by Hugh M. Lewis


Anthropology is not constructed of cards upon clouds.


Anthropological knowledge is characterized by four foundational principles--holism, comparative research, relativism and empiricism. Whatever else anthropology may or may not be, it can never escape the consequences of any of these four concepts and still remain genuine anthropological science. These are implicit presuppositions of all anthropological method and theory, and they reflect together a general attitude and approach to human reality that is characteristically anthropological.

It nevertheless remains true that the four precepts stand in uneasy relation, and in frequent opposition, to one another. A comparative approach often demands an analytical orientation that may defy the attempt at holistic comprehension, while the attempt at holism sometimes runs counter to empirical experience. Relativism and comparativism have always been cast in dialectical opposition to one another, even though each implies and necessitates the other.

No matter how native an anthropologist may become in the process of his learning about another culture, the anthropologist nevertheless continues to compare the foreign culture with his own, even if on an unconscious level or in a tacit manner without her/his awareness of it. Comparison is, on a basic, unmarked level, part and parcel of all ethnography--even the most emically contrived and subjectively delivered. Comparativism is inherent in the structural relationship between the anthropologist as observer and the "native" as observed. It is by comparison that anthropological understanding is achieved, and without it there could be no systematic construction of a science of human differences and identities.

Paradoxically, comparativism requires relativism--comparison cannot be realistic or successful without shedding our own ethnocentric biases that preclude our attempt to understand others. Relativism allows us to embrace other realities in such a way as to make the possibility of comparison, and by implication, translation, more of a realistic likelihood.

In a similar way, holism stands in contrast to comparativism and is required as well by relativism. Relativism entails that the relationships prevailing in a situation are somehow intermeshed in a way that cannot be easily separated. We speak of cultural integration of different institutions, practices and beliefs and see that change in one area may result in reverberations in many other areas. Similarly, we seek explanation for a particular cultural phenomenon not in terms of a single proximate cause, but in terms rather of a background of conditioning interrelationships which make the phenomena a possibility in the first place.

A strong emphasis upon empiricism entails that holism cannot be a purely rational affair of eidetic conceptual structures removed from the source of their experience in reality. We can generalize about basic structures of human action or social behavior based upon a handful of assorted examples, but when we hypothesis at a higher level of generality we cannot be too overly confident that our limited samples are enough, or are representative of the universal pool of experience. It is this single feature which keeps anthropology a science tethered closely to the ground.


A strong and strict empiricism has been one of the unacknowledged cornerstones of scientific anthropology's theory and method. Whether we are noting etic blinks or understanding emic winks, it is the act of encoding experience, a first-hand encounter with the world, and the primary weighting of this source of information over all others, which is the hallmark of anthropological research and which makes ethnography scientific. Our own theoretical or operational rationale does not matter so much, whether we are materialists or idealists, as does the consideration that we are to some extent, in some imperfect manner, engaged in the world, involved with it, seeking to learn about it first-hand.

Perhaps this empiricism has been so obvious in anthropology that its implications have mostly been taken for granted. Almost everyone would agree on the value of doing sound empirical research, although many would disagree on exactly what constitutes sound empirical research. And it does not matter what kind of anthropologist, all would hold as of primary importance first-hand observation in anthropological research.

To call anthropology an empirical science is not to claim that it is necessarily inductivist versus deductivist, or that it has its origins in the empiricist philosophers. It is this primary valuation of empiricism that anthropology as a science shares with all other genuine sciences. Anthropology, like any other science, must derive its knowledge and understanding from the world first and foremost.

The empiricism of anthropology and of science in general is largely an epistemological and an ontological proposition about reality and how we can know the world. It holds scientific knowledge to be essentially presentationalist, as opposed to representationalist, and that all such knowledge is derived primarily from the perceptual experience with the world which is independent of our awareness, separate from our own subjectivity.

Such a claim does not deny the fact of subsequent distortion of knowledge, or the secondary bias and reinterpretation of experience by representationalist structures of the mind, or that our experience of the world, even in its primary form, is constrained in critical ways by the limitations of our senses and the prejudices of our awareness. But the possibility remains of a relatively unbiased, "objective" view of reality which it is science's purpose to achieve, and even though this view itself might be from a theoretical point of view quite representationalist and derivative, it is held to remain fundamentally rooted to the basic percepts of common human experience.

This brings to bear another significant point about the empirical basis of anthropology--that is that all such basic knowledge must be "public" knowledge. It must be open to view and subject to multiple evaluations and the critical process. It must be communicated, or at least communicable, and it must be shared, or capable of being shared, between many different people. It cannot be privileged knowledge available to only a select group of people--all women, only to Americans, black only--nor can it be secret or specialized knowledge privy only to an insider's access.

It is precisely because the phenomenon that it reports upon are open to public scrutiny, that such knowledge can claim the possibility of being presentationally real and empirical. It must also be the case that all such knowledge must remain open to evaluation and scrutiny.

