Clovis Comet Revisited”), July ■ 2011 3

Volume 26, Number 3 Center for the Study of the First Americans Department of Anthropology July, 2011 Texas A&M University, 4352 TAMU, College Station, TX 77843-4352 ISSN 8755-6898 World Wide Web site http://centerfirstamericans.org and http://anthropology.tamu.edu

The Nanodiamond Controversy, Part I THE 5 A site that has the experts stumped The C. W. Harris site near San Diego is Folsom age without CLOVIS COMET Folsom tools or animal remains, coeval with Desert Culture sites yet lacks grinding implements. Perhaps, says anthropologist Ed REVISITED Knell, it was a shaft factory. 9 When you think of Clovis lithic technology, think blades F DIAMONDS REALLY ARE FOREVER, was, perhaps, warmer than most; over the Blades were once dismissed as as the saying goes, then natural speci- previous few years the world had been occasional finds at Clovis sites. Imens should provide excellent records shaking off its latest Ice Age. In North Now we know they were the basic of geological history—assuming we can de- America, the surviving large animal popula- component of every toolkit. In part cipher the clues hidden in their structures. tions were poised to recover, and an upstart III of her series, lithics analyst Charlotte Pevny tells and shows us That’s precisely what one group of scien- species called humanity—which until re- how toolmakers reduced cores to tists claims to have done for some of cently had been thin on the multipurpose blades. Expert the tiniest diamonds known, a ground—had established a knapper Bill Dickens kibitzes. collection of microscopic Kangerlussuaq, solid foothold on the conti- particles dating from the Greenland nent at last. Things looked 16 Discovery of Bison antiquus end of the Pleistocene ep- promising . . . until some- has far-reaching consequences och called nanodiamonds. thing smashed into the A remarkably well-preserved If they’re correct, the world like the fist of God. skeleton found on Orcas Island in Arlington Canyon, Washington State bears the very existence of these Santa Rosa Island Entire populations were classic marks of butchering—and minuscule gemstones flash-fried as the fireball a radiocarbon date that means chronicles one of the most exploded high over the the butchers were pre-Clovis dramatic catastrophes ever continent; others died in the humans. At stake is a basic to befall our species: a killer ensuing firestorms, and many premise in the model of peopling of the Americas. comet that slammed into the of the rest perished in the deep Earth, altering the climate and destroying winter that followed. The Clovis culture an entire human culture. vanished with the blast, and with it most of the Extraterrestrial Impact Hypoth- the surviving species of North American esis (EIH) camp, which cites more Fire from the sky? megafauna. than a dozen lines of evidence suggest- It was a day like any other, except that it At least, that’s the theory advocated by ing that an enormous extraterrestrial 2 Volume 26 ■ Number 3

(ET) body struck North America at the have been unable to reproduce the EIH criticism, since there’s only one known end of the Pleistocene. Along with trillions camp’s results, even when working at the way that they can form naturally: in a of nanodiamonds, they’ve documented same sites. One by one, critics have dis- high-temperature, high-pressure envi- peaks of aciniform soot, charcoal, mag- missed most of the evidence supporting ronment such as that of an ET impact. netic microspherules, iridium, and a num- the EIH. Some fault the original Impact But even nanodiamonds have recently ber of other purported impact markers at Team’s conclusions, charging that they come under increased scrutiny, with the Younger Dryas Boundary (YDB), a deliberately stretched the evidence to fit a critics becoming progressively more continent-wide stratigraphic layer that preexisting hypothesis; others condemn vocal in their claims that at least some dates to about 12,900 CALYBP. Unsur- their collecting and analysis techniques, of the particles cited by the EIH sup- prising, not everyone agrees with the or alleged that they’ve simply misinter- porters as nanodiamonds are not, in EIH interpretation; while something preted their findings. fact, diamonds at all. dramatic certainly triggered the 1,000- Nanodiamonds have largely resisted In this two-part story, we’ll take a look year climatic reversal known as the Younger Dryas, they consider mecha- nisms other than an ET impact more likely culprits.

The Mammoth Trumpet (ISSN 8755-6898) is published quarterly by the Center for the Study of the First Americans, Department of Anthropology, Texas A&M University, College Station, TX 77843-4352. Phone (979) 845-4046; fax (979) 845-4070; e-mail [email protected]. Periodical postage paid at College Station, TX 77843-4352 and at additional mailing offices. POSTMASTER: Send address changes to: Mammoth Trumpet JAMES KENNETT Department of Anthropology, Texas A&M University Kennett (above) in the field at Taroko 4352 TAMU Gorge, Taiwan, 2006; Kurbatov (below) College Station, TX 77843-4352 in Greenland, 2008. Copyright © 2011 Center for the Study of the First Americans. Permission is hereby given to any non-profit or educational organization or institution to reproduce without cost any materials from the Mammoth Trumpet so long as they are then distributed at no more than actual cost. The Center further requests that notification of reproduction of materials under these conditions be sent to the Center. Address correspondence to the editor of Mammoth Trumpet, 2122 Scout Road, Lenoir, NC 28645. Michael R. Waters Director and General Editor e-mail: [email protected] Ted Goebel Associate Director and Editor, Current Research in the Pleistocene e-mail: [email protected] James M. Chandler Editor, Mammoth Trumpet e-mail: [email protected] Laurie Lind Office Manager C & C Wordsmiths Layout and Design Tops Printing, Inc. Printing and mailing Web site: www.topsprinting.com World Wide Web site http://centerfirstamericans.com The Center for the Study of the First Americans is a non-profit organization. Sub-

SOLANA PYNE scription to the Mammoth Trumpet is by membership in the Center. We examined the EIH closely in a four- part series a while back (MT 23-1, -2, -3, Mammoth Trumpet, Statement of Our Policy -4, “The Clovis Comet”), even as inter- Many years may pass between the time an important discovery is made and the acceptance of research ested observers were scrambling to test it. results by the scientific community. To facilitate communication among all parties interested in staying abreast of breaking news in First Americans studies, the Mammoth Trumpet, a science news magazine, In the three years since, dozens of re- provides a forum for reporting and discussing new and potentially controversial information important search teams have published their find- to understanding the peopling of the Americas. We encourage submission of articles to the Managing ings. Many, including geoarchaeologists Editor and letters to the Editor. Views published in the Mammoth Trumpet are the views of Todd Surovell and Vance Holliday (MT contributors, and do not reflect the views of the editor or Center personnel. –Michael R. Waters, Director 25-2, “The Clovis Comet Revisited”), July ■ 2011 3

at the argument from both sides. Part I outlines recent research diamonds, they’re uncertain whether they were carried to supporting the presence of nanodiamonds in YDB deposits, Earth inside an impactor or created in place. They lean toward both in California and the latter. “We’ve performed geochemical analy- on the Greenland ice ses on nanodiamond residues,” says Kennett, sheet. In Part II we’ll “and the results indicate that the nanodiamonds tell you why other re- are not cosmic material, but instead formed from searchers believe that terrestrial material.” Experiments have proven this evidence is either

inconclusive, or dead ▲ An overview of the Kangerlussuaq sampling

wrong. site. ▲ Greenland team member Jøergen-Peder The nanodiamond Steffensen identifying the YD sampling site record near Kangerlussuaq. Nanodiamonds are so incredibly tiny that that lonsdaleite can be created from

millions can get lost in ANDREI KURBATOV oxygen-deficient, very hot carbon va- a liter of soil; it takes exquisitely por quenched in a warm substrate high-powered microscopes simply (such as soil) by a process known as to find them, much less analyze their carbon vapor deposition (CVD). A content and structure. They can take CVD origin for the nanodiamonds a number of forms, including cubic, seems to rule out the possibility that tabular, and near-spherical. Recent they might have been deposited by interest, however, has focused on means of some ET-related process shocked hexagonal nanodiamonds other than an impactor, such as a “cos- known as lonsdaleite. mic rain.” As far as we know, natural lons- The Arlington Canyon team has also daleite occurs only in association isolated a number of proxies suggest-

with meteorites and other cosmic im- SOLANA PYNE ing extensive biomass burning at the pacts. Lacking undeniable impact markers such as a crater, tek- YDB contact, including charcoal, carbon spherules, and acini- tites, and shocked quartz, many EIH supporters cite the form soot, which resembles tiny clusters of grapes. (This form lonsdaleite found in YDB deposits as convincing evidence that of soot is extremely rare in the geological record, known only to an impact may have triggered the Younger Dryas. Geologist be associated with ET impacts.) In addition, electron micros- James Kennett of the University of California, Santa Barbara copy studies revealed cubic nanodiamonds and metastable “n- summarizes the argument when he states, “Lonsdaleite has diamonds” embedded in the carbon spherules. Both cubic never been found associated with diamonds that formed through diamonds and n-diamonds are also impact markers; and like regular terrestrial pro- lonsdaleite, n-diamonds are never associ- cesses, so its presence in ated with terrestrial formation processes. the YDB suggests a cos- What’s more, the team has identified a phe- mic connection.” nomenon they attribute to mass wasting, Q.E.D.? Maybe, as- the sudden down-slope movement of mas- suming we can deci- sively rule out a terres- Three images of a hexagonal nanodiamond trial formation process (lonsdaleite) from the YDB layer, Greenland for lonsdaleite—which Ice Sheet. A, scanning transmission electron hasn’t been an issue so microscope (STEM) image; B, X-ray diffrac- far. A bigger problem tion image; C, crystallographic high- seems to be proving to resolution transmission electron microscope the critics that what the (HRTEM) image.

