Humans Use Forward Thinking to Exert Social Control

Abstract

Social control, the ability to exert influence over others, is critical in interpersonal interactions yet uninvestigated. Here, we used functional neuroimaging and a social exchange paradigm in which people's current choices either did, or did not, influence their partners' proposals in the future. Computational modeling revealed that participants used forward thinking and calculated the downstream effects of their current actions regardless of the controllability of the social environment. Furthermore, greater levels of estimated control correlated with better performance in controllable interactions and less illusory beliefs about control in uncontrollable interactions. Neural instantiation of trial-by-trial values of social controllability were tracked in the ventromedial prefrontal cortex (vmPFC), striatum, and insula for controllable interactions, but only in vmPFC for uncontrollable interactions. These findings demonstrate that humans use forward thinking, a strategy similar to model-based planning, to guide social choices; and that subjective beliefs about social controllability might not be grounded in reality.

Traits for depression related to agentic and external control

“…In this study, scoring higher on a depression inventory was not related to either type of control. Though, we note that contextual factors (Msetfi et al, 2016) and behaviour variation (Blanco et al, 2009; 2011) are known to influence effects mediated by mood and depression, indeed here we demonstrate a differential de-coupling of behaviour and judgements in our two groups(see also , Na, et al, 2019).…”

Navigating uncertainty in human social decision‐making: Consequences and coping strategies

“…Moreover, responders of an iterated UG learn from multiple rounds of interactions that their acceptance or rejection decisions can impact the amount of endowment allocated by proposers. According to this belief, responders plan ahead for future decisions to obtain higher allocations (Na et al., 2019). In an inspection game (Hampton et al., 2008; Hill et al., 2017), both boss and employee update the beliefs regarding the strategy adopted by one another as well as the influence of one's own strategy on the other's behaviors.…”

“…Importantly, however, the effect of impression updating on social behavior could be modulated by the type of partners to interact with, that is, actual (or highly credible) partners or pre‐programmed, unsophisticated algorithms. As mentioned above, responders of an iterated UG can plan ahead for future decisions to get higher allocations by learning from previous interactions (Na et al., 2019). However, since proposers are pre‐programmed to increase their allocations following specific rejection rates without any retaliation or negative reciprocity, responders may benefit from strategic rejection decisions (Na et al., 2019).…”

“…As mentioned above, responders of an iterated UG can plan ahead for future decisions to get higher allocations by learning from previous interactions (Na et al., 2019). However, since proposers are pre‐programmed to increase their allocations following specific rejection rates without any retaliation or negative reciprocity, responders may benefit from strategic rejection decisions (Na et al., 2019). Similarly, trustors' expectations of trustees' reciprocity closely reflect the actual reciprocity of trustees only when interacting with algorithms (e.g., Chang et al., 2011); otherwise, trustors tend to underestimate the actual reciprocity of trustees and pass up profitable opportunities to trust (e.g., Fetchenhauer & Dunning, 2009).…”