Gas chromatography/mass spectrometry, proteomic and scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM/EDS) analyses of residue on a stone flake from a 49,000 year-old layer of Sibudu (South Africa) indicate a mixture of ochre and casein from milk, likely obtained by killing a lactating wild bovid. Ochre powder production and use are documented in Middle Stone Age South African sites but until now there has been no evidence of the use of milk as a binder. Our analyses show that this ochre-based mixture was neither a hafting adhesive nor a residue left after treating animal skins, but a liquid mixture consisting of a powdered pigment mixed with milk; in other words, a paint medium that could have been applied to a surface or to human skin. The significance of our finds also lies in the fact that it establishes the antiquity of the use of milk as a binder well before the introduction of domestic cattle in South Africa in the first millennium AD.
Citation: Villa P, Pollarolo L, Degano I, Birolo L, Pasero M, Biagioni C, et al. (2015) A Milk and Ochre Paint Mixture Used 49,000 Years Ago at Sibudu, South Africa. PLoS ONE 10(6): e0131273. doi:10.1371/journal.pone.0131273
Editor: Nuno Bicho, Universidade do Algarve, PORTUGAL
Received: February 20, 2015; Accepted: June 1, 2015; Published: June 30, 2015
Copyright: © 2015 Villa et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: The Sibudu collections are housed in the Acacia unit of the Evolutionary Studies Institute at the University of the Witwatersrand and are catalogued 2931CA. All research data generated by this paper are within the paper and its Supporting Information files.
Funding: Research by PV and LP was funded by the National Science Foundation (BCS 0613319), the Paleontological Scientific Trust Scatterlings of Africa programme, the School of Geography, Archaeology and Environmental Studies and the Institute for Human Evolution, University of the Witwatersrand.
Competing interests: The authors have declared that no competing interests exist.
Paint is a mixture of solid pigment and a liquid vehicle that can be applied to a surface or to a body for decorative or protective  purposes. We report here an early case of a paint, preserved as a mineral and organic residue on the working edge of a stone flake from a Middle Stone Age (MSA) layer of Sibudu (South Africa), dated to c. 49,000 years ago . Gas chromatography/mass spectrometry (GC/MS), proteomic and SEM/EDS analyses indicate a mixture of ochre and casein from bovid milk . The presence of bovid milk is surprising at this time since domestic cattle are documented in South Africa only in the Iron Age, about 300 AD . However, hunting practices in the MSA suggest that milk could be obtained by killing a lactating female wild bovid . The novelty of our discovery is highlighted by what is known about ochre use from previous research.
Ochre use is ancient, predating anatomically modern humans in both Europe and Africa. Ochre, possibly in a liquid solution, was used by Neanderthals at Maastricht-Belvédère, Netherlands, 200–250 ka BP . Use of ochre in the African Middle Pleistocene is clearly documented, as shown for example at site GnJh-15 in the Kapthurin Formation, Kenya, 285 ka BP , in the 250 ka BP Lupemban Industry of Twin Rivers, Zambia , and in the early MSA at Sai Island, Sudan, 200 ka BP . There is an increase in the collection and use of ochre showing grinding, scraping or scoring marks at final Middle Pleistocene sites such as Border Cave layers 5BS and 5WA [10–11] and Pinnacle Point Cave 13B, South Africa (164 ka BP) [12–13]. In the Late Pleistocene use of ochre is a common feature of the South African MSA. Discoveries of ground ochre pieces, ochre powder or traces of ochre on grindstones or stone tools have been made at sites such as Apollo 11  Die Kelders , Blombos Cave [15–16], Diepkloof , Hoedjiespunt 1 , Hollow Rock Shelter , Klasies River [20–22], Klipdrift , Rose Cottage  and Sibudu [25–26]. Ochre pieces are also common in all of the South African Cape west coast MSA shell middens . Deliberately engraved and incised ochre pieces have been found at Blombos and Klasies River Cave 1 [22, 28–29].
Ochre powder is excellent for tanning hides because it reverses the process of decay [30–31] but only circumstantial evidence implies the use of ochre for hide preparation in the MSA. Ochre-stained bone awls from Blombos Cave are thought to have pierced hides, based on experimental replications . Residues and use wear on three Sibudu scrapers tentatively imply their use with ochre on hides  but a larger sample size is needed. Ochre had other potential roles in the MSA: ochre nodules seem to have been used as soft stone hammers  and at MSA sites like Rose Cottage, Sibudu and Umhlatuzana ochre powder was used in compound adhesives for hafting stone tools [24, 34–35].