This brings to bear a basic paradox of the inductive character of science as an empirical enterprise--no matter how many white swans we may count, we can never be finally satisfied that the next swan will also be white. In such an inductively based enterprise, we can never take the last logical leap of faith in our ascent from the particulars to the general. Whereas deductive approaches can fall back upon the power of logic, the best science can hope to do is to provide a statistical measure of relative certainty. But the virtue of this is that our deductive rationality is liable to fail us where and when our statistical methodology will not.

Much lip service has been paid to the hypothetico-deductive method, especially in the advancement of the natural sciences, without recognizing that the whole hypothetico-deductive approach must still rest upon an empirical substrate in science. The first and last referents of any hypothetico-deductive approach must remain empirical, or empirically derived, if the approach is to have any broader scientific relevancy.


Just as the empirical spirit and nature of science has so often been over looked, so also has the basic instrumentality of science and the role of scientific instrumentality in the advance of scientific knowledge been largely left implicit and taken for granted.

By the instrumentality of science is meant something different than "scientific instrumentality"--though instrumentality for both carries a larger set of connotations which describes a general orientation and overall approach within science. The instrumentality of science refers to its practical orientation and emphasis upon real, measurable results. If it works then its probably scientific. According to Thomas Kuhn, the measure of scientific advance and of paradigmatic unity for a field is its achieved progress written in terms of its past track record. But in this we must separate the idea of an empirically based instrumentality from the kind of ideological instrumentality which Gadamer treats and which underlies such sciences as capitalist or Marxist economics--the self-fulfilling prophecy of historical scientific demonstration after the fact. The instrumentality of science must be demonstrated in a field of counterfactuals--of contenders that rise to discredit and disclaim a theory. In other words any theory must successfully compete against other alternative contenders. Kuhn also refers to the "puzzle-solving" character of the scientific enterprise. The instrumentality of a science is its ability to successfully solve the problems which are defined by the scientific community as relevant, and these solutions are governed by a kind of selective regime in which there tends to be a more correct solution, which solves more problems or more sides to a problem, than other contenders.

 The correctness of such solutions can only be seen as limited case solutions which exist temporarily within the purview of the current or predominant scientific paradigm. New information encountered upon the margins of the paradigm accumulates to undermine the correctness of the current received scientific explanations, and open the field for yet another round of competition and debate. The selection of a newer theory will rest upon its demonstrated instrumentality in being able to account for and accommodate the newer exceptions to the older paradigm. The other side of the paradigmatic coin is the instrumentality of received, tried and true shared exemplars--those archetypical test cases upon which a particular theory rests as a demonstration. From a theoretical point of view, there is nothing which is prototypically or monolithic about a shared exemplar. It is rather a polythetic set of examples or models which serve as heuristically powerful and productive orienting models for subsequent scientific research. One common characteristic of shared exemplars seems to be their heuristic productivity in generating a host of derivative corollaries, spin-offs, questions and solutions to minor problems which becomes part of the footwork of a subsequent generation of normal scientists.

The idea of shared exemplars brings into focus the notion of scientific instrumentality. New ideas, discoveries, inventions and techniques allow us to see things in new ways, and to rethink old problems. Often new advances must await the development or refinement of new means for seeing or observing reality at a distance. The indirect observation and measurement of discrete phenomena, beyond the bounds of normal human-sized perception, has had to await the development of techniques and technologies which have pushed back the frontier of our perception. Exceptions our counterevidence which accumulates upon the periphery of our predominant paradigm often does so because of the detailed refinement of such scientific instrumentalities, and the collapse of the old paradigm and the possibility for a new way of seeing reality is made possible because of the same refinements in scientific instrumentalities.

The instrumentality of science and scientific instrumentality must be seen as going hand in hand in the achievement of progress in science, and are thus two sides of the same scientific coin of instrumentality in the general sense of the kind of practical, pragmatic, common sense, result-oriented, technical habits of the scientific mentality. Put another way, what works in science is what proves to be instrumental in the advancement and production of scientific knowledge--that which is not scientifically instrumental is usually regarded merely as trivial.

It has been this instrumentality of science that has guaranteed that its practitioners must be something more than ivory tower armchair philosophers of the world. It placed a premium upon empirical demonstration and upon the heuristic productivity of knowledge. In this regard, we cannot clearly separate the philosophy, or the theory, or ideology of science, from its praxis or its practical side.


The perennial quest for universals underlying human culture has been one of the primary theoretical concerns of anthropological science. Different kinds of universals have been proposed, some a little less arbitrary than other kinds. But no completely satisfactory set or theory of such universals has yet been discovered or defined in spite of decades of cross-cultural research, and for many the search, perhaps misled from the very beginning, has been all but abandoned.

And yet there remains something intuitively and hauntingly intriguing and reasonable about the concept of human universals--something strong and stable upon which we could build our science of humankind.

