Review: Supernatural Selection

How religion evolved

Books about human evolution often follow the format:

1. Intro explaining there are widespread misconceptions about what makes us human and how evolution could have selected for that trait (be it religion, language, the tendency to cooperate, or symbolic thought).

2. The bulk of the book meticulously defends the importance of the trait and the evolutionary steps that could have produced it.

3. A couple paragraphs on how the first evidence for the trait is 40,000 years ago in Europe, but if you squint there are kind of glimmers 70,000 years ago in Africa, before the Out of Africa migration. That’s a cleaner story; let’s go with that.

Written in 2010, Supernatural Selection by evolutionary psychologist Matt Rossano is a good example:

The Social Transformation of the Upper Paleolithic

Why are we (Homo sapiens sapiens) the only hominin species left on earth? Two hundred thousand years ago, when anatomically modern humans first emerged, there were at least four other hominin species around. Additionally, for the first 100,000 years or so of our ancestors’ existence, the archaeological record reveals nothing to distinguish them from other hominins, and certainly nothing to predict their eventual dominance.

For many decades, most archaeologists and anthropologists believed that the advanced cognition that gave Homo sapiens a decisive advantage over other hominins emerged suddenly in what was called an “Upper Paleolithic Revolution.” Around 40,000 ybp, the European archaeological record erupts with a trove of remains indicating a quantum leap in thought and behavior: sophisticated tools, stunning cave art, ceremonial weapons, abstract figurines, evidence of food storage, and elaborate burial sites with symbolic grave offerings.

Increased social complexity was a key component of this revolution. Cro-Magnon (our forerunners in Europe) settlement sites from the Upper Paleolithic (UP) are generally larger and more spatially structured than those of Neanderthals. The larger sites could indicate either more people at a particular site or greater seasonal aggregation of otherwise dispersed groups or both. In any case, it is an important social change from earlier periods. Cro-Magnons also participated in far more long-distance trading networks than Neanderthals. One recent study found that Cro-Magnons and Neanderthals differed only marginally in their hunting tactics and success rates. Instead, what gave modern humans an advantage was not technological or strategic but social. Modern humans were able to coordinate their foraging activity over larger ranges, and they took advantage of extensive, far-flung trading networks.

The UP also offers the first evidence of elaborate burial sites with abundant grave goods. These burials show that UP societies were becoming increasingly stratified. Obviously, not everyone was buried with such ceremony. We can surmise that those offered such an elaborate send-off were the elites of what was no longer a purely egalitarian hunter-gatherer world. Relatedly, it is at this time that prestige objects become regular features of the archaeological record. Such objects typically denote status in traditional societies. By the time of the Upper Paleolithic, Homo sapiens sapiens had become the most socially sophisticated species on earth. Other hominins simply couldn’t compete.

The idea of an Upper Paleolithic revolution implies that it took about 100,000 years for the human mind to catch up with the human body, and that it did so only after migrating from Africa to Europe. Recently, however, this idea has been challenged by those who believe that modern thought and behavior emerged gradually and that the process began in Africa. These researchers point out that some forms of sophisticated tools and behavior such as blade production, seasonal mobility, and the use of grindstones and barbed points can be found in the African archaeological record as far back as 100,000 ybp, possibly even earlier. Even more compelling are symbolic artifacts such as the intentionally inscribed red ochre plaques from Blombos Cave in South Africa (believed to be over 75,000 years old) and the perforated shell beads apparently used as personal ornamentation (between 130,000 and 70,000 years old). The aforementioned long-distance trading networks probably first began in Africa as well. The shell beads and certain tools (such as those of the Howiesons Poort industry) are made of raw materials that are not indigenous to the locations where they have been found.

