Prometheus Unfounded: Contradictions and Conundrums at Wonderwerk Cave

Wonderwerk Cave profile (Photo by M. Chazan)

I’ve previously opined on aspects of the claim for very early fire use at Wonderwerk Cave, South Africa here, here, here, and here. In brief, it’s Swiss cheese. Today I’m beginning to look at Peter Beaumont’s 2011 synthesis of the ‘evidence’ for hearths in the cave, published in Current Anthropology. Forgive me if this comes across as unusually pedantic–I find the author’s descriptions to be less-than rigorously scientific, and thus less-than helpful if one’s hoping to cast a critical eye on what amounts to his life’s work. I’ve found it all very difficult to wrap my brain around. See what you think.   
     Beaumont begins his discussion of what he calls ‘hearths’ by mentioning that in places he observed ‘poorly defined ash lenses’ and in other places ‘ash-rich’ deposits, which he thought had been ‘single hearths … largely destroyed (perhaps by trampling).’ It’s unfortunate that there could be so much ambiguity entailed in such a short paragraph.
     First of all, what really is the difference between a ‘poorly defined lens’ and an ‘ash-rich deposit’? Aren’t they both ‘ash rich’ if you can recognize the ash in profile? And what about the other sequelae of burning, charcoal and the reddened substrate. I would have expected any fire that could turn plant fuel to ash would have been sufficiently hot and of such a duration that it would also have reddened the sediments beneath the fire. Moreover, in many cases where reddened sediments and ash are visible in a stratigraphic sequence there is a higher than average chance that there will be a layer of charcoal-enriched or blackened sediment between the reddened substrate and the ash. Trampled or not, Beaumont’s inference that these ash lenses were hearths is hardly to be believed on the face of it.
     Presumably Beaumont wants us to believe that, in the case of the ‘ash-rich’ places, any vertical distinction that at one time would have been evident between the ash and the reddened sediments had been obliterated by treadage. Yet, if that were the case, how is it that he’s able to discern anything that might be called discrete (albeit poorly defined) ash lenses? For Beaumont to be able to observe ‘lenses’ comprising ash, those lenses must have escaped, in large part, the destructive results of trampling. And surely, if the ash ‘lenses’ had escaped the ravages of time and trampling such that they were visible in profile, the underlying reddened sediments would also have retained enough integrity to be visible, too! Yet, the author mentions nothing about the substrate, reddened or otherwise. Odd. On the other hand, one has to agree that in all likelihood it was trampling that transformed what had once been intact ash deposits elsewhere in the cave into something the author calls (merely) ‘ash-rich.’

[My recognizing problems with Beaumont’s after-the-fact verbal descriptions doesn’t ensure that the his inferences are incorrect. However, one does have to wonder. Doesn’t one? One would have thought that a perspicacious referee or editor would have noticed these vague and incongruous descriptions. Wouldn’t one?] 

     Alas, the abovementioned ‘hearths’ aren’t the only curiosities to be found in Beaumont’s treatment of putative fire use at 1+ Ma. In another example he describes stratum MU2, in excavation 5, where as much as 45 cm (!) of the stratigraphic column ‘is very largely composed of white ash with many burned bones and fire-damaged Middle Stone Age artifacts.’ This ash apparently ‘accumulated slowly’ and continuously between about 1,155,000 and about 70,000 years ago. [Get out your calculators!] 
     Depending on what Beaumont means by ‘very largely composed of’ [and it’s not at all clear], it sounds as if he’s suggesting that for 1,085,000 years a very wide area of the cave received little other sedimentary input than that of completely combusted plant material. That’s a prodigiously long time for a single depositional process to have endured, and a phenomenally long time for a large surface comprising a substance as mobile as ash to have survived without either blowing away or being adulterated by larger [especially inorganic], autochthonous clastic input. 
     Even more mysterious: Beaumont claims that the ash was the result of thousands of fires fueled by above 15 tons of fuel. It’s really hard for me to imagine that such a focus of hominid activity could have escaped the inevitable, and destructive, treadage that would have accompanied that use of that part of the cave for what amounts to a single activity–that of making and keeping fire–over such a vast expanse of time. I suppose it’s not impossible. But, probable? I really don’t think so. Plausible? Barely.
     Beaumont’s description of MU 2 in excavation 5 just makes no sense. If his account of its clastic composition is accurate, nothing but a long-lived colony of fire-loving faeries could have produced it! There must be alternative explanations. And, indeed there are, provided by an unlikely source–the latter-day excavators of Wonderwerk Cave, the very ones who have recently reprised Beaumont’s long-standing claim of fire use by Acheulean hominids.
     Here’s what Matmon, Chazan, Porat and Horwitz conclude in ‘Reconstructing the history of sediment deposition in caves: A case study from Wonderwerk Cave, South Africa‘ published in the Geological Society of America Bulletin (First published online October 14, 2011, doi: 10.1130/​B30410.1).

