Philip E. L. Smith
Problems and possibilities of the prehistoric Rock Art of Northern Africa
African Historical Studies, I, 1 (1968): 1-39
Problems and Suggestions
The problems which still await solution in the northern half of Africa in the field of prehistoric art are many. Some have already been discussed here; others can be mentioned very briefly. One of the most relevant is the question of indigenous cattle domestication in North Africa. Domesticated cattle are certainly present in the Bovidian Pastoral style art, probably from the beginning, but it does not seem possible at the moment to distinguish in the art itself a stage corresponding to an incipient domestication. Yet this possibility cannot be rejected. Certainly, the most recent findings in the central Sahara suggest a need to reexamine this problem, which has tended to be passed over in the last few decades in favor of postulated centers of domestication in southwestern Asia. Unfortunately even in the relatively well known Neolithic sequences of this latter region there is considerable uncertainty about the time and place of earliest domestication of cattle. But northern Africa has on several occasions been suggested as one of the centers of domestication 117, and Clark 118 has mentioned the possibility of domestication from a wild form Bos opisthonomus which existed there. Bosch-Gimpera has suggested in his arguments for an early dating for pastoral scenes in the Sahara that domestication had perhaps already begun in the Epipalaeolithic 119. This whole problem has also been discussed by Monod 120. Mori is now certain that domesticated cattle were present in the Acacus by ca. 5500 B.C., basing this claim in part on the discovery of part of a skull of the short-horned Bos brachyceros in the lower level of Uan Muhuggiag site, and, although he does not explicitly require that it involved indigenous domestication, he refers to “preparatory stages for true pastoralism.” 121 Clearly, this is one of the problems which have a direct bearing on the prehistoric art but which will probably have to be resolved by archaeological investigations; those now going on in the Sahara may be expected to shed considerable light on the matter.
A related problem is the matter of plant domestication since this question is raised by some of the scenes in the central Saharan art. Lhote has inferred the existence of agriculture in the Bovidian Pastoral period, because one scene shows a group of women said to be working in a field, and Clark seems tentatively to accept this suggestion 122. But as Monod points out quite properly 123, it is difficult in such ambiguous presentations to distinguish between digging and gathering, and the same criticism can be made of suggestions that grinding stones in archaeological levels, or scenes showing food grinding or preparation, necessarily involve domesticated plants.
Certain other problems have already been touched on in this paper, but the need for further investigation needs to be underlined. Do the differences in art styles represent the workof different groups, as Breuil seems to suggest? 124 To what extent are there transitions or overlaps between one period style and another? Breuil thought there was evidence of such overlap between the Round-Heads and the Bovidian Pastoralists at Tassili 125, and Huard concludes that the early Hunters art survived into the middle Pastoral period in the southwestern part of the Sahara 126. What was the nature of the environment and climate in the Saharan area in the pre-Pastoral phases? As Mori points out 127, very little is known of this matter before the Bovidian Pastoral period; yet this must be studied before we can talk intelligently about conditions for incipient domestication of plants and animals on the same level of competence as we can in southwestern Asia, where this problem is gradually being understood. The questions of the origins, dating, and duration of the Round-Head style still await documentation; it may well be “essentially African and Saharan” as Lhote claims, but much more information is necessary concerning the artifacts and physical anthropology before its implications can be properly grasped. The problem of what was happening in the Sahara between the middle of the third millennium B.C., when the Bovidian Pastoralists were apparently declining or dispersing, and the appearance of the horse-chariot in the area between ca. 1500 and 1200 B.C. also requires further research. In the eastern part of North Africa to what extent was the Nile a significant boundary in separating the art and cultures of the Libyan zone of the Sahara from those of the Eastern Desert? There are certain shared traits — e.g., the practice of “quartering” cattle bodies in the rock drawings — but in general this problem has tended to be neglected by most authorities in North African art who deal with the Sahara and whose interests fall off as one goes east of Tibesti; very little field work has been done between the Nile and the Red Sea since Winkler's time apart from some recent investigations byW. Resch 128.