Watts  suggests that the Pinnacle Point red ochre powders were intended as ingredients for body paints used during ritual performance. The inference of early paint manufacture has some support because an ochre-rich compound that may have been blended with marrow or fat was found stored in two Blombos abalone shells with ages of 101 ± 4 ka BP . The purpose of such paint remains speculative, but Henshilwood and colleagues propose that possible uses include decoration of various surfaces and skin protection. Some types of ochre are effective sun screens . Ochre traces on the surface or inside perforated marine shells from Blombos (more than 40 Nassarius kraussianus shell beads from the Still Bay layers) Sibudu (three perforated Afrolittorina africana from the Sill Bay layers) Border Cave (the Conus shell associated with the burial of BC3 an infant with an estimated date of c. 76 ka corresponding to the Howiesons Poort at the site) and North African sites suggest that they were worn against painted bodies, but it is also possible that either the shells or strings for threading them had been deliberately covered with red ochre [37–42].
Materials and Methods
Sibudu is a large rock shelter, approximately 40 km north of Durban and 15 km inland from the Indian Ocean. It has a 2.7 m deep MSA sequence spanning from ca. 77,000 to 38,000 years ago excavated from 1998 to 2011 under the direction of Lyn Wadley  (S1 Text. Site setting and excavation). The stone specimen, an unretouched flake of dolerite (length = 26.7 mm; width = 17.5 mm; thickness = 5.1 mm) was recovered from layer MOD dated by single grain OSL to 49.4 ± 2.1 ka . The residue is along the thin, unretouched, right lateral edge (Fig 1A).
(B) Detail of residue left after sampling for chemical and proteomic analyses. (C) View at 40 x. (D) View at 128 x.
Dolerite is an igneous rock that occurs in the vicinity of the site . The flake was found in square D6, quadrant d, where 98 flakes of dolerite were also found. Square D6 yielded 19 pieces of unutilized and 8 pieces of utilized ochre  405 blades and flakes, 3 cores and 32 retouched pieces. In layer MOD there is a total of 192 ochre pieces.
The excavation permit (number 007/09) was issued by Amafa KwaZulu-Natal Heritage Agency in accordance with KwaZulu-Natal Heritage Act 4 of 2008. The permit holder is Lyn Wadley. The Sibudu collections are housed in the Acacia unit of the Evolutionary Studies Institute at the University of the Witwatersrand and are catalogued 2931CA. The export permit of the flake and other pieces for analyses was issued by Amafa KwaZulu-Natal no. 0011/02. Analysis of the MOD lithic assemblage is provided in S2 Text.
Analysis of the pigment
A small fragment of the residue (dimensions ca. 0.05 x 0.05 x 0.01 mm3) was carefully hand-picked and mounted on a Gandolfi camera (R = 57.3 mm) for collecting the powder X-ray diffraction pattern using Ni-filtered Cu Kα radiation. Owing to the small dimensions of the sample, a 7-day exposure time was set. However, even after 7 days we did not register any sign of diffraction. We tried to capture some weak diffracted spots by means of an air-cooled CCD detector (more sensitive with respect to the photographic film used with the Gandolfi camera) mounted on a Bruker Smart Breeze diffractometer, by rotating the same crystals for several minutes along the φ axis with monochromatized Mo Kα radiation. However we still did not get any diffraction. This means that the pigment (Fig 1B–1D) consists of non-crystalline or poorly crystalline material.
The same fragment was studied with a Philips XL30 Scanning Electron Microscope (SEM) equipped with an EDAX DX4 Spectrometer for chemical (elemental) analysis. A total of 5 spot analyses were performed. All spot analyses showed the same elements (in order of decreasing relative abundance): Si, Fe, Al, K, Mg, Na, and minor Ti. The analyses can be grouped in two subsets, with a different ratio between iron, on one side, and the other major elements, on the other (Fig 2).
We conclude that the pigment consists of a mixture, in variable proportions, of ochre which gives rise to the peak of Fe, and a phyllosilicate possibly belonging to clay minerals, e.g. illite/montmorillonite, giving rise to the other peaks. A shale quarry, likely source of clayey ochre, is 1 km from the site .
Gas chromatography/mass spectrometry
A micro-sample (3.3 mg) of the amorphous residue on the MOD flake was analyzed by gas chromatography/mass spectrometry (GC/MS) using a combined procedure for the identification of lipids, waxes, proteins, resinous materials and polysaccharides possibly present in the same sample (S4 Text, Chemical analyses). To rule out the possibility of contamination due to burial conditions, some milligrams of sediment were subjected to the same combined analytical procedure.
