1. Gemegah 1999.
2. Bolles 1999 discusses a poet (Elisha Kent), a professor (Louis Agassiz), and a politician-geologist (Charles Lyell) and “how their clash of egos, ignorance, and imaginations led to the discovery of the Ice Age.”
3. Hrdlička 1923.
4. Meltzer 2006.
5. Wormington 1957.
6. Boldurian and Cotter 1999.
7. In this book we have had to deal with the thorny issue of reporting dates so that they are accessible to lay readers as well as archaeologists. We have chosen to report both radiocarbon dates and calendar dates whenever possible. Radiocarbon dating is done on organic remains recovered from sites, be it charcoal from a hearth or collagen from a bone. The date indicates the time of death of the living organism, based on the radioactive decay of carbon-14 that the organism acquired while alive, but it is expressed as radiocarbon years before present (RCYBP) and has a range depending on many factors, such as the purity of the sample and the ranges of individual readings within the sample. Although it was originally thought that the radiocarbon calendar would be the same as our calendar, the amount of radiocarbon in the atmosphere varies through time and therefore the derived dates cannot be used directly. There have been numerous studies that correlate radiocarbon dates with other means of dating, such as tree-ring, to yield absolute (calendar) dates and to derive calibration curves. We have chosen to use the OxCal calibration curve (the version current in September 2009) and have calibrated our dates directly from the online calculator at http://c14.arch.ox.ac.uk/embed.php?File=oxcal.html.
Even with the calibration curve worked out there are further complications, because radiocarbon dates are always given with a range and standard deviation (e.g., 12,240±40). This indicates that the date could be anywhere between 12,280 and 12,000 RCYBP. So what is the calendar year? This would also be a range of dates. However, to simplify our reporting of calendar dates we have usually chosen to give the calibrated date based on the actual radiocarbon date, disregarding the standard deviation. Therefore the calendar dates we give should be considered approximations rather than absolutes: we frequently use calendar years as general statements, especially when carbon dates were not available (because the dates we have either are based on other approximations, such as geological dating, or extend beyond the calibration curves, which end around 21,000 RCYBP). Readers who want to deal with more explicit calibrations are encouraged to do their own research with the available calculators.
8. W. A. Johnston 1933.
9. Hayes 1991.
10. Thompson 1948. See Dixon 1993 for a complete review of Alaskan fluted points.
11. Solecki 1951.
12. Wormington and Forbis 1965.
13. Stanford 1983a.
14. Dyke, Moore, and Robertson 2003.
15. Fladmark 1978.
16. Bednarik 1997.
17. For an overview of the controversies about the first Americans, see the papers in Shutler 1983. For a recent update on these issues, see the papers in Bonnichsen et al. 2005.
18. Jelinek 1965.
19. Frison and Bradley 1982.
20. Greenman 1963.
21. The earthly remains of an individual now known as Kennewick Man for a nearby town or as the Ancient One to Native Americans, were found by two students along the bank of the Columbia River in Washington State. The preliminary coroner’s inquest noted that his skull was shaped significantly differently from those of modern Native Americans and looked more like that of a European. Therefore, it was concluded that he was of European descent, although the weathered condition of his bones implied that he was not a recent casualty but perhaps a nineteenth-century trapper or settler. When a stone weapon tip known to be thousands of years old was found embedded in his hip bone, it was clear that the story of Kennewick Man was more complex and that he would provide important clues to the deep prehistory of the Columbia River Basin. A radiocarbon assay made from a sample taken from one of his finger bones suggested that Kennewick Man lived around 9,500 years ago, sending newspaper reporters into a virtual feeding frenzy, with some of them proclaiming that ancient Europeans discovered the Americas. The popular articles were of course inflammatory, and angry Native Americans clamored for the immediate reburial of the Ancient One.
The remains of several other individuals from different sites of roughly the same antiquity have been studied since the discovery of Kennewick Man, and they all fall into the same general physical type. The so-called Caucasoid features upon which the original assessment of Kennewick Man was made are indeed characteristic of Europeans today; however, these were common physical traits in early prehistoric populations in the Northern Hemisphere and are even retained by some modern south Asians. The apparent dominance of these physical features at an early date in America implies that the peopling of the New World occurred when early modern humans were first spreading throughout the world. It was only after the north Asian physical type developed that tribes dominating the ancestry of modern Native Americans moved into the Americas and intermingled with the pre-existing populations that they met along the way. For additional details, see Chatters 2001 and Benedict 2003.
22. There are two forms of DNA, one found in the cell nucleus and one which occurs in the cytoplasm surrounding the nucleus. The latter, termed mitochondrial DNA or mtDNA, is known as the “Eve gene” because it can be used to trace human heritage through time, presumably all the way back to Eve. This is possible because mtDNA is present in the maternal egg cell. Researchers have determined that mtDNA occurs in only a few dozen general patterns, which are called haplogroups. These groups have been assigned letter identifiers, and each one is believed to comprise the descendants of a single prehistoric female. A new haplogroup forms when a mutation occurs, which happens at the rate of 2.2–2.9 percent over a million years. The spread of related female lineages can be traced through the identification of mtDNA haplogroups and their distributions across the landscape. For instance, preliminary research indicated that most Native Americans could be traced back to four Asian haplogroups (A, B, C, and D), while Europeans find that they are descendants of nine haplogroups (I, J, and K, thought to be the original Paleolithic peoples, were followed by H, T, U, V, W, and X). But mitochondrial haplogroup X also occurs in North American populations more than 9,000 years old (see Smith et al. 2002) and in modern Native Americans (see Brown et al. 1998 and Schurr 2000), pointing to an ancient link between western Eurasia and North America. Apparently, mtDNA X originated in North Africa and spread into Europe and into North America sometime between 13,000 and 34,000 years ago, long before Clovis. An important bit of information from our point of view is that the easternmost occurrence of the X haplogroup is in the Altai region of south Siberia. If the ancestors of the Native Americans who carry the X marker and their clans traversed Siberia on their way to America, they did so without leaving any genetic markers behind. This is unlikely human behavior.
