Neandertal humanities

The dead speak to us through their leavings. Material objects left by ancient preliterate peoples—including their bodies—are the only possible carriers of significance to us today.

I study Neandertals, ancient people lived more than 30,000 years ago in Europe and western Asia. I have no Neandertal informants. No matter how often I ask them questions, they will never answer. They left no video, no books, no inscriptions. They left stones and cutmarked bones, a few postholes and pits, bodies and ash.

I seek to understand the lives of these ancient people. I struggle to gain insight into their humanity. Unquestionably they had rich social lives. No social primate can function without emotion, without affiliations or conflicts, without long memories of experiences with other individuals. No primate that depended on hunting and gathering, as all members of our genus have done, could survive without deep social bonds that enabled cooperation and mutual interest.

Yet the Neandertals remain enigmatic. In a few exceptional cases, they carved regular lines upon objects. An optimist can point to a handful of clear indications of intentional behavior, giving the humanist some hope of treating them like texts that can in principle be interpreted. A pessimist replies that whatever “texts” remain are written in an undecipherable script.

What would Neandertal humanities be like? The question is not trivial; it is emblematic of a deep problem in anthropology. Can we understand the subjective lives of other cultures? If so, what methods will bring us toward such a knowledge of ancient people? It may appear that all attempts to build knowledge about the subjective lives of ancient people are necessarily fictional.

Throughout this essay, I consider the problem of Neandertal intentional markings. This is a rich area of inquiry in archaeology, with more and more evidence of Neandertal “symbolic” behavior emerging every year. As much as we might wish it, intentional markings do not provide “windows” into the minds of ancient people. As described below, some past efforts to explain such markings cannot be distinguished from fiction. Yet, if the subjective lives of ancient people must forever remain unknown to us, then surely our attempts to interpret the lives of our contemporaries must be equally fictional. Consilience, as a general attitude about the ability of scientific knowledge to cross into humanistic realms, holds out the promise that we might study and come to an understanding of the humanity of ancient people, not only as social primates but also as subjects of experience. But I have little confidence that science can provide such knowledge, or that non-science interpretive approaches can create knowledge distinguishable from fiction.


Anthropology is a diverse field. Biological anthropologists, like me, study the human organism and its evolution. Like all organisms, humans exhibit behaviors that may adapt them to their environments. For humans, the social environment is a primary determinant of the adaptive constraints shaping behavior. Our social lives are a niche that we continually construct, our conscious selves are tools for our nervous system to navigate this social environment. Subjective experience, in this way of thinking, is fundamental to human biology. Without it, we could not function in our social environments.

Many anthropologists study human behavior not from a scientific perspective, but within an interpretive frame. The interpretive anthropologist recognizes that a person’s behaviors may not be fully translatable into different cultural contexts. Indeed, even interpretation within the subject’s own cultural context may be unsatisfying. Identifying the proximate or ultimate causes of behavior or belief is not the goal of interpretive theory; the interpretive anthropologist may not even find interesting such reduction of subjective experience to physical causes. Theory, to the interpretive anthropologist, is a productive and creative activity, a synthesis of observation with the interpretive frame. By observing the subjective experience of another person and her own encounter with the literature, the interpretive anthropologist builds something new.

I consider myself a scientist. Some interpretive anthropologists reject scientific approaches and methods, describing them at best as irrelevant to understanding subjective experience; at worst weapons of hegemony.

I consider myself a humanist. In academic discourse within anthropology, the interpretive approach is aligned with humanistic inquiry. My work is starkly different in its goals and methods from the interpretive anthropologist. Like many of my colleagues I pick my way through this minefield. Subjective human experience must have an evolutionary origin, which we seek to understand and discover.


Obviously, the broad concept of “consilience” is relevant to any interdisciplinary inquiry. Whewell’s “consilience of induction” Whewell:1840 is an important mode of argument in biological anthropology, just as in evolutionary biology Gould:structure:2002. Whewell held that consilience of inductions from two classes of facts “is a test of the truth of the Theory in which it occurs”. E. O. Wilson’s argument for consilience of different fields of knowledge Wilson:Daedalus:1998 is rather grander in scope than Whewell’s conception. From his 1998 essay on the subject (p. 133):

Consilience proved to be the light and way of the natural sciences. Physics, with its astonishing congruity to mathematics, came to undergird chemistry, which in turn proved foundational for biology. The successful union was not just a broad theoretical consistency, as articulated by Whewell, but an exact unfolding of principles pertaining to more complex and particular systems into the principles for simpler and more general systems.