Protein fraction (analysis of amino acids).
The amount of amino acids (0.77 μg) in the sample was significantly above the quantitation limit of the procedure (0.33 μg). The amino acidic profile (Table 1 below and Fig A in S4 Text) is consistent with the presence of casein, given the high percentages of leucine, proline and glutammic acid.
This was confirmed by submitting the amino acidic profile of the sample to Principal Component Analysis (PCA) together with a dataset of more than 100 reference samples of animal glue, casein, and egg, analyzed as raw materials and paint layers mixed with pigments, both naturally and artificially aged. The resulting score plot is reported in Fig 3, and shows that the sample is located in the cluster of casein. The absence of animal glue is confirmed by the absence of hydroxyproline. The amount of amino acids in the soil sample was below detection limit.
Polysaccharide fraction (analysis of sugar and uronic acids).
The analysis of the sugar and uronic acids fraction did not yield significant results: the amount of analytes was below the quantitation limit of the procedure (0.6 μg).
Lipid and resinous fraction.
Several fatty acids were identified (Table A and Fig B in S4 Text). In particular, linear monocarboxylic acids C12-C24 with a minor amount of azelaic acid and glycerol were detected, suggesting the presence of a lipid of plant origin. Sediment samples were used as analytical blanks. The amount of fatty acids in the sediment samples was several orders of magnitude lower than that in the flake sample. Therefore, we can exclude soil contamination. However, the flake fatty acid profile may have been altered by bacteria present in the sediment. Thus, we cannot be confident that the lipid material was actually of plant origin.
In conclusion, the occurrence of casein was inferred from the PCA score plot. This interpretation was tested and confirmed by proteomic analyses.
Proteins in the sample were identified following a minimally invasive proteomic procedure using a 4 mg sample of the residue (S4 Text, Chemical analyses). Raw data from nano liquid chromatography-tandem mass spectrometry (LC-MS/MS) were used to query the Swiss Prot database which contains protein sequences of modern species. A database search provided a good score for four bovine milk caseins: Alpha-S1-casein, (P02662), Alpha-S2-casein, (P02663), Beta-casein, (P02666), and Kappa-casein, (P02668). Two collagens [Collagen alpha-1(I) chain, (P02453), and Collagen alpha-2(I) chain, (P02465)] found in bovine species were identified in the sample (Table C in S4 Text).
Proteomics identification is straightforward but it relies on the presence of the sequence to be matched in the database or, at least, that a highly homologous protein from another species is available. Some peptides identical to the ones present in the known protein homologue are then expected to be produced in the enzymatic digestion of the protein of interest and can be matched. Although casein proteins are highly divergent across mammalian milks , they are conserved enough to allow identification, with some peptides identical to the ones present in the old protein homologue. Divergence, moreover, allowed genus-specific sequence information, since several of the matched peptides are specific enough to allow us to confidently identify the bovid origin of the caseins. In Fig F of S4 Text the alpha-S1 casein sequences from Bos taurus were aligned to those from some other potential sources of milk, i.e. human, sheep and goat. The matched peptides in the MOD flake (highlighted in grey and bold) clearly indicate that the casein belongs to an animal in the bovid family. Likewise, alignments of the other identified caseins with homologous proteins are similarly reported by  showing that several of the identified peptides are proteotipic of bovine caseins sequences.
Collagen was not identified in the fraction of the sample that was analyzed by GC/MS given the absence in the amino-acidic profile of hydroxyproline which is a marker of animal glue. This indicates that collagen was not uniformly mixed in the sample as one would expect if collagen was used as animal glue and that the collagen source was most likely a minute bone fragment or a fragment of connective or epithelial tissue (see Discussion).
Type I collagen is sufficiently variable between mammal genera to be taxonomically useful. Collagen from several bone samples of equid and bovid fauna from layer MOD was extracted and analyzed. Peptide matches from each sample and from the MOD flake were manually examined and aligned to the collagen sequences from modern species. Significantly comparisons with samples of equid and bovid fauna from layer MOD indicate that the collagens were from a bovid and not from an equid (Figs G and H in S4 Text).
In sum, independent analyses by two laboratories on different fraction of the residue and with different techniques support the identification of casein. Moreover proteomic analysis shows that the casein belongs to an animal in the bovid family.
