Could Empathy Encourage Hyperactive Authority Detection?

Misplaced Faith? an interview with Professor Luke Galen, inspires some interesting speculations that I’ll offer in response.  First, the mention of gender differences in the context of individual differences in the “sensus divinitatis” and agency detection is potentially important.  Several factors that may help explain gender differences in religiousness are currently under investigation (e.g., discrimination in patriarchal societies).  My first idea builds on the gender difference in empathy.  The androgen testosterone inhibits empathy, and feminine roles and norms emphasize empathy, so I wonder if the resultant gender difference could explain part of the difference that makes women more religious than men on average.

This builds on the hyperactive agency detection theory that came up in this interview.  This theory argues that people apply the theory of mind – the attribution of independent minds and motives to other living beings – in an overzealous way, sometimes mischaracterizing inanimate objects as living beings and attributing actions and desires to them.  I usually think of hyperactive agency detection as the self-protective assumption that a predator is the source of every bump in the night, at least until proven otherwise.  The striving for control over otherwise-uncontrollable events like natural disasters, incurable diseases, and difficult or complex challenges (e.g., war, gambling) may also lead to superstitious attempts to manipulate supernatural forces like luck or appeal to supernatural powers that might tip the outcome in one’s favor if appeased.  Even birds have been said to act superstitious under variable interval reinforcement schedules.

Yet other mechanisms may partially mediate effects of hyperactive agency detection on supernatural beliefs, such as the (innate?) motive to align with and internalize the goals of one’s (para-)social superiors.  As with emotion detection, more empathic or emotionally intelligent individuals may perceive the ingroup’s motives more easily and automatically, possibly even to the point of overdetecting more motives or motivational complexity than the group actually possesses.  In this general scenario, a highly empathic person wouldn’t have to overdetect the presence of agents, but might overestimate their activity or desire to act.  This hypothesis could begin to explain differences between deists (who believe in passive creator gods) and believers in gods that intervene actively.

As social animals, humans not only try to understand the motives and emotions of others, but also learn from them.  People construct sets of behaviors and infer laws through social learning from other agents, particularly ingroup members of higher authority like parents and presidents.  If the agents are absent or illusory, those behaviors and laws could take on a proto-religious significance for lack of direct feedback or operant shaping, especially if observant behaviors are still reinforced less directly by satisfaction of needs for control or affiliation, or by the approval of lesser superiors like older siblings or local authorities.

If greater empathy or emotional intelligence give individuals more opportunities to learn (and overlearn) from each other, then perhaps entire cultures could be shaped to serve society’s best guesses about what its highest authority figure desires.  To apply the prospect of overlearning via empathy to theistic cognition in general, I see god(s) in this role of highest ingroup authority.  If gods serve to explain uncontrollable events, it is probably natural to assume they are more knowledgeable and powerful than us.  Other characteristics like authoritarian submission or just world belief could incline people to identify with such a knowledgeable power – essentially applying theory of mind to the natural universe.  If by identifying with such an ultimate power, one could convince others of some borrowed knowledge or power, then one could attract subordinates, creating a social hierarchy while posing as a mere middleman.

Social networks have mediated the parasocial power of distant, famous leaders throughout human history, which includes many cases of deified monarchs.  Maybe this is not too short a span to have evolutionary implications, such that real lords have sensitized us (culturally, if not biologically) to their ability to project power and command servitude without personal presence.  Regardless of the origins of our personalities and social mechanisms, authority structures in modern society may take advantage of empathy, authoritarianism, and just world belief.  If so, religious authority structures are surely no exception.  Empathy might mediate reciprocal effects between gender and religiousness, and for that matter, authoritarian submission could mediate bonds between religiousness and conservatism, which relates to authoritarianism and just world belief.

