Why Darwinians Should Not Be Afraid of Mary Douglas—And Vice Versa

Introduction

Evolutionary psychology and human behavioral ecology have become mainstream academic approaches to human behavior and thinking. Yet, there is a widely shared feeling among evolutionary psychologists that the acceptance of the Darwinian approach is only the first step to be taken. The real goal is to replace completely what Tooby and Cosmides (1992) notoriously have dubbed, the Standard Social Science Model (SSSM). This SSSM holds that the human mind is a blank slate, imprinted solely (or almost solely) by culture.

According to Tooby and Cosmides, this model really was the standard model in the social sciences for much of the twentieth century. Others have pointed out correctly that this is a gross misrepresentation (Levy 2004): this social science model never was the standard in the social sciences. It is nonetheless hard to deny that some important social scientists—“culturalists”—did adhere to the so-called SSSM view, or, if the label is inadequate, to a view that stresses the “autonomy” of cultural and symbolic systems. Mary Douglas, for instance, seems to have been an adherent of the SSSM, if there ever was one. Being one of Durkheim’s followers, Douglas asserted the autonomy of the social realm. According to her, human behavior and thinking are almost completely shaped by systems of significant symbols. It is this kind of symbolic causality that is unacceptable for Tooby, Cosmides, and many other naturalistically inclined theorists. In comparison with unproblematic event-causation in nature, symbolic causation in a certain culture is a direct relation of dependency between autonomous cultural factors, such as categorical differences, and the psychological, and even neurophysiological, features of members belonging to that culture—for example, the way in which the category of impure food directly influences cognitive-affective and behavioral responses to food consumption. However, we will argue in this article that the symbolic causality, defended by Douglas, can be reconciled with—and even integrated in—a broadly conceived evolutionary framework for the social sciences.

Our “integrated hypothesis” takes a specific position as to the issue of how culture and social knowledge relate to human evolution and neurobiology. Consider the following three claims:

By and large, the SSSM defends the first claim, whereas evolutionary psychology endorses the third claim. In line with gene-culture co-evolutionary theory, our integrated hypothesis supports the second claim. We argue for this middle position in the case of disgust, and claim that this middle position is able to integrate Douglas’s SSSM view with the evolutionary psychology of disgust.

In her famous study Purity and Danger: An Analysis of Concepts of Pollution and Taboo, Douglas ([1966] 2002) developed, among other things, the idea that social organization and categorization produce strong disgust responses. Disgust is significantly influenced by culture, and the cultural influence on disgust originates in a symbolic framework. This claim is at odds with the “narrow evolutionary psychology” approach advocated by Tooby and Cosmides (Laland and Brown 2002). However, if we are right, Douglas’s theory may represent an important building block for gene-culture co-evolutionary theories and Darwinian theories of cultural evolution.

In this article, we will show how Douglas’s SSSM view of disgust can be integrated within a broadly conceived evolutionary theory against the backdrop of the problems of narrow evolutionary psychology to fully explain a complex human emotion like disgust. The structure of the article is as follows. Section 1 briefly outlines why the rejection of the SSSM by Tooby and Cosmides is ontologically unwarranted. In section 2 we then sketch how disgust is explained by “narrow” evolutionary psychologists. Section 3 discusses Douglas’s view of disgust. In this section we also examine how more recent and quite sophisticated accounts of disgust, like the “animal reminder” view developed by Rozin, Haidt, and their colleagues, still fall short in bridging the gap between Douglas’s account and evolutionary psychology. Sections 4 and 5 attempt to explicate why and how Douglas’s account of disgust can and should be given a place in any evolutionary theory of cultural disgust. We conclude in section 6 with a comparison of our own integrated theory with the animal reminder view of disgust.

1. Ontological Considerations

Tooby and Cosmides argue for the replacement of SSSM by what they call the Integrated Causal Model:

What should be jettisoned is what we will call the Standard Social Science Model (SSSM): The consensus view of the nature of social and cultural phenomena that has served for a century as the intellectual framework for the organization of psychology and the social sciences and the intellectual justification for their claims of autonomy from the rest of science. . . . In place of the Standard Social Science Model, there is emerging a new framework that we will call the Integrated Causal Model [ICM]. This alternative framework makes progress possible by accepting and exploiting the natural connections that exist among all the branches of science, . . . nothing is autonomous and all the components of the model must mesh. (Tooby and Cosmides 1992, 23)

Although they acknowledge in their criticism that the SSSM includes many important elements of truth and admit that their picture of the SSSM is somewhat synoptic, we think that, pace the caveats, their view is too myopic, and mistaken by not being fine-grained enough. We agree that SSSM exaggerates as to the autonomous status of sociocultural explanation. SSSM’s major defect is its failure to appreciate the role that evolutionary biology and evolutionary theory can play in the explanation of sociocultural phenomena such as disgust. However, to jettison the SSSM outright seems to us to throw away the baby with the bathwater. For our purposes, an ontological question about the nature of social reality is crucial.