Following the GC/MS and proteomic analyses which suggested the presence of organic material still preserved in the residue, the sample was considered suitable for direct AMS dating. The analysis was performed at the Oxford Radiocarbon Accelerator Unit (ORAU), University of Oxford, UK.
The direct AMS dating of the residue is 23.3 ± 0.4 ka BP (uncalibrated radiocarbon years before present). Lack of chemical cleaning, a decision imposed by the very small size of the sample (13 mg), meant that any contaminants would remain intact contributing to the overall 14C measurement. The determination therefore is considered minimum age only. The larger than normal standard error is a factor of the small amount of C measured in the sample (390μg), about 5 times smaller than the normal-sized graphites routinely measured at the ORAU.
The significance of this direct measurement lies in the fact that it establishes the antiquity of the use of milk well before the introduction of domestic cattle in South Africa in the 1st millennium AD (S3 Text. Direct dating).
Milk is an emulsion of fats within a water-based fluid which also contains carbohydrates, proteins and minerals. Casein is the major protein component of milk and is extracted by heating milk and adding acid . Casein then precipitates into small, solid globules. Use of casein as a glue requires mixing with quicklime or hydrated lime (calcium hydroxide, also called slaked lime) . Lime is obtained by crushing and heating limestone. However, recent geological mapping shows that there is no limestone in the Sibudu area . For use as a binder for pigments, lumps of casein can also be dissolved and thinned with water and wood ashes (mainly composed of calcium carbonate or calcium oxide). However calcium (Ca) is absent in the spectra of Fig 2. Given the absence of lime and wood ashes in the Sibudu residue, we reject an interpretation of the casein as either glue or a complex form of binder. Instead, we conclude that powdered ochre was simply mixed with milk in its liquid form and casein is the residue left after evaporation, since milk is a water-based liquid.
The identification of a milk and ochre mixture on a 49,000 year old flake from a late MSA layer of Sibudu raises a number of questions indicated below.
How was the milk obtained from wild bovids?
Some African bovids separate from the herd when giving birth and hide their calves until they are strong enough to keep up with the herd as a protection against carnivores . This is the case with the eland (Taurotragus oryx), the kudu (Tragelaphus strepsiceros), and the impala (Aepicerus melampus). The bushbuck (Tragelaphus scriptus) and the red and blue duiker (Cephalophus natalensis and Philantomba monticola) are solitary animals; they hide their young and go off to browse alone. All these species and the buffalo (Syncerus caffer, a herd animal) are represented in the MOD fauna.
Klein  notes that in the MSA deposits of Klasies River Cave I the giant buffalo (Pelorovis antiquus) a very large bovid with a horn span of 3 m or more, falls into two age groups: newborn or perhaps fetal individuals and physically mature individuals with worn molars. According to Klein, the explanation for this peculiar distribution may be that the Klasies people, faced with such a formidable prey, focused on females in advanced pregnancy or in the process of giving birth. Cows are easy prey when giving birth and would already have milk. It would not be difficult for hunters to locate lactating cows, particularly among seasonal breeders. This hunting pattern would result in access to milk and might explain the inhomogeneous presence of collagen observed by proteomic analyses. In the process of collecting milk from the mammary glands of a dead animal, a small amount of its epithelial tissue could have been taken as well.
Was the mixture a hafting adhesive?
The use of adhesives made from resin or plant gum combined with ochre has previously been suggested for the hafting of stone points and backed pieces to be used as hunting tools at Sibudu. Howiesons Poort and later MSA tools were studied by optical microscopy in combination with spatial distribution of residues on the tool [35, 51]. Replication experiments [34, 52] demonstrated the usefulness of ochre as a loading agent with products such as plant gum. Microscopic analysis showed that ground ochre was mixed with a plant exudate; experiments suggested that ochre-loaded adhesives are robust, dry fast and are not hygroscopic. Most archaeological residues occurred at the base of the point or on the back of backed pieces implying that the mixture was used for hafting. However optical microscopy alone cannot securely distinguish between plant gum (polysaccharides) and resin (terpenoid secretions from trees)  thus new analyses using GC/MS are appropriate.
The residue on the MOD flake occurs on the thin edge of an unretouched flake, not on the back or proximal part of the piece as would be expected if used for hafting. An alternative explanation is that the flake was used to prepare or apply the ochre-loaded mixture. GC/MS has been used to analyze six tools, dated between 65 and 38 ka BP, from Sibudu (n = 5) and Rose Cottage (n = 1). The residues were preliminarily ascertained by visual inspection. All samples were provided by Lyn Wadley. The purpose was to identify the residues on the MSA tools and to determine whether casein could have been used as glue in a variant of the recipes used at Sibudu.
