Critique of Phenomenology and Castaneda's 'anthropology' by Professor Marvin Harris


Excerpt from Marvin Harris, Cultural Materialism: The Struggle for a Science of Culture. New York: Random House, 1979, pp. 315-324.

Obscurantism is a research strategy whose aim is to subvert the possibility of achieving a science of human social life. Obscurantists deny the applicability of scientific research principles to the study of divergent and convergent sociocultural phenomena. Their aim orientation is to increase rather than decrease the semblance of disorder in the sociocultural realm and to cast doubt on all existing scientific theories without providing plausible scientific alternatives.

All nonscientific strategies are not necessarily obscurantist. As I said earlier, there are domains of experience, knowledge of which cannot be achieved by scientific research. The ecstatic knowledge of mystics and saints, the visions and hallucinations of drug users and schizophrenics, and the aesthetic insights of artists, poets, and musicians are certainly not obscurantist merely because they are not based on scientific research principles. The issue of obscurantism arises only when knowledge obtained through nonscientific means is deliberately used to cast doubt on the authenticity of scientific knowledge within the domains suitable for scientific inquiry. To be obscurantist, in other words, a research strategy must be antiscientific rather than merely nonscientific.

On the popular level, obscurantism has acquired many of the features of a social movement. Standards, inclinations, and attitudes associated with a large number of convergent nonscientific interests implicitly or explicitly deny the feasibility or utility of a science of social life. Obscurantism is an important component in the emics of astrology, witchcraft, messianism, hippiedom, fundamentalism, cults of personality, nationalism, ethnocentrism, and a hundred other contemporary modes of thought that exalt knowledge gained by inspiration, revelation, intuition, faith, or incantation as against knowledge obtained in conformity with scientific research principles. Philosophers and social scientists are implicated both as leaders and as followers in the popular success of these celebrations of nonscientific knowledge, and in the strong antiscientific components they contain.

Phenomenological Obscurantism

One of the most fecund sources of contemporary obscurantist attitudes in the social sciences is phenomenology, the neo-Kantian philosophy founded by Edmund Husserl. Like other neo-Kantians, especially Heinrich Rickert, Wilhelm Windelband, and Wilhelm Dilthey, Husserl sought to draw a sharp line between the physical and social sciences (between the Naturwissenschaften and the Geisteswissenschaften or Kulturwissenschaften--i.e., between the natural sciences and the sciences of the human mind or culture). Husserl proposed that ordinary natural science cannot be applied to sociocultural life because social acts involve a property not present in other sectors of the universe--namely, the property of meaning. According to Husserl, meaning can only be understood subjectively. Hence to understand social acts, one must understand what they mean as a subjective, "lived experience." By assuming that the subjective experiences of others are similar to one's own, observers can draw analogies between their own intentions and goals and those of other actors and in this way begin to explain social life. Husserl's philosophy, transmitted through the writings of Alfred Schutz (1967), form the foundation for the cognitivist obscurantist strategies known as ethnomethodology and symbolic interactionism.

At the beginning of this century anthropology had already come under the influence of the neo-Kantian movement. Generations of Boasian fieldworkers accepted the phenomenological demarcation of the human sciences and saw their primary mission to be that of finding out how natives think. In an attenuated form, therefore, the emic bias of phenomenology has always been an integral part of idealist strategies.

[Note: Earlier in the book, Harris discusses the terms etic and emic, which were introduced into anthropology by the anthropological linguist Kenneth Pike: "Emic operations have as their hallmark the elevation of the native informant to the status of ultimate judge of the adequacy of the observer's descriptions and analyses. The test of the adequacy of emic analyses is their ability to generate statements the native accepts as real, meaningful, or appropriate. In carrying out research in the emic mode, the observer attempts to acquire a knowledge of the categories and rules one must know in order to think and act as a native. . . . . Etic operations have as their hallmark the elevation of observers to the status of ultimate judges of the categories and concepts used in descriptions and analyses. The test of the adequacy of etic accounts is simply their ability to generate scientifically productive theories about the causes of sociocultural differences and similarities. Rather than employ concepts that are necessarily real, meaningful, and appropriate from the native point of view, the observer is free to use alien categories and rules derived from the data language of science. Frequently, etic operations involve the measurement and juxtaposition of activities and events that native informants may find inappropriate or meaningless." p. 32]Phenomenology is also perfectly continuous with the Durkheim-Parsons-Weber notion of social action and with the main currents of cognitive idealism as discussed in Chapter 9. Phenomenologists agree with Talcott Parsons that social action and emic goals are indissoluble:

"As Parsons has emphasized, the very notion of an action requires the idea of the actor's end or purpose. That is, for an action to be perceived, purpose and meaning must be perceived. Thus a change in the perceived meaning or purpose entails a change in the action that is perceived." (Wilson, 1970:67, fn.)