EIH proponents are call- JAMES KENNETT ing lonsdaleite really is lonsdaleite, a topic we’ll take up in detail sive amounts of sediment. All AMS dates from the lower 4 m of in Part II. the Arlington Canyon stratigraphic profile cluster around 12,950 ± 50 CALYBP, suggesting that the entire sequence was The Arlington Canyon evidence deposited quickly, possibly after fires denuded the region of In a recent paper in the Proceedings of the National Academy of vegetation. Science, Dr. Kennett and 16 colleagues reported on lonsdaleite While other researchers have since challenged their conclu- and other nanodiamonds detected in YDB sediments at Arling- sions, the evidence from Arlington Canyon is intriguing. Not ton Canyon on Santa Rosa Island, just off the California coast. only are nanodiamonds and proxies for biomass burning abun- While the team is confident that an impact event deposited the dant in the YDB sediments there, they’re roughly contempora- 4 Volume 26 ■ Number 3

neous with known wildfire events elsewhere on the Channel more cautious, pointing out tersely, “We mentioned in our Islands, as well as with the extinction of the Santa Rosa Island paper interpretations made by other scientists. We did not pygmy mammoths, Mammuthus exilis. This pattern of nano- study cryoconite holes in our work.” diamonds, fire evidence, and megafaunal extinction crops up at Actually, it’s hard to explain away the Greenland nano- terminal-Pleistocene sites all across North America. It’s not diamonds as the result of some event unrelated to the YDB clincher evidence for an impact, perhaps, but it’s certainly nanodiamond deposits observed elsewhere. It would require thought-provoking—especially since the only other known co- an astonishing coincidence, since this find represents the only occurrence of all three factors is at the famous Cretaceous- nanodiamond concentration in that interval of ice, and it appar- Tertiary boundary, the worldwide stratigraphic layer that ently dates to exactly the right time. Furthermore, the demarks the impact of the city-sized asteroid that finished off Greenland nanodiamonds are comparable in abundance, mor- the dinosaurs. phologies, and size range to other known YDB nanodiamonds. Some critics speculate that the team misidentified the par- The Greenland evidence ticles, suggesting that instead of being nanodiamonds they were In addition to the Arlington Canyon study, Kennett and 24 coau- actually quartz, rutile, copper, graphite, and carbonized thors recently reported in the Journal of Glaciology on the discov- spherules known as graphenes that can all display similar re- ery of a discrete, nanodiamond-rich layer in the Greenland ice sults. However, the Kurbatov group considers this highly un- sheet. In a 2008 expedition, spearheaded by Paul Mayewski and likely, since they identified the Greenland particles as Andrei Kurbatov of the University of Maine Climate Change nanodiamonds based on six different techniques of X-ray and Institute and accompanied by a NOVA television team, the team electron microscopy studies of their atomic structures. In par- pinpointed a layer of dusty ice near Kangerlussuaq that they ticular, the shapes appear to be wrong for most of those other determined is coeval with YD sediments elsewhere in the North- possibilities; and in any case, according to several different analy- ern Hemisphere. They couldn’t use conventional chronometric ses, the particles are made primarily of carbon. Furthermore, the methods to determine the age of fact that the Greenland the ice, given the dearth of datable particles identified as material; instead, the researchers lonsdaleite display a used as markers stratigraphic con- crystalline lattice spac- text and oxygen-isotope variations, ing of 1.93 angstroms focusing particularly on Oxygen-18 and a robust, multilay- (18O). As Kennett notes, “18O is a ered atomic structure proxy for temperature. The 18O val- seems to rule out ues suggest the diamond layer oc- graphenes, which are curred at or near the onset of a single-layered and have a major cold period, and the only significantly larger crys- known postglacial cold period is talline lattice spacing. the Younger Dryas.” One way to test their The team collected samples findings, Kurbatov suggests, is to drill new deep The Nova team in Greenland, ice cores elsewhere in led by Doug Hamilton. Greenland and in the Ca-

ANDREI KURBATOV nadian Arctic, and deter- along a trench 17 m long, representing about 6,000 years of ice mine if nanodiamonds peak in layers of the right age. The accumulation. Later analysis revealed a significant nanodiamond problem, as always, is expense. “Because of the serious price tag spike immediately below the dusty ice: at 5–50 parts per billion, associated with this activity, one of the compromises was to look concentrations were as much as 5,000,000 times higher than on the edge of the Greenland ice sheet,” Kurbatov says. “Ice background levels observed in the surrounding ice. One third of sublimation during the summer leaves a residue of dust particles the Greenland nanodiamonds were n-diamonds; another third embedded in the ice. Such a natural enrichment mechanism were lonsdaleite. Says Dr. Kurbatov, “We were very excited helped us to locate the nanodiamond layer, which is probably when we determined that the number of lonsdaleite particles harder to find in ice cores.” increased sharply and was associated with one layer.” Of course, it’s possible that something other than the hy- Intriguing possibilities pothesized YDB impactor enriched the ice with nanodiamonds. The EIH is a relatively simple concept, but as Kurbatov points For example, other researchers have found nanodiamonds in out, “It is changing existing concepts in several disciplines.” “cryoconite” holes, where meteors have melted depressions in That being the case, it continues to undergo rigorous testing— the ice, so it’s conceivable the spike resulted from a series of as any hypothesis that offers a new paradigm must do, before meteor strikes combined with some factor that concentrated it’s either added to the edifice of scientific knowledge or rel- the diamonds in one thin layer, as wind can do with lag gravels. egated to the dustbin of history. The current nanodiamond Kennett dismisses this as unlikely; cryoconite holes are rare controversy is just one aspect of that testing. It’s hard to say and isolated, and they found none in the study area. Kurbatov is continued on page 8 July ■ 2011 5

The C.W.Harris Site An Enigmatic Jewel

N THE BORDERS of an exclusive private community of million-dollar homes just north of San Diego, Califor- Onia, sits what researchers at the San Diego Archaeo- logical Center call “one of the most important archaeological sites in the United States.” Studied repeatedly over the last seven decades, the C. W. Harris site posts radiocarbon dates showing people lived here more than 10,000 years ago. Although its earliest occupation is as old as the Folsom culture, the site has been eclipsed by other sites, some earlier, occupied by people who left distinctly different artifacts. But C. W. Harris remains an enigma: Not only is it a Folsom-age site with no Folsom-type artifacts, there are no con- clusive data that tell us who its oldest occupants were, what they were doing at the site, or how they got there in the first place.

Sparse findings at its discovery First discovered in 1927, the C. W. Harris Site (CA-SDI-149) is the type site for the San Dieguito complex of Early American sites. In fact, it contains artifacts from three prehistoric cultural The C. W. Harris site excavation,1938 (W-198). periods, San Dieguito the oldest. Later components move up- Malcolm Rogers in foreground with E. L. Hardy. ward in age through the La Jolla period (7000–3000 years ago), ©SAN DIEGO MUSEUM OF MAN and into the late-Prehistoric period (3000 years ago to contact). ■ a perplexing near-absence of grinding implements such as Other sites in the overall complex include those of Pleistocene manos and metates, artifacts considered identifiers at so- Lake Mojave, to which it’s most closely linked, and several called “Desert Culture” inland sites. other sites that have yielded artifacts characteristic of early- Claude Warren and D. L. True, taking stock of the artifacts Holocene inhabitants of Southern California and surrounding in the C. W. Harris assemblage, judged them “quite different areas of the Southwest and northern Mexico. from those found on Great Plains sites.” Nevertheless they The earliest occupation of the San Dieguito complex is determined that the assemblage argued for a hunting characterized by: economy. The oldest part of the site, they concluded, was ■ large percussion-flaked biface knives and projectile points occupied by “an early hunting culture” with its closest affinity (but no fluted points similar to those of Clovis and Folsom- to “the early cultural levels of sites from the California desert age artifact assemblages); area.” The conclusions these researchers reached during the ■ numerous scrapers and one crescent-shaped stone consid- 1960s were a brave attempt to illuminate a site dogged by a ered by some researchers as an amulet; by others as spear- confusing history. shaft scrapers or food processors; by still others as hunting amulets, specialized projectile points used to stun water The first unsure steps fowl, or even ritual scarification knives (MT 25-3, “Study- Claude N. Warren was a young man when he plunged into the ing Crescentics: Form or Function?”); murky history of the site in the late 1950s. Now a professor 6 Volume 26 ■ Number 3