Evolutionary psychologists typically object that accepting the SSSM would force us to include in our ontology an autonomous social reality. Also Tooby and Cosmides clearly have qualms about this ontological issue in SSSM’s logic and view of culture:

The features of a particular culture are the result of emergent group-level processes, whose determinants arise at the group level and whose outcome is not given specific shape or content by human biology, human nature, or any inherited psychological design. These emergent processes, operating at the sociocultural level, are the ultimate generator of the significant organization, both mental and social, that is found in human affairs. (Tooby and Cosmides 1992, 32)

The worry is twofold: first, that the emergent sociocultural level is autonomous, entirely distinct from the biological level and, second, that the sociocultural factors “self-cause” other such factors and asymmetrically “downwardly cause” psychological factors. Although evolutionary psychologists do not flatly deny the influence of culture on evolution and human behavior, they cannot accept such a strong autonomy and such a direct causal impact in a short span of time accorded to social reality by the SSSM. In contrast, evolutionary psychologists all accept Edward O. Wilson’s leash-principle: “The genes hold culture on a leash. The leash is very long, but inevitably values will be constrained in accordance with their effects on the human gene pool. The brain is a product of evolution” (Wilson 1978, 167). In defence of the SSSM, we may now ask the question: Might social reality not be autonomous in some sense—compatible with the legitimate claims of evolutionary psychology?

The leash-principle of the ICM is usually interpreted as a reductionist principle (Dupré 2001): both culture and psychology reduce to evolutionary biology (and genetics). Correspondingly, the “unleash-principle” of the SSSM could, in the line of Tooby and Cosmides, be interpreted as an emergentist principle: sociocultural factors emerge at the group-level from the interactions of individuals, much in the same way as psychic or mental properties emerge from the complex, neurophysiological processes in the brain. According to these opposed principles, culture is either wholly heteronomous or strongly autonomous in relation to biological mechanisms and genes. There are, however, other conceptual possibilities available for conceiving this culture-genes relationship.

In accord with our reinterpretation of the status of SSSM explanation, we propose to (re)locate the ontology of the SSSM at a “supervenience” level by analogy with pertinent positions in the philosophy of mind (Kim 1996). This implies, firstly, that the sociocultural level is only weakly autonomous and, secondly, that sociocultural factors only superveniently cause other such factors or mental ones. On this interpretation, the nonreductionism of the SSSM means little more than that culture supervenes on genes. Whereas the ICM holds culture so tightly on a gene-leash that it becomes totally heteronomous, the SSSM—on our interpretation—provides culture with a weak autonomy of its own. Although culture can never become strongly autonomous and thus “unleashed,” it has enough autonomy to superveniently cause, at least partly, the structure and evolution of culture, mind, and genes. As Boyd and Richerson have put it succinctly: “Culture is on a leash, all right, but the dog on the end is big, smart, and independent. On any given walk, it is hard to tell who is leading who” (Richerson and Boyd 2005, 194). Of course, the doctrines of weak autonomy and supervenient causation in the philosophy of mind are highly controversial (Kim 1993, 1998, 2005), but they certainly give us a theoretical model to think in a more nuanced way about the relation between culture and genes than before. So, if we interpret the ontology of the SSSM along the lines of nonreductive physicalism, then social reality is, we submit, autonomous in a sense which is compatible with the rightful claims of the ICM.

Most importantly, we believe that this ecumenical position is supported by the nature of evolutionary theory. Evolutionary theory is designed to model how multiple levels of phenomena interact—genes, individuals, populations, and multipopulation-level assemblages of populations. This means that properties of the population do affect what evolves at the genetic level (Millstein 2006). For example, if populations are small, drift is a strong force. If populations are large, the gross mutation rate is high so that more favorable mutations will be introduced into the population per unit time. Of course, it is not that population-level phenomena are independent of individual-level phenomena. It is that both are important and interact.

With our reconciliatory ontology, we defend the equally justifiable claims of the SSSM, much in the same way as Boyd and Richerson try to bridge the gap between the individual and social sciences:

Disciplines such as economics, psychology, and evolutionary biology take the individual as the fundamental unit of analysis. These disciplines differ about how to model the individual and his psychology, but because they have the same fundamental structure, there has been much substantive interaction between them. . . . Other disciplines like cultural anthropology and sociology emphasize the role of culture and social institutions in shaping behaviour, and researchers in sociology, anthropology, and history find interaction with each other relatively comfortable. Bridging the gap between the individual and cultural disciplines has proved much more difficult. We believe that Darwinian models can help rectify this problem. Darwinian models are useful precisely because they incorporate both points of view within a single theoretical framework in which individuals and culture are articulated in a way that captures some, if not all, of the properties that their respective specialists claim for them. In population-based models, culture and social institutions arise from the interaction of individuals whose psychology has been shaped by their social milieu. (Boyd and Richerson 2005, 432-33)

In the light of the ontological framework set out in this section (see also Lachapelle [2000]), we now first evaluate the “narrow evolutionary psychology” approach to explain disgust, and then sketch our own theory, integrating Douglas’s SSSM view of disgust with gene-culture evolutionary theory.

