has ramifications that modern philosophy simply ignores.  Systems theory is truly a breakthrough in science, and very promising; it takes into account the way complex systems are more than the sum of their parts in a way reductionism is simply not geared to.  Communication and family-systems theory are ways of looking at health and pathology from a more accurate perspective; medical thinking is too black-and-white, health-or-pathology, and too reductionistic.  Treating the symptom with drugs or surgery is simply not always the answer.  Objects in the frame cannot change the frame (Watzlawick, Bavelas, & Jackson, 1967), that is the allopathic fallacy.  Second, if psychological problems really do exist in the context of systems, in patterns of communication of individual humans, then we don’t have to kill the patient to look inside it; we can observe the problem in live action without cutting anyone open.  The guts of the problem are right there in front of us, in the therapy room, in the interactions of the family.  That is a huge advantage over people who think that problems exist in individual brains, which must be either dead, in grave danger, or isolated in a very expensive machine fMRI machine which still cannot see much.  Third, even though we have to use inadequate, individual outcome measures to test our work, we are already coming out on a par with individual interventions (Sprenkle, 2002).  Imagine how we will do comparatively when outcome measures regularly consider the health of the whole family!  Last, we are just getting started.  Freud lived 150 years ago, and his ideas took off immediately.  They have had time to mature.  Family systems theory is barely 50 years old and has gotten a relatively slow start.  Who knows what we will be able to accomplish by the time our field is mature?

And Yes, Individual Versus Systems Psychotherapeutic Interventions Look Quite Different

The first psychotherapeutic intervention to take off was Freud’s psychoanalysis.  Over the course of years, patients (they still called them “patients”) would explore their past and especially their unconscious mind, increasing their insight, using free association and dreams, under the direction of the expert analyzer, the therapist.  The patient would come to think of the analyst as their father and eventually get over that idea, in the process transforming from pathologically neurotic to mildly neurotic.

The second great force in psychotherapy was behaviorism, which was largely a reaction to the heavily subjective, unquantifiable theories and process of the Freudians.  Behaviorists ignored what might or might not be going on in people’s minds and concentrated on what they did.  One of them noticed that dogs could be conditioned with rewards to salivate to a bell and a series of interventions based on that insight were created.  If you want to increase a behavior, you either reward or stop punishing the organism when they perform it.  If you want to decrease a behavior, you either punish or stop rewarding the organism when they perform it.  Simple!

The third great force was humanism.  The humanists did not like how the behaviorists treated humans like just another animal.  They also did not like how the Freudians dwelled so much on the past, pathology, and intellectual analysis.  They had a positive, emotional spin.   Humans were naturally healthy, not naturally sick.  Mildly neurotic was shooting too low.  They shot for self-actualization.  Psychological problems came from repressing emotions and not being understood.  They also believed in the talking cure, but they talked about what was happening right now, in the moment.  Their interventions looked like a series of interviews in which the client and therapist formed and explored an intense emotional relationship, and in which the therapist coached the client to be totally genuine.

The systems approach (the fourth great force?) was a reaction against all individual therapies.  It can be seen as a resurgence of behaviorism to some degree, in that it treated individuals’ minds as black boxes, at least in the early days (Watsalwick et al., 1967), but always with a systems, family-oriented view.  Later, some models incorporated humanist principles (e.g. Baldwin & Satir, 1987; Whitaker & Keith, 1982).  There are a great variety of systems interventions.  The only constants among them is that they concentrate on process over content—that is, they pay more attention to the rules of the conversation than the direct messages in the conversation—and they will always at least consider intervening with more than one family member at a time.  Therapists in the strategic school tend to use paradoxical homework as interventions—prescribing the symptom, for example.  Bowenian therapists also use homework in the form of reconnecting with cutoff family members.  Experiential therapists use the relationship with the therapist in the humanist style, plus coaching authentic communication.  Structural therapists also use homework to re-establish healthy boundaries between parents and children, going on a secret date, for example.  All of these are methods of causing an updating of the rules of a family system.

A Personal Reflection

I am happy to find myself fascinated by this topic—deeply and richly confused, as a friend of mine likes to say.  I also feel a little embarrassed to have written so much after reading so little, which is to say that ten weeks is not much time to make oneself an expert in such a deep topic!  That, and I feel grateful that there is so much to read—that so many very smart people have thought so much about how best to help people.  In the end, I still don’t know the answer to any of the questions I came across here.  But I have chosen my theoretical stance anyway.  The systems view is my hammer, and I am learning to swing it the best I can.


Baldwin, M. & Satir, V. (Eds.) (1987). The use of self in therapy. New York: Hawthorne.

Bateson, G., Jackson, D. D., Haley, J., & Weakland, J. (1963). Toward a theory of schizophrenia. In N. J. Smelser & W. T. Smelser (Eds.) Personality and social systems (pp. 172-187). Hoboken, NJ: John Wiley & Sons.

Becvar, D. S. & Becvar, R. J. (2006). Family therapy: A systemic integration. Boston, MA: Pearson.

Bowen, M. (1960). A family concept of schizophrenia. In D. D. Jackson (Ed.) The etiology of schizophrenia (pp. 346-372). Oxford, England: Basic.

Capra, F. (1997). The web of life. USA: Anchor

Carlson, N. R. (2005). Foundations of physiological psychology. Boston: Pearson.

Family therapy pioneers. (2008, September/October). Family Therapy Magazine, 7(5), 23-60.

Fisch, R., Weakland, J. H., & Segal, L. (1982). The tactics of change: Doing therapy briefly. San Francisco: Josey-Bass.

Jackson, D. D. (1954). The question of family homeostasis. Psychiatric Quarterly Supplement, 31, 79-90.

Jackson, D. D. (1965). The study of the family. Family Process, 4(1), 1-20.

Marx, K. & Engels, F. (1967). Manifesto of the communist party. New York: Penguin.

Minuchin, S. (1998). Structural family therapy. In D. M. Frank (Ed.) Case studies in couples and family therapy: Systemic and cognitive perspectives (pp. 108-131). New York: Guilford.

Nichols, M. P. & Schwartz, R. C. (2008). Family therapy: Concepts and methods. Boston: Pearson.

Satir, V. & Baldwin, M. (1983). Satir step by step: A guide to creating change in families. Palo Alto, CA: Science and Behavior.

Sprenkle, D. H. (2002). Effectiveness research in marriage and family therapy. Alexandria, VA: AAMFT.

TADS (2004). Fluoxetine, cognitive-behavioral therapy, and their combination for adolescents with depression. JAMA, 292(7), 807-820.

Watters, E. & Ofshe, R. (1999). Therapy’s delusions: and the exploitation of the walking worried. New York: Scribner.

Watzlavick, P., Bavelas, J. B., & Jackson, D. D. (1967). Pragmatics of human communication: A study of interactional patterns, pathologies, and paradoxes.

Whitaker, C. & Keith, D. V. (1982). Symbolic-experiential family therapy. In J. R. Neill & D. P. Kniskern (Eds.), From psyche to system: The evolving Therapy of Carl Whitaker (pp. 330-378). New York: Guilford.

Wilber, K. (1996). A brief history of everything. Boston: Integral.

Experiential Family Therapy: The Humanistic Family Therapy Model

By Nathen B. Lester, November 2009


I examine experiential family therapy in its philosophical context, its relationships with Becvar and Becvar’s (2006) core family systems theory assumptions, Watzlavick, Bevelas, & Jackson’s (1967) communication theory, and my own values system and beliefs about change.

Experiential Family Therapy: The Humanistic Family Therapy Model

Experiential family therapy is the meeting place of the humanistic psychology and therapies of the likes of Abraham Maslow, Carl Rogers, and Fritz Perls, and the family systems and communication theories of the likes of Gregory Bateson, Don Jackson, and Paul Watzlavick. It might be more accurate to say “meeting places,” however, as it clear that experiential family therapy has never been a single coherent system for conducting therapy. Becvar and Becvar, for example, tend to refer to “experiential approaches to family therapy” (e.g. 2006, pp. 158-159), rather than an “experiential model” and focus on individual practitioners rather than a general theory or set of interventions. It is also telling that they place Virginia Satir, one of the two major progenitors of experiential approaches to family therapy, in their chapter on communication approaches, rather than that on experiential approaches. There are at least three major branches of the family systems family tree – Carl Whitaker’s symbolic-experiential approach, Walter Kempler’s Gestalt-experiential approach, and Virginia Satir’s communications approach (Becvar & Becvar, 2006) – but experiential therapists’ focus on the unique self of the therapist gives the impression that there may be as many models as therapists (e.g. Baldwin & Satir, 1987; Napier & Whitaker, 1978), and the focus on the unique experiences and responses of both clients and therapists, from moment to moment, can give the impression that there are infinite variations.

In order to accommodate my newness to this subject and to write a reasonably short summary, however, I will treat experiential family therapy as one school of thought and use “experiential family therapy,” “experiential family therapists,” “experiential approaches,” and “the experiential model” somewhat interchangeably.  I will certainly miss many of the subtleties and may overemphasize the work of the importance of Satir, whose work has produced more literature, but these appear to be necessary sacrifices in order to concisely explore experiential family therapy in its historical and philosophical context.

Humanistic Psychology, Family Systems Theory, and Experiential Family Therapy

The humanistic, experiential, or “third force” of psychology and psychotherapy arose primarily in reaction to the way the second force, behaviorism, completely disregarded the inner experience of individuals and focused only on easily quantifiable behaviors. And while carrying forward the psychodynamic ideas of the first force – psychoanalysis – humanistic therapists were also reacting against the Freudian emphasis on the client’s past, analysis, and intellectual understanding.  The way that all of the various humanistic therapies were “experiential” was in the belief that therapeutic change occurs in the honest and accurate self-expression of the experience of both the therapist and client, in the here-and-now of the session, in the context of their genuine, empathic, therapeutic relationship (Greenburg, Watson, & Lietaer, 1998).  Humanistic therapies also tended to share beliefs in the value of self-actualization, the reaching of human potential, and in the natural ability and tendency of individuals to achieve it (Nichols & Schwarts, 2008).

In some ways, family systems theory can be seen as a resurgence of behaviorism, with its emphasis on observable behavior and insistence that the inner workings of another’s mind is unknowable, and best treated as a “Black Box” (e.g. Watzlavick, Bevelas, & Jackson, 1967, p. 43).  For the most part, however, family systems theory was an attempt to apply general systems theory to family systems (e.g. Jackson 1957, 1965); it improved on behaviorism by observing and theorizing about complex, interpersonal behaviors from the context of systems theory’s “reciprocal causality,” by using Russell’s theory of logical types (Watzlavick et al., 1967), and a postmodern, constructivist epistemology. In other words, they asked “How shall we best help people, now that we understand that their problems lack linear causality, are dependent on their interpersonal contexts and the confusion that can occur in the entanglement of multiple, contradictory levels of communication, given that we can only view and think about families through our limited, personal, subjective perspectives?” The answer, they decided, is that we can  help best by intervening at the level of family systems – humbly, without pathologizing – using paradoxical and structural interventions. That is, they refused to see individuals as sick or the cause of problems, focusing instead on the family system, tweaking or tricking it to force outmoded rules into adapting.

Those who ended up calling themselves experiential family therapists agreed with the family system theorists that families got stuck in inflexible, old rules, in the family-systems sense. They agreed that the family system was the ideal level of intervention, as it was the family that could either facilitate or disallow its members’ self-actualization. But they gave equal weight to the experiences –  especially the emotional experiences – of the therapist and of the individuals in the family. They took the humanist ideas that individuals tend towards, deserve, and can achieve self-actualization, and that learning and change happen best in the context of here-and-now experiencing in the context of a non-judgmental, compassionate relationship, and added it to the family-systems idea that families can fail to adapt to changing circumstances, or habitually communicate in self-contradictory ways, producing symptomatic behavior in their members in an unconscious attempt to maintain outmoded self-definitions.  Family systems therapists also concentrated on the behavior of the system and the alleviation of the symptom to the extent that they could sometimes justify bending the truth, or even outright lying to family members (e.g. Fisch, Weakland, & Segal, 1982; Burbatti & Formenti, 1988); the ends justify the means. This is in stark contrast to the standard of authenticity that experiential therapists hold themselves to.

The Experiential Model and Core Family Systems Theory Assumptions

Systems theory, one of the major philosophical foundations of all modern family therapies, came about in part as a reaction to traditional, linear and reductionist systems of thought and investigation.  Taking things apart (e.g. chemistry, anatomy) and thinking in terms of simplified, idealized situations (e.g. Newtonian physics) have always been powerful methods for gaining knowledge, but systems theorists saw that they were also quite limited because in natural, complex systems, the whole is greater than the sum of its parts and causality tends to exist in loops, patterns, and patterns of loops, rather than straight lines. These qualities tend to be invisible from the pre-systems perspective, so therapeutic models developed before systems theory tended to have the linear, reductionist, medical model as a philosophical foundation. Family therapy models, on the other hand, tend to be more closely related to systems theory, and each model has its own relationship with the tenets of systems theory.

Family systems theory, as opposed to general systems theory, also incorporates elements of postmodern philosophy. Postmodernism was, like systems theory, in part a reaction to traditional systems of thought and investigation, especially in how traditional thinkers assumed objectivity on the part of the observer. Postmodernists assert that since no one is objective, the closest we can come to the truth is perspective and opinion – meaning we must be humble about what we think we know.

In the subsections below, I will discuss the experiential family therapy model in light of the core family systems theory assumptions listed by Becvar and Becvar (2006, p. 8), which are an amalgam of general systems theory and postmodern philosophical tenets applied to families.  I will use the term “traditional” below to refer to approaches that predate systems thinking and postmodernism, such as modernist or positivist approaches, as well as older, myth- or theology-based approaches.

Asks, “what?” and here-and-now focus. Traditionally, the important question to answer was “why?” Insight into causality was considered the most powerful tool for therapeutic change. In family systems theory, the focus shifted to “what?”  This was ostensibly because “why?” was unknowable, but in practice, “why” was assumed — the family system is failing to adapt — so “what?” became the pertinent question; seeing the way in which the problem perpetuated itself – in the present, in the room with the therapist – was now the most powerful tool. Experiential therapists concur with this, both because of their systemic understanding that problems are maintained by the behavior of the family system, and because of their humanistic focus on the quality of experience in the moment.

Reciprocal causality and patterns. Family systems thinkers believe that traditional thinkers see linear causality in the world because of a mistake in punctuating events. The classic example is the husband who withdraws “because” his wife nags.  Instead of punctuating before the wife nags and after the husband withdraws, a systems thinker sees a reciprocally-causal pattern of events, in which the husband and wife are participating in an interactive pattern of nagging and withdrawing. Experiential therapists follow the systems model here, in seeing the shared influence but also introducing a kind of blame-free personal responsibility for one’s emotional states and behaviors with the exception of symptomatic behavior in children, which is seen as a somewhat unidirectional result of communication problems of the parents (Becvar & Becvar, 2006; Nichols, 2008).

