Ethnography Challenges the Belief-Motivation-Action Equation

An important part of explanation in anthropological writing has always been the process of setting people’s actions in the context of their beliefs, or knowledge. The aim is usually to explain behavior that might otherwise be hard to understand for an outsider who does not share the tacit assumptions, explicit theories, or experiences of the people whose lives the anthropologist is describing. Naturally, this means that collecting and presenting the beliefs and knowledge, including people’s explanations of their own actions, has always been an important part of the work of ethnography.

This kind of explanation—setting action in the context of beliefs—is something more than description, but it is not, I suggest, usefully thought of as a theory. Instead, it would be appropriate to describe this procedure as analysis. On the other hand, explaining using thoughts and beliefs does require a theory, whether or not it is explicitly discussed, in order to explain the relationship between three terms on which such analyses are based: (1) action or practice, including speech and the production of other representations; (2) mind or thought; and (3) the content to which the mind can be applied: beliefs, knowledge, sensations, emotions, desires, and so on.

This tripartite schema must be in place regardless of the position one takes on the question of the relationship of representations to the world. It is still required, for example, for “ontological” approaches that deny there is a difference between mind and world—jaguars can have beliefs and be mistaken about them and be misled by cunning hunters (cf. Viveiros de Castro 1998). The underlying theory that explains the relationship of these three terms (action, mind, and content) is best described as a theory of cognition, however suspicious many anthropologists have come to be of that term. Those anthropologists who have explicitly called their work “cognitive anthropology” (e.g., Boyer 1994; Sperber 1996; Whitehouse 2004) have focused on representations (including imagistic ones), but cognitive content does not need to be representational (Bloch 2005). Other anthropological theories of cognition are often taken for granted and tacit, but they are no less theoretical for that, and no less cognitive. A theory of cognition might include emotions and affect as well as beliefs, knowledge, memory, the content of utterances, texts, ritual, and so on. And to have a theory of cognition does not necessarily commit us to the kind of mind-body dualism usually blamed on René Descartes, or to the individuality of minds; one might well defend a theory of cognition based on the Strathernian dividual (Strathern 1988) or the Durkheimian superorganism (Durkheim 1979), for instance. Many anthropologists have done just that, of course.

A simple and widespread theory of cognition is what philosophers have called the belief-motivation or belief-desire model of action (Elster 1986). This is the theory that action is the product of decisions that arise when preferences are considered in the light of beliefs about the world in which the actor is constrained to act (table 1).

For example, if we have observed people making an offering to a god, we might use this theory to offer an explanation by supplying information about relevant beliefs and desires. This could be based on ethnographic or other data, if we have them, or we might have to speculate in a way that would at least be helpful in driving further research. The result might look something like table 2.

Like a mathematical equation with three terms, the logic of this model is that if we know two of the terms, we can derive the third. For example, if we know what someone believes and what they want, we can predict what they might do. If we know what someone did and we know the beliefs they hold about the world, we can at least begin to think about what they want. Say that we know that someone is ill and wants to get better and they decide in that context to conduct an elaborate ritual; we might conclude that they believe the ritual will help to heal their disease. Where the three elements are known but are not logically related to each other, then we might be looking at a case of irrationality.

Of course, the belief-motivation model, understood in this way, is a gross simplification. Its mechanistic form is also a turn-off for anthropologists who know that actions and representations are polyvalent. Real life is full of instances in which the belief-motivation model does not seem to apply, as ethnographers know only too well. The theory assumes too much about the consistency of beliefs over time and in different contexts, and about the relationship of action to habit and to the emotions.

From its beginnings, the Malinowskian fieldwork paradigm served precisely as an efficient way of finding examples of inconsistent statements of belief, or of statements of belief that were not consistently reflected in behavior. By observing life in its minutiae and collecting data over long periods, rather than simply relying on interviews or formalized accounts of mythology, ethnographers found that their subjects’ accounts of their own lives were frequently inconsistent. Fieldwork also brought the realization that very different forms of life also made sense and were often practically efficient, despite these inconsistencies. So rather than simply diagnosing irrationality as previous generations of anthropologists had done, anthropologists from Bronislaw Malinowski’s time on have increasingly sought to explain apparently irrational behavior or statements of belief by providing richer and more complex accounts of the relationship between action, mind, and content.