Wilson discarded Whewell’s reference to induction, and argued that the fact of consilience is a logical consequence of the reduction of chemical and biological theories to physical theories. Consilience was itself, in Wilson’s description, substantially easier to establish than reduction. A scientist can observe that two statements at different levels of physical complexity are consilient without being immediately able to reduce the complex system to a simpler one. For example, the ratio of stable carbon isotopes in a mammal’s tooth enamel reflects the photosynthetic pathway of plants eaten by the individual during the period of enamel formation. Eaters of grasses with a four-carbon photosynthetic pathway have enamel with a higher proportion of carbon-13 than do eaters of fruit and leaves from trees with a three-carbon photosynthetic pathway. This hypothesis relies on consilience of theories underlying the physical properties of carbon isotopes, the chemistry of photosynthesis in plants, and the biology of enamel formation in mammals. Yet we cannot presently reduce the embryonic development of ameloblasts (enamel-forming cells) to physical principles. Nor can we synthesize, or build up, the actions of ameloblasts from chemical or physical theories. In practice we make a much weaker assumption: that these cells will not betray well-known principles or act according to their own idiosyncratic physics. Discussing reduction and synthesis, Wilson admitted the difficulty (and in some cases, impossibility) of building predictive hypotheses about complex systems from simple physical principles. He referred explicitly to the “paradox of emergence” which necessitates that levels of organization have their own proper explanatory theories.

Our attempts to understand retrospectively the results of complex interactions may demonstrate the fact of consilience among explanatory theories that apply to different levels of interaction. Does the fact of consilience prove that the explanatory theories at different levels reduce to a common set of simpler (presumably physical) principles? To the extent that we believe in a single material reality, we must conclude Wilson is correct. When we have theories at different levels that are more-or-less true, they should give rise to consilient inductions. If theories at different levels give rise to different predictions about the same physical event, they are by definition incompatible. If theories always give rise to the same prediction about every possible event, they are by definition identical. Consilience is no proof of reducibility, but it may serve as a sign of reducibility.


But what if our theories are not close to the truth?

The question may seem impish: After all, if our theories are so wrong, why would we bother? But the idea that similarity of predictions is an indication of similarity of theories, which seems plausible for the recent history of physical sciences, simply doesn’t work in the social sciences and humanities. Replacing Newton’s theory of gravitation with Einstein’s was a tiny change for most easily observed physical systems, so tiny that we still teach Newton’s theory to our children as correct. The history of the social sciences is a march through dozens of all-encompassing social theories with barely any attempt to falsify or test for their mutual incompatibility. Today’s interpretive anthropologists flock among a confusing array of social theorists, each with different, partly incommensurable schemes of social explanation.

Consider for example the phenomenon of religious ritual. Counting only the last two hundred years, theories of the social and psychological causes of religious ritual have been promulgated by dozens of major scholars, including Tylor, Spencer, Freud, William James, Durkheim, Boas, Campbell, and our contemporaries Pascal Boyer and Scott Atran. Most among this group have attempted comparative procedures similar in form to the consilience of inductions. The subject of religion almost demands such an approach, as it touches on so many aspects of human interactions, and is understood through different religious traditions in different cultural groups.

An example close to the Neandertals is provided by George Barton, who argued that religious ritual remains as a legacy of prehistoric peoples’ awe of the divine feminine Barton:religion:1940.

Palolithic religion was, then, sex-mysticism. The psychologic unity of the race made it universal as its survivals in the historic period prove. This is the real origin of religion. It was not begotten by fear (Lucretius), nor by animism (Tylor), nor by ancestor worship (Herbert Spencer), nor by the mysterium tremendum (Otto), but by the mysterium feminium—a mysterium tremendum indeed, but scarcely that which Otto contemplated.