Wholistic. While traditional thinkers are reductionistic, seeing problems in individuals or even the psychodynamic or biological parts of individuals, family systems thinkers prefer to see problems in the context of the “whole” family system. Experiential family therapists may be more rigorous about this than those of most other models. Whitaker, for example, would often refuse to see clients without their families, and preferred to have three generations come in (Whitaker & Keith, 1982). Experiential therapists’ inclusion of emotion as an important aspect of therapy (Satir & Baldwin, 1983; Nichols & Schwartz, 2008) can also be seen as more “wholistic” than some other family therapists, who focus only on behavior and cognition.

Subjective/perceptual. One of the great overestimations of traditional science was the belief that humans could be objective in their gathering knowledge. This misconception was set straight by postmodern philosophers such as Von Glasersfeld (1984), rather than systems theorists; an individual’s understanding is limited and colored at every step of the process, from perception to description, by factors of which they are not aware. Experiential therapists align with the postmodernists here, focusing on and trying to communicate – as authentically (as opposed to objectively) as possible – their own subjective experiences, rather than on “knowing” or the illusion of observing from the outside.

Relational and contextual. Traditionally, problems were seen as pathology in an individual, or in a psychodynamic or biological part of an individual, but family systems theory came to see problems as existing in relational patterns, in communicative behavior and in the individual’s social context, usually the family. This tenet is strongly upheld by experiential therapists too; if individuals are naturally good and healthy in humanistic psychology, it has to be the context they exist in that produces symptomatic or problem behavior.

Relativistic and dialectical. Modern science overcame the absolutism of the premodern era to a great degree, but a vestige remained in the form of a belief in the reality of the conceptual categories “discovered” by scientists, and in the implied truth of theories which had gathered some supportive evidence.  Postmodernism came to see all opinions and ideas as part of a dialog between different perspectives, and thus not related to each other in hierarchical fashion – true versus false, or even more informed versus less informed – but existing side by side, equally valid.  Experiential therapists follow this new tradition, giving equal weight to the experiences of all family members as well as the therapist. Experiential family therapy, while directed by the therapist, can look very much like a conversation in which the therapist and family members come to understand each other’s values and experiences (e.g. Menninger Video Productions, 1993; Golden Triad Films, 2004).

Proactive. Family systems therapists tend to be proactive in comparison to the psychoanalytic school. That is, where in psychoanalysis the therapist and client would sit together regularly for years, talking about the client’s history and aiming at a restructuring of the client’s psychodynamics through intellectual insight, family systems therapists tend to do their work briefly, in 8-10 sessions, focusing on the client’s presenting problem with relative directness (Fisch et al., 1982).  Experiential therapists are proactive compared to psychoanalysis, in that they work more briefly and use direct emotional interventions such as Satir’s family sculpting (Satir & Baldwin, 1983) or Whitaker’s emotional confrontation (Neill & Kniskern, 1982). They use straightforward coaching and reframing to cause the emotional and communication shifts they believe are called for.  On the other hand, they may appear less proactive than some other family therapy models which focus more directly on  changing a single, problematic behavior (Nichols & Schwartz, 2008).

Experiential Family Therapy and Communication Theory Assumptions

The term “communication theory” can be confusing because of its different meanings for experiential family therapists, family therapists in general, and psychologists in general.  The communication theory created by Bateson and the Palo Alto/MRI team is primarily a theory of pragmatics, or the behavioral component of communication (Watzlavick et al., 1967), where a comprehensive theory of communication would also fully address syntax, or the structure of communication, and semantics, the meaning of communication (Carroll, 2008). This reflects the emphasis in family therapy of process over content.

The major elements of communication theory as described by Watzlavick and colleagues (1967) are, (a) since all behavior is communication, it is impossible to not communicate; (b) the two basic levels of communication are the digital/verbal/content level, which conveys the dictionary-definition-of-the-words information, and the analog/non-verbal/relationship level, which conveys information about how the information sender views the nature of their relationship with the receiver; (c) relationship-level communication exchanges can be seen as attempts to assert which communicator has the power to determine the nature of the relationship; and (d) since communication almost always occurs in an ongoing social context, the punctuation of events by participants is somewhat arbitrary, and often confusing and self-serving.

Most family therapy models use these insights to understand communication and experiential therapists are no exception. Indeed, Satir was part of the MRI team for many of the years they spent formulating their communication theory. What Satir ended up meaning by “communication theory,” however, while completely compatible with Bateson and MRI communication theory, had a very different focus: ways of preventing and clearing up miscommunications, and ways of giving and receiving information and requests in accurate, functional ways (see e.g. Satir, 1967, pp. 63-90). She emphasized, for example, the problems inherent in generalizing and assuming, and in various kinds of ambiguity in communication.

I do not yet know how much other experiential therapists, such as Whitaker and Kempler, used these communications theories, but it was probably not to the extent Satir did, as even the rest of the MRI team mostly moved on to other strategies, while Satir continued to develop it (Napier & Whitaker, 1978).

Core Assumptions of Experiential Family Therapists

The core assumptions of experiential family therapists are essentially the core assumptions of humanistic psychology, adapted to a family systems epistemology: Unless held back by their environment, individual humans tend towards and self-actualization and can reach their potential. The environment is primarily the system of relationships and communication that individual exists in.  It can stunt an individual’s growth by teaching them that it is not safe to fully feel or express the experience they are having, especially in the moment they are having it.  This serves to keep individuals from understanding themselves and others as they are, and from experiencing real intimacy with others, and through that, from being able to experience their own true individuality. The remedy for this is genuine, uncensored experience, communicated accurately and authentically in intimate relationships, including the relationship with the therapist. The therapist’s role is to model this kind of psychological congruence and intimacy while coaching family members to follow suit in their own unique way. This will allow both the individuals and their system to mutually support growth to the highest levels (see e.g. Becvar & Becvar, 2004; Nichols & Swartz, 2008; Satir, 1972; Greenburg et al., 1998).

Common Therapeutic Factors and Experiential Family Therapy

For decades, the charismatic leaders of various family therapy schools distinguished their work from each other’s and advocated for the superiority of their models (Napier & Whitaker, 1978); in recent years it has become increasingly common to view all therapeutic change through the lens of the Common Factors of many models, rather than their differences (see e.g. Asay & Lambert, 1999). There is considerable evidence that the bulk of therapeutic change can be attributed to (a) the qualities and resources of clients, such as their resiliency, motivation, or community, (b) the qualities and skills of the individual therapists, (c) the quality of the therapeutic relationship, including the compatibility of the client and therapist’s objectives, (d) the client’s hope or expectancy of change, and (e) other factors such as behavioral, cognitive, and affective coaching (Sprenkle & Blow, 2004). Additionally, Sprenkle and Blow assert that family therapy has three Common Factors that individual therapy does not: a view of problems in the context of social systems, intervention at the level of systems, and multiple, simultaneous therapeutic relationships (2004).

Experiential family therapists reflect all of the Common Factors in their process. They assume that it is the unique expression of the client’s strengths that is the engine for their personal growth. They emphasize that it is the ability of the therapist to model genuine expression of their own unique strengths and to form intimate, genuine connections with the clients that is the catalyst for change (Baldwin & Satir, 1987). Their belief in and respect for human beings should be a good vehicle for encouraging hope for change in clients.   Experiential therapists offer behavioral and affective coaching as well as cognitive reframing as techniques to enhance the intimacy of relationships and accuracy of perception and expression. They view problems as residing in family systems as opposed to in individual psyches. They intervene primarily at the level of relationships and systems and form intimate relationships with each person in the system.

A Personal Reflection on Experiential Family Therapy

At this early stage in my learning about family therapy models, I imagine that I will use experiential family therapy as my primary model with clients. I remain uncertain about how good a fit the standard techniques of experiential therapists, such as sculpting, will be for me, but I intend to remain open minded until I have used them skillfully for some time.  My belief system is very closely aligned with humanistic psychology; I too believe that humans are innately good and inclined to growth, and that it is the systems of interaction and oppression we inherit that impede that process. I grew up in a holistic-oriented family. I have actively internalized the belief systems and modes of communication of two mentors who are existential therapists, one of whom was a protégé of Fritz Perls. I have practiced intimacy, honesty, and emotional fluency and fluidity in my peer-counseling and peer relationships for many years. I believe that the ability to notice my experience as it is and to express it authentically in my language, behavior, and affect, and my ability to compassionately notice the expression of others’ experience is the key to my being able to have deep relationships. I believe that the depth of our relationships is a large part of what make our lives meaningful, beautiful, and useful.

There is a way that many family therapy models, in their focus on family systems, treat the individuals in those systems with less than full respect, actually lying to clients in some situations, to trick them into improving (e.g. Fisch et al., 1982) or, short of that, treating them as more or less fortunate cogs, Black Boxes, in their family systems.  I love the way Virginia Satir, especially, is not manipulative, except in her compassionate reframing. I appreciate the way experiential therapists recognize the uniqueness and lovable-ness of each individual, and the way they view that recognition as the catalyst for positive change.


Asay, T. P., & Lambert, M. J. (1999). The empirical case for the common factors in therapy: Quantitative findings. In M. A. Hubble, B. L. Duncan, & S. D. Miller (Eds.) The heart and soul of change: What works in therapy (pp. 23-25). Washington, DC: American Psychological Association.

Baldwin, M. & Satir, V. (1987). The use of self in therapy. New York: Hawthorn.

Becvar, D. S. & Becvar, R. J. (2006). Family therapy: A systemic integration. Boston, MA: Pearson.

Burbatti, G. L. & Formenti, L. (1988). The Milan approach to family therapy. Northvale, NJ: Jason Aronson.

Carroll, D. W. (2008). Psychology of language. USA: Thompson Wadsworth.

Fisch, R., Weakland, J. H., & Segal, L. (1982). The tactics of change: Doing therapy briefly. San Francisco: Josey-Bass.

Golden Triad Films. (Producer). (2004). Blended family with a troubled boy [DVD].

Greenburg, L. S., Watson, J. C., & Lietaer, G. (1998). Handbook of experiential psychotherapy. New York: Guilford.

Jackson, D. D. (1957). The question of family homeostasis. Psychiatric Quarterly supplement, 31, 79-90.

Jackson, D. D. (1965). The study of the family. Family process, 4(1), 1-20.

Menninger Video Productions. (Producer). (1993). Virginia Satir: The use of self in therapy [VHS].

Napier, A. Y. & Whitaker, C. A. (1978). The family crucible. New York: Harper & Row.

Neill, J. R. & Kniskern, D. P. (1982). From psyche to system: The evolving therapy of Carl Whitaker. New York: Guilford.

Nichols, M. P. & Schwartz, R. C. (2008). Family therapy: Concepts and methods. Boston: Pearson.

Satir, V. & Baldwin, M. (1983). Satir step by step: A guide to creating change in families. Palo Alto, CA: Science and Behavior.

Satir, V. (1967). Conjoint family therapy: A guide to theory and technique. Palo Alto, CA: Science and Behavior.

Satir, V. (1972). Peoplemaking. Palo Alto, CA: Science and Behavior.

Sprenkle, D. H. & Blow, A. J. (2004). Common factors and our sacred models. Journal of Marital and Family Therapy, 30(1) 113-129.

Von Glasersfeld, E. (1984). An introduction to radical constructivism. In P. Watzlawick (Ed.) The invented reality (pp. 17-40). New York: Norton.

Watzlavick, P., Bavelas, J. B., & Jackson, D. D. (1967). Pragmatics of human communication: A study of interactional patterns, pathologies, and paradoxes. New York: Norton.

Whitaker, C. & Keith, D. V. (1982). Symbolic-experiential family therapy. In J. R. Neill & D. P. Kniskern (Eds.), From psyche to system: The evolving Therapy of Carl Whitaker (pp. 330-378). New York: Guilford.

Visual Working Memory: Capacity, Resolution, and Expertise

[First published on Nathen’s Miraculous Escape, July 20, 2009.]


Visual Working Memory: Capacity, Resolution, and Expertise

Nathen B. Lester

For PSY 435, Dr. Awh

University of Oregon

June 9, 2008

Visual Working Memory: Capacity, Resolution, and Expertise

Ideas in psychology change and develop in much the same way that they do in a conversation, a conversation taking place over years, primarily in the form of detailed written accounts of hypotheses, experiments, and results. It may be slow and technical, but it has the same form: Assertions are made, evidence is presented, mistakes are pointed out, and new assertions are made. Incrementally, the amount of knowledge is increased.

One such conversation that is ongoing in the scientific literature is about the capacity of visual working memory and how it may be affected by the perceptual expertise of the viewer.

Because working memory capacity is correlated with scholastic ability, attentional control, and scores on intelligence tests (Cowan, Elliot, Saults, Morey, Mattox, Hismjatullina, &Conway, 2005), it is a topic of considerable interest among researchers in cognitive psychology. The following sections describe three articles which are examples of an exchange between researchers on this topic which gives rise to new knowledge as well as raising new questions.

The Capacity of Visual Working Memory for Features and Conjunctions

In their article of 1997, Luck and Vogel argued for a visual working memory capacity of approximately four objects, regardless of how complex those objects are. They found mounting evidence for this in a series of experiments using Phillips’ (1974) change-detection paradigm, where subjects are shown an initial array of objects, a pause, and then a test array, with the task of indicating whether the two arrays were identical or different.  They conducted several experiments using arrays of visual objects which varied in their number and types of features, including a single color only, two colors for each object, object orientation only, color and orientation together, and finally objects which varied in their color, orientation, size, and the presence or absence of a gap. In all conditions of all experiments, subjects could accurately detect changes in about four of the objects in an array. This is strong evidence for an object-based working memory capacity and against a feature-based working memory capacity: Four objects were consistently remembered, whether those objects’ combined features equaled four, eight, or even 16 features.

Several other variations were used to rule out possible alternate explanations. To test for the possibility that their verbal working memories were aiding in the task, subjects were given a verbal load of two digits to remember, with no effect on their performance. To rule out the possibility that capacity estimates were being limited by the very brief presentation of the initial array, that presentation was increased from 100 ms to 500 ms, with no change in the results. To address the fact that more decisions had to be made when viewing larger test arrays, which could lead to more errors, another condition had subjects indicate whether one randomly chosen object had changed. This also did not affect the results. Finally, the condition in which each object had two colors was run to test the possibility that there are separate working memory systems for each kind of feature. Subjects in this condition could remember about four objects whether that meant remembering four colors or eight colors, evidence against a feature-based working memory system for color distinct from the system that remembers the other types of features.

A Visual Short-Term Memory Advantage for Faces

Curby and Gauthier’s 2007 article was an attempt to show the effects of holistic processing on visual working memory capacity, using a variation of the change-detection paradigm. Because of the greater efficiency of holistic processing, they reasoned, objects like upright faces, with which subjects have a lot of expertise, will be stored more efficiently in working memory. They hypothesized that this should result in a larger working memory capacity for faces than for other complex objects. Their experiments resulted in three basic findings: (a) At 500 ms of encoding time, subjects were less accurate in detecting changes between faces than they were in detecting changes between cars or watches. (b) At 2500 ms encoding time, subjects’ accuracy was equivalent for all categories. (c) At 4000 ms encoding time, subjects were more accurate with the faces than they were with the other categories of objects.

