Friday, 9 November 2012

Hominid Hunting, Pinnacle Point, and Me. An epilogue.

Hominid Hunting has just covered the latest news from Pinnacle Point, the subject of yesterday's blurt. I couldn't let it go without comment. Hell, I think by now the author, Erin Wayman, probably expects it!
     It's 'awaiting moderation' as I write this. However, in case it's considered too technical or too contrary for publication, I'm gonna quote myself here, verbatim. Here's how it goes.
Hi, Erin. Pinnacle Point is quite an amazing archaeological locality. And the excavators, Brown and Marean, are doing excellent work. I know that it'll sound a bit like sour grapes, however, for years now there have been those who question not the finds, but the age estimates of these and other southern African sites. Like those at Pinnacle Point, the spectacular finds derive from caves, where organic preservation is often enhanced either by protection from the elements or the chemistry of the cave environment, or both. Unfortunately, for those parts of the site that are beyond the radiocarbon limit—around 40,000 years—the excavators have relied on less precise dating techniques. One in particular, optically stimulated luminescence (OSL), begins with the assumption that a given grain of quartz sand was exposed to sunlight for a time before final burial in the cave. Where caves are concerned this assumption is an untenable assumption, because no one is able to say ahead of time, with any certainty, that a single grain was or wasn't exposed to sunlight sufficiently powerful or for long enough. As a result, once the technique has yielded its raw results, there follows a complex mathematical dance based on all sorts of other assumptions, as a means of 'eliminating' the uncertainty of whether or not a given quartz grain had been sufficiently exposed to sunlight. Moreover, if the mathematical assumptions and the inherent complexity of the calculations are going to be in error they'll always overestimate the time since that quartz grain (and the artifacts in proximity that are of interest) was last exposed to sunlight. Logically, there's no under-estimating ages in caves using OSL. This leaves us with a question, one that I think the excavators of Pinnacle Point and elsewhere cannot logically escape. If the dates of unequivocally modern human behavior elsewhere in the world are less than about 45,000 years old, what are the odds that the anomalously early dates for similar behaviour in those southern African sites—all dated by OSL—are systematically overestimating their age? I'd say those odds were far better than the likelihood that all of the finagling involved in arriving at an OSL age estimate is yielding accurate and precise chronometric results.
[Update 0045 UTC November 10, 2012: Doh! I knew I was forgetting something! The bedrock in which the Pinnacle Point caves are developed is described as quartzitic. Wanna know how long it's been since the quartz grains in the bedrock last saw sunlight? Your guess is as good as mine. Suffice it to say that it's unlikely that they had been exposed to very much sunlight once they fell to the floor. All that matters is that the uncertainty introduced by one minimally exposed grain would introduce one hell of a wrench in any age estimate, even one that's been through the math mill.]  
I think that about sums it up. Now, why couldn't I have said that yesterday?


Thanks for dropping by! If you like what you see, follow me on Google Friend Connect or Twitter, friend me on Facebook, check out my publications at Academia, or connect on Linkedin. You can also subscribe to receive new posts by email or RSS [scroll to the top and look on the left]. And please don't forget. Oh, and you can always put me on your blogroll! By the way, I get a small commission for anything you purchase from if you go there by clicking through from this site. 


  1. Good epilogue. esp re: "Unfortunately, for those parts of the site that are beyond the radiocarbon limit—around 40,000 years—the excavators have relied on less precise dating techniques. One in particular, optically stimulated luminescence (OSL), begins with the assumption that a given grain of quartz sand was exposed to sunlight for a time before final burial in the cave."

  2. If you look at the actual OSL sample dates shown on the section diagram of the rock shelter, I find it odd that so much credence is given to the OSL dates, when what they show is that there are grains dating to the 60-69kya range in every single member of the shelter, and in any one member there's an age range between samples of 10ky. Either layers are mixed, or sample sizes are too small (n=3 in one case) to be calculating an average age with any confidence, or the dates themselves aren't telling you as much as you claim they are. Or is it just me.....?

  3. Sorry, clarification : by "you claim they are" I did not mean you, as in Rob. I meant the research team. Just in case there was misinterpretation!

  4. Hi, Sue. I'm glad you, too, noticed that all of the plots pointed to roughly the same age estimate. It's for us little people to accept the mathematical flimflammery of the OSL wizards and never to question their prescience, even when our intuition tells us something is very wrong.

  5. Is the photo of the strat in yesterday's blog from the part of the site in question? Assuming so, and not having read the original articles, it LOOKS from a cursory glance at the photo taken from many metres away that the strat is relatively homogenious in terms of repeating sequences and no obvious massive roof falls; can we assume relative uniformity in the deposition rate? Little quartz sand grains continuously raining down? I'd like to see that photo (or a profile line drawing)labelled with the locations and results of radiocarbon dates that are less than 40K, and the locations and dates of the OSL (which of course are problematic if they are so similar!). BUt if, say the top 2 m has gradually increasing dates to the limit of C-14 and young OSL, and we can assume some continuity of rate for the next 4 m and we get OSL estimates that are an appropriate multiple, that would give some support to the old dates. It sounds as if this either hasn't been done, or the results are not consistent with such a scenario (e.g., what also wouldn't support an old date would be if the C-14 dates were increasingly old to 40K but spread over say 5 m and there was only a metre and a half below that where all the old OSL dates were coming from....)

  6. Oops, see that Sue makes reference to a section diagram....


Thanks for visiting!

Note: only a member of this blog may post a comment.