Transactional Family Dynamics: A New Framework for Conceptualizing Family Influence Processes

Publisher Summary

In research on family influence processes, there is a growing dissatisfaction with existing models that assume unidirectional pathways and underemphasize dynamic processes. Aside from acknowledging the problem, few systematic proposals have been advanced for more sophisticated ways of thinking about these pathways of influence. Addressing this gap, this chapter proposes transactional family dynamics as a new way of understanding family influence processes. Transactional family dynamics refers to the myriad ways in which family members influence one another, that is, mutual influence processes within families. These processes may include complex patterns of influence in interparental, father–child, mother–child, and sibling relationships.

Introduction

people are not just onlooking hosts of internal mechanisms orchestrated by environmental events. They are agents of experiences rather than simply undergoers of experiences. (Bandura, 2001, p. 4)

In research on family influence processes, there is a growing dissatisfaction with existing models that assume unidirectional pathways and underemphasize dynamic processes. However, aside from acknowledging the problem, few systematic proposals have been advanced for more sophisticated ways of thinking about these pathways of influence. Addressing this gap, we propose transactional family dynamics as a new way of understanding family influence processes. Transactional family dynamics refers to the myriad ways in which family members influence one another, that is, mutual influence processes within families. For example, these processes may include complex patterns of influence in interparental, father–child, mother–child, and sibling relationships. Notably, our interest is in transactional processes—not unidirectional processes—that is, influence processes continuously moving in both directions over time.

Our interest in transactional family dynamics began with a review of the literature on child effects on families (Cummings & Schermerhorn, 2003) and empirical tests of child effects on marital conflict (Schermerhorn, Cummings, & Davies, 2005; Schermerhorn et al., 2007; Schermerhorn, Chow, & Cummings, 2007). We were intrigued to find that children's responses to interparental conflict predicted change in interparental conflict itself—either increases or decreases, depending on the nature of the child's response. Expanding our focus, we also found transactional links between interparental and parent–child relationships (i.e., mother–child and father–child) over time (Schermerhorn, Cummings, & Davies, 2008). We were impressed by the extent of the evidence for the transactional nature of these processes and by the multiple pathways of influence between multiple family members and relationships.

These findings, and emerging results from other laboratories, prompted us to think about the need for a new framework for conceptualizing the multitude of family influence processes. That is, rather than focusing narrowly on just one pathway (e.g., children's influence on marital conflict), we wanted our model to encompass the many pathways, and to integrate emerging empirical work suggesting the importance of multiple pathways of influence. The notion of transactional family dynamics refers to influence processes among multiple family relationships, including the influence of individual family members on family relationships, the influence of family relationships on one another, and family-wide influences. The transactional family dynamics framework also includes the reverse direction of effects. For example, with regard to the influence of individuals on family relationships, one would also be concerned with the influence of family relationships on individual family members. The aim of our approach is to provide a framework for representing these processes across multiple family relationships.

These processes unfold in “real time,” or moment-by-moment interactions, as well as “developmental time,” or long periods of time. Such processes encompass behaviors intended to influence other family members, but also include family members’ unintentional influence on one another. Thus, another key point of our approach is that dynamic processes of influence operate at multiple levels of analysis, including varying lengths of time, or time scales. Some time ago, Thelen and Ulrich (1991) called for investigators to develop dynamic accounts of behavior at many levels of analysis. Consistent with that message, our aim is also to describe and identify the nested, multiply caused phenomenon of family influence.

Although there are increasing calls for broader conceptualizations of families (Cox & Paley, 1997; Jenkins et al., 2005a), increasing the complexity of the study of families presents theoretical and practical challenges that remain to be addressed. Currently, much of the literature on families at least implicitly reflects a narrow conceptualization of families, for example, assessing only one direction of influence, or focusing on only one or a couple of family members. The narrow focus also presents a problem for the clinician, by endorsing therapies that may be ill-suited for real families because of failure to consider important directions of influence. Moreover, a gap in the study of family influence processes is the lack of an overarching theoretical framework to unite and integrate research concerning multiple directions of influence.

Thus, the development of a transactional family dynamics framework was motivated by the urgent need advocated by many in the discipline to move toward models that embrace the complexity of family relationships. In this context, it is important to consider the factors that contribute to the inherent complexity of mutual family influence processes. First, families have a hierarchical organization, with individuals nested within dyads and triads, which are nested within families. Figure 1a depicts this hierarchical organization. Second, families may include multiple family dyads and triads, and therefore, a multitude of influence pathways among them. Third, family influence processes unfold in real time interactions, as well as in the context of processes that may extend over longer periods of time (e.g., the development of emotional bonds or attachments; Bowlby, 1973). That is, time is hierarchically organized, with smaller time scales nested within increasingly long time scales, with varying possible lengths of time for potentially critical changes to occur (Cole & Maxwell, 2003). Fourth, there are different conceptualizations of influence and change in family relationships, including change from one time point to the next and overall patterns of change. Our framework recognizes and attempts to accommodate these complicating factors, for example, by classifying findings by family relationship and by conceptualization of influence and change.

We begin with a brief historical overview of some of the theoretical influences that have laid the foundations for this emerging approach to understanding family influence processes. We then describe transactional family dynamics at a theoretical level, providing a set of organizing concepts and principles. Next, selected research consistent with this framework is reviewed, toward showing how these seemingly disparate directions in research fit together, and underscoring the utility of the transactional family dynamics model. We conclude by highlighting some possible directions and hypotheses for future research. The aim of this chapter is to articulate and advance a model of transactional family dynamics as a framework for conceptualizing and studying family influence processes.

Allen et al. (2006) developed the metaphor of a dance for the concept of family influence processes. Arguing that the critical issue is not one of seeking to understand causality, but rather of seeking to understand the dance itself, Allen and colleagues pointed out that the dance might be lead by more than one family member. Moreover, our ability to distinguish cause and effect is very limited, as even the most sophisticated research designs and data analytic methods cannot prove that our causal models are correct, but rather, can only prove that they are incorrect. Moreover, Allen et al. argued, questions about causality are not even the right questions to be asking. That is, “Observing a dance doesn't tell us who's in the lead, and knowing who's in the lead doesn't tell us who decided what dance to do, or what music to play” (Allen et al., 2006). Instead, we should be asking questions about the dance itself—the complex process of leading and following, giving and taking. We elaborate this metaphor by pointing out that the metaphor is not mechanistic—there is no automatic correspondence between action and reaction (although there is certainly a strong relationship between the two). Thus, this metaphor allows for individuality—it affords space for interpretation of events, personality, and agency. That is, the metaphor allows for interpreting the music through dance.