Lepper 2009 suggests that the occurrence of mtDNA X, along with other North American haplogroups, in a small percentage of the Altai region population makes its appearance in the Americas less of a mystery. Before we consider this conclusion a fact, we must understand how X was transmitted from western Europe and why it does not occur in eastern Asia. Note that the Altai region is farther from Nome, Alaska, than from Paris. Another curious bit of information is that X is most common in tribes from the St. Lawrence Basin of eastern North America, becomes increasingly rare to the west, and is totally absent from the Arctic and sub-Arctic regions of North America (and Asia), a pattern that is similar to the distribution of Clovis artifacts. The relationship between mtDNA haplogroup X and either Clovis or Solutrean people is unknown at this time, as no human remains associated with either tradition have been tested, but we hope future research will resolve these issues.
23. Flemming 2004.
24. Kirch 2000 presents an excellent overview of the early evidence of boat travels in the Pacific. Strasser et al. 2010 discusses data from twenty-eight pre-ceramic lithic sites found on Crete. These sites have been dated by their geologic context to be at least 130,000 years old. Because Crete has been separated from the mainland for millions of years, the presence of these sites indicates “open-sea navigation and multiple journeys and pushes seafaring in the Mediterranean back more than 100,000 years” (145–90). It will be interesting to follow further research here: if this conclusion is correct, it could have major implications for the dispersal of modern humans.
1. See, e.g., Cotterell, Kamminga, and Dickson 1985.
2. Speth 1974.
3. Whittaker 2004.
4. Speth 1972, 1974.
5. Whittaker 1994.
6. Bradley 1982.
7. For an example of hand ax technology, see Bradley and Sampson 1986.
8. Bordes 1961.
9. Bradley and Giria 1996.
10. Collins and Lohse 2004.
11. Goebel 2004.
12. Bradley and Stanford 1987.
13. Ingbar and Frison 1987.
14. Jodry 1998.
15. Solutrean: Aubry, Nevis, and Walter 2003; Clovis: Bradley, Collins, and Hemmings 2010.
16. Kelterborn 1984.
17. Sinclair 1995.
18. For an example of controlled overshot flaking that is unlikely to be historically related to either Solutrean or Clovis, see Inizan and Tixier 1978.
19. For a description of one possible fluting method, see Gryba 1988.
1. Waters and Stafford 2007.
2. After the Last Glacial Maximum, between 25,000 and 13,000 years ago, the climate warmed and stabilized. This episode is known as the Bølling interstadial. It lasted from 14,700 to 12,700 years ago and preceded an unstable period whose climate shifted between warmer and cooler conditions. This ended with an abrupt plummet into a long-term cold, dry, and windy episode known as the Younger Dryas, from 12,800 to 11,500 years ago, which returned the environment to ice age conditions.
3. Bonnichsen and Will 1984.
4. E. Callahan, personal communication, 1999. See also Gardner 1974.
5. Collins, personal communication, January 2011.
6. Collins et al. 1989.
7. Broster and Norton 1993.
8. Mason 1962. At the time of Mason’s publication, reviewers cited the lack of both supporting radiocarbon dates and evidence of a progenitor of fluted point technology as reasons to consider his suggestion premature at best.
9. Brennan 1982.
10. Lepper 1983.
11. Renaud 1934. See also Figgens 1934; Cotter 1935.
12. There are a limited number of other possible variants, but none of these has been well dated or found in stratigraphic relationship to Clovis.
13. Carson-Conn-Short: Broster and Norton 1993; Williamson: McCary 1951; Adams: T. N. Sanders 1990; Sinclair: J. Broster, personal communication, 2009; Pine Tree and Topper: Goodyear and Steffy 2003.
14. Frison and Todd 1986.
15. D. G. Anderson and Gillam 2000.
16. Blades: Greene 1963; points: Stanford and Jodry 1988; mixes: see, e.g., Frison and Bradley 1999.
17. Meltzer 2002.
18. The McInnis Cache, found south of St. Louis, Missouri, comprises eleven bifaces, nine blades, and a broken Clovis point made of Burlington chert. Another Clovis-style biface made of Burlington chert was found on the Columbia River just below the Grand Coulee Dam in the 1950s by Erma Pryor. Mrs. Pryor gave this specimen to Marilyn Palmer. Subsequently, Mrs. Palmer gave the specimen to Peggy Carnes, who along with her daughter, Kelly Carnes, donated it to the Smithsonian. All of these specimens are housed at the Smithsonian Institution in Washington, D.C. Although out of context, the Burlington chert biface was probably made near St. Louis and probably carried in an unmodified condition to a location on the Columbia River. Its likely route of travel was up the Missouri River into western Montana and across the Lolo Trail into the Columbia River Basin. Perhaps a party of Clovis people pioneered the trail that Sacagawea led Lewis and Clark along 13,000 years later: the original voyage of discovery.
19. The reliability of radiocarbon dating is again under scrutiny, and the current debate, based on new data, indicates that it is very difficult to interpret and should be used with caution. This is especially true for dates between 11,000 and 15,000 years ago derived from non-archaeological sources, where scientists have discovered great inconsistency.
