The Evolution of Religion: Cognitive Byproduct Theory in Conjunction with Commitment Theory within a Multi-Level Selection Model to Account for the Ubiquitous Presence of Religion as a Human Preoccupation.
Igor Britan
Abstract
Genetic, memetic, and trait-group selection within a multi-level selection model are used to organize findings in anthropology, psychology, and cognitive science as they pertain to a discussion of human religion as a product of evolution by natural selection. Cognitive Byproduct Theory in conjunction with Commitment Theory provides a theoretical framework that accounts for the observed nature of religion today as it stems from roots in ancestral human populations. Further, morality is discussed as a human capacity independent of religion though requisite for its proper function.
Introduction—Evolution by Natural Selection
It has been said that humans occupy the intellectual niche. While other animals develop intricate biological solutions to Earthly constraints like nutrition, reproduction, and defense, humans seem to rely predominantly on their unique ability to think rationally. Rationality requires no definition, but I suppose it equates well with the scientific method. A rational thought is one which leaves the least room for error. It assumes as little as possible and produces a conclusion that is as factually accurate as can be with the amount of information provided.
It is surprising then that a substantial proportion of human activities appear to be devoted to utterly irrational behaviors—irrational in the sense that they aim to achieve a goal without any clear benefit and utilize a process of reasoning that ignores the most fundamental component of rational thought: the concept of falsifiability. It should require no mention that in order to propose a claim such as A is B, it must be at least theoretically possible to demonstrate that A is not B. Otherwise, the proposition is a meaningless one.
On the other hand, the world is cluttered with such ideas, ranging from relatively harmless astrology to utterly destructive Islamic and Christian fundamentalism—in other words ‘religion.’ Further, as far as I am aware, there is currently only one empirically-sound process by which any feature of any species can arise and that is the process of evolution by natural selection. The following is an evaluation of the current state of evolutionary science as it pertains to the birth and subsequent propagation of religion. Cognitive Byproduct Theory is presented in conjunct with Commitment Theory within the framework of Multi-Level Selection as they pertain to the evolution of religious concepts and behaviors at the genetic, memetic, and group-levels. Before I get into the substance of this presentation, consider for now the following preliminary discussion.
Genetic traits are only one kind of trait. Evolution by natural selection is not a concept that is limited to genes. Anthropologist Pascal Boyer defines three elements of evolution by natural selection, none of which necessitate genes as the unit on which the selective pressure operates (Boyer 2001, 34). The elements are a set of units, a mechanism by which these units can mutate, and a process of differential transmission that impacts the transmission rate of units beyond the rate of mutation—that’s all. Any such system will experience selective pressures and the units existing within the system will invariably represent the best adaptive solution to the problem presented by the transmission process.
Consider the following simple anecdote to emphasize the diversity of the process of evolution by natural selection. A few years ago, I purchased a Motorola cell-phone which I managed to break after dropping just one time. I proceeded to exchange it for another one which, disappointingly, I also broke after once again dropping it. Although shameful for me to admit, I then managed to drop and subsequently break two more phones within about one month (both after the first fall). My fifth phone, however, was a different story. I dropped the last phone at least twenty times during the course of my two year contract and, to my greatest joy, it did not break a single time. The units were the phones; the process of mutation is unknown to me but I imagine it had something to do with the factory assembly process; and the differential transmission was donated by my repeated negligence. In the end, the selective pressure I created within the system of my cell-phone usage yielded a species that possessed some advantage in the form of resistance to breaking when dropped.
The question is: if evolution by natural selection is such a broad concept, then how does one actually formulate a theory that describes an entity, in this case religion, as a product of evolution? The simple answer is that evolution by natural selection is currently the only empirically-tested process by which any degree of complexity can arise in the universe; thus any complex entity necessarily evolved by natural selection of some sort. To take it further, it can be said that any entity whatsoever exists in the form that it does directly as a consequence of whatever natural selection is taking place on the relevant level or levels pertaining to that entity.
On the other hand, one can take a less philosophic route and consider not just generalized conclusions. A satisfactory evolutionary account for the origin of any particular trait is one that sufficiently accounts for all three elements of evolutionary theory. Thus it is one that identifies the units being selected, describes the mechanism of mutation, and finally discusses the selection process with an emphasis on how it represents a problem solved by the adaptive quality of the trait in question.