The results (S4 Text, Chemical analyses) indicate that two of the Howiesons Poort segments, from layer GR in square C6c and from layer PGS in square B5a, contained a relevant amount of diterpenes indicating the use of a conifer resin, possibly from Afrocarpus (syn. Podocarpus) falcatus, and some lipid material. There is neither evidence of plant gums (no polysaccharides) nor proteinaceous material such as casein. No significant molecular markers of organic material were detected on the other samples; neither saccharides nor terpenes were detected. Fatty acids were present in very low quantities suggesting soil contamination. The use of conifer resin for hafting (Podocarpus elongatus) is also indicated by GC/MS analysis on a quartz backed flake dated to ca. 56 ± 10 ka BP in the Late Howiesons Poort at the site of Diepkloof .
As mentioned before, the use of casein as a glue requires mixing with hydrated lime  which is obtained by crushing and heating limestone. Recent geological mapping shows that there is no limestone in the Sibudu area  so, in conclusion, there is no evidence that the MOD flake residue was a hafting adhesive.
Was the MOD flake used to treat animal skin?
Three scrapers from layer SS at Sibudu, dated to 59.6 ± 2.3 ka by OSL , have ochre and animal products (fat, muscle tissue) on their edges. Ochre and fat were quite probably used for processing hides and some of the animal products were the result of use on animal skin . The presence of milk is however not consistent with use for processing hides.
What was the use of the MOD flake mixture?
Our analyses show that this ochre-based mixture was neither a hafting adhesive nor a residue left after treating animal skins, but a liquid mixture consisting of a solid (powdered) pigment mixed with milk; in other words, a paint medium that could have been applied to a surface or on the skin.
The edge of the flake may have been used as a mixing implement for combining the substances. A mixture of milk and ochre may have been used for body painting, skin protection or for painting on a rock surface like a stone slab. The oldest representational rock art in Africa is on seven small quartzite slabs from the Apollo 11 rock shelter in Namibia. The date of 27,500 years ago is an average of 3 radiocarbon samples of charcoal in hearths from the same layer . The pigments and binders of the slabs have not yet been analyzed. Traces of red coloring on ostrich eggshell fragments from the older MSA layers in Apollo 11  imply that there may have been simpler precursors to the representational art. Milk or casein has never been documented as media for pigment in San rock art . In fact, little is known about the methods or ingredients that San used for making paint and even less is known about paint binders [58–61]. Blood was identified in rock art paint at Rose Cottage . Some ethnographies and historic sources mention milk and egg as a binder for rock paint but there is as yet no direct evidence for its use .
Milk was not used as a binder with pigments for body paint by the San (Bushman) nor by the Hottentots and Nama herders of South Africa, Namibia and Botswana; animal or vegetable fat were the most commonly reported media for body paint . However clarified butter is mixed with ochre for body paint used by the Himba of Namibia . Only further research on pigments and binders of rock art in South Africa will allow us to identify similarities or differences that may support one hypothesis over the other.
Nevertheless, obtaining milk by killing a lactating wild bovid and then mixing it with ochre shows that MSA people experimented with coloring materials in creative ways and may have attributed a special significance and value to that product. In the present state of research we cannot demonstrate conclusively whether paint such as that found at Sibudu was a precursor to rock art painting or used for body painting. In either case this find suggests that toward the end of the Middle Stone Age new techniques of symbolic communication were evolving that became, some millennia later, key elements in the social life of hunter-gatherers in South Africa.
S1 Text. Site setting and excavation
S2 Text. Lithic analysis of layer MOD
S3 Text. Direct dating
S4 Text. Chemical analyses
PV is grateful to Scott Ortman, Payson Sheets and John Hoffecker for constructive comments. LB thanks Gennaro Marino for helpful discussions and critical reading.
Conceived and designed the experiments: PV. Performed the experiments: PV LP ID LB MP CB KD RV JJL. Analyzed the data: PV LP ID LB MP CB KD RV JJL. Contributed reagents/materials/analysis tools: ID LB MP CB KD RV JJL LW. Wrote the paper: PV LP ID LB MP CB KD LW.
- 1. Rifkin RR, d'Errico F, Boulliot L, Summers B. Assessing the photoprotective effects of red ochre on human skin by in vitro laboratory experiments. S Afr J Sci. 2015; 111(3/4): 1–8.