Genetic studies complement these archaeological findings. They show a dramatic population bottleneck followed by a relative population explosion among particular human groups in Africa sometime between 80,000 and 60,000 ybp. It was this subpopulation of Homo sapiens that eventually trekked out of Africa and conquered the world. A probable scenario is that cycles of drought (possibly accentuated by the Mt. Toba eruption) brought African Homo sapiens to near extinction. From a remnant of only about 2,000 breeding individuals, a small group of technologically and socially sophisticated humans began a precipitous expansion that eventually engulfed all of Africa and then the world. Central to their social sophistication was religion. At the same time of the worldwide spread of modern humans we see the first compelling evidence for the religious practices of shamanism, animism, and ancestor worship. Echoing the famous anthropologist Roy Rappaport, my view is that this is more than mere accident. Religion played a nontrivial role in the achievement of distinctively human society.

The bolded text glosses over a 25,000 year discrepancy; the compelling evidence for shamanism he refers to is seen 40,000 years ago in Europe, but Australia was settled as early as 65,000 years ago. In many parts of the world shamanism isn’t indicated until much more recently. In the Americas, not so much as a bead is dated before the Clovis culture about 13,000 years ago, despite being occupied for 10,000 years before that (and likely much longer). Or consider Australia. The iconic rock art associated with Dreamtime myths was not developed until the Holocene; the Rainbow Serpent, now worshipped across the continent, first appeared 6,000 years ago. Worldwide, shamanism was not the norm 20,000 years ago. If shamanism spread with those who left Africa, why is the coverage so spotty even tens of thousands of years later? Archeologists write books like The Birth of the Gods and the Origins of Agriculture, because it looks like the animistic gods took form just before the Agricultural Revolution. Others go as far as to argue abstract thought and recursive grammar did not develop until about 15,000 years ago¹.

At the end of the book, he summarizes the evolution of religion. Starting 500,000 years ago, our ancestors bonded by singing and dancing, producing ecstatic states. The next stage takes place among Homo sapiens just before Out of Africa.

During the African Interregnum (roughly 90,000–60,000 ybp), ecological degradation associated with rapid climate changes and possibly the massive Toba eruption forced our ancestors to undergo a social revolution. They formed increasingly larger, more complex social groups and established unprecedented intergroup trading alliances. The singing and dancing rituals of their predecessors were not enough to handle the demands of this more complicated social world.

Our ancestors expanded their social life to include rituals of initiation, trust-building, reconciliation, and shamanistic healing. These rituals taxed attention and working memory, making greater working memory capacity advantageous for survival and reproduction. This set the stage for the emergence of symbolic thinking and other uniquely human forms of cognition.

The major evidence for [this stage] is threefold. (1) Ethnographic and comparative evidence indicate that rituals of social bonding are extensive among traditional societies and widespread in the animal kingdom, including among our primate relatives. (2) The first archaeological evidence of expanded trading networks emerges at this time. (3) Neuroscience evidence indicates that ritual behavior requiring focused attention and the inhibition of prepotent responses activates brain areas essential to working memory.

The book argues that religion evolved just before Homo sapiens left Africa and that it was a crucial element of our dominance thereafter. But it can only point to expanding trade networks, ethnographic studies of modern humans and primates, and neuroscience in modern humans. Notably absent is any archeological indication of shamanic healing or rites of initiation. The final stage—the first with any direct evidence of religion—is given a negative spin:

Somewhere between 50,000 and 30,000 ybp, ancestor worship emerged, and, with it, the first religious myths and narratives. These narratives justified the inequities of increasingly stratified societies and laid the groundwork for the “elite versus popular” religious distinction of classic paganism.

The major evidence for [this stage] is archaeological. Innovations present with the onset of the Upper Paleolithic, such as sophisticated tools, cave art, food storage pits, time-keeping devices, and abstract figurines, indicate that full-blown language and episodic memory were present by this time. Elaborate burials and artifacts reflecting fertility concerns emerge predominantly during the Upper Paleolithic, indicating that ancestor worship was present. Ancestor worship and the social inequities it entails require compelling myths and narratives to justify and sustain them.

