Reviving the Past:

Revising My Approach to Rock Art

Grant S. McCall

University of Iowa

With comments by John Clegg

I intend this to be the beginning of a productive discussion. Please read the following (if you are interested and foolhardy enough) and I encourage a brief response after reading. I expect constructive criticism and advice that can refine our understanding of how rock art works. After a few weeks time, I will reprint the responses and provide a brief response of my own. You may respond without me reprinting your response. Simply indicate so in your response. To respond, you may either access my guestbook, or you may email me at grant-mccall@uiowa.edu.

I have been considering writing this piece for about four years. In fact, I had a version written once until I decided the revised version still did not say exactly what I wanted to say. I have been working on this version for nearly a year now. I guess the place to start is to say that this piece replaces online a piece that I wrote in the spring of 1997 in which I presented the case for a shamanistic interpretation of rock art. I explicitly intended the piece to be popular, which by all means I intend this one to be. And I largely succeed in my goal. Over the last four year, the piece that this replaces has received more than 1.2 million hits. This is certainly a somewhat anomalous number, and I know that it was not read 1.2 million times. Yet, I should say that when I wrote that piece, I never imagined that I would get such a response. It became circulated far beyond what I intended. As this process continued, I began to accrue this odd sense of guilt for contributing (in gross fashion) to the popularity of an idea that had suspect scientific merit in the capacity in which I described it. In other words, what I said was generally undocumented. The ideas that I expressed were beginning to take hold to an extent beyond what I was really comfortable with. So, not long after I finished the last piece, I began working on this one, secretly composing it an idea at a time in my mind until I got the energy put fingers to keyboard. Sometimes one needs to say something that he can not defend before he realizes the sorts of things that he should have been paying attention to in the first place.

In the following piece, I am attempting to theorize about the function of rock art within a diverse set of societies. Shamanism, to my mind, still helps explain a fair amount of rock art, especially in Southern Africa, if one accepts the theoretical viewpoint on which it is based. However, it is not enough to say this without addressing it further or showing why this is important. I think to really make sense of the rock art phenomenon, one need go a level higher on the theoretical ladder. I am going to make a functional argument for rock art in this piece in which I use ethnographic models to elucidate archaeological data. I also include a discussion of the dangers of the use of ethnography, and I show how I feel ethnography can be used safely. I describe how understanding the social, cultural, economic, and environmental context of a society is the key to interpreting its rock art. Accordingly, I also include a discussion of ethnographic analogy. Shamanism still has its place in here somewhere, but it is not nearly as ubiquitous as I might once have argued, and it gives specific information about what was going on in the society in which it occurred. I will try to show that by building complete ethnographic models and collecting extensive, solid archaeological data, hypotheses can be made about past behavior and these hypotheses are testable using archaeological data. This is how a scientific approach to the past can bring long dead people with no voice for speaking to our time to life, and this is my ultimate goal as an archaeologist.

Before I can proceed to this point, I must make a brief discussion of archaeological theory. In particular, how do archaeologists understand the function and meaning of artifacts in the archaeological record. Hence, I will proceed to a short and sweet discussion of these philosophical issues.

Middle Range Theory and Ethnographic Analogy for Dummies

The instant I mentioned archaeological theory, I imagine 1,199,997 of the 1.2 million readers of this piece stopping and going back to espn.com. And the three that I imagine continuing only did so in order to have material with which to assassinate me at the next vocational meeting. So, I shall proceed as summarily and attempt to speak as plainly as possible, which is extremely unique to discussions of theory (if this is even really a discussion of theory).

How does one determine what an objects is? Let alone how does one figure out what something means? And now I turn to New Archaeology architect, Lewis Binford (1981), in order to make sense of the issue. I need to stop here to express my fear. I fear Lew Binford, and I fear that some of the people that read this know him well enough to report me to him and the next time I look up, he will be quoting me in American Antiquity, attacking my use of his scholarship. So I am going to attempt to be as true to his ideas as possible, but this is a tall order.

Lew Binford in his office at SMU

Anyhow, according to Binford, the archaeological record represents a static and innately meaningless arrangement of material. Indeed, an archaeological site does not stand up and explain the human behavior that went on at it. In order to interpret this arrangement of objects, one must first have some set of knowledge that allows interpretation and attribution of meaning, which Binford refers to as “middle range” knowledge. This is required to link the static archaeological record to past dynamic systems of human behavior and taphonomy. This is not theory, but methodology. Binford uses the example of animal tracks to illustrate his point. Bear tracks are static remains left by dynamic bears. Middle range methodology is a way of seeing the bear through the tracks. In order to understand what "kinds" of bears (young bears, old bears, hunting bears, fishing bears, injured bears, etc.), we need to examine where the bear meets the track; we need to see the where the dynamic systems that produce the archaeological record meet the static archaeological record. This can only take place in there present.