emeritus at the University of Nevada, Las Vegas, Dr. Warren tunately for North American archaeology, Rogers published has spent half a century investigating the site and writing few of his findings and what eventually made it into print was articles about it. confusing to Warren and other investigators. In 1958, Warren was a graduate student at the University of California Los Angeles (UCLA) Eureka A fresh start with new tools and a new hire to the university’s fledgling Ar- When Warren began digging at the site in the late 1950s he chaeology Survey program. “After a few months had an invaluable tool that Rogers had lacked, the new sci- my professors told me to go to San Diego and Borax ence of radiocarbon dating. Warren was able to verify Rogers’s ✪ Lake find out what San Dieguito was,” he says. “I conclusion that San Dieguito was indeed the older component. went down there but I didn’t really know what I Following excavations in 1965 and 1967, Warren dated the Sacramento was looking for.” He found the available literature San beginning of the San Dieguito component at the C. W. Harris Francisco riddled with a baffling jumble of terms and no site at earlier than 9030 ± 350 RCYBP and ending after consistent definition of the site or theory of its 8490 ± 400 RCYBP. Calibrated dates demonstrate that origin. people lived there around 10,365 calendar years ago. Part of the problem was the terminology used by Tulare Despite the march of decades since sinking a shovel ✪ Lake China Malcolm J. Rogers, who first identified the complex in Lake into the C. W. Harris site, Warren still recalls that his ✪ 1919. Rogers, a geologist by training and an experi- first encounter with it was a “really amazing” experi- enced archaeologist, first explored a facet of this ence. The site, he says, was largely exposed in a cut- complex, which was ultimately named after the San Los Angeles bank that the river had sliced right through. “We Dieguito Valley and the river flowing through it, at Site could see San Dieguito artifacts in deposits,” he SDI-W-240 in Escondido, also in San Diego County. ✪C.W. Harris says, that were really “very deep,” some 14 ft Rogers, finding at the Escondido site a wide variety of San Diego below the ground surface. unifacially flaked stone tools he believed were scrapers, as- Warren’s long career at the site had been born. signed the designation “Scraper Makers” to the occupants. He Today, half a century after he first set foot on the C. W. Harris later suggested that these toolmaker inhabitants (whom we now site, he believes that although the cultural history of the site know as the San Dieguito culture) were the second occupation in has been well delineated, questions remain to be answered. the region. Rogers believed the site was earlier occupied by For a start, archaeologists can’t agree on what constituted the people from a shell-midden culture near the Pacific Coast, where daily activity of the earliest inhabitants. In true scientific form, Rogers also explored. Rogers placed these coastal-oriented interpretations abound, and they have varied through the years people in the context of what became known as the La Jolla as investigators test theories with new excavations or reexamine complex. artifacts pulled from earlier digs. Some researchers have seen It wasn’t long, Warren tells us, be- the very early San Dieguito as big- fore Rogers changed his theories game hunters, possibly following about human presence in North Clovis and Folsom hunters; oth- America (at the time his theories were ers have viewed them as more considered pretty eccentric). Warren generalized foragers; a few have reminds us that Rogers worked in the declared the artifacts collected 1920s, when American archaeology 0 5 there examples of specialized tool cm was in its infancy and few people be- production, possibly morphing lieved there were sites more than 2,000 into the later La Jolla complex and years old. Rogers’s geologic training its artifacts. convinced him that the remnants of cultures he was finding were far older A plan of attack for the than that, although he lacked the future means to prove it conclusively. Edward Knell, an assistant professor of anthropology at Califor- Retouched flakes and scraping tools. nia State University, Fullerton, seeks to determine if the site was EDWARD J. KNELL a specialized tool-production lo- A lucky break came for Rogers in 1927, when flooding cus by taking an extremely close look at the lithic artifacts exposed artifacts in the San Dieguito River valley on a farm collected from the C. W. Harris site during the 1950s. Besides a owned by C. W. Harris. After examining the C. W. Harris site, detailed study of how the lithic technology was organized, his as it was later known, he struggled for 10 years—Great Depres- plans call for high-power microscopic use-wear analysis of the sion years, remember—to get money to excavate it. When he artifacts. Ultimately, he hopes to make a contribution to the site dug there in 1938 the well-preserved stratigraphy he found led literature by showing what the tools found there were used him to revise his earlier conclusions. It was now clear to him for—reading from the record of the lithic technology and use that the San Dieguito component lay beneath the La Jolla wear on the stone tools to determine what people did there. component, thus making San Dieguito the older people. Unfor- Dr. Knell’s special field of interest is the Paleoindian period. July ■ 2011 7

Although his primary work has concentrated on Great Plains Preliminary use-wear analysis of biface tools under low magni- components of the Alberta and Cody complexes, the C. W. fication suggests they were used on softer materials. Prelimi- Harris site has long been at the forefront of theories concerning nary use-wear analysis of scrapers—the collection contains early-Holocene origins and the cultural historical relationship more than 100 of various types collected from different levels— of the Great Plains indicates they were primarily used to process sites and to the La 0 0 5 3 harder materials, but the exact nature can’t be cmcm Jolla and San Dieguito identified under low-power magnification. complex. “People Which leads to another mystery: No faunal know about the C. W. remains have been recovered from the site. In Harris site,” Knell his paper Knell poses the question: “So what else says, “but they don’t were they using all those scrapers for?” quite know what to make of it.” What A theory to test strikes him in the dis- His working hypothesis, which he hopes to test cussions about the with detailed use-wear analysis of the tools, is site is that “nobody— that the scraping tools were used to manufacture nobody—has thus far published a detailed Later-stage bifaces. analysis where they EDWARD J. KNELL actually tried to assess what the San Dieguito people actually hafts from hard materials. Many of the scrapers show evidence were doing at the C. W. Harris site. . . . I’m just hoping to of continual resharpening and reshaping, a pattern that Knell contribute to that debate.” In fact, he has already done so with argues could result from “processing hard materials that rap- a paper he presented at the 2010 Society for American Archae- idly deplete tool edges.” Shaping hard wooden shafts for stone ology meeting in St. Louis on tools would certainly wear scraper an assemblage-level analysis edges. “I want to stress my work is of Claude Warren’s 1958 and very, very preliminary,” Knell, a care- 1959 excavations at the Harris ful scientist, cautiously adds. He em- site. His purpose was to assess phasizes that further testing is how the occupants from the necessary. oldest site component orga- If this hypothesis holds true, nized their stone tool technol- Knell says, it will strongly suggest ogy. Among his conclusions: that the C. W. Harris site was “a campsite as is generally perceived, ■ biface production and the and a whole lot more—a specialized use of scraping tools were shaft-shaping factory and activity major on-site activities, and center producing tools used to the bulk of the toolstone shape shafts for projectile points, as found on the site originated locally; Knell at the Hell Gap site, Wyoming. ■

the flaking patterns and EDWARD J. KNELL morphology of the few (four) biface cores suggest that well as the points themselves.” He is excited by the possibili- “flake blank production was minor in the scheme of on-site ties. In his paper he conjectures that “the remaining tools production activity”; could be by-products of other produc- 03tion-related activities, food consump- ■ substantial quantities of thin- cm tion/campsite activities by the work- ning flakes from 49 biface ers, or both.” blanks indicate biface manu- More surprises may be in store at the facturing was a major activity site as new technology is employed on there; most of the bifaces, the older artifacts or on other parts of however, are “manufacturing the site. Some surprises, in fact, have failures,” tools broken during already emerged as the result of CRM manufacture and discarded. studies mandated in anticipation of de- Knell pays particular attention to tools he believes were used on Scraping tools. site: 35 retouched flakes, 11 scrapers, 11 utilized flakes, 4 re- velopment of the San Dieguito Valley

touched flakes probably used as EDWARD J. KNELL during the 1990s. The San Diego Ar- multipurpose tools, 2 bifacial tools, a crescent, and a chopper. chaeological Center reports that a new wave of archaeological 8 Volume 26 ■ Number 3