Phenomenologists deny the possibility that etic behavior stream actions are worth studying independently of the actor's meaning or purpose:

". . . social actions are meaningful actions, that is, . . . they must be studied and explained in terms of their situations and their meanings to the actors themselves." (Jack Douglas, 1970:4)

Thus, phenomenology shares a common starting point with other idealist and emic approaches. But it relentlessly drives toward sectarian conclusions many idealists are not prepared to accept. Combining their commitment to the "lived experience" with an attack on positivism, phenomenologists reject the possibility of separating observers from the observed. Observation is itself to be approached as a lived experience in which the subjective meanings of both observer and participant are constantly "reflected" on. Moreover, the participant observer can never find the truth of the lived experience, apart from the consensus about such things found in the community in which the observer participates. Truths are always relative and social.

"Truth is never a feature of the sensations of a discrete individual; it is always to be recognized in the knowledge of members of communities. (Silverman, 1975:75) Truths are always recognized with (as) the system of intelligibility of a community. Truths are always for and within a community. . . ."(lbid..77)

At first reading, this assertion about the social nature of truth appears to be quite reasonable and innocuous. Who would deny that truths are always established in conformity with socially specified rules of accountability and significance? Science itself is clearly nothing but a "system of intelligibility of a community." But in the strategy of phenomenology the social nature of truth has profoundly obscurantist implications. Since phenomenology equates social action with the emics of mental and behavioral phenomena, thereby denying the knowability of etic behavioral and mental events, the insistence on the social nature of truth reduces to the proposition that sociocultural theories are true only to the extent that they are reflections of the "system of intelligibility" of the people being studied. This differs from and is antagonistic to the cultural materialist strategy for coping with the social nature of scientific theory. Cultural materialism grants that scientific truth is a social product, but it denies that the corpus of scientific theory necessarily differs from culture to culture. The community that establishes the authenticity of scientific theories is not a community of participants in any given culture, but rather the transcultural community of scientific observers. For this community, as for the communities in the phenomenologist's formula, emic truths must be viewed relativistically, altering with each culture's system of intelligibility. But for this community the realm of sociocultural truths is not exhausted by emics; there are also etic truths, and these do not alter in conformity with each culture's system of intelligibility. Rather, they alter only in conformity with the agreed-upon data collection and theory-testing procedures of the community of scientific observers.

Thus phenomenology, like other varieties of cultural idealism, conflicts with cultural materialism because phenomenologists deal only with emic phenomena. But the conflict goes deeper and is less amenable to solution than in the case of cognitivism, since the phenomenotogists insist that the etic behavior stream is unreal or completely subordinate to the reality of each culture's system of knowing.

Inherent in this approach is a capacity for boundless confusion and deception about the nature of human social problems. Whereas other idealist strategies merely ignore or misrepresent the causes of poverty, sexism, and other central dilemmas of human social life, phenomenological idealism denies that such causes exist. By reducing and confining all sociocultural events to the motivations and plans of immediate experience and communal consensus, phenomenology eventually even leads to the denial of the existence of sociocultural systems and of the universal components of such systems (such as infrastructure, structure or superstructure). For many phenomenologists, such things as ruling classes, imperialist powers, capitalism, or socialism also have no existence apart from the communities of participant observers who happen to believe in them. Processes such as evolution, adaptation and exploitation are also unreal. Phenomenologists dismiss such entities and processes as "reifications." The only social reality worth talking about is the everyday lived experience in which individuals encounter one another and interact in terms of arbitrary symbols and conventional meanings. The task of social "science" is to penetrate and explicate these symbols and meanings, nothing more.

The Phenomenology of Don Juan

Carlos Castaneda's widely read books about the alleged lived experience of Don Juan, a Yaqui Indian shaman, exemplify the obscurantist consequences of phenomenology. At the University of California in Los Angeles, Castaneda studied under the ethnomethodologist Harold Garfinkel, in turn a student of Alfred Schutz--mentioned above. (Garfinkel [1967], who was on Castaneda's dissertation committee at the University of California at Los Angeles, is famous for experiments designed to prove that the essence of social reality consists of conventional meanings attached to everyday activities by communal consensus. The experiments consisted of having students board buses and decline to pay the fare, or having them go home and sit at the dinner table and refuse to pass the salt.) Inspired by his phenomenological mentors, Castaneda resolved to do fieldwork that would involve him in the symbols and conventional meanings of a lived experience entirely different from that of Western social reality.