Note that the mutation concept mentioned earlier, though quite clearly related both in terminology and spirit to the most famous instance of natural selection—that at the genetic level—is actually often quite different. In the case of my cell-phone, the process of mutation in fact had little to do with mutation in any biologically-meaningful sense of the word. The cell-phones that came and went were presumably variants of each other but, not being a biological species, they bore no familial relation to each other, as is demanded by the genetic interpretation of the word ‘mutation.’
Defining Religion
Religion is a word that almost needs no definition. For this reason it is also a word that requires the most extensive of definitions if one is to consider it in an empirical fashion. Certain researchers take advantage of the implicit meaning of religion and leave the interpretation largely to the reader, while others are very meticulous in identifying their parameters.
Thinkers as far back as our written record can take us have found an interest in studying religion. It wasn’t until the early 19th century, however, that evolutionary theories became a subject of interest for the anthropologists of the day (Dow 2006, 1). Anthropologist James Dow describes the state of anthropology as it pertained to the study of religion from the early 19th into the 20th century as primarily culture-oriented, putting forward a series of “lower-to-higher” theories, identifying the origin of origin in so-called savage states (non-Western states) and the enlightened end-product in themselves (2006, 1).
These theories increased in complexity but ultimately failed. The underlying flaw of such theories was that they neglected the possibility that a religious predisposition is equivalently present in all humans; that it is perhaps only manifested differently as a result of some unseen factors. Dow quotes evolutionary psychologist Stephen Pinker in expressing that these early theories all made the same naïve mistake—considering the mind as a blank-slate rather than a collection of dispositions, abilities, and constraints (2006, 2).
Biologist Nicholas Wade suggests in his treatise on religion, The Faith Instinct, that the best place to seek an unadulterated view of religion as it first existed is in societies that geneticists have identified using genetic markers as most unpolluted by external DNA. These culturally isolated hunter-gatherer societies, in Wade’s view, best allow a glimpse of religion in its most primitive form (2009, 5).
Wade believes that by 50,000 years ago, the neural circuitry requisite for the processing of religious concepts was firmly in place. Wade compares religion to language, defining it as a capacity to be filled. “People are born with innate instincts for learning the language and the religion of their community.” Similar to language, Wade consider religion predominantly a form of communication. Although both language and religion can of course be practiced in solitude, both, in his view, exhibit their most important features in the company of others (2009, 6).
A more succinct definition is provided by Anthropologists Joseph Bulbalia and Andrew Mahoney. In studying the correlation between religion and generosity, the two defined religion simply as “beliefs and practices respecting gods.”—A one sentence definition that seems to allow for little exception. On a related note, Bulbulia and Mahoney also provided a convenient and workable definition of the term “religiosity.” In contrast to the beliefs and practices connoted by ‘religion,’ ‘religiosity’ refers instead to whatever psychological properties produce the beliefs and practices termed ‘religion’ (2008, 1).
Neuropsychologist Fred H. Previc, defines religion as “one of five major behavioral phenomena that evolved with modern humans and are found in all human cultures—the others being language, advanced tool-making, music, and art”. In Previc’s view, religion must have a genetic basis of some sort, as suggested by findings correlating religiosity—as defined above—with models of genetic inheritance. He concedes, however, that no theory has yet to offer a complete and general pathway linking religion to particular brain systems (2005, 1).
Previc assigns three distinct components of religion: beliefs, practices, and experiences. He holds that all religious concepts are intrinsically-bound to the concept of the supernatural and defines the gradient of religiosity as follows. “…persons in societies having strong religious beliefs are more likely to interpret anomalous experiences as religious in nature” (2005, 2). Previc offers an optimistic finding, making the claim that religion appears to rely predominantly on domaminergic transmission in the ventromedial cortex—a finding the specifics of which are largely irrelevant to the underlying implication that perhaps someday a discrete relationship will be uncovered connecting human religiosity to the structure and function of the human brain.
For the purposes of this thesis, I have synthesized the preceding definitions of religion. Religion is to be defined as the beliefs, practices, and experiences of individual humans as they relate to gods and other supernatural concepts. More generally, religion can viewed as any instance of acceptance of an unfalsifiable claim, though this definition is much too broad for the time being. With this definition in mind, I now move on to a comprehensive analysis of the emergence and propagation of religiosity in the human species with respect to the only empirically-sound mechanism of complexity—evolution by natural selection.