Multiplying the Terms of the Equation

The classic belief-motivation model relates the key terms—motivation, belief, and action—by assuming that each category is internally coherent and that the mind that works with all three is capable of drawing logical inferences. It works only to the extent that subjects have preferences that are not contradictory, for example, and only to the extent that beliefs are consistent over time, and only to the extent that people are capable of acting in a way that fulfills their decisions. I have said that ethnographic fieldwork led ethnographers to posit a more complex relation between minds, content, and action than that envisaged by the belief-motivation model. This is not the place for an exhaustive survey of these contributions, which are extremely diverse. What I want to highlight here is that one common way in which these more complex theories of cognition work is to split one or more of the terms of the belief-motivation model into multiple subcategories: different genres of thought and speech, different domains of action, different kinds of mind. As most anthropologists since the early twentieth century have accepted the psychological unity of all humans, their attempts to generate more complex accounts of cognition have focused on understanding multiple forms of content and action rather than suggesting that there are different kinds of mind. A widespread version of this approach that is still influential distinguishes between ordinary language on the one hand and religious language on the other. On this view, ordinary language reflects a person’s beliefs and preferences in a straightforward way (that is, in the way the belief-motivation model would suggest), while the special genre of religious language must be understood in a different way. Perhaps religious representations are systematically metaphorical, or “symbolic,” and must be translated into literal language before they can be understood. Perhaps speech in a religious context has a “performative” role. By adding genres in this way, it is possible to account for—to explain—important areas of thought and action that are apparently irrational while leaving subjects’ capacity for pragmatic action intact.

Consider a classic example: Edmund Leach’s essay “Virgin Birth,” published in 1966. The Trobrianders and some Australian groups are said to believe that a woman can get pregnant without any involvement from a man. To interpret such statements in the same way as ordinary belief, argues Leach, is to treat these people as if they are stupid. Statements about the role of men in producing children, he goes on, should be treated as a special category of speech, which he calls “dogma.” Dogma is a special genre of language that expresses consent to the prevailing social system. In matrilineal systems, the brothers of children’s mothers and the progenitors of those children often come into conflict. When a Trobriander asserts the reality of virgin birth, that assertion serves to affirm the matrilineal social system by denying the validity of fathers’ claims to rights in children. In other contexts, such as animal husbandry, the same people show themselves perfectly capable of understanding the biological conditions of paternity, Leach insists.

Leach’s concept of dogma shows that it is possible to complicate the belief-motivation model by introducing a distinction between two subcategories of action, without giving up on the idea of explicability (table 3). Once the expressive content of dogma is rendered in literal form (“I do not wish to challenge the prevailing social order”), it can be fed into the standard belief-motivation model like any other belief.

However, some other anthropological theories of mind do not preserve strict rationality in this way. Consider, for example, Pierre Bourdieu’s (1977) theory of doxa. Bourdieu distinguishes between two modes of thought and action: objectified and unobjectified. Objectified thought is based on well-ordered, internally consistent, and ends-oriented beliefs and is associated with certain unusual activities, such as formal education or capitalist bookkeeping, in which establishing a disinterested view is valued. The content of what Bourdieu calls “practical” thought, by contrast, is not objectified; it is naturalized through its embodiment in habitus and is therefore unquestioned and perhaps unquestionable (table 4).

Though practical thought reflects economic or strategic interests, and therefore has a certain instrumental rationality when judged by its effects, it is not strictly rational when judged in terms of its content because it may not produce beliefs that are internally consistent or consistent over time. Kabyle determinations on questions of kinship are unstable because they depend on contextual calculations of cost and benefit that are not—and cannot be—apparent to the person making the calculation, if Bourdieu is right.

In this sense, then, Bourdieu’s doxa is unlike Leach’s dogma in an important sense. A further difference is the way the boundaries between different categories of content or action are mapped onto distinctions between groups in society or between societies. For Leach, dogma was a mode of speech in which anyone could participate some of the time, a mode, as he explicitly argued, that was just as available to Cambridge dons as it was to Melanesians. Bourdieu, by contrast, argued that objectified thought was more characteristic of particular classes of people—scholars and capitalists are two important examples. It is therefore more typical of societies in which members of such classes are numerous than of those in which they are few in number or absent. That still allows for a smooth gradation of societies ranging from those with the least space reserved for objectified thought to those with the most.

Other anthropological theories of cognition draw a firmer distinction between different kinds of society determined on the basis of their forms of cognition. I have in mind here the influential idea that a certain kind of reflective belief is only found in societies influenced by Christianity, or more generally by monotheism. Malcolm Ruel (1982), for example, complained that anthropologists frequently assume that the people they study have, and are able to reflect on, well-formed, consistent propositional beliefs. This, he argued, betrays the formation of anthropology as a discipline with roots in Christian societies. Christianity is a tradition in which adherents are expected to hold such a set of propositional beliefs—the creed—and to be defined by them. Many societies have been deeply influenced by Christianity, but the ground rules of Christian belief do not apply in all societies or all religious traditions, Ruel warned.