These mechanisms may explain benevolent and parental aspects of traditional god images.  Since parents are the original objects of social learning and authoritarian submission, parental schemata might accommodate god(s) by emphasizing power and absence.  Other characteristics might carry over from parental schemata, such as nurturant intent and various anthropomorphisms.  The moniker “our heavenly father” makes this blending of schemata explicit, as does belief that god(s) could beget human children (e.g., Jesus, ancestors of monarchs).  Furthermore, referring to god(s) as “lord” may reinforce the general anthropomorphization of god(s) and the believability of monarchs’ hereditary divinity claims.  It would be difficult to discern whether the popularization of religion has anthropomorphized god images over time or whether god images began as anthropomorphic and helped to popularize religion.  Again, regardless of origins, reciprocal causation is plausible in modernity.

In light of the many predispositions that might influence perceptions of god(s), it seems strange to argue, as some cognitive scientists of religion (CSR) have, that a sensus divinitatis could offer any positive information about the existence of god(s).  Did a god create us, and especially our social superiors, in “his” image, along with the impulses to detect him?  Or did people create god(s) in the images of themselves, their parents, and their ingroup leaders, using no more than the same cognitive architecture that has served many essential, secular purposes throughout humanity’s existence?

As I’m sure has been said before, identifying the cognitive architecture that explains perceptions of birds doesn’t independently provide any more proof than disproof of birds’ existence.  Supportive evidence is needed, and gods would require a lot more scaffolding to support.  Granted, prominent CSR theories may refer to a more vague higher power than most religions – this resembles ietsism (Dutch for “somethingism”) more than it resembles the Abrahamic faiths that contextualize most of CSR – but frankly, I suspect this is a compromise with the heterogeneity of god representations across world religions, which is difficult for even Eurocentrists to overlook.  If it is important to also acknowledge the possibility of independent evidence that would support the existence of god(s), or to acknowledge that all empirical evidence as we know it is mediated by human perception, then arguably it is important to revisit epistemology in general.  In my opinion, this overextends the scope of manageably focused, productive conversation, as does bringing the matter of god’s actual non/existence into the discussion of any psychological predisposition toward theism.  These related topics all warrant independent study.

That being said, another point raised in the podcast deserves at least tangential recognition.  The involvement of the temporal lobe in religious/spiritual cognition is fascinating, because the temporal lobe plays integrative roles for the semantic interpretation of sights, words, and feelings.  This overlap of structures and functions would seem to suit the confabulation of complex constructs in general.  In particular, the temporoparietal junction (TPJ as Galen refers to it) serves several functions relevant to theism, including moral judgment, agency detection, differentiation of others from the self, and out-of-body experiences (the last probably being more a matter of dysfunction).  While hardly disproving the existence of god(s), the capacity of these brain structures to intuitively overdetect moral authorities that transcend the self and corporeality itself certainly calls these popular god concepts into question.  In other words, regardless of the non/existence of god(s), it seems quite plausible that we have made several assumptions about the nature of god(s) that suit our neurological inclinations.

Finally, and further beyond the scope of evidence, I had a small difference of interpretation regarding artificially induced involuntary movement and the overattribution of control to oneself.  Without knowing the data independently of Galen’s description, I can only say that his description seems to suit the counterargument: that people have some capacity to underdetect external agents, even as they might override routine self-control.  If personal control is antithetical to divine control, it seems worth acknowledging in fairness if these experiments would suggest that the illusion of self-control could obscure external control.  Yet by the same token, one might need to reconsider infamous legal defenses like “the devil made me do it”.  The overattribution of self-control would also support biological reductionist arguments against free will, which is another important component of many popular religious belief systems.  It could be quite a conundrum if the same brain mechanism held contradictory implications for the potential validity of a religion (i.e., if the sensus divinitatis exists, but free will doesn’t, or vice versa, given religious claims that both exist), but this would be a theological conundrum, and probably not much of a problem for psychology.

Reliability and Religion: A response to Misplaced Faith?

Professor Galen’s podcast is refreshing in many ways. Claiming that recent scientific theories about religious belief are neutral has been the orthodox position in both philosophy and science for some time now. Galen questions the validity of this position. I will first formulate what I take Galen’s position to be and then offer some critical remarks.