The Basis of Religion—Cognitive Byproduct Theory
In 2001, evolutionary anthropologist Pascal Boyer published a fascinating book which he bravely titled Religion Explained: the Evolutionary Origins of Religious Thought. As promised by the title, Boyer aims throughout the book to explain religion as a function of natural selection and strives to every conceivable end to remain as objective as is demanded by such an undertaking. Armed with a lifetime of experience in the field of Anthropology and a wealth of sources, Boyer is able to pin-point what he believes are the fundamental laws governing religious concepts—and for that matter all concepts.
Boyer’s primary claim is summarized by the following excerpt:
“The explanation for religious beliefs and behaviors is to be found in the way all human minds work. I really mean all human minds, not just the minds of religious people or of some of them. I am talking about human minds, because what matters here are the properties of minds that are found in all members of our species with normal brains.”
Contrary to the objectives of his peers, Boyer’s intention in Religion Explained is not to theorize on the nature of particular religious beliefs and behaviors or to provide a tale of selective-advantage to illustrate the inevitability of some facet of religion; rather, he aims to explain the most fundamental level at which these phenomena operate (2001, 2). Specifically, Boyer emphasizes the fact that human minds are clearly not prepared to acquire just any notion or to participate in just any behavior (2001, 3). Conversely, the totality of available research indicates that human minds are far from blank slates at birth. Our minds are filled with dispositions and constraints of various sorts that culminate in the sort of species we are, much like those of giraffes culminate in the typical behavior associated with a giraffe (2001, 3). Presumably, if we had the ability to study the thought processes of giraffes then we might demonstrate that those, too, are governed by a similar set of dispositions and constraints, much like ours are proposed to be.
Boyer’s thesis in Religion Explained represents the largest component of what is known as the Cognitive Byproduct Theory (CBT) of religion’s evolutionary origins. On the surface, CBT appears to offer the conclusion that the majority of fitness advantages associated with religion are, in fact, outside religion itself (Dow 2006, 6). Indeed I agree that on the surface this seems so, but, as I will demonstrate, this is actually far from the case.
The first step to grasping the implications of CBT is to consider Boyer’s analysis of what supernatural concepts are actually like. Boyer begins with a thought-experiment, asking his readers to look over a list of claims and to evaluate whether or not any known religions have been built around these claims (2001, 51). Consider for example the following four.
- There is only one God, he is omnipotent, but He ceases to exist on Wednesdays.
- The gods are watching us and they notice everything we do. But they forget everything instantaneously.
- When people die their souls sometimes come back in another body.
- We pray to this statue because it listens to our prayers and gives us what we want.
If you guessed that 3 and 4 were the only viable statements on that list then you were completely right (2001, 52). And if you considered it a bit odd that without any concrete knowledge you were able to pick out those statements which no religion has yet found an identity in, then you may also be onto the idea I am trying to convey. Indeed it is a prediction of Boyer’s thesis that any human with a functional brain will be able to intuitively discern between acceptable and unacceptable claims, as far as religion is concerned and beyond (2001, 54).
Further, the observation that statements 1 and 2 just don’t cut it is indicative of an even more profound aspect of CBT. Boyer firmly holds that what ultimately makes religious concepts such that they are is their adherence to intuitive human expectations (2001, 55). In the mind of any human, Boyer explains, it stands to reason that things don’t simply cease being for one day every week, as is suggested by claim 1. The same reasoning applies to claim 2, which suggests something that any person would find absurd, that an omnipresent agent is able to observe everything we do and yet remember none of it. A less intuitive reason that dismisses claim 2 is the fact that it yields no behavioral consequences, though this is a matter for later discussion.
Although all four claims are quite clearly equivalently poorly-grounded in empirical facts, the latter two claims accomplish what the former two do not—that is they adhere far more so to the typical human’s intuitive expectations. Though the idea that a soul—interpreted as a person’s essence, I suppose—might die and come back in another body does indeed introduce a non-obvious claim, it does not introduce one that is in direct opposition to that which we consider given. The same goes for claim number 4 which stipulates a statue endowed with the ability to interpret demands and to respond to them. The statue behaves essentially as a human being and so requires minimal mental exercise in its cognitive representation (Boyer 2001, 56).