2. Disgust and Pathogens

Many aspects of disgust are universal. As Darwin already knew, the ability to express the emotion of disgust appears to be universal (Darwin 1871; Ekman, Friesen, and Tomkins 1971). The same is true for the ability to recognize such expressions as expressions of disgust (Philips et al. 1997). Moreover, many disgust elicitors elicit disgust all over the world: bodily secretions such as (other people’s) feces, vomit, saliva, and decaying flesh or fish are reported as universal or quasi-universal elicitors of disgust.

These universal characteristics of disgust suggest that disgust has evolved through natural selection. But what could be its function? Tooby and Cosmides themselves are pretty explicit on this point:

Imagine an ancestral human who had neural circuits that made dung smell sweet—that made him want to dig in whenever he passed a smelly pile of dung. This would increase his probability of contracting a disease. (. . .) In contrast, a person with different neural circuits—ones that made him avoid faeces—would get sick less often. He will therefore have more time to find food and mates and will live a longer life. The first person will eat dung and die; the second will avoid it and live. As a result, the dung-eater will have fewer children than the dung-avoider. (. . .) As this process continues, generation after generation, the dung-eaters will eventually disappear from the population. Why? They ate dung and died out. The only kind of people left in the population will be those like you and me—ones who are descended from the dung-avoiders. No one will be left who has neural circuits that make dung delicious. (Cosmides and Tooby 1997, p. 4, Evolutionary psychology: a primer)

Dung, it is suggested, is dangerous, and aversive emotions (such as disgust) toward dangerous things are adaptive. Part of the reasoning of Tooby and Cosmides seems downright correct. Feces carry many viruses and bacteria (Curtis, Aunger, and Rabie 2004). The same is true for many other universal or almost universal disgust elicitors: saliva can transmit herpes and syphilis, fleas and rats transmit the plague, fresh water contaminated by animal urine causes leptospirosis (Weil disease), and nasal mucus harbors tuberculosis. So, according to this line of reasoning, disgust is adaptive for humans because it prevents the (repeated) self-administration of toxic stuff. The adaptiveness of disgust seems even more plausible when its form is taken into account. The specific structure of disgust (stomach revulsion and other typically bodily reactions) closely fits its proposed function, and makes it rather unlikely that it can be explained as a nonadaptive byproduct of another (adaptive) trait. Disgust is an “aversive” emotion, leading to a desire to have the elicitor removed. When something disgusts you, it almost literally turns your stomach. Moreover, it produces lesser increases in heart rate than other negative emotions (e.g., fear), partly because the adaptive behaviors associated with disgust (sensory shutdown, vomiting) require lower levels of cardiac activation than flight reactions (Levenson and Ekman 2002).

However, the first stumbling block for this evolutionary hypothesis is the fact that disgust is significantly influenced by culture. Darwin himself noted that the people living on the Isla Grande de Terra del Fuego were disgusted by other things than he was:

A native touched with his finger some cold preserved meat which I was eating at our bivouac, and plainly showed utter disgust at its softness; whilst I felt utter disgust at my food being touched by a naked savage, though his hands did not appear dirty. (Darwin 1872, 198)

Although some disgust elicitors are almost universal in humans, many others are particular for one or a few cultures. It seems as if the disgust elicitors are not only genetically inherited, but also socially transmitted.

A second problem for a naive adaptationist hypothesis is that disgust is a typically human emotion. Other animals show distaste or proto-disgust, but only humans exhibit disgust (Rozin et al. 1999), probably in part because this emotion involves a lot of conscious reasoning. It may be quite literally a gut feeling, but it very often starts in the head.

Closely related to the typically human character of disgust is, thirdly, its domain-flexibility and lack of isolation. Although many people think that disgust is intimately tied to taste and/or food, disgust can equally well be activated by visual, auditory, tactual, and olfactory stimuli. Hence, disgust does not seem to be very modular, whereas most evolutionary psychologists believe that adaptive disgust is a Darwinian module, or at least that the automatic appraisal mechanism is modular, because only such modular mechanisms can be adaptive (Evans 2001). Rozin and his colleagues have argued that disgust is far too domain-general to be modular. Think, for instance, of moral and social disgust (Rozin, Haidt, and McCauley 2008): for some vegetarians, meat is disgusting, but the meat-eaters themselves are equally so. Close contact with a serial killer may provoke disgust, and this disgust is (neurophysiologically and behaviorally) very similar to the “core disgust” that focuses on food (Zhong and Liljenquist 2006, Chapman et al. 2009).

These stumbling blocks make clear that a more sophisticated evolutionary hypothesis than the rather naive hypothesis about disgust endorsed by Tooby and Cosmides would be welcome. Actually, it’s not so difficult to come up with such a more sophisticated hypothesis within the framework of evolutionary psychology. Consider the following elements.

