John Gottman Research Articles

Interlocking patterns of conditional probability: The reinforcement matrix.

Abstract To further integrated behavioral and ecobehavioral models, we introduce the idea of the reinforcement matrix. We define the reinforcement matrix as relatively stable patterns of interlocking contingencies. We can conceptualize this matrix at the ontological level, consisting of varying reinforcement density of dynamic interaction. These interactions occur between interlocking elements of the social and nonsocial environment. Multiple layers of interlocking patterns increase stability and resistance of the ecology to change. These multiple environmental influences act on the organism and lead to gradual shifts in behavior over time. These behavioral shifts are economical in nature based on fluctuations in reinforcement rates, amount of reinforcement, movement toward least effort, and decreasing delay and uncertainty. While relatively stable, the entire process drifts slowly overtime. We identify and discussed factors that lead to this drift. role of metacontingencies, which can lead to rapid shifts or even disturbances in the matrix are discussed. We make suggestions for the study of communities and the development of technology. ********** The experimental analysis of behavior is a rigorous, extensive, and rapidly advancing branch of biology. (Skinner, 1974, p.255) Behavior is the locus of the interaction between organism and environment (Baum, 1994; Hineline, 1992; Skinner, 1981). Through this interaction behavior is selected ontogenetically by environmental consequences (Campbell, 1956; Kimble, 1961; Skinner, 1953, 1981), which correlate with the occurrence of the behavior over time (Baum, 1973). Thus the nature of reinforcement is primarily to serve the organism as a feedback system (Baum, 1973; Hineline & Wanchissen, 1989; Timerlake, 1993, 1995). In this view, variation of behavior is a natural occurrence. Variations that lead to less effort, increased reinforcement density (i.e., higher pay off or higher amount), or decreased aversive stimulation to which the organism is susceptible are selected for continuation in similar contexts (Herrnstein, 1970; Rachlin, Logue, Gibbon, & Frankel, 1986). For theories (Hineline) this process is called adaptation. (4) Variations that correlate with the opposite of the above are selected against recurrence. Also, in novel situations component repertoires can be rapidly integrated in a process called contingency adduction (Andronis, 1983). In this process behavioral classes combine, the product of which then undergoes the selection process (Andronis; Johnson & Layng, 1992; Layng & Andronis, 1984). Moment to moment attempts to maximize the adaptive factors may not lead to maximization over time. We call this process melioration (Herrnstein, 1990). Due to stable environmental elements, relatively stable patterns of interaction form. For example, John Gottman in studying couples discovered that non-distressed couples had a positive to a negative interaction ratio of 5:1 (Gottman, 1993; 1996). In this view we say that the environment-organism-environment system closed-looped (Hineline & Wanchissen, 1989) with contexts of high probability of reinforcement increasing the likelihood of responding (stimulus control). As expected, stimulus control degrades at different rates depending on the species (Balda & Kamil, 1989; Krebs, Healy, & Shettleworth, 1990). Interconnecting patterns of stable probability of reinforcement may lead relatively to stable patterns of interaction over time and even the development of bias (Herrnstein, 1990; McDowell, 1989; Snyder & Patterson, 1995). Bias in responding may develop after prolonged exposure to nonequivalent reinforcer parameters (McDowell; Noll, 1995). This is similar to an idea called homeostasis in organism-based theories. Shifts in stable reinforcement effects can have profound effects not just on the target but on all connected to the pattern. …

A Random Sample of Current Research

This article is a review of clinical research composed by Lee N. Johnson, PhD. Dr. Johnson discusses these research articles and their relevance to the practicing clinician. The following articles are discussed: the prediction of marital stability and marital happiness, an investigation in the timing of divorce, both authored by John Gottman; the third article reviews a 20 year prospective study of risk factors associated with suicide (by Brown, Beck, Steer, & Grisham); the fourth is a review of a study on the prediction of treatment duration in private practice (by Mueller & Pekarik); the fifth is a discussion of outcome effects of therapy (Shadish, Matt, Navarro, & Phillips). Dr. Johnson reviews an article that presents a self-report measure for adolescent parental attachment and its relation to adjustment (Vivona).

Attachment and Negative Absorption States: Implications for Treating Distressed Families

Emotion has long provided family therapists with important clues to family relationships, problems, and ways of intervening. Therapists intuitively make ongoing distinctions between emotions that should be downplayed and other emotions that should be amplified or heightened. Most therapists will move quickly to control angry and accusatory exchanges that threaten to distance and alienate family members from one another. In contrast, less obvious, more adaptive emotions, ranging from assertive anger and fear through sadness, are hidden in hostile exchanges. In the current paper, we draw on John Gottman's descriptions of dysfunctional emotion in distressed marriages, as a way of describing the presenting state of families who seek treatment. We then suggest that attachment theory and research provides an explicit guide for therapists, not only in understanding dysfunctional emotion, but more importantly, in pointing toward the hidden emotions that hold the potential for therapeutic change in families. As such, attachment theory provides an important tool for discriminating between functional and dysfunctional emotions, and guides the therapist toward emotionally focused interventions that hold the potential for moving family relationships toward more healthy and adaptive functioning.

The private practice of misleading deflection.

The private practice of misleading deflection.