Episodic memory is the ability to recall past or imagine future “episodes” in your life. Artifacts that indicate such an ability—or full-blown language, ancestor worship—are not found globally by 50,000-30,000 ybp. Those dates refer to the posited Upper Paleolithic Revolution, which holds that something neurological changed then. The yawning hole in that model is the evidence for Behavioral Modernity is primarily found in Europe. This is pointed out in the Sapient Paradox, which notes the conundrum of religion being absent in much of the world until about 10,000 ypb:

“Once again, such evidence for religious practices is first seen, along various trajectories of development in different regions, at about the time of the sedentary revolution [10,000 ybp]. (The specially early case of the European Upper Palaeolithic should again be recognized, with its cave art and its figurines, and might indeed be interpreted as an exceptionally early case of incipient sedentism.)” ~Neuroscience, evolution and the sapient paradox: the factuality of value and of the sacred, Colin Renfrew, 2008

The most common rejoinder is to posit that full religion, language, and episodic memory may have existed much earlier. Some ochre deposits date back half a million years have been found. Maybe it was used to decorate bodies and perform rituals, the same as happens worldwide now. But Rossano doesn’t do this. He accepts language, episodic memory, and ancestor worship emerged 50,000-30,000 ybp, but fails to add the caveat that the evidence for those behaviors is not worldwide until 10,000 ybp. Remember, Rossano is trying to tell the story of the evolution of religion, but these years are an afterthought. If a crude form of religion evolved from 90,000-60,000 ypb, did evolution stop then? Did episodic memory and full language qualitatively change the type of religion that could exist afterward? Is it fair to call pre-linguistic religion religion at all? Did the myths and symbolic thought that became part of religion 50,000-30,000 ypb exert evolutionary pressure? Did they spread? That would be an interesting story!

Apart from the confused timeline, this book is not helping evolutionary psychology avoid charges of just-so stories. Notice that he casually asserts myths were invented to justify income inequality. Yes, 40,000 years ago, coinciding with the first flowering of full-blown language, the pressing rhetorical concern was the same as the shiny social issue of the 2010s when Supernatural Selection was written. On what evidence? Do our oldest myths function primarily to explain social inequality? One contender for the oldest myth, likely stretching back 40,000 years, are creation myths. These explain the human condition, of which relative social status is only a footnote. Creation stories are primarily concerned with why there is something rather than nothing, how humans are different from other animals, and the nature of the soul. There is a vast literature connecting self-awareness to the language faculty. Indeed, the only tribe in the world without recursive (full) language also seems to have only a limited narrative sense of self, as you’d expect without modern episodic memory². The genesis of myths could just as well have to do with the emergence of self as income inequality. This is even in accord with the myths passed down from the Ice Age.

I hesitate to call this post a review, as it’s more of a narrow disagreement about timelines. Rossano tries to have it both ways with the Upper Paleolithic. He admits it offers the first direct evidence for religion, language, and episodic memory but treats these developments as an afterthought. What gives? Overall, though, the book is a fine overview of the evo-psych case for religion, which Rossano has spent years thinking about. He’s a psychologist, not an archeologist, so it makes sense that he would fix his model to the timeline the fewest will complain about. This is why it’s so difficult to answer who we are and where we came from. Those trying to change consensus usually need their feet in at least two disciplines to articulate discrepancies and move the needle. The conceit of this blog is that comparative mythology should be added to the usual mix of linguistics, archeology, genetics, and psychology. If evolutionary timelines are measured in just tens of thousands of years, then myths may be informative.

Supernatural Selection, Andrew and Dall-E

1. See the intro of the Snake Cult piece for a collection of a few such claims.

2. “Discussions about the role of language in shaping memory (and cognition in general) have a long history. Tulving (2005) asserted that language is not a necessary condition for EAM, but acknowledged that it supports and enriches its development, while others proposed a stronger connection between language and EAM (Nelson & Fivush, 2004; Suddendorf, Addis, et al., 2009). In the last few years it was found that a tribe of Amazon Indians – the Pirahã – lacks many of those attributes which are commonsense for our understanding of the selfhood and existence across time. They do not think of past and future and consequently cannot imagine the life of persons from history or past (Everett, 2005, 2008). The language spoken by the Pirahã reportedly only comprises two rudimentary temporal markers (Everett, 2005; Suddendorf, Addis, et al., 2009).” Memory, autonoetic consciousness, and the self, Markowitsch and Staniloiu, 2011