The Yaqui Indians provided Castaneda with a suitably exotic context for studying the "separate reality" of another culture, especially as he singled out and sought to penetrate and participate in the the most exotic aspect of this culture--the activities and thoughts of the community of Yaqui sorcerers and shamans. To a certain point, therefore, Castaneda's phenomenological journey merely qualifies as a typical cultural idealist study of the mental superstructure. An exclusive preoccupation with mental and emic superstructure adds up to an ineffectual, stunted, and scientifically undesirable strategy, but it does not necessarily add up to an obscurantist strategy. The obscurantism of Castaneda's approach arises from his presentation of the emic reality associated with shamanic consciousness as a challenge to the legitimacy of the epistemological principles upon which science is based.

Castaneda reports that Yaqui shamans believe they can fly through the air, change into animals, kill an adversary by sorcery, and see through opaque objects. None of this is news. Many anthropologists have provided vivid accounts of shamanistic exploits without becoming national celebrities and without being accused of obscurantism. Castaneda's account differs from the others because he tells his story from the "inside," deliberately letting the emics and his own subjective feelings dominate the narrative. The aim of the narrative is to get the reader to participate in the shaman's system of intelligibility and thereby to demonstrate that reality is the creature of social consensus. If we can be persuaded to participate in the shamanic consensus, we will believe that shamans can fly. (Just as we will believe drug-induced hallucinations when they are happening to us.)

The influence of Garfinkel's phenomenology is apparent in the little-read technical addenda to Castaneda's first book, The Teachings of Don Juan. Castaneda here portrays his apprenticeship to Don Juan as a search for the validating consensus that converts the nonordinary component element of his experiences from illusion to reality. (In other words, if two people have the same fantasy, it is no longer a fantasy.) Since these nonordinary component elements were not subject to ordinary consensus, their "perceived realness" would have been only an illusion if he had been incapable of obtaining agreement on their existence. For Castaneda, the "special consensus" came from the sorcerer himself:
"In Don Juan's teachings, special consensus meant tacit or implicit agreement on the component elements of non-ordinary reality. . . . This special consensus was not in any way fraudulent or spurious, such as the one two persons might give each other in describing the component elements of their individual dreams. The special consensus Don Juan supplied was systematic. . . . With the acquisition of the systematic consensus the actions and the elements perceived in non-ordinary reality became consensually real. . . ." (1969:232)

Eventually, through a process by which Don Juan put Castaneda into the proper state of mind, hundred-foot gnats and man-sized butterflies ceased to be illusions. They became instead another reality--another ordinary reality--for "the classifications 'ordinary' and 'non-ordinary' [became] meaningless for me":

"there was another separate but no longer unordinary realm of reality, the 'reality of special consensus.'" (lbid.:250)

To expose the flaw in Castaneda's phenomenological exercise, I should like to compare the expository technique of the Don Juan books with that of an even more compelling phenomenological account rendered in another medium; that of the classic Japanese movie Rashomon. In this movie the viewer witnesses four different versions of the "same" scene. The principal actors are a man, his wife, a stranger, and an onlooker bidden in the bushes. Each of the actors narrates a different version of the lived experience, and each version appears on the screen as the lived reality. Manly heroism in one version is abject cowardice in another; chastity in one is carnal heat in another; magnanimity in one is brutality in another; and so on. Each narrative unfolds as a graphic, vivid reality, and the audience is left on its own to decide which version, if any, actually represents the event--or indeed, if there ever was an "event" to begin with.

For a cultural materialist there are only two possible solutions of Rashomon's contradictions and ambiguities: one of the versions is etically correct and the others are false; or they are all etically false. For the phenomenologist there is a third solution: all versions are equally true. This third possibility arises because in the phenomenological strategy, there is no way to distinguish etic from emic events. If the participants are not lying, then what they saw within their system of intelligibility must be accepted as true.

The very fact that Rashomon (or Don Juan) can be presented as a problem of multiple truths, however, proves that the problem of which version, if any, is true can be solved. In order to convince the audience that truth is relative to consensus, the film maker actually obtains a consensus concerning the truth of what the camera sees during each version. The camera shows vividly and unequivocally that there is a seduction, a rape, a murder, a duel, and so forth. These events are the analogues of etic events in the strategy of cultural materialism. Since it is possible to obtain a consensus about what happened in each episode, even though each version contradicts the others, one must conclude that a film maker could have filmed the actual event and achieved the same kind of consensus. Such a film would not constitute the whole truth, but it would provide us with a sound basis for deciding which one of the other versions was most nearly correct or whether they were all equally false. Except by trickery or incompetence, a camera could never show them all to be equally true.