20. Charlie Lake Cave: Fladmark, Driver, and Alexander 1988; Alaska: Bob Gal, National Park Service archaeologist for Northwest Alaska, personal communication, 2009.
21. C. V. Haynes Jr. et al. 1999.
22. T. A. Morrow and J. E. Morrow 1994.
23. Frison and Bradley 1999.
24. Beck and Jones 1997.
25. Adair-Steadman: Tunnell 1977; Lindenmeier: Wilmsen and Roberts 1978; Hanson: Frison and Bradley 1980.
26. C. V. Haynes Jr. and Huckell 2007.
27. Bement and Carter 2003.
28. G. Haynes and Stanford 1984.
29. A new date from the bone indicates that it was of Clovis age and not related to the later bison kill. See Bement and Carter 2010.
30. McNett 1985.
31. Ferring 2001.
32. Stanford 1983b.
33. Collins et al. 1989.
34. Collins 2002.
35. A Clovis point that retains significant portions of the flake blank surface was included in the Anzick Cache in Montana, and at least four were encountered at Blackwater Draw. Most appear to have been made on thin flakes. The reduced surface is the dorsal side of the flake or blade, and the less modified surface is the ventral side. Pressure flaking to flatten the bulb of percussion and minor edge retouch from shaping are the only modifications on the ventral faces. The dorsal surfaces were finished with the same steps and techniques used to produce a bifacial specimen.
36. Wilke, Flenniken, and Ozburn 1991.
37. To develop a comparative sample for the presence of controlled overshot flake scars on non-Clovis bifaces, we examined 418 bifaces from Archaic and Late Prehistoric contexts at the Gault site. These and the underlying Clovis bifaces were produced in the same location, and the knappers had access to the same sources of high-quality flint. Any differences between technologies cannot be explained by raw material availability. We studied bifaces recovered by Prof. J. E. Pearce during excavations in 1929 (see Pearce 1932). The vast majority of these were unfinished, but we did not include fragments. We classified overshot flake scars as: error (the flake removal resulted in the termination of production), unknown effect of overshot flake removal (did not harm the biface), and possibly intentional. Although the artifacts were recovered from post-Clovis deposits, there is the possibility that some were actually Clovis in origin and were mixed into the later deposits in a number of ways (the excavations were done with shovels, and mixing of strata was frequent). Two of the bifaces we examined look to have been Clovis based on a suite of technological traits, including controlled overshot flaking, and a distinctive yellow patina seen on many Clovis artifacts. Nevertheless we included these in our sample. Out of a total of 418 bifaces, 409 (98 percent) did not have overshot flake scars and 9 (2 percent) had them. Of these 9, 3 were abandoned because of overshot failure, 1 had what looked like an intentional overshot, and 5 had overshot scars that did not result in abandonment but also were not evidently intentional (most were at the tip or base of the biface). All told, overshot flaking at Gault in post-Clovis technologies is extremely rare and does not represent an intentional part of any of the biface technologies extending from Late Paleo-Indian through Late Prehistoric times.
38. Bradley 1997.
39. This possibility is even greater in the Folsom fluting method, which demands even more expertise and has higher failure rates. For other evidence that Folsom fluting may have taken on a ritual function, see Bradley 1982.
40. Collins 2002.
41. Broster and Norton 1993.
42. For more information, see Bradley, Collins, and Hemmings 2010, 10–55.
43. End scrapers are tools that have been retouched on the end of a flake or blade. Although they are called scrapers, microscopic analysis of use traces has indicated that this tool type had multiple uses—generally for preparing animal skins for rawhide or leather but sometimes for working wood and other more durable materials such as bone and antler.
44. Shoberg 2010. The butchering reference is in figure 6.4, p. 143.
45. Tomenchuk and Storck 1997.
46. Wilke, Flenniken, and Ozburn 1991.
47. Nami et al. 1997.
48. Hemmings, Dunbar, and Webb 2004.
49. Hannus 1997.
50. Gramly 1993. The East Wenatchee Cache is also known as Richey-Roberts.
51. Lahren and Bonnichsen 1974.
52. Owsley and Hunt 2001. There are some real problems with these two skeletons relative to their association with the Clovis Cache. The bone foreshafts have a good, solid Clovis age radiocarbon date of 11,040±35 RCYBP. The toddler has a date of 10,680±50 RCYBP, while the other child has a date of 8,600±90 RCYBP. The toddler’s bones are covered with red ocher, but the other child’s remains are not stained. Unfortunately, the entire site, including the artifacts and remains, was disturbed by earthmoving equipment, and the exact locations of the burials relative to the cache are unknown. It may be that they were not associated with the Clovis Cache but were incidentally buried nearby and the red ocher staining the toddler’s bones is purely coincidental.
53. Recent studies indicate that these items were made from elk antler rather than bone; see J. Morrow and Fiedel 2004.
54. Lahren and Bonnichsen 1974.
55. Stanford 1997.
56. C. V. Haynes Jr. and Huckell 2007.
57. C. V. Haynes Jr. and Hemmings 1968.
58. Boldurian and Cotter 1999.
59. While at Point Barrow, Alaska, between 1881 and 1883, J. Murdock, an ethnologist from the Bureau of American Ethnology, collected thirteen hafted mauls and thirteen flaking hammerheads. Murdock 1892 shows a small hammer with a bone head. This hammerhead is undamaged, but others in the collection and hammers used in flintknapping experiments are broken in the manner of the Blackwater Draw specimen.