Curby and Gauthier’s (2007) explanation of these results was that for complex objects, perceptual encoding processes cause a bottleneck for creating representations in visual working memory. At the shortest encoding time, this results in fewer objects in memory from an array of faces than from arrays of less complex objects such as cars or watches. At the 2500 ms encoding time, the benefits of efficient, holistic processing brought the number of faces encoded up to the number of other objects. By 4000 ms of encoding time, those benefits allowed more faces to be stored in visual working memory than any other kind of object tested. In other words, given enough time, the benefits of the more holistic processing of faces outweighs their disadvantage of being more complex. Finally, based on this, they reasoned that the limits on the storage of complex objects in working memory hypothesized by Alvarez and Cavanagh (2004) are ameliorated to some degree by this efficient processing of faces.

Perceptual Expertise Enhances the Resolution but Not the Number of Representations in Working Memory

Scolari, Vogel, and Awh’s 2008 article was largely a correction and clarification of the meaning of Curby and Gauthier’s (2007) results: The benefit of expertise is not in the number but in the resolution of objects in working memory. The difference in subjects’ ability to detect changes in a face out of an array of faces compared to changes in, for example, a car out of an array of cars, was actually a measurement of comparison errors made between the memories formed and the test display, not the number of objects held in working memory. That is, when Curby and Gauthier (2007) thought they were measuring the quantity of objects in working memory, they were actually measuring their quality.

Scolari et al. (2008) managed to show this in one experiment using four categories of objects: faces, inverted faces, shaded cubes, and colored ovals. Within-category changes between the initial display and the test display replicated Curby and Gauthier’s (2007) results. In other words, subjects were more likely to detect a change from one upright face to another, as compared to a change from one cube to another or one inverted face to another, possibly reflecting the benefits of holistic processing for faces. Changes across categories, however, which eliminated the possibility of comparison errors, replicated Luck and Vogel’s (1997) results. In other words, when the changes were big, a face changing into a cube, for example, working memory capacity estimates were at about four objects, regardless of the complexity of those objects.

Scolari et al. (2008) also found that individual differences in subjects’ performance in the cross-category change detection tasks were correlated to their performance in their simple color-change detection task. This suggests that performance on these tasks produces a more pure estimate of working memory capacity than does performance on within-category tasks. Additionally, the cross-category individual differences were not correlated to the within-category individual differences, indicating that being able to hold a certain number of objects in working memory and being able to store details about those objects are distinct abilities, drawing on different resources.

Two Additional Voices

Two other articles are worth briefly describing to flesh out this conversation. The first of these is Alvarez and Cavanagh’s 2004, “The capacity of visual short-term memory is set both by visual information load and by number of objects.” This article was in part a reply to Luck and Vogel (1997) and set the stage for Curby and Gauthier’s error: Alvarez and Cavanagh measured what they thought was the complexity of the objects they used in their change-detection tasks and found that visual working memory capacity was limited by the complexity, and not just the number, of the objects therein.

The problem was that their operational definition of “complexity,” which was based on visual search rate, was confounded with similarity. That is, Alvarez and Cavanagh (2004) judged categories of objects more complex because they were more difficult to tell apart. This was pointed out by Awh, Barton, and Vogel (2007) in “Visual working memory represents a fixed number of items regardless of complexity.” Setting the stage for Scolari et al. (2008), Awh et al. (2007) showed that it was actually comparison errors, caused by object similarity, and notobject complexity that were producing the lower estimates for visual working memory capacity.


Useful conversations rely on careful logic, clearly defined terms, and up to date information. In one way, the exchange here can be seen as corrective of errors in these areas.

Luck and Vogel (1997) used sound and thorough reasoning in combination with a series of very straightforward experiments to present evidence for an object-based visual working memory capacity of about four items, and, simultaneously, evidence against the idea that visual working memory is limited by the number of features each object has. Alvarez and Cavanagh (2004) presented what they thought was evidence for the complexity of objects being an additional limit to capacity, but Awh et al. (2007) showed that apparent limit to be a resource artifact from the difficulty of distinguishing highly similar items from each other. Almost certainly unaware of Awh and colleagues’ work, Curby and Gauthier (2007) set out to refine the work of Alvarez and Cavanagh (2004), stating that the complexity of objects may limit capacity, but expertise can overcome that limit to some degree. Then Scolari et al. (2008) pointed out that Curby and Gauthier (2007), relying on Alvarez and Cavanagh’s reasoning, had made the same error: Their results did not mean what they had thought. When comparison errors are eliminated, it is obvious that the number of objects stored in working memory is not affected by those objects’ complexity.

While in some ways side-tracks based on faulty reasoning, these works of Alvarez, and Cavanagh (2004), and Curby and Gauthier (2007), have also been useful in extending our knowledge. What we mean by the word “capacity,” for example, has been refined. “Capacity” has been used in a variety of ways, especially in Curby and Gauthier (2007), who used it to mean either the total number of “slots” available in working memory, the number of slots that happened to have been filled in a certain experiment, an amount of total information, and a kind of rate-based encoding capacity, in the vein of “objects encoded per second.” It is now clear to those with up-to-date information, that when discussing visual working memory, “capacity” should refer to the number of slots available for visual objects.

It should also be clear that we currently have a model for visual working memory that has at least two factors: capacity for storing objects and the resolution of those objects. Further, the information-load bottleneck may be a real phenomenon, but it is not about the time it takes to store complex items; it appears to be about the time it takes to build representations of sufficient resolution to avoid comparison errors. Furthermore, expertise, which may result in more holistic processing of visual stimuli, does seem to increase subjects’ ability to encode high-resolution memories, even if it does not increase the number of objects which can be stored.

Many good questions have been raised as well. Since search-rate based measurements have been ruled out, what are good operational definitions of “complexity” and “information load” in regard to visual objects? Is there a kind of “resolution capacity,” and how would it relate to expertise, given a good operational definition of complexity? How does this relate to the “consolidation time” estimates between 50 ms and 500 ms in Curby and Gauthier (2007, p. 627)? What is the role of expertise in the resolution of memories of objects other than faces? The most interesting question follows from the correlation of working memory capacity, intelligence scores and academic achievement (Cowan et al., 2005): Since individual differences in the resolution factor are not correlated with individual differences in the capacity factor (Scolari et al. 2008), what is the relationship between the resolution of visual working memory and intelligence?


Alvarez, G. A., & Cavanagh, P. (2004). The capacity of visual short-term memory is set both by visual information load and by number of objects. Psychological Science, 15, 106-111.

Awh, E., Barton, B., & Vogel, E. K. (2008). Visual working memory represents a fixed number of items regardless of complexity. Psychological Science, 18, 622-628.

Cowan, N., Elliot, E. M., Saults, J. S., Morey, C. C., Mattox, S., Hismjatullina, A., & Conway, A. R. A. (2005). On the capacity of attention: Its estimation and its role in working memory and cognitive aptitudes. Cognitive Psychology 51, 42-100.

Curby, K. M., & Gauthier, I. (2007). A visual short-term memory advantage for faces.Psychonomic Bulletin & Review, 14, 620-628.

Luck, S. J., & Vogel, E. K. (1997). The capacity of visual working memory for features and conjunctions. Nature, 390, 279-281.

Phillips, W. A. (1974). On the distinction between sensory storage and short-term visual memory. Perceptual Psychophysiology, 16, 283-290.

Scolari, M., Vogel, E. V., & Awh, E. (2008). Perceptual expertise enhances the resolution but not the number of representations in working memory. Psychonomic Bulletin & Review, 15, 215-222.

The Relationship between Clarity of Enunciation and Idea Density

[First published on Nathen’s Miraculous Escape, June 12, 2009.]


This study was an attempt to determine whether there is a relationship between individuals’ clarity of enunciation, rated subjectively, and linguistic ability, measured as idea density, as in Findings from the Nun Study (Riley, K. P., Snowdon, D. A., Desrosiers, M. F., & Markesbury, W. R., 2005; Snowdon, D. A., Kemper, S J., Mortimer, J. A., Greiner, L. H., Wekstein, D. R., & Markesbury, W. R., 1996). Idea density, the number of propositions per 10 words, had no significant correlation with clarity of enunciation in 33 digitally videotaped dyadic conversations between adult participants.

The Relationship between Clarity of Enunciation and Idea Density

Since 1992 I have wondered whether exceptionally clear enunciation was an indication of intelligence. I seemed to notice enhanced and separated consonants more often in those that I considered radically intelligent. My Organic Chemistry teacher, for example, said the word “little” exactly as written, with the crisp, unvoiced, alveolar stop [t], and not, as most other people said it, sounding like “lid’l.” Also, if he said a word that ended in a stop consonant followed by a word that started with a stop consonant, he stopped for each consonant; in “stand together,” the [d] and [t] would be separate, not, as in others’ speech, “stantogether.”

There has apparently been no research on a possible connection between cognitive or linguistic ability and clarity of enunciation. Most speech clarity research has focused on intelligibility, in relation to speech disorders, or dysarthrias, of various kinds on the speech-production side (e.g. Ansel & Kent, 1992), or with hearing impairment and hearing aids on the reception side, (e.g. Amyn, Rakerd, & Punch, 2006). There is also some research on the effects of the increased speech clarity in infant-directed speech, on early language acquisition. In infant-directed speech, for example, mothers’ vowel sounds were found to be more distinct from each other than in normal speech (Kuhl, Andrusky, Chistovich, Chistovich, Kozhevnikova, Ryskin, Stolyarova, Sundberg & Lacerda, 1997), and the more distinct the vowel sounds were, the better their infants’ speech perception (Liu, Kuhl, & Tsao, 2003). Whether distinctly produced speech is correlated with any other traits or tendencies of the speaker, though, appears to have gone uninvestigated.

Idea density

Having no direct measure of cognitive ability available, this study used idea density as a proxy. Idea density is a measure of linguistic ability that is associated with knowledge, vocabulary, and education level, and is defined as the average number of ideas, or propositions, per ten words in a text (Snowdon, Kemper, Mortimer, Greiner, Wekstein, & Markesbury, 1996). A text with high idea density, then, is complex and provides the reader with a lot of information.

In linguistics, a proposition is an idea expressed in a narrative, and considered a basic unit of memory for texts (Kintsch & Keenan, 1973). Propositions include verb, adjective, and adverb phrases, noun and clause conjunctions, and indications of temporal and causal relations (Turner & Greene, 1977). Linguists can take a text and construct what they call a propositional text base, which is a list of propositions coded in such a way that all of the information from the original text can be reconstructed.

In a longitudinal study of 180 nuns who entered their convents between 1931 and 1943, the idea density of their autobiographies, written at an average age of 22, correlated with their cognitive functioning and neuropathology in their old age and at death: Lower idea density predicted decreased cognitive functioning, and dementia (Snowdon et al., 1996), low brain weight, cerebral atrophy, and the neural plaques and tangles associated with Alzheimer’s disease (Riley, Snowdon, Desrosiers, & Markesbury, 2005).

It would be useful to know if idea density is also correlated with something as easily recognizable as clarity of enunciation. To that end, the hypothesis of this study was that clarity of enunciation would be significantly and positively correlated with linguistic ability, as measured by the idea density in conversational speech.

Methods and results


Participants were 110 adults, recruited by undergraduate psychology majors at the University of Oregon for a required class project, and recorded in 55 dyadic conversations on digital video recorders of varying quality. Conversations were recorded in residences, not in a lab. Forty-four participants were excluded from analysis: 34 because the video file was either not provided, would not play, or did not match the accompanying conversation transcription, 4 because participants were eating during the conversation, 4 because they were non-native English speakers, and 2 because their longest utterances contained no more than two words. This left 66 participants (34 female) between the ages of 18 and 57 (M = 24.4, SD = 6.3), 3 Latino, 1 Native American, and 62 White. Education level ranged from less than high school to graduate degree. All participants signed informed consent forms and filled out simple demographics forms prior to being filmed. No formal debriefing was given.


Ten minutes of each 15-minute conversation was transcribed by the student who recruited and recorded the participants, using Elan transcription software. Transcriptions were to include all speech by each participant during the 10-minute interval. The only punctuation marks used were [/] to indicate falling intonation at the end of an utterance, and [?] to indicate rising intonation at the end of an utterance. Annotations were composed of groups of utterances by a single participant separated by pauses of less than 2 seconds. Thus, an annotation could be of nearly any length, and could contain any number of utterances.

Idea density coding

Idea density was coded by the researcher, using the coding scheme presented by Riley et al. (2005), in which idea density is the average number of propositions per 10 words. Propositions include verb, adjective, and adverb phrases, noun and clause conjunctions, and indications of circumstance such as time, place, and causality. (See Appendix A for all coding schemes, and see Turner & Greene, 1973, for a thorough presentation of the construction of a propositional text base.)

For a very short example of proposition counting, consider the utterance “We’re going to see Meaghan’s art show.” This sentence contains 6 propositions: (a) the predicate phrase “We see show,” (b) “we” means the speaker and at least one other person, (c) the show is an art show, (d) the show is Meaghan’s, (e) this event is to happen in the future, and (f) the people meant by “we” will have to move to another location in order to see the show. Notice that even in this simple example there is some ambiguity; proposition (f) may or may not have been intended by the speaker.

For each participant, idea density was coded for the annotation of their speech closest to 10 seconds long. In cases where much of that annotation was taken up by laughing or unintelligible speech, the annotation closest to 10 seconds with the largest number of words was used. Annotations ranged from 17 to 51 words (M = 33.40, SD = 8.59), and formed a platykurtic distribution (kurtosis = -.69, SE = .58). Number of propositions ranged from 18 to 27 (M =18.30, SD = 4.46), and formed a platykurtic distribution (kurtosis = -.51, SE = .58). Idea density ranged from 3.6 to 7.2 (M = .56, SD = .08), forming a somewhat skewed (skew = -.14, SE = .3) and platykurtic (kurtosis = -.24, SE = .59) distribution. The levels of skew and kurtosis present were probably acceptable, all falling well within 2 standard errors.

Enunciation coding

Clarity of speech is usually coded using digital editors and spectrographic analysis to measure vowel space expansion and consonant enhancement. The limitations of this study, however, made it necessary to code enunciation more subjectively: Two coders, one of whom was the researcher, and neither of whom were blind to hypothesis, listened to the first 2 minutes of each conversation and coded the clarity of speech of each participant on a 3-point scale: 0 = noticeably unclear speech, 1 = average clarity of speech, and 2 = noticeably clear speech. The raters exhibited poor reliability (Cronbach’s α = .35), indicating the need for more training on the coding scheme, but, because of the time constraints of this study, each participant received a speech clarity score equal to the average of the two coders’ ratings as they were.

The averaged speech clarity scores had a mean of 1.40 (SD = .46) and formed a skewed (skew = -.46, SE = .30), platykurtic (kurtosis = -.52, SE = .60) distribution. The level of skew and kurtosis were probably acceptable, falling within 2 standard errors.


The question under investigation was whether clarity of speech was correlated with linguistic ability, as measured by idea density, the average number of ideas per word, in the conversational speech of this sample. The answer is no, they were not. The correlation between clarity of speech and idea density was not significantly different from 0 (r = .03, p = .81). This indicates that clarity of speech and idea density were not related.