- 2. Jacobs Z, Wintle AG, Duller GAT, Roberts RG, Wadley L. New ages for the post-Howiesons Poort, late and final Middle Stone Age at Sibudu, South Africa. J Archaeol Sci. 2008; 35: 1790–1807.
- 3. Colombini MP, Modugno F, Eds. Organic mass spectrometry in art and archaeology. Chichester: Wiley; 2009.
- 4. Robbins LH, Campbell AC, Murphy ML, Brook GA, Srivastava P, Badenhorst S. The advent of herding in Southern Africa: Early AMS dates on domestic livestock from the Kalahari desert. Curr Anthropol. 2005; 46: 671–677.
- 5. Klein RG. Middle Stone Age man-animal relationships in Southern Africa: Evidence from Die Kelders and Klasies River Mouth. Science. 1975; 190: 265–267.
- 6. Roebroeks W, Sier MJ, Nielsen TK, De Loecker D, Parés JM, Arps CES, et al. Use of red ochre by early Neandertals. Proc Natl Acad Sci. 2012; 109: 1889–1894. doi: 10.1073/pnas.1112261109. pmid:22308348
- 7. McBrearty S, Brooks AS. The revolution that wasn’t: A new interpretation of the origin of modern human behavior. J Hum Evol. 2000; 39: 453–563. pmid:11102266
- 8. Barham LS. Possible early pigment use in South-Central Africa. Curr Anthropol. 1998; 39: 703–710.
- 9. Van Peer P, Fullagar R, Stokes S, Bailey RM, Moeyersons J, Steenhoudt F, et al. The Early to Middle Stone Age transition and the emergence of modern human behaviour at site 8-B-11, Sai Island, Sudan. J Hum Evol. 2003; 45: 187–193. pmid:14529653
- 10. Watts I. Ochre in the Middle Stone Age of southern Africa: ritualized display or hide preservative? S Afr Archaeol Bull. 2002; 57 (175): 64–74.
- 11. Grün R, Beaumont P, Tobias PV, Eggis S. On the age of the Border Cave 5 human mandible. J Hum Evol. 2003; 45: 155–167. pmid:14529650
- 12. Marean CW, Bar-Matthews M, Bernatchez J, Fisher E, Goldberg P, Herries AIR, et al. Early human use of marine resources and pigment in South Africa during the Middle Pleistocene. Nature. 2007; 449: 905–908. pmid:17943129
- 13. Watts I. The pigments from Pinnacle Point Cave 13B, Western Cape, South Africa. J Hum Evol. 2010; 59: 392–411. doi: 10.1016/j.jhevol.2010.07.006. pmid:20934093
- 14. Thackeray AI. Middle Stone Age artefacts from the 1993 and 1995 excavations of Die Kelders Cave 1, South Africa. J Hum Evol. 2000; 38: 147–168. pmid:10627401
- 15. Henshilwood CS, Sealy J, Yates R, Cruz-Uribe K, Goldberg P, Grine FE, et al. Blombos Cave, Southern Cape, South Africa: preliminary report on the 1992–1999 excavations of the Middle Stone Age levels. J Archaeol Sci. 2001; 28: 421–448.
- 16. Henshilwood CS, d’Errico F, Marean CW, Milo RG, Yates R. An early bone tool industry from the Middle Stone Age at Blombos Cave, South Africa: implications for the origins of modern human behaviour, symbolism and language. J Hum Evol. 2001; 41: 631–678. pmid:11782112
- 17. Dayet L, Texier P-J, Daniel F, Porraz G. Ochre resources from the Middle Stone Age sequence of Diepkloof Rock Shelter, Western Cape, South Africa. J Archaeol Sci. 2013; 40: 3492–3505.
- 18. Will M, Parkington JE, Kandel AW, Conard NJ. Coastal adaptations and the Middle Stone Age lithic assemblages from Hoedjiespunt 1 in the Western Cape, South Africa. J Hum Evol. 2013; 64: 518–537. doi: 10.1016/j.jhevol.2013.02.012. pmid:23628196
- 19. Evans U. Hollow Rock Shelter, a Middle Stone Age site in the Cederberg. South Afr Field Archaeol.1994; 3: 63–73.
- 20. Singer R, Wymer J. The Middle Stone Age at Klasies River Mouth in South Africa. Chicago: Chicago University Press; 1982.
- 21. Deacon HJ. Two late Pleistocene-Holocene archaeological depositories from the Southern Cape, South Africa. S Afr Archaeol Bull. 1995; 50: 121–131.