In situ biface and flake cluster at has no doubt its earliest people Unit #1 of the C. W. Harris site “were living there seasonally, complex (Sdi-4935B), 1990. using local toolstone and making lots of tools.” His long-held testing with new technology has conviction that people came to disclosed some tantalizing find- the Americas at different times ings. Pollen analysis, for example, and along different routes is be- reveals that as a result of increasing confirmed as sites older ing aridity during the period of than benchmark Clovis and early occupations, grasses gradu- Folsom cultures emerge. ally replaced amaranth and pig- Today we don’t know where weed. Traces of deer and rabbit the San Dieguito people came blood detected in blood-residue from. Warren is confident we’ll

analysis of tools are the first solid ©SAN DIEGO MUSEUM OF MAN know tomorrow. evidence that occupants exploited faunal resources. –George Wisner Microspectroscopy of a tool revealed that one end is coated How to contact the principals of this article: with asphaltum, thereby confirming that natural tar was locally available. And excavation revealed intact rock hearth Edward J. Knell features and grinding tools dating to the late–La Jolla period. Assistant Professor Department of Anthropology Satisfaction after a long journey California State University, Fullerton Warren is delighted with the renewed interest in the site, now P.O. Box 6846 that new technology and new ideas can be brought to bear on Fullerton, CA 92834-6846 solving its mysteries. “There is no question that the site is e-mail: [email protected] old,” he says. Today he stands by conclusions he reached Claude N. Warren more than 40 years ago, that the San Dieguito complex of sites e-mail: [email protected] is unique and shouldn’t be lumped in with Great Plains sites or Desert Culture sites as some researchers have tried to do. Suggested Readings “The assemblages of the San Dieguito complex include large, Beck, C., and G. T. Jones 2010 Clovis and Western Stemmed: Population Migration and the Meeting of Two technologies in the leaf-shaped knives and rather crudely made points and a Intermountain West. American Antiquity 75(1) pp. 81–116. series of scrapers that are typologically and technologically Warren, C. N. 1967 The San Dieguito Complex: A Review and distinct from the material of the Great Plains,” he wrote in Hypothesis. American Antiquity, 32(2) pp. 168–85. 1967. Put simply, Warren remains convinced that the earliest Warren, C. N., and D.L. True 1961 The San Dieguito Complex and occupation is as old as Folsom; but because the artifacts don’t Its Place in California Prehistory. Archaeological Survey Annual resemble Folsom at all, he believes San Dieguito and Folsom Report 1960–61, pp. 246–338. Department of Anthropology and are culturally distinct. Sociology, University of Calfiornia, Los Angeles. Our blurred vision of the C. W Harris site is sure to come San Diego Archaeological Center Website: into focus with future investigations. In the meantime Warren www.sandiegoarchaeology.org/PDF/Harris%20Exhibit.pdf

cult to isolate. “This is very demanding work,” he explains. So how is the EIH faring as a scientific theory these days? The Clovis Comet Revisited “That’s a question for others to answer,” Kennett avers. “For us, the evidence has only become stronger—and so far, we see no continued from page 4 new evidence that refutes it.” what consensus will ultimately be reached, given the difficulty Join us next issue for Part II, when we’ll examine arguments of testing such a brief event, which produced only trace from the other side of the theoretical fence. amounts of materials that are, as Kennett observes, very diffi- –Floyd Largent

How to contact the principals of this article: James P. Kennett Andrei V. Kurbatov Professor of Oceanography Climate Change Institute Department of Earth Science University of Maine University of California 303 Bryand Global Sciences Center Santa Barbara, CA 93106 -9630 Orono, Maine 04469-5790 e-mail: [email protected] e-mail: [email protected] July ■ 2011 9

But it wasn’t until about 45,000 years ago, during the Upper Paleolithic in Eurasia (a.k.a. the Late Stone Age in Africa) that blade technology really took off. Until recently, many archaeologists have argued that this occurred solely because we put the sapiens in sapiens; that is, our evolved intellectual capacity spawned new technologies. As mentioned above, however, it seems that’s not quite true. Some archaeologists believe that such factors as increased mo- bility or perhaps access to high-quality raw material spurred on the UP/LSA blade revolu- tion; others, like Ofer Bar-Yosef of Harvard and Steve Kuhn of the University of Arizona, suggest that blades became popular and stayed in fashion because toolmakers increasingly relied Collins examining an artifact on hafted tools and composite or multicompo- recovered during excavation at nent tools, which require interchangeable parts. the Gault site, Texas. Blades are really good for this . . . but, as usual,

GAULT SCHOOL OF ARCHAEOLOGICAL RESEARCH I’m getting ahead of the story. Back home in the New World, there are es- sentially nine instances—at various places and times—when blade technology comes into play in North America. These Part III: nine instances are strictly associated with modern humans. In the mid ’90s, Bill Parry (CUNY) provided a fine summary of Producing Clovis Blades these blade industries, as they are sometimes referred to, which stretch from Alaska into California, across to various LADE TECHNOLOGY has woven its way in and out of the places along the Mississippi Valley, and down into Meso- archaeological record for tens of millennia, appearing on america. They’re associated with long-ago mobile hunter- different continents and at different latitudes, and per- gatherers like Clovis folks, as well as more sedentary, haps even “invented” by a different species of Homo. Though the complex societies found in Mexico, Central America, and the ability to produce blades and reliance on blade technology are larger valleys of the Ohio, Illinois, and lower Missouri rivers. often associated with anatomically modern humans, new sites Not only is there a broad “space-time continuum” associated and dating techniques have demonstrated that our nearest homi- with these industries, technological diversity (e.g., variation nid neighbor, Homo neanderthalensis, was clearly capable of in blade size and manner of production) exists as well. The producing blades in the Old World more than 250,000 years ago, upshot is that there is no historical continuity. Clovis blade before the advent of modern Homo sapiens sapiens. By and large, technology didn’t evolve down the line into the blade technol- though, they made the same stone tools using the same tech- ogy seen, let’s say, thousands of years later during the Pov- nique for more than 100,000 years. French prehistorian François erty Point period in parts of Louisiana and Mississippi. There Bordes says of the Neanderthal lithic industry, “They made are, however, a few things in common among these assorted beautiful tools stupidly,” meaning repeatedly or by rote. Nean- and sundry industries: the products and by-products . . . and, derthals, however, may have been more innovative than we of course, the particular kind of debitage that archaeologists previously believed. and lithics analysts define as blades. 10 Volume 26 ■ Number 3

Not all rectangles are squares, and not all flakes are during other types of lithic reduction. Jay Johnson of the Uni- blades versity of Mississippi provided some pretty useful criteria in Back in grammar school, we learned that a square is a special addition to length for defining prismatic blades: They should type of rectangle with four 90° angles and four sides equal in have prepared platforms, parallel lateral edges, and exterior length. (The equal length is what makes it a square, right?) Well, a similar rule exists for debitage: All blades are flakes. In Part II of this series, we discussed a number of different types of flakes: biface thinning flakes, overshot flakes, normal flakes, etc. This list goes on to include blades; after all, just like a flake, a blade is struck from a “parent” core. In its simplest form, a blade is defined as a flake that is twice as long as it is wide. Early on, the definition of blades followed this deceptively straightforward rule. The trouble is that it’s not really that easy. (You’re starting to anticipate this about lithics,

A regular cortical blade showing the exterior and interior blade surfaces and a lateral view that shows how radically curved Clovis blades can be. Note the parallel lateral margins and the previous blade removal on the exterior surface that creates a triangular cross section. This blade has a cortical margin that creates a naturally backed blade, which makes it easy to grip. Proximal end aren’t you?) A suite of other characteristics besides the 2L-W 05FRANK WIER guideline are essential for defining what’s referred to as a “true” cm blade technology—a prismatic blade technology. This is criti- flake scars that are parallel to the long axis of the blade itself. cal because long, narrow flakes can be incidentally produced Add to these a trapezoidal cross section and two ridges on the

brittle than the material being worked (you want to drive off a flake, not shatter your hammer). The platform on the core should have an angle of less than 90°, and the striking angle (the angle of the blow) must be less than 90°. Aim at a spot about 3– 6 mm from the core edge, as far back as the thickness of the Knockin’ on Rock flake you want to make. Use an easy, natural swing and follow HREE TECHNIQUES used to create blades, or any type of flake through; a great deal of force usually isn’t necessary, but devel- Tfor that matter, are direct percussion (using either a hard or oping a good aim is (just as in golf—trying to hit exactly the soft hammer), indirect percussion, and pressure flaking. Direct right spot on a platform can be as frustrating as trying to hit a percussion is perhaps the easiest of the three, and if you’ve little white ball). ever watched a skilled knapper, it looks Soft-hammer percussion so darned simple. Ha! Try it sometime, uses a material as a hammer- though. Making stone tools is a most stone that’s softer than the amazing experience. Replicating, ex- material being flaked, such perimental archaeology if you will, is as antler, bone, or a “soft” truly the best way to understand the rock like dolomite or lime- whole process of reducing lithics, stone. Because the hammer whether producing bifaces or blades. is softer, more force may be But watch out for sharp edges! required to remove a flake. Soft-hammer percussion is Dr. Bill Dickens explains the impor- useful in thinning, flattening, tance of preparing and maintaining and shaping bifaces; the plat- the correct platform angle on an form (the edge of the biface) experimental blade core. generally forms an angle well

BOTH PHOTOS: CHARLOTTE PEVNY under 90°. A soft hammer Hard-hammer direct percussion produces flakes by strik- slightly compresses when striking a core, thereby diffusing the ing a stone (one suitable for producing flakes) with another force of the blow over a larger area. When using a soft hammer, stone (one suitable for use as a hammerstone). A swing it in an arc. Aim most of the force onto the edge to detach hammerstone should be of a material that’s denser and less a long flake, and direct a little of the force downward to begin July ■ 2011 11

Moreover, to constitute a “true” prismatic blade technology, a lithic assemblage must contain blade cores (the parent rock from which a blade is removed) and specific types of debitage 05 that are exclusive to producing blades. For instance, if a potential cm core doesn’t have a suitable starting point for removing blades, the toolmaker must prepare and shape it to detach blades. Core preparation is a key element in defining a blade technology. Either bifacial or unifacial flaking can be used to create a ridge— Platform or crest—to guide the first blade removal. Detaching this crested blade establishes ridges for subsequent blade removals. As blades are successively removed, the platform from which they are struck may need to be “rejuvenated” to reestab-

Conical blade core, showing the platform surface and blade removal faces. Blades are removed in a single direction from the proximal, or platform, end downward toward the tapering distal end. The platform surface is scarred from multiple deep flake removals as a result of maintaining the platform. Several scars that terminate in a hinge or step fracture will need to be cleared by detaching a core-tablet flake before more blades can be removed. The dark outline (bottom right) shows the next potential blade that could be removed.