Boyer uses examples like the preceding to demonstrate the fact that religious claims are clearly not just any sort of unverifiable beliefs; rather there is a relatively limited subset of forms that religious concepts can take. These limitations, as CBT would have it, are artifacts of human nature itself (Dow 2006, 7). In particular, the preceding claims are exemplary of the role of something Boyer calls “ontological category violations,” as they pertain to defining the limitations of religious concepts.
Citing research on human children and even going as far as to mention some of the other more intelligent mammals, Boyer emphasizes that as far as cognitive science is concerned, humans do not acquire concepts in just any random manner. Indeed, Boyer and his colleagues hold that concepts are introduced into the mind via something he calls a “template.” A person does not simply encounter a concept, save it for later use, and retrieve it at will; the process is in fact far more complex—though at the same time perhaps simpler (2001, 57-61).
Boyer’s conception of ontological categories—and their subsequent violations—is intricately bound to the idea of templates. An ontological category is an abstract concept, such as an animal, tool, person, or number. Each such category has attached to it a particular template. According to Boyer, every time a person encounters a new instance of some category, such as a new animal or a new person, he or she instinctively begins to mentally fill out the appropriate template (2001, 60).
If a person is told, for example, that object X is an animal, that person accesses the animal template and instinctively knows precisely what features are relevant in describing such an object. Since object X is not a tool, it isn’t instinctive to question what its purpose is. On the other hand, any person with a functioning brain will instinctively know to ask about the animal’s mode of reproduction, its nutrients of choice, and all other categories necessitated by the animal template. This person will also instinctively infer that any independent movement made by object X is necessarily a function of some desired outcome. Only the inanimate object template would permit the consideration that the object moves only if acted on by an outside force (Boyer 2001, 59-61).
The concept of inference mentioned above effectively captures the essence of Boyer’s CBT. Boyer defines an inference as the generation of a fact that was not explicitly conveyed by an informational input. For example, take the input that object X is an animal. All of the prescriptions of the animal template constitute inferences. One does not have to be told that object X requires food in order to survive and yet this fact is conveyed by the simple observation that object X is an animal (2001, 60). In my opinion, this point is Boyer’s most significant.
This discussion extends to religion in that the most successful religious concepts appear to be those that satisfy the greatest amount of default inferences. This is the reason that statements 1 and 2, discussed earlier, failed to qualify as viable religious concepts. Not only were they not proven empirically but they directly violated our intuitive understanding of how the world functions. Empirical proof, in fact, is a relatively modern invention—the vast majority of human contemplation over the years has proceeded via an intuitive process of evaluating whether a concept does or doesn’t make sense (Boyer 2001, 69).
According to Boyer, those religious concepts that we observe today are those that have survived a de facto process of selection, in accordance to the intuitive inferences they were able to satisfy (2001, 75). An extension of this thought ties into the concept of meme theory.
Evolutionary Biologist John S. Wilkins defines a meme as “the least unit of socio-cultural information relative to a selection process that has favorable or unfavorable selection bias that exceeds its endogenous tendency to change (1998, 5).” What makes Wilkins’ definition a good one is that it defines the ‘meme’ in a fashion that conforms to the constraints of the established evolutionary model.
To allow for the assumption that religious claims qualify as functional memes allows the transformation of this discussion into a legitimate evolutionary evaluation. Before I proceed further, it is necessary to discuss the idea of “multi-level selection” as advocated by David S. Wilson in Darwin’s Cathedral.
Recall the three elements of a successful theory of evolution by natural selection. These are the identification of units, an account of a mutational process, and a description of the mechanism of differential transmission. Boyer’s ideas intimately conform to the stipulations of Wilson’s multi-level selection model to culminate in a successful discussion of selection on at least two levels.
On both levels, the units are memes and the process of mutation is the process by which humans formulate their claims in the first place. The reason that two levels are identified instead of just one stems from the difference between the mechanisms of differential transmission on these two levels.
The differential transmission operating on one level, as already discussed in great depth, is the human tendency to accept only a limited subset of possible concepts, in relation to the degree that these concepts adhere to our intuitive expectations. In line with Boyer’s thesis, those concepts that stray too far from what a human will intuitively accept as plausible are selected against while intuitive concepts will persist. This doesn’t strictly apply to religion, however, and so permits our departure from this point.