- 22. d'Errico F, García Moreno R, Rifkin RF. Technological, elemental and colorimetric analysis of an engraved ochre fragment from the Middle Stone Age levels of Klasies River Cave 1, South Africa. J Archaeol Sci. 2012; 39: 942–952.
- 23. Henshilwood CS, van Niekerk K, Wurz S, Delagnes A, Armitage SJ, Rifkin RF, et al. Klipdrift Shelter, southern Cape, South Africa: Preliminary report on the Howiesons Poort layers. J Archaeol Sci. 2014; 45: 284–303.
- 24. Gibson NE, Wadley L, Williamson BS. Microscopic residues as evidence of hafting on backed tools from the 60 000 to 68 000 year-old Howiesons Poort layers of Rose Cottage Cave, South Africa. South Afr Humanit. 2004; 16: 1–11.
- 25. Wadley L. Cemented ash as a receptacle or work surface for ochre powder production at Sibudu, South Africa, 58,000 years ago. J Archaeol Sci. 2010; 37: 2397–2406.
- 26. Hodgskiss T. Ochre use in the Middle Stone Age at Sibudu, South Africa: Grinding, rubbing, scoring and engraving. J Afr Archaeol. 2013; 11: 75–95.
- 27. Parkington JE. Moderns and middens: Shellfishing and the Middle Stone Age of the Western Cape. S Afr J Sci. 2003; 99: 243–247.
- 28. Henshilwood CS, d’Errico F, Yates R, Jacobs Z, Tribolo C, Duller GAT, et al. Emergence of modern human behaviour: Middle Stone Age engravings from South Africa. Science. 2002; 295: 1278–1280. pmid:11786608
- 29. Henshilwood CS, d'Errico F, Watts I. Engraved ochres from theMiddle Stone Age levels at Blombos Cave, South Africa. J Hum Evol. 2009; 57: 27–47. doi: 10.1016/j.jhevol.2009.01.005. pmid:19487016
- 30. Audouin F, Plisson H. Les ochres et leurs témoins au Paléolithique en France: Enquéte et expériences sur leur validité archéologique. Cahiers du Centre de Recherches Préhistoriques 1982; 8: 33–80.
- 31. Rifkin RF. Assessing the efficacy of red ochre as a prehistoric hide tanning ingredient. J Afr Archaeol. 2011; 9: 131–158.
- 32. Wadley L, Langejans G. Preliminary study of scrapers around combustion features in layer SS, Sibudu, 58 000 years ago. S Afr Archaeol Bull. 2014; 69: 19–33.
- 33. Soriano S, Villa P, Wadley L. Ochre for the toolmaker: Shaping the Still Bay points at Sibudu (KwaZulu-Natal, South Africa). J Afr Archaeol. 2009;7: 41–54.
- 34. Wadley L, Hodgskiss T, Grant M. Implications for complex cognition from the hafting of tools with compound adhesives in the Middle Stone Age, South Africa. Proc Natl Acad Sci. 2009;106: 9590–9594. doi: 10.1073/pnas.0900957106. pmid:19433786
- 35. Lombard M. The gripping nature of ochre: the association of ochre with Howiesons Poort adhesives and Later Stone Age mastics from South Africa. J Hum Evol. 2007; 53: 406–419. pmid:17643475
- 36. Henshilwood CS, d'Errico F, van Niekerk K, Coquinot Y, Jacobs Z, Lauritzen SE, et al. A 100,000-year-old ochre-processing workshop at Blombos Cave, South Africa. Science. 2011; 334: 219–222. doi: 10.1126/science.1211535. pmid:21998386
- 37. Henshilwood CS, d’Errico F, Vanhaeren M, van Niekerk K, Jacobs Z. Middle Stone Age shell beads from South Africa. Science. 2004; 304: 404. pmid:15087540
- 38. d’Errico F, Henshilwood CS, Vanhaeren M, van Niekerk K. Nassarius kraussianus shell beads from Blombos Cave: Evidence for symbolic behaviour in the Middle Stone Age. J Hum Evol. 2005; 48: 2–14.
- 39. d’Errico F, Vanhaeren M, Wadley L. Possible shell beads from the Middle Stone Age layers of Sibudu Cave, South Africa. J Archaeol Sci. 2008; 35: 2675–2685.
