Early ochre mining in Southern Africa

I was reading back through Bednarik's "Concept-mediated marking in the Lower Palaeolithic," for some background on the ochre-shellfish post, and I ran across this quote in Bednarik's rejoinder to Randall White:

[A]n estimated 100 tons of iron ore was mined at just one Middle Stone Age site, Ngwenya, carbon-dated to about 43,200 B. P. (Dart and Beaumont 1971)

That seemed striking -- that is, I hadn't seen any other references to that site -- so I tracked down a bit more information about Ngwenya. The LSA/MSA hematite quarry was described in short publications by Raymond Dart and Peter Beaumont (1967, 1969, 1971). In recent years, the quarry had been mined by a modern operation, and had been a locus of iron ore mining since LSA times at least. Dart had some interest in the idea of Stone Age mining for iron ore, having found an LSA mine in Zambia many years before. Some old flaked tools and other artifacts were found at Ngwenya, and Dart got some some archaeologists, including Adrian Boshier, to investigate.

The Swaziland National Trust has an informative, although unreferenced, website discussing the findings:

Boshier found three ancient mines named Lion, Castle and Stag caverns. He collected stone tools in and around them made of dolerite, which is foreign to the area. The tools were unlike those normally found on a stone age site, they were more specialised, consisting of choppers, picks and hammer stones. Professor Dart identified them as mining tools. The tools were not confined to the surface layers but were scattered throughout huge depressions which would have been solid haematite. In one of the mines they were lying among and beneath thousands of tons of red iron oxide known as haematite. Enquiries by Boshier among the Swazis elicited the information that haematite deposits had been mined in historic times and that it was the custom to fill in the excavation to avoid offending the spirits of the underworld. Boshier theorised that if the holes had been refilled with haematite then the mine workers must have wanted something other than haematite.

The suspicion, mentioned by Dart and Beaumont (1967), was that specularite (a dark-colored sparkly hematite) was the real objective of the quarrying -- they were also led to this idea by the fact that Castle Cavern had surface exposures of hematite, so it would hardly be worth mining deep into Lion Cavern to get it.

Anyway, once the initial excavations found a substantial antiquity for quarrying at the site, they went deeper to see how old the activity really was:

From six to eight feet, there were undoubted Middle Stone Age artefacts together with some possible Later Stone Age tools. From eight feet to worked bedrock at over eleven feet, the deposit yielded some 23,000 artefacts which belong unquestionably to a middle stage of the Middle Stone Age. Occasional stone mining tools were also found. Well-defined ash levels indicated that the assemblage was in situ. Quartz, white quartzites, grey-and-white dappled quartzite, black indurated shales, and greenish cherts were the main materials used. Most of these rock-types occur on a ridge overlooked by and about a quarter of a mile from the cavern. Some of the exposures there bear clear evidence of having been flaked. The dappled grey-and-white quartzites come from exposures about a mile and more northwest of the site (Dart and Beaumont 1969:127).

The extent of the Middle Stone Age activity at the site suggested that a vast amount of hematite had been removed:

At least 50 tons of haematite rich in specularite must have been removed from Lion Cavern; two-thirds of it during the Middle Stone Age (Dart and Beaumont 1967:408).

Their dating of charcoal nodules (Dart and Beaumont 1967) in the "middle to lower levels of the Middle Stone Age level" were 22,280 +/- 400 BP and 28,130 +- 260 BP. This was later revised (Dart and Beaumont 1971) to a larger number, based on further survey. Likewise, Dart and Beaumont (1971) provided an earlier date at or above the limits of radiocarbon, 43,200 BP.

None of this informs us about the earliest use of red pigments, which is far older, or the earliest evidence of quarrying, which is also far older. It's probably even irrelevant to the development of the MSA in southern Africa, which, again, is far older.

But in the context of the recent literature, highly focused on the "first" evidence of ochre use, or ocher engraving, or "symbolic behavior," a site like this one can get lost in the noise. Systematic utilization of one site for one purpose is not recent in the archaeological record, but when we find evidence of such places, they can be the most informative about the activities of people in the past, their transfer of information with each other, and their acquisition of resources across relatively long distances.


Bednarik RG. 1995. Concept-mediated marking in the Lower Paleolithic. Curr Anthropol 36:605-634.

<p class=”cite”Dart RA, Beaumont P. 1967. Amazing antiquity of mining in Southern Africa. Nature 216:407-408. <a href=”http://dx.doi.org/10.1038/216407a0</a>doi:10.1038/216407a0</a></p>

Dart RA, Beaumont P. 1969. Evidence of iron ore mining in Southern Africa in the Middle Stone Age. Curr Anthropol 10:127-128.

Dart RA, Beaumont P. 1971. On a further radiocarbon date for ancient mining in southern Africa. S Afr J Sci 67:10-11.