Hoyt Alverson
Professor of Anthropology
Dartmouth College

http://www.dartmouth.edu/~anthro

(This Guest Editorial first appeared in 1987)

"Science" as typically understood is based on three assumptions: (1) there is a "real" world with an order that is separate from human interpretations of that order; (2) both the "real" world and human interpretations of the real world independently affect human experience, and (3) the effects of the "real" world can be represented by logically rigorous but otherwise relatively simple symbolic languages known as truth conditional languages..

These assumptions are just that—assumptions: they cannot be proven. They have guided scientific inquiry, and their use has led to models of reality that have been very successful in accounting for human experience. But the success of scientific models is not the same thing as absolute truth, which would depend on proving the assumptions themselves true. Scientists are willing to accept that some models successfully account for experience in the long run because they represent what is real, and, conversely, that other models fail because they misrepresent reality. This is what is meant by "proof" in science.

We can now ask the crucial question: is a science of culture—a science of anthropological nature—a possibility? If we answer yes, we are committed to setting severe limits on what culture—that is, on what being human—may be conceived to be. Regarding the first assumption, culture in general, including its changes, evolution, and particular geographical or historical manifestations, would have to exhibit an invariable determinate structure. With regard to the second assumption, the effects of this "structure of culture" upon human experience would have to be separable from any effects resulting from the biological and cultural processes of perception and cognition. With regard to the third assumption, this "structure of culture" would have to be represented by a scientific model using a logically rigorous yet simple symbolic language.

For several reasons, many anthropologists argue that the assumptions of science concerning the conditions that reality must meet to be suitable for scientific explanation make a science of culture impossible and paradoxical. (1) The world’s cultures are simply too varied and too dependent on environmental and historical contingencies to be reduced to a single invariable structure. (2) It is practically and logically impossible to separate the structure of culture from the biological and cultural processes of perception and cognition, by means of which human beings know anything at all (that is, a species that is defined by culture cannot pry itself away from culture in order to behold it as a spectator). (3) Many anthropologists, and others, claim with good evidence that because of the species-specific endowment of rational thought and language, human beings are genuinely "free," that is, their actions are not determined by specific processes and forces. (Structure, by contrast, is understood to determine acts or events, even if only probabilistically). If the actions of human beings are potentially or actually not determined by outside phenomena there is no sense in which culture could be reduced simply to structure. Freedom (not to be confused with randomness or uncertainty) means simply that other phenomena do not have even a probabilistic effect on people’s actions.

If these objections are well founded, the most one can hope for in the study of culture, or of any human behavior, is a phenomenology—a continuous description or interpretation of what observed culture means to its participants and to its observers, anthropological or otherwise. Any study of what something means presupposes and requires the culture-bound, indeed person-bound, experience of a human subject. Meaning only exists relative to people and experience. The structure that science presupposes is presumed to exist independently of all people and any experience. Those who seek a science of culture are appropriately called "universalists" or "antirelativists." Those who seek a phenomenology of culture are often known as "relativists." A third point of view is that some aspects of culture exhibit structure in the sense set forth here, while other aspects are considered undetermined, free. This "dualist" position will often assert that what people do is structured—determined—but that their thoughts and their ideas about right and wrong are free, or that "the economy obeys laws, but individuals have free will." Such dualism is essentially a confession of ignorance about the limits of determinism and freedom.

Intellectual honesty and the practical demands of research encourage people to choose sides—to be either a relativist, committed to the interpretive view of culture, or a universalist, committed to belief in the existence of structure and the explanation of it. The common-sense, dualist position is not commended by serious thinkers. Its value is largely therapeutic, masking the dissonance that comes from not knowing what to believe.

It should by now be apparent that one’s assumptions concerning the existence of structure in culture, or the existence of freedom in human action, determine whether one believes that there can be a science of culture or not. Note that the possibility of developing a science of culture has nothing to do with the use of mathematics, the precision of one’s assertions, or the elegance of one’s models. If a phenomenon actually has structure, then a science of that phenomenon is at least conceivable. If a phenomenon exhibits freedom, is not necessarily ordered, then a science of that phenomenon is inconceivable. The human sciences, including anthropology, have been debating the issue of structure versus freedom in human cultural behavior for the past two hundred years, and no resolution or even consensus has emerged.

Some persuasive models of culture, and of particular cultures, have been proposed, both by those working with scientific, universalist assumptions, and by those working with phenomenological, relativistic assumptions.

To decide which of these approaches is to be preferred, we must have a specific set of criteria for evaluation. Faced with good evidence for the existence of both structure and freedom in human culture, no coherent set of criteria for comparing the success of these alternative models is conceivable. The prediction of future action, for example, is a good criterion for measuring the success of a model that purports to represent structure; it must be irrelevant to measuring the success or failure of a model that purports to describe freedom. For the foreseeable future, and maybe for the rest of time, we may have to be content with models that simply permit us to muddle through.

Return to Schultz/Lavenda Index