Idea density was also not significantly correlated with gender, level of education of the parents of the participant, or the number of propositions per annotation. On the other hand, idea density was negatively correlated with number of words spoken per annotation (r = –.38, p <.01), positively correlated with age of participant (r = .25, p = .04), and had a very marginal positive correlation with the education level of participant (r = .21, p = .09).

Clarity of speech was marginally correlated with only one other variable, level of participant education (r = .24, p = .06). See Table 1 for all correlation values.


This study showed no support for the hypothesis that clarity of enunciation was positively correlated with linguistic ability. The dimensions may, therefore, be orthogonal.

It is true that the methodology used was limited in several ways. The quality of the sound recordings was highly variable. The coding of enunciation was subjective, unreliable, and coders were not blind to condition. Idea density was coded by one individual, so there was no way to check his reliability. Additionally, the difference between the correlations of enunciation and idea density for one coder (r = .173) and the other (r = –.157) was marginally significant (p =.06) using Fisher’s r to z transformations to make the comparison.

Clearly, these results may not be the best indication of the relationship between these two variables, and improved methodology might reveal different results. On the other hand, it may be prudent to look elsewhere for correlates of idea density. The only moderately strong correlation with idea density in this study, for example, was the negative correlation with number of words spoken. The idea that high rates of speech might indicate low verbal ability is somewhat counterintuitive and intriguing.

Another possibility is that the idea density coding used by Riley et al. (2005) and Snowdon et al. (1996) is not appropriate for conversational speech, or might need some modification; there may be ways that spoken and written language differ that need to be taken into account. One annotation in this sample (and which was not used for this reason), for example, contained only one word, “Yeah,” giving this participant an average of 10 propositions per 10 words. Surely, a tendency to speak in one-word utterances is not an indication of linguistic ability!

Other possible differences between written and spoken language are the prevalence of filler words and run-on sentences in spoken language. Here is an example from the sample: “No they were talking about Eddie/ And they were like OH YEAH EDDIE was getting a hold of me like he wants to say goodbye to all his homies before he goes/ And I was like they were like he hasn’t called you? And I was like no?” Chances are, had this passage been written instead of spoken, “said” would replace “was like” and “were like,” and the “and” at the beginning of 3 of the 4 utterances would not appear. Although in this case the changes would balance each other, it may be that on the whole, written and conversational language differ significantly in their average idea density, and perhaps even in the cognitive factors that produce high or low idea density in each. Unfortunately, Riley et al. (2005) and Snowdon et al. (1996) did not publish their idea density statistics (though not surprising, since they published in neurobiology and medical journals, not linguistics or psycholinguistics journals), or that analysis could have begun in this study.

Flawed as it is, this study represents the only evidence to date that I am aware of about the relationship between clarity of speech and linguistic ability, and, so far, it seems unwise to judge someone intelligent based only on the crispness of their enunciation.


Ansel, B. M., & Kent, R. D. (1992). Acoustic-phonetic contrasts and intelligibility in the dysarthria associated with mixed cerebral palsy. Journal of Speech and Hearing Research, 35,296-308.

Amlani, A. M., Rakerd, B., & Punch, J. L. (2006). Speech-clarity judgments of hearing-aid-processed speech in noise: Differing polar patterns and acoustic environments. International Journal of Audiology, 46, 319-330.

Kintsch, W. & Keenan, J. (1973). Reading rate and retention as a function of the number of propstitions in the base structure of sentences. Cognitive Psychology, 5, 257-274.

Kuhl, P.K., Andruski, J. E., Chistovich, I. A., Chistovich, L. A., Kozhevnikova, E. V., Ryskin, V. l., Stolyarova, E. I., Sundberg, U., & Lacerda, F. (1997). Cross-language analysis of phonetic units in language addressed to infants. Science, 277, 684-686.

Liu, H. M., Kuhl, P. K., & Tsao, F. M. (2003). An association between mothers’ speech clarity and infants’ speech discrimination skills. Developmental Science, 6, F1-F10.

Riley, K. P., Snowdon, D. A., Desrosiers, M. F., & Markesbury, W. R. (2005). Early life linguistic ability, late life cognitive function, and neuropathology: Findings from the Nun Study.Neurobiology of Aging, 26, 341-347.

Snowdon, D. a. Kemper, S J., Mortimer, J. A., Greiner, L. H., Wekstein, D. R., & Markesbury W. R. (1996). Linguistic ability in early life and cognitive function and Alzheimer’s disease in late life: Findings from the Nun Study. Journal of the American Medical Association, 275, 528-532.

Turner, A. & Greene, E. (1977). The construction and use of a propositional text base.University of Colorado, Institute for the Study of Intellectual Behavior; Boulder, CO.

Table 1

Correlations of all variables

Variables                    1.         2.         3.         4.         5.         6.         7.         8.         9.         10.

1. Propositions

2. Words                     .84**

3. Idea Density            .17       -.38**

4. Speech Clarity        -.04      -.07      .03

5. Coder 1 SC             -.08      -.20      .17       .88**

6. Coder 2 SC             .06       .15       -.16      .71**   .29*

7. Gender                    .05       .14       -.15      .08       .07       -.02

8. Age                         .12       -.01      .25*     .18       .21       .03       -.02

9. Participant Ed.        .11       -.03      .21       .24       .24       .16       -.16      .52**

10. Mother Ed.            -.08      -.01      -.13      -.03      -.01      -.04      -.02      .06       .23

11. Father Ed.             .04       .02       .03       .00       -.05      .04       -.08      .12       .28*            .47**

Note. SC = speech clarity. Ed. = education. Gender was coded 0 = female, 1 = male.

* Correlation is significant, p < .05

** Correlation is significant, p < .01

Appendix A

Idea density coding

The following from Turner and Greene (1977):

1. Modified arguments of predicate propositions

2. Connected arguments of predicate propositions

3. Predicate propositions

4. Modifiers of predicate propositions

5. Modified arguments of circumstantial propositions

6. circumstantial propositions

7. Other connective propositions within clause

8. Repeat

Enunciation coding

0 = noticeably unclear speech

1 = average clarity of speech

2 = noticeably clear speech

Gender coding

0 = female

1 = male

Education level coding

1 = Less than high school dimploma

2 = High school diploma

3 = Some college

4 = Undergraduate degree

5 = Some graduate school

6 = Graduate degree

Differentiating the Effects of Social and Personal Power

[First published on Nathen’s Miraculous Escape, June 11, 2009.]

Thanks to my advisor and collaborator, Sean M. Laurent, and my second reader, Sara D. Hodges


The current research attempted to differentiate the effects of social power (i.e., having control of others’ outcomes) from personal power (i.e., control of one’s own outcomes) on variables related to perspective-taking. Using methodology adapted from Galinsky, Magee, Inesi, and Gruenfeld (2006), 224 participants were primed with high social power, high personal power, or low power, and then completed two perspective-taking measures. While these measures did not significantly vary with power condition, evidence was found that supports the differentiation of personal power from social power. Low power, as experienced by participants (rather than as manipulated) had a strong negative correlation with personal power, and a weak negative correlation with social power. Personal power and social power were not significantly correlated.

Differentiating the Effects of Social and Personal Power

To some extent, all human beings want power—even if it is just enough power to ensure one’s safety, allow oneself to prosper, and obtain or keep freedom for oneself and one’s groups. This desire for power, to get just enough or to get even more than one needs for one’s own comfort and safety, may explain the frequent power struggles that populate our history books. And given the strong desire humans have to attain power, it is no wonder that psychologists are fascinated with the construct; understanding power is central to understanding human behavior.

Social power

Power can be defined in many ways, but in psychological research it is usually defined in relational terms, and is called “social power.” Individuals are said to have social power when they can control the outcomes of another person, influencing that person’s states, providing that person with rewards, inflicting punishments on that person, or governing the flow of resources to that person (e.g., Fiske & Berdahl, 2007; Galinsky, Magee, Inesi, & Gruenfeld, 2006; Keltner, Gruenfeld, & Anderson, 2003; Smith & Trope, 2006). Common to all definitions of social power is that there exist relationships between people, where some people control and are important sources for valued resources for other people. These valued resources are not limited to any particular physical or social object; a valued resource might range from a kind word to a grade in a class, a promotion, an offer of protection, or a glass of water for a thirsty child.

The approach-inhibition theory of social power ties high and low power states to psychological approach and inhibition mechanisms (Keltner et al., 2003). According to this theory, people who are high in social power will tend to have an approach orientation, will typically experience more positive affect, and will be trait-congruent—that is, they will act in alignment with their own traits, be less careful, more automatic in their social cognitions, and attend primarily to elements of the environment which might be useful or pleasurable, including a tendency to see others as a means to their own ends. People who are low in power, on the other hand, will tend to have an inhibition orientation, where they tend to experience more negative affect, be more cautious and observant of potential risks in their environment, and tend to see themselves as means to others’ ends.

Research has provided some support for approach-inhibition theory: Compared to people low in power, people high in power have been found to be more trait-congruent, more sensitive to potential rewards, and to experience more positive affect (Anderson & Berdahl, 2002). They are more prone to take risks and be optimistic about risk outcomes (Anderson & Galinsky, 2006). Higher power has also been associated with increased abstract information processing (Smith & Trope, 2006), increased reliance on the availability heuristic (Weick & Guinote 2008), and increased stereotyping (Fisk, 1993), while low power has been shown to increase metastereotyping (Lammers, Gordjin, & Otten, 2008) and impair executive functions (Smith, Jostmann, Galinsky, & van Dijke, 2008). Supporting an approach orientation, high power is associated with a tendency to act rather than not act (Galinsky, Gruenfeld, & Magee, 2003), to be more expressive (Snodgrass, Hecht, & Ploutz-Snyder, 1998), and to be less influenced by a variety of situational variables such as pressure to conform (Galinsky, Magee, Gruenfeld, Whitson, & Liljenquist, 2008). Furthermore, high power has been linked to devaluing others’ work for self-gain (Kipnis, 1972), to more extreme, less accurate judgments of opponents (Keltner & Robinson, 1997), and in some cases, to sexual harassment (Bargh, Raymond, Pryor, & Strack, 1995).

While many of the behaviors associated with power are self-serving and might serve to entrench power holders in their positions, at least some of these effects can be shifted by values and accountability; for example, individuals with a communal relationship orientation were found to be more generous when primed for high power (Chen, Lee-Chai, & Bargh, 2001), and familial responsibilities reduced several kinds of risky reward seeking in high need-for-power individuals (Winter & Barenbaum, 1985). It may also be true that some effects may not be the direct result of having power—instead they may be the result of the responsibilities that come along with power: For example, high power individuals have been shown to be better at individuating others unless burdened with organizational responsibility (Overbeck & Park, 2001).

Personal power

Van Dijke and Poppe (2006) asserted that striving for power is better explained by distinguishing between social power (i.e., having control of others’ outcomes) and personal power (i.e., having control of one’s own outcomes). While social power is related to authority, status, and dominance (Hall, Coats, & LeBeau, 2005), personal power is also related to autonomy, independence, agency, and competence (Overbeck & Park, 2001; Van Dijke & Poppe, 2006). Van Dijke and Poppe presented evidence that, given the choice, individuals act to increase their personal power instead of their social power, and will at times even voluntarily reduce their own social power to obtain personal power. Van Dijke and Poppe argued further that when individuals act to increase their power over others, they are motivated not by a wish to control others, but by a wish to not be controlled by others, and that increasing social power is a means to that end.

While most researchers have been careful to define power in social terms, it is evident that personal and social power have at least sometimes been confounded in research on the psychological effects of having power. That is, because personal and social power may tend to vary together, if they are not carefully separated, the effects of one type of power may be masked by the effects of the other. This may be especially true in experimental paradigms where social power is manipulated by giving some participants control over others or where the idea of control over others is primed. In these cases, personal power might be inadvertently manipulated at the same time. For example, one commonly used power-priming manipulation is to have participants write about a time in which they had power over someone else, or when someone else had power over them. While manipulating social power, this manipulation will tend to prime personal power as well, because people low in social power will tend to have less control of their own outcomes than do those high in social power, and those high in social power will tend to have more power over their own outcomes. This leaves open the possibility that any observed effects of this power manipulation could be due to the participants having control of others’ outcomes, having control over their own outcomes, or some combination of the two.

If personal power is the most likely reason that people seek power over others, then it is important to distinguish the effects of personal power from the effects of social power. It may be that having social power is associated with more negative or antisocial outcomes than is having personal power. If this is true, then encouraging people to seek personal power but not social power may be indicated.

Power and perspective-taking

Perspective-taking, conceptualized as the act of imagining or ability to imagine the experiences of others (Galinsky et al., 2006), is an important skill, associated with social competence and self-esteem (Davis, 1983). It is considered a necessary component of moral development and moral reasoning (Kohlberg, 1976; Walker 1980), and of empathy (Eisenberg, Murphy, & Shepard, 1997). Furthermore, perspective-taking has been linked to reductions in stereotyping (Galinsky & Moskowitz, 2000), increases in self-other overlap (Davis, Conklin, Smith, & Luce, 1996), and helping behaviors (Batson, 1991).

Perspective-taking, then, is clearly a beneficial behavior, and a necessary part of healthy social functioning. Social power, however, seems to have a negative effect on the ability to take others’ perspectives; Galinsky et al. (2006) found that after participants were primed with social power, they were less likely to take an outside visual perspective (i.e., less likely to draw an “E” on their foreheads in the direction that would be easily readable from another person’s perspective), less able to recognize other people’s facial emotional expressions, and were more likely to believe that others knew what they knew, even when they were presented with contrary evidence.

The present research

The purpose of the current research was to take a preliminary step toward differentiating the effects of having social and personal power on perspective-taking. Toward this end, Galinsky and colleagues’ (2006) methodology was used and extended somewhat, in order to test hypotheses about personal as well as social power. In the original paradigm, participants were first primed for either high social power or low power by writing a power-related essay. In the current paradigm, another condition was added where participants were primed for personal power. Following the priming tasks, participants performed one of two perspective-taking tests (these were also the same dependent variables used by Galinsky et al., 2006). One of these was the Diagnostic Analysis of Nonverbal Accuracy for adult faces (DANVA; Nowicki & Duke, 2001). In this test, participants view photographs on a computer screen, and are asked to identify the emotions shown on facial expressions of the photographed target. This task is thought to be a measure of interpersonal sensitivity related to perspective-taking, because the ability to recognize what emotions others are feeling is an important part of being able to take their perspectives. The other task asked participants to draw an E on their own foreheads; this measure assesses whether participants tend to spontaneously adopt the visual perspective of a person facing them, rather than their own internal perspective.

As a first outcome, we expected to replicate the work of Galinsky and colleagues, showing that the high social power group would perform worse on the DANVA and draw Es from their own perspective more often than the low power group. In terms of what personal power would predict, our hypotheses were not as clear. It was possible that personal and social power would activate approach equivalently and thus have the same effect on perspective-taking. It was also possible, however, that personal power would affect perspective-taking more negatively than social power, because people in the personal power condition might be less cued in to relationships with others. We reasoned, however, that the personal power group would perform somewhere between the low and social power groups on measures of perspective-taking, because similarly to social power, envisioning personal power should create associations with rewards and freedom, but different from social power, it should not contain the same dominance or hierarchy power cues.