Indeed, the whole problem of establishing archaeological zones or provinces in the North African Neolithic and later periods still awaits proper treatment. Some attempts have been made — e.g., that by McBurney and Hey 129 — in distinguishing an eastern and a western province. The existence of a “Neolithic of Sudanic tradition” has been suggested by some writers as a kind of counterweight to the Neolithic of Capsian tradition. Certainly the majority of North African prehistorians today, including Balout, Hugot, and Gobert, no longer see events in the Maghreb as consisting solely of a monolithic Neolithic of Capsian tradition which expanded into the Sahara. Things are unlikely to have been so simple at this time when, in the earlier stages at least, most groups seem to have been sub-Neolithic with most of the subsistence emphasis apparently still on hunting, fishing, and collecting. A more precise definition of the culture areas or provinces on this horizon will in turn provide a more reliable background for discussing the prehistoric art found in the areas concerned.
It is of course always easier to offer suggestions for research than to apply them, and those offered in this paper are no exceptions. Very obviously we require further studies of the spatial distribution of art elements, fauna, and artifacts. Lhote's efforts in painstakingly listing the geographical distribution of all animal species shown in the rock art and their frequencies and associations, as well as in attempting to correlate them with local geographicalfeatures and with palaeobotanical data where the latter are available, is an admirable example of this; in many cases it has enabled him to offer environmental reasons for presences or absences of particular species in the art and thus to eliminate explanations based on other possibilities 130. But it seems to me that there should be more exhaustive zoological studies of the animals depicted in the art, particularly of those thought to be domesticated. Although a number of zoologists have in the past given opinions on the fauna shown in North African art, it has generally been on the level of species identification. To my knowledge few or none of the palaeozoologists of various countries who have in the last few decades brought about such advances in our knowledge of the processes of prehistoric domestication of animals have been called on to examine the art of North Africa, especially the Sahara, in this light. Yet one would have thought that this would be extremely useful from their own points of view as well as from that of the archaeologist, for, especially in the highly naturalistic paintings of cattle herds, information on non-skeletal mate-rials, such as sexual dimorphism, body profile, and coloration, are often available and should be invaluable to supplement the osteological remains from archaeological deposits. And one might mention in passing that in more than one instance in past studies of North African art the animals in question have been wrongly identified, usually by non-zoologists, so that antelopes have been interpreted as horses and phacoceros as hippopotamus.
Finally, I return to a topic which was mentioned briefly before and offer some comments which may be of use in evaluating the ever increasing quantities of rock art data now available. Prehistoric archaeologists, especially in the New World, have for some years been occupied with analyzing artifacts in terms of such features as attributes or modes as a way of organizing and studying their materials 131. Related studies have been carried out in Europe (see, for example, the attempt by Gardin to apply rather similar concepts to artifacts and iconography) 132. Rock engravings and paintings are, after all, artifacts; twodimensional ones, it is true, but nevertheless vestiges of cultural behavior, which should be susceptible to analysis like their three-dimensional kin. The methods suggested by the archaeologists mentioned above may be relevant to their analysis.
Rouse suggests that artifacts can be described usefully in terms of modes and attributes 133. Modes in this sense are any standards, concepts, or customs governing the behavior of artisans which are passed from one generation to another, or from one community to another; attributes (e.g., raw materials used, shape, artifact decoration) are the manifestation of such modes in artifacts, the means by which the customs and concepts of manufacturing and using artifacts are expressed. Archaeologists can use the concept of modes in several ways: to refer to the behavior of the artisans in making and using the artifacts (“procedural modes”) or to refer to ideas and standards the artisans have expressed by means of artifacts (“conceptual modes”). From then on there are several ways of treating the data to establish types, if one wishes to go beyond the level of modes: by noting the significant modes on punch cards or by some other means of data handling; by dividing the specimens on the basis of first one set of modes, e.g., raw materials, then by another set, e.g., shapes, until all artifacts of the same kind have been analyzed. The number of modes selected depends, of course, on the number present and expressed by the attributes; in “simple” artifacts they will be few and can all be used, whereas in “complex” artifacts fewer may be chosen.
