16th August 2013, 06:19 PM
As Seedygirl can attest, I'm afraid my site sampling strategy usually gets modified continuously while I'm digging a site, given the fore-knowledge that budgets only stretch so far I usually try to get maximum bang for the bucks. One usually has a fair idea what's going to be dateable (either from finds, stratigraphy or if it's going to end up on a potential scientific dating shortlist), so I generally concentrate sampling on those features, have never seen the point of epic palaeoenvironmental analyses of undated deposits. And sealed, finds-rich/faunal-rich etc assemblages get high up the list - Seedygirl's recently been doing me the goods on a staggeringly good late C17th pit assemblage where sadly the seeds were slighly disappointing but the finds and small bones were amazing, and the fact that the seed reults were in contrast poor was in itself of interest - getting a wide dataset from a small number of 'star' contexts has in my experience been far more archaeologically valuable than the usual palaeo report where there's a few charred seeds and a bit of chaff scattered across a load of poorly dated contexts. And some contexts are just rubbish sampling targets, such as ditches filled with secondary materials from who-knows-where. Always understand your site formation processes before even thinking about fetching sample tubs.
Definitely helps if the site's being dug by a PO who'll be writing the final report and has a good concept as the excavation develops as to where it'll be going and what data can feed into it - that's the best sampling strategy, and as Seedygirl pointed out the same samples can have multiple purposes (palaeoenv/finds/small bone/industrial etc), just needs someone to sit down and think things through and maximise the potential return on effort. And results from one context can affect what else to sample - if you've got a site with similar features over a significant time-depth, and e.g. a late Iron Age pit has lots of grain, try and sample a run of similar pits throughout the history of the site to the early Norman, on a small budget a coherent group of comparable data (in this case grain in pit fills through time) is going to be far more valuable than a random series of pit, ditch and posthole fills which tells no one anything useful
Definitely helps if the site's being dug by a PO who'll be writing the final report and has a good concept as the excavation develops as to where it'll be going and what data can feed into it - that's the best sampling strategy, and as Seedygirl pointed out the same samples can have multiple purposes (palaeoenv/finds/small bone/industrial etc), just needs someone to sit down and think things through and maximise the potential return on effort. And results from one context can affect what else to sample - if you've got a site with similar features over a significant time-depth, and e.g. a late Iron Age pit has lots of grain, try and sample a run of similar pits throughout the history of the site to the early Norman, on a small budget a coherent group of comparable data (in this case grain in pit fills through time) is going to be far more valuable than a random series of pit, ditch and posthole fills which tells no one anything useful