Of course I am aware that a filmed version of an event involves selective viewing, and that interpretations of pictures, like interpretations of lived scenes, are influenced by a person's total perceptual and cognitive framework. Yet one does not need to obtain the total and absolute truth about a scene in order to refute the obscurantist claim that contradictory versions of scenes may all be equally true. The fallacy involved here is a variant of the search for empirical certainty discussed in Chapter 1. It does not follow from our inability to obtain absolutely certain knowledge that all knowledge is equally uncertain. By using recording devices under explicitly operationalized conditions, the community of scientific observers can get closer to what happened etically even though they may never get to the absolute final truth. Cultural materialism is committed to getting closer and closer to this etic reality: phenomenology is committed to getting further and further away from it.

Again: "What Does It Matter?"

No one can object to Castaneda's artful presentation of the different reality of shamanic consensus. Unfortunately, however, his attempt to get closer to the emics of the shamanic world is shackled to a mischievous attempt to mystify what was happening while he was cultivating the shamanic consciousness. In fact, there is so little about the etic who, what, when, where of Castaneda's experiences in his books that substantial doubts arise as to whether or not Don Juan exists--doubts which Castaneda has never taken the trouble to dispel (Time, 1973; Harris, 1974:246ff; Beals, 1978; New West, January 29, 1979).

Internal inconsistencies in the chronologies of the earlier and later volumes, the absence of a Yaqui vocabulary, the close parallel between Castaneda's visionary experiences and those reported in other works on shamanism, testimony of his ex-wife, friends, and colleagues, and Castaneda's failure to defend himself against the accusation that he deceived his Ph.D. committee at UCLA, make it very unlikely that Castaneda was ever an apprentice to Don Juan (De Mille, 1976). This is not to assert that Castaneda's knowledge of shamanism in a more general sense is defective, nor that his vivid descriptions of shamanic consciousness are without redeeming value. Castaneda probably has much first hand as well as literary knowledge of shamanic practices, and he has communicated that knowledge in a uniquely effective manner. The only problem is that without the etic context, we do not know whose system of intelligibility is represented. We cannot rule out the possibility that Castaneda never interviewed any Yaqui Indian shaman, and that the apparent authenticity of his shamanic experiences derive entirely from his own shamanic gifts and literary and imaginative powers.

Since Castaneda shows no interest in defending himself against this speculation, many of his admirers have been obliged to consider the question of whether it matters to them if the Don Juan stories are fact or fiction. Professor David Silverman (1975:xi), lecturing in the Department of Sociology at UCLA, had no problem disposing of this issue: "It does not matter to me in the least whether any or all of the 'events' reported by Castaneda ever 'took place,'" just as it does not matter to Levi-Strauss if his "book on myths is itself a kind of myth" (p. 169). Levi-Strauss rationalizes his indifference to fact or fiction on the basis of his conviction that his own mind works the same as any Indian's mind; Silverman rationalizes his indifference on the ground that any phenomenological account is interesting in its own right. Castaneda's books are a phenomenological account or a "text." Since truth is always relative to a system of intelligibility, there is always an "invented" imaginative, or fictional component in such "texts." "What text is not a construction?" asks Silverman.

Going one step further, the novelist and literary critic Ronald Sukenick sees everything that happens as a story or a story of a story, and so forth. And stories are "neither true nor false, only persuasive or unreal." This has been the great revelation inspired equally by Zen, the Book of the Dead, witch lore, Sufism, various Eastern disciplines, Western mystical tradition, Jungian speculations, Wilhelm Reich, and Carlos Castaneda:

"All versions of 'reality' are of the nature of fiction. T'here's your story and my story, there's the journalist's story and the historian's story, there's the philosopher's story and the scientist's story. . . . Our common world is only a description . . . reality is imagined. . . ." (Sukenick, 1976:113)

This invitation to intellectual suicide returns us full circle to the epistemological anarchism of Paul Feyerabend (another one of Castaneda's admirers). The rebuttal of Sukenick, like the rebuttal of Feyarabend (see p. 22), comes in two parts, intellectual and moral. Let me attend to the intellectual part first. Does Sukenick seriously believe that all versions of reality are fictions? If so, then he believes his version of reality is a fiction. Since he believes that everything he says is a fiction including what he says about reality, only a fool would believe such a man about anything.

More astonishing than the intellectual obscurantism of phenomenology is its moral opacity. Morality is the acceptance of principled responsibility for the way in which one's actions or lack of actions affect the well-being of other members of the human species. The absolute precondition for any kind of moral judgment is our ability to identify who did what to whom when, where, and how. The doctrine that all fact is fiction and that all fiction is fact is a morally depraved doctrine. It is a doctrine that conflates the attacked with the attacker; the tortured with the torturer; and the killed with the killer, It is true that at Dachau there was the SS's story; and the prisoners' story; and that at Mylai, there was Calley's story and there was the kneeling mother's story; and that at Kent State there was the guardsmen's story and the story of the students shot in the back, five hundred feet away. Only a moral cretin would wish to maintain that all these stories could be equally true.

A Response to Harris's Critique