So, a more nuanced theory of disgust within an evolutionary psychology framework is possible, although demonstrating its viability in detail remains a challenge.

Hence, it seems that evolutionists shouldn’t bother too much dwelling on an outmoded culturalist explanation of disgust, such as Douglas’s. Yet, we think that there is one important reason not to neglect Douglas’s culturalist view. Evolutionists, and evolutionary inspired disgust theorists like Rozin, acknowledge that learning disgust is to some extent also learning a culture. And one can safely assume that learning a culture essentially involves many of the factors found in traditional SSSM explanations by cultural anthropologists and sociologists (Dupré 2005). Moreover, some Darwinians, especially those found in “narrow evolutionary psychology,” tend to forget that getting the proximate mechanisms right often helps finding a correct answer to the ultimate why-questions. So, if Douglas’s culturalist hypotheses regarding disgust turn out to be on the right track, they can eventually refine the evolutionary accounts of how disgust and culture are acquired. In addition, as we will argue, such a refinement may prove fruitful for a Darwinian theory of cultural evolution in general. Before tackling these issues in sections 4 and 5, we briefly sketch Mary Douglas’s theory of disgust and its pivotal claims.

3. Disgust and Categorical Boundaries

According to Douglas (1966 [2002], 1970), rules are created against contact with a thing that cannot be given a place in a society’s system of categories. If something exists at the border of cultural categories, or in between such categories, it should not be eaten (or touched, or seen). Douglas emphasizes that the objects of cultural taboos and purity beliefs are perceived as dangerous or powerful. She argues that these objects are not dangerous in themselves: they are not per se unhealthy. They are only dangerous because of the place they occupy in a symbolic system. Such a symbolic order influences not only people’s behavior (through rules and regulations), but it also has an important effect on the emotions and preferences of human individuals. Take, for instance, the taboo on eating shellfish in the Jewish culture, which creates disgust for eating shellfish in Jewish people, even though shellfish is quite nutritious and people from many other cultures consider it a delicacy. According to Douglas, this taboo exists because shellfish are anomalous creatures in the Hebrew categorical system. The Jewish folk biological classification has no place for shellfish. These animals do not obey the “God-given” categories. Shellfish are off limits since they don’t quite meet the requirements for true fish. They live in the sea, but they don’t have scales and fins.

In Douglas’s view, taboos should not solely—or even primarily—be interpreted in medical terms. They are not simply hygienic measures, even though some of them might have sanitary advantages. Halal or kosjer food may perhaps contain less pathogens, but the rules regulating food consumption have other origins: “Even if some of Moses’s dietary rules were hygienically beneficial, it is a pity to treat him as an enlightened public health administrator, rather than as a spiritual leader” (Douglas 1966 [2002], 37). According to Douglas, this symbolic interpretation also holds for the near universal disgust for feces. Like so many other bodily fluids, such as urine, menses, mucus, and earwax, feces are on the verge of the inside and the outside of the body. Hovering somewhere between our body and the outside world, they belong to neither of them. As such, feces attest to an awkward ambiguity that jeopardizes the way we categorize our world. Feces, then, are not unclean because they contain lots of pathogens (at least not in the first place), but because they are categorically unclear. They resist to be fitted neatly into our comfortable and familiar worldview. Therefore, Douglas claims, the taboo on feces and, in general, on ambiguous entities and situations, originates in the typically human “reaction which condemns any object or idea likely to confuse or contradict cherished classifications” (ibid., 45).

The idea that disgust has something to do with a symbolic or categorical system is shared by some of the most important researchers in the field. Rozin and Haidt (Rozin, Haidt, and McCauley, 2008), for example, have developed the “animal reminder” view of disgust: we feel the need of distinguishing ourselves from animals, and are disgusted by anything that marks our animal nature or that makes the categorical animal-human distinction blurry. All animals urinate, hence human urine is disgusting. Animals do often bleed, hence human blood disgusts us. But tears as a body product we generally don’t find disgusting, simply because it is the only uniquely human body product. Although Rozin and Haidt explicitly acknowledge that sociomoral disgust is not derived from animal-reminder disgust, they defend a highly similar explanation for this social-moral disgust: if physical disgust is about distinguishing ourselves from animals, then social disgust is about distinguishing ourselves from “demonic” wrongdoers and “beastly” criminals. We simply want to keep the category “human being” clearly defined.

On this view, the purity of which Douglas writes, is not just some or other neutral categorical purity, but rather the purity of being cleaned from animal traces. In this way, Rozin and Haidt avoid one of the major problems of Douglas’s account, namely that her explanation sometimes tends to be tautological: if something pollutes, it doesn’t fit, and if it doesn’t pollute, it fits (Miller 1996). According to their view, if something pollutes (or is disgusting) and doesn’t fit, it doesn’t fit because it reminds us of our animal origins. Furthermore, Rozin and Haidt, unlike Douglas, do not take an anti-evolutionary stance. They do believe that core disgust may have been naturally selected as a food-rejection system: core disgust may be an adaptation to solve the adaptive problem of microbial threats. However, this biological and uniquely human adaptation was exapted for other kinds of rejection: “This harnessing, or accretion of new functions may have happened either in biological evolution or in cultural evolution. Human societies take advantage of the schemata of core disgust in constructing their moral and social lives, and in socializing their children” (Haidt et al. 1997, 127). As to why we would care about the boundaries between animals and humans, they contend that fear of death is a likely motive. Any reminder of our animal nature is also a reminder of our own mortality. And indeed, much of what we find disgusting relates immediately to death. Think for instance of blood, amputated members, or corpses.

