Dear William,

Intriguing question.  At face value, one answer is, Why not?  But presumably you are thinking of sheetwash, etc., as natural processes that selectively remove usually smaller-fraction (lighter) flakes and therefore bias surface assemblages.

Difficult to provide much detail on this site, but I'm right now doing work along these lines.  Clearly, my surface assemblages are impoverished in finer fractions, probably the result of taphonomic winnowing compared, for instance, to experimental assemblages that are similar in toolstone, technology, cobble size, etc.

The key thing is to have comparative material (of similar toolstone, etc.) from excavated context.  My data are from obsidian quarries that I didn't excavate, so surface-only.  A somewhat similar quarry some distance away has a similar size-distribution in its surface assemblage but a very different one--much higher small-fraction proportion--from excavation.  So too a nonquarry site to which material from my quarry probably was exported.

I used size-distributions from excavation at those sites to estimate the magnitude (30% or more) and pattern (definitely inverse to flake size) of winnowing.  Then I multiplied my surface proportions by scaling factors (higher at smaller sizes) to estimate--it's an estimate, not a certainty--original size distribution.

Can provide more detail if you're interested.  ALso, would be interested to know if you're doing similar work with flake-size distributions.

MS

I would add that sampling bias has a great impact on the character of the assemblage. How many times was the site collected? Were the collections systematic? What was the interval between collection transects? What were the site conditions like? There have been multiple studies showing that rainfall between collection events, wind, surveyor experience all have a bearing on the nature of surface collected assemblages.  Larry Todd and some of his students did some interesting work a few years back using nested sampling strategies or the "modified Whittaker" approach to examine size bias within surface collections.

Burger, Oskar, Lawrence C. Todd, Paul Burnett, Tomas J. Stohlgren, and Doug Stephens

2002–2004 Multi-ScaleandNested-IntensitySamplingTechniquesfor Archaeological Survey. Journal of Field Archaeology 29(3 & 4):409–423.

Hi William

Big "depends".

If the site has recently been uncovered and the depoists relatively undisturbed, then the data could prove interesting. Otherwise it would be 'noise'.

Ignácio de la Torre has an interesting paper on the natural accumulation of manuports - although he did not make a basic comparison of very close areas to discover the extent of the background 'noise'; which was a pity as this could have been used to determine how much of the manuport data he found was 'ecoport' and how much probable manuport.

I would hope that a site with a cache of stones that are not directly linked to surface rock nor any upstream source, that appears to be recently uncovered or still partially covered and that has a marked concentration of stones in just a few m2 when compared to the surrounding grid, would imply manual selection.

At least that was my goal when drafting "The Mass Distribution of Manuports..." as this is one of the only ways of looking at hominin behaviour pre tool manufacture.

Best

Alan

Aggregate analysis does not consider tools or manuports directly, but rather focuses on debitage. The surveys I am concerned with are the typical surveys required of CRM projects in the USA such as surface collections along 2 m transects in plowed fields following a rain, not "crawl" surveys or gridded controlled blocks. Thus, my concern is with the sample diversity recovered from such surveys. Smaller flakes that may be missed when people are walking and that larger flakes are likely to be overrepresented in the sample collected.   I realize they likely will be skewed towards core reduction since aggregate analyses look for general trends from the assemblage analyzed. Thus,  can an aggregate analysis still be applied in a meaningful way.

I would like to add to this the problem of "skimming the best flakes off the top"---this can happpen at ANY point in time---and whether a site is buried or on the surface. LOL