I take Galen to argue that religious belief is unreliably formed. His point mirrors claims made by philosophers of religion who claim that recent theories of religious belief support the claim that these beliefs are reliably formed. The meaning of unreliability is widely discussed in philosophy. It is usually predicated of belief-forming mechanisms and means that the mechanism produces many false beliefs. Because they produce many false beliefs, beliefs produced by unreliable mechanisms cannot be considered rational.

In the discussion over recent scientific theories about religious belief, unreliability claims see the (potential) unreliability of religious mechanisms as following from their evolutionary history. The mechanisms at the root of religious beliefs are claimed to be the way they are as a result of evolutionary processes like natural selection. Since natural selection selects for traits that increase fitness and cares little about truth, the mechanisms are unlikely to be aimed at truth. Some authors have concluded to unreliability in this way.[i]

Galen’s argument for unreliability is different. He claims that there is independent evidence that the mechanisms at the root of religious belief produce many false beliefs. The evidence he offers is threefold.

  • First cognitive science shows that there are important individual differences in religious mechanisms. It is well established that, on average, women are more religious than men. People with a more analytic style of thinking also seem to be less likely to form religious beliefs. Tanya Luhrmann’s absorption theory, finally, states that some people are more gifted to form religious beliefs.
  • Second some mechanisms that contribute to forming religious beliefs have been connected with the production of false beliefs. Luhrmann’s absorption was connected to falsely detecting agency. Higher religiosity was correlated with blending of ontological categories.
  • Finally, (and this is the strongest evidence according to Galen) religious beliefs are malleable. When people encounter counterevidence for their beliefs they are inclined to double down on those beliefs rather than revising them. People suffering from social deficits are also found to be more likely to have religious beliefs. furthermore, manipulations of the brain can give rise to misattributions of agency; for example to attribution to a supernatural agent.

Galen claims scientific theories can lead to metaphysical conclusions. I think his argument is better understood as epistemological. From a claim of unreliability no strong metaphysical claims about what exists can be drawn. When a religious belief is produced by an unreliable mechanism, the object of that belief (in this case God or another supernatural being) can still exist. I believe Galen’s argument can be reformulated and summarized as follows:

  • There is strong evidence that religious beliefs are unreliably produced.
  • Beliefs that are unreliably produced are not rational.
  • Therefore, religious beliefs are not rational.

The argument is logically valid as the conclusion follows from the premises. The second premise is widely assumed in recent epistemology but is not obvious. A minority position states that rationality of beliefs is a function of how well they cohere with other beliefs. In this case religious beliefs can be considered rational because they cohere well with beliefs about order in the universe and beliefs about a supernatural origin of morality. Another minority position is pragmatic and states that beliefs are rational if they work, that is if they allow an individual to better make sense of her environment. In both cases, whether the belief in question is unreliably produced or not is (largely) irrelevant.

Denying premise 2 may seem implausible but in many cases we do tend to deny it. In the podcast the interviewer noted that many political beliefs are formed by similar unreliable mechanisms as religious beliefs are. These beliefs are sometimes claimed to be rational in virtue of their coherence with other beliefs or in virtue of their pragmatic use.

Nonetheless, the second premise is widely accepted. Therefore the crux of Galen’s argument is in premise 1. Galen’s first group of evidence is not really evidence for unreliability. Contrary to what Galen claims, I believe the strongest evidence for the premise are the second group of reasons. Mechanisms that are like other known unreliable mechanisms or that have been connected to the production of false beliefs are likely unreliable. The third group of reasons say more about what people do with their beliefs  than how they form their beliefs. Doubling down on religious beliefs when confronted with counterevidence presupposes that the individual already had a religious belief on forehand. Claiming that social deficit increases religious belief is also hard without presupposing that some belief was already there. Compensating lack of social interactions by interacting with an invisible, divine, being is easier if the individual already has some prior belief. Without it, jumping to beliefs in invisible beings seems a long jump. Misattributing agency also comes a lot easier if the individual already has some idea about the agent to whom actions can be attributed.

