Elucidating the role of the posterior medial frontal cortex in social conflict processing

Abstract

A fundamental function of the brain is learning via new information. Studies investigating the 22 neural basis of information-based learning processes indicate an important role played by the 23 posterior medial frontal cortex (pMFC) in representing conflict between an individual's 24 expectation and new information. However, specific function of the pMFC in this process 25 remains relatively indistinct. Particularly, it's unclear whether the pMFC plays a role in the 26 detection of conflict of incoming information, or the update of their belief after new 27 information is provided. In an fMRI scanner, twenty-eight Japanese students viewed 28 scenarios depicting various pro-social/anti-social behaviors. Participants rated how likely 29 Japanese and South Korean students would perform each behavior, followed by feedback of 30 the actual likelihood. They were then asked to rerate the scenarios after the fMRI session. 31 Participants updated their second estimates based on feedback, with estimate changes more 32 pronounced for favorable feedback (e.g., higher likelihood of pro-social behavior than 33 expected) despite nationality, indicating participants were willing to view other people 34 favorably. The fMRI results demonstrated activity in a part of the pMFC, the dorsomedial 35 prefrontal cortex (dmPFC), was correlated with social conflict (difference between 36 participant's estimate and actual likelihood), but not the corresponding belief update. 37 Importantly, activity in a different part within the dmPFC was more sensitive to unfavorable 38 trials compared to favorable trials. These results indicate sensitivity in the pMFC (at least 39 within the dmPFC) relates to conflict between desirable outcomes versus reality, as opposed 40 to the associated update of belief. 41 42

Testing the reinforcement learning hypothesis of social conformity

“…Consistent with the previous works, we found a significant conformity effect; the second ratings of the participants were significantly influenced by the group rating even after the regression-to-the-mean effect was controlled (t(24) = 2.18, p = 0.02, d = 0.43). We also found highly significant regression-to-the-mean effect (t(24) = -15.81, p < 0.001, d = 3.16), which is consistent with our previous studies [4,5]. During the reinforcement learning task, participants selected options with a higher reward probability 59.9 % of the trials, which is significantly higher than the chance (50%; t(24) = 6.22, p < 0.001, d = 1.24), indicating that the participants were generally able to accurately keep track of the fluctuating reward probabilities based on reward outcomes.…”

“…The dependent variable was rating change (participant's second rating -first rating). The first rating was considered as one of the predictor variables to control for the regression-to-the-mean effect [4,5]. On the rare occasion (0.32% across all participants), seven participants pressed the decide button (right thump button) before the right or left key (i.e., they submitted their rating, most likely accidentally, when the cursor was still invisible).…”

“…In a study on social conformity that used transcranial magnetic stimulation, it was shown that the pMFC plays a causal role in preference change [3]. The original fMRI findings have been replicated by several studies (e.g., [4][5][6]). A recent meta-analysis revealed that the insula and pMFC are consistently positively related to social conflict (i.e., the difference between individual and group opinions) and that the ventral striatum is negatively related to social conflict [7].…”

Testing the reinforcement learning hypothesis of social conformity

“…In a study on social conformity that used transcranial magnetic stimulation, it was shown that the pMFC plays a causal role in preference change (Klucharev, Munneke, Smidts, & Fernandez, 2011). Using the similar experimental paradigm, several studies have replicated the original fMRI findings (e.g., Campbell‐Meiklejohn, Bach, Roepstorff, Dolan, & Frith, 2010; Izuma & Adolphs, 2013; Korn et al, 2014; Korn, Prehn, Park, Walter, & Heekeren, 2012; Wake, Aoki, Nakahara, & Izuma, 2019), and a recent meta‐analysis revealed that the insula and pMFC are consistently positively related to social conflict (i.e., the difference between individual and group opinions) and that the ventral striatum is negatively related to social conflict (Wu, Luo, & Feng, 2016).…”

“…The dependent variable was rating change (participant's second rating ‐ first rating). The first rating was considered as one of the predictor variables to control for the regression‐to‐the‐mean effect (Izuma & Adolphs, 2013; Wake et al, 2019). On the rare occasion (0.32% across all participants), seven participants pressed the decide button (right thump button) before the right or left key (i.e., they submitted their rating, most likely accidentally, when the cursor was still invisible).…”