This approach has not been altogether unknown in studies of North African rock art, of course; thus, Mori uses the combinations of the “four elements determining and characterizing” Saharan rock drawings - - patina, style, technique, and dimensions — in his classifications for chronological and other purposes 134. Other writers have also used units considerably more complex than the gross ones of, say, “engraved hippo,” “painted cow”; for instance, the elaborate subdivisions of varieties of cattle horns proposed by Rhotert 135. Nevertheless, a far more refined system than those hitherto applied is necessary if we are to extract the maximum of information from the data. Here the code devised by Gardin for the analysis of such artifacts as tools, ornaments, and iconography is particularly suggestive for describing the data in prehistoric rock art, whether on the level of technique of manufacture, of form of individual units or of large groups or “scenes.” 136 This is not only a way of categorizing the data for descriptive purposes; with such methods associations which today are not discernible, or barely so, can be examined for their possible significance far more easily than can be done by traditional means of recording. The reader is referred to Gardin's paper for the details of his scheme and for his use of “distinctive features,” which seem analogous in many ways to the “attributes” of other archaeologists. The application of such methods of analysis and description to North African rock art would admittedly be a long and costly one in practice, but it does not seem beyond the limits of techniques available today, certainly no more so than those currently being suggested for the analysis of artifacts collected from excavated sites. In the long run it will be by the adoption of methods such as these — which, after all, are intended to aid and supplement, not to replace the judgment of the investigator — in conjunction with the data from excavated sites that we shall come to appreciate fully the role that the art of this huge region can play in an understanding of the cultural events and cultural processes of prehistoric and protohistoric North Africa.
117. Ed. Dechambre, “Le Sahara, centre primitif de domestication,” Séances de la Société de Biogéographie, c.r. (1951), 147-151.
118. J. Desmond Clark, “Africa South of the Sahara,” Courses toward Urban Life, R. J. Braidwood and G. R. Willey, eds. (Chicago, 1962), 15.
119. P. Bosch-Gimpera, “The Chronology of the Rock Paintings of the Spanish Levant,” Prehistoric Art of the Western Mediterranean and the Sahara, L. Pericot Garcia and E. Ripoll Perelló, eds. (New York, 1964), 129.
120. Monod, “The Late Tertiary,” 200.
121. Mori, “Contributions,” 175.
122. Clark, “Africa South of the Sahara,” 16.
123. Monod, “The Late Tertiary,” 186.
124. Breuil, “Les roches peintes,” 150.
125. Ibid., 150.
126. Huard, “Art rupestre,” 140.
127. Mori, “Contributions,” 175.
128. W. F. E. Resch, “Neue Felsbilderfunde in der agyptischen Ostwuste,” Zeitschrift fur Ethnologie, 88 (1963), 86-97.
129. C. B. M. McBurney and R. W. Hey, Prehistory and Pleistocene Geology in Cyrenaican Libya (Cambridge, 1955).
130. Lhote, “L'évolution de la faune,” 83-118.
131. E.g., see Spaulding, “Dimensions,” 437-456, and I. Rouse, “The Classification of Artifacts in Archaeology,” American Antiquity, 25 (1960).
132. J.-C. Gardin, “Four Codes for the Description of Artifacts: An Essay in Archaeological Technique and Theory,” American Anthropologist, 60 (1958), 335-357.
133. Rouse, “Classification.”
134. Fabrizio Mori, “Short Conclusions on the Discussion of the Chronological Problem of Saharan Rock-Art,” Prehistoric Art of the Western Mediterranean and the Sahara, L. Pericot Garcia and E. Ripoll Perelló, eds. (New York, 1964), 248.
135. Rhotert, Libysche Felsbilder.
136. Gardin, “Four Codes,” 335-357.