27th May 2008, 01:15 PM
Hmm
We had a similar layer on a site recently. We stuck a few narrow slots through it, found features underneath, then excavated more of the layer in a grid of small squares around the points through which the features were hypothesised to have cut. The distribution of artefacts confirmed that the features did indeed cut through the layer, which was a very worthwhile result as it discounted various other interpretations for the layer (colluvium, flooding event, etc.). However, we still couldn't find the features' edges until physically below the layer (and nor could several emminent site visitors). I could have found the edges with a lot of soil micromorphology samples, but I don't think the client would have been keen. There's a steep curve of diminishing returns here, and we have sufficient unquestionably-stratified artefact assemblages from the lower parts of the features for reliable dating.
Essentially, we applied the first law of archaeological resource allocation: 'Yes, but what does it tell us?'
It's worth noting that this kind of stratigraphy defies the rigid lines inherent in British archaeological recording. For example, an horticultural soil may develop over many years while pits are continuously excavated. One might invent an arbitrary palaeo-land-surface for each identified cut, but that would still be a fudge as pedogenic processes will have destroyed those surfaces, even if they could be defined as events (which all British 'contexts' are). In the end, you have to find peace with cut lines on your sections that just disappear into nothingness without reaching the surface (as geologists are happy to do).
We had a similar layer on a site recently. We stuck a few narrow slots through it, found features underneath, then excavated more of the layer in a grid of small squares around the points through which the features were hypothesised to have cut. The distribution of artefacts confirmed that the features did indeed cut through the layer, which was a very worthwhile result as it discounted various other interpretations for the layer (colluvium, flooding event, etc.). However, we still couldn't find the features' edges until physically below the layer (and nor could several emminent site visitors). I could have found the edges with a lot of soil micromorphology samples, but I don't think the client would have been keen. There's a steep curve of diminishing returns here, and we have sufficient unquestionably-stratified artefact assemblages from the lower parts of the features for reliable dating.
Essentially, we applied the first law of archaeological resource allocation: 'Yes, but what does it tell us?'
It's worth noting that this kind of stratigraphy defies the rigid lines inherent in British archaeological recording. For example, an horticultural soil may develop over many years while pits are continuously excavated. One might invent an arbitrary palaeo-land-surface for each identified cut, but that would still be a fudge as pedogenic processes will have destroyed those surfaces, even if they could be defined as events (which all British 'contexts' are). In the end, you have to find peace with cut lines on your sections that just disappear into nothingness without reaching the surface (as geologists are happy to do).