I recently found an excellent blog, Homo religiosus — The Natural History of Religion, written by the German scholar Dr. Michael Blume. After I linked to his blog, Michael came over here for some reading. He also had a question, which I answered, and he responded. The issue we are discussing — higher fertility rates linked to religion, is important so rather than have our discussion buried in a comment section, I thought it would be best to bring this discussion to the foreground and let it run here.
Michael: I appreciate your discussions of evolutionary hypotheses of religion very much! But I wonder why I couldn’t find anything about the (on average) higher numbers of children by religionists, even if controlled for education, income and other factors. Religious tend to be fruitfull and multiply throughout the generations – which is a direct fitness advantage.
Cris: My short answer is this: I am generally hesitant to look at modern religions — their forms, beliefs, practices (and fertility rates), and then project those back into evolutionary time. Doing so poses serious methodological problems. Indeed, it is my contention that there was no such thing as “religion,” in any organized or systematic sense, until well after the Neolithic Revolution approximately 12,000 years ago.
The first evidence of organized and systematic religion, in the modern sense, comes from Mesopotamia and Egypt, and thus dates to approximately 5,000 years ago. By this time, the forces of culture appear to have been far more important to humanity than the forces of selection. Thus, while organized religions may have increased fertility rates over the last 5,000 years, I am not certain it had anything to do with fertility rates during the preceding 150,000 years, which is when Homo sapiens first appears in the archaeological record.
Although I dislike the term “cultural evolution” (because I think the concept of “evolution” should be restricted to biology), it seems to me that most analyses of post-Neolithic, historically known, or modern religions are looking at “cultural evolution” rather than biological evolution. Cultural factors such as religion, in other words, were driving fertility rates but not doing much to shape the genomes or phenotypes of humans over the last 5,000 years.
Before there were any modern religions (i.e., before 5,000 years ago), there were shamanisms (the spiritual traditions which everywhere preceded the rise of organized religions). Thus, if I wanted to test the hypothesis that spiritualism in the form of shamanisms affected fertility rates, I would be looking at the reproductive fitness of shamanistic groups.
Because nearly all shamanistic groups over the past 100,000 years were hunter-gatherers, I think you will find that they deliberately depressed fertility rates through various methods, including extended lactation, abortion, and infanticide. All hominids and humans hunted and gathered for nearly 99% of their existence on earth, and whatever their spiritual/religious practices were, one thing is certain: they could not afford many offspring and took great care to limit group size.
This would seem to argue against religion playing much of a role in reproductive fitness, at least until quite recently in human history.
Michael: That’s an interesting perspective! Nevertheless, I would like to point out three main arguments against this sidestep:
1. If an evolutionary function is observable among contemporary humans, one would have to explain why it shouldn’t have been in older times. In fact, Sarah Hrdy emphasized the role of cooperative breeding (related to mythologies and religious “as-if kin”) especially (!) among the cultures of hunters and gatherers! I discussed her convincing stand here.
2. Among contemporary, egalitarian hunters and gatherers without religious institutions etc. as e.g. the !Kung San, the Hadza etc., religious mythologies and commandments “are” emphasizing pro-family values, children etc., as well.
3. A great deal of the available symbols and statues of pre-historic times is dealing with topics of sexuality and especially fertility, e.g. female figurines giving birth. This would hardly have been the case if reproduction had been a no-topic or even a negative one in our ancestor’s mythologies.
Cris: Your points are well taken and I do not want to sidestep them. What I want to do is adhere to chronology and be parsimonious with data. I am on guard, in other words, against drawing unwarranted inferences. Such inferences often arise when we assume that the way something functions today functioned the same way in the past. More fundamentally, the problem is assuming that something as complex as “religion” even existed in the past. The final problem arises from the assumption — pervasive in evolutionary biology and psychology — that the current function of a physical feature or behavioral trait indicates that it was targeted by selection and therefore is adaptive.
In their classic article, “The Spandrels of San Marcos and the Panglossian Paradigm: A Critique of the Adaptationist Programme,” Stephen Jay Gould and Richard Lewontin examined these assumptions and made a compelling case for caution. We should not, in other words, be telling “just so” stories based on current functions that appear to be adaptive in modern settings. From this perspective, religion is what Gould and Lewontin would call a “spandrel.” I would also call religion a spandrel (or byproduct of other neural functions), in part for reasons I discussed in Religion as Evolved Adaptation: The Fallacy of Backwards Projection.
I am familiar with Sarah Hrdy’s work and admire it greatly. On reading your post about her excellent book, Mothers and Others: The Evolutionary Origins of Mutual Understanding, one of the first things I noticed was that you discuss it primarily in the context of modern forms of religion — none of which (except shamanisms) were present in our evolutionary past (i.e., before 5,000 years ago):
Hrdy rightfully observes that religious mythologies is legitimizing distinct family models and organizing cooperative childcare. Not only Mother Church (!) provided assistance and later education to families and children (including the often-quoted “widows and orphans”), as did the Muslim Umma (from arab. umm = mother). Religious personnel is frequently called in the terms of as-if kin, e.g. as Father (Pater), Mother, Brother, Sister, the “Nun” etymological closely related to the “Nanny”. Shamans are communicating with the ancestors, emphasizing familial as well as mythological relationships.