FRANK WIER Face lish the proper angle before another blade can be detached. This exterior blade surface that define this cross section, and you’ve platform-rejuvenation flake is often referred to as a core-tablet got yourself a bona fide blade. flake. Expended or exhausted wedge-shaped and conical

detaching the flake. It’s EXTREMELY IMPORTANT to prepare the Indirect percussion is a valuable technique for use when platform beforehand by abrading. To avoid breaking the piece making blades, especially small blades or blades of consistent you’re working, it’s equally important to support it (your thigh size, that require directing a blow at a precise spot on the core protected with a sturdy piece of leather or a folded towel makes platform. With this method you don’t directly strike the core a good work surface). platform. Instead, you hit a punch, which can be antler or wood You use soft- and hard- or today even metal, with a hammer, thereby hammer percussion to re- directing force at exactly the spot needed to move a flake by striking detach uniform flakes. The question that has the stone. With pressure plagued archaeologists and lithics analysts is, flaking, on the other Who or what held the core while the Paleo- hand, you remove a flake indian knapper’s one hand was holding the from a stone by pressing a punch and the other hand was striking the smaller, pointed pressure hammer? Theories have varied widely; most flaker (an antler tine prob- rely on the idea of some sort of vise—perhaps ably served Paleoindians holding the core between one’s toes or feet, as wonderfully) against a pre- adept Oriental artisans do today. pared platform. Pressure flaking gives you optimum Dickens has removed a large control and is therefore endthinning flake from the biface. usually used to finish the fine edges of tools. Pressure flaking removes flakes similar in Indirect percussion explains how Clovis and Folsom knappers shape to those made by soft-hammer percussion, but smaller. probably removed fluting flakes. After all, a fluting or end- Place the pressure flaker on the edge of the stone tool, then thinning flake is a flake usually twice as long as it is wide and very apply firm pressure downward and in the direction you want bladelike in that respect. Removing a fluting flake from the base the flake to follow. The edge must be dull enough not to crush of a projectile point requires extremely accurate aim. This is under the flaker, but sharp enough to prevent its slipping off. especially true if the base of the point is concave and the “ears” Again, abrading is a good technique to use to prepare the edge. of the point obstruct the platform of the fluting flake. Not only Not an easy technique to master. And beware of driving the does a punch improve the likelihood you’ll detach the flake, it detached flake into your finger or hand! also lessens the risk you’ll break the point. 12 Volume 26 ■ Number 3

about bifacial and blade cores, but there are other types lurking out there. With an informal, or amorphous, core, flakes are removed in many directions and from many platforms around the circumference of the core. Informal cores are usually considered Blades vs. Bifaces in the World less efficient because flakes aren’t removed consistently in any one direction from a single platform.) Mary’s research suggests Series of Clovis Technology that informal cores are actually more efficient producers of flakes than bifacial cores. N THE PAST, blade technology generally has been considered Metin Eren of Southern Methodist University and his cohorts Ia way to “get more bang for your buck,” an efficient means of recently tackled the issue of core efficiency to better understand producing a fairly standardized product (e.g., long, sharp the Middle to Upper Paleolithic technological transition in Eu- blades) that yields the maximum amount of usable cutting rope. Their experiment compared the efficiency of prismatic edge from a piece of toolstone. And, as we learned last time in blade cores with discoidal cores. (Yes, another core type. We part II of this series, the biface is a resourceful, long-life tool, archaeologists just love to “type” things. The discoidal or which at the end of its use life can be used as a core to produce Levallois core, a type of bifacial core shaped like a disc, is almost flakes (which can then be made into other tools). Thus the exclusively associated with Neanderthal lithic technology.) Metin biface itself can be made into a knife, projectile, or other and company found that reducing discoidal cores yields more bifacial tool. So which technology, blade or biface, is really usable flakes per gram of raw material than reducing prismatic more efficient? blades. This question hasn’t been an easy one for archaeologists to Hmmmm. . . . Informal beats out bifaces. Discoidal trumps answer, and attempts to answer it have produced mixed results blades. The top two long-standing contenders, biface and blade from researchers. For example, in recent experiments Mary technologies, knocked out of the efficiency race. So what’s really Prasciunas of Westland Resources, Inc. found no significant more efficient? More importantly, and to return to our subject difference in the yield of usable flake edge from reducing bifacial matter at hand, “What Does It Mean to Be Clovis?” what does cores compared with informal cores. (So far, we’ve only talked this mean for Clovis, which incorporates both biface and blade cores also can be considered by-products of producing blades. Collins was the first to try to understand and define Clovis blade technology by comparing blades from known Clovis con- A brief history of Clovis blade technology texts (like those found in the Blackwater Draw and Green caches First reported in the ’60s, blades were identified in two separate and at Clovis sites like Murray Springs) with blades from other Clovis caches at Blackwater Draw Locality 1 in New Mexico. But sites that may or may not be Clovis (like the Keven Davis cache they were rarely noted elsewhere in the U.S., and most archae- in Texas) or sites with surface context (like the Adams site in ologists considered the Blackwater Draw and Green blade Kentucky). His research prepared him to define “a distinctive caches (as they are referred to) mere constellation of attributes” isolated occurrences. Paleoindian ar- 0 5 for known Clovis blades chaeologists just didn’t believe blades cm (in addition to the basic were a common component of the Clovis characteristics that define toolkit. Of course, the evidence at the blades). These include time was sparse, and over the years that small abraded platforms, misperception has been set right, along longitudinal curvature, with several other misconceptions con- smooth interior surfaces, and length. Clovis blades Regular noncortical blades are loooooong, frequently showing the exterior surface. Because longer than 100 mm— these were removed by direct that’s over 4 inches on a percussion, the margins are gener- regular basis! Contrast ally parallel. Note the previous blade that with the average removals on the exterior surface length of Alaskan micro- that create trapezoidal and triangu- blades of around 20 mm, lar cross sections. Proximal end less than an inch.

SCOTT MINCHAK Since Collins’s initial re- cerning Clovis blade technology. Things really started getting search we’ve dispelled several other misunderstandings about interesting in the late ’90s for two reasons: 1) Texas State archae- Clovis blade technology. For example, we used to believe Clovis ologist Mike Collins published his seminal book on Clovis blades were produced using indirect percussion; we also believed blades, and 2) new Clovis sites, with an abundance of blades and that conical cores were typical of Clovis blade technology and that blade by-products in good stratigraphic context, were thor- wedge-shaped cores were the anomaly. Both false perceptions oughly excavated. were the fault of a simple lack of data. Identifying new Clovis sites July ■ 2011 13

technologies? Well, my good friend and colleague, doctoral transported stone very efficiently. On the Southern Plains, where candidate Tom Jennings of Texas A&M—with the help of high-quality, large nodules of Edwards or Alibates chert were knapper extraordinaire Bill Dickens—set out to do some ex- readily accessible, Folsom knappers used bifaces as cores. In perimenting to address this question. areas where raw-material size and abundance varied, say farther À la Clovis, we directly compared the flake-production to the north, then informal cores were the core of choice. efficiency of bifaces with Clovis knappers, however, didn’t the blade-production effi- rely heavily on informal cores, ciency of wedge-shaped though they occasionally appear at cores. Our data appear to a few sites and in a few caches. show that the package Reducing biface and blade cores is size of raw material, which the staple of Clovis lithic technol- considers weight, not the ogy, attested to by their abundance count of usable flakes, de- in caches (they’re part of the mo- termines which core-re- bile toolkit) and at archaeological ducing strategy to use. Minimizing stone trans- Cortical and noncortical blades port weight was a primary litter the ground after a concern for mobile Early knapping session by Dickens. Paleoindian hunter-gath-