On a second level of selection, Boyer identifies yet another aspect of religion—its recall value. By referencing studies conducted by himself, his colleagues, and others, Boyer concludes that concepts that violate their primary ontological category are far more memorable than concepts that do not. Religious concepts, it so happens, all unanimously appear to violate precisely their most fundamental ontological category (2001, 81-82). People that are omnipotent (Gods), artifacts that can listen (religious statues), people that lack bodies (ghosts), etc. are all fundamentally in contradiction with the established understanding of the typical behavior of these entities. At the same time, however, they preserve all relevant ontological expectations. The fact that gods and ghosts clearly operate as a function of the human template is evidenced intuitively. As Boyer explains, an analysis of any of the world’s religions will reveal that all conceptions of gods, ghosts, and spirits invariably assume human-like properties (2001, 76). Likewise, he points out that conceptions of sacred artifacts also conform to all intuitive expectations of artifacts, except of course for those aspects that require violation for inclusion as a religious concept.
With these two levels of selection in mind, while noting the possibility and likelihood of additional levels, Boyer concludes that the religious concepts in existence today are those that have survived selection on all relevant levels and are thus the best alternatives that humans were able to come up with given their natural constraints (2001, 61).
What we can take away, then, from Cognitive Byproduct Theory is the observation that religious concepts are subject to a selection process like any concepts with regard to conception, memory, recall, and transmission. We can be certain that if CBT has gotten it right that empirical testing will demonstrate the efficacy of modern religious concepts at conforming to all of the constraints of the typical human mental machinery. What CBT does not provide is an answer to the question of why religious concepts came into existence in the first place and why they were subsequently propagated by some process of multi-level selection into phenotypic ubiquity as they were.
As defined earlier, religious concepts appear to be predominantly concerned with the actions of beings termed gods. In Boyer’s view, fully in line with the other predictions of CBT, the human preoccupation with gods is merely a manifestation of our capacity to infer agency in our environment. When presented with a seemingly purposeful object or event, it is a natural human disposition to assume in it the workings of an intentionally-minded agent (Boyer 2001, 96-97).
It is here that Cognitive Byproduct Theory demonstrates the byproduct aspect of its name. CBT makes the claim that all relevant mental systems for processing religious thoughts were in place to perform non-religious functions (Dow 2006, 6). Indeed this is so in Boyer’s view. The systems needed to create religious interpretations of environmental inputs are all systems required to perform basic survival functions (Boyer 2001, 97). The systems include agency-detection, inherited from our ancestry as animals on the lookout for predators at all times. As Wilson notes in Darwin’s Cathedral, “a rabbit that hears a rustling in the leaves and assumes a fox will survive better than one that does not (2002, 25). Other required systems are those responsible for interpreting physical causation, keeping track of who’s who, linking structure to function, among others. Perhaps the most important of these is the mental module responsible for moral reasoning, which I will expand upon shortly. The conclusion of CBT, however, is unequivocal, that all of the afore-mentioned mental systems and the others involved in the processing of religious concepts absolutely do not possess an evolutionary origin specific to religion (Dow 2006, 6).
CBT effectively explains what enables the representation of religious concept in human minds. It does not, however, satisfy the requirements of a complete evolutionary account because its framework lacks a discussion of the selective advantage religion is certain to have conveyed to practicing populations at the time of its emergence. After-all, there is a near infinite array of possible concepts that can utilize the human mental machinery, despite all of its identified limitations.
The following sections propose a way to fill the void left by Cognitive Byproduct Theory. Two prominent alternative theories proposing a natural-selection-driven evolutionary process responsible for human religion are Commitment Theory (CT) and Environmental Regulation Theory (ERT). Unlike CBT, these two theories are concerned not with the underlying dispositions responsible for the observed nature of human religion, but rather with the selective advantage that religion appears to confer. CT will be evaluated in depth while ERT will be left for subsequent revisions. A key consideration to keep in mind is that CT and ERT are essentially interchangeable as far as they represent the sort of theory that would appropriately satisfy the goal of providing a comprehensive analysis of religion with regard to its reliance on the process of evolution by natural selection.
Trait-Group Selection—Relating CBT to Commitment Theory
In 1974, Evolutionary Biologist David Sloan Wilson, then at Michigan State University, published a paper in which he coined the term “trait-group.” Wilson defined a trait-group just as it sounds—a group of organisms affected by a particular trait (1).
According to Wilson, in a population of randomly mixing interbreeding individuals, traits that relate to mating, competition, feeding, predation, and others can be considered according to their impact on the trait-group as a whole (1974, 1). This is in contrast to the idea that traits, in the evolutionary sense, are subject only to selection at the individual level. To the contrary, Wilson pointed out; certain traits affect more than just the individual possessing the trait, particularly in organisms that conform to certain specifications (1974, 2).