For example, suppose you are the archaeologist of the distant post-apocalyptic future and you come upon a site with a fork. How do you know what a fork is? For us, interpreting the fork is easy. At every meal we see our peers use a metallic four-pronged object to pick up food and insert it into their mouths. In fact, we do it ourselves constantly. We have gained a body of knowledge through years of exposure to know what a fork is, how it works, and I suppose what it means (which is probably not that much in this case). Unfortunately, our stalwart post-apocalyptic archaeologist has not had this luxury. She has never seen a fork before, nor does her own culture have such an object. So, how does she find the meaning of this object?

There are several ways in which to do this, which all involve what Binford calls “actualistic” study. The first way is to take the fork and experiment with it. Come up with a list of possible uses and test each of them experimentally. Furthermore, our archaeologist can look at the wear patterns on the experimentally tested forks and compare these with archaeologically observed patterns on forks. This works well with forks but not with rock art. The next is to examine other living societies and see how the patterns of forks that are incorporated into the archaeological record relate to behavior that went on to produce the pattern. After observing a modern society, one might be able to infer eating through forks in the archaeological record. This is a simplified intepretation, but it gives the idea. If no modern societies use forks, (which is really the more common case in this sort of problem, and what I will be dealing with), the archaeologist can at least theorize about how a fork might fit into such a society, and what sorts of things take its place. Obviously, this is not a terribly reliable way of constructing interpretations of the fork. One might well come to many incorrect conclusions (perhaps the fork is a hair comb, or a projectile device), however we as archaeologists hope we can eventually determine the correct use of the fork.

The problem is that much of archaeological inference relies on convention rather than this sort of middle range methodology. Archaeologists are interested in testing higher level theoretical statements, that perhaps involve measuring the distribution of forks in the archaeological record. For example, an hypothesis might be that the development of social inequality relied on increased numbers of forks. This is obviously a ridiculous example, but I hope it illustrates my point. The problem in testing this assumption is that archaeologists often assume that what they categorize as a fork by convention is actually a fork. Often this convention is wrong. Without middle range research, there is no way of knowing that a fork is a fork and not a comb or a projectile point. However many skip this step, which was Binford's original point.

As I alluded to earlier, this is a fairly straightforward process with respect to artifacts like forks. However, rock art is not that simple. In looking at rock art, we must deal not only with what the type of artifact did, but also we are interested in what the imagery present meant. In sensu Binford, we already have done the middle range research to link rock art to behavior in the past. This was NOT a difficult process. Rock art is the static derivative of people painting or engraving rock surfaces in the past. This is confirmed by ethnographic evidence for people who were not convinced by self-evidence. So, the long and the short of it is that we are not after this kind of middle range theory. I only discussed it in the first place to make this point clear and to illustrate the order and nature of the problem we are addressing; to define the question we are asking. We are after a way of linking rock art to how it functioned within a past society.

The problem here is the direction of the question. I admit outrightly that I am going about this backwards. I am not setting out to test theoretical assertions using archaeological variables. To Binford, the question may be something more like, "What is the relationship between varying manifestations of rock art (which is already operationally defined) and some other sociocultural variable?" This is a valid thing to do, and perhaps the better thing in the long run. However, that is not what I am out to do. I am trying to understand what rock art “meant“. For a pure scientist, this isn't a good question, but I don't care. This is an example of applied science, and as Odell (2001) comments, the distinction is slim in archaeology. For me, the question is, "How did rock art function in past societies and what did it mean?" As Binford (2001) comments, we can only assess the merit of our techniques on the basis of the goal we are trying to accomplish.

Perhaps this is still an answerable question using the methodology of Binford. For example, we may look for a relationship between any number of variables and rock art. One may ask, "What is the relationship between rock art and social inequality?" Here one must measure the occurence of rock art in some regard, figure out a way to measure social inequality (by operationalizing it using middle range methodology) and compare the two. This provides insight for why I am not going use this approach. This kind of data just doesn't exist. I am not going to make some altruistic for more rigor and detail in collecting data. It is practically limited. Operationilizing and measuring rock art is an extremely difficult task. Furthermore, this sort of data are unavailable to me, since I will be simply summarizing and reassessing others' research here. However, this is a direction that we may wish to take in the future, and I think it may be preferable to what I end up doing.

We should be clear in stating that what we are dealing with here is analogy plain and simple. We are looking for how one type of artifact functioned in other societies and applying this by analogy to the artifact we wish to understand in the archaeological record. Naturally, Binford (1967) has also commented on the use of analogy as an archaeological tool. He suggests that we should not use analogies from ethnography to “interpret” features of the archaeological record ipso facto. Rather, he offers that we should use ethnographic analogy to generate hypotheses that “explain” a feature of the archaeological record and that are testable using the same archaeological data whence the question arose. Thus, the pressure is taken off coming up with indisputable logic for interpreting the archaeological record using ethnography. Hypotheses can come from anywhere-- ethnography, biology, history, classics, dreams, religious visions, bad mushroom trips, etc.-- as long as they can be substantiated by the archaeological data. I think Binford had good reason for taking this view. However, I think it eliminates a productive avenue for the evaluation of an analogy. His model assumes that archaeological data is conclusive and that it monolithic in its interpretation across observers. 20th century science is predicated on the fact that nothing can be proven completely and no two people see the same pattern in a certain arrangement. For my money, analogies that have logical links are better than ones that don’t. The more independent sources of verification for an analogy, the better off we are, I say. The important point here is that we must generate analogies that are testable using archaeological data. Binford did hit that one on the head, and this is the approach that I will take.

