Whenever people interact with each other in social settings, they seem to spontaneously synchronize their body movements in time, space and form [2, 3]. This phenomenon, commonly referred to as interpersonal synchrony, is defined as the “temporal coordination of micro-level social behavior” [8] and emerges in daily-life joint actions such as parent-infant [2, 7, 8], teacher-student [2] as well as therapist-client interactions [17]. Research investigating the effects of synchrony suggest that a period of synchronized activity is related to an increase in various categories of interpersonal affect such as, for example, trust [11].
This finding could, in turn, explain synchrony’s large positive effect on interpersonal relationships [7, 8, 9]. For example, research suggests that the therapeutic relationship between a client and his/her therapist is a more predictive of treatment success than the used therapy interventions [10]. This therapeutic relationship is, to a significant degree, established through interpersonal synchrony of body movements, speech and emotions [9]. In the light of above-mentioned findings, this effect of synchrony could, thus, be mediated by an increase in trust. However, until now, the dynamics of trust and synchrony have not been systematically investigated in detail, possibly due to a difficulty in defining and modelling trust [4]. Here, the aim of this paper is to investigate the adaptive processes between synchrony and trust using a computational model. To do this, a temporal-causal network approach will be adopted [20, 19] to model the cognitive states of trust [6] and synchrony [5] as well as their interactive dynamics.
The remainder of the paper is organized as follows. Section 2 of the paper discusses background knowledge, sections 3 and 4 present a conceptual and numerical representation of the model, respectively. Section 5 gives its experimental evaluation and section 6 describes the mathematical analysis of the model. Section 7 concludes the paper.
Recently, synchrony has received attention as a form of non-verbal communication that can affect the affiliative relationships of the conversational partners. For example, in their seminal study,
1
Wiltermuth and Heath [21] investigated the potential benefits of synchrony-inducing group activities (e.g. singing, marching, dancing) for psychological boundaries. They found that synchronous behaviors led to more cooperative behavior in subsequent group economic exercises, possibly by strengthening social attachment [21] and increasing trust among group members [12].
While these findings suggest a positive relationship between interpersonal synchrony and trust, literature about the underlying mechanisms of these dynamics remain unclear. This might be partly due to a lack of a coherent definition of trust within the literature as [4] identified 72 different definitions of trust, depending on factors like the research field or context. Thus, in order to systematically investigate the dynamics between synchrony and trust, it is imperative to first establish and use a common definition and framework of trust. In this paper, we will adopt the socio-cognitive theory of trust, developed by [4] because this framework acknowledges the complexity of trust by introducing cognitive processes on several levels and layers of the concept. Combined with an overarching, context independent definition of trust, this framework creates the opportunity to investigate trust from various research areas (e.g. psychology, sociology, economics, computer science). In the following, we will selectively describe relevant parts of the theory and point out their relation to our study. For a comprehensive review of the theory, see [4].
According to this socio-cognitive theory, trust can be defined as “a mental state, a complex attitude of an agent x towards another agent y about the behavior or action alpha relevant for the result (goal) g” [6]. Furthermore, they posit that trust can be divided into two subcategories: on the one hand, trust in someone or something that has to act and produce a performance and, on the other hand, the global trust in the occurrence of a global event or process which is affected by external factors. In the present study, we will focus on the former type of trust, which is composed of the basic beliefs and expectations of agent x towards agent y with regard to a goal g. More specifically, [6] describe the basic beliefs as “a positive evaluation of y that is necessary, x should believe that y is useful for this goal of its, that y can produce/provide the expected result, that y can play such a role in x’s plan/action, that y has some function” [6]. In addition, expectations are described as follows: “A positive expectation is the combination of a goal and of a belief about the future. X both believes that y can and will do g; and x desires/wants that y can and will do g” [6].
Furthermore, the theory describes two types of “delegation”, which is the plan of agent x to achieve the goal through agent y, thus, constructing a “multi-agent plan and [agent] y has a share in this plan: y’s delegated task is either a state-goal or action-goal” [6]. While strong delegation is based on y’s awareness of x’s intention to exploit her action, in weak delegation y is generally not aware of the fact that x is exploiting her action [6]. For our example scenario’s, we will focus on weak delegation since it does not presuppose any explicit agreement between the agents: agent x relies on y to reciprocate her body movements and predicts that she will do it in order to reach the goal of establishing synchrony. This reliance, in turn, is the result of the degree of positive beliefs and positive expectations described above.
Using the socio-cognitive theory of trust as a framework, the cognitive states of trust can be expressed and modelled on various levels. While empirical research conducted studies to investigate the isolated effects of synchrony on trust [11, 12] and vice versa [15], this paper will be the first study of our knowledge to computationally model the adaptive processes occurring between synchrony and trust. To do this, we will use the network-oriented modelling approach, which is a useful tool to understand the structure and dynamics of adaptive cognitive processes [20, 19].