I just finished reading “The Adaptationist-Byproduct Debate on the Evolution of Religion: Five Misunderstandings of the Adaptationist Program.” The article was written by Richard Sosis, a professor of anthropology at the University of Connecticut, and it was published by the Journal of Cognition and Culture 9 (2009) 315-332.

This article will mostly consist of a summary of Sosis’ article (I am putting page numbers from the Sosis’ article next to various parts of my summary). Sosis is convinced that the often contentious debate as to whether religion is an adaptation or a byproduct, and the premature declaration that it is a byproduct, is hampering serious interdisciplinary efforts to scientifically study religion. He holds that these disagreements stem largely from disagreements as to the meanings of “core ideas upon which the evolutionary study of religion is founded.” Nonetheless, he is hopeful that these debates can be largely resolved after we take the time to clarify these core ideas.

Many people will probably not take the time to read Sosis’ fine article because they will presume religion could not possibly be an adaptation because the practices and beliefs of many people strike them as bizarre (I also find many such practices and beliefs bizarre).  Richard Dawkins and many other prominent writers have taken this position that religion is not an adaptation; rather, they find it to be an annoying and sometimes dangerous byproduct of evolution (I’ve written about this byproduct position here and here). In fact, this byproduct position is the dominant position among scientists studying religion from an evolutionary perspective.

Although I agree with many of the positions of Richard Dawkins, and I admire him for his energy and willingness to publicly confront the anti-science claims stemming from religious beliefs, I have also criticized his characterization of religion as a byproduct. I believe that Dawkins has come to this byproduct conclusion as a leap of faith (or a leap of frustration) rather than as a result of disciplined scientific inquiry.

It appears to me that the scientific examination of religious has only recently got underway, and there is a long way to go before we can decide, as a scientific conclusion, whether religion is a byproduct or adaptation. I was impressed by the conceptual clarity of the above-cited article by Richard Sosis, and also by the game plan he offers for resolving this thorny dispute.

Before going further, here’s a summary of an “adaptationist position” Sosis developed with Candace Alcorta:

Religion may best be understood as an adaptive complex of traits incorporating cognitive, affective, behavioral, and developmental elements. We argue that these traits derive from pre-human ritual systems and were selected for in early hominid populations because they contributed to the ability of individuals to overcome ever-present ecological challenges. By fostering cooperation and extending the communication and coordination of social relations across time and space, these traits served to maximize the potential resource base for early human populations, thereby increasing individual fitness. The religious system, I contend, is an exquisite, complex adaptation that serves to support extensive human cooperation and coordination, and social life as we know it.

(Page 317)

Sosis recognizes that cognitive scientists have amassed an impressive body of work revealing underlying cognitive structures associated with religious behavior. For instance, Justin Barrett has noted that people often impose anthropomorphic limits on God. Guthrie has proposed a hyperactive agency detective device (which makes it easy to see faces in clouds and creatures in the closet). Guthrie and others propose that human hyperactive vigilance was necessitated to ensure survival (we assume that the rustling sound from a bush is caused by a predator rather than the wind). Whitehouse has identified two broad modes of religiosity, one characterized by a rousing rituals and the other by “mundane, routinized rituals.  Other research has identified that the cognitive expectations and assumptions we use while engaged in ritual activity is common to all goal directed actions. There are many commonalities between religious activity and nonreligious activity, which “serves as a starting point for their contention that religion is a byproduct.”

Adaptationists have joined this discussion much rather recently. For instance, in 2002, David Sloan Wilson’s Darwin’s Cathedral posed a “significant challenge to the byproduct position,” focusing on “the secular utility of religion at the group level.” D.S. Wilson held that religion can confer considerable benefits upon its constituents.

Only recently have prominent thinkers started to question the byproduct advocates. Cognitive research claiming that religion is a byproduct has criticized adaptationist approaches to religion. These pro-byproduct criticisms tend to fall into five categories. (Page 318).

I. It has been maintained that there is no such thing as “religion.” Since it does not exist, how can it be an adaptation?