- 40. d’Errico F, Vanhaeren M, Barton N, Bouzouggar A, Mienis H, Richter D, et al. Additional evidence on the use of personal ornaments in the Middle Paleolithic of North Africa. Proc Natl Acad Sci. 2009;106: 16051–16056. doi: 10.1073/pnas.0903532106. pmid:19717433
- 41. Grün R, Beaumont P. Border Cave revisited: a revised ESR chronology. J Hum Evol. 2001; 40: 467–482. pmid:11371150
- 42. Millard AR. Bayesian analysis of ESR dates with application to Border Cave. Quat Geochronol. 2006; 1: 159–166.
- 43. Wadley L., Jacobs Z Z.. Sibudu Cave: background to the excavations, stratigraphy and dating. South Afr Humanit. 2006; 18: 1–26.
- 44. Wadley L, Kempson H. A review of rock studies for archaeologists, and an analysis of dolerite and hornfels from the Sibudu area, KwaZulu-Natal. South Afr Humanit.2011; 23: 87–107.
- 45. Hodgskiss T. An investigation into the properties of the ochre from Sibudu, KwaZulu-Natal, South Africa. South.Afr Humanit. 2012; 24: 99–120.
- 46. Lemay DG, Lynn DJ, Martin WF, Neville MC, Casey TM, Rincon G, et al. The bovine lactation genome: insights into the evolution of mammalian milk. Genome Biol. 2009; 10 (4): R43. doi: 10.1186/gb-2009-10-4-r43. pmid:19393040
- 47. Kawasaki K, Lafont AG, Sire JY. The evolution of milk casein genes from tooth genes before the origin of mammals. Mol Bio Evol. 2011; 28: 2053–61.
- 48. Forest Products Laboratory. Casein glues: their manufacture, preparation and application. U.S. Forest Service Research Note, FPL-0158; 1967. pp. 1–15.
- 49. Clarke BM, Schoeman P, Kota MW. The geology of the Verulam area. Pretoria: Council for Geoscience, D&V Premier Print; 2007.
- 50. Stuart C, Stuart T. Field guide to the mammals of Southern Africa. Cape Town: Struik; 1997.
- 51. Lombard M. Finding resolution for the Howiesons Poort through the microscope: micro-residue analysis of the segments from Sibudu Cave, South Africa. J Archaeol Sci. 2008; 35, 26–41.
- 52. Wadley L. Putting ochre to the test: Replication studies of adhesives that may have been used for hafting tools in the Middle Stone Age. J Hum Evol. 2005; 49: 587–601. pmid:16126249
- 53. Burger P, Charrié-Duhaut A, Connan J, Albrecht P. Taxonomic characterisation of fresh Dipterocarpaceae resins by gas chromatography–mass spectrometry (GC-MS): Providing clues for identification of unknown archaeological resins. Archaeol Anthropol Sci. 2011; 3; 185–200.
- 54. Charrié-Duhaut A, Porraz G, Cartwright CR, Igreja M, Connan J, Poggenpoel C, et al. First molecular identification of a hafting adhesive in the Late Howiesons Poort at Diepkloof Rock Shelter (Western Cape, South Africa). J Archaeol Sci. 2013; 40: 3506–3518.
- 55. Wendt WE. Art mobilier’ from the Apollo 11 Cave, South West Africa: Africa’s oldest dated works of art. S Afr Archaeol Bull. 1976; 31: 5–11.
- 56. Vogelsang R. Middle-Stone-Age-Fundstellen im Südwest Namibia. Africa Praehistorica 11. Köln: Heinrich Barth Institute; 1998.
- 57. Rudner I. Paints of the Khoisan artists. S Afr Archaeol Soc., Goodwin Series 1983; 4: 14–20.
- 58. Rudner I. Khoisan pigments and paints and their relationship to rock paintings. Ann S Afr Mus. 1982; 87: 1–181.
- 59. Arocena JM, Hall K, Meiklejohn I. Minerals provide tints and possible binder/extender in pigments in San rock paintings (South Africa). Geoarchaeol. 2008; 23: 293–304.
- 60. Bonneau A, Pearce DG, Pollard AM. A multi-technique characterization and provenance study of the pigments used in San rock art, South Africa. J Archaeol Sci. 2012; 39: 287–294.
- 61. Prinsloo L, Tournié A, Colomban P, Paris C, Bassett ST. In search of the optimum Raman/IR signatures of potential ingredients used in San/Bushman rock art paint. J Archaeol Sci. 2013; 40: 2981–2990.
- 62. Williamson BS. Direct testing of rock painting pigments for traces of haemoglobin at Rose Cottage Cave, South Africa. J Archaeol Sci. 2000; 27: 755.