erers. Makes sense, right? CHARLOTTE PEVNY sites of all types—kills, camps, and Who wants to lug around a bunch of heavy rocks? Though it quarries. Perhaps Clovis groups weren’t always über highly may seem counterintuitive, reducing informal or amorphous mobile foragers, like their Folsom successors. If so, then their cores is the most transport-efficient technology; small informal core technology wouldn’t value transport above other consid- cores (<1000 g) are more efficient than biface or blade cores. erations, such as the distribution of raw material and package Folsom hunter-gatherers—the most mobile of the mobile— size. like Gault in Texas and Topper in South Carolina, and reanalyzing of mobile hunter-gatherers and are used to walking). existing collections, especially those recovered from caches on Clovis blade caches vary widely in location (they’ve been the Southern Plains, have demonstrated that wedge-shaped cores identified as far away as northern Minnesota in the Pelland are definitely the “norm” and cache) and in raw-material that direct percussion was the Whole core-tablet flake Partial core-tablet flakes type. David Kilby has done technique most often used to legwork similar to Mike remove Clovis blades from Collins’s (and then some) the parent core. and come to the conclusion Clovis blade technology that the blades in these is a strong presence on the caches and others match Southern Plains and its pe- the defining characteristics riphery. In Texas it’s well of Clovis blades. Farther represented at the Gault east, blade technology is ro- and Pavo Real sites and the bustly represented at the Keven Davis cache. Small Topper site in South Caro- blades and a core-tablet flake were found at the Core-tablet flakes Aubrey site, also in Texas. rejuvenate the platform The Anadarko blade cache, by removing scars and another find rich in speci- other flaws, making a mens, is located just north fresh surface for striking 05 in adjacent Oklahoma. BILL DICKENS cm off blades. Blades and tools made on blades were recovered from the kill-associated camp at lina. A few other sites in the East, such as the Carson-Conn-Short Murray Springs, Arizona, west of Blackwater Draw. Blade and Johnson sites in Tennessee, hold promise for additional cores (and broken bifaces with endthinning) have even been comparative information on Clovis blade technology. recovered from the El Bajio site in Sonora, Mexico (MT 24-3, “Beyond the Border: Paleoamericans in Sonora, Mexico”). Producing Clovis blades The location of these sites and caches isn’t an accident: Based on our discussion above, we know that Clovis blades They’re all within walking distance of high-quality Edwards are removed in a definite sequence from a prepared core. A chert (that is, if you happen to be a member of a group specific set of skills is employed to create a desired product— 14 Volume 26 ■ Number 3

a Clovis blade isn’t just a random flake with a length twice its a cone or pyramid shape. In the case of a wedge-shaped core, width. As we mentioned in Part II of this series, the by- blades are removed from along the intermediate axis and are products related to initial core preparation (e.g., normal typically curved in profile. On a conical core, blades are typi- flakes) aren’t always easy to assign to either reducing bifaces cally removed from only one or two platforms (if blades are or producing blades, mainly because of the limitations im- removed from the top and bottom it’s considered a bidirec- posed by certain raw materials. But tional core). Blades some of these by-products, termed cor- from wedge-shaped 05 tical bladelike flakes and cortical ir- cm cores, on the other regular blades, are definitely more hand, often are re- bladelike than others. Even though moved from several they may not be related to blade pro- Platform faces, resulting in mul- duction, they can be twice as long as tidirectional cores. they are wide, with triangular or trap- Blade removal As blades are de- face Cortex ezoidal cross sections. Where do these tached, the platform angle changes and the Wedge-shaped core from the Gault platform is crushed, site, Texas, showing the platform, which makes it neces- blade removal face, and lateral or sary to reestablish a side view. The platform is free of viable platform. To hinge and step fractures, which further complicate the implies that the platform either had toolmaker’s task, a been freshly rejuvenated or required failed blade removal little maintenance. may terminate in a FRANK WIER step or hinge fracture imposters come from? Well, take an an- on the face of the core; gular piece of Edwards chert, for instance; the intersecting this flaw must be removed from the face before more blades faces frequently provide naturally occurring ridges for de- can be detached. In making Clovis blades, the platform is taching the initial blade. If there’s no ridge, then small flakes rejuvenated by removing core-tablet flakes. Sometimes re- are removed to create the edge or crest, which directs the moving a core-tablet flake from a conical core detaches the removal of the first blade, referred to as a crested blade. entire platform surface with a single blow—a phenomenon Detaching a crested blade observed at the Gault site. Core- leaves two ridges on the face tablet flakes are fairly easy to rec- of the core, which guide ognize because they bear the other blade removals. scars of previous blade removals Flakes and blades like to fol- around their perimeter. The plat- low ridges on core faces, be form of a wedge-shaped core is they bifacial or blade cores; Platform rejuvenated by removing one or in Lithics Speak, these two flakes, or occasionally a core- ridges are known as arrises. Crested blade tablet flake, from the platform sur- After the blade face and face. Once the proper angle and platform are established— an acceptable platform are rees- tablished on the core, blades are Conical blade core removed until the core is ex- prepared by removing hausted. The complex 3-D puzzle one end and bifacially of refits from the Pavo Real site flaking two faces to create clearly demonstrates this repeti- a ridge to guide the tive process of removing blades removal of the initial and rejuvenating the core. crested blade. Noncortical regular blades

FROM COLLINS 1999 are the desired product from coni- ideally 90° for conical cores, cal cores. Noncortical regular an acute angle greater than 45° for wedge-shaped cores— blades are also the intended product from wedge-shaped blades can be detached. Depending on the exact shape of the cores; in practice, however, blades removed from wedge- initial piece of raw material and the pattern of blade removal, shaped cores are often cortical regular blades that retain two types of blade cores may be formed, conical and wedge- cortex on one side of their exterior face. The cortical edge shaped cores. Clovis blades are removed around the circum- retained on some blades is a natural “backing” that makes it ference of a conical core parallel to its long axis; blade scars easier to grip an unhafted blade without cutting yourself on commonly meet at the distal or bottom end of the core, creating sharp edges. These “regular” blades (regular, because they July ■ 2011 15

unequivocally fit the definition of classic variety of microwear analysts like scrape a wide variety of materials like Clovis blades) were then used by Clovis Marvin Kay of the University of Arkan- meat, hide, bone, and plant resources. folks for a variety of activities. Mike sas, Scott Minchak of Texas A&M Uni- Usually they were used “as is” with little Collins refers to Clovis blades as “Swiss versity, and Marilyn Shoberg of the additional modification. Sometimes they Army knives,” and he’s got solid data to Gault School of Archaeological Re- were shaped into a form appropriate for a justify the playful nickname—extensive search. specific task, like boring or drilling. We analyses over the last 10-plus years by a Clovis blades were used to cut and haven’t found evidence, at least not yet, that Clovis blades were ever segmented or broken for making composite tools The Simon Clovis Cache similar to those found in Alaska (e.g., harpoons) and parts of the Old World More than 100 full-color images grace the pages of this book, which describes the (e.g., sickles). setting, history, and lithic artifacts of the Clovis cache discovered in 1967 on the Camas Prairie near Fairfield, Idaho. In 1997 the assemblage was donated to the How to contact the author of this article: Herrett Center for Arts and Science of the College of Southern Idaho, and since then Charlotte Donald Pevny, PhD, RPA it has been studied by hundreds of students and scientists. R. Christopher Goodwin & Associ- Author Steve Kohntopp, a professional archaeologist and lifetime resident of Idaho, ates, Inc. describes in detail the topographical setting of the site and recounts the results of 309 Jefferson Highway, Suite A laboratory analysis and field work at and near the site. He also gives an overview of New Orleans, LA 70121 utilitarian and ritual caching e-mail: [email protected] practiced by different cultures. The appendix to The Simon Clovis Cache includes photos Suggested Readings and line drawings of the 32 arti- Bar-Yosef, O., and S.n L. Kuhn 1999 The facts in the Herrett collection: Big Deal about Blades: Laminar Tech- nologies and Human Evolution. American completed projectile points, Anthropologist 101:322–38. and specimens in various stages of bifacial manfacture. The pho- Collins, M. B. 1999 Clovis Blade Technol- tos show the actual color of the ogy. Texas Archaeology and Ethnohistory various toolstones, which in- Series. The University of Texas Press, Austin. clude quartz crystal, chalce- dony, and mahogany obsidian. Green, F. E. 1963 The Clovis Blades: An

To order your copy, com- JAMES C. WOODS Important Addition to the Llano Complex. American Antiquity 29:145–65.

plete the form below and send JAMES W. HENDERSON You must use this form. This book is only available it with your payment to CSFA. You must use this form. This book is only available Johnson, J. K. 1983 Poverty Point Period directly from CSFA and cannot be ordered through TAMU Press. Blade Technology in the Yazoo Basin, Mississippi. Lithic Technology 12:49–56. Unit price Qty. The Simon Clovis Cache (2010, paperback, Kilby, J. D. 2008 An Investigation of Clovis 80 6-by-9-inch pages, ISBN 978-0-9786137-4-7) $27.50 Caches: Content, Function, and Techno- Subtotal: $ logical Organization. Unpublished Ph.D. U.S., $5 + $1 each additional volume dissertation, Department of Anthropol- Foreign, $10 + $2 each additional volume ➙ Shipping & handling: ogy, The University of New Mexico, Texas residents add 8.25% sales tax: Albuquerque.

