In Religion in Human Life, Edward Norbeck (1974:6) observes that “religion is characteristically seen by anthropologists as a distinctive symbolic expression of human life that interprets man himself and his universe, providing motives for human action, and also a group of associated acts which have survival value for the human species.” Various formulations could be subsumed under that general description, such as Lessa and Vogt’s (1972:1) notion that “religion may be described as a system of beliefs and practices directed toward the ‘ultimate concern’ of a society,” or Geertz’s (1973:90) concept of religion as “a system of symbols” that integrates a culture’s world view and ethos. Those definitions, however, could logically embrace existentialism, communism, secular humanism, or other philosophies which most anthropologists would be reluctant to call religion. How then is religion distinguished from comparable sets of beliefs and behaviors that fulfill similar functions?
As Norbeck (1974:6) explains, “the distinguishing trait commonly used is supernaturalism, ideas and acts centered on views of supernatural power.” The concept of the supernatural has been firmly tied to the anthropological definition of religion since the origins of the discipline. Edward Tylor (1958:8), for example, argued that “it seems best…to claim, as a minimum definition of Religion, the belief in Spiritual Beings.” Frazer (1963:58) maintained that “religion involves, first, a belief in superhuman beings who rule the world, and, second, an attempt to win their favour.” Malinowski (1954:17) observed that sacred “acts and observances are always associated with beliefs in supernatural forces, especially those of magic, or with ideas about beings, spirits, ghosts, dead ancestors, or gods.” The concept of the supernatural continues to dominate anthropological conceptions of religion today. Marvin Harris (1989:399), for example, declares that “the basis of all that is distinctly religious in human thought is animism, the belief that humans share the world with a population of extraordinary, extracorporeal, and mostly invisible beings.”
There is a fundamental problem with the term “supernatural,” however: it is so varyingly conceived in the different cultures of the world that it lacks a common, unambiguous definition. The Yanomamo, Roman Catholic, !Kung San, and Buddhist conceptions of the “supernatural” realm, for example, are widely divergent and even contradictory in some aspects. The problem is that the term “supernatural” is an emic concept, meaning that it is defined in terms of the categories and concepts regarded as meaningful and appropriate by the members of particular cultures; it is not an etic concept, one defined in terms of the categories and concepts regarded as meaningful and appropriate by the community of scientific observers (Lett 1990). As an emic concept, the term “supernatural” has as many definitions as there are cultures; as an etic concept, it has no recognized, agreed-upon definition.
Nor could any such objective, scientific definition be offered for the term “supernatural,” for the simple reason that the word is propositionally meaningless. The term “supernatural” is purportedly used to designate a reality that somehow transcends the natural universe of empirical reality, but what does it mean to “transcend empirical reality?” If such a thing as “nonempirical reality” exists, how could we, as empirical beings, even know about it? (Revelation and intuition, after all, are demonstrably unreliable-witness the mutually exclusive claims to knowledge made by different people on revelatory grounds.) If such a thing as “nonempirical reality” exists, by what mechanism is it connected to empirical reality? (How, in other words, do supernatural beings and forces have an impact on the natural world?) Further, if such a thing as “nonempirical reality” exists, why is there not a single shred of objective evidence to indicate its existence? As the physicist Victor Stenger (1990:33) points out, there is no rational reason whatsoever to even hypothesize the existence of the “supernatural:”
At this writing, neither the data gathered by our external senses, the instruments we have built to enhance those senses, nor our innermost thoughts require that we introduce a nonmaterial component to the universe. No human experience, measurement, or observation forces us to adopt fundamental hypotheses or explanatory principles beyond those of the Standard Model of physics and the chance processes of evolution.
The term “supernatural” thus purports to describe a reality that we could not know or recognize, one that could not have any impact on the reality we do know and recognize, and one for which we have no evidence whatsoever; it is, in short, unintelligible. The philosopher William Gray (1991:39) eschews the term “supernatural” and suggests instead that religious statements can be described as “metaphysical,” by which he means statements that refer to facts that could not possibly be observed. But what would an “unobservable fact” be? To substitute “metaphysical” for “supernatural” is simply to play a semantic game. Terms such as “supernatural,” “metaphysical,” and “nonempirical reality” are, in fact, oxymorons. It would make just as much sense to talk about the “unreal real.”