Although Rozin and Haidt’s theory goes a long way in the direction of a genuine biocultural account of disgust, it is still confronted with a series of problems. First, the theory does not completely account for the facts. Many things we share with animals do not disgust us (breathing, moving), and we are less disgusted by the animals that resemble us (monkeys and apes) than by genuinely freaky animals such as insects, spiders, and squids. And most nonhuman animals do not sweat as conspicuously as we do, but we do find sweat disgusting. Second, the difference between humans and other animals does not pose an adaptive problem, so caring about it by humans through the formation of disgust reactions does not seem particularly adaptive. Of course, one could object that mortality is something that reduces our genetic fitness, and that the awareness of it is a side effect of our adaptive (and uniquely human) cognitive capacities (Montell 2002), even though avoiding any reminder (even the quite distant ones) of our animal, finite nature does not help much in solving the problem of our mortality. Still, this avoidance of “animal reminders” seems to be too costly not to be selected against. And lastly, fear seems to be a more “appropriate” response to mortality than disgust.

We acknowledge that these specific problems are far from devastating for the model of Rozin and Haidt. Moreover, it is hard to deny that their account culminates in a nuanced combination of biological and cultural elements. However, the animal reminder view still remains too crude and too vague. When Douglas’s insights are integrated into the biocultural theory of disgust, we claim that higher levels of sophistication and accuracy can be reached in evolutionary thinking. In addition, we will argue that such integration will allow us to partly correct the animal reminder view.

4. Symbolic Order and Adaptive Disgust

Most cultural anthropologists dismiss evolutionary explanations outright because they believe that “culture” overrides “nature” when making sense of human behavior. Boyd and Richerson illustrate this antinaturalism precisely with the example of food taboos and disgust:

Many cultural anthropologists make fun of the idea that human behaviour is adaptive, and delight in citing examples of what seem like capricious and arbitrary differences between cultures. For example, Marshal Sahlins cites the fact that the French relish horsemeat, while Americans find it inedible as dog flesh. How could it be, he asks, that it is adaptive to eat horse in France, but not in America. Moreover, such examples can be multiplied endlessly—in many societies dog meat is a delicacy. Culture, not biology, rules. (Richerson and Boyd 2005, 148)

Douglas’s theory of disgust also seems to illustrate this SSSM “dogma.” On her view, the cultural categories are the crucial factors in understanding disgust elicitors. If some of the disgust elicitors are found in almost every culture, the best explanation is that some things (e.g., feces), by their own structure (e.g., unclear whether in or out the body), cannot, or only with difficulty, be placed in a categorical structure.

Douglas emphasizes that disgust is a response to a threat to the known and predictable order (the “cherished classifications”). Disgust protects (1) the subject’s capacity to organize his or her universe, and (2) the organization/order itself. Although she explicitly frames these claims in an anti-adaptationist framework, they do not need to be anti-adaptationist as such. Our central hypothesis is that it may very well be adaptive to feel disgust when the categorical or symbolic order is not respected, (a) if this order itself is adaptive, and (b) if disgust genuinely protects it. Douglas has made it sufficiently clear that disgust does protect order, and the huge amount of data amassed by disgust researchers can easily be interpreted as further substantiating her claim. However, most disgust researchers shy away from claims about the biologically adaptive value of order-protection. They seem to think that one order cannot be intrinsically better—more adaptive—than another. Hence, it is not intrinsically better to relish horsemeat rather than dog flesh, from a naive adaptionist perspective.

Now, while one may concede that one order is not more adaptive than another, it does not follow from such a concession that having a categorical order as such is adaptively neutral (or even maladaptive). Yet, one might object, can having a rather strict categorical order be genuinely adaptive? Why would it be in the interest of our genetic fitness to be culturally strict or rigid? We know that such a rigid categorical scheme sometimes comes with significant costs. For instance, when people operate in a new cultural environment, they are often overwhelmed by feelings of disorientation, surprise, and confusion. The anthropologist Oberg called this phenomenon “culture shock” (Oberg 1960), and noticed that the victims of such a culture shock often feel deeply disgusted by certain aspects of the new and different culture. They start malfunctioning on many levels; culture shock is even a determinant of increased risk of schizophrenia (Martens 2006).