Concluding,  I agree with Galen that there is evidence for unreliability but disagree over what evidence is the strongest. His arguments are also rooted in a particular position in epistemology, and may need some refinement.  However, Galen has raised an interesting argument for the unreliability of mechanisms involved in religious cognition and as he rightly points out, CSR cannot be taken as completely irrelevant for the status of religious beliefs.


[i] Though they are a minority position, some unreliability claims have been made. Examples of this approach are: Wilkins, J. S. and P. E. Griffiths (2012). Evolutionary Debunking Arguments in Three Domains. A New Science of Religion. G. W. Dawes and J. Maclaurin. London, Routledge: 133-146 and Goodnick, L. (2016). “A De Jure Criticism of Theism.” Open Theology 2(1).

Self-Report: We Can Do Better (And Are!)

The Religious Studies Project interview with Dr. Luke Galen conducted by Tommy Coleman was an excellent cross-section of some of the a long way to go in figuring out ways to both incorporate nonbelievers into our work as well as to signal when our findings only apply to a particular belief group, instead of all humans (ie. “Increased religiosity helps prevent recurring depression for religious believers” instead of “Religion prevents depression”). The idea that we need to explicitly include nonbelievers in our samples has begun to find solid ground, both in Dr. Galen’s work and others’ (e.g. Galen, 2012; Streib & Klein, 2013), but there have been some issues with developing this idea further. Dr. Galen alluded to one of the major issues in continuing to include nonbelievers, namely the increasing usage of the word “spirituality.” Does it include well-being and having a sense of meaning in life? Feelings of Awe and transcendence? Believing in Ghosts, angels, and demons? Yes, and this lack of clarity is a major problem for studies which try and link “spirituality” with mental health and well-being.

I strongly agree with Dr. Galen’s assertion that the amount of criterion contamination found in most discussions and measures of “spirituality” is problematic, and this point was well-highlighted in Tommy’s point about, “long walks on the beach.” A question that covers so much territory may not even be an accurate reflection of general well-being for people who prefer walking in the woods, let alone serve as a marker of the supernatural component implicit in “spirituality.” While I don’t think Dr. Galen presented a hard-experimentalist view completely dismissing self-report, the criterion contamination introduced by our fuzzy definition of “spirituality” and poorly-constructed self-report measures seem to be bundled up into a problem that exists for self-report measures in general. Just as in the study of moral reasoning, experimental designs which attempt to tap implicit beliefs risk ignoring the fact that humans also seem to be able to exert some conscious control over their beliefs and thus can’t be treated as simply heuristic machines (Cunningham et al., 2004; Turiel, 2008).

Instead, it seems best to attempt to fix the problem of poor self-report measures more directly. We can do this by making measures which don’t use double-barreled questions which nonbelievers can’t straightforwardly answer, explicitly addressing the issue of “supernatural spirituality,” and ensuring that aspects of the measure which tap more general well-being concepts are sufficiently differentiated from supernatural concepts. Additionally, to construct better measures we’ll need to include large enough samples of nonbelievers during all stages of scale development to ensure that the resulting measures are valid for both believers and nonbelievers.

I bring all of this up because there are already measures which have been (or are being) published which meet these criteria, so I can flagrantly advertise them. Cragun, Hammer, and Nielsen’s Nonreligious-Nonspirituality Scale (in press) addresses the problem of fuzzy-spirituality by clearly specifying that respondents should only respond in regards to their beliefs regarding the supernatural aspects of spirituality and not the more general well-being aspects. In addition, their scale was developed for use with believers as well and seems to validly measure the extent of their nonreligious and nonspiritual beliefs, allowing for comparisons between believers and nonbelievers which might not be feasible with “beach walking” measures of spirituality.