Total: $ Parry, W. J. 1994 Prismatic Blade Tech- Ship to (please print clearly): nologies in North America. In The Organi- zation of North American Prehistoric Name Chipped Stone Tool Technologies, edited by P. J. Carr, pp. 87–98. Archaeological Address Series 7, International Monographs in Prehistory, Ann Arbor. City State Zip Sanders, T. N. 1990 Adams: The Manu-

e-mail address (in case we have a question about your order) daytime phone facturing of Flaked Stone Tools at a Paleoindian Site in Western Kentucky. Per- Payment method Check Money order MasterCard VISA Discover Card/NOVUS simmon Press Monographs in Archaeol- ogy, Buffalo, New York. Credit card no. Expires M/M Y/Y Signature Waters, M. R., C. D. Pevny, and D. L. Carlson 2011 Clovis Lithic Technology: Investiga- Need help placing a credit card order? Call (979) 845-4046 tions of a Stratified Workshop at the Gault Site, Texas. Texas A&M University Press. 16 Volume 26 ■ Number 3

The well-preserved skull of the Ayer Pond male Bison antiquus, radiocarbon dated to 11,990 ± 25 RCYBP (UCIAMS-53549). Note the healed injury to the left nasal. Wilson notes that B. Antiquus specimens found in the San Juan Islands had a body comparable in size to any of this species, “but the horn cores were reduced in size, in comparison with animals of comparable antiquity from the Great Plains. This is likely an insular effect—island animals often undergo size reduction or other changes, due to dietary or other factors. So there is much more still to be said (and learned) about these particular bison.” Pre-Clovis Butchers EXCEPT AS NOTED, ALL PHOTOS: STEPHEN KENADY of Bison antiquus

ORKMEN digging out a peaty wetland to create a Perhaps the most complete B. antiquus specimen pond on the south end of Orcas Island in Washington ever found WState made a discovery that is contributing to our The Ayer Pond bison was found in 2003, but the bones re- understanding of when and how the first Americans discov- mained in a cardboard box stored in the garden shed of one of ered and peopled this New World. In 2003, when a trackhoe the workers until 2005, when the remarkable discovery came to operator noticed a bone sticking out of the side of his excava- the attention of archaeologist Stephen Kenady with Cultural tion, he and his coworkers became Resource Management in Sedro- Bellingham curious. Digging into ancient Orcas Island Woolley, Washington. To him the peat with their bare hands, they importance of the bones was obvi- recovered the well-preserved ous, so in 2007 Kenady, with archae- skull and 97 other bones and Olympic ologists Dr. Randall Schalk, from Peninsula bone fragments of Bison antiquus, Seattle Cascadia Archaeology, and Robert an extinct form of bison that lived Mierendorf, from the National Park in this region during the Ice Age. Service, conducted a geoarchae- Despite the absence of stone tools, ological investigation of the discov- a team of scientists has found power- ery site assisted by volunteers fully convincing evidence that hu- Norm Exton and Marshall mans butchered this animal 400–800 Sanborn. Using a trackhoe, they years before Clovis hunters appeared in excavated four trenches more or America. less perpendicular to the edge of the pond. An ash Although this astonishing find hasn’t found universal accep- layer above the level where the bones were found proved to be tance throughout the scientific community, it has caught the tephra from the 6730 RCYBP eruption of Mount Mazama. The attention of so many scientists and members of the general bones therefore had to be considerably older than this. public that Discover magazine included it in the 10 top science The Ayer Pond bison skeleton, though missing many mem- stories for 2008. bers, nonetheless is remarkably intact. Kenady notes that July ■ 2011 17

▲ Loose relative refit of major skeletal

elements of Ayers Pond bison. ▲ The Ayers Pond Bison antiquus skeleton. Shaded areas identify unmodified bones; bones in red bear green-bone fractures, points of impact, cutmarks, or polish. Unshaded bones are missing.

ington, is one of scores of large and small islands in the San Juan archipelago, which skirts the northwest coast of North America. These islands are the remains of a much more extensive land mass that existed prior to the melting of the Pleistocene glaciers. It’s under- standable that the western coast of the Ameri- cas is widely considered to have been one of more-complete bison skeletons have been assembled (at La the avenues traveled by the original colonizers of this hemi- Brea, for example) as composites of bones from many different sphere. individuals. In the Ayer Pond specimen, Dr. Michael Wilson of The Orcas Island bison isn’t the first discovery of late-Pleis- Douglas College in New Westminster, British Columbia, notes tocene megafauna made in this region. Nearby Vancouver Island that “we have the luxury of an has yielded three Bison antiquus associated individual, and one finds. In the San Juan Islands at least with a cranium, at that—and all of 11 others have been found as well as it beautifully preserved.” the remains of ground sloth and giant Kenady’s team found sediment short-faced bear. The dated bison layers that closely matched what from three locations date from 11,990 the workmen had described. The to 10,800 RCYBP; the supporting eco- bones lay well below the ash layer system therefore survived at least and the peat in a layer of gray clay that long. The profusion of ancient and mucky silts that rested on a bones found in this region attests to layer of blue sand, which con- the abundant megafauna that once tained marine shellfish remains. thrived here; their exceptional state Posterior Anterior As relative sea level dropped, this region changed from a marine to ▲ Angular green-bone fractures, right distal humerus. a nonmarine environment and the depression became a freshwater ▲ Tibia with V-section chop/cut mark aligned with pond. A radiocarbon date on the astragalus split. bones places the death of the bi- ➙ son at around 11,990 RCYBP, of preservation can be attributed to numerous wetlands which agrees with the geological ➙ with low acidity. age of the deep layer where the Paleoamerican presence has also been documented in bones were found. the northern Pacific coast region. Clovis points and early stemmed lanceolate points have been stray finds in the A setting on the pathway to Puget Sound area, and points similar in form to Western the Americas Stemmed and Windust types were found at the DeStaffany Orcas Island, lying between the site in the San Juan Islands. Documented evidence for a southern end of Vancouver Island pre-Clovis human occupation includes coprolites and other in Canada and the coast of Wash- traces of human activity at the Paisley Caves in Oregon (MT 18 Volume 26 ■ Number 3

25-4, 26-1, “Paisley Caves” ) and the Manis mastodon, found been removed. This “gourmet butchering” pattern, they note, on the Olympic Peninsula in Washington State with a putative is typical of Paleoamerican bison kill sites. antler spear point embedded in a rib. In examining the breaks and markings on the bone, it became clear to the investigators that “nearly all fractured, cut, or Who (or what) killed the Ayer Pond bison? polished bone surfaces had the same color as adjacent unmodi- Bison antiquus was an extinct ancestor of modern bison, a fied bone.” This means that all bone modifications share “a monster 25% larger than its modern descendants. Fluted common history of exposure to processes which have effected spear points found with Bison antiquus bones at Folsom, New color change.” And because the workmen that discovered the Mexico, in 1926 provided the first clear evidence bones excavated them with- for the presence of people in America during the out using metal tools, appar- Ice Age. ent cutmarks on the bones couldn’t have been made by Examples of possible cutmarks shovels or trowels. (right) and polish (below). Kenady and his team don’t credit such Evidence from bone- indicators of possible butchering breakage patterns because they can be the result of The larger limb bones natural agents. “show spirally-fractured ‘green bone’ breaks.” No projectile points or stone tools of Green-bone fractures, char- any kind were found with the Ayer Pond acterized by sharply de- bison, so what makes the investigators fined edges and smooth think early Paleoamericans had any- fracture surfaces, identify thing to do with its demise? Various fea- bone that is from a living or tures of the site and the bison skeleton very recently dead animal. eliminate the possibility that natural Dried, older bone shatters agents were involved and lead to the con- in a very different way. clusion that the animal was killed or at Green-bone fractures, es- least scavenged by human hands. The pecially when associated evidence falls into three categories: site taphonomy, patterns of with points of localized impact, suggest that humans may bone breakage, and traces on the bones that meet the criteria for have intentionally broken the bones to obtain the rich marrow stone-tool impact marks and cutmarks. or to quarry stock for making bone tools (MT 23-1, “Early Mammoth Bone Flaking on the Great Plains”). Natural pro- Evidence from taphonomy cesses, however, can also account for green-bone fractures. The collection of beautifully preserved Ayer Pond bison bones That’s why the taphonomy of the site as well as other evidence includes both large and small bones. Features of the skull and for bone modification must be taken into consideration. teeth indicate the animal was a mature male 7–8 years old. That both large and small bones of the skeleton were found together Evidence from markings on the bones

argues against the Kenady and his co-investigators in- possibility that they ➙ terpret a variety of markings on the

were washed into the bones as evidence of human modifi- pond by rushing wa- ➙ cations, including “points of impact” ter, which would have and cutmarks. transported smaller A point of impact—damage bones farther than caused by a localized sharp blow or larger bones, espe- crushing force—is identified by a ring-shaped fracture radiating from Impact fractures. a single point. Kenady and his team observed 13 points of impact, of cially the skull. Moreover, there are no scratches on the bones which 5 were associated with concentric ring fractures. Four that would indicate they had been tumbled roughly along a bones exhibit polish, defined as a bone having “a rounded stream bed before coming to rest at the bottom of the pond. visibly shiny edge or point.” Four bones had parallel, straight, The Ayer Pond bison skeleton includes the skull, one tho- shallow V-shaped scratches less than 1 mm deep. Finally, V- racic vertebra (the spinal member that anchors a rib), the right shaped cuts more than 2 mm deep were found on a distal right lower front leg, left lower front foot, and both lower hind legs. tibia and fragments of the right astragalus; in fact, it was one cut Kenady and his co-researchers observe that the bones that are that severed the astragalus into two pieces. present consist largely of “non-meaty, low-utility elements,” The right radio-ulna—the lower front limb (forearm) bones of which suggests the highest-value parts of the carcass have the Ayer Pond bison—exhibits two impact fracture points that July ■ 2011 19

