Brain Activity Dissociates Mentalization from Motivation During an Interpersonal Competitive Game

Abstract

Studies demonstrating selective brain networks subserving motivation and mentalization (i.e. attributing states of mind to others) during social interactions have not investigated their mutual independence. We report the results of two fMRI studies using a competitive game requiring players to use implicit 'on-line' mentalization simultaneously with motivational processes of gains and losses in playing against a human or a computer opponent. We delineate a network, consisting of bilateral temporoparietal junction, temporal pole (TP), medial prefrontal cortex (MPFC) and right fusiform gyrus, which is sensitive to the opponent's response (challenging>not challenging the player) and opponent type (human>computer). This network is similar to a known explicit 'off-line' mentalization circuit, suggesting its additional involvement in implicit 'on-line' mentalization, a process more applicable to real-life social interactions. Importantly, only MPFC and TP were selective to mentalization compared to motivation, highlighting their specific operation in attributing states of mind to others during social interactions.

Fig. 1

Domino game paradigm. Panel a depicts the domino game sequence and corresponding consequences. In the beginning of each game round the player (participant scanned) must decide what chip he/she will play next (decision-making interval) and move the cursor to this chip when instructed (ready interval). The chip can either match the opponent's chip (i.e. have one of the numbers match those on the opponent's chip, 6:3 in this example; upper panel, 6:1) or not (lower panel, 5:2). After placing the selected chip face down next to the opponent's chip, he/she awaits the opponent's response (anticipation of outcome interval). The opponent can either challenge the player's choice (‘show’; red arrows) or not (‘no-show’; pink arrows). Based on the player's choice and the opponent's response there are four possible consequences for each round (response to outcome interval): show match (overt gain); no-show match (relative loss, as the player could have been rewarded if challenged); show non-match (overt loss) and no-show non-match (relative gain, as the player could have been punished if challenged). Note: colors of boxes/wording correspond to panel b color schemes. Panel b outlines the ANOVA design for the response to outcome interval analyses. Red tones correspond to mentalization/opponent's response effect and green to motivation. Panel c describes the suggested mentalization and motivation mechanism of each event. The opponent's chip and samples of matching and non-matching chips are highlighted for demonstration purposes only. In the actual scan the game board and all chips are colored. Also all chips are the same size and color

Fig. 2

Players' choices as a function of time during experiments I and II. Non-match index (number of non-match choices divided by the number of non-match and match choices) for games played against human in experiment I (orange), and against human (light gray) and computer (dark gray) opponent in experiment II are plotted for each minute of the game (averaged for all games for all subjects). No significant differences were found between the two groups (experiment I and II) when playing against human and between human and computer opponents in experiment II

Fig. 3

Main effects of mentalization and motivation (experiment I). Panel a depicts the activation map of random effects ANOVA showing brain regions with a significant main effect of mentalization/opponent's response (n=19, q(FDR)<0.005). These regions included the temporoparietal junction (TPJ), temporal pole (TP), medial prefrontal cortex (MPFC), ventrolateral prefrontal cortex (VLPFC) and fusiform gyrus (FG). Other brain regions presented in this map are midbrain, cuneus, postcentral gyrus and posterior cingulate cortex. Panel b shows the brain networks activated during mentalization (red), motivation (green) and overlapping regions (yellow). As shown in the graphs (panel c), all the mentalization regions had significantly more activation during ‘show’ (red bars) than ‘no-show’ (pink bars) events regardless of losses and gains. Some of the brain regions (TPJ, FG and VLPFC) also demonstrated significant main effect of motivation, such that activations related to gains (dark green) were higher than activation during losses (light green). (Note: since bilateral regions, such as TPJ and TP, showed the same patterns of activations, graphs presented here are averaged percent signal change of right and left activations.) **p<0.001; *p<0.05; L=left; R=right hemisphere

Fig. 4

Mentalization network (experiment II). Activation map of random effects ANOVA showing brain regions with significant main effect of mentalization during the human-opponent runs (n=18, q(FDR)<0.05). Notably, these brain regions are almost identical to the brain regions activated by the same analysis in study I (see Fig. 3a.) Panel b depicts brain regions showing a significant main effect of mentalization during the human opponent runs (n=18, qFDR<0.05) masked with regions showing a significant main effect of opponent type (p<0.05) such that activations during human runs are greater than during computer runs. Note that right VLPFC does not appear in this map. Panel c shows the response to outcome interval activations during the human-opponent games (light gray bars) and the computer-opponent games (dark gray bars). All regions but the VLPFC showed significant main effect of opponent type (i.e. human vs. computer) such that signal was greater during the human-opponent compared to the computer-opponent games. (Note: since bilateral regions, such as TPJ and TP, showed the same patterns of activations, graphs presented here are averaged percent signal change of right and left activations.) TPJ, temporoparietal junction; TP, temporal pole, FG, fusiform gyrus; VLPFC, ventrolateral prefrontal cortex; MPFC, medial prefrontal cortex; **p<0.005; *p<0.05; k>50