Archive for April, 2011

Meeting 3.6 4/4/11

We began the seminar this week with a general discussion of what religious prosociality actually is, mostly by defining what it isn’t. First, religious prosociality does not consist of a domain-general trait of altruism towards all other individuals. Rather, it is bounded, in that it tends to be specific to other members of the group. By circumscribing the parameters of group membership, prosocial behaviors can be directed specifically towards other members of the in-group, who are in turn more likely to reciprocate.

This definition of bounded prosociality does not necessarily require group level selection. There may be group level processes occurring, but individual level selection for religious prosociality may work in the following way: There exists an evolutionary arms race between free-riders and cooperators, and the features of religious membership may serve to undermine the ability of cheaters to exploit their social environment. This is because religious membership often tends to be costly – it requires an initial or ongoing investment in the community in order to be considered part of the prosocial network.

We referred to the Norenzayan & Shariff (2008) Science review article for evidence for this hypothesis. Several studies showed that individuals act more altruistically towards members of their own group. For example, in an experiment where subjects were given access to a group account, and were allowed to withdraw money (implicitly) at the expense of other group members), less money was withdrawn if the other members of the group were part of the same religious community. As we discussed, there may be group –level selection at work here, but this behavior enhanced not only the overall “fitness” of the group, but the individuals playing the game as well. This is because the experimental manipulation that was used penalized all players for if the sum total of withdrawal was too high.

We discussed the methodology used to test these hypotheses, which included a critique of self-report measures. Specifically, there is a very justifiable concern that there is a great deal of impression management occurring when self-report is used to assess religious prosocial behaviors, due to the very nature of the trait being studied. We also discussed the possibility of impression management behaviors themselves being an interesting variable to compare to various aspects of religious behavior. For instance, would those that are more religious be more likely to report, over-report, or at least bias their interpretation of their own behaviors as being more charitable?

We also briefly discussed the psychological trait of “religiosity”, which is essentially a continuous dimension of personality. Religiosity is content-free, in that it does not specify what religion an individual will adopt, but rather the likelihood of and devotion to religious commitment. This trait is also heritable, as has been demonstrated empirically. More on this in a future meeting.

In discussing some of the results presented in our reading, we were concerned with the taxonomy of prosocial behavior. Is there really a difference between secular and religious prosociality? For example:

These results clearly indicate no significant difference between the two experimental conditions. What mattered in this study was the presence of a moral prime, and not the nature thereof.

However, one aspect that may be unique to religious, as opposed to secular morality and prosociality is the “belief” or “faith” aspect. While bracketing any ontological debate, we discussed the distinction of true vs. feigned altruism. In sum:

In order to successfully function in any society that values prosociality as expressed through reciprocal relationships, one must at least appear to be a cooperator. Anyone who is able to create the impression that they are a trustworthy person, regardless of whether they are being deceptive or not, will be able to at least temporarily exploit the altruistic behaviors of other group members. In a secular society, this might be a strategy individuals are willing to take, if they think they can “get away with it”. However, if there is a fear of divine retribution for cheating from a god or gods, this behavior will be much less likely in religious groups. This requires an “interventionist” god who is interested in the daily lives and behaviors of individuals. The Abrahamic religions meet this criteria, supporting the hypothesis that monotheism might have co-evolved with increasingly larger, more complex, and therefore more anonymous societies, by having a supernatural agent to detect cheating and enforce punishment, either in life or afterlife.

Post contributed by Z.H.

Read Full Post »

Apologies for going out of order, I will get meeting 3.6 posted as soon as it is ready. In the meantime, here is what we were up to yesterday.

No readings, this week. Participants were asked to brainstorm concrete examples of evolution in religion and we discussed four scenarios in class.

The Assignment:
Bearing in mind our 3 cardinal questions:
What is inherited?
What varies?
What is selected?

We’ll want to look at particular cases in the history of religion: please bring one or two scenarios you think would be tractable to evolutionary analysis.

Variable reproductive incentives and the rise of Christianity in the late Roman Empire
Clerical celibacy and declining tribalism in Medieval Christianity
Enforced tithing and ingroup definition in the Latter Day Saints and South Pacific Congregationalists

We’ll want to work out a few concrete examples, come up with discrete theories, and hopefully imagine a research program that could test them.


Class began with a caveat that some traits appear to be universal in religions or in societies in general (e.g., encouraging pro-social behavior). These traits may be heritable, but represent a single historical change, and are therefore not amenable to comparative analysis. We agreed that incest taboos at large were universal, but that the specific nature of those taboos, the degree of consanguinity (blood relation) required, and competing factors were not universal, and might be interesting.
We created a table to summarize our thoughts. It had columns for “What Varies?” “What is Inherited?” and Selection was divided into three subsections – Cultural Bias, Group Level Biological Bias, and Individual Biological Bias; where bias indicated differential survival biased toward presence, absence, or quality of a trait.

The Table:

Common Themes:
Overall, we found that it was extremely difficult to come up with discrete traits, traceable (if not equivalent) across variation, inheritance, and selection. Both conceptual clarity and clear communication proved challenging, though all four traits explored showed some promise.
The units of inheritance tended to be Doctrine and Knowledge/Technology, which seemed to be uncontroversial. Irreversibility and default states had a surprisingly common influence.

Post Contributed by LM

Read Full Post »