60. Saunders et al. 1990.
61. The rod was originally identified as ivory, in Frison and Stanford 1982. The analysis in the supporting material for Waters and Stafford 2007 determined that the artifact is made of either bone or antler.
62. S. D. Webb 2006.
63. Dunbar 1991.
64. Wilmeth 1968.
65. Cressman 1946.
66. Tankersley 1997.
67. Hemmings 2004.
68. Stanford, Bonnichsen, and Morlan 1981.
69. Hannus 1989.
70. Bradley 1995.
71. Hester 1972.
72. Collins et al. 1991.
73. Rawls 2009.
1. Hopkins et al. 1982.
2. W. A. Johnston 1933.
3. Mandryk et al. 2001.
4. West 1996.
5. Hopkins et al. 1982.
6. See Colinvaux 1981; Colinvaux 1996.
7. Elias et al. 1996.
8. World Wildlife Fund 2008.
9. With minor reservations, we concur with the interpretations summarized in West 1996. Where we part company is in the interpretation of the existing Arctic data relative to the origin of Clovis culture. See also West 1981.
10. See, e.g., Abramova 1984; Kashin 1991.
11. Pitulko et al. 2004.
12. Although edge grinding is unreported in Siberian assemblages, a putative fluted point from the Uptar Site in northeastern Siberia, dating to less than 9,000 years, has heavy basal grinding. The point is not fluted but instead has an impact scar. When inverted 180 degrees, it fits into the same bi-pointed projectile type as was found at the site. See King and Slobodin 1996.
13. Derev’anko 1998.
14. Goebel, Waters, and Dikova 2003.
15. Dikov 1996; Powers 1973.
16. Mochanov and Fedoseeva 1996.
17. Vereshchagin and Ukraintseva 1985.
18. See Vartanyan 1995; Vartanyan, Garutt, and Sher 1993.
19. Vereshchagin and Ukraintseva 1985.
20. Hoffecker, Powers, and Goebel 1993.
21. D. D. Anderson 1970.
22. Odess and Rasic 2007.
23. Crass and Holmes 2004.
24. West 1996.
25. Larson 1968.
26. Humphrey 1966.
27. Alexander 1987. For additional information on the dating of fluted points in Alaska, see Kuntz and Reanier 1994; Dixon 1999.
28. Hamilton and Porter 1975.
29. Kuntz and Reanier 1995.
30. See, e.g., Dixon 1999.
31. Ted Goebel and Kelly Graff, personal communication, November 2010.
32. Gryba 1988.
33. Fladmark, Driver, and Alexander 1988.
34. Kuntz, Bever, and Adkins 2003.
35. Kuntz and Reanier 1994.
36. Hamilton and Goebel 1999.
37. Justice 2002.
38. Lohse and Sammons 1994.
39. The Sluiceway Site was found and tested by Bob Gal and Tom Hamilton in 1992 and later excavated by Dennis Stanford in 1994 and 1998. During the summer of 1998 a large Sluiceway manufacturing site was found by Jeff Rasic on Tuluaq Hill in the Western Brooks Range, and subsequent surveys in the area and examination of collections produced thirty-four sites with Sluiceway artifacts. See Rasic 2008.
40. See Rasic 2008.
41. Hamilton and Goebel 1999.
42. Cook 1996.
43. See Goebel, Powers, and Biegelow 1991; Holmes 1996; Holmes, VanderHoek, and Dilley 1996.
44. Crass and Holmes 2004.
1. Fitzhugh 2001.
2. Not all technologies develop successfully to deal with new situations, and sometimes they transform for unknown reasons. Henrich 2004 examines the change in technology in Tasmania, which has resulted in a continual loss of knowledge, and finds no environmental or adaptational explanation.
3. There is a lengthy history of archaeological research seeking to find “pre-Clovis” in the Americas. See Dixon 1999 for a comprehensive review of these research projects and a critical appraisal of the North American sites considered older than Clovis.
4. For example, we do not include the evidence of an early occupation at the Monte Verde site in Chile, as we do not see a connection between the technology used there and the development of Clovis in North America. We contend that the paleolithic people of Monte Verde had a completely different ancestral heritage, stemming from a population from another place and time than the ancestors of Clovis people.
5. Adovasio and Page 1997. Everything in the text about Adovasio is from here.
6. Although some scholars have argued that the stratum IIa charcoal samples were contaminated by coal particulate or soluble material incorporated into the groundwater to explain their unexpected antiquity, a detailed independent study of the microstratigraphy found no evidence of either contamination or fluctuating groundwater levels and concluded that the early dates are probably as accurate as the later ones. See Goldberg and Arpin 1997.
7. Boldurian 1985.
8. Another argument against the earliest Meadowcroft dates is that the plant and animal remains in the lower levels are all from species adapted to modern woodlands rather than a late Pleistocene environment. However, this incongruity does not invalidate the radiocarbon results. The sample of organic remains from these levels was extremely small. Only eleven animal bones could be identified, from a white-tailed deer, southern flying squirrel, deer mouse, passenger pigeon, snake, and toad. While these animals are indeed characteristic of the Carolinian biotic province (basically a modern habitat; for an excellent discussion, see Colthurst and Waldron 1993), they have broad ecological tolerances and have been found throughout the Carolinian and Canadian biotic ecotone (the boundary of two or more major ecological zones) and in equally old paleontological records farther north, in the Upper Ohio Valley. In fact, animals such as the white-tailed deer are extremely adaptable and thrive in a wide variety of environmental niches, making the significance of their presence ambiguous. J. E. Guilday and P. W. Parmalee, the archaeozoologists who analyzed the Meadowcroft fauna and fauna from many other sites of similar antiquity in the eastern United States, suggest that the period of transition between the ice age and modern environments may have taken place during a relatively short time in western Pennsylvania (see Guilday and Parmalee 1984). Such rapid habitat shifts can be observed today. For instance, the current warming of our climate has pushed the Baltimore oriole northward sufficiently that its range no longer includes Baltimore, and ravens have replaced it as the Baltimore totem animal. Similarly, robins previously unknown above the Arctic Circle have recently begun to delight the Inuit children at Point Barrow, Alaska, by heralding the onset of summer.