The concern here is that if you can’t agree on a definition, how can you know what it is that you are studying? Some researchers are convinced that “religion” is a Western academic and political category, and not a “natural kind” with any independent existence. Scholars have not been able to agree on a single definition of religion, even though belief in supernatural agents is a regularly recurring component. Sosis admits that religion is “an inherently fuzzy category with unclear boundaries.” It consists of “recurrent core features that receive varied emphasis across cultures.” Sosis holds that religion is best defined by reference to these recurrent features rather than by any overarching definition, and these prominent features include such things as “ritual, myth, taboo, emotionally charged symbols, music, altered states of consciousness, commitment to supernatural agents, and afterlife belief, among others.” There is debate about which core features should be included in this list. Nonetheless, defining religious activity through this type of list is fruitful for various reasons:

1. It helps to avoid “endless disputes concerning whether Marxism, science, patriotism, sports and so on are religions,” even though it is clear that religion shares some core elements with these other activities. Sosis notes that “most of religion score elements are not unique to religion (e.g., ritual, myth, music and taboo).

2 Making these features explicit will clarify that some group activities emphasize certain core features and others emphasize others.

3. When you break religion down into its core elements, “it becomes obvious that these elements did not evolve together. Sosis cites the work of Eugene D’Aquili et al., who have concluded that “ritual has antecedents in other species,” and thus it is more ancient than other core elements such as myth. Therefore, one should not ask when religion evolved, because it did not appear all at once. Religion “consisted of uniting cognitive processes and behaviors that for the most part already existed.” The right question, then is “when did the features of religion coalesce?” There is no clear answer to this yet

(Page 320). Sosis prefers this “bottom-up” method of defining religion because it lets us ask more productive questions.

Even if religion is simply a Western construct, it is a collection of cognitive processes and behaviors that form an appropriate unit of evolutionary analysis. Specifically, it is an adaptive system, similar to–the less complex than–the respiratory, circulatory or immune systems, all of which are Western constructs and probably lacking in the lexicon of traditional populations, yet no less interpretable through an evolutionary lens. Rather than debate whether “religion” is a natural category and wallow in its murky definitional waters, we should recognize the religious system, consisting of a recurrent set of core elements, as the appropriate unit of evolutionary analysis. To clarify, I am not claiming that we should abandon the study of individual core elements of the religious system, such as supernatural agent police, ritual, music, or emotionally charged symbols. Quite the contrary; studying those core elements is essential. The point here is that it is the religious system–the coalescence of these elements–that must be the focus of an adaptationist analysis.

(Page 321).

Sosis points out that “adaptation” is an elusive concept. “Adaptation” refers to both a process of phenotypic modifications by natural selection, as well as the products of that process. throughout his article, Sosis uses only the second meaning. In that sense, “adaptations solve particular ecological problems organisms face in acquiring energy for growth and reproduction. More specifically, adaptations are traits that exist because of a process of phenotypic modification by natural selection for a particular gene-propagating effect.”

Sosis offers additional definitional work. What is a “trait”? This term lacks a universally accepted definition. Essentially, “traits” are “quantifiable features of organisms,” though there is no “one-to-one relationship between genes and traits. Further complicating things, traits are “integrated with one another” and therefore organisms are not simply collections of traits. A trait need not be a physical feature: “behaviors and cognitive processes can also be analyzed as traits.”

Putting these two concepts together, a trait is “adaptive” if it confers reproductive benefits upon its bearer in a particular environment.” However, showing that a trait is adaptive “does not establish that the trait is an adaptation.” Therefore, the adaptationist analysis of “the religious system” needs to identify the system’s function (if any) and determine “specific selective pressures that are responsible for its evolution.” Sosis defines “the religious system” as the appropriate “trait” to be studied, reminding us that it is a “collection and interaction of defined cognitive, emotional and behavioral elements.”

II. The psychological mechanisms that produce religious thoughts and behaviors were not designed to produce religion; therefore religion is a byproduct.

Sosis argues that it has not yet been demonstrated that religious beliefs and behaviors are “inevitable spandrels or byproducts of the psychological mechanisms that produced them.” Even if a proximate explanation (For instance, the hyperactive agency detection device–HADD) causes a belief in God, this doesn’t tell us whether “the religious system” is an adaptation. All adaptive systems consist of building blocks or parts that cannot be given an evolutionary analysis in isolation.

For instance, to evaluate whether the human respiratory system is an adaptation to mediate the movement of oxygen and carbon dioxide in and out of the body, a detailed analysis of the larynx would be important, but insufficient to reveal the selective pressures that ultimately shaped the respiratory system. .. . We would have to know how [the larynx] fits within the respiratory system and other functions unrelated to respiration it may sustain (such as its role in human vocalization). Similarly, we cannot evaluate whether or not the religious system is an adaptation by examining its independent parts in isolation. We must consider the religious system more comprehensively, focusing on how the constituent parts contribute to the system, how the parts interact with each other to achieve functional goals, and other functions that the parts might play.