But maybe the costs of this strict cultural system are relatively minor to the costs of an unrestricted cognitive flexibility. At least, that is what Joseph Carroll argues for. In his view, our species’ cognitive flexibility is quite costly. According to Carroll, humans need a mechanism to counter the maladaptive side effects of our cognitive flexibility. He underscores that we need a coherent and strict scheme of the world to set the boundaries for our capacity to experience that world (Carroll 1998). This reasoning is in line with E. O. Wilson’s thinking about the costs of an unlimited cognitive flexibility. According to Wilson, the large human brain has adaptive (survival) value. It has helped us to become one of the most successful species ever on this planet, because the brain is the crucial factor in our species’ ability for adaptation to very different environments. But the downside of this adaptation (or set of adaptations) is that we are the one species that has to wrestle with a new set of adaptive problems: because the human brain is so flexible, it also causes confusion and uncertainty (Wilson 1998). To achieve our flexibility, evolution had to cut human cognition loose from any rigidly programmed set of instinctual mechanisms and behaviors. Yet, this unlimited flexibility would come with substantial costs, unless it was counteracted by another rigidly programmed structure at a “higher” level. And that higher level might very well be the cultural level with its rigid categorical system. Theorists like Scott Atran and Dan Sperber have made a similar point, albeit in a markedly different context. Sperber and Atran (Atran 1990; Sperber 1990) argue that the active structuring of the world, which is found in folk biology and folk psychology, is helpful in guiding inductive inference. Essentialist and strict categories guide inductive inference more successfully than other representational systems, even if such categories are ontologically unwarranted. The continuous flow of contingencies is stabilized through generalizations and concepts, and these generalizations and concepts are to be seen as fast and frugal heuristics to address the vast complexities of the world we live in (Todd and Gigerenzer 1999). Hence, there seems to be a good evolutionary reason why we are fond of our categories, and why we dislike categorical ambiguity (Barrett 2001).

Remarkably, Wilson’s evolutionary claim echoes the view of many antinativist anthropologists, such as Clifford Geertz, who once famously wrote that “[u]ndirected by culture patterns—organized systems of significant symbols—man’s behavior would be virtually ungovernable, a mere chaos of pointless acts and exploding emotions, his experience virtually shapeless” (Geertz 1973, 46). This echo is especially striking, given the fact that Tooby and Cosmides (1992, 26) notoriously criticize this quote, because they believe it establishes that the social world is the primary cause of the mental organization of adults. But although we do not adopt Geertz’s strongly autonomous symbolism in anthropology, the general drift of his explanation makes perfect evolutionary sense against the backdrop of our own integrated theory.

On our hypothesis, disgust might be adaptive in case the “cherished classifications” are transgressed, on the twofold condition that these classifications themselves are adaptive and that disgust really safeguards them. Now, if Wilson’s hypothesis is correct, then disgust can be seen as an affective force to counter the maladaptive effects of unlimited flexibility (complementing other less emotional and more cognitive mechanisms, such as self-control). A rigid categorical system shields off the lack of certainty and predictability caused by too much flexibility. If avoidance of categorical fuzziness is adaptive, then disgust reactions to protect a categorical order are adaptive too. For a brainy species like ours, disgust comes in handy for setting categorical restrictions to what we can think, do, and imagine, for otherwise we would be trapped in confusion and uncertainty.

A rigid categorical system has other advantages as well. Apart from being very brainy animals, we are also extremely social animals. Humans can only thrive in an environment in which other humans are present when they share more or less the same cultural order, including the same categorical system. For a social animal, transparent communication is crucial, and a shared system of rather rigid categories makes communication both easier and more transparent (Deacon 1997). Although not all forms of social transmission make use of symbols, social transmission in humans is typically symbolic, and symbolic communication allows for a great deal of social transmission of adaptive information without direct observation (Henrich and McElreath 2003). Therefore, a rigid symbolic system is certainly to be preferred over a less rigid symbolic system as a less rigid system would cause many copying errors in the already error-prone psychological mechanisms of cultural replication (Claidière and Sperber 2007). Sharing our thoughts requires a shared system of categories, and the more rigid this system is, the better it can be shared. So, making cultural transmission less “noisy” implies making it more adaptive. Furthermore, because such a shared system of categories strengthens the cohesion of the group, it comes in quite handy at times of intergroup conflict (Hirschfeld 1996; Richerson and Boyd 2005). Human communication is not only grounded in cooperative and shared intentions, but to a certain extent it also creates shared intentions and efficient cooperation (Tomasello 2008). And the fine-tuning of these intentions and cooperative behavior within a group can be expected to fare better (e.g., in cases of intergroup conflict) with more robust categories (and other conventions) than with less robust ones.