Jean Pouillon (1979, 1982) argued along similar lines, although he thought that the key to the distinction is not Christianity, with its emphasis on acceptance of the creed, but the cosmological model shared by all the Abrahamic religions. These traditions, he reasoned, all feature a bipartite cosmology that distinguishes between the world we know through experience and the transcendent world “beyond,” to which we have no direct access (table 5). Belief, in the religious sense, is what we do in respect of knowledge about that world beyond. It requires commitment and trust because we cannot ever verify its contents. Other religious systems, he claimed, such as animism and polytheism do not share this cosmological premise and therefore have no need for a fundamental distinction between knowledge and belief.

I could provide many other examples of anthropological theories of cognition, but these few examples will suffice for my current purpose, which is to draw attention to the common basis of different approaches with different conclusions. In each case, ethnography that challenges models of rationality is explained using a theory of action, mind, and content that explains the complexity of relations between the terms by dividing one or more of them into multiple subcategories.

The Cultural Nature of Metacognition and Its Importance for Anthropology

Flavell proposed that metacognition improves both according to a universal timetable in younger children and as a result of explicit education about metacognition in older children. For educators, this means that teaching needs to provide opportunities for students to learn to reflect on their learning. For anthropologists, though, there is another, and I think a very important, message. If metacognition can be taught, if it can be passed from one person to another, and if it can vary, then there can be cultures of metacognition. And if there can be cultures of metacognition, then there can be ethnography of metacognition. As far as I can tell, though there has been a profusion of work on metacognition by educational and developmental psychologists, it has taken an exclusively normative approach to metacognition; that is to say, they have been interested in establishing which pedagogies of metacognition are most effective in promoting learning in children. It falls to anthropologists, then, to make the full range of existing forms of metacognition into an ethnographic object.

The specifics of Flavell’s model of metacognition have been debated and developed by educational psychologists and philosophers since he proposed it, and some have suggested the cake should be cut up in different ways. For anthropologists, however the elements are defined, the important point remains: thought and action often depend in consequential ways on how the people concerned think about thought. This is analogous but parallel to the ways in which anthropologists have long formulated their own theories of cognition. Just as Leach can speculate about different genres of language and Bourdieu can elucidate different forms of thought and action, all anthropological subjects are also doing the same thing, and the ways in which they think about their thought will be important for the way they actually think.

Anthropologists might still conclude that people are not aware of all the processes that govern their thought; that would not be surprising. I am by no means arguing for an ethnographic foundationalism (see introduction, this volume) when it comes to cognition, in which taking local or specific forms of metacognition seriously means giving up on formulating and refining a general theory of cognition. However, the role of metacognition in shaping people’s actions in relation to the content they think about, and especially the potential for thought about thought to be part of ethical projects, means that whatever general theories of cognition we might endorse, specific forms of metacognition will often have to be an important part of our explanations. Let me now turn to post-truth to illustrate how this approach can work in practice.

SJWs Always Lie

SJWs Always Lie, subtitled Taking Down the Thought Police, presents a straightforward argument—though there is a twist in the final chapter. It is written in a curious bathetic style that mixes grand world-historical claims with highly personal examples drawn from Day’s quarrels with specific “social justice warriors.” Day’s central claim is that freedom of speech, logic, reason, and Western civilization are under attack. The aggressors are SJWs. They dominate the state, media, education, and other establishment institutions and are motivated by a highly simplistic moral narrative, or “the Narrative.” Day writes, “The Narrative is the story that the SJWs want to tell. It is the fiction they want you to believe; it is the reality that they want to create through the denial of the problematic reality that happens to exist at the moment. And there is no one definitive Narrative. Instead, there are many Narratives, all of them subject to change at any time, thereby requiring the SJW who subscribes to them to be able to change his own professed beliefs on demand as well” (2015, 277).

In a cameo foreword that introduces Day’s argument, Milo Yiannopoulos explains that, because the Narrative is aspirational, not representational, it “cannot survive contact with reality” (2015).¹ The Narrative abhors complexity, but real life is always complicated. This explains what Day calls the First Law of SJWs: SJWs always lie. (Day distinguishes SJWs from other groups of people who habitually lie: propagandists, marketers, and sociopaths [2015].)

As SJWs always lie, on Day’s account, they must constantly face inconvenient evidence; it does not discourage them, because they display “a willingness to deny science, history, logic, their past words, or any other aspect of reality that contradicts their current Narrative” (2015). Apart from facing down evidence and reason, in order to defend their Narrative, SJWs must also contend with heroic refuseniks such as Day himself. This is why, Day asserts, SJWs are so hostile to freedom of speech in all the venues in which they seek to disseminate their Narrative (i.e., every venue in which they find themselves, since they do so at every opportunity).