In our case, we are making analogies between rock art in the archaeological record and a range of ethnographic cases. This presents numerous problems particular to our goals. For one thing, a scant few societies make rock art, and the ones that do cannot be taken to represent the whole of rock art as a range of phenomena in the archaeological record. In addition, unlike in the case of the fork I belabored for so long, with rock art meaning is not as closely related to the function of the artifact. Rock art as an ethnological phenomenon may “mean” something in and of itself. However, presumably all examples of rock art had a meaning unique to themselves. When one looks at a petroglyph, how does one determine what the image meant?

Analogy is the only real way of interpreting rock art or how it functioned. You can’t experiment with it. You can’t look at wear patterns on the artifact. Ethnographic analogy is more pervasive than most people realize. For instance, when you look at a petroglyph and see a hunter with a bow and arrow, and you think to yourself, “That is a guy with a bow and arrow,” you have just made an ethnographic analogy (shame on you).

A painting of a man with a bow and arrow from Sevilla in the Southwestern Cape of South Africa

You have looked at the inherently meaningless and static archaeological record in front of you and imposed meaning on the basis that your culture has a conception of bows and arrows. You have seen bows and arrows within your cultural experience, and the petroglyph in front of you fits the ethnographic understanding of these objects that you have. You have made an analogy between what you know and what the image in front of you looks like. The archaeologist MUST examine societies that have a link to those that made the rock art of interest and attempt to place the imagery of the rock art into the cross-cultural framework. In fact, we understand all other aspects of other cultures in analogy to our own, or to another that understand. This is bad news for the archaeologist. This is a very flimsy way of making inferences, and any number of archaeological thinkers are waiting to attack this kind of logic. After all, how does one decide what societies are analogous? This is a terribly prickly philosophical question. In addition, for rock art of real age, perhaps as old as 40,000 year b.p., there are no extant societies which have a real historical link. On closer examination of all ethnographic analogies, most seem seriously flawed. I am going to make a preemptive strike and acknowledge that this is a bad way of doing things. In addition to the problems I have already addressed, Martin Wobst (1978) points out that this argument by analogy is a logical circularity. By imposing interpretations that we have derived from ethnography on archaeological problems, we lose the ability to learn anything new from the archaeological record. We miss the differences and diversity of how people lived in the past, which is the point of archaeology to begin with. We just reiterate things we have already learned. He refers to this as the tyranny of the “ethnographic record”. This is a problem that is hard to get around, however I think the emphasis placed on hypothesis testing is a good start. Still, I acknowledge this as a problem. that I am working in an unscientific direction with regard to the problem I am addressing, and that the methodology that I do use in the end is flawed. But I would argue that we should not discard an approach simply because it is flawed. Something that is flawed is not totally useless. Instead, since we must use ethnographic analogy as an approach, we should at least understand its limitations and it value. In this spirit, I shall discuss quickly how analogies are made and what the most productive way forward may be.

Ethnographic analogies in the study of rock art can be lumped into a few distinct categories.

Edward B. Tylor

(1) Psychic Unity: Stemming from the unilineal evolution of E.B. Tylor, this notion is that all societies can be placed on a linear evolutionary scheme ranging from primitive to modern. Therefore, societies at the same stage of development have the same cultural traits. In terms of rock art, societies that make rock art now, have similar technology, and are generally of the same developmental stage as older societies that made rock art are analogous. Modern rock art is the same thing as 40,000 year old rock art, because the artists were at the same evolutionary stage and therefore had the same culture. I am not doing this idea justice at all (nor do I intend to), but it is clear that this sort of analogy has its share of deep flaws. However, it is surprising to me how pervasive this idea is within rock art today. I guess I am going to lump what I am going to call mentalist analogies in the same category, at the risk of angering many of my friends. This includes structuralist arguments. These go something along the lines of common structures exist in each individual human mind that manifest themselves in terms of common cultural structures. Claude Levi-Strauss was one of the original perpetrators of this idea. He showed how mythology had cross-cultural similarities supposedly on the basis of similar structures in the mind. I think that this is the current theoretical underpinning of the shamanism argument, which I am somewhat less than comfortable with. My main complaint against structuralism is that it is very difficult to address empirically. After all, how does one identify common structures in the brain, let alone measure them, let alone prove them as the causal factor in rock art similarity across space and time. This obviously holds some appeal for good reason, but I think the future of this kind of research are limited to simple conjecture, and “just so” stories.