If you look at Europe and beyond, religious communities today are increasingly focussing on their central competences – successful religions bestowing massive reproductive advantages upon their adherents in contrast to the less-reproducing seculars. We don’t have to speculate about the evolutionary mechanism: We are able to observe it, today and worldwide.
I do not deny that historically and ethnographically known shamanisms at times concerned themselves with fertility and reproduction. My reading of the shamanistic record, however, does not indicate that increased fertility and child rearing was a primary preoccupation of shamans and their clients. Although I have read most of the literature on shamanisms, I have yet to encounter a shamanistic practice or tradition that is linked to alloparenting, which is what Sarah Hrdy is discussing.
One last point regarding historical demographics, which admittedly are a bit speculative but are supported by genomic studies. Most historical demographers estimate the world population — just before the Neolithic Revolution and domestication of plants-animals — to be in the range of 3-5 million people (though I have seen even lower estimates, around 1.5 million people worldwide). All these people were hunter-gatherers and for a world-wide population count, these numbers are astonishingly low.
I think it no accident that the first organized and systematic religions appear in conjunction with the Neolithic Revolution and subsequent population explosion. Evolution, however, involves more than just fertility rates or reproductive fitness. Let’s take Judaism as a simple example of a religion that resulted in increased fertility rates. How did this increased fertility affect genomes or phenotypes? Which traits were under selection within Judaic groups? Is there any evidence of a selective sweep — or allele fixation — within Judaic groups?
Michael: Well, if I may condense your answer, it sounds like: I am personally certain that religion is a spandrel and if somebody is demonstrating evolutionary adaptivity in the modern times, I just point out that this is no proof with regard to pre-history. (And in doing that, you just sidestepped those many, clearly fertility-related female figurines – which are widespread and much, much older than 5000 years! 😉 How do these fit into the idea of a non-fertile spandrel?)
In contrast, Darwin defined religiosity in his “Descent of Man” as “belief in supernatural agents” and assumed very early beginnings in brain-based animism (here following Hume’s “Natural History of Religion”). He even did some comparisons with animals, especially dogs. And he assumed that the believed surveillance of supernaturals corresponded to changing behaviors, higher in-group loyalty and, thus, evolutionary success – as do many psychologists and sociologists today.
Just take one of the many examples of Sarah Hrdy: The belief in Chimerism. If people believed that children had more than one father, or if they believed that children were reborn ancestors, higher readiness to alloparental care would be the outcome. In fact, it IS among historic and contemporary hunters and gatherers!
But if we would assume that religiosity started as a “spandrel” (or an exaptation?), we would have to explain how it “got” the reproductive advantages we are observing today. To me, it seems to be more parsimonious to probe whether the observable effects of today could reach into the past than to decree that any contemporary effect should be ruled out from the start for the sake of an indirect route.
Cris: Nice points. I am not a Kantian so do not make a priori assumptions about anything, including spandrels or exaptations. Increased fertility resulting from modern religions does not demonstrate adaptiveness. It simply means that people who practice modern religions have more children. As I stated earlier, evolution is not simply the product of increased fertility.
Evolution, in the classic sense, throws up variations (through random mutations) which can subsequently be targeted by selection, with the result being increased survival/reproduction directly related to the variation that confers a fitness advantage. The usual result of this process is physical and/or behavioral change.
With this in mind, another step must be taken for your analysis to be convincing. Don’t you need to demonstrate that the increased fertility is having genomic or phenotypic effects? Given that culture over the last 5,000 years has substantially altered human mortality profiles, and dramatically diminished mortality rates, selection has been severely muted. I fail to see what effect — if any — all this religiously increased fertility is having on genomes or phenotypes or behaviors. More people means more variation, but when culture shields people from mortality selection, all we get is lots and lots of genetic variation (i.e., genetic noise) with nothing being fixed. And without fixation or a selective sweep, there is no evolution.
Thus, I think the null hypothesis for evolutionary explanations should be that before 10,000 years ago, the primary driver of change in humans was biological evolution and that after 10,000 years ago (post Neolithic Revolution) the primary driver of change in humans is related to cultural forms, which include political systems, economic systems, legal systems, religious systems, technological systems, etc. As is evident from the exponential population explosion caused by the Neolithic Revolution, selection — which at step one is about survival/death — was barely operating. More and more people were surviving, and reproducing, and most of this had to do with one simple fact: the domestication of plants and animals (i.e., agriculture).
I also think it important to rely in the first instance on the data we have from 150,000 years ago to 5,000 years ago. Do we have any evidence of “religion”? No. The venus or fertility figurines to which you point do not amount to religion. Nor do burials or cave art. These scanty items of evidence suggest shamanisms but not religions. And there are fundamental differences between shamanisms on the one hand and religions on the other. Until you can point me to shamanistic practices or traditions that bear directly on fertility and which resulted in increased birthrates, I will have to remain skeptical of the claim.
With respect to your final question — why, if supernaturalism is a spandrel, do we observe increased fertility among modern religious groups? The answer, I think is fairly simple: religions do not exist in a cultural or social or historical vacuum. Most religions encourage their members to marry and make babies for reasons of political, economic, and military security.
Over the last 5,000 years, small groups of people with peculiar religious ideas tend to disappear rather quickly. Such groups tend to be crushed or conquered by larger groups. To avoid this dismal fate, these small groups encourage reproduction. They may make more babies, but I fail to see how this amounts to evolution.