The plant remains indicate that a mixed conifer-hardwood forest dominated by oak, hickory, pine, and walnut trees and with a hackberry understory surrounded Meadowcroft. While these species argue against a boreal or tundra environment, they are compatible with a mosaic or patch environment rather than a uniform periglacial one. This interpretation is supported by Meadowcroft’s low elevation and southern exposure, whose milder ambient temperatures would have allowed earlier colonization by some woodland life forms. Further, the plant and animal associations during the last ice age were “disharmonious” by today’s standards, combining, for example, patches of tundra, woodland, and boreal communities of the kind represented in the Meadowcroft area at the time of the earliest occupations of the shelter.
9. McAvoy and McAvoy 1997.
10. Wagner and McAvoy 1997, 2004.
11. See McWeeney 1997, 2000.
12. Collins 1999.
13. M. F. Johnson 1997.
14. See Lowery 2007; Lowery et al. 2010.
15. Wah 2003.
16. R. C. L. Wilson, Drury, and Chapman 2000 has an excellent summary of the geologic and environmental processes at the end of the LGM.
17. Lowery et al. 2010. Tilghman Island is the namesake of the paleosol, which Lowery named.
18. Lowery 2008.
19. Hobbs 2004.
20. Specialized studies of the Cinmar site include a radiocarbon assay by Tom Stafford of Stafford Research, Inc., Lafayette, Colorado; use-wear studies by Marvin Kay of the Department of Anthropology, University of Arkansas, Fayetteville; a search for the source of the projection point’s rhyolite by Jeff Speakman of the National Museum of Natural History, Smith sonian Institution; and lithic analysis by Bruce Bradley of the Department of Archaeology, University of Exeter.
21. See Stewart 1987; Bondar 2001. For further reading on eastern meta-rhyolites, see Stewart 1984.
22. Shea 1999. See also Grosman et al. 2011.
23. Hallin 1983.
24. Faure, Walter, and Grant 2002.
25. Weaver et al. 2003.
26. Lowery et al. 2010.
27. J. D. Webb 2006.
28. Dunbar and Hemmings 2004.
29. Redder 1985.
30. Daniel and Wisenbaker 1987.
31. Dunbar et al. 2005.
32. Hemmings 2004.
33. Baker and Broster 1996.
34. See Waters and Stafford 2007. Several years after the first dates were obtained from the hearth features, an attempt was made to validate the original results. New charcoal was collected from the cut bank and assayed to circa 9,000 years old. But there is a 9,000-year-old occupation level above the proto-Clovis level, and since the cut bank had eroded since the first assay, the charcoal collected on the second dating attempt was probably either from the wrong level or redeposited down the bank by surface erosion. Whatever the cause of the younger date, it is clear that the artifacts recovered from the lower level are technologically much older than the 9,000 years level, and there is little reason to disregard the dates from the original assay.
35. Adovasio and Page 1997.
36. Overstreet 2004.
37. Overstreet et al. 1997.
38. Overstreet 2004.
39. Arnold 2002.
40. Turner et al. 1997.
41. Mill Iron: Frison 1996; Jim Pitts: Donohue and Sellet 2001.
42. M. C. Wilson and Burns 1999, 235.
1. Jelinek 1965.
2. See, e.g., Churchill and Smith 2000.
3. For a comprehensive discussion of the Aurignacian, see Bar-Yosef and Zilhão 2006.
4. Pettitt 2008.
5. See Sandars 1968.
6. The traditions in Portugal and Mediterranean Spain share general biface and blade technology, shouldered points, and other artifact forms, such as laurel leafs and end scrapers, with the northern Solutrean. They share other characteristics, such as arrow points, with some North African technologies, which Otte 2002 sees as evidence of Solutrean origin.
7. Pettitt et al. 2003.
8. P. E. L. Smith 1966.
9. See, e.g., Aubry et al. 1998.
10. See, e.g., Straus and Clark 1986.
11. Hogg and Johns 1995.
12. Rigaud and Simek 1990.
13. Sackett 1981.
14. Combier and Montet-White 2002.
15. Walter and Aubry 2001.
16. P. E. L. Smith 1966.
17. Rigaud and Simek 1990.
18. Straus 1986.
19. Clottes and Courtin 1996.
20. Straus and Clark 1986.
21. Bird 2002.
22. See, e.g., Butzer 1986.
23. Adovasio, Soffer, and Kléma 1996.
24. Adovasio, personal communication, 2001.
25. Straus 1992.
26. Pokines and Krupa 1997.
27. Chadelle, Geneste, and Plisson 1991.
28. Straus reports that of the nineteen concave base points found at La Riera Cave, eighteen were from levels 4–7 and the other was out of context. Straus and Clark 1986.
29. P. E. L. Smith 1966. See especially figure 74–13.
30. Sinclair 1995.
31. Bordes, personal communication, 1969; Collins 1973; Tiffagom 2006.
32. Collins insisted it was necessary to show a pattern of heat alteration that occurred during the manufacturing process, after initial flaking and before final flaking, as indicated by differences in luster of adjacent flake scars.
