Unlike previous studies with negative scenarios (negative intention and negative outcome is more salient), here, we introduced a modified third-party social judgment task that involves both good/bad intention and good/bad outcome, and further demonstrated a causal role of the rTPJ in intention-based social judgments under mixed scenarios. We first identified a consistent intention-oriented attribution regardless of the fairness of outcomes in the baseline condition across groups. Moreover, cathodal tDCS over the rTPJ compared to the anodal or sham stimulation diminished the goodness ratings towards good/bad outcomes when the intentions are hyperfair and showed no difference with ratings under bad/unknown intentions, yielding impaired intention-oriented attribution. However, such modulatory effect of tDCS was only existed under a desirable intention condition. Our results provide the first evidence that deactivation in the rTPJ disrupted intention-oriented attribution mainly by lowering the goodness ratings toward outcomes when the intentions are hyperaltruistic, but not by enhancing the permission of good outcomes when the intentions are bad/unknown. Our findings argue for a causal role of the rTPJ in modulating intention attribution that may rely on the contextual salience.
Dissociation of intention and outcome
In our study, no ratings difference between hyper fair and unfair outcome was observed no matter before or after stimulation. The judgment difference we have observed in a baseline condition are mainly driven by the intention-oriented attribution, suggesting the dissociation of intention and outcome processing on the one hand, and the higher weight of intention on social judgments on the other hand. It is possible that participants may show strong demanding social characters, thus leading to higher weight of intention attribution than the outcome. The outcomes (good/bad) here indeed didn’t impact their own compensations given that they are working as a third-party. Meanwhile, the outcome itself didn’t convey moral values or utilities 15,19,22,24. Notably, our results were qualitatively similar to our previously published work based on a similar paradigm, in which a strong intention-oriented attribution were also demonstrated 26. Moreover, a weak effect of outcome attribution was also observed when the initial intention was hyperfair in our previous work 26, but was absent in the current study. Such difference could be attributed to the different trial settings in presenting the final payoffs before 26 or after (current study) initial offers. Though another study from our group have identified similar behavioral effects as the current one after reversing the temporal order of final payoffs and initial offers (unpublished data), the potential influence from temporal sequence cannot be fully ruled out and worthy of further investigation in the future studies.
Comparison with previous studies
First and foremost, we have observed a largely reduced role of the intention on social judgments when inhibiting the rTPJ. Similar impairment on intention attribution was also observed in others’ studies 15,21. However, the disruption of intention attribution in our study was mainly driven by the diminished goodness ratings when the initial intentions are good regardless of the desirability of outcomes, but not by enhancing the permission of good outcomes when the initial intentions are bad/unknown. Indeed, no any modulatory effects of the tDCS on the negative intentions and negative outcomes were observed in our study that were consistently shown in the previous studies 15,19,21. Such discrepancy may come from the contextual difference in delivering the outcomes: in previous studies, the outcomes were usually framed as socially undesirable events, such as death or serious disease 15,19,21, while the outcomes in the present study are framed with both fair and unfair allocations that are linked with individual benefits. In general, the latter one is more socially desirable and positively framed than the prior ones. Previous studies have shown stronger emotional arousal toward the same events when they are framed as negative compared to negative 38. One recent study has specifically indicated the role of the TPJ and the TPJ-mPFC connection in modulating decision preference toward the same offers but were framed differently in a social framing task 39. It is very likely that the impaired intention-oriented attribution shown in our study relies on the positive and desirable context. Moreover, the difference in expectation or top-down attention may also matter across contexts. In the traditional tasks, strong expectation violation may occur when the protagonist’s belief is incongruent with foreshadowing (negative intentions), but not when the dictator’s proposal is unfair. Indeed, fair proposal (positive intentions) could be even more salient or unexpected in our current design, thus leading to stronger modulatory effects.
The second difference from previous studies is that no improvements or even no modulatory effects on the intention attribution were observed when activating the rTPJ via anodal tDCS. The absence of anodal tDCS effects on judgments of either good or bad intention may be due to the ceiling effect 40: for the present task, anodal tDCS may have little benefit on the intention attribution as the rTPJ could have been sufficiently activated among normal people, while cathodal tDCS may show significant inhibition effects. Another reason could be the individual baseline difference of cognitive function 41. There is one study has shown that anodal tDCS over the right posterior parietal cortex (PPC) only improved short-term memory for participants with poor baseline performance 42.
Salient context-based intention modulation
Besides the essential role for intention and outcome processing, the rTPJ was also implicated in the deployment of attention (“attention hypothesis”) to an unexpected event as firstly suggested by Buccino et al, in which non-intended actions showed more activation in the rTPJ compared to intended actions but not the reverse (“intention hypothesis”) 43, and then followed by others 44–46. Based on our findings, though rTPJ plays a similarly disruptive role in modulating social judgments under scenarios with good intentions compared to bad intentions, such effects cannot be merely attributed to attentional mechanism or expectancy violation. The reason is an equally modulatory effect of the cathodal tDCS on the bad outcome was found as the good outcome under the good intention though the former one is more unexpected. When we bring previous studies and our current findings together, it is obvious that the cathodal tDCS over rTPJ consistently disrupted the intention-based social judgments not only by enhancing the permission of neutral outcomes when the intentions are bad 15,21, but also by decreasing the positive evaluations of the good/bad outcomes when the intentions are good. Such results may argue for a causal role of the rTPJ in modulating intentional attribution that is also closely relying on the salient contexts.
Significance of our findings
The current results may shed light on the understanding of deficits in mindreading among neurodevelopmental disorders, such as ASD, and other populations with psychiatry disorders. Previous studies have indicated common deficits in intention attribution among ASD and terrorists, however, the ASD tend to judge people as more trustworthy than the normal people 9,10,47, while the terrorists tend to overly rely on the outcome regardless of the intentions 8. Our findings may reconcile the conflicts, in which the deficits in intention attribution among ASD may be related to enhanced permission of neutral or bad outcomes under bad intentions, while the deficits in intention attribution among the terrorists could be related to the devaluations of hyper fair and unfair outcomes under good intentions. This is same to the people with psychiatric disorders, such as schizophrenia, who may have difficulty in understanding others’ either good or good intentions and show biased social judgments 48,49.
Limitations and future directions
Our study has some limitations. Firstly, it is still unclear whether the effects we observed in the current experiment is specific to an economic domain with mixed scenarios or can be generalized to other non-economic domains. Future work is necessary to address the behavioral similarity and difference between the traditional story-based moral judgment task and the current task within a same person and under a same scenario. Secondly, it is possible that other brain regions apart from the rTPJ may be stimulated by tDCS and underlay the results given that we did not include a control stimulation site. Future studies may include a control stimulation site and further confirm the uniqueness of stimulating rTPJ in modulating intention-based outcome evaluation. Lastly, the absence of anodal tDCS effects and tDCS effects on judging negative intentions could be context-specific and need to be treated with caution as discussed above. Future studies using different tDCS protocol parameters and social judgment paradigm are needed to replicate the present findings and identify possible boundary conditions.