Despite attempts to prohibit the free speech of dissenters who challenge the Narrative, Day avers that “a broad-spectrum, reality-based resistance to the mirage is now taking shape, a resistance that will eventually undermine and replace all the old institutions that have been invaded and captured by the SJWs. All it takes to be part of it is a refusal to accept the religion of social justice, a refusal to bow down before the gods of Equality, Diversity, Tolerance, Inclusiveness and Progress” (2015, 250).

Members of this resistance can expect to be attacked for their pains. Day explains that when SJWs are confronted with their lies, they will “double down” (this is the Second Law of SJWs) by elaborating on their lies and “project” (and this is the Third Law) by accusing their opponents of faults of which they are themselves guilty.²

Dialectic and Rhetoric

In the final chapter, Day abruptly changes tack. He begins by introducing a distinction—derived from Aristotle’s Rhetoric—between different kinds of language. Both of these forms of speech are used, he explains, for persuasion. Dialectic is based on logic and proceeds by way of syllogism. Rhetoric is based on emotional appeal and works with false syllogisms that merely associate certain conclusions with particular emotions. Day explains that Aristotle defended the use of rhetorical language as a means of communicating with those who are not capable of understanding dialectic. He quotes Aristotle: “Before some audiences not even the possession of the exactest knowledge will make it easy for what we say to produce conviction. For argument based on knowledge implies instruction, and there are people whom one cannot instruct.”

That does not mean, Day points out, that these people cannot be taught in any way, just that they can only be reached by playing on their emotions. Not surprisingly, he asserts that

SJWs fall squarely into the category of people who cannot be instructed and cannot be convinced by knowledge. This is the key to understanding their astonishing ability to cling to their Narrative in the face of evidence that obliterates it as well as their insistence on clinging to it even as it shifts and contradicts itself. The reason SJWs can believe seven impossible and mutually contradictory things before breakfast is their inability to be instructed by knowledge; as long as each of those seven things happens to be in line with whatever their emotions are at the moment, SJWs will not see the inherent contradictions that thinking people do.

Finally, halfway through this final chapter comes the surprising twist. Day gives a practical example in order to illustrate the difference between rhetoric and dialectic. “If I say ‘SJWs occasionally lie’ in response to an SJW’s false statement,” he writes, “this is proper dialectic but poor rhetoric, as it is likely to fail to persuade a rhetoric-speaker of the actual truth, namely, that the SJW is lying in the present circumstance. The better rhetorical statement is ‘SJWs always lie,’ which is not dialectically sound (or if you prefer, untrue), but despite its lack of soundness, it is more likely to persuade the rhetoric-speaker to believe the relevant truth, which is that the SJW is lying.”

So Day concludes his book by saying that the argument he has been advancing is not after all true, in the logical sense, but is merely emotionally effective in persuading people who are incapable of being moved to change their position.

He goes on to explain that there are three kinds of people. There are those who can understand dialectic, who are “intellectually honest and capable of changing their minds on the basis of information.” Then there are those who are only capable of understanding rhetoric, like the SJWs. If one tries to communicate in dialectic to a “rhetoric-speaker, he hears it as rhetoric. Or, not infrequently, as complete gibberish.” Finally, a third category of person has the intelligence to be able to communicate in both modes, and to switch between them at will. Day, of course, sees himself as being in that superior category: “I speak dialectic to those capable of communicating on that level, and I speak rhetoric to those who are not.” He continues, “However, because many SJWs attempt to cloak their rhetoric in pseudo-dialectic, you can use sound dialectic to strip them of that pseudo-dialectic cloak on behalf of those capable of following the real thing, while communicating directly in rhetoric to the SJWs. This requires a degree of fluency in both discourse-languages as well as the ability to switch back and forth between them at will, a skill that takes some time to develop.”

On this view of the world, information presented by the mainstream media, or other establishment institutions, mainly serves a political or “ideological” purpose, rather than an informative one, so it can safely be ignored; in fact, it must be ignored. Day represents the “rebels” of the alt-right as soldiers in the info trenches, pitting propaganda against propaganda by providing competing representations that can win out over those promoted by the liberal establishment.

