40. ~FACTS VS. ARTIFACTS: AN EXPERIMENTAL STUDY OF THE MALPAIS INDUSTRY Michael J. Harner The Malpais "Industry` of the deserts centering around the lower Colo- rado River has presented a difficult theoretical problem over since it was first described by Malcolm J. Rogers (1939). Some archaeologists, mainly in oral discussion, havo questioned whether tho Malpais specimens wore really produced by the hand of man. Others, tentatively accepting Malpais as a cultural product, have raised the question of whether Malpais is evidence for a pr tearly lithic" occupation in North America, or whether despite its crudity, it represents a relatively late culture (Willey and Phillips, 1955, p. 734). Theso questions have mainly been caused by the facts that Malpais speci- mens are: 1) extremely crude in appearance; 2) exclusively surface finds of questionable human association and ago. The crudity of the Malpais ma- terials led Rogers (1939, P. 22) himsolf to state: Roalizing that archaeologists are sometimes led into improper interpretations through the enthusiasm engendered by a personal dis- covory, the writer wishos to stato that ho finds the formulating of a cultural pattern from hand choppers, primary flakes, and a few re- touched flakes as difficult to accept as will the critical roador. If some precursor of homo sapions wore involved, tho problem would be sinplified; but that mny type of. modern man could exist with such a primitive working equipment is difficult to believe. The Malpais specimens illustrated by Rogers (1939, p1. 3-5), unimpress- ive as they are, are probably among the best-looking ones collected. Only a very small proportion of the Malpals specimens that the writer has soen or personally collected in the lower Colorado River valley have any even half-way convincing evidence of human uso. There is a total absence of such distinguishing artifacts as blades, projectile points, or grinding tools. In fact, Malpais seems distinguishable by what is absent as much as by what is present. Many of the specimens bear a remarkable resemblance to the Old World eoliths and also to tho "pseudo-eoliths" manufactured by European archaeo- logists simulating geological processes (for a review and bibliography of the oolith question, see Boule ot Vallois, 1952, pp. 98-105). Unlike the European eoliths, however, Malpais materials occur on flat expanses of desert pavement and the surface locations of some Malpais sites make it difficult to dismiss all of these crude lithic materials as products of stream and gravel action. Although the lithic specimens called Malpais are often widely scattered over the desert gravels, this is not always the case. Sometimes they occur by the hundreds in concentrations of a few meters diameter in the midst of an expanse of otherwise undifferentiated desert pavement. Such desert pavements consist of the same gravels as the Malpais - 39 - specimons, but lack their sharply fractured surfaces. These are the desert areas which are characteristically located well above surfaces ever reached by flash floods or other moving water. In short, it is extremely difficult to explain the origin of some of those sites in terms of water and gravel action. If this situation is to be empirically verified, however, detailed reports are needed on a number of such sites by both archaeologists and geologists. Since I could not, to my own satisfaction, account for the origin of all the Malpais specimens in terms of geological causation, I was faced with either tentatively accepting omec of the Malpais materials as evidonce of a legitimate stone industry or finding somn previously unconsidered origin for tho specimens. Such an origin was suggested to me by Arnold R. Pilling, follow graduate student in the Department of Anthropology at Berkeley. His idea was that the Malpais materials might be the result of fracturing produced by aborigi- nal man through the dropping of hot stones into water for stono-boiling pur- poses. Another possible factor involving heat was the aboriginal practice in southern California and the Southwest of roasting moscal in stone lined pits.1 In order to test Pilling's hypothesis, I obtained information on tho heat-producing abilities of various kinds of firos from Dr. R. Keith Arnold, U. S. Forest Service Experiment Station, Berkoloy, and fron Mr. Ralph Fennor, consulting forester to the same station. Arnold and Fenner have both boon conducting experiments on the subject. Arnold reported that either a