Exploring The Significance Of Different Plants In Neolithic Settlements

In his 1978 report “Archaeobotany and Early Farming in Europe”, Robin Dennell excavates and analyzes two Neolithic settlements, Chevdar and Kazanluk, in Bulgaria. As an archaeobotanist, he looked specifically at plant remains, hoping to “discover which plants were intentionally cultivated, which were the most important, and how plant foods were prepared for consumption and storage”. His first step in exploring the significance of different plants was to determine which plants were processed there and how; processed plants hold obvious importance as food crops. Analyzing three sample groups collected from among the two sites, he was eventually able to distinguish which grains seemed to be processed and the various stages of grain processing. In order to do so, Dennell examined three aspects of each sample: its context, its composition, and the size of the grain. Using these, he deduced the uniformitarian processes they likely resulted from. In the first sample, composition and context played equally important roles in Dennell’s analysis. Composed almost entirely of valued food crops such as emmer, barley, and legumes, weed seeds were practically nonexistent in this batch and spikelet and straw fragments were absent completely. Dennell greatly stressed the importance of the context of a sample because “threshing, winnowing, sifting, and other harvesting techniques will tend to remove some grains and seeds from a cereal crop and thus produce a biased sample”.

Composition itself could be misleading and insufficient evidence without knowing more about the environment it came from. Seeing as how these grains were found principally in ovens in Chevdar, it is likely that they had been “carbonized either while being dried before storage or being cooked”. Here, context played a major role; if not for the ovens, there would be little hints as to how these grains were burnt. The relation to the ovens directly links the production of this sample to human processing. The relative purity of this sample and the location in which it was found makes it highly likely that this sample represented grains that were nearly finished processing. With the second sample, Dennell demonstrated why the size of the grain can be an important factor in determining what processing activity occurred. This sample was found on floors in both Chevdar and Kazanluk. The sample also lacked any spikelet or straw fragments, suggesting that by this stage of processing, this grain had already been threshed and winnowed. However, there was also much more weed seeds present than in the first sample and the cereal grains that were present were rather small. While the extra weed seeds might be seen as the leftovers of grain cleaning, Dennell suspected that the weed seeds and smaller grains were left behind as the larger grains were sieved out in the processing activity. Here, Dennell put his hypothesis to the test by sieving a batch of wheat himself, producing a test sample that was “remarkably similar in composition to the samples from the floor deposits at Chevdar and Kazanluk”.

To Dennell, these results seemed to fit quite well with his interpretation of the first sample. While the first sample represented grains that were cooked or dried at the end of processing, the second sample was the undesirable parts of the crop that were removed through sieving. In examining the third sample, Dennell had to go beyond just the context, composition, and grain size and apply his own knowledge of grain processing in order to build his interpretation. This sample consisted mainly of straw and spikelet fragments as well as many weed seeds and a few cereal grains of all sizes. Found inside refuse dumps in Kazanluk, this sample was most notable in the fact that all plant material within was burnt while the pits were not. This suggests that these were burnt beforehand and trashed after.

The presence of large cereal grains among the other residue as well as the seeming removal of the majority of cereal grain makes it unlikely that this sample was just the result of winnowing (there would be only seed and grain) or threshing (there would still be all the grain along with the residue). Instead, Dennell deduced that “the inhabitants may have combined threshing and winnowing into a single process, as was done until recently in parts of Scotland”, again utilizing uniformitarian assumptions to predict the cause of this sample’s condition. Dennell saw this sample then as the first stage of processing, where the crop was singed and dehusked, removing the spikelet and straw fragments before being sieved to clean and then roasted in ovens to be stored.