In Wilson’s view, the evolution of altruistic traits—those that benefit the group at a cost or no benefit to the individual—can evolve by the process of trait-group selection (1974, 2). Consider the following scenario as per Wilson’s model.
A population of organisms differs only with regard to expression of trait-A. Type-A individuals express it while type-B organisms do not. As these organisms differ in only this respect, any difference in fitness between the two types is assumed a direct consequences of possession of trait-A. If trait-A is an altruistic trait, then it will yield a benefit to other members of the trait-group—both type-B and type-A individuals (Wilson 1974, 4).
Wilson proposed a mathematical relationship between average group fitness and the proportion of type-A individuals in the group. Using complex analytical tools, he demonstrated that sufficiently altruistic groups can benefit from positive selective pressure in relation to less altruistic groups. That is, despite the lack of selective pressure for trait-A at the individual level, the trait can still be favored by natural selection at the group level and thus promoted into phenotypic ubiquity.
Wilson was cautious, however, in prescribing trait-group selection as the predominant mechanism of any sort of evolution. He emphasized, in fact, that only certain kinds of organisms are subject to significant evolutionary input from selection at a group level (1974, 4).
Wilson’s model of trait-group selection stipulates that the organisms under consideration must be part of a particular sort of population. This population must be such that its members constantly isolate themselves into separate groups which compete with each other for some time period but frequently merge and reshuffle (Wilson 1974, 4). Thus if an altruistic trait is injected into the gene-pool of such a population of organisms, then trait-group selection can become a powerful force in selecting groups with a higher-proportion of altruistic individuals, as these groups will have a higher average fitness than competing groups.
The model breaks down, however, if these conditions are not met. If the population in question remains large, un-split, and not involved in inter-group competition, then trait-group selection may not be an applicable consideration. There is reason to believe, however, that a substantial period in human evolutionary history created an environment in which trait-group selection played an important role (Wilson 2002, 20-24).
This reasoning stems primarily from the sort of animals that anthropologists consider ancestral humans to have been. Following the period from our divergence with our significantly more ape-than-human-like common ancestor with chimpanzees, humans experienced a transition from a society dominated by alpha-male hierarchy to one that was largely egalitarian (Wade 2009, 45). A theory advocated by some anthropologists holds that this may be related to the invention of weapons making physical strength more or less obsolete (Wade 2009, 47).
The egalitarian nature of early human populations satisfies the first condition of Wilson’s trait-group selection model—that all members of the population start off on a level playing field. Biologist Nicholas Wade supports the claim that individual differences in ancestral human groups were effectively suppressed, emphasizing the differences between groups as a dominant factor of selection. He cites the findings of anthropologists studying modern hunter-gatherer societies, who find that the few such civilizations remaining are all exceedingly disapproving of selfish behavior and wealth disparity among members (2009, 70).
The second stipulation of the model, that the population continuously break ups, rejoins, and reshuffles its members into smaller groups is met with flying colors as well (Wilson 2002, 70). Indeed the early populations of humans were in a constant state of warfare between small groups of a few hundred people, competing for a limited set of resources (Wade 2009, 72). Archeological evidence suggests that as much as 13 to 15% of deaths during this time period occurred as a result of inter-group warfare (Wade 2002, 72). Wade suggests as comparison the 1% that died in the 20th century, despite two world wars.
The deadly competition among early humans fostered an environment in which groups were constantly created and destroyed (often literally destroyed by competing groups), but more often they simply broke apart or joined others (Wade 2009, 73). The fact that the early human environment appears to have been a significant promoter of Wilson’s trait-group selection process, suggests strongly that this time period in human evolution was one during which pressure at the group level promoted the selection of traits that benefited groups rather than individuals.
If the analysis of Wilson and Wade is correct and the process of trait-group selection was indeed the mode by which altruistic traits were introduced and cemented into the human repertoire, then the question is: what sort of altruistic traits were these? One answer is that they were religious ones.