Barton turned to the study of myth and religion in small-scale societies, as well as the mortuary practices of Paleolithic people, to make the argument that ritual had its origin in the worship of females:

Dr. Lewis has pointed out that in the case of such burials the body of the dead is covered with red earth and often either rests upon or is partly covered with the shells of various kinds of shellfish. He rightly infers that the image represented the mother-goddess— the goddess of life—that the red earth represented blood, the vehicle of life, and that the shells were emblems of the mother goddess because, like her, they contained life or produced a living creature. Barton:religion:1940

Neanderthals never to our knowledge painted figures or other representations on the walls of caves. Nor did they accompany burials with pigments as did some later European peoples. But some red marks on the walls of caves are now known to be contemporary with late Neanderthals, raising the possibility that they were the authors of this behavior. Pigment crayons, blocks of mineral pigment that have been worn by rubbing on some surface, have long been known from Neanderthal sites. A recent forensic analysis of such crayons has shown that some of them were rubbed on a soft surface, such as skin or hide Soressi:dErrico:2007. At two sites, small cup shaped pieces of stony stalagmite have been found with traces of mineral pigment inside them. The Neanderthals appear to have been painting, or mixing mineral pigment for some other purpose. Evidence of shellfish exploitation is rarely found that Neanderthal sites, possibly because the lower sea level during Neanderthal times meant that shorelines were far from today’s sites in most of Europe. Nevertheless a handful of sites show that Neanderthals were using shellfish and other marine resources Barton:Gibraltar:2000, and at one site Neanderthals were gathering empty shells with natural holes, apparently stringing them together. One Iberian site preserves an exceptional shell bearing a natural band of pigment on one side and on the other, painted by a Neanderthal, a red ochre band mirroring the opposite side Zilhao:shells:2010.

With this kind of evidence emerging from the archaeological record of Neandertals, perhaps we can test such hypotheses as Barton’s for the origin of religious ritual. Barton’s work does not stand alone: Several scholars have pursued similar models for the origin of religion, relating red pigment, female symbolism, and the origin of religious ritual or broader social interactions (e.g., Wreschner:1980, Knight:1995). Menstruation in this view ties blood with the social power to govern sexual relations, as women in a number of societies undergo an involuntary seclusion at the time of menstruation. Fertility is a mysterious and powerful process within many societies, and menstruation visibly connects fertility with blood. A connection of religious ritual and the sacred feminine reads like the Da Vinci Code of Paleolithic archaeology, depending on observations drawn from ethnography, folklore, political economy, art, and Paleolithic archaeology. In Knight’s model, as an example, the appearance of complex social interactions, symbolic behavior and language is explained by a sexual revolution in which women gained meat and protection from men by controlling sexual access Knight:1995. Knight and others have tied these phenomena to ethnographic and economic observations among African and Australian small-scale societies, tying them further to the iconography of rock art.

What should we make of this seemingly plausible hypothesis of a causal connection of religious ritual, menstruation, motherhood, shells, the color red? I suggest that the idea is an “umbrella hypothesis” Langdon:1997, an attempt to explain a large set of true or uncontroversial propositions as results of a single underlying cause. An umbrella hypothesis is not necessarily false, or even problematic. But the strength of support of the hypothesis is not a function of the facts adduced in its favor. Indeed, the hypothesis does not necessitate that any of these propositions be true; if one were false, some other proposition might otherwise have been found as support for the hypothesis. An umbrella hypothesis subtends an internally consistent model of causal relations among facts.

How can we explain the association of red pigment with burials and Paleolithic art without a causal relation among menstruation, religious ritual and art? Red ochre, the second most common pigment in archaeological sites after black, is much easier to obtain and survives better in archaeological contexts than most other color pigments. Likewise shells are archaeologically durable and widely collected and traded in nearly all preindustrial societies. Symbolic association of the color red with blood is ubiquitous, even inevitable. But other correlates of red vary widely across cultures, in some generally positive, others negative, some masculine and others feminine Wreschner:1980. To be sure, some traditional African groups have placed enormous social and economic importance on red cosmetics, associating them with menstruation and status. These elaborate traditions leave heavy material traces, from the acquisition, distribution and use of red pigments, and not a trace of such elaborate material traditions can be found in the Paleolithic record. Instead, we have traces of red ochre use, scattered across much of the Old World during the last 300,000 years, with a handful of co-occurrences with shells.