Franz Boas taking part in a Kwakiutl ceremony for the benefit of the camera-- one of the first hunter-gatherer ethnographers

(2) Diffusionism: This perspective is most linked to the "historical particularism" of Franz Boas, the father of American anthropology. It is all the more helpful to prove that a society that made rock art in the past is connected to a modern society. In its current form, this is usually referred to as the “direct historical approach.” A good example of this is the San rock art of Southern Africa. The rock artists of Southern Africa have direct links to modern San groups. Therefore, modern San ethnography can shed light on rock art. This idea, too, has its share of flaws. For instance, how does one go about finding a modern society with a link to Upper Paleolithic Europe. Lest we begin laughing at this idea, there are people who do this. They draw maps with a lot of arrows on them, and show how the atlatl, bow and arrow, and other such items spread throughout the world. And I certainly do not condemn this activity in principle. The societies that exist today did have to come from somewhere, so there must be links to the past. However, these links are nearly impossible to find, and would not be of great use in the case that they were found. Cultures change a great deal in 40,000 years, and it is impossible to make an analogy on this sort of basis over such as long period of time. Even making an analogy from modern San groups to ancient Southern African rock artists is problematic enough. All too often, the considerable ethnographic diversity of modern San is ignored. Cultures change a tremendous amount even within a few thousand, or a few hundred years.

A.R. Radcliffe-Brown as a professor at the University of Cape Town

(3) Functional unity: Beginning with A.R. Radcliffe-Brown, who stated that societies tend to have similar cultural traits because they have similar circumstances. In modern times, this logic is employed by cultural ecology, saying that cultures are similar because of similar ecological situations (Steward 1955). In other words, some societies must meet the same sort of demands in order to survive and tend to deal with these demands in the same sorts of ways. This is really just another hack at the psychic unity discussed above, and has many of the same flaws. However, this is the method that I believe offers the most hope for analogy, particularly distant analogies into the Upper Paleolithic. Using this sort of method, there is actually a logic to making an analogy between San rock art in Africa to Chumash rock art in California. Both peoples were under the same kinds of economic pressures, so they dealt with them in the same sorts of ways. Rock art is one of the ways of dealing with these economic pressures, and therefore might represent some commonality in function across cultures with similar economic circumstances. This is a key idea for my further discussion.

So, how does an interpretive framework coalesce out of this maelström of flawed analogical reasoning? I guess the answer is carefully. I think both diffusion and functional unity arguments are important. On the basis of logical merit, I do not give either an advantage over the other. However, I think there extremely few good sources of diffusion links that can be helpful on the sort of timescale necessary in discussing rock art. This leads into the next part of my argument. Maybe we as archaeologists can have a better hack at the meaning of rock art if we understand how it functioned within ancient societies. MAYBE rock art functioned in SIMILAR ways in societies without direct historical links, and MAYBE we can learn something about how rock art might have functioned in ancient societies on the basis of modern ethnography, MAYBE even ones that do not produce rock art currently. If not, then we are all screwed, because I am out of other ideas, really.

The next section of this paper is a discussion of modern hunter-gatherer ethnography aimed at providing both functional and historical contexts for the production of rock art. From this base, I hope to work toward a few models of rock art meaning that may be potentially testable using actual, honest-to-God archaeological data.

Hunter-Gatherer Ethnoarchaeology

I mentioned that there are a scant few cultures that produce rock art today. Though I may be making a somewhat politically incorrect assessment, I frankly question the connection of the modern rock art producing cultures to any prehistoric traditions. This makes any direct sort of analogy totally problematic. Another problem that I perceive (and certainly my critics will jump on the title of this section) is that not all rock art was produced by hunter-gatherers. However, the majority of rock art was, and the logic I am going to argue for still explains rock art that wasn’t. So, for the time being, I am going to ignore this notion.

Hunter-gatherer ethnoarchaeology has an interesting history. The systematic study of modern hunter-gatherer peoples began because of work with primates (and this begins to give an idea of the level of political correctness on which this argument is made). Anthropologists were studying primates in Africa to try to make analogies to early hominids. On a mechanical level, these same anthropologists noticed that there were still surprisingly people that lived as hunter-gatherers, which was pretty neat. Second, on a more theoretical level, if early hominids are somewhere between the great apes and modern humans, shouldn’t modern humans be included in this process of making analogies? Hunter-gatherers were the obvious choice because they were economically most similar to what anthropologists felt early hominids in the Pleistocene were like. Furthermore, a vast number of archaeologically studied people were also hunter-gatherers, and knowing something about the anthropology of hunter-gatherers seemed useful in that regard. Suddenly, studying hunter-gatherers seemed like a really good idea.

The first attempts at this were to construct the “modal” hunter-gatherer society. Julian Steward (1955) began this process. In short, anthropologists sought to construct a model of the average, generic, Brand X hunter-gatherer society. By making a model of the modal hunter-gatherer society, anthropologists could use this model to interpret the past. This seemed to be a step in the right direction. The Man the Hunter conference (which also shows the level of political correctness at work) was the high water mark of this kind of reasoning. Man the Hunter was convened in Chicago in 1966 by Richard Borshay Lee and Irven DeVore and the proceedings published in 1968. This was a conference of cultural anthropologists, archaeologists, evolutionary anthropologists, and even primate ecologists designed to build a model of the modal hunter-gatherer society and figure out how to use it to sort out human evolution and hunter-gatherer archaeology.

