For a long time, the Manis Mastodon site near Sequim, Washington was the elephant in the room of the Northwest Coast early period. The apparent bone point embedded in a mastodon rib was seemingly hard to explain by any non-cultural means, yet maddeningly short of definitive proof, and so was politely ignored. The point has always been a thorn in my side too, which is why I have posted on it three times, once over a year ago, and twice recently.
Maybe I am a bit obsessed with it because if I rise gently from my sofa in Blog World Headquarters, being careful not to spill fine single malt on my pyjamas, then through my window I can see Sequim in the extreme distance, seemingly mocking me.
So all the more cathartic that today, with the publication of a convincing re-analysis of the mastodon rib by Michael Waters et al. in the respected journal Science, we can say that the site is, indeed, evidence of humans hunting Mastodon on the Northwest Coast 13,800 years ago. That’s about eight hundred years pre-Clovis. Like I said before: it’s real. It’s old. It’s on the coast. Wow.
In common with many papers in this journal, it is very short: not much more than a single page. There is quite a lot of additional meat in the online supplementary information, which I have seen. Even so, I haven’t had time to digest the whole article, and my day job is being very demanding right now, so in addition to the general points I raised two weeks ago, here are some more points of interest.
1. The ancient DNA and protein analyses did not work well. Nonetheless they are able to establish (more or less) that the point itself is made of mastodon bone. From the x-ray above, there is no question it is penetrative and not some kind of pathology. The identification as “bone” (as opposed to tusk ivory) is referenced to “High-resolution x-ray computed tomography (CT) scanning (15) revealed that the osseous object embedded in the rib is dense bone shaped to a point (Fig. 1 and movies S1 and S2).” Identification as mastodon obviously rules out antler and the rutting elk theories. It would be nice to have tusk ivory completely and irrefutably ruled out by examination of cellular structure of the bone point, say.
2. The sort of eruptive effect seen on the external pictures of the rib is unexplained – perhaps displaced rib cortical bone? Close examination of the radiograph at the top of the post shows part of the rib surface bent down along the lower margin of the projectile, and perhaps part was forced upwards as well. The rim of the eruptive zone does seem very clean where it meets the rib though.
3. The projectile itself is very narrow. By the scale above, it is only about 4 mm in diameter. The author’s note that to even contact the rib, the projectile would have had to penetrate 25 to 30 cm of muscle, and connective tissue. That’s a foot of meat! And it doesn’t include the thick skin and hair. And yet, it still had enough grunt to penetrate the rib by over 2 cm. I’d be more skeptical of this if I hadn’t recently seen Nick’s experiments of thrusting antler points through about 40 cm of ballistics gel, occasionally penetrating into an underlying piece of lumber. One feature of those experiments to my eye was how tamely the shaft followed the point into the gel – and at Manis, presumably the bone projectile did not measure 32 cm X 0.4 cm: it must have thickened up, then met a shaft or foreshaft. All the same, for a long, straight bone point then dense mastodon long bone would presumably be just the ticket. While mentioned in the article in passing, it would be quite interesting to see comparison to the Clovis bone rods, one of the largest collections of which is also from Washington State. These early osseus artifacts are usually interpreted as hafting aids, though we hashed out their potential as bone points in this space a while back – see this quick picture from Lyman et al 1998.
In any case, comparison of the new image to the previous one shows the older one was actually quite misleading. The new CT scans include some 3-d images (and apparently an online rotational animation of the CT scan in three dimensions, but I haven’t seen those yet EDIT: see here). It takes a bit of effort to decode the one in the article (below) but I think it becomes clear that the bone point itself has a markedly narrow length:diameter ratio and is quite straight. Narrow is good if the goal is to pass between ribs, and/or to penetrate very deeply to reach vital organs. The pointy end is not that clear, but to my eye seems to be chisel shaped or beveled, though it might well be crushed or warped by the aforementioned linear foot of mastodon meat it just penetrated. In any case, I think it extremely unlikely that this is a random splinter of bone, it is far more likely to be the end of a projectile point. Waters is quoted in this Guardian article:
Waters placed the mastodon in an industrial-grade CT scanner at the University of Texas. “It’s more powerful than a hospital one. They’re taking slices every 0.06mm, half the thickness of a piece of paper,” he says. “The 3D rendering clearly showed that the object was sharpened to a tip. It was clearly the end of a bone projectile point.”
4. One surprise is that they assert the animal did not, as previously thought, survive for months after this wound: there is no sign of wound healing. This actually makes a tidier story of a single event “kill-butchery site”, for the authors also accept without comment at least some of the previous studies (e.g., D. Gilbow thesis) of the bones showing impact fractures, spiral fractures, cut marks, and a systematic pattern of disarticulation. Hunting in a wetland is, of course, quite characteristic of early American adaptations, and is also the pattern at the nearby, contemporaneous, Ayer Pond. Daniel Fisherhas been able to do remarkably fine-grained seasonality studies on mastodon tusks – which may have an almost-daily encoding of growth. It seems like a missed opportunity not to attempt this at Manis, thus establishing the season of death. It would also have been most welcome for the supplemental materials to illustrate (with photographs) the bone characteristics said to result from butchery. The compelling illustrative material from the Ayer Pond report sets a good example in this regard, allowing readers to clearly see the specimens in question and follow the argument closely. It’d be great to see the Manis skeleton as a whole receive that treatment.