33. Whittaker 1994. Modern knappers frequently use French boxwood billets for striking blades, but it is unlikely that these would have been available to Solutrean knappers.
34. Aubry et al. 2003.
35. To be counted as exhibiting overshot flaking the biface must have had at least one overshot flake scar that removed part of the edge opposite the platform. Many pieces exhibited more than one such scar.
36. Pelegrin 1990.
37. Clottes and Courtin 1996.
38. Toggling harpoons have detachable points tied to a line that either attaches to a float or can be held by hand. When the harpoon penetrates the animal, the shaft detaches, leaving the point inside. Force on the attached line turns the harpoon point sideways, greatly increasing its attachment to the animal. These points were designed to stay in. While this design is superior to that of other points used to procure sea mammals (if a line isn’t attached, many animals are lost when they sink in deep water), other forms have been used successfully, such as the self-barbed spear point. There are also cases where toggling harpoons were used to hunt terrestrial animals.
39. Garcia 1927.
40. P. E. L. Smith 1966; Rigaud and Simek 1990.
41. Bradley, Anikovitch, and Girya 1995.
42. Garrod 1926.
43. Otte 2002.
44. Strasser et al. 2010.
1. Toth and Schick 2005.
2. Amvrosievka: Krotova and Belan 1993; Horner: Frison and Todd 1987.
3. D. H. Thomas 1998.
4. See, e.g., Bordes 1972; Movius 1975; Leroi-Gourhan 1957; Straus and Clark 1986; Aubry et al. 2003; Combier and Montet-White 2002.
5. J. M. Adovasio, personal communication, 2007.
6. By generalized we mean reconstructed through the study and analysis of a large number of individual pieces. This sequence is sometimes called the production strategy, and by the French chaîne opératoire. Especially helpful in these reconstructions are pieces on which we are able to refit the flakes and thus see the sequence of individual choices. These are rare, however, so we often have to rely on reconstructions based on pieces that were abandoned during the flaking process, frequently fragments.
7. We have used early, middle, and late phase to describe Clovis and Solutrean biface reduction, but not all biface technologies can be readily subdivided in this way. Therefore for this analysis we have opted for the terms primary and secondary flaking. Generally, the early and middle stages fit with primary and the late stage with secondary flaking. While we have attempted to use observations that archaeologists would generally agree on, there could certainly be alternative interpretations of specific attributes.
The characteristics selected for use in this analytical approach vary depending on the technologies being compared. For example, the comparison of blade technologies does not include such traits as thinning and finishing. Ideally, the characteristics of a technology are determined through rigorous peer review; although DSA has not undergone this process (yet), we feel it is useful and should be further developed.
8. See Bradley, Collins, and Hemmings 2010 for specific examples and definitions of these technical terms.
9. For the Solutrean: Collins 1973; for Clovis: Wilke, Flenniken, and Ozburn 1991.
1. Demars and Laurent 1992: 40–41, figure 63 (“micrograttoir de Laugerie-Haute”). These scrapers also appear in the Grimaldi Magdalenian and become common in the Azilian and other Epipaleolithic assemblages.
2. Clovis: Collins 1999; Eurasian Upper Paleolithic: see, e.g., Giria and Bradley 1998.
3. Clovis and Solutrean peoples produced bladelets and larger blades from the same core types, whereas Asian microblades were produced from a variety of specially prepared wedge-shaped microblade cores.
4. Bradley, Collins, and Hemmings 2010.
5. Tunnell 1978; see also Collins 1999.
6. Bradley, Collins, and Hemmings 2010.
7. Lowery and Stanford 2008.
8. Wilke, Flenniken, and Ozburn 1991. However, the appropriately strict criteria developed by Mike Collins to confirm heat treatment were not applied. See Collins 1973.
9. Stanford 1997.
10. Hemmings 2004.
11. Hemmings, Dunbar, and Webb 2004.
12. Picat 2009.
13. Guthrie 2005; Villaverde Bonilla 1994 is an excellent two-volume analysis and illustration of the art of Parpallo Cave.
14. Excavations in South Africa have revealed extensive bead use and manufacture in 70,000-year-old deposits in Blombos Cave. See Bouzouggar et al. 2007. While very early compared to Eurasia, this is nevertheless likely the work of modern humans rather than an earlier hominid form.
15. Stone pendants have been recovered from the Uptar Site in western Beringia, but their association with early dates is questionable at best. See King and Slobodin 1996.
16. Amick, Hoffman, and Rose 1989.
17. Kilby 2008.
18. Mochanov and Fedoseeva 1996.
19. D. D. Anderson 1970.
20. These floors are just post-Solutrean in time, and there was a minor Solutrean use of the site, but the associated artifacts are Magdalenian.
21. See chapter 6 in Bradley, Collins, and Hemmings 2010.
22. Straus et al. 1980.
23. Owsley and Hunt 2001.
24. Straus 1990. Straus 1977 points out that concave (indented) base points are restricted in distribution to a narrow (circa 25 kilometer) coastal zone of northern Spain near the Cantabrian Sea (Bay of Biscay) and the French border. Smith 1966 suggests that concave points might have been used for hunting specific game. Alternatively, Straus 1977 hypothesizes that they might have had social or territorial connotations. We agree that both situations may have been in play, but we suggest that the concave base points show up in this narrow coastal zone due to its proximity to the coast and that they were fashioned for harpoon-related gear. These points do not appear in Solutrean sites in the interior of Spain or France; moreover, at La Riera Cave they are restricted to the levels that contain the majority of fish and shellfish remains. Thus, we think coastal-adapted people may have seasonally used the uplands for hunting, while other Solutrean groups used the cave for their activities and employed different types of hunting gear armed with shouldered points.
