Why quantify?

Quantification in archaeobotany has been used in two rather different ways. First, it has been used directly as a means of measuring the economic or dietary importance of different species, which led to debate on the reliability of different quantification methods as an indication of ‘importance’ (see for example Hubbard 1975). Ultimately, this debate proved fruitless, partly because there are different criteria for determining what is important and partly because archaeobotany moved on to address other questions in addition to that of numerical importance (Dennell 1976). To this end, the second approach sees quantification as a means of numerical description, which permits interpretation in a variety of ways, which may or may not relate to ‘importance’ (Jones 1991). Quantification remains an important aspect of archaeobotany as it forms the basis of data presentation and analysis, ranging from the tabulation of ‘raw’ data to the use of multivariate statistics

What to count?

The quantification of archaeobotanical remains may be semi-quantitative or fully quantitative, and it may be conducted at different levels of scale. The simplest form of quantification is probably to record the presence or absence of different taxa at a site (or even regional) level. Most archaeobotanists would go further than this, and the archaeobotanical ‘sample’ is a common level for quantification. The choice of whether to adopt semi-quantitative or fully quantitative methods is dependent on the type of material, the questions to be addressed, the level of reporting required and the resources available.

Simple presence or absence in a sample is rarely sufficient: first, presence or absence is very dependent on sample size and, secondly, is very affected by small amounts of sample contamination or mixing. Semi-quantitative ‘scales of abundance’ using a three or a four point scale (Hall and Kenward 1990, 299) can be a very effective way of quickly assessing the basic composition of samples and it can be based on a rapid ‘scanning’ of the samples. It is also very suitable for plant material that is difficult to count, e.g. awn or pod fragments, tubers and for waterlogged remains of leaf, stem etc. (see the identification section for waterlogged tissues). Finally, it is often used as a means of assessing samples for potential fully quantitative study as, for example, for an initial assessment report.

Fully quantitative methods, on the other hand. involve counting individual specimens, and decisions must be made on what exactly to count. If all fragments are counted as found, this can lead to misleading results through ‘double counting’ of the same seed or other plant part. A way of avoiding this is to choose a specific feature of the seed or plant part to count. The choice of which feature to count should be based on three criteria: (1) it should be a feature that survives well archaeologically (to avoid under-representation), (2) it should be easily identifiable (to ensure reliable identification), and (3) it should be unique (to avoid double counting). For cereals, suitable unique features are the embryo end of the grain (not the embryo itself as these are sometimes found detached and are not then easily identifiable), the base of the glume (for glume wheats, where this is the most durable part of the chaff), the rachis node (for free threshing cereals, where the rachis is more durable than the glume bases), and the culm node (the most durable part of the straw). These features also have the advantage that the ratio of grains to glumes and rachis nodes in the whole plant is known (at least approximately) and so the ratios obtained for the archaeological sample can be compared to that expected in the intact plant, thus permitting the identification of ‘chaff-rich’ or ‘grain rich’ samples. Similar countable features can be identified for pulse seeds and other plant types. This method is akin to the ‘diagnostic zones’ method for archaeozoological quantification, proposed by Watson (1979).

References

• Dennell, R. 1976. The economic importance of plant resources represented on archaeological sites. Journal of Archaeological Science 3: 229-247.
• Hall, A. and Kenward, H. 1990. Environmental evidence from the Colonia. The Archaeology of York 14 pp.289-434 London: Council for British Archaeology.
• Hubbard, R, 1975. Assessing the botanical component of human palaeoecnomies. Bulletin of the Institute of Archaeology 12: 197-205.
• Jones, G. 1991. Numerical analysis in archaeobotany, pp. 63-80 in W. van Zeist, K. Wasylikowa and K.-E. Behre (eds.) Progress in Old World Palaeoethnobotany. Rotterdam: A.A. Balkema.
• Watson, J.P.N. 1979. The estimation of the relative frequencies of mammalian species: Khirokitia 1972. Journal of Archaeological Science 6: 127-137.