We are not looking at a consilience of red ochre burials, shells, and mother-goddess ideation, we are looking at a coincidence of these things in some material and cultural contexts.

From a purely interpretive perspective, it may not matter that the coincidence is accidental. The strong association of menstruation and red pigment in the ideological systems of some societies is a fact that has braced the social lives of thousands of people, across hundreds of years. People’s subjective experiences within such cultural contexts have included these social facts, by which people have navigated their social lives. Interpretation of these facts is a form of play, sometimes reversing hypotheses of cause, connecting them to the work of social theorists, reading them as if they were a text through an interpretive frame. But to apply this frame outside the context where it is observed — to extend it to Neandertals or other Paleolithic peoples — is fiction.


The scientist wants more than to interpret the associations where they occur; she wants to predict, to deduce, to infer unknowns. And for this, she needs far more than the observation of their occurrence together. She must show the associated facts to be causally connected. A statistical association between the elements of the hypothesis, the red pigment, shells, symbols of the feminine, and rituals related to menstruation in recent human cultures might test the hypothesis of causal connection among them. A consistent record of ideational connection from informant interviews might also contribute to such a test, or a deeper record of archaeological shell use by Paleolithic people. Such evidence helps to substantiate that the inference as applied to ancient humans may be more than fiction. Such evidence drawn from several different lines of inquiry might contribute to a consilience of inductions. Because such an approach must be tied to more lines of evidence, we must necessarily proceed more slowly than the purely interpretive anthropologist. We must also spend more time on each line of evidence establishing the direction and strength of causal relations among observations.

More than 20 years ago, most archaeologists held that only modern human groups show clear evidence of intentionally marking objects for the purpose of communicating. The upper Paleolithic record of artistic expression is vast and unparalleled by anything that came before it. When a visitor today walks through a cave with art painted on the walls, she experiences a reality of iconic expression that simply did not exist anywhere in the world before 50,000 years ago. Painted art, sculpture, musical instruments, and other nonvocal modes of communication appeared in the archaeological record of Europe, Australia, and Southern Africa within the span from 80,000 to 30,000 years ago. Some archaeologists have dubbed this phenomenon the human revolution, noting that provides the first evidence that human minds had become truly subjects of experience Mellars:Stringer:1989.

This view has become increasingly problematic as evidence of the capabilities of Neanderthals and other archaic humans has grown dErrico:2003. As noted above, the exceptional evidence of intentional marking by Neandertals contributes to this record. More evidence comes from the growing record of behavioral sophistication among African peoples of the Middle Stone Age period, earlier than 70,000 years ago. At several sites, including some relatively far from the seashore, pierced shells have been found Bouzouggar:2007. Ostrich eggshells were incised with geometric patterns in southern Africa as early as 80,000 years ago Texier:Diepkloof:2010. A tradition of pigment use, including the complex preparation of mixtures, is of equivalent age in southern Africa as in Europe, in both cases dating long before the origin of modern human populations.

Faced with these facts, archaeologists have problematized the origin of marking behavior. At one extreme some suggest that we should maintain the null hypothesis that archaic people had cognitive capabilities identical to those of living humans. From this point of view evidence of marking, pigment use, or the accumulation of nonutilitarian objects such as shells, supports the interpretation that the symbolic communication abilities of humans extend far back into the Pleistocene. At the opposite end of the scale, we can observe that many of the archaeological cultures of the last 10,000 years failed to show systematic evidence of marking, gathering of exotic objects, or other archaeological indicators of nonvocal communication Speth:2004. The lack of such evidence does not indicate that these recent peoples were incapable of such activities; it merely reflects an ecological or social context in which people did not value the production of such material objects.

