4.1 Binding agents
Evidence for preparation of ochre colourants with either liquid or semi-liquid binding agents has been identified via mass spectrometry (Izzo et al., 2022; Villa et al., 2015), and ethnography (Borg & Jacobsohn, 2013; Hill, 2001). For example, ochre unguents have been unearthed from the funerary beds of Mayan royalty (Vázquez de Ágredos Pascual et al., 2015, 2018). In another instance paint from red pigmented burial artefacts was found to have been mixed with animal fat (DoménechCarbó et al., 2020). Gas chromatography-mass spectrometry (GC-MS) and pyrolysis gas chromatography-mass spectrometry (Py‐GC/MS) were used to identify organic components used as binders in Mayan burial treatments (Izzo et al., 2022). These authors identified vegetable drying oils, potentially derived from chia seeds, possibly mixed with bituminous substances, and suggest they were used as binding media by some Maya for corpse treatments. Moreover, GC-MIS and SEM with energy-dispersive X-ray spectroscopy (SEM/EDS) analyses on stone flakes from Sibudu (49 ky BP: South Africa) have identified a milk casein and ochre mixture, providing the first evidence for milk as a binder (Villa et al., 2015). Their results indicate the liquid was used to form paint, not an adhesive, that may have been applied to human bodies.
Ethnographic research of paint pigments and their preparation in Papua New Guinea reports that iron pigments are mixed with different binders depending on their purpose or intended endurance. For instance, when painting objects that are meant to perish or that will be repainted, water is used as a binder; however, when the paint is meant to be enduring, such as for use on houses, the binder will contain a resinous plant sap (Hill, 2001). If a sheen is required tree resin (tigaso: Campnosperma brevipetiolata) may be added. When oils and resins age, the paste may become brittle due to cross-linking or polymerisation, and thus some Papua New Guinea peoples use fruit pulp, while tigaso, blood, and pig fat (saturated fatty acids that do not dry through polymerisation) are also added as binders for sacred objects in the Highlands. When pig fat is used the paste will fail to dry and remain sticky (Hill, 2001). Other ethnographic reports note ochre is mixed with a greasy substance when it is applied at the last stages of hide tanning (Rudner, 1982 in Dubreuil & Grosman, 2009).
Given the presently limited evidence for the chemical composition of ochre mixtures used in prehistory, particularly on corpses, it was decided for this experiment to use albumen from domestic chicken eggs, as it would be sufficiently viscous to adhere to the piglets, but not so fatty as to fail to dry or to attract carnivore activity. Until chemical analysis shows otherwise, it is hypothesised here that binders may have been chosen based on 1) what was available geographically and seasonally, and 2) their functional or symbolic intention. This is not to suggest that the binder is of secondary importance. In fact, it may be that in certain instances when ochre treatments have a demonstrative function (e.g., in tanning) that it is the binder and not the ochre itself that has a preservative effect.
4.2 Perinate/neonate bone and microbiome development
In a previous experiment using juvenile piglets of varied age, only the stillborn sample of those buried reached partial skeletonisation; the others retained significant amounts of flesh, and moreover, adipocere, which was hypothesised to have resulted from high subcutaneous fat and the anoxic, wet burial environment (White & Booth, 2014). The only other stillborn sample in this experiment was sub-aerially exposed, reaching skeletonisation after eight weeks. Both stillborn samples were rated a tunnelling score of 0, indicating no microbial damage. Older piglets took four to six weeks longer to skeletonise, some of which were also rated low tunnelling scores (White & Booth, 2014). Three of the four clearly stillborn individuals in Booth et al.’s (2016) study of human archaeological remains were rated an Oxford Histological Index (OHI) of 5, indicating excellent preservation; the other perinate and neonate individuals were variable in their manifestation of bioerosion. Barker (2019) found high levels of bioerosion in their small dataset of human juvenile remains, save for the only foetal sample, which was rated an OHI 5; however, this sample was also the only to derive from a non-depositional context. The only other potentially stillborn or prematurely born individual exhibited strong evidence of bioerosion, though it may be that they survived long enough to be fed and thus their gastrointestinal flora were sufficiently developed to produce osteolytic bacteria (Barker, 2019).