25. Straus and Clark 1986.
26. Frison and Bradley 1980.
27. Straus, Meltzer, and Goebel 2005 presents a compilation of Beringian-Siberian Paleolithic typological and technological traits intended to show that traits we think link Clovis and the Solutrean are also found in Siberian Paleolithic cultures, thereby negating our hypothesis. Among the problems with their comparison is that they include middle through late Upper Paleolithic sites from an immense area stretching from southwestern central Siberia through central Alaska, and by including the Yana Site and Culture Zone 3 at Broken Mammoth they borrow from a span of 17,700 years. What couldn’t one find in such an immense area through such a huge time range? By contrast, we are comparing well-defined technologies from two restricted areas: the Solutrean from southwestern France and the area of Spain north of the Cantabrian Range, and the Mid-Atlantic pre-Clovis through Clovis. For additional discussion of Straus, Meltzer, and Goebel’s evaluation of the Solutrean hypothesis, see Bradley and Stanford 2006.
28. G. Haynes 2002.
29. We concur with Straus, Meltzer, and Goebel 2005 about what we should find in pre-Clovis assemblages. Although the pre-Clovis record in eastern North America is still sparse, the artifacts that have been found do meet these expectations. The Beringian record does not.
1. Straus and Clark 1986.
2. Ibid., 350–65.
3. Binford 2001.
4. Altuna 1986.
5. Freeman 1973 estimates the average weights of usable meat from various species of animals hunted by the Paleolithic peoples of northern Spain.
6. Binford 2001.
7. Pokines and Krupa 1997.
8. Ortea 1986.
9. Binford 2001, table 7.13.
10. Butzer 1986.
11. Straus and Clark 1986, 367–83.
12. Mol et al. 2008.
13. Lowery 2001.
14. Erlandson 2001.
15. Straus and Clark 1986, 350–65.
16. Cleyet-Merle 1990.
17. Straus 1990.
18. Straus 1977 suggests that the variation in projectile point types might reflect either seasonal or functional use of different hunting gear or the use of different point types by different Solutrean groups.
19. Straus 1990.
20. Lyman, Clark, and Ross 1988.
1. CLIMAP 1976.
2. COHMAP 1988.
3. Mix, Bard, and Schneider 2001.
4. For an in-depth but readable discussion of the issues of understanding the ice age physical environment, see R. C. Wilson, Drury, and Chapman 2000.
5. Robinson, Maslin, and McCave 1995.
6. The amount of 18O incorporated in the organism’s shell increases relative to the amount of 16O as the water gets colder. A 1°C drop in sea surface temperature results in an increase of 0.2‰ in the δ18O.
7. Late Proterozoic, between about 600 and 800 million years ago; during the Pennsylvanian and Permian, between about 250 and 350 million years ago; and the late Neogene to Quaternary, the past 4 million years.
8. Bond et al. 1993.
9. Heinrich 1988. See Alley 2000 for an informative discussion of Heinrich and Bond events.
10. Dansgaard et al. 1993.
11. Dyke et al. 2002.
12. Bowen et al. 2002.
13. Bellamy 1995.
14. Butzer 1986.
15. Straus, Bicho, and Winegardner 2000.
16. For the latest summary of the results of excavations at Solutré, see Combier and Montet-White 2002.
17. Rigaud and Simek 1990.
18. Ortea 1986.
19. Altuna 1986.
20. Reeves et al. 2002.
21. Clark 1946. Due to sea level rise there is little direct evidence of seal hunting in Europe before the later Paleolithic period.
22. “Recent ethnoarchaeological work among coastal foragers in Torres Strait indicates that for reasons of transport efficiency fish and especially shellfish are likely to be processed close to shore, and that (depending on processing costs) few elements susceptible to archaeological preservation will be moved more than a kilometre or so off the beach,” Bird 2002.
23. O’Connor and Veth 2000.
24. Butzer 1986.
25. Ibid.
26. Miller et al. 2002.
27. Dyke, Moore, and Robertson 2003.
28. Prest 1984.
29. Dyke et al. 2002.
30. There is an unconfirmed recent report that a mammoth tusk was dredged up from the Grand Banks. If verified, this would mean that the Grand Banks sustained plant and animal life. The environment would have been much like that of modern-day Baffin Island but with lusher vegetation thanks to the prolonged sunlight of its lower latitude.
31. See J. E. Sanders 1962; Shepard 1963.
32. Richards and Judson 1965.
33. Curray 1965.
34. Bumpas and Peirce 1955.
35. Dyke et al. 1999.
36. Milliman, Pilkey, and Ross 1972.
37. L. J. Jackson et al. 2000.
38. See Bothner and Spiker 1980; Emery et al. 1967.
39. S. T. Jackson et al. 1997.
40. Lowery 2002.
41. Whitmore et al. 1967.
42. Hemmings 2004.
43. Westley and Dix 2008.
44. Perennial ice cover: Falkowski and Raven 1997; freeze-thaw of ice pack: D. L. Wilson, Smith, and Nelson 1986.
45. De Vernal and Hillaire-Marcel 2000.
46. T. Webb et al. 1996. We generally agree with this study’s findings, but in our opinion there are problems with the interpretations made from similar studies. We are looking for the edge or maximum extent of permanent ice, which would sometimes shift rapidly and at other times might be stationary for hundreds of years. This is a process rather than an event, and the difficulty of mapping a dynamic process likely accounts for the wide variation in interpretations of the location of the ice edge. During warming phases, as seen in the Greenland ice core, the ice edge retreated north, and during colder periods it extended south. Perhaps the maps do represent the very maximum extent of the LGM ice, but to accept that would mean presuming that the organisms found in deep-sea cores lived at the same spot were the core was taken, and there is little discussion in the paleoceanogaphic literature about taphonomic processes—how things such as dead organisms, bones, etc., become deposited in the sediments where they are found. It seems unlikely to us that these tiny organisms represent the sea surface conditions directly above the core; more probably, they were deposited down-current from their point of origin. This would skew the results in the North Atlantic by overrep-resenting organisms that dwell in warmer water whose remains were carried north and east by the Gulf Current: it would look like the ice edge was farther north than its real location.