Social learning in humans is characterized by a series of distinctive skills, such as imitation, teaching by (verbal) instruction, and joint attention. Many of these distinctive skills are well suited to—at least partly—maintain an ordered categorical system. In contrast with social learning in nonhuman animals that tends to enable the learner to reinvent the routine or behavior for itself, human cultural learning usually does not require a significant degree of individual trial-and-error discovery (Jablonka and Lamb 2005; De Smedt, De Cruz, and Braeckman 2009). In other words, human social learning is less error-prone than other forms of animal social learning, and many of the underlying cognitive mechanisms (e.g., shared attention) and anatomical adaptations (e.g., the huge white sclera on both sides of the iris that makes it easier to follow one’s eye direction) were probably also selected because they reduced the error-rate (Kobayashi and Kohshima 2001). Moreover, our capacity to understand the goals of other individuals adds to the faithfulness of social transmission (Tomasello 2008).

Now, if all of this is true—which it probably is—why would humans still need disgust to protect the ordered categorical system? Are the other mechanisms and adaptations that support gaze following, cooperation, and shared attention not sufficient to maintain socially stable categories? We think they are not. First, all cultural evolutionists stress the importance of faithful social transmission for cumulative cultural evolution to occur. But they have to admit to their critics such as Sperber that social learning in humans is still quite error-prone (Richerson and Boyd 2005; Sterelny 2006). Because there is a kernel of truth in Williams’s saying that “the natural selection of mutation rates has only one possible direction, that of reducing mutation rates to zero” (Williams 1966, 139), selection will probably occur for a not too costly emotional tool that can help to further reduce the transmission noise (infra). Second, disgust is known to be a strong motivator that encourages hypervigilance with regard to contamination. There are a number of factors that motivate us to communicate and to agree on joint goals, but disgust is—in our view—the emotion that motivates us to stay away from categorical ambiguity that might hamper communication and/or cooperation. Disgust is an unpleasant emotion, so that when people feel disgusted because their own behavior or thinking has transgressed a cultural standard, they are inclined to avoid this kind of behavior and/or thinking in the future. But disgust might also mediate locally adaptive behavior in another way: when other people infringe upon cultural conventions, they are corrected by facial or verbal expressions of disgust from the group. This response is quite effective because prolonged social disgust serves to punish and ostracize the “disgusting” person (Curtis and Biran 2001).

All the evidence above shows that keeping categories robust and unambiguous, certainly makes sense from an evolutionary point of view. Consequently, having an emotional defence, like for instance disgust, to safeguard these categories from fuzziness and to keep our categorical system rigidly in place, might also add to our genetic fitness. That is not to say that the function of disgust is limited to its being such an emotional defence. The traditional “narrow” evolutionary explanation of disgust accounts well for many instances of disgust, even if (or precisely because) the exact content and range of disgust is largely determined by social and individual prepared learning. For instance, Joe and Nathalie Henrich (2010) have recently shown that the patterns of food taboos in three Fiji-villages represent a culturally evolved adaptation, influenced by various cognitive biases, that protects people from dangerous marine toxins. These patterns probably emerged, and are now maintained, by the operation of the cultural learning mechanisms identified by dual inheritance theory. However, we contend that this proto-function of (core) disgust has in the course of human evolution been exapted into two other functions, a cognitive one and a social one, making disgust even more advantageous for the brainy and social animal we are.

5. Why Disgust?

One crucial question as to the second part of our central hypothesis has been left unanswered: why would disgust be the best emotional defence against categorical fuzziness? For the sake of clarity, we will tackle this question by splitting it up into two subquestions: (a) why does it take an emotion to fulfil this function, and (b) why would disgust be the most fitting reactive attitude to ward off categorical fuzziness?

One could label our argument “a sociocognitive exaptationist story” ² of disgust. This exaptationist story can be summarized as follows. The evolution of disgust started with the distaste or proto-disgust of our primate ancestors. In our primate relatives and in young children, distaste still serves as a rejection response to bad-tasting (and unhealthy) food. Human core disgust is a more cognitively elaborated version of this distaste; it is an adaptive reaction to food items and other stuff that might be contaminated with pathogens. The shift to sociocultural disgust (the disgust that is elicited by categorical ambiguity) is an exaptation of this core disgust. In this section, we will argue how and why this exaptation might have occurred. It sketches how disgust helps in transmitting information and how disgust itself is transmitted. But it also tries to show why and how disgust got co-opted to play auxiliary roles in the sociocultural domain.

6. The “Integrated Hypothesis” and the Animal Reminder View Compared

As far as gene-culture co-evolutionary theories provide (part of) the framework for a unification of the behavioral sciences (Gintis 2006), every scientifically sound social constructivist hypothesis can be informed by evolutionary theory. At the same time, we acknowledge that there are properties of the phenomena studied by social constructivists that cannot be explained in terms of the concepts of evolutionary theory alone. For that reason, the “integrated hypothesis” of disgust we have developed in this article differs substantially from the reductionist ICM that Tooby and Cosmides propose. Quite unlike the narrow evolutionary approach, our blend of Darwin and Douglas underscores the importance of symbolic causality for human behavior and emotions. In conclusion, we compare our hypothesis with its best contender, the animal-reminder view.