Two hundred and twenty-four undergraduates (172 females) from the University of Oregon’s Psychology and Linguistics human subjects pool participated in this experiment. The mean age was 19.6 years and the age range was between 17 and 49 years. In addition to partially fulfilling course requirements by participating, participants received entry into a lottery drawing for $100.


Participants each drew an E on their own foreheads as a measure of perspective-taking (Galinsky et al., 2006; Hass, 1984). An E that read in the correct direction for an outside observer facing the participant was treated as an indication that perspective-taking had occurred. This was presented to participants as a coordination task, rather than a test of perspective-taking. For a full description of the methodology, see Galinsky et al. (2006).

Participants also completed the Diagnostic Analysis of Nonverbal Accuracy (DANVA2-AF), a task where participants guess the emotions shown in 24 photographs of young adults’ faces: happiness, sadness, anger, or fear (Nowicki & Duke, 2001). Each image appeared on a computer screen for 2 seconds, followed by 5 seconds for the participant to mark their answer on a paper test form.

Next, participants completed four questionnaires that had been slightly modified from their original forms such that participants were instructed to respond based on how they were currently feeling, rather than assessing themselves globally. The first of these was a measure of Unmitigated Agency (UA; Spence, Helmreich, & Holahan, 1979), focusing on the self to the exclusion of others. On this scale, participants rated themselves on a five-point scale anchored by two contradictory characteristics (e.g., “Not at all arrogant” and “Very arrogant”).The second was the WHO-5 Well-being Index (WHO-5; Bech, 1993), where participants responded to questions such as “I feel cheerful and in good spirits” on a 5-point scale anchored by “Does not describe the way I feel at all” and “Describes how I’m feeling well.” The next questionnaire was Hegelson’s Unmitigated Communion Scale (UC; Hegelson & Fritz, 1998), which measures a focus on others’ needs to the detriment of the self. Participants responded on a five-point scale anchored with “Strongly agree” and “Strongly disagree” (e.g., “I always place the needs of others above my own.”)The last measure assessed conservative attitudes (Short Right-Wing Authoritarianism Scale; RWA; Altemeyer, 1998; Zakrisson, 2005). In this measure, participants rated their agreement with statements such as, “Our country needs a powerful leader, in order to destroy the radical and immoral currents prevailing in society today,” on a six-point scale anchored with “Strongly agree” and “Strongly disagree.” One additional, single-item measure of self-esteem was used: “How much do you like your name, in total?” This was measured on a 9-point scale, and was taken from Gebauer, Riketta, Broemer, and Maio (2008), who found that higher scores on this single-item measure indicated higher self-esteem.

As a manipulation check, participants rated themselves on several power-related adjectives or short statements (7-point scale where 1 = “Not at all” and 7 = “Extremely”). The following adjectives were used: High Social Power—dominant, in-charge, leader, organizer; High Personal Power—self-directed, independent, controls own choices, free; Low Power—accountable (reverse-scored), submissive, restricted, powerless. As an additional manipulation check, participants completed a post-experimental questionnaire about how powerful they actually felt during the manipulation. Questions included, “As you wrote about your life, to what extent did you feel as if you had power over someone else?” and, “As you wrote about your life, to what extent did you feel as if another person had power over you?” Each of these questions was rated on a 7-point scale anchored by either “Not at all,” and “Very much,” or “I agree completely,” and “I disagree completely.” See Appendix A for the complete list. Demographic questions were also included.


After giving written consent to participate, participants were run in groups of 1-4. In Study 1, participants were then seated at computers in separate rooms (in Study 2, participants were first informed about a $100 drawing and resource allocation task—see below). Participants were randomly assigned to one of three conditions: high social power (HSP), high personal power (HPP), or low power (LP). HSP and LP participants completed an experiential priming task commonly used to manipulate a sense of high or low social power (Anderson & Berdahl, 2002; Galinsky et al., 2003; Galinsky et al., 2006; Smith & Trope, 2006) in which they wrote about a time when they had power over someone (HSP) or when someone had power over them (LP). HPP participants wrote about a time when they could do whatever they wanted, without interference. See Appendix B for the complete set of instructions.

Study 2 added a second task to amplify the prime (Galinsky et al., 2006). HSP participants were instructed that they had been randomly assigned to a “resource allocation task” in which they were to divide seven chances to win in a drawing for $100 between themselves and one other randomly selected participant. LP participants were instructed that they had been assigned to the resource allocation task, but were not to do the allocating. Instead, they would guess how many chances out of 7 another participant would give them. HPP participants were not assigned to the resource allocation task, but asked which of several prizes they would prefer, should they win the drawing. See Appendix C for the full text of these tasks.

Following this, participants completed the “E” task and the DANVA, counterbalanced. Next, participants responded to the questionnaires described above: UA, WHO-5, UC, RWA, adjectives, and self-esteem measure. This was followed by the post-experimental manipulation check questions and demographics questions, as well as written suspicion probes. Participants were then probed for whether they had guessed the connection between the manipulation and measures debriefed, thanked, and excused.


The central research question of this study was how personal power would affect perspective-taking compared to low power and social power. To address this, participants were primed with either low power (LP), high personal power (HPP), or high social power (HSP). Based on past research (Galinsky et al., 2006), we expected high social power to decrease perspective-taking relative to low power. Additionally, we hypothesized that high personal power would decrease perspective-taking, but not to the same extent as high social power.

Contrary to our hypotheses, the primary dependent variables (the direction participants drew their Es and scores on the DANVA), did not vary significantly across condition. Because previous research (i.e., Galinsky et al., 2006) has shown that the same methodology used in the current study reliably affected perspective-taking, we undertook to discover why this finding was not replicated in our sample. As a first step, we assessed the reliability of the DANVA. This analysis revealed low internal reliability of the measure (Cronbach’s α = .345). To improve the reliability of the measure, we used the criterion of removing any item that had an item-total correlation lower than .1 (see Patterson & Stockbridge, 1998, for a similar method). Eleven of the 24 DANVA items were thus removed, based on this criterion, increasing reliability to .49. The item-total correlation of one further item dropped below .1 after adjusting the scale, so this final item was removed, leading to a final scale that contained 12 items. The reliability of the adjusted DANVA was still low compared to the alpha of .77 reported for college students by Nowicki and Carton (1993), but improved somewhat (Cronbach’s α = .504). Rerunning the previous analysis using the adjusted DANVA, however, returned the same result: no significant differences across condition in DANVA scores.

Next, we assessed whether there were any significant differences across conditions on any of the questionnaire measures that were included (UA, UC, WHO-5, RWA, Self-esteem). Although the majority of these questionnaires did not differ, the global self-esteem measure that asked participants how much they liked their names differed significantly across conditions, F(2, 217) = 5.79, = .004. Post-hoc tests revealed that participants in the low power condition (M = 7.80,SD = 1.52) liked their names significantly more than participants in the high personal power (M= 7.34, SD = 1.80) and high social power (M = 6.75, SD = 2.21) conditions, = .005, and that the high personal and high social power conditions differed marginally, = .058. That is, as measured by name-liking, LP participants had significantly higher self-esteem than both HPP and HSP participants, and HPP participants had marginally higher self-esteem than HSP participants. That self esteem varied by condition is most likely a contrast effect. Intuitively, participants primed with low power should have lower self esteem than those primed with high power, unless being made to think of themselves as low in power caused them to react and reassert themselves, increasing name-liking. This idea is supported by other research, where well-being and self-esteem were positively correlated, and name-liking predicted well-being better than explicitly measured self-esteem (Gebauer et al., 2008), while in this study, name-liking was positively correlated with well-being only in the HPP condition.

Manipulation Check

Because our primary hypotheses were not supported, and improving the reliability of the DANVA did not affect this finding, we turned to another idea: Perhaps the reason for the null effect is that power was not reliably manipulated. To test this, we checked the effectiveness of our manipulation, which had been assessed in two ways. On one questionnaire, participants rated themselves on a series of low, personal, and social power-related adjectives. On another questionnaire, participants answered either two or three questions about how much low, personal, and social power they had felt during the power-priming task.

Correlations between the four power-related adjectives and the questions meant to assess each type of manipulated power were examined. In each case, the correlations supported averaging measures together, forming three final composite manipulation check scales. For each, the reliability was adequate (low power α = .63, personal power α = .61, social power α = .64). Next, three separate one-way analyses of variance (ANOVAs) were run for each of the composite measures.

In the first, the composite measure for low power differed significantly across condition, (2, 178) = 20.32, < .001. Post-hoc tests showed the low power condition (M = 3.57, SD = 1.14) scored significantly higher (ps < .001) than the high personal power (M = 2.47, SD = 1.07) and high social power (M = 2.58, SD = .93) conditions, which did not differ from one another. That is, participants in the low power condition felt significantly less powerful than participants in the other two conditions, who did not differ in their experience of low power.

In the next analysis, the composite measure for personal power also differed significantly across condition, (2, 178) = 8.69, < .001. In this case, the low power condition (M = 4.63, SD = 1.05) scored significantly lower than the personal power (M = 5.36, SD = 1.04) and social power (M = 5.23, SD = 1.01) conditions (ps < .005), and again, the personal and social power conditions did not differ. That is, participants in the low power condition felt less personal power than participants in the other two conditions, which did not differ.

Last, on the composite measure of social power, the three conditions again significantly differed,(2, 178) = 12.415, < .001. Post-hoc tests revealed that on this measure, participants in the social power condition (M = 4.14, SD = 1.00) felt significantly more social power than participants in the personal power (M = 3.28, SD = 1.05) and low power (M = 3.44, SD = .97) conditions (ps < .001), which did not differ in the amount of social power they felt.

This pattern of results suggests that the experimental manipulation had the intended effects: Participants in the low power condition felt less powerful than participants in the other two conditions, and also felt less personal power than participants in the other two conditions. At the same time, on both of these measures, participants in the personal and social power conditions did not differ, as expected. On the last measure, which assessed feelings of social power, people in the social power condition felt like they had more power over others than did participants in the low power and personal power conditions, which did not differ.

Exploratory Correlational Analyses

After finding that the results of our primary tests did not support past research or our hypotheses, we decided to explore the data further. For example, although the manipulation check measures suggested that people in the different conditions actually did experience power in different ways, it is possible that the subjective experiences of power as measured by the manipulation checks themselves (i.e., rather than the manipulation) might shed some light on the differential effects of low, personal, or social power on perspective-taking and other outcome measures. Examining patterns of correlations between the manipulation check composite items (i.e., experienced power) and other variables allowed us to explore these questions.

Initially, correlations of the experienced power items (i.e., the composite manipulation check items) with each other were computed. As can be seen in Table 1 (along with the correlations of all the major variables), low power was negatively correlated with both social power (r = -.14, p= .055) and personal power (r = -.55, < .001). Although personal and social power were not significantly correlated across condition, within conditions, an interesting pattern emerged. In the social power condition, there was a significant and positive association between personal and social power (r = .39), but in the personal power condition, they were negatively correlated (=

-.21). These correlations significantly differ using Fisher’s r to z transformation, z = 3.34, p < .001. This provides further evidence that the experience of personal and social power can be differentiated, and that the manipulation served to do so effectively (see Table 2 for correlations of measures within conditions). This suggests that participants who had been reminded of a time when they were in control of themselves evaluated social power differently than those reminded of a time when they had power over someone else; in the social power condition, personal and social power were associated, but in the personal power condition, they were negatively associated. Perhaps, from a standpoint of social power, more social power means more freedom, but from a standpoint of personal power, more social power just means more responsibility.


The current study set out to determine how personal power, a type of power that in the literature has been potentially confounded with social power, would differentially affect perspective-taking compared to social power and low power. We hypothesized that, as in past research (Galinsky et al., 2006), participants who were primed with social power would exhibit less perspective-taking than people primed with low power. In addition, we hypothesized that being primed with personal power, or the feeling of not being under anyone else’s control and able to take charge of one’s own destiny, would have effects on perspective-taking somewhere between the two. Our results, however, did not support these hypotheses. The primary dependent variables—the direction of an E drawn by participants on their own foreheads and a facial emotional expression recognition task—did not significantly differ across condition. Further exploration of the data, however, did allow a few conclusions to be drawn.

Using composite measures of power that were created from what were originally manipulation check items, it appears that there are some clear distinctions between social and personal power, particularly in how each type of power relates to the experience of not having power at all. In support of this, experiences of low power and high personal power were strongly negatively correlated, while low power and high social power were only marginally correlated. This was true even though the definition of low power provided was in terms of social power: being under the control of someone else. This indicates that low power and personal power might be part of a single dimension, while low power and social power are less related. This supports the view that personal power is a broader category than social power. Furthermore, it suggests that similarly to how people may seek social power as a way of increasing their personal power (e.g., Van Dijke & Poppe, 2006), people who have little power also are less concerned about others having power over them than lacking power over themselves. In further support of this, the same social and personal power measures were positively correlated in the social power condition, and negatively correlated in the personal power condition, indicating that participants primed with social power failed to make a clear distinction between social and personal power, while participants primed with personal power not only made the distinction, but conceived of the types of power as negatively related. If this finding can be strengthened by future research, it could have implications for power theorists: Not only do people seem to prefer personal to social power (Van Dijke & Poppe, 2006), but a reminder of personal power is all it takes to make the distinction salient. 

Although it is difficult to interpret null results, the lack of significant findings in this case implies several possibilities. It is possible that, despite our best efforts, our experimental procedure was flawed in some way. For example, some element in our instructions may have been restricting the range of outcomes for the DANVA scores and E direction: In the previous research that we were trying to replicate by using this design element, Galinsky et al. (2006) found that 33% of participants in the high social power condition drew a self oriented E compared to 12% for the low power participants, while we found 29% of high social power participants drew a self-oriented E and 31% of low power participants, a difference in magnitude that was not only much smaller, but in the opposite direction. Galinsky and colleagues also found that high social power participants made an average of 4.54 errors on the DANVA and low power participants made 3.11, while we found that high social power participants made an average of 4.20 errors, and low power participants made an average of 4.16 errors. These differences suggest that while the social power manipulation might have had the intended effect, participants in the low power condition did not respond in the same way in the current study as in Galinsky et al. (2006).

Another possibility is that the relationship between power and perspective-taking, at least as perspective-taking is measured by the variables used here and in Galinsky et al., may not be a stable one. It may be subject to change, for example, based on minor situational variables such as the wording of auxiliary instructions. Last, our inability to replicate past research on the topic might imply a difference between the populations of the human subjects pools at the University of Oregon who take part in studies for partial class credit, and the paid participants found at business and public service graduate schools used by Galinsky and colleagues. Perhaps power was particularly salient to their population or particularly unimportant to ours.

Personal and social power

One limitation of this research is that personal and social power were still confounded in both the low power condition and the high social power condition. Ideally, some way might be found to further disentangle these types of power, although this might be difficult. Theoretically, low and high social power are not distinct from personal power, but social power is one of the routes by which people seek personal power. The question of whether it is possible to separate these effects may depend on how we conceptualize personal power.