Julian Steward as a professor at Utah

Before I continue on to the problems with this kind of logic, I will stop and put it into its theoretical context. Why should hunter-gatherers be similar in the first place? The espoused answer was the “functional unity” idea that I mentioned in the last section. Hunter-gatherers were similar because they had similar economies and interacted with their environments in similar ways. This was cultural ecology as first enunciated by Julian Steward, which became the dominant framework for anthropology in the 1960’s and 1970’s. Steward’s new definition of culture was interaction with the environment. Culture was a way of surviving in a given environment. Hunter-gatherers were similar because they made a living from their environment in similar ways. Man the Hunter shared this framework and expanded upon it.

I will not get into the gory detail of the model that emerged, but I will provide a summary. Man the Hunter established the model of the mobile, band society where men hunt and women gather. Resources are shared equitably, no individual is allowed to accumulate food (i.e. store food) or property, and political organization remained at the egalitarian level. Kinship descent is patrilineal and patrilocal. This model was largely derived from the !Kung San (now referred to as the Ju/’hoansi) of Kalahari. For what it is worth, this is where I do my ethnographic fieldwork today.

From the Man the Hunter conference, there emerged a few main foci of ethnographic attention: the Kalahari, central Australia, and Northern Alaska and Canada. To a lesser degree and for different reasons, the central African rainforest, the Amazon, and the Northwest Coast of North America also became emphasized. The core societies that I mentioned in the first sentence were picked because of their perceived similarity (which is handy in constructing a model of the modal hunter-gatherer). However, the more anthropologists studied these societies and the more they considered other hunter-gatherers such as those that I have ascribed lesser emphasis, the more one notion occurred. In addition to the remarkable similarity in some regards, there was extraordinary diversity among these groups. Essentially every characteristic of the modal hunter-gatherer society that I mentioned above was contradicted in hunter-gatherer ethnography in multiple situations.

For example, the Northwest Coast Indians were essentially sedentary, living in large permanent villages. Political organization was far from egalitarian, with extreme emphasis on prestige and significant inherited social inequalities. Other inconsistencies emerged in almost all of the other case studies, as well. Pretty soon, the only society that fit the model was the society on which the model was based: the Ju/’hoansi. After the initial reaction of trying to label examples that did not fit the model anomalies, anthropologists broadened their models to include two types of hunter-gatherer. Here, the nomenclature gets tricky. These two types have been labeled “simple” and “complex”, “immediate-return” and “delayed-return”, “egalitarian” and “non-egalitarian”, and a host of other titles. I choose “egalitarian” and “non-egalitarian” as my terms for the sake of descriptive accuracy, simplicity, and lack of implied value judgment. Egalitarian hunter-gatherers inherited the model discussed above. Non-egalitarian hunter-gatherers were more sedentary, they stored food, they had political and economic inequalities, which were often hereditary, with defined senses of personal property and less of an ethic of sharing. The Northwest Coast Indians became the model for non-egalitarian hunter-gatherers.

This division of hunter-gatherers improved the models a great deal, but it still left a lot to be desired. Hunter-gatherer anthropology is still struggling with this problem today. In 1995, Robert Kelly published a book that argued for the recognition of diversity among hunter-gatherer societies. The approach outlined in the book was not to view hunter-gatherers as either/or, +/- oppositions, but to view them as a continuum. So, for example, a hunter-gatherer society is not just mobile or sedentary, but varying degrees of both. The same could be true on a more abstract level for egalitarianism. The idea is not to cram the diversity of the hunter-gatherer ethnographic record, which is considerable, into two strained models, and not to ignore cases because they were not similar to existing models. This is concerned not with understanding how hunter-gatherers are similar, but the causes of similarity as well as difference. I see this as a productive way forward.

Furthermore, ecological theory has the potential to explain diversity. Ecological theory must consider more than why cultures are similar. It must also consider why they are diverse. If there are reasons why cultures are similar, there must be reasons why they are diverse, and these reasons are knowable. In terms of build models for interpreting the past, we must reckon the ecological and economic reasons why cultures are similar and, just as importantly, why they are different. Then, when we approach the archaeological record, we may recognize the circumstances that cause a certain cultural adaptation among modern peoples and be able to infer that people in the past may have had similar adaptations. This is how to interpret rock art- not to say that X society had rock art and use what it meant to them to interpret other rock art that looks similar. Rather, by understanding how modern hunter-gatherers adapt to their environments, we may begin to understand how rock art might be such an adaptation.

Hunter-gatherer ethnography helps us build models with which to interpret the past. This is one example of a way in which ethnoarchaeology can aid in addressing complex issues in the past. In the next section, I will examine some rock art sites that have been well documented archaeologically and use a few models derived from hunter-gatherer ethnoarchaeology to interpret them.

Understanding Rock Art: Archaeological Data and Ethnographic Models

I am going to keep my discussion of the archaeology of rock art mainly to the area that I know best: Southern Africa. The first case study I will examine is that of the Hungorob Ravine in Namibia. The Hungorob Ravine is one ravine over from the much more famous Tsisab Ravine in the Brandberg of Namibia. The Tsisab Ravine is famous for historical rather than sensible reasons. It is home to the notorious “White Lady of the Brandberg” painting, which was discovered in the first part of the last (the 20th) century by Reinhardt Maack. It became famous as scholars such as Abbe Breuil decided that the painting showed contact with early European explorers from before the Roman empire. Of course, the White Lady was neither white, nor a lady. However, the endemic racism of the day inspired great fame.