5. It does raise the question of whether there were any trees or bluffs around to fire the projectile from, which entered the animals middle back from above at about a 60 degree angle. The authors suggest that perhaps it was a lance used in a thrusting motion, which I actually have some sympathy with – we made the suggestion at the SAA conference in Vancouver that some early NW Coast organic projectiles may have been killing lances for dispatching large fauna from boats, in caves, or other awkward places of inter-species rendezvous. Nick Waber is also including this general idea in his forthcoming thesis. An atl-atl dart launched at close range at a downed mastadon can’t be ruled out, however.
6. A suite of radiocarbon dates were run, showing definitively that the rib (found out of context) is exactly contemporaneous with the rest of the mastodon (found in situ). The table above shows the four new dates used, including the new one on the rib. A number of other dates were also run, all by Tom Stafford through the respected lab at Irvine (UCIAMS). One curious feature of the dates is that, judging by the lab numbers (e.g., 11350, 29113), which are issued sequentially at Irvine, and comparing them to lab numbers from projects I have worked on, then these new Manis radiocarbon dates were run in about 2005. This shows that the Manis reanalysis has been going on for quite some time.
7. The authors’ conclusions are apt, but strike a slightly sour note in some ways:
The absence of stone projectile points at Manis, Hebior, Schaefer, and Orcas Island and the presence of an osseous projectile point at Manis suggest that osseous projectile points may have been the pre- dominant hunting weapon during the pre-Clovis period. Bone and ivory points and other tools are common in the Upper Paleolithic of Siberia and in late Pleistocene sites in Beringia (22–24). They are durable and lethal hunting weapons that continued to be used during and after Clovis (16, 23, 25). The invention and spread of a new hunting weapon at 13 ka—the Clovis lithic point— may have accelerated the demise of or doomed the last megafaunal species.
As Dale Croes and David Rice commented in my earlier post, perhaps we really are dealing with a widespread pre-Clovis bone point tradition out here on the NW Coast, and elsewhere (though what about those pre-Clovis “Miller Points” and El Jobo points, etc.?). In particular, they note Alex Krieger’s “pre-projectile point hypothesis”, which Waters et al. do not mention. Speaking of authorship, it is most welcome to see Carl Gustafson as a co-author, considering his close connection to the site over the last 25 years. I’ve moaned about the Manis uncertainty at length but, you know, without Gustafson we probably wouldn’t have anything to talk about at all.
But as I was saying, the rather cavalier assertion that the invention of Clovis projectile points somehow hastened the extinction of Megafauna seems once again to place undue privilege, or undue burden, on a sort of “Clovis Exceptionalism” which I think is unhelpful in early period archaeology. In the Science news report (which is behind a paywall) Gary Haynes perpetuates the centrality of Clovis as an archaeological construct:
Haynes adds that the oldest Clovis sites are “only” 8 centuries younger than Manis. Thus the rib “may actually indicate the earliest beginning of the Clovis era, or an immediately proto-Clovis stage of human dispersal,” he says. “Proto-Clovis” peoples in small numbers may have filtered south from Beringia as early as 14,000 years ago, he says, although their impact was negligible until the arrival of Clovis technology. To Waters, such talk of “proto-Clovis” amounts to “grasping at straws.”
I mean, we don’t talk about Clovis as “pre-Folsom”, so I’m going to start a campaign to just outright banish the term “pre-Clovis” except strictly as a temporal marker – those folks were not trying to become Clovis, Clovis was an unknown cultural expression eight centuries in their future.
8. Speaking of Clovis Exceptionalism, I was very interested to see that one of the non-anonymous reviewers thanked in the article is Stuart Fiedel, who has been one of the most outspoken critics of any pre-Clovis occupation in the Americas. Either he is in agreement with this paper, or his critique of the paper was insufficient to modify the authors’ strong conclusions, or to prevent publication in this top journal. The paper thus seems almost “immunized” against certain genres of criticism.
Though really, with the equally convincing paper from earlier this year, also by Michael Waters et al., on the pre-Clovis Buttermilk Creek Complex (Friedkin Site) in Texas, it would be harder than ever to drive pre-Clovis advocates en masse over a cliff. And the fundamental insecurity at the heart of the “Clovis First” arguments was always that it only took one legitimate pre-Clovis site to pull aside that curtain. I think it’s fair to say that we now have a handful of solid pre-Clovis sites, including at least four (Paisley Cave, Monte Verde, Ayer Pond and Manis Mastodon) on the western margins of the Americas, of which two are only a few tens of kilometers apart right here in the Salish Sea. It’s a good time to be an archaeologist on the Northwest Coast.
Michael R. Waters, Thomas W. Stafford Jr., H. Gregory McDonald, Carl Gustafson, Morten Rasmussen, Enrico Cappellini, Jesper V. Olsen, Damian Szklarczyk, Lars Juhl Jensen, M. Thomas P. Gilbert, Eske Willerslev 2011.
Pre-Clovis Mastodon Hunting 13,800 Years Ago at the Manis Site, Washington. Science 334 (21):351-353.