Personal power, or the ability to control one’s own outcomes, has been related to independence, agency, competence, feelings of expertise, autonomy, and personal causation (Overbeck & Park, 2001; Van Dijke & Poppe, 2006). Personal causation, “the initiation by an individual of behavior intended to produce a change in his environment” (DeCharms, 1968, p. 6), is related to personal power, but only tangentially so; an individual with personal power would likely exhibit this initiating behavior, but the behavior is not synonymous with personal power. Agency defined as self-efficacy (see, for example, Bandura, 1989, and Van Dijke & Poppe, 2006) is very closely related to personal power, but another psychological definition of agency—as a focus on self and separation (e.g., Abele, Uchronshi, Suitner, & Wojciszke, 2008; Ghaed & Gallo, 2006; Hegelson & Fritz, 1999)—confuses the issue in much the same way that operationalizing social power as dominance orientation has been a problem for social power theorists (e.g., Hall et al., 2005). Similarly, autonomy, which is commonly understood to mean self-governance, is a near synonym of personal power. Self-determination theorists, however, distinguish autonomy, defined as the extent to which an individual personally endorses the activities they engage in, from independence, defined as the extent to which people are reliant on others’ resources, assistance, or permission (e.g., Chirkov, Ryan, Kim, & Kaplan, 2003; Ryan & Deci, 2000). This leaves open the possibility that individuals whose behaviors are tightly controlled by others, but who enjoy the tasks they are given, or who feel that the control is justified and wise, could be labeled personally powerful, which stands in contrast to our working definition of personal power.

It may be then that personal power needs to be defined as a combination of other factors: In order to control one’s own outcomes, there must be the freedom to do so (i.e., independence as defined above), but also the ability to do so, or competence. If this is the case, then only one of these preconditions to personal power is confounded with social power—the extent to which we have independence is constrained by the social power others hold over us, while competence is a dimension of personal power unrelated to social power.

With this understanding of personal power, it may be possible in future research to create experimental conditions of both high and low personal power that do not vary as a function of social power. One possibility is using a manipulation of task difficulty, as in studies on the effects of control deprivation (Ric, 1997; Ric & Scharnitzky, 2003) and learned helplessness (Hiroto & Seligman, 1975; Simkin, Lederer, & Seligman, 1983). In these paradigms, participants in a “no-control” group are given some form of unsolvable task, putting them in a position of non-social, low personal power; their inability to control their own outcomes stems from the difficulty of the task, not from the control of someone else. A symmetrical non-social, high personal power condition could be created using practice to achieve competence at a task, or relative mastery of a task as a quasi-independent variable.

In its two-faceted sense, personal power is a broad concept, including social power as one of its limiting or enabling components. It should also be possible and instructive, however, to create experimental conditions in which social power varies independently from personal power. Participants in both low and high social power conditions could be instructed to perform the same task, perhaps relaying orders from superiors to inferiors in a hierarchy, and the number of superiors and inferiors could be manipulated: Low social power participants could be placed low in the hierarchy, with many superiors and only one underling, while high social power participants could be placed in the opposite position, second from the top, with many underlings.

There is also a need for control groups in power research. While high and low power are conceptualized as having symmetrical effects, with high power increasing positive affect and automatic social cognition, for example, and low power increasing negative affect and controlled social cognition (e.g., Keltner et al., 2003), most power research has not used control groups. In research that has used a control group, results have sometimes been asymmetrical. For example, in some experiments the low power condition was not different from the control, and in others it was the high power condition that was not different from the control (Smith & Trope, 2006). Without controls, it is impossible to tell whether it is high or low social power that is responsible for effects that are found, or whether the effects are symmetrical, as predicted by approach-inhibition theory. Control groups could also be helpful in determining what outcomes are a direct result of power and which are mediated by stress or mood.

Future research

How does personal power affect perspective-taking compared to social power? To the extent that these types of power are confounded, their effects will be equivalent, but it is also possible that their effects will remain equivalent even if they can be disentangled. Intuitively, personal power should produce an approach orientation. It may be, however, that personal power negatively affects perspective-taking more than social power: Personal power, on a broader continuum, has more extreme positions of power available; imagine being at the top of your hierarchy and a master of the skills required of you, versus at the bottom of your hierarchy and, additionally, inept. Social power, on the other hand, involving relationships by definition, may be associated with increased perspective-taking compared to personal power; at least someperspective-taking is necessary when dealing with subordinates. There is some evidence for this line of reasoning: Tjosvold and Deemer (1980) found that independence decreased perspective-taking compared to dependence, which in turn decreased perspective-taking compared to interdependence. To the extent that personal power is akin to independence, low power to dependence, and social power to interdependence, then, personal power should decrease perspective-taking more than low power, which in turn should decrease perspective-taking more than social power.

On the other hand, it may be that personal power affects perspective-taking less than social power does. The subjective sense of power has been shown to mediate at least some of the effects of social power (Anderson & Berdahl, 2002), and the cues of status and dominance that social power provides may be important to social power’s effect on perspective-taking.

Broadening our view beyond perspective-taking, this general line of research—distinguishing between the psychological outcomes of social and personal power—could be useful in expanding approach-inhibition power theory to include personal power. Many questions remain. Does high and low personal power come along with the same biological markers as social power, such as increased testosterone or glucocorticoids (Sapolsky, 2004)? How would a non-social, personal power condition compare to social power on measures of relative action, abstract thinking, stereotyping, executive functioning, and risk taking? In which areas are the effects of power the results of status and dominance cues and in which are they the results of freedom and efficacy?

The “empowerment” of oppression theorists such as feminists and other post-modern philosophers, anarchists and other libertarian philosophers, post-modern sociologists, and some alternative-education proponents, is often construed as more than just a human right. It is a benign form of power, (theoretically) coming with all of the benefits and few of the drawbacks and ethical quandaries inherent in hierarchical power. And empowerment, the ability of individuals and groups to be independent, self-governing, and to take part in decision-making processes, has much in common with personal power. Power corrupts, but empowerment does not. By distinguishing the psychological effects of having personal power and social power, we can begin to examine this assertion as an empirical question.


Abele, A. E., Uchronski, M., Suitner, C., & Wojciszke, B. (2008). Towards an operationalization of the fundamental dimensions of agency and communion: Trait ratings in five countries considering valence and frequency of word occurrence. European Journal of Social Psychology, 38, 1202-1217.

Altemeyer, B. (1998). The other “authoritarian personality.” In M. P. Zanna (Ed.), Advances in Experimental Social Psychology, Volume 30 (pp. 47-92). St. Louis, MO: Academic Press.

Anderson, C., & Berdahl, J. L. (2002). The experience of power: Examining the effects of power on approach and inhibition tendencies. Journal of Personality and Social Psychology, 83,1362-1377.

Anderson, C., & Galinsky, A. D. (2006). Power, optimism and risk-taking. European Journal of Social Psychology, 36, 511-536.

Bandura, A. (1989). Human agency in social cognitive theory. American Psychologist, 44, 1175-1184.

Bargh, J. A., Raymond, P., Pryor, J. B., & Strack, F. (1995). Attractiveness of the underling: An automatic power -> sex association and its consequences for sexual harassment and aggression.Journal of Personality and Social Psychology, 68, 768-781.

Batson, C. D. (1991). The altruism question: Toward a social-psychological answer. Hillsdale, NJ: Erlbaum.

Bech, P. (1993). Rating scales for psychopathology, health status, and quality of life : a compendium on documentation in accordance with the DSM-III-R and WHO systems. New York : Springer-Verlag.

Chen, S., Lee-Chai, A. Y., & Bargh, J. A. (2001). Relationship orientation as a moderator of the effects of social power. Journal of Personality and Social Psychology, 80, 173-187.

Chirkov, V., Ryan, R. M., Kim, Y., & Kaplan, U. (2003). Differentiating autonomy and independence: A self-determination theory perspective on internalization of cultural orientations and well-being. Journal of Personality and Social Psychology, 84, 97-110.

Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journals of Personality and Social Psychology, 44, 113-126.

Davis, M. H., Conklin, L., Smith, A., & Luce, C. (1996). Effect of perspective taking on the cognitive representation of persons: merging of self and other. Journal of Personality and Social Psychology, 44, 113-126.

DeCharms, R. (1968). Personal causation: The internal affective determinants of behavior.New York: Academic Press.

Eisenberg, N., Murphy, B. C., & Shepard, S. (1997). The development of empathic accuracy. In W. J Ickes (Ed.), Empathic accuracy (pp. 73-116). New York: Guilford Press.

Fiske, S. T. (1993). Controlling other people: The impact of power on stereotyping. American Psychologist, 48, 621-628.

Fiske, S. T., & Berdahl, J. (2007). Social power. In A. W. Krulanski & E. T. Higgins (Eds.),Social psychology: Handbook of basic principles (2nd ed.) (pp. 678-692). New York, NY: Guileford Press.

Galinsky, A. D., Gruenfeld, D. H., & Magee, J. C. (2003). From power to action. Journal of Personality and Social Psychology, 85, 453-466.

Galinsky, A. D., Magee, J. C., Gruenfeld, D. H., Whitson, J. A., & Liljenquist, K. A. (2008). Power reduces the press of the situation: Implications for creativity, conformity, and dissonance.Journal of Personality and Social Psychology, 95, 1450-1466.

Galinsky, A. D., Magee, J. C., Inesi, M. E., & Gruenfeld, D. H. (2006). Power and perspectives not taken. Psychological Science, 17, 1068-1074.

Galinsky, A. D., & Moskowitz, G. B. (2000). Perspective-taking: Decreasing stereotype expression, stereotype accessibility, and in-group favoritism. Journal of Personality and Social Psychology, 78, 708-724.

Gebauer, J. E., Riketta, M., Broemer, P., & Maio, G. R. (2008). “How much do you like your name?”: An implicit measure of global self-esteem. Journal of Experimental Social Psychology,44, 1346-1354.

Ghaed, S. G., & Gallo, L. C. (2006). Distinctions among agency, communion, and unmitigated agency and communion according to the interpersonal circumplex, five-factor model, and social-emotional correlates. Journal of Personality Assessment, 86, 77-88.

Hall, J. A., Coats, E. J., & LeBeau, L. S. (2005). Nonverbal behavior and the vertical dimension of social relations: A meta-analysis. Psychological Bulletin, 131, 898-924.

Hass, R. G. (1984). Perspective taking and self-awareness: Drawing an E on your forehead.Journal of Personality and Social Psychology, 46, 788-798.

Hegelson, V. S., & Fritz, H. L. (1999). Unmitigated agency and unmitigated communion: Distinctions from agency and communion. Journal of Research in Psychology, 33, 131-158.

Hegelson, V. S., & Fritz, H. L. (1998). A theory of unmitigated communion. Personality and Social Psychology Review, 2, 173-183.

Hiroto, D. S., & Seligman, M. E. P. (1975). Generality of learned helplessness in man. Journal of Personality and Social Psychology, 31, 211-327.

Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, Approach, and Inhibition.Psychological Review, 110, 265-284.

Keltner, D., & Robinson, R. J. (1997). Defending the status quo: Power and bias in social conflict. Personality and Social Psychology Bulletin, 23, 1066-1077.

Kipnis, D. (1972). Does power corrupt? Journal of Personality and Social Psychology, 24, 33-41.

Kohlberg, L. (1976). Moral stages and moralizations: The cognitive-developmental approach. In T. Lickona (Ed.), Moral development and behavior (pp. 31-53). New York: Holt, Rinehart & Winston.

Lammers, J., Gordjin, E. H., Otten, S. (2008). Looking through the eyes of the powerful.Journal of Experimental Social Psychology, 44, 1229-1238.

Nowicki, S., & Carton, J. (1993). The measurement of emotional intensity from facial expression.Journal of social Psychology, 133, 749-750.

Nowicki, S., & Duke, M. P. (2001). Nonverbal receptivity: The diagnostic analysis of nonverbal accuracy (DANVA). In J. A. Hall & F. T. Bernieri (Eds.) Interpersonal Sensitivity: Theory and Measurement (pp. 183-198). Philadelphia, PA: Lawrence Erlbaum Associates.

Overbeck, J. R., & Park, B. (2001). When power does not corrupt: Superior individuation processes among powerful perceivers. Journal of Personality and Social Psychology, 81, 549-565.

Patterson, M. L., & Stockbridge, E. (1998). Effects of cognitive demand on judgment strategy on person perception accuracy. Journal of Nonverbal Behavior, 22, 253-263.

Ric, F. (1997). Effects of control deprivation on subsequent use of stereotypes. Journal of Social Psychology, 137, 333-342.

Ric, F., & Scharnitzky, P. (2003). Effects of control deprivation on effort expenditure and accuracy performance. European Journal of Social Psychology, 33, 103-118.

Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55, 68-78.

Sapolsky, R. M. (2004). Why zebras don’t get ulcers: The acclaimed guide to stress, stress-related diseases, and coping (3rd ed.)New York: Henry Holt and Company.

Simkin, D. K., Lederer, J. P., & Seligman, M. E. (1983). Learned helplessness in groups.Behavior Research and Therapy, 21, 613-622.

Smith, P. K., Jostmann, N. B., Galinsky, A. D., & van Dijk, W. W. (2008). Lacking power impairs executive functions. Psychological Science, 19, 441-447.

Smith, P. K., & Trope, Y. (2006). You focus on the forest when you’re in charge of the trees: Power priming and abstract information processing. Journal of Personality and Social Psychology, 90, 578-596.

Snodgrass, S. E., Hecht, M. A., & Ploutz-Snyder, R. (1998). Interpersonal sensitivity: Expressivity or perceptivity? Journal of Personality and Social Psychology, 74, 238-249.

Spence, J. T., Helmreich, R. L., & Holahan, C. K. (1979). Negative and positive components of psychological masculinity and femininity and their relationships to self-reports of neurotic and acting out behaviors. Journal or Personality and Social Psychology, 37, 1673-1682.

Tjosvold, D., & Deemer, D. K. (1980). Effects of controversy within a cooperative or competitive context on organizational decision making. Journal of Applied Psychology, 65, 590-595.

Van Dijke, M., & Poppe, M. (2006). Striving for personal power as a basis for social power dynamics. European Journal of Social Psychology, 36, 537-556.

Walker, L. J. (1980). Cognitive and perspective-taking prerequisites for moral development.Child Development, 51, 131-139.

Winter, D. G., & Barenbaum, N. B. (1985). Responsibility and the power motive in women and men. Journal or Personality, 53, 335-355.

Weick, M., & Guinote, A. (2008). When subjective experiences matter: Power increases reliance on the ease of retrieval. Journal of Personality and Social Psychology, 94, 956-970.

Zakrisson, I (2005). Construction of a short version of the right-wing authoritarianism (RWA) scale. Personality and Individual Differences, 39, 863-872.

Appendix A

Post-Experimental Questionnaire

All questions were answered on a 7-point scale. Questions 1 through 3 were anchored by 1 = “Not at all” and 7 = “Very much.” Question 4 was anchored by 1 = “Not at all powerful” and 7 = “Very powerful.” Questions 5 through 8 were anchored by 1 = “I agree completely” and 7 = “I disagree completely,” and reverse coded.

For the following 8 questions, please select the answer that best describes your experience.