The notion of universals comes squarely up against the paradoxes of cultural relativism, and anthropological science as a quest for universals and anthropology as a relativistic orientation toward humanity constitutes a fundamental contradiction within anthropology. On one hand, we are supposed to seek an insider's perspective in the appreciation of foreign world views and values with the understanding that there may be no set of universally shared values or world-views free of cultural bias. On the other hand we are told to find a common language of translation, a universal language of anthropologese, that will allow us to adequately reinterpret the cultural codes of any point and place into our own language of choice, without loosing anything which may be of substantive, indeed vital, importance. As anthropologists, we are supposed to be able to entertain some privileged point of view, a "bird's-eye view," of the bottom line in human reality. Inherent to the very act of crossing critical cultural boundaries to do long-term participant-observation is the belief that something of scientific value, something general if not partially universal, can be carried away from the experience. And yet the more the field experiences accumulate, the more aware we are made of all the exceptions to the general rules which once so monolithically framed our science.

If we are to look for, and eventually find, human universals, then we must have some reasonable, if not quite realistic notion, of what it is we should be looking for and what we may or may not actually find.

If we start with a few basics, like language is a human universal, or likewise the human capacity for culture, its acquisition, elaboration, development and transmission, is sort of another human universal, the most we seem to have accomplished is to emphasize the self-evident. Similarly, we can stake a claim to certain universal cognitive structures or even something as indeterminate as rationality, we do not move much further in the direction of explaining the foundation of human reality except to place a priority of mind over matter. Of course, humankind is also a symbolic species, as well as a tool-making and tool-using. And humankind everywhere has long had the same basic sets of needs--nutrition, sexual reproduction (some would even add to things like "love, societas, even intellectual stimulation to the list). Humans share a similar basic range of emotions, aggression, and prejudice. And yet, if we tally the list, what kind of scientific formula are we finally left with.

First we can safely say that there is no such thing as an exceptionless universal. For every principle we come up with, some exception to the rule, can be found. This does not disqualify the universal principle in anyway, but only sets it within a reasonable scope of a bounded, and therefore scientific, generality. If we presume things to be universally, exceptionlessly true, then we are setting it beyond the purview of our scientific control.

Not being an exceptionless universal, does not thereby make it any less valuable to anthropology.

There is another idealist point of view which sees the concept of culture as the metaphor of our own making which we apply to certain kinds or qualities of human phenomena on the basis of certain tacit presuppositions of cultural difference and identity. Culture does not exist out there somewhere so much as it is our own invention, itself a part of a certain anthropological theory about human reality, a conceptual category into which we fit certain varieties of experience.

On one hand, universals cannot be so empirically self-evident as to beg the obvious--all human have two hands, so many teeth, so many bones, etc. On the other hand, universals cannot be so rationalistically self-defining as to be tautological--culture are the things that compose culture, the things that compose culture are universal--for instance the definition of universals as the needs or institutions which compose culture.

What we are searching for in terms of universals are bridging, or mediating principles or theories, which link the concrete of empirical phenomena with the abstractions of our own definitions. We are looking for such middle-range theories that will serve as adequate, broad-based predictors of certain social phenomena. We are not talking about the prediction of discrete events so much as probabilistic tendencies of co-occurence. Rather than prediction in the normal scientific sense of the term, we should perhaps be referring to estimation of likelihood. Similarly, the kind or level of control that we seek in the verification of our predictions does not have the precision or laboratory-like insulation from extraneous or unknown variables. The kind of control we seek is rather more one of relative distance, defined historically, geographically, culturally or socially. When we speak of an island colony or of a peasant village as a kind of natural laboratory, we have in mind a pristine state of affairs which exists in an a-historical vacuum, perfectly sealed off from outside contact.

Universals must address some abstract principle which can be held to sufficiently account for a set or specified range of cultural phenomena, regardless of the circumstances or extraneous influences which may result in a great degree of variability in the expression of that phenomena. Furthermore, universals must be able to account for general directions, and unfolding patterns of change. Such principles must be able to explain not only how a given kind of change has occurred, but, from a social and historical perspective, why that change occurred.

Furthermore, we must distinguish for analytical purposes different levels at which our theory must account for different ranges of experience. One level is that of the existential life-world of the individual, the ego-centered and ego-derived plethora of phenomenological experience including social relationships, patterns of interaction, characteristics of individual personality, as well as the interior psychological realm of the individual. Another level is that of the social institution seen as some kind of corporate organismic whole, somehow above and beyond the life experiences of any one individual. Finally, we must recognize a cross-cultural, trans-historical level beyond the bounds of any corporate organization of people, and encompassing effectively humankind in its entirety.

We must see that universals operating at one level of experience and abstraction, may not articulate upon another level. And yet, if we see human reality as more or less integrated, then we must approach the problem of the articulation of universal theory upon different levels of experience.


Blanket Copyright, Hugh M. Lewis, 2005. Use of this text governed by fair use policy--permission to make copies of this text is granted for purposes of research and non-profit instruction only.

Last Updated: 03/07/05