To clarify, studying the independent parts that constitute religion–ritual, myth, supernatural agent believed and so on–is crucial. And the argument here is in no way a critique of reductionism. I am simply pointing out that to claim that the religious system is not an adaptation because the cognitive systems that produce supernatural agent belief might not have evolved to produce such beliefs is misleading and inaccurate. It is the religious system, not the constituent parts, that produces functional effects and is the appropriate unit of an adaptationist analysis. The proper byproduct account of religion, which has yet to be offered, must explain why the religious system’s constituent parts recurrently coalesce across cultures. The most likely evolutionary scenario is that cognitive, emotional, and behavioral elements were exacted for use in a complex system of communication, cooperation and coordination, namely the religious system.

An exaptation is “a preexisting trait that acquires a new role for which it was not originally designed by natural selection.” Sosis points out that exaptations “have functional effects but exapted traits are not modified when taking on a new role. If a trait that evolved for a particular function (e.g., feathers for heat retention) remains unchanged as it “moves into a novel ecological niche” it is an exaptation. If the move into the new niche “spurs structural design changes, feathers would be considered secondary adaptations.” (323). Many adaptationists agree that the cognitive and emotional mechanisms constituting religious beliefs and behaviors did not evolve for this purpose, but they argue that the co-opting of these existing structures “for novel solutions to ecological challenges is a hallmark of evolutionary adaptation. A key point, then, is whether these that have been exapted by the religious system “have been adaptively modified by the new socioecological niche created by religion. If yes, the religious system is an adaptation (or “secondary adaptation,” in Stephen Jay Gould’s terminology); if no, the religious system is an exaptation. While I would put my money on the former, this remains an open question in need of further research.”

III. Adaptationist accounts of religion are “just so stories.”

“Just-so story” is a common critique of evolutionary psychology research. Sosis does not take this criticism as an argument against religion (or any trait) as an adaptation. “It is an argument for better scientific standards, and one which I fully endorse.” He quotes David Sloan Wilson:

“Properly understood, “just-so story” is just another phrase for “untested hypothesis” and should be treated as a rallying cry for another turn of the crank, specifically more hypothesis testing.”

(Page 324) Sosis holds that there is no agreed-upon protocol for deciding what counts as an adaptation, although Paul Andrews has proposed six evidentiary standards that are been used by biologists (e.g., phylogenetic comparisons, fitness maximization, and beneficial effects in ancestral environments). He cautions, however, that to prove one’s case, “alternative explanations for the emergence of trait characteristics must be eliminated.” This can get tricky, as he discusses at page 324, and byproduct advocates argue that religion lacks all of these characteristics because religion does not seem to be very good at solving any particular adaptive problem. Sosis counters that it is impossible to evaluate this claim that religion is not very good at solving a problem when the problem itself remains unspecified.

In fact, if we consider the religious system as a complex adaptation for cooperation, built on highly flexible cognitive processes, there is indeed considerable historical ethnographic and experimental evidence that religion does this quite well.” [Many citations here, at page 325].

Sosis admits that it is often difficult to eliminate all alternative explanations for the existence of a trait. But what is good for the goose is also good for the gander:

“What is often unappreciated is that the standards of evidence necessary to support the position that a trait is a byproduct are no less burdensome than establishing that a trait is an adaptation. Indeed, adaptationist hypotheses must be tested as alternative explanations. Needless to say, such standards of evidence are rarely met, especially by those claiming that religion is a byproduct.”

(Page 325). Sosis then quotes David Buss:

Hypotheses about functionless byproducts must meet rigorous scientific standards that include a functional analysis of the original adaptations responsible for producing the functionless byproducts and the existing human cognitive and motivational mechanisms responsible for the co-opting. Without this specification, the mere assertion that this or that characteristic is an exaptation encounters the same problem that [Stephen Jay Gould] leveled against adaptationists– the telling of “just-so stories.”

(325-6).

IV. If religion were an adaptation, everyone would be religious. Since everyone is clearly not religious, religion must not be an adaptation.

Sosis argues that this critique is tempting to ignore because of its “blatant ignorance of evolutionary processes.” The fact that many parts of the world have high rates of atheism does not disprove the adaptationist. “Moreover, neither adaptationists nor byproduct theorists endorse genetic determinism. On the contrary, evolutionary theory assumes that environmental input during ontogeny is critical for the expression and adaptive functioning of many traits, including religious belief.”

