Decoding the Charitable Brain: Empathy, Perspective Taking, and Attention Shifts Differentially Predict Altruistic Giving

Abstract

Altruistic behavior varies considerably across people and decision contexts. The relevant computational and motivational mechanisms that underlie its heterogeneity, however, are poorly understood. Using a charitable giving task together with multivariate decoding techniques, we identified three distinct psychological mechanisms underlying altruistic decision-making (empathy, perspective taking, and attentional reorienting) and linked them to dissociable neural computations. Neural responses in the anterior insula (AI) (but not temporoparietal junction [TPJ]) encoded trial-wise empathy for beneficiaries, whereas the TPJ (but not AI) predicted the degree of perspective taking. Importantly, the relative influence of both socio-cognitive processes differed across individuals: participants whose donation behavior was heavily influenced by affective empathy exhibited higher predictive accuracies for generosity in AI, whereas those who strongly relied on cognitive perspective taking showed improved predictions of generous donations in TPJ. Furthermore, subject-specific contributions of both processes for donations were reflected in participants' empathy and perspective taking responses in a separate fMRI task (EmpaToM), suggesting that process-specific inputs into altruistic choices may reflect participants' general propensity to either empathize or mentalize. Finally, using independent attention task data, we identified shared neural codes for attentional reorienting and generous donations in the posterior superior temporal sulcus, suggesting that domain-general attention shifts also contribute to generous behavior (but not in TPJ or AI). Overall, our findings demonstrate highly specific roles of AI for affective empathy and TPJ for cognitive perspective taking as precursors of prosocial behavior and suggest that these discrete routes of social cognition differentially drive intraindividual and interindividual differences in altruistic behavior.

Significance statement: Human societies depend on the altruistic behavior of their members, but teasing apart its underlying motivations and neural mechanisms poses a serious challenge. Using multivariate decoding techniques, we delineated three distinct processes for altruistic decision-making (affective empathy, cognitive perspective taking, and domain-general attention shifts), linked them to dissociable neural computations, and identified their relative influence across individuals. Distinguishing process-specific computations both behaviorally and neurally is crucial for developing complete theoretical and neuroscientific accounts of altruistic behavior and more effective means of increasing it. Moreover, information on the relative influence of subprocesses across individuals and its link to people's more general propensity to engage empathy or perspective taking can inform training programs to increase prosociality, considering their "fit" with different individuals.

Keywords: fMRI; mentalizing; multivariate pattern analysis (MVPA); prosocial decision-making; social cognition; theory of mind (ToM).

Figure 1.

fMRI donation task and behavior. A, In each trial, participants were presented with a mission statement of a real-life charity (reading phase, up to 30 s). Subsequently, a continuous rating scale (€0 to €50) was presented, and participants were asked to decide how much of their endowment of €50 they were willing to donate (decision phase, up to 10 s). Intertrial intervals varied from 4 to 8 s. B, Histogram of participants' average donations (range of €6.69 to €37.68; mean ± SD: €22.30 ± 9.09), confirming considerable variations in participants' average generosity. C, Taking advantage of the variability of generosity across trials, we defined high and low donations for each participant based on their average donation. Graph represents mean ± SD for individuals' high and low donations used to decode participants' generous and selfish giving decisions from neural response patterns. D, Stimuli were selected based on independent behavioral pretests. Graph represents mean ± SD of donations to each charity for the fMRI sample. Substantial variations in donations to each charity ensured that neural decoding of generosity did not merely reflect properties of individual stimuli.

Figure 2.

Neural decoding of generous donations and of stimulus-driven reorienting of attention. A, A whole-brain searchlight decoding approach was used to identify neural signatures that reliably encoded participants' generous and selfish giving decisions across participants (p < 0.05, FWE-corrected at voxel level, k > 5 voxels). For details, see Table 2. B, Post hoc permutation tests further illustrated that decoding accuracies in predictive brain regions (displayed as triangles), based on response patterns of all voxels in a cluster, are unlikely achieved by chance (all p values <0.007). Boxplots represent null distributions of classification accuracies in the cluster across 1000 permutations. Central marks of boxplots indicate medians, which were found to be 50% (chance level) for all clusters. Edges of boxes indicate the 25th–75th percentiles. Whiskers extend to extreme data points. C, Cluster in the right TPJ/pSTS that encoded (nonsocial) stimulus-driven reorienting of attention (displayed in turquoise) in a whole-brain searchlight analysis (within attention task decoding, p < 0.05, FWE-corrected at voxel level, k > 5 voxels). The predictive cluster overlapped with two clusters in the right TPJ and pSTS that predicted generous and selfish donations in a separate donation task (displayed in orange, within donation task decoding, see A). D, ROI-based cross-task decoding examined whether neural responses that encode attention shifts are reactivated during generous donation decisions. Activation patterns in the right pSTS coding for stimulus-driven attention shifts contained significant information on generous giving decisions (illustrated by triangle, p < 0.05, corrected, permutation test). This finding suggests that the domain-general process of reorienting of attention toward (socially relevant) stimuli may account for part of the predictive information in the pSTS in the donation task. No evidence for shared neural code was found for the TPJ cluster (p > 0.05). Graph represents cross-task decoding accuracies in both clusters (triangles) and for cluster-based null-distributions (1000 permutations). Boxplots represent medians and 25th–75th percentiles. Whiskers indicate ranges of decoding accuracies.

Figure 3.

Neural decoding of empathy and perspective taking for beneficiaries of charities. Left, Data are mean ± SD for participants' (A) empathy and (B) degree of perspective taking across donation trials as reported after scanning, indicating participants' engagement in both processes and comparable degrees of variance. C, During donation decisions, empathy for beneficiaries was encoded in response patterns in the right AI (red), but not in the right TPJ (p < 0.05, FWE-corrected). Pairwise comparisons confirmed that predictive information on empathic responses in AI was significantly higher than those in TPJ (p = 0.002) that varied around chance. D, The degree to which participants represented the perspective of the charities' beneficiaries was encoded in the right TPJ (green), but not the right AI (p < 0.05, FWE-corrected). Information on perspective taking in the TPJ was significantly higher than those decoded from the AI that varied around chance (p = 0.008). Graphs represent median predictive accuracies (white bar) achieved in the right AI and TPJ. Boxes represent the 25th–75th percentiles. Whiskers indicate range of multivariate regression accuracies. *p < 0.01.

Figure 4.

Neural decoding of the relative influence of empathy and perspective taking in donation decisions. A, Neural signatures for perspective taking obtained in an independent fMRI task (EmpaToM) predicted the relative weight of perspective taking as input into donation decisions (β_{perspective taking}) (mean Fisher z-transformed r = 0.34, p < 0.001, Bonferroni corrected), but not of empathy (r = −0.001, p > 0.05, corrected). Graphs represent median predictive accuracies (white bar) achieved in these cross-subject decoding analyses (SVR). Boxes represent the 25th–75th percentiles. B, Independent empathy-related neural signatures (EmpaToM) reliably predicted the degree to which participants relied on affective empathy during donation decisions (β_empathy) (r = 0.16, p < 0.001, corrected), but not contributions of cognitive perspective taking (β_{perspective taking}) that varied around chance (r = −0.04, p > 0.05, corrected). *p < 0.001.