Of course, by the time he makes these claims about categories of person defined by intelligence and the information war, it is not at all clear to what degree we are to take any of what he says at face value. Indeed, as a result of his open admission that he is involved in a sort of Gramscian war of position, Day’s alt-right sympathizers often appear confused about which of his statements are meant in earnest and which are merely designed to provoke social justice warriors. In other words, to put it in his terms, they are confused about whether he is speaking dialectic or speaking rhetoric. This gives the exchanges between Day and his followers on his websites a sense of vagueness, uncertainty, and playfulness, and a disinterested reader will be left wondering whether even Day has a firm handle on the boundary between his sincerely held beliefs and his attention-seeking posturing.

SJWs Always Lie as a Metacognitive Theory

SJWs Always Lie is typical of much of the discourse that characterizes the alt-right and other populist movements. It is intended as a manual for use in “resistance” to the allegedly dominant culture, and the tools it provides to do that are not really first order, in the sense that they are directly related to the content of the dominant culture that the protagonists find objectionable. Instead, Day addresses second-order questions about the way in which the opponents think. By multiplying terms just as the anthropologists I discussed earlier do, Day aims to provide an explanation, a key to unlock the secrets of rhetoric. Like Bourdieu, he introduces subcategories of belief (table 6). Like Bourdieu’s doxa, the SJW Narrative is, in Day’s telling, wishful thinking—and it is related to preferences in a way that is not epistemically justified. Like Leach, he also introduces subcategories of language: dialectical language for ordinary communication, and rhetoric, which, like Leach’s dogma, has a political effect.

One important difference between Day’s model and those of the anthropologists is that Day also multiplies the kinds of minds that have beliefs and preferences and act on them. For Day, there are three categories of mind, each of which apprehends rhetoric and dialectic in a different way. As I noted earlier, since the time of Malinowski, the “psychic unity” of human beings has been treated as axiomatic by most anthropologists. There are, however, anthropological precedents for this kind of move, most notably in the anthropology of Lucien Lévy-Bruhl and James Frazer.

I want to draw two sets of conclusions from this brief discussion, first in relation to post-truth, and second in relation to anthropological explanations of behavior.

Day is only one moderately influential thinker in the American alt-right, and his thinking can hardly stand for all of the phenomena around the world that have been classified as “post-truth.” However, I think this case is enough to suggest that paying attention to the metacognitive content of alt-right and post-truth discourse can bring certain insights to the specificity of post-truth thought that the explanations that I discussed briefly at the outset ignore. Day’s account is not simply an expression of a failure of trust in politicians, nor of the operation of a paranoid imagination in the absence of the regulating virtues of critical thinking. It is an elaborate expression of thought about the nature of thought in the contemporary moment.

SJW’s Always Lie is substantially taken up with what Flavell called metacognitive theories. There are theories about the kinds of content that we can encounter: representations are either rhetoric or dialectic. Then there are theories about the kinds of thinking agents that apprehend this content: there are social justice warriors and plucky rebels such as Day, and this distinction is crosscut by the distinction between the three levels of aptitude: those who can understand rhetoric only, those who can understand dialectic only, and those who can switch between the two.

It seems to me that any attempt to understand Day and his followers in terms of universal characteristics of thought such as cognitive biases, or in terms of the transformation of information economies, will have missed a significant part of the picture and will fail adequately to comprehend the differences between the cynicism of the contemporary alt-right and other forms of cynicism. It is hard to say how much generalization will prove possible, but less elaborate versions of elements of Day’s scheme, such as the distinction between mainstream media and independent media are certainly very widespread. In order to understand the specificity of positions such as Day’s, with all its irony and instability, it is necessary to understand the specific configuration of theories of minds, content, and motives that underlie them. In other words, in order to explain their thought and action, it is necessary to produce an account of their own explanations of their own and other people’s thought and action, of their own culture of metacognition.

Understanding this aspect of post-truth, if it turns out to be widespread, will also be important for those who are concerned to counter it. The form of metacognition that Day is teaching could make his audience almost impervious to persuasion at the level of first-order content. Any challenging claim can be dismissed as rhetoric if its author can be identified as an SJW—in this sense it is a “closed predicament,” in Horton’s sense (1967), complete with processes of secondary elaboration. It seems to me that the system is more vulnerable to second-order criticism, that is, criticism that aims at the metacognitive theory.

For anthropology, there is an additional lesson. Our explanations of people’s actions are likely to be lacking if we do not pay adequate attention to those people’s own explanations, including their conceptualization and evaluation of mind, action, and content. Since metacognition is formally identical to anthropological theories of cognition, there is a danger that we will simply apply our own categories to the first-order content, overwriting local metacognitive schemas with our own favored theories. That would be a mistake, not because anthropological theory must be subordinated to the thought of anthropological subjects, as some ethnographic foundationalists would claim, but because the thought of any agent is significantly conditioned by his or her thought about thought.