Moreover, the dynamic and fixed attitude model differed in the partner-specificity of some behavioural patterns. In the fixed attitude model, affiliative partner choice is purely based on rank-distance (via fixed LIKE attitudes that are based on rank distance) and as a result behavioural rates within the dyads of the same rank-distance showed low variation. In contrast, in the dynamic attitude model affiliative partner choice is dependent on the specific identity of actors of earlier received affiliation. As a result, affiliative relationships were highly individualized and behavioural rates within dyads of the same rank-distance showed high variation. To quantify this, we calculated the standard deviation (SD) over the behavioural rates for all dyads of a certain rank-distance and then averaged the SD over all rank-distances. Second, we describe the fixed attitude model (symmetry-based model), in which LIKE attitudes are also used to choose affiliation partners, but where LIKE attitudes are symmetric and only dependent on rank-distance and thus are fixed.

More specifically, receiving and executing affiliative behaviour increases satisfaction levels and decreases arousal and anxiety. While receiving submissive behaviours results in decreased arousal and anxiety, receiving aggressive behaviour or perceiving aggression nearby results in increased levels of anxiety and arousal. Executing aggressive behaviour increases ego’s own arousal, while ego’s own anxiety level decreases or increases depending on the outcome of the conflict. Besides studying the time course of affiliative behaviour in specific dyads in a group, another suggestion for future research concerns the role of partner choice.

Dynamic Attitude Model

The EMO-model assumes that anxiety and satisfaction are two important emotional dimensions regulating different types of behaviour. A high arousal state increases the probability to perform some active behaviour and also increases the probability to scan the environment for group members. At increased LPS, the Shannon index of the LIKE attitudes, the proximity scores and affiliative behavioural frequencies were decreased (see black box-plots in Figure F2 in S1 Supplementary Material). Hence, at increased LPS, individuals restricted their affiliative behaviours, their proximity and high LIKE attitudes to fewer group members than at lower LPS.

At increased LPS, rates of leaving, i.e. moving away from someone in close proximity, were generally decreased, simply because individuals spent less time in close proximity of distant-ranking individuals. Taking the emotional experiences from past interactions with specific partners into account over a longer time span is called emotional bookkeeping [19,21,47]. Here, individuals assign and update long-lasting emotional attitudes towards specific partners based on earlier interactions, without remembering these Running Law Firm Bookkeeping: Consider the Industry Specifics in the Detailed Guide events specifically. The effects of these long-lasting emotional attitudes are explored in the current paper. The effect of the time span over which emotions elicited by earlier interactions are integrated into these attitudes is explored in another study using the EMO-model [48]. In a modified version of the GrooFiWorld model, FriendsWorld [89], group members with whom an individual engages most in grooming (top quartile) have been defined as “friends” and model individuals preferably approach such “friends”.

4. trial balance

Individuals are characterized by their dominance strength (myDOM), which does not change over time or after interactions [43,44,56,68]. Individuals differ in their schedule time (myTIME), (current) scanning probability (myPscan) and in the current width of the view angle (myVIEW_ANGLE), which change dynamically over the course of the simulation. The individuals are characterized by a number of state variables (Table T1 in S2 Supplementary Material and see below).

In the fixed attitude model, there is no feedback from affiliative behaviour onto LIKE attitudes. Therefore, the fixed attitude model serves as a control model to assess the importance of this dynamic feedback from behaviour to emotional attitudes for the emergence of certain group patterns. The purpose of the model described in this paper was to explore certain capacities of information integration that may be used in primates to develop and maintain social relationships, and their effect on the emergent properties of affiliative relationships. While the model was intended (and parameterized) to specifically investigate social behavior of macaques, we believe that general model processes (such as the emotional regulation of behavior and emotional bookkeeping) may be very similar across many primate species. The resulting group level patterns may also apply to other New World or Old World monkeys and apes, and possibly even to prosimians. We studied affiliative partner selectivity, i.e. the degree of preferring certain individuals as affiliation partners.