We initially hypothesised that different histotaphonomic signals would be identified in samples deriving from intentional burials in comparison to those left exposed. Interestingly, our piglets, despite having been given different burial treatments and being left to decompose in different environments, were all well preserved, exhibiting either no bioerosion or minor, localised areas of demineralisation, particularly along the periosteal surface. This homogeneity in ratings may result from the stillborn/perinatal status of the piglets, all of which died prior to suckling. Inhibited bioerosion is often reported for stillbirths and short-lived perinates in the archaeological record (Booth, 2016; Booth, Redfern & Gowland, 2016; White & Booth, 2014).
While the minor evidence for bioerosion in our dataset may derive from the burial environment, the low level of demineralisation may also result from hereditary or congenital illnesses, such as perimortem sepsis. Moreover, the infant skeleton is composed of woven bone with minimal Haversian systems (Pfeiffer, 2006), and thus lacks Haversian and Volkmann canals through which microbiota are hypothesised to disseminate (Carlson et al., 2022; Turner-Walker, 2012; Yoshino et al., 1991). Alternatively, it may be that the demineralisation seen in the virtual sections results from a non-microbial form of diagenesis that cannot be visualised with microCT images at this resolution. However, and importantly, results from the whole body human taphonomic experiment by Mavroudas et al. (2023) suggest the stage at which microbial bioerosion is initiated may take place later in the post-mortem interval. Therefore, a longer interval between deposition and exhumation may be required to see evidence of bioerosion.
4.3 Insect repellant properties, and bodily odours
Although albumen is an organic substance that may attract flies, it was observed that the entire bodies of the paste-treated piglets (Piglets 1 and 4) were not initially attacked (unlike Piglets 2, 3, and 5), with flies attracted specifically to facial orifices. When applied as a powder, the flies accumulated over the entire corpse, scouting for areas of ingress. Two potential reasons for this, both mechanical, are hypothesised, 1) the thickness of the paste itself provides a barrier preventing flies and other insects from penetrating the paste to oviposit on the skin or to enter its orifices, and 2) the thickness of the paste may provide a temporary barrier that mitigates the release of gases that attract creatures such as insects, but which repel humans. These hypotheses are partially supported by the experiment because fewer insects were attracted to Piglets 1 and 4; however, the areas on which they landed were still the mouth and nose, despite the fact that both were packed full of paste. Moreover, although the piglets were not particularly odourous during the experiment (potentially due to their recently frozen state), insects still landed around the face. This suggests that although humans could not smell the gases, insects could, and were attracted to these moist, relatively open areas. Female flies are uniformly attracted to the facial region, but will also accumulate around the anus and genitals if they are not covered (Goff, 2009). However, in the present experiment, they were not attracted to the anal or vaginal orifices. It may also be that the effect is chemical and results from the binder.
Limited experimental research suggests that ochre treatment does not deter carnivore activity by masking smells of carcasses in graves (Rausing, 1991); though, in Rausing’s experiment the carcasses were sprinkled with ochre, not encased in a paste. Izzo et al. (2022) hypothesise that scented ointments and ochres used in Mayan burials served to mitigate the odours produced by putrefaction, which would be particularly important for funeral customs that last many days, while also functioning symbolically to represent blood. However, given that Piglet 4 eventually burst due to a build-up of gasses trapped by the ochre casing which disarticulated the skeletal remains, if ochre was used to mask smells for funerary rituals, this may have been only effective for a short period of time, particularly in warm climates or during hotter seasons.
