Author Reply: Coregulation is a State of a Temporal Interpersonal Emotion System

As Troxel (2013) elegantly puts it, coregulation has emerged as an important research topic because it takes us beyond the traditional focus on the individual and considers “the inter-dependency in attachment relationships, itself, as a potential pathway explaining the regulatory, and ultimately health-promoting properties of close relationships” (Troxel, 2013, p. 211). Troxel also rightly points to the central challenge of distinguishing coregulation from related constructs. We propose that progress can be made by better characterizing what “interdependence” means. In other work we have argued that interdependence can be defined by the presence of a temporal interpersonal emotion system (TIES; Butler, 2011). Specifically, emotion is often framed as an intrapersonal system comprised of subcomponents such as experience, behavior, and physiology that interact over time to give rise to emotional states. Whenever emotions occur in the context of social interaction or ongoing relationships, however, a tem-poral interpersonal emotion system (TIES) comes into existence, in which the subcomponents of emotion interact not only within the individual but across the partners as well. Once this occurs, emotional interdependence has been established, and understanding the outcomes that emerge demands a know-ledge of the dynamic principles of TIES.

We propose that coregulation is one potential state of an interpersonal emotion system defined by coupled dampened oscillations between partners’ emotions that converge on a stable level. This implies a precise statistical model that can be distinguished from other, similar possibilities. Stress buffering, for example, would differ in that the interaction partners would not show any statistical dependence (coupling) in their emotional responses. Stress buffering could still be seen as a form of interdependence, in that the emotional response of the target person could not be adequately represented without including the support-provider’s state as a sort of attractor, drawing the stressed individual back towards baseline, but it would be a form of interdependence that does not include emotional coupling. Thus, ultimately, we believe that distinguishing between coregulation and other important forms of interdependence rests on improving and clarifying the statistical models used to investigate TIES (see Butler, 2011, for a review of existing models). Similarly, we believe many of the other questions raised by Troxel (2013) could be at least partially tackled with more sophisticated modeling. For example, individual differences could be represented as initial conditions in a dynamic system, and the set points around which coregulation takes place could be operationalized as dyadic attractors.

Nevertheless, our models can only be as sophisticated as our data. As Troxel (2013) notes, distinguishing between various forms of interdependence requires, at a minimum, assessing some aspect of emotion over time in both partners. In addition, coregulation should only become apparent when one or both partners have been disturbed away from their emotional set point. Data that meets these requirements are rare, but do exist. We encourage researchers who have such data (i.e., emotion measured repeatedly from both partners following a stressor) to make use of the coupled oscillator model described by Boker, Bunter, and their colleagues (Boker & Laurenceau, 2006; Butner, Diamond, & Hicks, 2007) to investigate if and when a coregulatory pattern emerges.

For new data collection, Troxel (2013) suggests an elegant approach adapted from research on circadian rhythms that involves removing exogenous cues in order to “unmask” a free-running endogenous system. In the context of coregulation, the attachment figure represents the exogenous cues, and when he or she is removed—if coregulation is in fact an important aspect of dyadic emotion regulation—we should see changes in the functioning of the endogenous emotion systems of both partners, attributable to the removal of the exogenous regulation provided by each other. To test this we imagine a sort of “strange situation” for adults in which securely attached couples must deal with a stressor together, both when paired with someone else’s partner (i.e., a stranger) and alone. Continuously assessed emotion in both partners could be used to model coregulatory, stress-buffering, and emotional escalation states in the dyadic contexts. In addition, intrapersonal emotion dynamics (e.g., negative escalation, flexibility) could be modeled across conditions, providing a comparison of emotion regulatory dynamics when alone, with a secure partner, or with a stranger.

Finally, one fundamental question that needs to be answered is, “What is the time frame for coregulation?” If coregulation is essentially an emotional process, then we should probably be looking for coupled dampened oscillations in partners’ emotions over approximately 5–20 minutes following a stressor, since that is roughly the time frame for emotional episodes. This speculation demands empirical validation, however. In addition, the issue of time frame is relevant to distinguishing between coregulation and entrainment of diurnal rhythms due to shared routines and habits. We hypothesize that coregulation may exist as an overlay on entrained rhythms. Specifically, coregulation may take the form of dampened oscillations in the high-frequency range (fluctuations occurring over minutes and hours), while entrained rhythms may occur in the form of slower oscillations (fluctuations occurring over days or weeks). If this were the case, both entrained rhythms and coregulation may contribute to emotional and physiological homeostasis in close relationships, but they could be distinguished based upon time frame. This would allow investigation of both the shared and unique influences of coregulation and entrained rhythms on emotional and physical well-being.