Connotatively, the term “supernatural” presents additional problems: it is not sufficiently comprehensive to embrace beliefs and behaviors that are virtually identical in form and function to so-called “religious” beliefs and behaviors, but which would not commonly be called “supernatural.” Gods, demons, angels, and souls, for example, could easily be called “supernatural,” and so too, perhaps, could incubi, succubi, ghosts, goblins, fairies, sprites, trolls, and leprechauns. But what about witches, clairvoyants, telepathists, psychokinetics, extraterrestrials, psychic surgeons, vampires, werewolves, spirit channelers, fire-walkers, astrologers, the Loch Ness Monster, and Sasquatch? Would those too be called “supernatural?” Would anthropologists call beliefs in such beings and forces “religious?”
At least one recent anthropological text on religion recognizes this problem. In Magic, Witchcraft, and Religion, Lehmann and Myers (1989:3) argue that it is time for anthropologists to abandon the restrictive connotations of the term “supernatural:”
Expanding the definition of religion beyond spiritual and superhuman beings to include the extraordinary, the mysterious, and unexplainable allows a more comprehensive view of religious behaviors among the peoples of the world and permits the anthropological investigation of phenomena such as magic, sorcery, curses, and other practices that hold meaning for both pre-literate and literate societies.
Lehmann and Myers fail, however, to suggest an alternative term to replace the word “supernatural.” Fortunately, there is an obvious alternative available, one that is winning increasing acceptance both inside and outside anthropology, namely the word “paranormal.” (Note 5) The term refers ostensibly to phenomena that lie beyond the normal range of human perception and experience, although in practice it does not denote simply anomalous phenomena. Instead, it describes putative phenomena whose existence would in fact violate the rules of reality revealed by science and common sense. From an etic point of view, therefore, the notion of the “paranormal,” like the notion of the “supernatural,” is propositionally meaningless. Unlike the term “supernatural,” however, the term “paranormal” is not restrictive in its connotations, and that is its principal advantage. “Paranormal” is a useful umbrella label for the complete set of emic beliefs concerning the unreal real. The term embraces the entire range of transcendental beliefs, covering at once everything that would otherwise be called magical, religious, supernatural, metaphysical, occult, or parapsychological.
Therein lies the real common denominator in all paranormal beliefs: not that they are all “supernatural,” but that they are all irrational, by which I mean that every single paranormal belief in the world, whether labeled “religious,” “magical,” “spiritual,” “metaphysical,” “occult,” or “parapsychological,” is either nonfalsifiable or has been falsified. (The vast majority of all paranormal propositions-such as the Judeo-Christian proposition that “God” exists-are nonfalsifiable and hence propositionally meaningless; a smaller percentage-such as the Judeo-Christian proposition that a universal flood covered the earth sometime within the past 10,000 years-are falsifiable but have invariably been falsified by objective evidence.)
The simple fact of the matter is that every religious belief in every culture in the world is demonstrably untrue. Regardless of whether the religious practices are organized communally or ecclesiastically, regardless of whether they are mediated by shamans or priests, regardless of whether the intent is manipulative or supplicative, the one constant that runs through all religious practices all over the world is that all such practices are founded upon nonfalsifiable or falsified beliefs concerning the paranormal.
Irrationality is thus the defining element in religion. Religion and science are not at odds because religion wants to be “supernatural” while science wants to be “empirical;” instead, religion and science are at odds because religion wants to be irrational (relying ultimately upon beliefs that are either nonfalsifiable or falsified), while science wants to be rational (relying exclusively upon beliefs that are both falsifiable and unfalsified).
I am aware that many anthropologists are likely to react negatively to the pejorative connotations of the word “irrational.” The term, however, is simply descriptive and therefore entirely appropriate. It is unarguably irrational to maintain a belief in an allegedly propositional claim when that claim is either propositionally meaningless or has been decisively repudiated by objective evidence. Whether it is laudable or forgivable to do so is another question: it is not, of course, a factual question, but neither is it a question that scientists can entirely avoid.