The cave sediments comprise a sequence of fine sands and silts that were transported naturally by wind to the environs of the cave and later into the cave by water. Transport within the cave occurred by low-energy water sheetflow, which distributed and deposited the sediment in its final location. Field observations and grain-size distribution analysis of the sediments inside and outside of the cave imply the following sediment transport scenario: eolian transport of Kalahari sand to the Kuruman Hills, slope wash of the eolian sediment into the intermontane valleys, fluvial transport of the sediment from the intermontane valleys to the entrance of the cave, and final deposition of the sediment inside the cave by low-energy water action.

 These are the verbatim conclusions. Unfortunately for the authors, this sequence of transport processes leading to the input of allochthonous sands and silts at Wonderwerk Cave can equally explain the presence of anything that would be as easily transported as sands and silts. Indeed, anything lighter than fine sand–if it shows up at the doorstep–would have been subsequently transported into the cave by sheetwash. Sheesh! They’ve done my work for me! 
     Their conclusions also solve a riddle that I found while wandering through the images from Wonderwerk. This one shows non-conformable and curiously wavy strata. I’ve drawn yellow rectangles where I see evidence of what appear to be erosional events overlain by non-conformable strata. These observations support the conclusions of Matmon et al. It appears that there have been numerous erosional episodes during the build-up of sediments in Wonderwerk Cave. The wavy contacts suggest an agent even more energetic than sheetwash. If these observations are borne out it’s clear that at times the input of material and liquid from outside the cave was considerable.     

After Berna et al. 2011

Make of it what you will. 
     Beaumont describes ‘grass mats’ in various stages of combustion that occur here and there in the cave. Regardless of how they arrived at the cave’s doorstep, in they went–wind-whipped dry grass, partly combusted grass, grass ash [try saying that five times really fast without saying something unfit for polite company], small bits of burned bone made as light as fine sand by partial combustion. You name it! Matmon et al.’s conclusion opens the door to serious questioning of Wonderwerk’s depositional history. How can they claim, unequivocally, that any wind-transportable allochthonous sediments came to rest in the cave by the hands of hominids?
     Seriously! They are way past the bounds of logical inference when they claim that any of the tiny particles that Berna et al. describe in exquisite micromorphological detail were left there as a result of hominid behaviour. Somebody’s gotta tell them. I’m trying me best. But they appear not to be listening.
     So, get out there to the meetings and to your classrooms and call out the litany of overwrought inferences of hominid behaviour that keep emanating from Wonderwerk Cave!     

One thing’s for sure: the excavations at Wonderwerk Cave are looking more and more like job security for this Subversive Archaeologist.

Please show your support for this blog. Remember that when you purchase from Amazon by clicking any of the links on the right, you’ll be getting great discounts and supporting the Subversive Archaeologist’s field activities at the same time!

Cultural Deposits in Homogeneous Natural Sediments VS. Homogenized Cultural Deposits in Natural Sediments

     Thursday’s touchstone blurt has prompted three interesting comments–two of which from long-time acquaintances; one of which was left by Edward Harris, progenitor of the Harris Matrix and author of Principles of Archaeological Stratigraphy [which the author has made freely available online–just click on the title]. 

     All three expressed dismay that any archaeological sites are still excavated in arbitrary levels, and they all added a gentle prod to the effect that I should not be so quick to accept the practice’s continued use, even for sites in California. The problem really boils down to this. California archaeologists aren’t dealing with cultural deposits in homogeneous natural sediments. They’re dealing with completely homogenized (bioturbated) natural sediments and cultural traces. That’s quite a different story, as I’ll argue further down. But right now: nail buffing time.
     The Matrix man [or so he calls himself] sent me an email in addition to leaving a comment on Thursday’s blurt, which I copy here for posterity

Hello Rob

Thank you for your kind remarks and I made a comment on your interesting blog.

How can we advance the revolution (the Matrix is 40 next February!). Perhaps I should learn how to do a blog.

Anyway, I was speaking with colleagues from Playa Vista in California, who suggest they could not have done that dig without the matrix!

Here is the first one ever made and also a little article. You can post them if you like. The first-ever Matrix (read, as we now know, “stratigraphic sequence”) was made in 1974 for the site excavated that year at the south gate into the Roman town of Winchester, England.