Rifkin (2015) assessed the efficacy of ochre treatments (dry powder, ochre and animal fat, and ochre and clarified butter) as a mosquito repellent when applied to human skin, finding results similar to the present experiment. The untreated surface received 33.655% of visits and 42.89% of bites. The dry ochre powder received 19.79% of visits and 29.81% of bites. Notably, the ochre and fat treated surface received 30.4% of visits and 48.81% of bites, more bites than the untreated surface. However, the ochre and clarified butter treated surface received both the fewest visits, 16.18% and the fewest bites, 24.36% (Rifkin, 2015). The greater attraction of mosquitos to the ochre and fat treated surface than the ochre and clarified butter treated surface, “probably occurs because fat degrades into various carboxylic acids, triacylglycerols and CO2. These chemicals mimic the olafactory [sic] signatures of human breath, feet and sweat, which are responsible for attracting mosquitoes” (Rifkin, 2015:67/70 citing Douglas et al., 2005 and Syed & Leal, 2008). Hill (2001) also noted that animal proteins and plant carbohydrates in paints can attract moths. For example, the Papua New Guinea National Museum found untreated cellulosic string bags were unaffected by insect activity, though bags treated with pig fat were attacked by webbing clothes moths (Tineola bisselliella). Lastly, Trájer (2022) suggests that because insects like tse-tse flies and mosquitoes are attracted to the L-lactic acid content of human sweat (Coutinho-Abreu et al., 2021; Dekker et al., 2002; Vale, 1979), covering surfaces with many sweat glands with an ochre mixture may have served as an insect repellent function.
4.4 Antibacterial/antiseptic properties and use as a tanning agent
It has been argued that ochre may have been used to process animal hides because it has antiseptic, antibacterial, and/or antifungal properties (e.g., Audoin & Plisson, 1982; Couraud, 1988; González & Ibáñez, 2002; Rifkin, 2011; Velo, 1984) due to the assumed ability of iron oxides to impede collagenase as reported by Mandl’s (1961) work with metal salt solutions (Watts, 2009). It is also suggested that ochre can absorb grease (Audoin & Plisson, 1982; Ibáñez & González, 1996; Philibert, 1994). However and importantly, as noted by Watts (2009), the belief that iron oxides in ochre provide an antiseptic or tanning effect may arise from a basic misunderstanding of chemistry, with authors (e.g., Keeley, 1980; Knight, Power & Watts, 1995; Wadley, Williamson & Lombard, 2004) assuming that iron oxides, which are relatively insoluble, have the same properties of some soluble iron salts (e.g., iron sulphate), which are reported as potential tanning agents (Tonigold, Hein & Heidemann, 1990 in Watts, 2009; Martín Ramos, Gil & Martín-Gil, 2022). However, Trájer (2022:2, citing Kaiser & Sulzberger, 2004 and Pal & Sharon, 1998) cites experimental evidence demonstrating that iron-oxides are solar activated by “photochemical conversion-initiated free radical production”, which can impart an antimicrobial effect. Moreover, although Rifkin (2011) found that red ochre high in iron content functioned as a better hide preservative than yellow ochre or kaolin (in agreement with Audoin & Plisson (1982)), the author also noted that removal of fat was necessary to avoid decomposition. Watts (2009) argues that the function Wadley et al. (2004) and Wadley (2005) who cite Audoin and Plisson (1982), find for ochre’s antibacterial properties are better explained by its ability to desiccate that to which it is applied (Phillibert, 1994). Clearly further experimental research is required to ascertain the exact circumstances under which ochre may have a sanitising effect. Given all three ochre-treated piglets were completely skeletonised by the end of the experiment, we reject the idea that ochre itself has tanning or anti-microbial properties.
Lastly, the antibacterial effects of ochre have also been debated in other fields of research, such as ornithology. It has been suggested that iron oxides have some antibacterial and antiparasitic properties, and that bearded vultures that bathe their feather in iron oxides do so for sanitary purposes (Arlettaz et al., 2002; Tributsch, 2016). However, there is no evidence that supports this. Instead it was found that iron oxides are not toxic to mallophaga that eat feathers (Frey & Roth-Callies, 1994), nor to other ectoparasites (Negro et al., 2002). This is in keeping with the fact that bacteria actively search for iron, which is necessary for growth, metabolism, and replication (Miller & Britigan, 1997; Negro et al., 2002). Given that previous research failed to take into consideration that iron oxides may require UV radiation to activate (Pal & Sharon, 1998; Kaiser & Sulzberger, 2004), Margalida et al. (2019) performed an experiment on iron oxides/ochre incubated in both the light and dark to assess their antibacterial effects on several gram-positive and -negative bacteria (Bacillus licheniformis, Kocuria rhizophila, and Escherichia coli), finding that none of these species were affected by iron oxides, regardless of light conditions.