With regard to the three elements of a natural-selection-driven evolutionary process, the aforementioned discussion has accommodated the first and the third. Wilson’s model of trait-group selection identifies both the units and the mechanism of differential transmission; however it does not discuss the process of mutation. The units are the trait-groups themselves while the selection aspect of the transmission is provided by the difference in average fitness among groups of varying levels of altruistic, type-A, individuals. The mutation aspect is the subject of some contention (Pyyasiainen and Hauser, 2009, 1). Some researchers, like Dow, interpret the universality of religion as proof of its genetic foundation (2006, 20). Boyer, on the other hand, considers the question entirely irrelevant (2001, 3). My analysis reveals that the question is indeed partially irrelevant. As CBT suggests, religion is indeed grounded in universal human cognitive dispositions, though, on the surface, religion is best represented as a memetic trait.
A theory that fits in with a memetic evolutionary model of religion is Commitment Theory (CT). Originally sketched out by economist Robert Frank, CT was adapted and expanded by cultural anthropologist Richard Sosis. At the root of CT is the observed paradoxical nature of the religion phenomenon. With Commitment Theory, Sosis attempts to account for religion’s tendency to simultaneously waste time while it appears to improve the longevity of practicing groups.
To arrive at the latter conclusion, Sosis referenced studies on small Israeli groups called Kibbutzes. These groups are essentially isolated communities similar to the various utopian movements that sprung up in late 19th century in the United States. The most useful aspect of these groups is their varying level of religious commitment. Sosis most profound contribution to an evolutionary model of religion was the significant correlation he obtained by relating the Kibbutz’s level of commitment to religion and the length of time it managed to exist as a functional society (2009, 4).
Sosis accounts for this observation by proposing that human religion acts a complex signaling system, intricately connected to the Theory of Costly Signaling elsewhere in the animal kingdom (2009, 4). He provides the example of an antelope encountering a lion. Instead of running, it proceeds to leap up and down in full view of the approaching predator, effectively signaling that the lion need not waste its time attempting to capture an antelope so clearly physically fit (2009, 5).
Costly Signaling Theory stipulates that an appropriate signal is one that is too costly to fake, such as that of the leaping antelope, which, if faked, could easily result in a dead antelope. In a similar fashion, Sosis advocates religion as a source of such signaling among humans—a way for members of a group to tell each other exactly how committed they are to the group (2009, 6). Ultimately, Commitment Theory presents religion as a promoter of cooperation (Dow, 2006, 7).
In Religion Explained, Boyer, too, advocates religion as a potential promoter of cooperation. A profound extension of Boyer’s inference system hypothesis is the idea that concepts motivate behavioral change. Consider for example the popular religious notion of a morally-invested all-access agent. Although perhaps the most famous instance of this sort of concept is the Judeo-Christian God, Yahweh, the idea is actually present in a very large proportion of the world, in various forms (Boyer 2002, 165). The importance of such a concept is the way in which it influences behavior. Boyer puts it as follows: “To assume that there is a fully-informed agent around is likely to change my behavior. But then if others assume that there are such agents it will change their behavior too, which is why their representations are of great interest to me.”
But how exactly does such an idea change one’s behavior? Recall that intuitive inferences are invariably made by all humans who encounter any particular concept. Typically, these inferences are bound to be the same in all members of a given population. The most prevalent all-access agents, according to Boyer, appear to be notions of deceased ancestors with whom individuals continue to interact after their death. Boyer explains that when an individual subscribes to the notion that his dead relatives have an interest in the way he lives his life, it is up to him or her to interpret exactly what that means. It isn’t farfetched, nor is it devoid of empirical evidence, that any normal individual will assume that the grandfather watching his behavior would prefer it if he did not steal, murder, lie, etc. Such inferences are intuitive in that they do not need to be spelled out explicitly. Thus in Boyer’s view, religious concepts take the form that they do because it is in that form that they are most intuitively integrated into the human mind (2002, 168). The reason that they have become a ubiquitous component of the human species is then because they represented a form of altruistic behavior in an environment that promoted the selection of such behavior, as per Wilson’s trait-group selection model (Boyer 2002, 169; Wilson 1974, 3).
Morality
If religion is a promoter of morality, what then is morality? It is tempting to make the claim that religion is in fact a source of morality but that is far from the actual state of affairs. Precursors to moral intuitions have been identified in a plethora of other animals (Krebs 2008, 1). Furthermore, despite the varying nature of religious beliefs and the certain isolation experienced by various scattered societies across the globe, moral intuitions appear to share some remarkable similarities across the board (Krebs 2008, 3).