Edward C. Harris, MBE, JP, PHD, FSA, Executive Director

Cell: (441) 704-5480; eMail:
National Museum of Bermuda
Incorporating Bermuda Maritime Museum
Street: 1 The Keep, Sandys MA 01, Bermuda
Mailing: P.O. Box MA 133, Sandys MA BX, Bermuda
Tel: 1 (441) 234-1333    
Fax: 1(441) 234-1735

 To see his article on time and the Matrix, click here

First-ever Harris Matrix–South Gate, Roman Winchester, 1974 (Photo courtesy of E. Harris).

In what follows I’ll try to explain (and not explain away) the conundrum faced by every archaeologist in almost any landform and at almost any altitude in California. Here goes!
     Regardless of the characteristics of the parent material on and in which pre-Columbian California archaeological occupation sites have formed, and regardless of the rate of aggradation throughout the period during which the cultural materials were abandoned, there is a Nemesis against which it is futile for an archaeologist to struggle: fossorial rodents–specifically the ground squirrel (Sciuridae) and the pocket gopher (Geomyidae). These animals are prodigious breeders and tireless burrowers. And because they have been present throughout the epoch of human occupation, they have had a universal and profound, time-transgressive effect on the vertical and horizontal distribution of archaeological materials in California sites. This has been documented in numerous ways and in sites across the Golden State. 
     These animals are capable of transporting any clast smaller than their burrow’s diameter anywhere from millmeters to (eventually) meters from it original location (if that was, in fact, its ‘original’ location). At any time they are regularly disturbing the top 50 cm or so. Given their habit of pushing their backdirt upward and out onto the surface, the smaller archaeological pieces ‘liberated’ in this fashion tend to remain in the upper 10 cm or so. And larger clasts are transported generally downward because over time they are undermined and drop a few cm at a time. This can happen to boulders and everything smaller that can’t be displaced along the burrow itself. 
     The habits and ubiquity of these rodents have ensured that no site in California, and certainly no site with a longish occupation sequence, can be expected to present observable stratification or, for that matter, a vertical column that contains the oldest traces at depth; the youngest near or at the surface. Moreover, the process size-selective upward and downward displacement I’ve just outlined means that in general there is a bi-modal vertical distribution of artifacts and natural clasts.
     Donald L. Johnson’s three drawings below illustrate what can and does occur as a matter of course in (quite literally) any unconsolidated dirt anywhere in California with the exception of [and I’m guessing here] the area above the tree line. It’s necessary to read the detailed figure captions to make sure you understand what’s going on in each series. However, I’m certain that those of you who aren’t from around here will be astonished and might eventually come to accept why it is no surprise that arbitrary levels are still the state of the art in California.

Johnson. Donald L. 1989. Subsurface Stone Lines, Stone Zones, Artifact-Manuport Layers, and Biomantles Produced by Bioturbation via Pocket Gophers (Thomomys Bottae), American Antiquity, 54,  370-389.

In most non-shell-midden sites in California the average stratigraphic profile records only the following:

a. the position of the large clasts that weren’t released during excavation.

b. the presently visible rodent krotovinas

c.  a humic layer at the top of the sequence that represents where the latest decomposed vegetation has resulted in illuviation

d. any pedogenetical horizonation that may still be in evidence below where burrowing has occurred, which is often the case in what are known as ‘prairie soils’ in relatively xeric landscapes, where illuviation of carbonates creates a relatively rodent-unfriendly consolidated layer at depth. 

That’s it!
     Alas! There’s no remedy for the effects of burrowing rodents on most archaeological sites in California. In fact, the effect I’ve sketched here is so widely distributed and such a virtual certainty that when attempting to ascertain the extent and time-depth of most pre-Columbian occupation sites savvy archaeologists know that one is most likely to find an example of every temporally sensitive cultural indicator smaller than the diameter of a rodent burrow in the top 10 cm of the site, regardless of the age of a site’s cultural sequence.
     Such is life in California archaeology. I’ll stand up for my California colleagues any day when there’s any question of the professionalism evinced by excavating in arbitrary levels. There’s no other choice. We who’ve worked in California know full well that there’s little point in digging a unit to ‘sterile’ in a site that’s been subject to rodent bioturbation. We know that, more than likely, all we’re recovering is bagfulls of out-of-context cultural artifacts.
     My new, but long-admired, colleague, Edward C. Harris, MBE mentioned that the people who worked at Playa Vista in southern California told him they couldn’t have done their work without using the Harris Matrix. I’d be willing to bet any money that they were referring to their excavations of the historic remains on that massive project, and not any pristine (nominally speaking) pre-Columbian occupation sites.
     This has been fun. Have a good week, everyone.