forest fire or an ordinary wood-burning campfire is easily capable of heating stones equivalent in size to those of the desert gravels to a temperature of 1000 degrees Fahrenheit, if they are on the ground underneath the fire. Fenner supplied the additional information that in a brush fire an extremely concen- trated pile of brush (such as a rat's nest), can likewise heat such cobbles to 1000 degrees F. Howeverr, he ruled the possibility out that a grass fire would be able to do so. After obtaining the above information, the following experiment was conducted. Seventeen Malpais specimens of jaspor and chalcedony from UCAS site Yuma 27 (Arizona) were placed in an electric laboratory oven which had been heated to a temperature of 500 degroes F. (because an electric oven rises in temperature very slowly compared to an ordinary wood firo, the oven was pro-heated to simulate better the conditions of an actual fire.) After approximately 20 minutes, the ovon reached a temperature of 1000 do- grees F. It was held at this temperature for 2 hours. The oven was then permitted to cool to 900 degroos F. and was opened. It was found that the 17 specimens had fractured into 54 fragments plus a number of minute chips. The fragments and chips thus produced by the experiment appeared to resemble the original specimens in shape, though, of course, they wore smaller. The fragments wore then reheated to 1000 degrees F. (from 900 degrees F.), indi- vidually removed and quickly plunged into a kettle of water at room temperature. 1Tho experiments by Blackwelder (1927, p. 138) indicate that diurnal changes in temperature can be dismissed as a significant factor in rock fracturing. - 4o - No significant additional fracturing or flaking was produced by the water immersion, oven though the stones were rod-hot when dropped into the water. Thus the experiment suggests that stone boiling is not an essential stop in the heat-fracturing of rock. A nunbor of additional oven-runs have since boon made, using Malpais specimens from tho sane site and producing more than 600 fragrents. The laboratory procedure has been the same as outlined, except that the oven has boon permitted to cool to 500 degrees F. before removing the specimens; and the water irneorsion stop has been eliminated. These subsequent runs have made it evident that Malpais-like fractures are not produced as fre- quently as the results of the initial experiment indicated. Oven-produced surfaces on the Malpais specimens tend to have less chipping along the edges and the fragnonts as a whole tond to be less conchoidal in their appearance. Although the fragments wore not exeminod microscopically, additional dif- ferences should be observable microscopically. While the original Malpais specimens sometimes exhibit bulbs of percussion, none of the oven-run Malpais specimens thermally fractured in such a way as to produce a new bulb of percussion. It should be noted, however, that this was not the case with a non- Malpais rock of obsidian. This was subjected to identical laboratory treat- mnt. The obsidian produced several chips with bulbs of percussion. By reconstructing the rock from its component chips, it was possible to deter- mine that every bulb of percussion was the direct result of percussive blows. These blows had been struck with an iron harmer when the rock was broken off of its parent obsidian mass. In other words, tho hoat fractures, in producing bulbs of percussion, were apparently following lines of stress set up by previouis percussive action. Whether such lines of stress could be built up by natural percussive forces was not determined, but seems pro- bable. In sumrary, thermal fracturing has failed to reproduco sonm of the char- acteristics which typify Malpais specimens. If the experiments adequately simulated the conditions of an actual fire, and if certain of the sites bearing Malpais lithic materials should prove to be unexplainable) on a geo- logical basis; then it seems likely that wo will bo faced with the conclusion that a certain proportion of the Malpais lithic materials are evidence of a legitimate stone industry. There are two "iftas" in the above statement and both can be resolved only by competent fieldwork and reports. First is the question of whether the laboratory ovon adequately simulates the actual conditions of a real fire set at an actual site with local fuels. If fieldworkers could find it feasible to set campfires at Malpais sites and report the fracturing