For all real intents and purposes, the Hungorob Ravine has better rock art than its neighbor. There are four main sites scattered within the ravine with numerous satellite sites surrounding each main site. The area has been extremely well-documented archaeologically by former Namibian state archaeologist, John Kinahan (1991). The sites with paintings began to be occupied around 4,500 b.p. The residents were hunter-gatherers, with no signs of domestication. Around 2,000 b.p., pottery was introduced into the area, however the residents seemed to have remained hunter-gatherer. Around 500 b.p., stock was introduced and the residents became herders. The paintings likely date to the 4,500 b.p. date (Kinahan 1991). Rock art was done into the pottery phase, which I have established by pottery correlation from Kinahan’s (1991) data. However, it takes a peculiar form. It becomes engravings of four rows of dots, and never any other subject matter. This is the limit of detail that I will give for this area in this piece, given its scope. I will gladly give more detail upon request.

A painting of herd of gemsbok from the Brandberg

Kinahan (1991), using faunal data (animal remains from things the residents ate), has established that the sites were occupied during the dry winter months. During the wetter times of year, the residents occupied the gravel plains outside of the ravine where larger game was much more plentiful, but which lacked permanent water sources. The ravine had the drawback that food was relatively scarce. The residents relied heavily on small mammals and unpleasant vegetable foods. Furthermore, it seems that there was a relatively high number of people located in the small area (the ravine is extremely narrow) during the time period in which the paintings were done. This is where the archaeological data ends and the ethnoarchaeological inference begins.

I suggest that the occupants of the ravine that made the rock art were egalitarian hunter-gatherers, and Kinahan’s (1991) data support this. Most convincingly, there is a direct and continuous archaeological link between these people and modern egalitarian hunter-gatherers of Southern Africa in the Kalahari. The direct historical approach supports my claim. In addition, the mobility patterns are residential rather than logistical, meaning that people moved to the resources rather than having parties bring the resources to the consumers (Binford 1981). This is typical of egalitarian groups. There are no signs of data that might suggest anything but egalitarianism, such as inequalities in grave goods, food storage, sedentism, or the like. These seem to be people typical of the egalitarian model of hunter-gatherers, as well as having strong links to modern egalitarian hunter-gatherers.

One of the characteristics of egalitarian hunter-gatherer societies is reciprocal access to resources, both among individuals and groups. This means that if the food runs out in your territory, your water resources dry up, etc., you have the right as an individual to go live elsewhere where resources are better with friends or relatives with whom you have a specific type of social tie (Wiessner 1982). The reverse would also be true. Among the Ju/’hoansi of the Kalahari, this is called a “xaro” relationship, and it is essentially a way of sharing risk (Wiessner 1982). Furthermore, in some cases, groups will come together and share resources, mainly having to do with water, during times of extreme scarcity. The long and the short of it is that your neighbors will not let you starve or die of thirst if they can help it as long as they have the assurance that you will do the same for them. This is a very important characteristic of egalitarian hunter-gatherers, and it cuts down risk of shortage significantly without employing storage techniques that have both economic and social drawbacks.

This model is important because we know that this is the social and economic context in which the rock art is being done in the Hungorob Ravine. People were crammed together because water was scarce, and food seems to have been in short supply. I suggest that the rock art functioned as a mechanism for facilitating smooth access to the local resources for cooperating bands and individuals. Rock art gave people a common social, cultural, and ideological framework that helped hold individuals and groups together during strained times. It ensured the continuing function of the reciprocity practices that are absolutely vital for the egalitarian hunting and gathering mode of production. The act of painting may have acted to link individuals and groups to places, as well as giving people common frameworks, which were the basis for social ties. In short, the people in the Hungorob Ravine could not survive without the egalitarian reciprocity mechanisms, and I suggest that rock art was a tool that helped maintain them.

Is there archaeological evidence for this? After all, I have put forward a hypothesis that is useless unless it is testable. First, in terms of the largest grain of data, a strong majority of the figures in the rock paintings are human. This puts an emphasis on social interaction from the start. Furthermore, the four core sites are heavily dominated by scenes of social interaction. This also emphasizes these sites as locations of heavy social interaction and this also coincides with the location of the scarce resources that people would have been coming together to exploit. Social interaction is shown disproportionately at sites where social interaction would have been the greatest in this model. The data all fit together nicely. However, this is an example where models can affect the kind of data that need to be collected. While Kinahan (1991) has recorded the rock art in some detail, it would be easily possible to address this problem quantitatively rather than descriptively. One would like to ask questions like, what is the statistical breakdown of the depiction of social interaction across these sites? I feel I am safe making my conclusion from Kinahan’s qualitative descriptions, but more data would help this cause.

