doi: 10.1038/nature05510 SUPPLEMENTARY INFORMATION

  • Published 2007


Bayesian design predict that the computational architecture of the kin detection system will weight the coresidence duration cue less (or not at all) when MPA—the higher quality cue—is available to the system. This predicts that outcome measures for the two motivational systems (altruism and sexual aversion) should be high when MPA is present, and that the MPA and coresidence cues will interact. As can be seen from Figure S1 and Tables S1a, b presented below, both are true. The scatterplots shed light on the procedures the brain uses to compute the kinship index, that is, on the kinship estimator. When present, MPA produces elevated levels of all outcome measures (black lines) indicating that, upon detection, the kinship estimator translates this highly reliable cue into a variable whose magnitude represents high relatedness. In contrast, when MPA is absent and coresidence is used as a cue to relatedness, the kinship estimator appears to more slowly ratchet up the magnitude of this variable, generating a gradual increase in altruistic motivations and sexual aversions (red lines). The point at which the lines cross suggest that 1418 years of coresidence duration for MPA absent individuals sets the kinship index to the same magnitude as for individuals with cues to MPA. Table S1b provides descriptive statistics and comparisons for each dependent measure according to whether MPA is absent or present. The predicted interactions exist between MPA and coresidence duration for both motivational systems (altruism and sexual aversion). Interactions for three of the four outcome measures (altruism and the two sexual disgust measures) were significant (Table S1a shows that moral opposition elicited the same pattern of effect sizes as the other three measures, but had >150 fewer subjects than the other measures did; indeed, the interaction effect sizes for moral opposition and altruism are the same). Consistent with these interactions, each of the four outcome measures show significant correlations with coresidence duration when MPA is absent, but no correlation when it is present; for each measure, the drop in effect size was significant (see SI#2). The similar pattern of data across different outcome measures indicates that both altruism and sexual aversion are regulated by a common internal regulatory variable—a kinship index. It also suggests that the kinship index is set on the basis of two distinct cues. While caution is required in interpreting any correlational research, the data show that two of the cues used are duration of coresidence and MPA—or else other cues that are highly correlated with them. SUPPLEMENTARY INFORMATION doi: 10.1038/nature05510

Cite this paper

@inproceedings{2007doi1S, title={doi: 10.1038/nature05510 SUPPLEMENTARY INFORMATION}, author={}, year={2007} }