47. Productivity estimates are based on the presence of hard-bodied organisms in the sediments on the sea floor. Unfortunately, there is no way to determine the variety and number of soft-bodied creatures that were well adapted to extreme cold temperatures. For instance, scientists have only recently discovered several hundred heretofore unknown cold-adapted species, both hard- and soft-bodied, in both polar seas (Casey 2009). In part based on estimates of LGM productive estimates, Westley and Dix 2008 views our hypothesis as unlikely but concludes that the LGM paleoecological evidence has not been examined in enough detail to support or invalidate it.
48. Sancetta 1992.
49. See Stirling 1997; Gilchrist and Robertson 2000.
50. Chapman and Shackleton 1998.
51. Chapman and Maslin 1999.
52. See, e.g., Butzer 1986.
53. Chapman and Shackleton 1998.
54. Jensen 1944.
55. Dyke et al. 2002.
56. Bond et al. 1992.
57. See, e.g., Broecker 1994.
58. See, e.g., Sancetta 1992.
59. Abrantes 1991.
60. Clottes and Courtin 1996.
61. Fuller 1999.
1. Beyond the personal experience Dennis gained during his years of working in the Arctic, the information on Arctic hunting and survival in this chapter is taken from the following sources: Oozeva et al. 2004; Krupnik 1993; Steinbright 2001; Nelson 1969. We are especially indebted to Richard Nelson, whom Dennis first met on the frozen Beaufort Sea near Point Barrow, Alaska, some forty years ago. We draw on Nelson’s deep knowledge of Inuit ice hunting to provide practical insights into the probability of successful Paleolithic voyages on the ice age oceans.
2. For the latest reconstruction of the extent of sea ice during the LGM, see de Vernal et al. 2006.
3. Modern trash caught in the oceanic gyres is an environmental hazard, but Ebbesmeyer and Scigliano 2009 gives a fascinating account of how human drifters, lost tennis shoes, messages in bottles, and other non-oceanic debris have helped researchers to document their size, velocity, and orbital speed.
4. See Byock 2001 for the cultural conflicts and legal infrastructure related to driftwood in early Icelandic society.
5. Montenegro et al. 2006 suggests that during the LGM westward winds were stronger and eastward winds weaker than present-day winds.
6. D. N. Thomas 2004 is an excellent non-technical source of information on the formation, conditions, and rich biotic habitats of pack ice.
7. Hough 1898.
8. The adventurer Tim Severin, to test the possibility that the sixth century Irish monk St. Brendan had successfully sailed an open skin boat from Ireland to Newfoundland, researched and built such a boat to re-enact the legendary journey. After a false start with a relatively inexperienced crew, he signed on a Faroese fisherman, Trondur, who was well versed in problems related to ice and survival on the North Atlantic. It took only fifty days to make the crossing, and the boat remained in the water for those eight weeks. For additional fascinating details, see Severin 2000.
9. Maggs et al. 2008.
10. See Straus 1979; Straus and Clark 1986.
11. Additional evidence of the Solutrean use of marine and freshwater fish can be found in Morales-Muñiz and Roselló-Izquierdo 2009.
12. For an excellent summary of the early developments of watercraft in the Pacific, see Kirch 2000. See also Erlandson 2001; O’Connor and Veth 2000.
13. Oda 1990. See also A. Anderson 1987.
14. It is likely that Crete was never connected to the mainland, but some researchers assume continental bridges existed from time to time. See Facchini and Giusberti 1992.
15. D’Errico 1984.
16. Fuller 1999.
17. Otte 2002. But cf. Straus 2001.
18. Renfrew and Aspinall 1987.
19. Excavation: Simmons 1988; controversy: Bunimovitz and Barkai 1996.
20. C. M. Smith and Haslett 2000. On these boats’ technical details, see Roberts 1995.
21. Chronicled in the Relación Sámano-Xeréz, cited in Edwards 1965.
22. Bergen, Niekus, and Vilsteren 2002.
23. Zunig 2002.
24. Casson 1994.
25. Aikens and Higuchi 1982.
26. Lee and Robineau 2004.
27. See McCartney and Veltre 1996 for the latest summary of the archaeology of Ananiuliak Island.
28. Dixon et al. 1997.
29. See Orr 1962; J. Johnson et al. 2000.
30. Erlandson et al. 1997.
31. See Erlandson et al. 1996; Erlandson et al. 1999.
32. Holland 1999.
33. S. Loring, personal communication.
34. Loring 1980.
35. See Jodry 2005; Morris and Goodyear 1973; Yerkes and Gaetner 1997.
36. Wheeler et al. 2003.
37. See, e.g., Straus 2000.
38. Krupnik 2005.
39. Chesemore 1975.
40. Lehn 1979.
1. Prudden 1903.
2. Schurr 2004.
3. Reidla et al. 2003.
4. Izagirre and de la Rúa 1999.