V. Religion is so costly. How could it be adaptive?

The cost it takes to be involved in religious practice can be extreme, and this cost includes “time, resource and opportunity costs, as well as physical and emotional pain.” Celibacy and self-sacrifice are mentioned as examples. This “cost puzzle” of religion can be solved by recognizing the potential payoff. What if turns out to be true that religion is:

An evolving system of communication, which offers mechanisms that can promote in-group trust and overcome commitment problems. … The costliness of religious activities or specifically what I’ve referred to as the four “B’s” – religious belief, behavior (rituals), badges (such as religious attire) and bans (taboos)– enables them to serve as reliable and honest signals of group commitment [this passage is from one of Sosis’ articles published in the 2006 edited volume, Where God and Science Meet: How Brain-Dead Evolutionary Studies Alter Our Understanding of Religion.]… Only those who are committed to the group will be willing to incur the time, energy and opportunity costs of religious adherence, but by doing so they demonstrate their commitment and loyalty to the group, and can thus achieve a net benefit from successful collective action and other status benefits available to trusted signalers. Contrary to Lee Kirkpatrick’s assessment the application of costly signaling theory to religious phenomenon is one of the most active areas of adaptationist scholarship on religion.

(Page 327)

Sosis concludes by asserting that religion cannot meaningfully be treated as “a seamless whole.” Further, the fact that religion is a “fuzzy category” does not preclude an adaptationist analysis. He admits that this adaptationist analysis will be “exceedingly complex,” but the byproduct analysis would be “equally challenging.” One of the most difficult issues for evolutionary researchers is determining how the purported exapted traits that constitute the religious system interact with one another.

Although much progress has been made, much more work needs to be done. In the meantime, it is clear where Sosis is placing his bets. In his final paragraph, he paraphrases Steven Pinker’s and Paul Bloom’s comment on the functional design of the eye:

It is impossible to make sense of the coalescence of elements that constitute the religious system without noting that it appears as if it was designed for the purpose of uniting individuals under common purpose. Systems that can do what the religious system does are extremely low-probability arrangements. By an unimaginably large margin, most biologically possible arrangements cannot unite unrelated organisms under common purpose, achieve extraordinary self-sacrifice, and motivate large-scale cooperation and coordination. All of this suggests that the religious system is an adaptation. Now we must begin to properly evaluate this possibility.

(Page 328-9).

[Note I: Here is the original passage from Steven Pinker and Paul Bloom’s “Natural language and Natural Selection” :

It is impossible to make sense of the structure of the eye without noting that it appears as if it was designed for the purpose of seeing — if for no other reason that the man-made tool for image formation, the camera, displays an uncanny resemblance to the eye. Before Darwin, theologians, notably William Paley, pointed to its exquisite design as evidence for the existence of a divine designer. Darwin showed how such “organs of extreme perfection and complication” could arise from the purely physical process of natural selection.

The essential point is that no physical process other than natural selection can explain the evolution of an organ like the eye. The reason for this is that structures that can do what the eye does are extremely low-probability arrangements of matter. By an unimaginably large margin, most objects defined by the space of biologically possible arrangements of matter cannot bring an image into focus, modulate the amount of incoming light, respond to the presence of edges and depth boundaries, and so on. The odds that genetic drift, say, would result in the fixation within a population of just those genes that would give rise to such an object are infinitesimally small, and such an event would be virtually a miracle. This is also true of the other nonselectionist mechanisms outlined by Gould and Lewontin. It is absurdly improbable that some general law of growth and form could give rise to a functioning vertebrate eye as a by-product of some other trend such as an increase in size of some other part. Likewise, one need not consider the possibility that some organ that arose as an adaptation to some other task, or a spandrel defined by other body parts, just happened to have a transparent lens surrounded by a movable diaphragm in front of a lightsensitive layer of tissue lying at its focal plane. Natural selection — the retention across generations of whatever small, random modifications yield improvements in vision that increase chances of survival and reproduction — is the only physical process capable of creating a functioning eye, because it is the only physical process in which the criterion of being good at seeing can play a causal role. As such it is the only process that can lead organisms along the path in the astronomically vast space of possible bodies leading from a body with no eye to a body with a functioning eye. ]

[Note II: I intentionally created the image that accompanies this article to illustrate my recurring concern that the use of the word “adaptation” often invites a gut emotional judgment that can only be cashed on in terms of what is “good” from the user’s personal perspective. I am deeply concerned that when applied to human traits, “adaptation” invites contentious debate that is no more likely to be resolved than debates about aesthetic judgments. Dandelions are certainly no less an adapted species than the types of flowers we like to display at weddings and funerals].