The role of right temporoparietal junction in processing social prediction error across relationship contexts

Abstract

<jats:title>Abstract</jats:title> <jats:p>How do people update their impressions of close others? Although people may be motivated to maintain their positive impressions, they may also update their impressions when their expectations are violated (i.e. prediction error). Combining neuroimaging and computational modeling, we test the hypothesis that brain regions associated with theory of mind, especially right temporoparietal junction (rTPJ), underpin both motivated impression maintenance and impression updating evoked by prediction error. Participants had money either given to or taken away from them by a friend or a stranger and were then asked to rate each partner on trustworthiness and closeness across trials. Overall, participants engaged in less impression updating for friends vs strangers. Decreased rTPJ activity in response to a friend’s negative behavior (taking money) was associated with reduced negative updating and increased positive ratings of the friend. However, to the extent that participants did update their impressions (more negative ratings) of friends, this behavioral pattern was explained by greater prediction error and greater rTPJ activity. These findings suggest that rTPJ recruitment represents the integration of prediction error signals and the capacity to overcome people’s motivation to maintain positive impressions of friends in the face of conflicting evidence.</jats:p>

The representation of mental state information in schizophrenia and first-degree relatives: a multivariate pattern analysis of fMRI data

“…, 2012 ; Koster-Hale and Saxe, 2013 ; Tamir and Thornton, 2018 ; Thornton et al. , 2019 ; Park et al. , 2020 ; Richardson and Saxe, 2020 ).…”

Neurocomputational mechanisms engaged in moral choices and moral learning

“…There are similarities and differences between these different forms of learning processes. All of them can be formalized by different types of PE: norm PE (Xiang et al, 2013), PE about harmful outcomes occurring to others Nostro et al, 2020), and PE by observation of other's moral actions (Bellucci et al, 2019;Park et al, 2020). When learning a new set of moral rules and when learning the moral character of strangers, what is learned is moral information.…”

“…RL models help to explain how a history of pairing social phenomena with positive or negative outcomes can influence and bias complex moral behaviors (Buckholtz, 2015;Gȩsiarz and Crockett, 2015;Christopoulos et al, 2017;FeldmanHall et al, 2018a,b;FeldmanHall and Dunsmoor, 2019). For example, RL mechanisms describe learning about others' moral values based on their preference to punish fairness violation (FeldmanHall et al, 2018a,b), or learning others' moral traits, such as generosity (Hackel et al, 2020), honesty (Bellucci et al, 2019) and trustworthiness (Fouragnan et al, 2013;Park et al, 2020) as well as learning moral norms (Xiang et al, 2013;Gu et al, 2015;Hétu et al, 2017). More generally, learning procedures described by Pavlovian and instrumental conditioning provide valuable frameworks for understanding learning in moral contexts, and account for how histories of past decisions influence future moral choice (Gȩsiarz and Crockett, 2015).…”

Neural support for contributions of utility and narrative processing of evidence in juror decision making

“…However, similar to past findings on difficult moral judgment ( 39 ), we observed brain activations primarily in the posterior aspects of the social-cognition network, not typically associated with social value. Instead, these posterior regions of the social cognition network are hypothesized to play a role in prediction ( 40–42 ). This role would be consistent with participants considering the consequences of future punishment and wrongful conviction.…”