The strongest alternative to our “integrated hypothesis” is probably not the narrow evolutionary explanation of disgust given by some evolutionary psychologists and human sociobiologists. Most disgust researchers have adopted Rozin’s framework for their studies. In line with our own hypothesis, Rozin explicitly acknowledges that there exists an important cultural influence on the formation of disgust. According to him, many elicitors of disgust have indeed their pathogenic character in common, but also other objects and persons elicit disgust because they remind us of our all-too-animal nature. Things like corpses, blood, and our own saliva in a cup turn our stomach because they are reminiscent of the fact that we are not made of mind and ratio alone.

First, and for the purpose of this article foremost, our hypothesis is the more evolutionary one. Whereas Haidt and Rozin primarily refer to the embodied cognition literature to explain the diversions of disgust, we have tried to give some evolutionary arguments for why these diversions might often be adaptive, and for how they might have evolved from more ancient emotional reactions. Second, our hypothesis might also provide an evolutionary underpinning of the animal-reminder view that seems more evolutionarily viable than the evolutionary background that Rozin and Haidt have sketched.

We believe that the animal-reminder view (see section 3) is not necessarily wrong: disgust can very well be elicited by objects or processes that remind us of our animality. However, the “animal reminders” do not primarily fill us with disgust because animals remind us of our own mortality, as Haidt and Rozin believe. Against the backdrop of our “integrated hypothesis,” we conjecture that it is not our animality/mortality as such that is responsible for disgust reactions, but our insignificance or meaninglessness. The “animal reminders” elicit disgust because animals remind us of the insignificance of our lives. This conjecture follows directly from our “integrated hypothesis”: the existential insignificance of which we are reminded is closely related to the insignificance—and thereby caused disgust—we experience when our “cherished classifications” and categories are muddled up and blurred, as Douglas points out. Let us elucidate this conjecture somewhat further.

Normally, we take ourselves seriously and think of ourselves as important beings. When this illusory bubble is pricked, we come to realize that our existence is as insignificant and meaningless as that of animals, machines, and stones. We may well be cognitively aware that, sub specie aeternitatis, all our actions and projects are ultimately meaningless and arbitrary (Nagel 1986), but we only genuinely experience that insight when we are confronted with concrete evidence—for example, “animal reminders”—of the contingency of our life. When so confronted, the emotion we experience is disgust. This is so because the meaninglessness of our existence of which we are reminded closely resembles the meaninglessness we experience when infringements are made upon our cultural symbolic order. A confrontation with our animal nature might lead to the distortion of those fragile distinctions that we cherish and that structure our lives. Of course, we do not automatically experience this meaninglessness whenever we are convinced that our life is insignificant and meaningless. But when we are hit by an experience that involves the contamination of the personal with the arbitrary and meaningless, a (mild) nausea often is the result.

To be sure, this does not imply that disgust has nothing to do with mortality. In our view, however, mortality only elicits disgust when our death (or the death of others) echoes a lack of meaning, when mortality expresses meaninglessness. This important component of our “integrated hypothesis” becomes clear as soon as the relationships between mortality, fear, and disgust are taken into consideration: mortality as danger evokes fear, whereas mortality as a reminder of our existential insignificance evokes disgust. In this context, it is telling that Jean-Paul Sartre in his novel Nausea connects disgust with the unfamiliarity of everything, including one’s own existence (Sartre 1959). The protagonist of that novel, Antoine Roquentin, discovers that his “pure disgust” is caused by the confrontation with “pure being,” that is, the experience of things separated from words or categories, from their familiar significance.

In some passages, Haidt and Rozin come very close to what we are saying:

The European existentialists felt nausea and dread as they contemplated the senseless slaughter of World War II. They felt nausea because, in the West, meaninglessness is perhaps the greatest threat to the self. When people are so casually stripped of life, or of dignity, the implication is that life is cheap, and there is nothing of value to be respected. The Americans in our data similarly felt threatened and disgusted by the senseless murders that happen around them every day, and by people who strip others of their dignity, including racists, rapists, and child abusers. (Haidt et al. 1997, 129)

So, one may wonder whether there is then a fundamental difference between our “integrated hypothesis” and the animal-reminder view. Certainly, Haidt and Rozin also acknowledge that disgust and meaninglessness are intimately intertwined. But whereas we connect this meaninglessness to the jeopardizing of the structure imposed on the world by cultures, they seek for an explanation in the rather vague claim that disgust, or some subset of its embodied schemata, is the emotional response to the heterogeneous class of threats that one cannot simply run away from or fight off (Haidt et al. 1997). We essentially differ from them in that we claim that the existential insignificance is closely related to the insignificance and disgust we experience when our “cherished classifications” are distorted. Moreover, we have tried to show (1) how higher cognitive versions of disgust develop continuously from deep roots in “core disgust” (Clark 2010), and (2) how these higher versions are probably culturally evolved adaptations. In that sense, our “integrated hypothesis” of disgust really integrates Mary Douglas’s important work on symbolic structures and cultural categories into the Darwinian approach. And that is why Darwinians should not be afraid of Mary Douglas—and vice versa.