Kinahan (1991) describes the social scenes as depictions of trance dances in several cases, and interprets much of the other imagery as trance metaphors taken from David Lewis-Williams’ (1981) model. I agree with this interpretation in the context of the model that I have attempted to build. The problems stem from operationalizing trance scenes. How are they defined? Ethnographic descriptions of the trance dance involve traditional healers/religious leaders (shamans) dance for long periods of time to very intense, loud, and rhythmic music. Exhaustion contributes to the person’s collapse, along with other neuropsychological factors. In the trance, the person goes through several stages involving visions of geometric abstractions similar to what is seen in a migraine, conflated imagery, and often the feeling of becoming a certain species of animal. The shaman uses the religious potency accrued in this ritual to attempt to rid people of various medical problems, affect the weather, or solve other supernatural problems. Lee (1979) gives an ethnographic account stating that trance dances occur significantly more often when groups are combined around scarce resources, and social tensions run particularly high. The trance dance is an extremely effective way of diminishing the growing social tensions. This is obviously very similar to the situation archaeologically visible at the Hungorob Ravine.

I have very little patience for those that challenge the ethnographic evidence from the Kalahari concerning the trance dance. It is ethnographic fact. Numerous reliable ethnographers have witnessed this ritual, including myself. To say that there is no ethnographic evidence for the trance dance is ridiculous. I certainly recognize the right to challenge to applicability of this model because of the logical problems raised by making ethnographic analogy, and I definitely see numerous problems in isolating what imagery is convincingly related to the trance dance and how to operationalize it. However, I think the trance dance is recognizable within rock art. At the four core sites, there are depictions that look exactly like trance dances observed in the Kalahari, and a good deal of imagery which can be plausibly explained as trance imagery, even at the periphery sites. Furthermore, it fits into the model I have laid out very well. The depictions of the trance mirror the increased occurrences of the trance dance caused by the greater social tensions that accompany group fusion. This is not wild, unfounded, and unprovable speculation that certain images are related to trance dances. This is a model taken from solid ethnoarchaeological research that started in archaeological data and was tested using archaeological data. I have tried to put the trance dance in the social, cultural, and economic contexts of a long lost people that we recognize through archaeological data.

In summary, rock art acted to lessen the social tensions growing due to the scarcity of resources. It is clear that the rock paintings were not an idle record of events that served no real purpose to the society other than as an aid to memory. Instead, it is clear that the rock art was vital to the economy of the hunter-gatherers that occupied the ravine. If social tensions got the better of the egalitarian social structure and norms, it spelled disaster. Without the social relief that the trance dance and rock art brought, the people of the ravine would have few means available to them. This shows how archaeological as well as ethnographic data can contribute to the construction and revision of ethnological models. In addition, it shows how interpreting archaeological data with the help of ethnographic models can revive the stories of long dead peoples and make their struggles real. This is the most important and most often neglected aspect of archaeology.

Polly Wiessner doing ethnoarchaeological research with the Ju/hoansi in Namibia

This sort of approach has the potential to interpret rock art on many levels. Scholars have long marveled at the stylistic continuity of Southern African rock art. In some parts of the subcontinent, there appears to be as many as 8,000 years of stylistic continuity, and the rock art tradition has been going on for more than 20,000 years, and likely much more. Approaching the makers of the art as egalitarian hunter-gatherers offers a good explanation of this. In her analysis of projectile points, Wiessner (1983) argues that egalitarian societies exhibit extremely slow rates of change. This is due to the nature of egalitarian societies. There is strong social pressure against material innovation. It is not good to make material that is different than what other have made and have the ability to make. In addition, it would be inappropriate for people try to improve their social status by coming up with new styles, and people are actively discourage from attempting to do so. The result is that styles change extremely slowly. This offers a good explanation for the stylistic change in the rock art traditions for most of Southern Africa.

Handprints from the hallowed Eland's Baai site in the Southwestern Cape of South Africa

The exceptions to this are also quite interesting. For example, in the Southwestern Cape region of South Africa, the incorporation of a herding economy is recognizable around 1,000 b.p. Corresponding with this is an abrupt change in the rock art style away from the fine line style that had dominated the region for several thousand years continuously during occupation by hunter-gatherers. The style changes to more gross finger painting technique. Individualistic motifs begin to be more predominant. In particular, the handprint takes an important role. Tony Manhire (1997) offers a good explanation of the handprints. He has shown that the handprints were done mostly by young adults between the ages of 10 and 13. This is an utterly remarkable finding. This suggests strongly that the handprints were a part of some sort of initiation rituals. This also points to territoriality and links of individuals and groups to certain areas by marking territory

This is a very interesting revelation with respect to ethnographic models of non-egalitarian herding societies. Non-egalitarian herding societies stress individual prestige and territoriality. It is extremely important that groups maintain through aggressive intimidation access to large stretches of vital grazing land. The initiation ritual in which the handprints seem to have been a part helped to develop a sense of belonging and duty to the group, a sense of individual prestige, and it made a tangible link of individuals to a certain territory. Hence, the appearance of this motif is perfectly consistent with the switch to herding. We build the picture of a people that go through an initiation at puberty (in this case, likely males) that builds a sense of individual prestige, duty to the group, and link to a specific geographical local. After this, the young males are primed to build their prestige by defending their territory from encroaching neighbors through whatever means necessary, even at the risk of their own death. The archaeological data the show a switch to herding help interpret the rock art and the rock art gives clues about the dynamics of the switch to herding. Ethnographic models help explain the situation further. In this case, the ethnoarchaeological interpretation contradicts the shamanistic interpretation, which sees the handprints as being a part of the trance dance where the shaman performs healing by laying his or her hands on the patient covered with potent sweat or nasal blood. This interpretation is consistent with neither the archaeological data nor the ethnographic models. Once again, this case contributes both to growing ethnological models of herding societies and our understanding of the color of the lives of forgotten people.

