The axiom that rock art is notoriously difficult to date serves only to paint a partial picture of the inconsistent and contested chronological records of rock art in Africa. For example, where research has focused on interpretation, chronology has been less prominent and as such the capacity for judging meaningful relationships between sites and imagery has been inhibited; by contrast where chronologies have led research agendas, the temporal and spatial relationships are much clearer, but chronologies are hotly disputed. A significant obstacle is the challenge in directly dating rock art, and current research is exploring ways forward in refining these techniques. Here, we give an overview of dating methods and developed chronologies to date in rock art regions across the continent.
Determining the age of rock art depictions has always been one of the main goals of research, and a wide range of techniques have been developed to try to assign a date for rock art images throughout the world. There are two main approaches to rock art dating: relative and absolute.
Relative chronologies aim to organise the images from the oldest to the more recent, even if their exact dates are not known, providing the relative position of groups of depictions over time. It uses methods such as the analysis of superimpositions (the figures on the top have to be younger than those underneath), the study of depictions of animals already extinct or newly introduced in areas (for example, camels in the Sahara) or objects that have a known timeframe of usage, such as ships, firearms, chariots, etc.
In some cases, graffiti can be found alongside figures, and linguistic studies can help to determine the age of both texts and depictions. Of course, one of the best ways to establish a relative date for rock art is when it is covered by archaeological deposits that can be dated through radiocarbon methods or archaeological materials. However, that date only states the minimum age of the rock art depictions, as they could have been covered hundreds or thousands of years after the images were painted or engraved. Moreover, these situations are very rare, as most of the time rock art is situated in the open air, without associated archaeological sites.
The second type of dating is called absolute dating and provides concrete dates (or more often, intervals of age) in which the images were made. Although this could seem the perfect solution for the dating of paintings, the methods of absolute dating can only be used in very specific contexts, and often have serious problems of accuracy and reliability. The most efficient is based on the radiocarbon (carbon-14) method, measuring the amount of carbon remaining in pigments, although in many depictions the remnants of paint are so sparse that the size of the samples is too low to be measured. Besides, this method is destructive, as part of the painting has to be removed from the original site to collect the sample.
The study of engravings is even more difficult, as it is based on the changes that affect their surface after they were made. The most obvious is a progressive change in colour, making the older engravings seems “darker” than the newer. When the changes in colour are due to chemical, internal processes, the term used is patina, while when they are caused by the accumulation of particles the process is known as varnishing. Finally, the progressive weathering of the engraving’s traces can also be measured. By studying the pace of these natural processes, scientists can propose dates for the engravings. However, as with carbon-14, these methods have several problems that affect their suitability: the study of these natural processes is slow and costly, and the results can only be applied to small, concrete areas. There are also problems with the accuracy of the measurements, and risks of contamination of samples are high. Therefore, although they are promising, the absolute dating methods are still far from providing generalized and relatively easy ways to reliably date rock art.
In practice, the dating of rock art is achieved through a combination of relative and - when possible - absolute dating methods, although relative techniques are still the more broadly used. Concepts such as style, technique and iconography are still key tools of analysis, but as absolute dating accuracy and availability grow, the chronological framework of rock art will be substantially improved. As scientific research advances, the techniques of rock art dating become more refined and trustworthy, and in the future, significant progress is expected, which will mean a revolution for world rock art chronologies.