While Dr. Galen’s assertion that the well-being of nonbelievers has been underestimated due to incorrectly grouping them with believers who might be experiencing religious and/or spiritual struggles seems to be an accurate depiction of the literature at the moment, this also seems likely to be a problem of improperly interpreted self-report measures rather than with self-report in general. There is initial evidence pointing to a U-shaped curve of well-being related to the strength of a person’s (non)belief (Streib & Klein, 2013). Investigating this idea using the level of control afforded by in-lab experimental studies will be important, but it will also be important to leverage the generalizability of broad self-report studies. We just need a measure of “spiritual” struggles which actually works with the kinds of struggles which might point to lower levels of belief for both believers and nonbelievers.

At the risk of continuing to over-toot the horns of projects that I’m involved with, the Religious and Spiritual Struggles Scale (Exline, Pargament, Grubbs, & Yali, 2014) seems like it will work in that regard. While two of the sub-scales explicitly contain supernatural items, the scale is modular and our early analyses indicated that atheists experience less spiritual struggles than agnostics, when excluding the explicitly supernatural scales (Uzdavines, Bradley, & Exline, 2014). We are currently working on confirming that the scale is measurement invariant with fine-grained belief identification groups (ie. Atheists, Agnostics, Theists, etc) before investigating the link between non-supernatural “spiritual” struggles and well-being, but our early analyses show that it is invariant when considering nonbelievers and believers as two broad groups.

Which is all to say; those of us within psychology of religion who study secularity are privileged to be working in a time where secular beliefs and nonbelievers are starting to be taken seriously within the field as a whole. Maintaining a high level of rigor in the methodology we employ, while important in and of itself, is even more crucial because of the history of criterion contamination within the field that Dr. Galen discussed in this interview and in his own work. “Spirituality” is an overly broad term and, when interpreted incorrectly, can lead to conclusions that more religion leads to more well-being without considering that more nonreligion might also lead to more well-being. It will take much more work to shift the field towards accepting religious nonbelief as a discreet and important category, separate from religious belief even if we still need to clarify our terminology.

But rigorous does not only mean experimental. Self-report can provide interesting avenues of investigation, but more care needs to be taken in building self-report measures which minimize criterion contamination and allow nonbelievers to indicate their level of nonbelief or well-being without having to dance around double-barreled questions. Fortunately, the rapidly expanding breadth of research communities dedicated to investigating secularity should allow the field of secular studies to continue pooling ideas and methodology to illuminate the nature of nonbelief and nonbelievers.


Cragun, R. T., Hammer, J. H., & Nielsen, M. (in press). The Nonreligious-Nonspiritual Scale (NRNSS): Measuring Everyone from Atheists to Zionists. Science, Religion, and Culture.

Cunningham, W. A., Johnson, M. K., Raye, C. L., Gatenby, J. C., Gore, J. C., & Banaji, M. R. (2004). Separable neural components in the processing of black and white faces. Psychological Science, 15(12), 806–813.

Exline, J. J., Pargament, K. I., Grubbs, J. B., & Yali, A. M. (2014). The Religious and Spiritual Struggles Scale: Development and initial validation. Psychology of Religion and Spirituality, 6(3), 208–222.

Galen, L. W. (2012). Does religious belief promote prosociality? A critical examination. Psychological Bulletin, 138(5), 876–906.

Streib, H., & Klein, C. (2013). Atheists, Agnostics, and Apostates. In K. I. Pargament, J. J. Exline, & J. W. Jones (Eds.), APA handbook of psychology, peligion, and spirituality (Vol 1): Context, theory, and research (Vol. 1, pp. 713–728). Washington, DC, US: American Psychological Association.

Turiel, E. (2008). The Development of Children’s Orientations toward Moral, Social, and Personal Orders: More than a Sequence in Development. Human Development, 51(1), 21–39.

Uzdavines, A., Bradley, D. F., & Exline, J. J. (2014). Struggle and the nonreligious: Do weaker forms of nonbelief increase susceptibility to spiritual struggle? In Religious and spiritual struggles: New research frontiers. La Mirada, CA.