In Moral Minds: How Nature Designed Our Universal Sense of Right and Wrong, evolutionary psychologist Marc D. Hauser sets out to demonstrate the overwhelming amount of evidence suggesting the presence an innate moral disposition in all normally-functional humans. The evidence he uses is primarily cross-cultural analysis to identify underlying similarities and studies on children to indicate the innate presence of a moral system at birth.
Boyer agrees, labeling morality an inference system just like any other. As an agency-detection system picks up relevant environmental cues and transforms them into behaviorally-relevant inferences, so does the moral system identify morally-relevant situations and produce moral dispositions and behavioral inclinations by means of emotional responses (Boyer 2002, 205).
Fully in line with Boyer’s Cognitive Byproduct Theory of religion is the idea that religion does not create morality but rather uses the moral inference system as one of the systems required in its representation. In other words, without the innate human disposition toward moral decision-making, religion would not be able to propose moral instructions because such ideas would have no room for representation within the human mind (Boyer 2002, 181). The brain must possess the capacity to represent a meme in order for that meme to take hold in that brain. To take the chicken and the egg analogy, the egg must have come first because there is no such thing as a chicken that does not start out as an egg. Likewise, there is no such thing as an idea that doesn’t correlate to a mental capacity capable of representing that idea.
Conclusion
I have outlined a theory that I believe is most representative of the current availability of empirical evidence to provide a framework that presents religion as a product of evolution by natural selection. Three types of evolutionary units were described: genes, memes, and trait-groups. In line with multi-level selection theory, these three units experience selective pressure on several related and unrelated levels.
CBT accounts for the evolution of genetic traits that endow individual humans with the capacity to process religious concepts. Differential transmission at this level is unrelated to religion and derives entirely from considerations of mating and survival. Trait-group selection, by such means as Commitment Theory, on the basis of the proportion of religious memes in the group accounts for the mechanism by which such memes grow to dominate entire populations of human beings. CBT in conjunction with meme theory provides the framework in which potential religious memes can originate and take hold by requiring their adherence to unavoidable human dispositions.
This analysis, while extensive, is by no means comprehensive. While Cognitive Byproduct Theory appears to bear the support of a virtual library of findings in cognitive science, Commitment Theory is but a mere suggestion as far as empirical backing is concerned. Ultimately, what this analysis hopes to provide is the motivation to improve upon such theories and to perhaps one day achieve a complete understanding of how evolution by natural selection resulted in a species that almost unanimously subscribes to time and resource-wasting ideas and behaviors based on unfalsifiable principles that seem absolutely absurd to a modern observer equipped with the ability to perform basic logical reasoning.
References (in order of first usage)
- Boyer P. 2002. Religion Explained: The Evolutionary Origins of Religious Thought. 1st edition. New York (NY): Basic Books.
- Dow JW. 2006. The Evolution of Religion: Three Anthropological Approaches. Method & Theory in the Study of Religion 18: 1-23.
- Wade N. 2009. The Faith Instinct: How Religion Evolved & Why it Endures. 1st Edition. New York (NY). The Penguin Press.
- Bulbulia J, Mahoney A. 2008. Religious Solidarity: The Hand Grenade Experiment. Journal of Cognition and Culture 8: 1-25.
- Previc FH. 2005. The Role of the Extrapersonal Brain Systems in Religious Activity. Consciousness and Cognition 15: 1-39.
- Wilson DS. 2002. Darwin’s Cathedral: Evolution, Religion, and the Nature of Society. 2nd Edition. Chicago (IL). The University of Chicago Press.
- Wilkins JS. 1998. What’s in a Meme? Reflections from the Perspective of the History and Philosophy of Evolutionary Biology. Journal of Memetics – Evolutionary Models of Information Transmission 2: 1-30.
- Wilson DS. 1974. A Theory of Group Selection. National Academy of Science 72: 1-4.
- Pyysiainen I, Hauser M. 2009. The Origins of Religion: Evolved Adaptation or By-Product? Trends in Cognitive Science 14: 1-6.
- Sosis R. 2009. The Adaptive Value of Religious Ritual. American Scientist 92. 1-6.
- Krebs DL. 2008. Morality: An Evolutionary Account. Perspectives on Psychological Science 3: 1-23.
- Hauser MD. 2006. Moral Minds: How Nature Designed Our Universal Sense of Right and Wrong. 1st edition. New York (NY). HarperCollins Press.