1. As you wrote about your life, to what extent did you feel as if you had power over someone else?

2. As you wrote about your life, to what extent did you feel as if another person had power over you?

3. As you wrote about your life, to what extent did you feel as if you had power over yourself?

4. As you wrote about your life, how powerful did you feel?

5. As you wrote about a situation in your life, you felt as if you had control over someone else’s destiny.

6. As you wrote about a situation in your life, you felt as if someone else had control over your destiny.

7. As you wrote about a situation in your life, you felt as if you had control over your own destiny.

8. As you wrote about a situation in your life, you felt very powerless.

Appendix B

High Social Power Writing Task Instructions

On the next screen, you will be asked to spend about 5 minutes thinking about and writing about a time in your life when you felt like you had power over someone else. That is, you will write about a time when you had control over someone else’s actions or something that they wanted. Any time or event that fits this description will do. When it happened or the context in which it happened is not important – for example, it could concern school, family, work, or your personal life.

This task is important. Therefore, before continuing and writing about this time in your life, spend a moment RIGHT NOW to think of the time or event in your life that best fits this description. When you have this time or event firmly in mind, you may click “continue” and begin writing

Low Power Writing Task Instructions

On the next screen, you will be asked to spend about 5 minutes thinking about and writing about a time in your life when you felt like someone else had control over your actions or something that you wanted. Any time or event that fits this description will do. When it happened or the context in which it happened is not important – for example, it could concern school, family, work, or your personal life.

This task is important. Therefore, before continuing and writing about this time in your life, spend a moment RIGHT NOW to think of the time or event in your life that best fits this description. When you have this time or event firmly in mind, you may click “continue” and begin writing.

High Personal Power Writing Task Instructions

On the next screen, you will be asked to spend about 5 minutes thinking about and writing about a time in your life when you felt like you had complete control over your own life and were free to do whatever you wanted, without interference. Any time or event that fits this description will do. When it happened, or the context in which it happened, is not important – for example, it could concern school, family, work, or your personal life.

This task is important. Therefore, before continuing and writing about this time in your life, spend a moment RIGHT NOW to think of the time or event in your life that best fits this description. When you have this time or event firmly in mind, click “continue” to progress to the next screen.


Appendix C

High Social Power Resource Allocation Task Instructions


You have been assigned to the resource allocation task and are in charge of deciding how many tickets for the drawing for $100 that you and the other person involved in this task each will receive.

Decide on a division of 7 tickets between yourself and the other person:

7                      7 for yourself and 0 for the other person

6                      6 for yourself and 1 for the other person

5                      5 for yourself and 2 for the other person

4                      4 for yourself and 3 for the other person

3                      3 for yourself and 4 for the other person

2                      2 for yourself and 5 for the other person

1                      1 for yourself and 6 for the other person

0                      0 for yourself and 7 for the other person

Low Power Resource Allocation Task Instructions


You have been assigned to the resource allocation task, but you do not have control over the allocation of the tickets for the drawing. Instead, predict how many tickets the person making the allocation decision will take and how many will be given to you.

Remember not to pretend you are determining the allocation. You should try to predict what the person making the decision is going to do.

0          0 for you and 7 for them

1          1 for you and 6 for them

2          2 for you and 5 for them

3          3 for you and 4 for them

4          4 for you and 3 for them

5          5 for you and 2 for them

6          6 for you and 1 for them

7          7 for you and 0 for them

High Personal Power Resource Allocation Task Instructions


As part of this study, with your consent, you will be entered into a drawing at the end of the study. You will have seven full chances to win.

There are four prizes to choose from. Please choose which of these prizes you would prefer, should you win. If you choose number 4, a text box will open for you to specify which business you would like a gift certificate from.

1          $100 cash

Appendix C continued

2                      $110 gift certificate at Market of Choice on Franklin Blvd.

3                      Bose Triport In-Ear Headphones, worth $99.95

4                      $105 gift certificate to local or online business of your choice <specify>

5          Opt out of drawing

Table 1

Correlations and reliability (Cronbach’s α) of major variables

E           DANVA UC            UA         WHO-5      RWA     Name     LP comp   HPP comp  HSP comp


DANVA     -.039         .504

UC             -.028         .224*      .651

UA              -.075         -.195*     -.232*    .693

WHO         -5 .047      -.004       -.031      -.132*      .698

RWA          -.068         .058        -.021      .017         .065       .820

Name         -.122         .000         -.115      -.008        .121       .132

LP comp     .123          .159         .157*     -.001        -.225*    .011       -.011         .632

HPP comp  .021          .013         -.111      -.018        .201*     -.074      .046          -.554*    .608

HSP comp  -.006         .084       -.024      .109       .207*     .164*     .081          -.143      .089           .635

Note. Reliability coefficients are on the diagonal. E = E direction (coded 1 = perspective-taking, 0 = no perspective-taking); DANVA = adjusted DANVA scores; UC = Unmitigated Communion Scale; UA = Unmitigated Agency Scale; WHO-5 = the WHO-5 Well-Being Index; RWA = Right-Wing Authoritarianism Scale; Name = self-esteem measure; LP comp = composite low power measure; HPP comp = personal power composite measure; HSP comp = composite social power measure.

* Correlation is significant, p < .05

Table 2

Correlations of composite power variables with other dependent variables

Condition  ____

Variables LP HPP HSP Overall

LP comp HPP comp                   -.421a*    -.561a*    -.502a*    -.554a*

HSP comp                                    -.233a -.043a -.100-.143a

HPP comp  HSP comp              .089abc -.213a .388b*     .089c

E direction LP comp                 .212a .234a -.056a .123a

HPP comp                                   .021a -.112a .   160a .021a

HSP comp                                   -.098a -.171a .267b*     -.006ab

DANVA LP comp                      .020a .038a .397b*     .159ab*

HPP comp                                  .205a -.200b .115ab .013ab

HSP comp                                 -.139a .232.140ab .084a

UA          LP comp                     .040a .016.096-.001a

HPP comp                                 -.126a .042a -.061-.018a

HSP comp                                .146a -.008a .083.109a

WHO-5   LP comp                   -.142a -.278a*    -.369a*    -.225a*

HPP comp                                 .239a .204a .191.201a*

HSP comp                                .104a .219a .371a*     .207a*

UC          LP comp                    .303a*     .211a .143.157a*

HPP comp                                -.011a -.351a*    -.097-.111a

HSP comp                                 .039a      -.002a     -.135a     -.024a

RWA      LP comp                    -.043a .115a -.103.011a

HPP comp                               -.039a -.071a -.082-.074a

HSP comp                                  .135a .207a .223.164a*

Name-liking    LP comp         -.036a -.262a*    -.001-.011a

HPP comp                                 .005a .163a .126a         .046a

HSP comp                                  .026a .284a*     .128.081a

Note. Within rows, correlations with different subscripts differ significantly, < .05, using Fisher’s r to z transformations. LP comp = composite low power measure; HPP comp = personal power composite measure; HSP comp = composite social power measure. For condition, LP = low power condition; HPP = personal power condition; HSP = social power condition. DANVA = adjusted DANVA scores; UA = Unmitigated Agency Scale; WHO-5 = the WHO-5 Well-Being Index; UC = Unmitigated Communion Scale; RWA = Right-Wing Authoritarianism Scale; Name-liking = self-esteem measure.

* Correlation is significant, p < .05

The Illusion of Control

[First published on Nathen’s Miraculous Escape, December 24, 2008.]

I do not believe in free will. I am what you might call a free-will agnostic. Hundreds of years of thought and debate have brought us no closer to agreement about whether an individual’s consciousness is capable of true agency. I can understand, though, how the idea that one might not control even one’s own actions is distressing; a perception of control makes stressful situations easier to cope with, while a perceived lack of control, and especially perceived loss of control, make stressful situations more distressing (Sapolsky, 2004). Note that it is our perception, however, not necessarily the reality of control, which is important. As it turns out, our perceptions of control are not reliable. The only situation in which we can be sure about how much control we have is when we demonstrably have no control at all, and in those situations people often still believe they have control.

Psychologists call this phenomenon the illusion of control: The tendency to overestimate how much control one has over an outcome. Participants in an experiment by Tennen and Sharp (1983), for example, believed that they controlled a blinking light about one third of the time, when in fact there was no connection between their button and the light. Even more surprising, experimental participants came to believe that they were good at guessing the outcome of the notoriously random coin toss (Langer & Roth, 1979). All it took were a few successes early on. Ellen Langer, in an early series of experiments (1975), found that the key to producing the illusion of control is the presence of skill cues, like practice, familiarity, choice, competition, and personal involvement. That is, if a task reminds you of a time when you used skill to influence what happened, you are likely to believe you have more control over the outcome of the task than you do. This tendency creates an inaccurate perception of reality which can have serious negative consequences but also may have beneficial effects.

The desire to be accurate in our perceptions and judgments is a basic human motivation, among psychological motivations second only to our need for self-esteem (Aronson, Wilkert, & Akert, 2007). This desire serves us well, especially in cases when a discovery of error on our part tempts us to distort reality to save our self-esteem; it would be much more difficult to learn from our mistakes if we were never compelled to admit them. In the case of the illusion of control, however, the mistaken belief in control will often not even come to our awareness. Usually there is no experimenter at the end of our task, gently debriefing us on how our button is not connected to the light. There are cases when the mistaken belief in control is benign enough—the experimental participants thinking they controlled a light were not hurting themselves or others, nor does a basketball player, swaying to the left, trying to move his shot toward the basket—but there are situations where having an inaccurate sense of control can be unfortunate.

The illusion of control can lead to overconfidence, a bad quality for those in dangerous or critical situations. A surgeon, for example, who believed she had more control over bleeding than she did, might kill her patient with a bad decision. Pilots overconfident in their control of an aircraft’s elevation or attitude are risking their own and their passengers’ lives. I expect that those who have lives explicitly in their hands are highly trained to know exactly where their control ends, so that they will take proper precautions, but it may be that the illusion of control causes injuries and fatalities in these and similar fields. And in less explicit situations, such as driving on the highway, the training may not be as thorough. A truck driver, for example, who fell asleep behind the wheel, likely did so believing that he had control over his staying awake. And even if life is not at stake, there can be loss; dice gamblers, for example, often believe that their strategy of playing increases their odds, when it demonstrably does not (Henslin, 1967, as cited by Langer, 1975). There may be interpersonal costs, too. If I believe that I am in control of a relationship, I might not give my partner the attention she needs, and eventually lose her. The illusion of control may play a part in addiction, too: The well-known first hurdle to recovery is admitting a lack of control.

I suspect that the best antidote to the illusion of control is to use knowledge of the phenomenon and how it operates to maintain skepticism about your level of control. Awareness alone should help. If the question of how much control I have comes up, I will be more skeptical, knowing that people tend to overestimate their level of control. The knowledge that the illusion tends to be stronger when skill cues are present should help, too: If I am feeling more confident about an outcome because I’ve had practice at a task or even because I am allowed to make a choice in it, this might trigger my memory of the illusion of control and lead me to carefully consider base rates. The trick is maintaining an empirical mindset, where questions like “How much control do I have here?” present themselves to the consciousness as questions, and knowledge of the ways we fools ourselves should help with that.

Another antidote to the illusion of control is more troubling: depression. It turns out that depressed people are realistic in their assessments of control where non-depressed people are not (Alloy & Abramson, 1979, as cited by Seligman, 2002). In another version of the blinking light experiment, non-depressed participants believed that they had control of the light about a third of the time, but depressed participants could tell that they had no control. This came to be called depressive realism. Happy people, on the other hand, tend to have a flexible locus of control. That is, they tend to believe that they are in control only when it serves their well-being to believe so. When things are going well, happy people will think they are responsible, but when things are not going well, they think that it is because of passing circumstances (Seligman, 2002). (Understand that this is not a conscious strategy, but a non-conscious tendency.)

It may be that depression causes depressive realism, but it also may be the other way around: A tendency to believe we have control keeps us from getting depressed, while a tendency to be realistic about control causes us to become depressed. Everyone is a victim of circumstances to some degree—social psychology is a mountain of evidence for that—and lack of control is stressful, especially if a situation is unpleasant. That stress may be ameliorated to some degree by an incorrect belief in control, as long as that belief has no serious negative consequences. In that case, the inability to fool ourselves may be the problem for depressed people. And if in fact we do not have free will, and any sense of control we have is illusory, that illusion may be what is keeping us happy.


      Alloy, L. B., & Abramson, L. Y. (1979). Sadder but wiser.

Journal of Experimental Psychology, 108,

      Aronson, E., Wilson, T. D., & Akert, R. M. (2007).

Social Psychology. 

    Upper Saddle River, NJ: Pearson Prentice Hall.
      Henslin, J. M. (1967). Craps and magic.

American Journal of Sociology, 73

    , 316-330.
      Langer, E. (1975). The illusion of control.

Journal of Personality and Social Psychology, 32,

      Langer, E. & Roth, J. (1979). Heads I win, tails it’s chance: The illusion of control as a function of the sequence of outcomes in a purely chance task.

Journal of Personality and Social Psychology, 32,

      Sapolsky, R. M. (2004). Why is psychological stress stressful? In

Why zebras don’t get ulcers: The acclaimed guide to stress, stress-related diseases, and coping 

    (pp. 252-270). New York: Henry Holt.
      Seligman, M. E. (2002). Happy but dumb? In

Authentic Happiness, 

    (pp. 37-39). New York: Free Press.
      Tennen, H. & Sharp, J. P. (1983). Control Orientation and the Illusion of Control.

Journal of Personality Assessment, 47,


Social Psychology Journal: My Changing Bias

[First published Nathen’s Miraculous Escape, December 21, 2008.]
      I first took the race Implicit Associations Test (or “race IAT”), a test of non-conscious attitudes toward African Americans, several years ago after reading about it in Malcolm Gladwell’s Blink. In the IAT paradigm, a computer program measures how quickly you associate positive and negative words with African Americans compared to Whites. I was unhappy with the program’s conclusion, that I moderately favored Whites over African Americans, as this bias runs directly counter to my belief system. My disappointment and embarrassment were tempered somewhat, though, by the fact that Gladwell, an African American, reported receiving the same results that I had. Being implicitly racist is an effect of growing up in a racist society1; seventy percent of those who took the race IAT before me had been biased against African Americans to some degree (Project Implicit, 2008). This bias can change, though. Rudman, Ashmore, and Gary (2001), for example, found that race IAT scores improved for volunteers that took a seminar on prejudice and conflict who liked the African American teacher and formed friendships with African Americans during the semester.

      Although I seem to have a weak implicit bias as measured by the IAT, when it comes to my conscious beliefs, I have thought of racism as a big problem for as long as I have understood the concept. It is easy to trace the sources of that belief: First, I come from an egalitarian-minded family, for whom any assertion of differences between races was met with suspicion or contempt. My parents strictly controlled my media exposure, too;2 my two earliest memories of race in the media are watching Roots and a documentary on the Little Rock desegregation process. I remember being confused and enraged by the way people treated each other. Second, my closest friends throughout my life have been even more vehemently anti-racist than my parents, leading me to take part in just the kinds of discussions and self-examination that should lead to shifts in explicit prejudice (Rudman et al., 2001).