I have offered a few cases to try to show the power of ethnoarchaeological research in building behavioral model and reviving the richness of the lives of a nameless people from a lost time. Ethnography has the ability to contribute immensely to our understanding archaeological data by providing analogies that help provide interpretations. Without ethnography, we have little way of understanding the past. However, without building the thick context for the models we take from ethnography that consists of holistic documentation or social, cultural, economic, and environmental background, our models have no foundation and quickly lead us away from the scientifically solid truth and into the realm of unfounded speculation.

Conclusions

Many make the argument that approaching archaeology (or anthropology, for that matter) as a science takes the humanity out of our subjects. I see it quite the opposite. Using scientific methods to approach the past gives us the tools to put the humanity back into people with no other voice. Others may argue that what I have done is not really science. This is only partially true. I have used scientific data and techniques to provide a descriptive interpretation of the past. Binford may claim that archaeology should not concern itself with reconstructing culture histories. However, I would claim that this is a among the primary goals of the discipline. People want to know that the past was like, and we have a chance to give them a link to long dead stories that otherwise would not have been told. Maybe this is not as important as building anthropological or archaeological theory, but it is still extremely important. The past is describable and knowable, and the way culture works is the same. Understanding both gives us insight on the human condition. This is no doubt why rock art has such a dominant interest within our society today. It gives us a more real link to our past, and gives us a more tangible way of putting ourselves within the framework of what it means to be human. We see ourselves in the rock art that fascinates us. Science does not cheapen this. Rather, it gives us the only real way of understanding the truth. Archaeology that ignores these questions is incomplete, and the techniques I have described here are a small step in this direction.

As I mentioned at the beginning, I would appreciate any feedback to this article. I intend this to be the beginning of a discussion. Discussion is the most productive way of building our understanding of how rock art works with regard to the processes of culture. Please contribute your comments to my guestbook and, in a few weeks time, I will reprint them at the end of this article and provide a brief reply. If you do not wish your response to be reprinted, please indicate so. I certainly appreciate and look forward to all responses.

Comments

I am pleased to receive this response from John Clegg. I will provisionally proceed and publish his insights while I consider his comments. My first instinct is that he has correctly caught me in a few of my verbal excesses! I think incorporating his perspective as a complementary view from perhaps the opposite side of the same intellectual geneology is fascinating. I look forward to constructing a full reply! Furthermore, I look forward to any additional comments others may have.

September 25th, 2002 GM

References

Binford, L.R. 1967 Smudge pits and hide smoking: the use of analogy in archaeological reasoning. American Antiquity, 32, 1-12.

Binford, L.R. 1981 Middle range research and the role of actualistic studies. In: Bones: Ancient men and modern myths, pp. 21-30. Academic Press, New York.

Binford, L.R 2001 Where do research problems come from? American Antiquity 66 (4): 669-678.

Kelly, R.L. 1995 The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Smithsonian Institute Press, Washington D.C.

Kinahan, J. 1991 Pastoral Nomads of the Central Namib Desert. New Namibia Books, Windhoek.

Lee, R.B. 1979 The !Kung San: Men, Women, and Work in a Foraging Society. Cambridge UP.

Lee, R.B. and I. DeVore 1968 Problems in the study of hunters and gatherers. In: R. B. Lee and I. DeVore, eds., Man the Hunter, pp. 12-20. Aldine, Chicago.

Lewis-Williams, J.D. 1981 Believing and Seeing: Symbolic Meanings in Southern San Rock Paintings. Academic Press, London.

Manhire, A. 1997 The Role of Hand Prints in the Rock Art of the South Western Cape. South African Archaeological Bulletin 53: 98-107.

Odell, G. 2001 Research problems r us. American Antiquity 66 (4): 679-685.

Steward, J. H. 1955 Theory of Culture Change: The Methodology of Multilinear Evolution. University of Illinois Press, Urbana.

Wiessner, Polly 1982 Risk, reciprocity and social influences on !Kung San economics. Pp. 61-84 in Eleanor Leacock & Richard Lee (eds.) Politics and History in Band Societies. Cambridge UP.

Wiessner, Polly 1983 Style and Social Information in Kalahari San Projectile Points. American Antiquity 48: 253-276.

Wobst, H.M. 1978 The archaeo-ethnography of hunter-gatherers, or the tyranny of the ethnographic record in archaeology. American Antiquity 43: 303-309.