re- sults, the first "if" undoubtedly could be resolved. The second "if" is the question of whether or not the occasional concentrations of Malpais lithic materials can bo explained on a geological basis alone. It seems unlikely; but in order to be certain, detailed studies of a number of such concentrations by archaeological-goological teams are needed. The Malpais materials present a problem of dating as well as one of origin. It may bo useful to suniarize here the presont status of the dat- ing problem. - 41 - One of Malcoln Roger's arguments (1939, pp. 19-29) for the antiquity of the Malpais materials is that they are invariably coated with a layer of accumulated iron or manganes0 oxide. This coating is conmonly called "desert varnish" and generally is most noticeable on the oxposed surfaces of desert rocks. The speed at which it accumulates has been a subject of controversy and it now appears that local conditions considerably affect the rate of accumulation. In the course of the writer' s archaeological survey work, desert varnish, or a coating identical in appearance to it,l was discovered on the exposed surfaces and exposed broken edges of potsherds and a pottery vessel found in the desert west of Yuma (Harnor, 1953, p. 5). In view of this information, caution should be exercised in utilizing desert varnish as an indicator of considerable antiquity in the lower Colorado River region. The nondescript nature of Malpais and the necessity of having a statis- tically adequate sample of specimens considerably complicate attempts at dating it. This is one of the problens involved in attempting to equate Malpais with the artifacts of the Ventana Cave volcanic debris layer (dated pro-8000 B. C.; Haury, 1950, figs. 116, 117). Rogers apparently felt that this correlation could be made (Haury, 1950, tablo 12). However, careful reading of the Ventana report leaves the suggestion that the question is still open. Rogers also recomanded (1939, p. 21) assigning the Tule Springs, Ne- vada, site (now dated "older than 23,800" years; Libby, 1954, p. 740) to Malpais on the basis of some "oxidized., silicified limestone" spocimons which wore not found in situ. The one chip found in situ at Tule Springs is of obsidian (Simpson, 1933, pp. 8-9). [Recent research at the site has not revealed any more chips in situ (personal conversation with M. R. Harrington, August 31, 1955).] However, Rogers had reported, "Either chert, silicified limestone, or some metamorphic phase of the saeo were exclusively" [italics mino] the materials of which the Malpais specimens wore composed in that area of Nevada. In conclusion, the results of the experiments suggest that although firo-cracked rocks may have smetimos boon mistaken for Malpais specimens, fire-fracturing cannot reproduce all of the characteristics found on some Malpais specinons. In addition, follow researchers should be cautioned that the Malpais lithic materials, whatever their origin and age, cannot be satis- factorily recognized (let alone dated) without a sample of at least several hundred specimens. Such specimens must be virtually unmixed with other lith- ic assemblages, becauso the Malpais materials are essentially indistinguish- able from much of the waste chipping that normally occurs in lithic industries. 1 Identified as iron oxide by Professor Norman E.A. Hinds, Department of Geological Sciences, University of California, Berkeley. - 42 - BIBLIOGRAPEY Blackwelder, Elliot 1927. Fire as an Agent in Rock Weathering. Journal of Geology, vol. 35, pp. 134-140. Boulo, Marcellin, et H. V. Vallois 1952. Los Homes Fossiles: Elents do Paloontologie Humaino. Masson et Cie., Paris. Earner, M. J. 1953. Gravel Pictographes of the Lower Colorado River Region. Reports of the Univ. of Calif. Archaeological Survey, no. 20, pp. 1-32. Haury, E. W. 1950. The Stratigraphy and Archaeology of Ventana Cave, Arizona. The University of Arizona Press, Albuquerque and Tucson. Libby, W. F. 1954. Chicago Radiocarbon Dates V. Science, vol. 120, pp. 733-742. Rogers, M. J. 1939. Early Lithic Industries of the Lower Basin of the Colorado River and Adjacent Desert Areas. San Diego MusmUm Papers, no. 3. Simpson, G. G. 1933. A Nevada Fauna of Pleistocene Type and its Probable Association with Man.. American Museum Noitatps, no. 667, pp. 1-10. Willey, G. R., and Philip Phillips 1955. Method and Theory in American Archaeology II: Historical- Developmental Interpretation. American Anthropologist, vol. 57, pp. 723-819. - 43 -