Implicitly, I’ve had more trouble, as my IAT results indicated, but that was more difficult to discern. I was in my 20s before I had an experience that revealed my non-conscious racism. I was waiting for my connection in the Sacramento Greyhound station at 4 am, very tired and anxious, and I dozed off for a moment. When I started awake, disoriented, I was vaguely afraid that someone had taken my bag, and scanned the room for who it might have been. My eyes went directly to the one African American man in my field of vision. Then I quickly realized two things. My bag was still with me, and the man I had looked at was a highly unlikely candidate for a thief. He was a middle-aged, conservatively dressed man, talking quietly with his three young children. I was embarrassed and disturbed. For the first time I had the sense that I was subtly possessed by racism. The stress I was experiencing had undermined my controlled processing, and so my implicit attitude had leaked out.3

The roots of my automatic stereotyping are also easy to see: I grew up in a rural, white town, with many racist peers and little opportunity to individuate African Americans. I did not seek out opportunities to do so, either, which left me prey to illusory correlations with stereotypes: Of the two African Americans my age, one happened to be the best break dancer I knew and the other the fastest sprinter. My high school had three African Americans attending, one of whom had a locker next to mine. He had a conversation with his twelve-inch penis every day as we got dressed for gym.4 Later, for college, I moved to a more diverse area, but also started seeing more TV and movies about gang violence and crime. There were also many super aggressive African American panhandlers in downtown Oakland, while I waited for my bus at night. There was when I worked in a restaurant in a poor, African American neighborhood in San Leandro; I remember reminding myself that it was all socioeconomics, but the fact that I could count on bigger tips, on average, from the White people who came in made it hard not to feel happier when they did.

On the other hand, most of my coworkers at that restaurant, whom I became very fond of, were also African American, and I spent a lot of time getting to know them individually. Many of my fellow students at the trade school I attended at San Francisco State were African American as well. Those situations provided me with all of the elements that should result in decreased prejudice against African Americans—mutual interdependence, common goals, equal status, many informal, interpersonal contacts, and a social norm of equality (Aronson, Wilson, & Akert, 2007, p. 451)—as well as the opportunity to change my implicit attitudes, as the participants in Rudman and colleagues’ (2001) seminar on prejudice had, by liking my African American supervisors and making friends among this “out-group,”5 during the same period that I had my two most anti-racist housemates ever, and lots of conversations about race.

During the period of my social psychology journal, I started thinking that I stood a good chance of doing better on the race IAT. During the last several years I have stepped up my thinking and emotional work around racism, and have also had more fortunate experiences: I’ve been in an Eliminating Racism support group and two of my favorite counselors are African American men. I had an African American man as a recording studio client, who, though a gangsta rapper, drug dealer, and pimp, I came to know as a sophisticated and complex person. I took a sociology class called America’s Peoples, which looked at the history of oppression in America and included a lot of class discussion. I took a Family and Human Services class called Diversity in Human Services, which was a series of lectures by minority speakers about what it was like to be a member of their group, again with a lot of discussion and introspection. Both of these classes had opportunities for perspective taking, which Galinsky and Moskowitz (2000) found to decrease both implicit and explicit stereotyping. I have good relationships with my honors thesis advisor and my practicum supervisor, both of whom are African American. Finally, I have been elated ever since the election of Barack Obama, an African American, to the presidency of the United States, and watching his speeches with great appreciation.

I was surprised and very frustrated to find that second attempt at the race IAT showed the same results as my first had: I moderately favored Whites over African Americans. Perhaps I shouldn’t have been surprised; it is notoriously difficult to change your score. On the other hand, the results of the test, as they are available online, are not very fine-grained; all response times are measured to the millisecond, but all that variability is funneled into seven categories—strong, moderate, mild, or no preference for or against African Americans. It was quite possible that I had come very close to improving my results without knowing it, so I engaged in counter-factual thinking. Perhaps I was a victim of stereotype threat; perhaps my fear of confirming the stereotype that White people are racist caused me to do so.6 Or maybe it had been sequence effects that had foiled me; the version of the test I had taken had started with negative words associated with African American faces. I thought that I would have done better if it had started with positive words associated with African American faces. But it hadn’t and I hadn’t.7

After I calmed down, I took an IAT for preferences between the presidential candidates, Obama and McCain, at the same time as testing preferences for Whites or African Americans. This time, the results said that I moderately preferred Obama over McCain, but also that I stronglypreferred African Americans to Whites. The IAT is supposed to be quite reliable, so I was surprised about the change8—something was going on that could change my results pretty radically. I started thinking that part of my score was my skill on the task—that one or more of my inept answers were serving as extreme outliers, 9 biasing my scores one way or the other. I took the race IAT for a third time, and this time scored as not biased at all, making me the happiest yet, but even more curious. Maybe I spiked my appreciation for African Americans by priming myself with images of Obama and McCain,10 and in the future my score would again show my moderate bias. Or perhaps my greater facility with the test had brought out my true non-bias.11 Or perhaps itwas the sequence effects, after all, because the third time had paired African American faces with good words first. Finally, if the results could vary this much, maybe my single attempt at the test years ago had been an aberration as well, misreporting my true bias or lack of bias.12

Over the next few days, I took the race IAT three more times, and each time scored the same—unbiased—so I feel pretty safe now to say that this is my true score. I feel pleased about that but I don’t believe that it means I am free from bias or prejudice toward African Americans. Strictly speaking, it just means that I don’t associate negative words like “hate,” “pain,” and “evil,” or positive words like “laughter,” “glorious,” and “love,” with images of African American faces any quicker than I do with images of White faces. Still, assuming that my first score of moderately biased was my true score of several years ago, it shows some progress—my non-conscious processes have caught up with my conscious beliefs to some degree. And even if my original score was misleading, I am still pleased with my unbiased score.

[Comments by instructor, Sean Laurent:]

Good you should be pleased. Still, there are ways, if one is consciously trying, to “beat” the IAT (there has been some discussion in the literature about this), such as described above.

This is a well-crafted paper, full of insight and very interesting to read. Did you actually keep a journal for the two weeks, or just reflect on implicit and explicit racism in your life as you wrote the paper. Either way, it came out well, and tells a very nice story, linking ideas in your life to specific experiences and showing that you have reflected on the topic. Well done. 48/50 I might like to use this paper as an example of a reaction paper that is done correctly.

1 Is it? This is kind of a sweeping statement – it implies that the problem of why people are implicitly racist is solved.

2 You might mention why this is important in terms of knowledge of racial stereotypes driving certain effects like shooter bias, etc.

3 Very vivid storytelling here.

4 That’s funny….

5 This last part seems tacked on and could be safely reorganized or made into another sentence.

6 Good point, possible.

7 Usually this doesn’t matter much.

8alternatively, it could be that you have very strong associations with Barak, and that while he was serving as an exemplar of the group, this overrode the automatic stereotyping)

9 Also possible.

10 Quite possibly.

11 Or you might have deliberately or unconsciously slowed your responses to BOTH categories in order to speed your pairings of whites with negative, etc.

12 Probably not. Remember, from articles like the shooter bias, that even a strong knowledge of the cultural stereotype can drive implicit effects, so…


      Aronson, E., Wilson, T. D., & Akert, R. M. (2007).

Social Psychology. 

    Upper Saddle River, NJ: Pearson Prentice Hall.
      Galinsky, A. D., & Moskowitz, G. B. (2000). Perspective-taking: Decreasing stereotype expression, stereotype accessibility, and in-group favoritism.

Journal of Personality and Social Psychology, 78, 

      Gladwell, M. (2005).


    . New York: Little, Brown.
      Project Implicit (2008). Retrieved November 16, 2008, from


      Rudman, L. A., Ashmore, R. D., & Gary, M. L. (2001). “Unlearning” automatic biases: The malleability of implicit prejudice and stereotypes.

Journal of Personality and Social Psychology, 81,


Reaction to “On the Confirmability and Disconfirmability of Trait Concepts”

[First published on Nathen’s Miraculous Escape, November 7, 2008.]

Some time ago, I was present for a conversation between two of my friends in which it came out that one had lied to the other. It was immediately clear that there was nothing the one who had lied could do to make the other believe anything else he said. It could easily take years of acting with perfect integrity, I thought, to build that trust back, if it could be built back. This is a major point of the article, “On the Confirmability and Disconfirmability of Trait Concepts,” by Rothbart and Park: Where favorable trait ascriptions, like “honest” and “truthful,” are easy to lose, unfavorable trait ascriptions, like “deceitful” and “deceptive” are difficult to lose. This is a problem not only for individuals like my friend who lied, but for whole groups of people laboring under the unjustly and inaccurately received negative trait ascriptions called stereotypes.

Rothbart and Park had participants rate 150 trait adjectives on how easy they were to acquire and lose in three dimensions: how often the environment provides opportunities to show or disprove each trait, to what extent each trait had clear, corresponding behaviors to indicate it, and how many times a behavior would need to be seen before a corresponding trait could be inferred or disproved. What they found in part was that two of the dimensions, the first and third were related to how favorable a trait was rated. That is, when it comes to how often the environment provides opportunities to show or disprove each trait, favorable traits are easy to get and easy to lose, whereas unfavorable traits are hard to get and hard to lose. On the dimension of how many times a behavior would need to be seen before a corresponding trait could be inferred or disproved, favorable traits were hard to get and easy to lose, where unfavorable traits were easy to get and hard to lose. Notice that in both of these relationships, unfavorable traits were difficult to lose; there are few occasions provided to disprove them, and disconfirming behaviors seem more ambiguous than confirming behaviors.

This pattern may apply to the traits ascribed to groups as well as those ascribed to individuals, as trait ascriptions are part of our schemas about individuals as well as groups. Schemas are cognitive information-sorting systems that function in part as filters on our perception, causing schema-consistent behaviors to be memorable and schema-inconsistent behaviors to go unnoticed more easily and be forgotten more easily, if noticed (Aronson, Wilson, & Akert, 2007). This may be a problem-compounding effect for those who are victims of a negative cultural stereotype.1 Rothbart and Park’s (1986) participants were not asked to consider race, and reported that it was difficult to imagine someone disconfirming a negative trait. It seems that it would be that much more difficult for a person of color to disconfirm a negative trait that coincided with the stereotype of their group. At least three factors would converge against it: The environment doesn’t provide many opportunities to disconfirm unfavorable traits, disconfirming behaviors are much more ambiguous than confirming behaviors, and stereotypers2 are less likely to notice disconfirming behaviors and more likely to forget them.

There is another troubling point. Rothbart and Park (1986) predicted that, once acquired, a negative stereotype will persist in a stereotyper’s mind unless they have contact with the stereotyped group, for otherwise there is no opportunity to observe disconfirming behavior. That reasoning is congruent with the evidence they collected, but is somewhat contrary to, or at least complicated by, the evidence collected by Correll, Park, Judd, and Wittenbrink (2002) investigating “shooter bias,” the increased tendency for experiment participants to mistakenly shoot African American target images more than white target images. The authors of this study found that the bias was predicted in part by the level of contact with African Americans. In keeping with Rothbart and Park’s prediction, increased contact with African Americans did correlate with decreased personal endorsement of stereotypes of African Americans, but also with this increased, non-conscious acting on a stereotype, that an African American is more likely to have a gun.3

There are caveats to consider in this line of thinking. First, “gun-possessing” was not among the traits rated by Rothbart and Park’s participants, and so it is a stretch to compare shooter bias with even the related traits that were included, like “violent,” “mercenary,” and “hostile.” Second, trait ascription may be more closely related to the personal endorsement of a stereotype than the unconscious application of a stereotype, like shooter bias. Third, Rothbart and Park used imaginability as their operationalization of their three major factors. In other words, to determine how many occasions the environment provides for confirming or disconfirming a trait, for example, they asked participants how easy it was to imagine such occasions for each trait. Though Rothbart and Park found that their operationalization provided reliable statistics in that they were stable from participant to participant, imaginability may or may not accurately predict how difficult a trait will be to gain or lose. Fourth, Rothbart and Park predicted that contact with a stereotyped group would be necessary to disconfirm a trait, which is probably true, but it does not follow that contact would be sufficient to disconfirm a trait.4

Still, that stretch is worth entertaining because of how closely all three elements, trait ascription, belief in stereotypes, and acting on stereotypes, are related conceptually. The possibility that contact with a stereotyped group could simultaneously be involved in decreasing belief in the stereotype while increasing the amount of unconsciously acting on the stereotype begs for more investigation. One possible place to start is to check if the effects replicate in other stereotypes, perhaps having participants rate job applications attributed to either whites or African Americans, to see how bias in the ratings might correlate with participants’ personal endorsements of African American stereotypes and contact with African Americans.5 There is also the possibility that “unconsciously acting on a stereotype” breaks down so that on tasks involving time-forced errors, like the shooter-bias video game, bias correlates positively with level of contact, while on tasks that leave time for conscious deliberation, like the rating of job applications, bias would correlate negatively with level of contact.6 This could be checked by decreasing the amount of time each participant had with each application to see if time-forced biases resulted.

It does seem clear that, at least in people’s imaginations, unfavorable trait ascriptions are difficult to shed, and that this probably applies both to individuals and to groups. The complexities of that process remain to be investigated, or have, perhaps, been investigated in work by Rothbart and Park (and others) since their 1986 paper. It also remains how Rothbart and Park’s findings relate to others’ such at Correll and colleagues’ (2002): Just how deep does changing one’s mind about a stereotype go? Under what circumstances will changing one’s mind about the truth of a stereotype produce a change in behavior? Are some behaviors immune to influence by such a mind change? If so, what processes will affect these behaviors?



Aronson, E., Wilson, T. D., & Akert, R. M. (2007). Social psychology. Upper Saddle River, NJ: Pearson Education.

Correll, J., Park, B., Judd, C. M., & Wittenbrink, B. (2002). The Police Officer’s Dilemma: Using ethnicity to disambiguate potentially threatening individuals. Journal of Personality and Social Psychology, 83, 1314-1329.

Rothbart, M., & Park, B. (1986). On the confirmability and disconfirmability of trait concepts.Journal of Personality and Social Psychology, 50, 131-142.


Very good, tight paper, with lots of interesting questions and thoughts that might be pursued. This is just about the epitome of what I’m looking for in a reaction paper – strong evidence of critical thinking, application of constructs across domains, integration of disparate material, and demonstration of knowledge that you probably would not have had before taking this course. 50/50


Btw, may I have permission to post this paper as an example of a great reaction paper for future classes? J


1 Awk., rephrase.

2 Strange use of this word; I might rephrase.

3 This last sentence is hard to parse…I’d rephrase. Also, did Correll et al. show that Ps thought AAs were more likely to have a gun, or were they just more likely to shoot even when a gun wasn’t present?

4 Good point.

5 This would address the first part: whether contact reduces personal bias and/or belief in the cultural ST – but how would this relate to further automatic application of the same ST?.

6 An interesting idea, but in a non-salient domain, do you think you would see effects? I guess the question is: how much does knowledge of cultural ST or endorsement of ST relate across domains – such as “blacks are poor workers” with “blacks are more likely to have a gun.”