Archaeologists have begun to consider models for the gradual emergence of such cognitive abilities in archaic humans. Within Africa the sophistication and standardization of stone tool technology increased across the last 300,000 years. The appearance of small projectile points, long-distance movement of raw material, and increasing pigment use during the last 100,000 years seems a logical extension of a much longer-term process Mcbrearty:Brooks:2000. Likewise European Neanderthal populations show a gradual increase in sophistication of tool raw material acquisition, stronger evidence of a diversity of material culture, and pigment used in marking behavior over the same time period. Just as genetic evidence has made the genealogical connection between living people and archaic people more complex, so the archaeological record has made the interpretation of behavioral capabilities in these ancient people more complex.

Our growing knowledge of Neandertal marking behavior shares little with the interpretive approaches of cultural anthropology. We see only hints and traces of the behavior of a few Neanderthals at a handful of instants in their lives. With the cave art of the Upper Paleolithic, the art historian can use interpretive methods to add value to our understanding. She may not be able to read the minds of the artists, but she can arrive at an understanding of their stylistic processes, the relation of representation to their lives, and the picture of their social existence. The archaeologist Dale Guthrie brings a contextual understanding to Paleolithic art built upon his experience as a hunter and his consideration of Paleolithic groups Guthrie:2006. The consistent representation of what he calls high testosterone activities helps to place the role of the artist within their hunting society. The representations of animals attend to small details of their biology, such as rutting male deer, animals at the peak fat distribution of autumn, and prowling carnivores, testify to the intimate ecological knowledge of the artists. Meanwhile, the distribution of handprint sizes shows that children were often among the artists, an inference supported by the relatively crude form of a large majority of representations seldom illustrated in books about cave art. The understanding attained by the art historian about these ancient people is not dialogic, not negotiated, nor is it based directly on testimony. But it is undeniably humanistic.

With Neanderthals and their contemporaries we can approach their humanity only through a fog. Drips of red ocher in the dust of an ancient site Roebroeks:ochre:2012 do not speak of their maker. They do not even have a maker in the sense of an author of intention behind their production. Repeated marks upon a bone may reflect intention, but ascribing intention requires ruling out alternatives that may in practical terms be impossible to exclude for a singular object.

Still, our inability to arrive at a symbolic or iconic understanding of Neanderthal production does not preclude a truly humanistic understanding of them. A humanistic interpretation of the cave art of the Upper Paleolithic cave art depends on much more than the content of the images themselves. Such an understanding can be built by juxtaposing the content of the art with a sophisticated knowledge of hunting practice tied to a depth of ethnographic knowledge about human foraging and social interactions in small groups. When we have a blind spot of interpretation, it is because we assume too much, not because the paintings tell us too little. In other words humanistic understanding in this applied the context does depend on a consilience of inductions in the Whewellian sense.


The remains of the Neandertals and other ancient peoples do speak to us. We can develop some knowledge about the subjective lives of these people. But this knowledge is necessarily limited, and can be attained only very slowly.

Anthropologists approach the problem of understanding subjective experience in two ways. Scientific anthropologists approach the problem with great caution, proceeding slowly and by degrees building understanding of small parts of the overall picture. Interpretive anthropologists implicitly exercise even greater caution, by assuming that subjective experience may never be comprehensible even in principle. They therefore focus on the aesthetics of an improvised encounter between observation and theory.

I do not deny that the interpretive approach can yield interesting and compelling stories. The methods and style of humanistic and interpretive anthropology are not sterile; they actively bring new perspectives on the observations made by ethnographers and other cultural anthropologists. But it is not in the nature of such interpretive methods to comport with the requirements of hypothesis testing. I assert that these approaches really cannot be reconciled, in the Wilson’s sense of consilience. The aims of interpretive anthropology are productive and aesthetic in nature. This is not a misunderstanding within anthropology, it is a real difference of priorities. Interpretive anthropology cannot be reduced, even in principle, to scientific principles of biology or psychology.

Sometimes I despair that the noise of exuberant interpretive work may drown out the quiet signal as we slowly build up multiple lines of evidence about the lives of ancient people. We are building a humanistic understanding of Neandertals and other ancient people. We may never know the songs of Neandertals, but we will someday know whether any songs could have existed. We may never find a Neandertal portrait, but we will assemble an understanding of how pigment marking once fit within now-extinct social systems. It diminishes the humanities to restrict them to mere interpretation.