Chronologies in northern Africa
The importance of dating has been at the forefront of much research in North African rock art, yet there is no consensus of opinion about the precise dating of the phases of rock art in this region. The construction of rock art sequences has been predicated on stylistic analysis, on which most scholars agree in principle, and broadly conforms to the following:
- Early Hunter, Wild Fauna Period or Bubalus Period: 12,000 – 8,000 years ago
- Round Head Period: 10,000 – 8,000 years ago
- Pastoral Period: 7,500 – 4,000 years ago
- Horse Period: 3,000 – 2,000 years ago
- Camel Period: 2,000 years ago – present
For the most part, there is general agreement on the dating of the Horse and Camel period, but the earliest phases of rock art - the Bubalus/Round Head/Pastoral periods, are much more contentious. Researchers Henri Lhote and Alfredo Muzzolini have argued that the Pastoral period begins around 6,000 years ago, while others, such as Fabrizio Mori, propose an earlier date of 8,000 years ago. Domestic stock is thought to have existed in the Sahara from around 7,500 years ago, making a date of 8,000 years old for paintings of cattle (which may have been wild) feasible. Dating the end of this period is just as problematic, with Lhote and Mori placing it between 3,000-4,000 years ago and Muzzolini about 1,000 years later. However, the largest discrepancy relates to the Bubalus and Round Head periods, where significantly disparate chronologies have been proposed by Lhote, Mori and Muzzolini. The most extreme of these is the Bubalus period, where Mori advocates an Early/Long chronology proposing that the earliest images date back 12,000 years, while Muzzolini suggests a Recent/Short chronology proposing a much later emergence of art in the region at 6,000 years ago.
The application of direct dating will be central in resolving these substantial differences and has yielded some fruitful results on archaeological sites in making meaningful correlations. However, for the time being, the chronologies of Saharan rock art remain ambiguous and contested.
Chronologies in southern Africa
Although a vast amount of work has been undertaken in rock art research in southern Africa, the focus has been largely placed on the meaning of images and their place within a rich ethnographic record. Little attention has been given to investigating temporal and spatial relationships, and as such there is a conspicuous absence of secure regional chronologies for southern African rock art.
Based on observations in the field, relative sequences were proposed by Patricia Vinnicombe (1976) for the southern region of the Drakensberg and by Revell Mason (1933) and Harald Pager (1971) for the northern region. Subsequently, several sites in South Africa have been analysed using a method termed the Harris Matrix. In its original context the Harris Matrix was a diagrammatic tool used in archaeological fieldwork to determine the temporal succession of deposits and stratigraphic relationships. Its use in South African rock art has been limited and often on isolated sites, and one of the drawbacks in this methodology is the lack of information on time lapses between successive images. While a few radiocarbon and Accelerator Mass Spectometry (AMS) dates have been secured for some sites, the absolute dates need to be matched with relative chronologies, which to date are still not firmly established.
Much rock art research has focused on the San rock paintings found in southern Africa (Botswana, Lesotho, Malawi, Mozambique, Namibia, South Africa, Swaziland and Zimbabwe). The subject matter of San rock art is extremely varied but regional characteristics can be seen; for example, depictions of eland dominate much of the rock art in South Africa, while kudu and elephants prevail in Zimbabwe. The earliest San art is thought to be up to 10,000 years old and continues possibly well into the twentieth century in the Drakensberg mountains.
The spread of farming to South Africa by Bantu-language speakers around 3,500 years ago brought another style of rock art, until it ceased around the mid-1900s. The art is predominantly in white, and applied using fingers, which gives it a rough appearance. Subject matter is varied but dominated by humans and animals. Later rock art reflected contact with European settlers/colonisers, attested to by the numerous depictions of steam trains, soldiers, settlers and guns.
Chronologies in central and eastern Africa
Little research has been conducted, either on absolute or relative rock art chronologies in central Africa, and is therefore still poorly understood by comparison with northern and southern Africa. Promising recent radiocarbon dates have been reported from paintings in Angola, but the absence of contextual information with these dates renders them unproductive. Dating rock art by association with archaeological deposits has not proved entirely successful either. Relative chronologies have been suggested, and while there are regional variations, in general two styles of rock art have been identified and associated with a relative chronology: Twa and Sandawe.
Twa-style art dates to around 3,000 to 1,000 years ago and was made by hunter-gatherer groups in eastern and central Africa. Depictions of animals are rare in Twa art and usually consist of geometric designs, often concentric circles. Ancestral Sandawe art, found mainly in central Tanzania, is thought to date from around 9,000 years ago to the recent past, and includes fairly large naturalistic images of animals in both hunting and domestic scenes. Figures are frequently holding bows and arrows, and are often depicted with elaborate hairstyles and/or headdresses.