Kenady and his coauthors attribute to “blows by a heavy object with a gritty surface, likely a cobble chopper.” Large grooves with V-shaped cross sections they interpret as “cleaver-like chop marks.” They acknowledge that whereas some researchers regard such gouges as “definitive of butchering,” others caution against drawing conclusions from isolated occurrences. The American buffalo, Bison The relatively small number of cutmarks might be dis- bison, is an impressive beast. A missed by some authorities as evidence of butchering, but bull may stand 6 ft high at the AND TOURISM. Kenady and his coauthors observe that “experimental butcher- shoulder and weigh upwards ing studies using large ungulate bones have shown no relation- of a ton. ship between carcass processing intensity and creation of It pales beside its giant cutmarks.” Moreover, they note that cutmarks are rarely found ancestors, though. Paleo- Bison antiquus, formidable prey on bones from Paleoamerican butchering sites compared with american hunters had to for Paleoamerican hunters sites of later time periods. contend with Bison antiquus. The kinds of bones, and even the particular places on the Extrapolating from skeletal remains, Larry Todd at Colorado MANITOBA CULTURE HERITAGE bones, on which marks are found are almost as revealing as the State University estimates that a large bull stood 7½ ft tall and marks themselves. On a butchered bison at the Casper site in weighed 2,400 pounds! Its horns were equally daunting, 4½ ft Wyoming, for example, the bones exhibiting the greatest inci- from tip to tip. (The long-horned bison, a rarer species, had a dence of impact marks were the upper and lower limb bones. span of 6 ft.) Kenady and his coauthors note that the Ayer Pond specimen B. antiquus is a hypermorph, the term paleontologists use “fits this pattern closely.” to identify megafauna that flourished in the Pleistocene— Moreover, on the Casper bison impact marks occur in the giant beaver, Irish elk with 11-ft antlers, ox-size ground same places on the bones as on the Ayer Pond bison bones. The sloths. Why they grew so huge is still debated. Perhaps it was same patterning holds true on bison bones at the Hell Gap site a result of the super-fertile periglacial environment, or a in Wyoming and the Stewart’s Cattle Guard site in Colorado. In response to the increasing size of predators—saber-tooth both these instances the damage was attributed to Paleo- cats, dire wolves, and the giant carnivorous bear Arctodus. americans breaking the bones to get at the rich marrow. Paleoamerican peoples surely deserve respect: they killed DRAWING COURTESY OF HISTORIC RESOURCES BRANCH, Further arguing the case that humans and not natural and butchered B. antiquus using only flintknapped tools. agents modified the Ayer Pond bison bones, Kenady and –Ed. colleagues point out that two impact-fracture points on the right radio-ulna lie on the front-facing surface of the bone. butchered the carcass, or could the damage be the result of These blows could only have been struck “with the limb other processes, such as predation by carnivores? segment rotated and sup- Kenady and his colleagues studied ex- ported so that the [front] sur- amples of bone modification by wolves, dogs, face was uppermost.” They wolverines, African lions, hyenas, bears, and note that “similar fracture pat- the extinct giant short-faced bear, Arctodus terns have been interpreted as simus, to determine whether any of these evidence for butchering on predators and scavengers could have caused other bison” and that “almost the damage to the Ayer Pond bison bone identical fractures are known assemblage. Of all these candidates, only from Plains bison kills span- wolverines and short-faced bears regularly ning the Holocene.” Wilson ar- fractured bones as large as those from bison, gues their case most strongly: but they inevitably left large bones heavily “If we reject the Ayer Pond tooth-marked. Considering that no large butchering evidence we also teeth marks were observed on the Ayer Pond call into question a lot of the bison bones, these animals don’t appear to butchering evidence from have been involved with this carcass. Plains Paleoindian bison kills.” It’s noteworthy that Kenady and his team have dismissed as indicators of a human Alternative explanations presence such bone modifications as polish In the absence of stone tools, and parallel scratch marks. “Polish can re- the case for human involve- ment depends upon interpret- Mierendorf in an exploratory trench ing the Ayer Pond bison bones adjacent to Ayer Pond, 2007. The arrow themselves—the presence and marks where the bison was found. absence of certain bones, their spatial orientation, breakage patterns, and various marks found sult from many other processes such as abrasion from wind or on the bones. Can these data prove categorically that humans waterborne abrasives, licking by carnivores, and persistent 20 Volume 26 ■ Number 3

rubbing,” Kenady explains. “Likewise parallel scratch marks odon and mammoth bones found at the Heisler site in Michi- can be the result of movement across a coarse surface caused gan (MT 6-4, “Clues to Paleoindian Survival: Underwater by many natural agencies, and therefore they do not necessar- Caches May Have Supplied Meat in Winter”). Kenady and ily indicate human activity.” This is a team of careful scientists colleagues, however, regard the caching interpretation as who are confident they unlikely because most of the bones in the Ayer have excluded all natural Pond assemblage yielded relatively little agents as the possible meat. They admit the possibility, however, cause of the extensive sys- that high-yield bones may later have been re- tematic damage found on moved from the cache and only the least use- the Ayer Pond bison skel- ful parts were never retrieved. eton. Kenady and his coauthors conclude their analysis by stating that “butchering by humans is the Conclusions explanation that is most consistent with all of the Drawing on all the various physical evidence that is currently available.” If lines of evidence, Kenady they are correct in their interpretation, then the and his colleagues sug- date of 11,990 ± 25 RCYBP for the Ayer Pond Bison gest that the Bison antiquus assumes enormous significance, for it antiquus was killed, or at becomes irrefutable evidence for the presence of

least butchered, by pre- MEREDITH GRIFFITH/ISLANDS’ SOUNDER NEWSPAPER pre-Clovis humans along the Pacific Coast, which cursors of Clovis people in turn lends credence to on the frozen surface of the increasingly popular theory that immigrants Kenady x-raying the bison skull to search for ▲ from Northeast Asia en- possible lithic fragments embedded in a healed tered this hemisphere nasal injury, June 2010. following a coastal route instead of waiting for the Wilson on Orcas Island presenting the first Ice-free Corridor to public lecture on the Ayer Pond bison, 2006. ▲ open. Further investigations in this region may Ayer Pond. Any carcass parts left over would someday turn up evi- have frozen rapidly and so might have escaped dence of similar-age kill the notice of scavengers, which would account sites with stone or bone for the absence of tooth marks on the bones. tools. This would pro- When the ice melted in early spring, the remains vide the “smoking spear of the carcass would simply have settled gently point” needed to con- to the bottom of the pond, which would account vince doubting scientists for the tightly clustered large and small bones in that pre-Clovis Paleo- the pond deposits. americans were present

An alternative hypothesis is that human INEKE J. DIJKS in the Pacific Northwest hunters killed and butchered the bison elsewhere, then and that they killed and butchered the Ayer Pond bison. transported the remainder of the carcass to the pond, where –Bradley Lepper they deliberately submerged it to serve as a meat cache. This borrows from the idea proposed by University of Michigan paleontologist Dan Fisher to account for clusters of mast- Suggested Readings Kenady, S. M., M. C. Wilson, and R. F. Schalk 2007 Indications of Butchering on a Late-Pleistocene Bison antiquus from the Maritime How to contact the principals of this article: Pacific Northwest. Current Research in the Pleistocene 24:167–70. Stephen M. Kenady Kenady, S. M., M. C. Wilson, R. F. Schalk, and R. R. Mierendorf Cultural Resources Management 2011 Late Pleistocene Butchered Bison antiquus from Ayer Pond, Orcas Island, Pacific Northwest: Age Confirmation and 5319 Cedar Ridge Place Taphonomy. Quaternary International 233:130–41. Sedro-Woolley, WA 98284 e-mail: [email protected] Kenady, S. M., R. F. Schalk, M. Wolverton, M. C. Wilson, and R. R. Mierendorf 2011 A New Perspective on the DeStaffany Site in Michael C. Wilson, Chair the San Juan Islands, Washington. Current Research in the Pleis- tocene 25:105–08. Department of Earth and Environmental Sciences Douglas College Wilson, M. C., S. M. Kenady, and R. F. Schalk 2009 Late Pleis- Bison antiquus P.O. Box 2503 tocene from Orcas Island, Washington, and the Biogeographic Importance of an Early Postglacial Land Mammal New Westminster, BC V3L 5B2, Canada Dispersal Corridor from the Mainland to Vancouver Island. Qua- e-mail: [email protected] ternary Research 71:49–61.