A Psychophysiological Intervention for Marital Discord

The research of John Gottman and co-workers has demonstrated a correlation between the degree of physiological arousal of couples, especially husbands, during marital conflict and the eventual end of the marriage. He proposed that this phenomenon has its basis in gender differences: males require a longer time to recover from the arousal, resulting in husbands emotionally withdrawing from the relationship due to the aversiveness of the arousal. The wife eventually also withdraws; thus, the seeds of divorce are sown. The present article proposes a psychophysiological intervention, relaxation therapy with biofeedback assistance, for marital discord, based on Gottman's theory. This intervention involves the use of various relaxation techniques, and may also involve a descnsitization hierarchy. An example hierarchy is presented. It is also proposed that Gottman's model and the proposed intervention may be applicable to non-heterosexual couples.

The Relational Systems Model: An Analysis of Gottman's Findings on Marital Success

ABSTRACT This paper reviews John Gottman's research on marriages through an overlay of the recently developed Relational Systems Model (RSM) by Ray Bardill. Gottman's findings, which identified three styles of marriages that sustain themselves over time and identified those actions that indicate the resolution of conflicts or the eventual ending of the marital relationship, can be interpreted by Bardill's RSM model. Such an interpretation can be made through the concepts of self-differentiation; the realities of self, other, context and spirituality; multigenerational transmission of values and beliefs; and emotionalized right/wrong. The RSM model effectively illuminates Gottman's work and develops additional categories for illustrating why people stay in or leave relationships. The introduction of the emotionalized right/wrong concept produces a self-regulating component for the individual and a self defining component for the family. In keeping with Gottman's research, the RSM model provides integration...

Genetic Influence on Family Systems: Implications for Development

In this article, my major objective is to describe recently observed connections between genetic processes and family systems. In describing these connections I hope to show how new findings in genetics may give us a clearer idea of the role that families play in the differences among us: differences in healthy and adaptive development and differences in pathological and maladaptive development. I will provide some new data, some of it yet unpublished, from behavioral genetic research. I go somewhat beyond the data currently available, and lay out some very preliminary speculations about genes, families, and development. In doing so, I wish to take advantage of two important trends in the study of human development. The first of these is an increasing emphasis on individual differences in development rather than on stages or transitions in development that are common across individuals. The second trend is increasing attention to biological factors that shape changes in individuals across time and that account for differences in the rates and patterns of change. I will argue here, as have several before me, that a detailed study of family process in relationship to biological factors provides an excellent opportunity to test and revise developmental theories that combine biological and psychosocial data. John Gottman (1994) has pioneered the integration of family process and psychophysiological data. Janice Kielcolt-Glaser has taken the initiative in relating family process and immunological data (Dura & Kielcot-Glaser, 1991). A chapter by Troost and Filsinger (1993) in Sourcebook of Family Theories and Methods outlines a number of theories relating evolutionary processes to family process. This essay centers on genetics as a central link between psychosocial and biological modes of thought about development. Certain properties of genes permit powerful research designs and strong inferences about development. First, we know how individuals acquire their genes. This permits us to say, with security, that every child shares exactly 50% of his or her genes with each biological parent, approximately 50% with each sibling, exactly 100% with its identical twin, and approximately 25% with each half sibling. We also know that genes are finite in number. We can search for single genes that have major effects on behavior, the so-called single gene effects. We can also search for sets of genes that collectively produce certain types of behavioral variation, the polygenic effects. Of particular importance, we can ask of any two behaviors--say parenting behavior in a mother and antisocial behavior in her child--whether they are influenced by the same genes or by different genes. Moreover, the pathways by which genes express themselves in behavior are beginning to be mapped out. In this essay I will emphasize one new idea in this domain: Genes influence behavior by influencing an individual's exposure to certain types of environmental experience. For example, we have preliminary evidence that genes influence the level of irritability and anger that adolescents receive from their parents (O'Connor, Hetherington, Reiss, Plomin, in press; Plomin, Reiss, Hetherington, & Howe, 1994). That is, heritable characteristics of adolescents have systematic effects on the negative parenting that they receive from their mothers and fathers. This genetic influence may play an important role in how genes influence differences among adolescents in levels of self-control, self-esteem, and social responsibility. Taken together, these concepts of genetic endowment, discrete genetic effects, and mechanisms of genetic expression open up a new world of questions and powerful research designs for family scholars. It is also important to set aside, at the outset, certain misconceptions about genes. First, genes do not determine behavior; they are not biological juggernauts. The word influence is much more appropriate, considering the many biological and social events that intervene between the structure of genetic DNA and complex human behaviors. …

A User's Guide to the Gottman-Williams Time Series Analysis: Computer Programs for Social Scientists.

Esther Williams and John Gottman describe a complete set of programs they have written in Fortran IV to enable even beginners to use all the techniques presented in John Gottman's Time-Series Analysis: A Comprehensive Introduction for Social Scientists. There are three packages, available on IBM card source desks from the authors, for (1) time and frequency domain model fitting, including detecting cycles, (2) forecasting and interrupted time-series analysis, and (3) multivariate time-series analysis, in both time and frequency domains. The packages have been tested for portability. Modification specific computers are noted in the guide.