Detecting milk proteins in ancient pots


Deciding whether to farm cattle for milk or beef was just as complex in the past as it is today. Compared with meat production, dairying is a high-input, high-output, high-risk operation indicative of an intensive, sophisticated economy, but this practice is notoriously difficult to demonstrate in the archaeological record1. Here we provide evidence for the presence of milk proteins preserved in prehistoric vessels, which to our knowledge have not been detected before. This finding resolves the controversy that has surrounded dairying on the Scottish Atlantic coast during the Iron Age2,3,4,5 and indicates that farming by the early inhabitants of this harsh, marginal environment was surprisingly well developed.


The analysis of sorbed lipid residues in prehistoric ceramics has provided a powerful new indicator of how vessels were used6,7,8. Although proteins are more diagnostic of specific foodstuffs than lipids, they are difficult to extract from archaeological ceramics9. We have developed an immunological detection method, the digestion-and-capture immunoassay (DACIA)10, which overcomes this difficulty by dissolving the ceramic then capturing the liberated proteins for immunodetection.

We obtained sherds from nine coarseware cooking vessels, dated to the middle of the first millennium bc, from the fill of an Early Iron Age house at Cladh Hallan, South Uist, in the Outer Hebrides, and analysed them by DACIA. Extracts were tested using a monoclonal antibody raised against heat-degraded and dephosphorylated bovine α-casein (about 1.4% w/v milk), which was specific for bovine milk.

Immunological analysis of archaeological materials has been criticized for the lack of negative controls11, so we included an extensive array of reference samples (Fig. 1). Seven of nine of the interior sections of sherds recovered from Cladh Hallan tested positive for casein and the amounts were comparable to those found on experimentally buried milk sherds (Fig. 1). DACIA analysis failed to detect the presence of bovine α-casein in the associated sediment or exterior surfaces of the samples.

Figure 1: Amounts of bovine α-casein present in samples of pottery and soil, as determined by duplicate assay using digestion-and-capture immunoassay with a monoclonal antibody raised against this protein.

Error bars, one standard deviation. The assay is specific only for cows' milk and is able to detect as little as 100 ng protein per g of ceramic (0.1 p.p.m.). Experimental coarseware pots (ceramic 'blank') were used to boil either milk (milk pot) or beef (meat pot) repeatedly and were buried for 1 year in upland soil. Ethnographic pots were obtained from Pakistan (HM) and India (EMP); each had been recently used to prepare dairy products. Cladh Hallan (CH) vessels (inset) were collected from a single site (fill of house 112, South Uist, Outer Hebrides). Domestic cooking pots from Easingwold, North Yorkshire (E), contained large amounts of well-preserved animal fats.

The large number of neonatal cattle remains found at this site (42% of individuals) has been attributed to the deliberate culling of young calves in order to preserve fodder in an adverse environment2,3 or to sustain a high-input dairying economy4,5. The presence of bovine α-casein on a substantial number of sherds (Fig. 1) lends support to the latter interpretation. Our successful characterization of protein residues after 2,500 years demonstrates the potential of DACIA as a high-resolution technique for determining how archaeological ceramics were used.


  1. 1

    Halstead, P. Anthropozoologica 27, 3–20 (1998).

    Google Scholar 

  2. 2

    McCormick, F. in Life on the Edge: Human Settlement and Marginality (eds Mills, C. M. & Coles, G.) 49–53 (Oxbow, Oxford, 1998).

  3. 3

    Gilmour, S. & Cook, M. Antiquity 72, 327–337 (1998).

    Article  Google Scholar 

  4. 4

    Parker-Pearson, M., Sharples, N. & Mulville, J. Antiquity 70, 57–67 (1996).

    Article  Google Scholar 

  5. 5

    Parker-Pearson, M., Sharples, N. & Mulville, J. Antiquity 73, 149–152 (1999).

    Article  Google Scholar 

  6. 6

    Rottländer, R. C. A. in Proc. 24th Int. Symp. Archaeometry 403–405 (Smithsonian Institution, Washington DC, 1986).

  7. 7

    Evershed, R. P. World Archaeol. 25, 74–93 (1993).

    CAS  Article  Google Scholar 

  8. 8

    Dudd, S. N. & Evershed, R. P. Science 282, 1478–1481 (1998).

    CAS  Article  Google Scholar 

  9. 9

    Evershed, R. P. & Tuross, N. J. Archaeol. Sci. 23, 429–436 (1996).

    Article  Google Scholar 

  10. 10

    Craig, O. E. & Collins, M. J. J. Immunol. Meth. 236, 89–97 (2000).

    CAS  Article  Google Scholar 

  11. 11

    Tuross, N., Barnes, I. & Potts, R. J. Archaeol. Sci. 23, 289–296 (1996).

    Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Oliver Craig.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Craig, O., Mulville, J., Pearson, M. et al. Detecting milk proteins in ancient pots. Nature 408, 312 (2000).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing