AN ICE AGE QUARRY-WORKSHOP the West Athens Hill Site Revisited the UNIVERSITY of the STATE of NEW YORK Regents of the University

Home , Beech is Back, Fisher Site, Flint Mine Hill, Hiscock Site, Kalkberg, Lithic stage

AN ICE AGE QUARRY-WORKSHOP The West Athens Hill Site Revisited THE UNIVERSITY OF THE STATE OF NEW YORK Regents of The University

ROBERT M. BENNETT, Chancellor, B.A., M.S...... Tonawanda ADELAIDE L. SANFORD, Vice Chancellor, B.A., M.A., P.D...... Hollis DIANE O’NEILL MCGIVERN, B.S.N., M.A., Ph.D...... Staten Island SAUL B. COHEN, B.A., M.A., Ph.D...... New Rochelle JAMES C. DAWSON, A.A., B.A., M.S., Ph.D...... Peru ANTHONY S. BOTTAR, B.A., J.D...... North Syracuse MERRYL H. TISCH, B.A., M.A...... New York GERALDINE D. CHAPEY, B.A., M.A., Ed.D...... Belle Harbor ARNOLD B. GARDNER, B.A., LL.B...... Buffalo HARRY PHILLIPS, 3rd, B.A., M.S.F.S...... Hartsdale JOSEPH E. BOWMAN,JR., B.A., M.L.S., M.A., M.Ed., Ed.D...... Albany LORRAINE A. CORTÉS-VÁZQUEZ, B.A., M.P.A...... Bronx JAMES R. TALLON,JR., B.A., M.A...... Binghamton MILTON L. COFIELD, B.S., M.B.A., Ph.D...... Rochester JOHN BRADEMAS, B.A., Ph.D...... New York

President of The University and Commissioner of Education RICHARD P. MILLS

Chief of Staff Counsel and Deputy Commissioner for Legal Affairs KATHY A. AHEARN

Chief Operating Officer Deputy Commissioner for the Office of Management Services THERESA E. SAVO

Deputy Commissioner for Cultural Education CAROLE F. HUXLEY

Director of the New York State Museum CLIFFORD A. SIEGFRIED

Director, Research and Collections Division JOHN P. HART

The State Education Department does not discriminate on the basis of age, color, religion, creed, disability, marital status, veteran status, national origin, race, gender, genetic predisposition or carrier status, or sexual orientation in its educational programs, services, and activities. Portions of this publication can be made available in a variety of formats, including braille, large print or audio tape, upon request. Inquiries concerning this policy of nondiscrimination should be directed to the Department’s Office for Diversity, Ethics, and Access, Room 152, Education Building, Albany, N.Y. 12234. Requests for additional copies of the publication may be made by contacting Publications Sales, Room 3140, State Museum, Albany, N.Y. 12230. AN ICE AGE QUARRY-WORKSHOP The West Athens Hill Site Revisited

by Robert E. Funk New York State Archaeologist (Emeritus) Anthropological Survey, New York State Museum

New York State Museum Bulletin 504 2004

The University of New York The State Education Department © The New York State Education Department, Albany, New York 12230

Published 2004

Printed in the United States of America

Copies may be ordered from: Publication Sales 3140 CEC New York State Museum Albany, New York 12230 Phone: (518) 402-5344 FAX: (518) 474-2033 Web address: http://www.nysm.nysed.gov./publications.html

Library of Congress Catalog Card Number: 2004101578

ISBN: 1-55557-173-5 ISSN: 0278-3355

Cover Illustration: New York State Museum’s “Ice Age Hunters: New York’s First People” life group.

iv EDITORIAL ACKNOWLEDGEMENTS

On September 19, 2002, just a few short days before his untimely and unexpected death, Bob Funk gave Beth Wellman for editing the manuscript that now constitutes this book. In the intervening months, a number of individuals have played important roles in bringing manuscript to publication. Bob’s son Al allowed me to access Bob’s home office to look for notes, photographs, and maps associated with the West Athens Hill site and the manuscript. Beth Wellman worked tirelessly in initial editing, assembling and identifying figures and photographs, and checking proofs during the publication process. David Anderson, Jim Bradley, Jim Petersen, and Art Spiess all provided many useful comments and suggestions on the manuscript. Ted Beblowski scanned and electronical- ly edited the various line drawings and photographs. Jack Skiba checked proofs and managed the publication process. Leigh Ann Smith designed the cover. And, of course, Jim Petersen wrote a wonderful foreword placing Bob’s research at West Athens Hill into a larger context of Paleoindian research in the Northeast. I extend sincere thanks to every- one who made this publication happen. I think Bob would be very pleased.

John P. Hart April 2004

v TABLE OF CONTENTS

List of Figures ...... vii List of Tables ...... ix Foreword ...... xi Prologue...... l Introduction ...... 1 Location and Discovery ...... 4 Regional Geography and Geology ...... 5 Site Description and Setting ...... 10 Research Goals...... 12 Excavation ...... 13 Sedimentary Contexts ...... 15 Summary of Geological History ...... 23 Artifacts from Areas A and B ...... 25 Artifacts from Area C ...... 57 Use-Wear Patterns on Artifacts ...... 71 Artifacts from Areas D, E, and F...... 76 Palynology...... 77 Radiocarbon Dating...... 78 Stratigraphy, Typology, and Intrasite Settlement Patterns ...... 79 Comparisons ...... 89 The Position of West Athens Hill among Northeastern Paleoindian Sites ...... 117 A Speculative Reconstruction of Paleoindian Quarry and Workshop Activities...... 123 General Conclusions...... 125 Aknowledgments...... 132 Appendix: Descriptions of Cherts Commonly Found on Prehistoric sites in Eastern New York...... 133 References Cited...... 136

vi LIST OF FIGURES

1. Map of New York State showing location of West Athens Hill and other Paleoindian sites mentioned in the text...... 2 2. Geologic map of the middle Hudson Valley...... 8 3. Idealized west to east cross-section of bedrock formations in the vicinity of West Athens Hill...... 9 4. Map of the West Athens Hill site, showing major loci (Areas A, B, and C) and excavated areas...... 10 5. View looking west from the summit (Area A) at West Athens Hill, showing the Hans Vosen Kill valley, the Helderberg escarpment, and the Catskill Mountains in background...... 11 6. View of excavations in Area B, looking north...... 13 7. View of excavations in Area A, showing presumed quarry pit, looking north...... 16 8. West profile of section W10N0...... 17 9. West profile of section E20N0...... 17 10. Stratigraphic profiles in Area B...... 18 11. Fluted point uncovered in stratum 1, section E10N20...... 18 12. View of excavations in Area C...... 20 13. Exposed bedrock feature in Area C...... 20 14. Map of the quarried chert outcrop in Area C...... 21 15. Close-up view of the areas on the bedrock mass showing anvil-pitting...... 22 16. View of chert quarry pit in Area F...... 22 17. Archaic and Transitional projectile points and a drill from Areas A and B...... 26 18. Stage 1 bifaces from Areas A and B...... 28 19. Scattergram of width vs. thickness for Stage 1 bifaces from Areas A and B...... 29 20. Scattergram of length vs. width for Stage 1 bifaces from Areas A and B...... 29 21. Stage 2 bifaces from Areas A and B...... 30 22. Scattergram of width vs. thickness for Stage 2 bifaces from Areas A and B...... 31

vii 23. Stage 3 bifaces from Areas A and B...... 33 24. Scattergram of width vs. thickness for Stage 3 fluted point preforms from Areas A and B...... 35 25. Scattergram of length vs. width for Stage 3 fluted point preforms from Areas A and B...... 35 26. Scattergram of width vs. thickness for Stage 3 non-fluted point preforms from Areas A and B...... 36 27. Fluted points from Areas A and B...... 37 28. Scattergram of width vs. thickness for Stage 4 fluted points from Areas A and B...... 38 29. Fluted point fragments and preforms from Areas A and B...... 39 30. Scattergram of width vs. thickness for probable fluted point fragments from Areas A and B...... 40 31. Miscellaneous bifaces from Areas A and B...... 41 32. Scattergram of width vs. length for miscellaneous bifaces from Areas A and B...... 42 33. End scrapers from Areas A and B...... 44 34. Side scrapers from Areas A and B...... 45 35. Large side scrapers from Areas A and B...... 46 36. Flake knives from Areas A and B...... 49 37. Chipped and rough stone tools from Areas A and B...... 50 38. Rough stone tools from Areas A and B...... 51 39. Chipped and rough stone tools from Area C...... 58 40. Map showing distribution of artifacts in Area C...... 59 41. Scattergram of width vs. thickness for Stage 1 bifaces from Area C...... 60 42. Scattergram of length vs. width for Stage 1 bifaces from Area C...... 60 43. Scattergram of width vs. thickness for Stage 2 bifaces from Area C...... 61 44. Scattergram of width vs. thickness for Stage 4 fluted points and Stage 3 preforms from Area C...... 64 45. Artifact clusters possibly representing localized activity concentrations in Area B...... 83 46. Bar graph comparing the relative importance of the hypothesized activities at Paleoindian sites in central and eastern New York ...... 112

viii LIST OF TABLES

1. Summary statistics for Stage 1 bifaces from Areas A and B...... 27 2. Summary statistics for Stage 2 bifaces from Areas A and B...... 27 3. Summary statistics for fluted point preforms from Areas A and B...... 31 4. Summary statistics for Stage 3 bifaces other than fluted point preforms from Areas A and B...... 32 5. Metric and non-metric data for finished Stage 4 fluted points from Areas A and B...... 34 6. Summary statistics for Stage 4 fluted points from Areas A and B...... 36 7. Summary statistics for probable fluted point fragments from Areas A and B...... 38 8. Summary statistics for end scrapers from Areas A and B...... 42 9. Summary statistics for side scrapers from Areas A and B...... 43 10. Summary statistics for flake knives from Areas A and B...... 47 11. Trait list for chipped stone tools in the R. Arthur Johnson collection...... 48 12. Weights of a sample of hammerstones from Areas A and B...... 48 13. Trait list for Areas A and B...... 52 14. Frequencies of lithic materials used for artifacts other than Normanskill chert from Areas A and B...... 53 15. Summary statistics for block fragments from Area B...... 54 16. Summary statistics for a sample of cores from Area B...... 55 17. Summary statistics for core fragments from Area B...... 55 18 Frequencies of debitage classes from Area B...... 56 19. Summary statistics for Stage 1 bifaces from Area C...... 57 20. Summary statistics for Stage 2 bifaces from Area C...... 61 21. Summary of data for finished (stage 4) fluted points and Stage 3 preforms from Area C...... 63 22. Summary statistics for Stage 4 fluted points and Stage 3 preforms from Area C...... 64 23. Metrical data for possible fluted points in process from Area C...... 64

ix 24. Summary statistics for end scrapers from Area C...... 65 25. Summary statistics for side scrapers from Area C...... 65 26. Summary statistics for samples of quarried chert blocks from Area C...... 67 27. Summary statistics for a sample of block fragments from Area C...... 67 28. Summary statistics for a sample of cores from Area C...... 68 29. Summary statistics for a sample of core fragments from Area C...... 68 30. Weights of hammerstones from Area C...... 69 31. Trait list for Area C...... 70 32. Depth distribution of general artifact categories in Areas A and B...... 80 33. Depth distribution and frequencies of artifact types with provenience data from Area B and artifact lists from Area A, C and the surface...... 81 34. Comparisons of frequencies of artifact classes in Area B between the cluster groups and between the combined clusters, Area C, and the site as a whole...... 86 35. Comparisons of artifact types in assemblages from Paleoindian sites in central and eastern New York...... 99 36. Comparisons of fluted point size and weight between sites...... 100 37. Comparisons of the size and weight of Stage 1 and Stage 2 bifaces...... 101 38. Comparisons of end scraper size and weight between sites...... 102 39. Comparisons of side scraper size and weight between sites...... 103 40. Artifact to debitage ratios and percentages for Paleoindian assemblages...... 104 41. Measurements of Paleoindian site productivity...... 105 42. Frequencies of artifacts made of non-local cherts in chipped stone assemblages from Paleoindian sites in central and eastern New York...... 106 43. Types of exotic lithic material in the chipped stone assemblages in central and eastern New York...... 107 44. Frequencies of exotic lithic material by artifact types in the chipped stone assemblages in central and eastern New York...... 108 45. Wear patterns on chipped stone artifacts from Paleoindian sites in New York and New England...... 109-10 46. Relative importance of activities at Paleoindian sites in central and eastern New York inferred from the frequencies of stone tools...... 113 47. Summary of data for Paleoindian sites mentioned in the text...... 119-22

x FOREWORD: WEST ATHENS HILL, THE PALEOINDIAN PERIOD, AND ROBERT E. FUNK IN NORTHEASTERN PERSPECTIVE

By James B. Petersen

INTRODUCTION region, with his mentor and predecessor as New York State Archaeologist, William A. Ritchie (Ritchie and Funk 1973), were both It is a distinct honor to write the foreword published by the NYSM. for this highly significant publication on the I have recently heard from my colleague West Athens Hill Paleoindian archaeological Jim Bradley that Bob was planning to under- site, a report by the late Robert E. Funk, take a broader systematic review of all avail- Emeritus State Archaeologist of New York. I able Early Paleoindian (fluted point) period knew Bob for more than 25 years and he had a evidence in the Hudson Valley. Pieces of this very substantial impact on my early develop- project were already done. Sadly for all of us, ment as an archaeologist, as he did for many Bob did not live long enough to see this bigger others, avocational and professional archaeol- project through to fruition, and it would have ogists alike. Bob’s high energy, positive atti- been undoubtedly another highly significant tude, good humor, and generosity are leg- scholarly contribution to regional, national, endary. He had a lot more research that he and international archaeology on a level well wanted to get done at the time of his prema- beyond New York State. ture passing, including a few old projects and This foreword has been designed to serve various new ones as well. Nonetheless, Bob several distinct purposes. First, I present a has left this world a better place because of his personal overview of the Paleoindian period presence, kindness, and many works. (and presumed antecedents) in northeastern Bob Funk had finished the bulk of the North America, or the “Northeast,” to contex- work needed to see his reanalysis and expand- tualize recent literature and perspectives that ed reporting of the West Athens Hill site ready Bob did not necessarily have the time (or incli- for publication prior to his untimely death. nation) to include in his publication on West Nearly everything had been completed for the Athens Hill. Secondly, I briefly summarize the present volume as another in a series of influ- significance of the West Athens Hill site and ential archaeological publications by the New its importance to Paleoindian studies in gen- York State Museum (NYSM). Bob’s landmark eral. Finally, in closing I present a brief per- volume on Hudson Valley prehistory (Funk sonal tribute to Bob and Bob’s spirit. 1976) and the equally legendary volume con- The reader is left to judge how well these cerning Native American “settlement pat- goals are achieved, but if nothing else, there terns” in New York State and the broader

xi can be no doubt about the significance of Bob Settlement Patterns in the Northeast, including Funk’s Paleoindian (and other) archaeological the first substantial reporting of West Athens research in New York State. This is just one Hill by Funk (pp. 9–36). The same might be more reason to celebrate Bob’s life and his said when Bob (Funk 1976:205–229) presented notable contributions to the study of archaeol- an extensive overview of Paleoindians in New ogy related to northeastern Native Americans. York State and elsewhere, as well as his (Funk 1978:16–19) broader overview of early regional prehistory in the Handbook of North THE PALEOINDIAN PERIOD IN American Indians. NORTHEASTERN NORTH At the time, just a few major Paleoindian AMERICAN PREHISTORY: PROCESS sites were known regionally (e.g., Barnes, Bull AND PARTICULARS Brook, Debert, Holcombe, Plenge, Potts, Reagen, Shoop, and Williamson). While still Background and Terminology important for understanding Paleoindian variability, these early studied sites represent As with various other areas of North a very small sample. Moreover, at the time American archaeology related to Native there was very little sense of internal Americans (or “First-Nations people” in Paleoindian chronology, just a handful of Canada), the study of the Paleoindian period questionably appropriate radiocarbon dates, has been practically revolutionized in the past and essentially no subsistence data at all to 25–30 years. A significant and substantial speak of for the region. Comparative number of these gains have been made in the researchers had to look at evidence from far “Northeast,” which is defined here as the area afield in their Paleoindian syntheses, includ- covered by Volume 15, Northeast, of the ing the Plains and Southwest areas. Handbook of North American Indians, broad- Now, 30 years later the Northeast has pro- ly speaking a “culture area” (Trigger 1978). duced more substantial Paleoindian evidence The Northeast thus roughly extends from that is of national and international signifi- about the Mississippi River eastward to the cance, leading many researchers from beyond Atlantic coast on an east–west axis, and from the region to look here for help in understand- about the mouth of the Ohio River near the ing Paleoindians generally. Evidence from Kentucky–Tennessee border and North both the American and Canadian portions of Carolina–Virginia border northward to the the Northeast has become important, even southern margins of the Canadian Shield on a essential, for understanding Paleoindians on a north–south axis. The Plains culture area bor- continental scale, along with broader issues ders the Northeast to the west, while the such as the peopling of the Western Subarctic and Southeast bound it to the north Hemisphere during the late Pleistocene epoch. and south, respectively. In part, this “revolution” is reflected in com- In fact, it is arguably the case that archaeo- parison of fluted point distribution studies logical and paleoenvironmental research over over the past 20–30 years or so, reflecting how the past 25–30 years have transformed our col- the sheer numbers of fluted point discoveries lective knowledge of the Paleoindian period have increased in the Northeast and else- proportionally more in the Northeast than in where too (cf. Anderson and Gillam 2000; any other part of North America. The Brennan 1982). But distribution studies only Northeast is no longer poorly known in terms tell part of the story, since the quality of of the Paleoindian period generally, at least Paleoindian characterization and contextual not in relative terms. The case was rather dif- data have been greatly enhanced in the ferent in 1973 when Ritchie and Funk pub- Northeast and the contextual data have lished their seminal volume, Aboriginal enabled great strides in different areas, as

xii described below. In correlation with better ability, not so long ago only a single common- contextual data, we have gained significant ly used designation, the “Paleoindian period,” new information about absolute dating of was recognized, and this was generally equiv- Paleoindian sites in the Northeast and unfor- alent to what we can now call the Early tunately, new knowledge about dating prob- Paleoindian period based primarily on diag- lems too. Advances in regional dating are nostic fluted points, given evidence then especially striking when data from the far available. Moreover, a long occupational hia- Northeast are compared to the paucity of tus during the early-middle Holocene was radiocarbon–dated Paleoindian sites in most once suggested for the immediate post–fluted other portions of North America. For example, point period in northern portions of the the Southeast and even the Midsouth and Northeast due to environmental restrictions, Midwest are still very poorly dated for the ca. 10,000–5000 B.P., but this has been since Paleoindian period (e.g., Goodyear 1999; refuted. Recent identification of Late Lepper 1999). Paleoindian and later (“Archaic”) evidence Notable Paleoindian and other early sites has filled in the putative hiatus, but not every- in the broad Northeast now include a few one fully agrees on the details (cf. Bradley dated stratified sites such as Meadowcroft 1998; Funk 1977, 1978, 1993, 1996; Mason 1981; Rockshelter, Cactus Hill, Shawnee Minisink, Petersen 1991; Petersen and Putnam 1992; Templeton, and Thunderbird. Stratified sites Robinson 2001). Brief discussion of temporal of this antiquity are very rare, however, and periods both before and after the Early instead we must typically depend on the iso- Paleoindian period is included here, as Bob lation of early components in shallow, partial- Funk briefly mentions in the present publica- ly disturbed contexts. Some of the most tion. Of these, the earliest, preceding period is notable recent discoveries of the latter type by far the most tentative. include Adkins, Arc, Cummins, Fisher, The “Pre-Clovis” period is conjectural for Hiscock, Israel River, Lamb, Nobles Pond, many analysts, and elsewhere it has been Parkhill, Rimouski, Thedford, Vail, and called the “Early Lithic Stage” (or period) by Varney Farm. These and other sites now allow Bob Funk, following broad precedent (Funk suggestion of a three–four part sequence for 1993:142–143), among various other names. the span of the Early Paleoindian period, ca. The “Pre-Clovis” period as used here applies 11,000–10,000 B.P., or the time characterized to still rare and enigmatic, but now seemingly by diagnostic fluted points of one type or incontrovertible, evidence for early human another. Likewise, at least a two–three part presence in the Northeast, at least in souther- Late Paleoindian period, ca. 10,000–9000/ ly areas, before the Paleoindian period proper. 8500 B.P., has been defined, in correlation with Recognition of a “Pre-Clovis” period in the unfluted lanceolate points. In addition, some Northeast will be considered heresy by some, researchers now recognize a “Pre-Clovis” but available evidence requires recognition of period (or another terminological equivalent) some sort of cultural manifestation before the that preceded the advent of fluted point man- time of fluted points, but it is not more for- ufacture before 11,500–11,000 B.P. and thus, mally named here. Thus, there were at least the Early Paleoindian period per se. (These two and likely three recognizable temporal and all other radiocarbon dates are left uncal- subdivisions of early prehistory in the ibrated here, due to a lack of an absolute cali- Northeast before the subsequent “Archaic” bration convention, the fact of date compres- and “Woodland” periods (Funk 1993:134-140), sion for some portions of the overall period, dated ca. 9000/8500–3000 B.P. and ca. and to allow direct comparison with dates 3000–500/400 B.P., respectively. reported by Funk.) Obviously, we need to be primarily con- Contrasted with present recognition of cerned here with early prehistory as it pertains chronological subdivisions and temporal vari- to the context of West Athens Hill. Besides

xiii broad chronological periods, more substantial Paleoindian life ways, namely subsistence, Paleoindian evidence from the Northeast now mobility, and aspects of technology, among helps us refine other aspects of Paleoindian others. Changing paleoenvironments were lifeways per se, filling in some but not all of formative within the context of late the details. These data include new recogni- Pleistocene and early Holocene climactic fluc- tion of both broad and regional technological tuations. Modern conditions became much data on different scales, regional variations in more typical after the onset of the Holocene. diagnostic artifacts, long-distance mobility Greater stability pertained thereafter (Carr and/or interaction (of some sort), and other and Adovasio 2002:2–3; Chapdelaine and issues discernable primarily through the LaSalle 1995; Custer 1996:97–100; Dent study of stone (“lithic”) artifacts, long the hall- 2002:68–71; Dineen 1996; Eisenberg mark of Paleoindian studies in general. New 1978:17–35; Ellis and Deller 1990; Funk settlement pattern and subsistence data have 1976:207–212, 1993:43–62; Karrow and Warner been accrued for the Northeast over the past 1990; Petersen and Putnam 1992:19–20; Ridge 25–30 years, though these are still very limit- 2003; Storck 1988b:261–263; Tankersley ed. I go on to these and other matters, draw- 1996:21–22). ing from past personal syntheses dealing with Paleoenvironments in the Northeast pro- Paleoindian developments and variability vide an important backdrop for Paleoindian (e.g., Petersen 1995:208–214; Petersen, studies, but they are not detailed here. We still Bartone, and Cox 2000), along with the work need to guard against the simple assumption of various colleagues, selectively cited, and that hunter-gatherers strictly live at the mercy broadly updating these as appropriate. of local and broader environmental contexts, especially if in doing so we downplay or eliminate social contexts. There is a natural incli- Paleoenvironments and Paleoindians in the nation toward environmental determinism in Northeast modeling Paleoindian lifeways, but we need To begin, I need to again strike a caution- to address this, the most “sacred” of “sacred ary note concerning Paleoindian studies in cows.” Otherwise we run the risk of denying general and those in the Northeast in particu- any role for the historical, contingent, and lar. As with all archaeological research, we functional aspects of social interaction within need to continually critique various interpre- the broader realm of Paleoindian behavior, a tations derived from our ongoing research point that is returned to below in the discus- and continually refine our perceptions, in sion of Paleoindian lithic resource acquisition. spite of many recent promising discoveries The Early Paleoindian period, and pre- (e.g., Gramly and Funk 1990:25–26; Ritchie sumably the “Pre-Clovis” period before it, and Funk 1984:2–3). For example, we might witnessed the first human settlement across ask again, How well do we understand the northeastern North America, and these pio- origins of Paleoindians and the Paleoindian neering circumstances surely complicate the period in general? What can be said about its story. Likewise, the typical assumption of very duration and variability in time and space? simple, almost “primordial” Paleoindian Can it be profitably subdivided, spatially and social systems in all portions of the Northeast temporally? What do we think we know and (and elsewhere), derived in part from general what do we really know about settlement and “ethnographic analogy,” may well underesti- subsistence regionally? What do we know mate these people and the nature of their life- about other cultural “subsystems” such as ways. mortuary patterns, etc.? To be explicit, I am not saying that we Paleoenvironmental conditions certainly should avoid or throw away ethnographic influenced or even determined some details of analogy-in fact, quite the contrary-but for

xiv these and other reasons, we need to proceed because of oft-disturbed archaeological con- cautiously as we look beyond the available texts. In addition, the recently recognized evidence to interpret and reconstruct various problem of date compression during one or aspects of Paleoindian lifeways. However, all more portions of this temporal span (e.g., of our models probably represent dangerous ca.10,600–10,200 B.P.) means that full temporal simplifications without substantive corrobo- resolution of chronological subdivisions by ration. As many would admit, it is certainly radiocarbon dating may be impossible difficult to generalize spatially and temporal- (Curran 1996:5; Fiedel 1999; Goodyear ly for most aspects of the Paleoindian period. 1999:434–435; Levine 1990; Spiess, Wilson, and A significant amount, perhaps most, of the rel- Bradley 1998:238; Wilson and Burns 1999:233). evant evidence has been likely effaced by The Early Paleoindian period is much bet- environmental and cultural factors over time, ter dated than the Late Paleoindian period and we must interpret the available evidence regionally. Most analysts now recognize dis- within the context of very limited, incomplete, tinctive Late Paleoindian period manifesta- and biased samples. tions in different portions of the region, unlike It seems that the Northeast witnessed only 20 years ago. The putative “Pre-Clovis” broad regional patterning as a more or less uni- period that preceded the Paleoindian period tary (and contemporaneous?) set of proper is much less clear relative to both the Paleoindian developments, at least to begin Early and Late Paleoindian periods regionally. with, where we have sufficient evidence for Nonetheless, significant gains have been comparison. Such broad uniformity is some- made here too, depending on one’s views times taken as evidence that the “Clovis” flut- about acceptance of “Pre-Clovis” evidence ed point makers and the Early Paleoindian generally (Adovasio 1998; Adovasio, period represent the earliest human occupation Boldurian, and Carlisle 1988; Adovasio, et al. of North America and not the successor of 1998; Adovasio with Page 2002; Goodyear some sort of “Pre-Clovis” occupation(s) (e.g., 1999:435–436; Goodyear and Steffy 2003; Kelly and Todd 1988; Tankersley 1998a). McAvoy and McAvoy 1997:176–181, Figures Alternatively, the Clovis–type fluted point and 5.65, 5.68, and 6.1, Table 6.7). Problems inher- its broader technological repertoire may just ent in radiocarbon dating and the difficulty of represent the first in a long series of wide- isolating cultural remains in reliable, single- spread recognizable archaeological “horizon component contexts continue to plague styles,” with Clovis arising out of earlier, still chronology building for all of these periods, somewhat amorphous “Pre-Clovis,” occupa- however. tion(s). Although not necessarily as widespread Through the efforts of R. Michael Gramly, as Clovis, many other horizon styles, including Vance Haynes, James Bradley, Arthur Spiess, later Paleoindian ones, variably characterized and others, the chronology of the Early all subsequent Native American regional and Paleoindian period in the Northeast has been continental prehistory, until the arrival of increasingly better resolved (Curran 1996; Europeans and other intruders historically. Haynes, et al. 1984; Levine 1990; Spiess, Wilson, and Bradley 1998:232–238; Tankersley, et al. 1997:Table 3). Some analysts have sug- Paleoindian Dating and Chronology gested a single, generally uniform chronologi- Absolute dating and related chronology cal sequence for the Early Paleoindian period building for the Paleoindian period have wit- all across the Northeast (e.g., Gramly and nessed significant progress over the past few Funk 1990:21–22). Others recognize subre- decades in the Northeast. Apparently, we will gional differences. For example, Bradley, never have a plethora of reliable absolute Spiess, and others, using a combination of dates for Paleoindian sites and site samples absolute dates and inference, have produced a

xv three–four part chronological sequence for the began by ca. 11,000 B.P. in the Northeast prop- Early Paleoindian period, specifically for New er. It seemingly did not begin much earlier, if England and the adjacent Maritime Provinces. at all, before this, in spite of what some have This sequence includes the Bull Brook–Debert claimed (cf., Bonnichsen, Keenlyside, and /Vail–Neponset/Michaud–Nicholas phases Turnmire 1991; Bonnichsen, et al. 1985; Curran and point types (Bradley 1998; Spiess, Wilson, 1996:4; Custer 1996:94–97, 2001:69–71; Gramly and Bradley 1998:Table 6). and Funk 1990:6; Haynes, et al. 1984; A generally similar sequence for the Early Keenlyside 1991:164–166; Spiess, Wilson, and Paleoindian period, again with three (or four) Bradley 1998:232–238). This starting date of ca. temporal subdivisions, was previously estab- 11,000 B.P. is slightly later than what most rec- lished by Chris Ellis, Brian Deller, Peter ognize as reliable dates for the earliest Storck, and others for the lower Great Lakes. “Clovis”-related dates in the Plains and far- This sequence includes the Gainey– ther south and west, although the inception of Parkhill–Crowfield phases and Gainey– Clovis is argued about there too. Barnes–Crowfield point types. However, it Terminal dating of the Early Paleoindian has yet to be reliably dated in Ontario and (fluted point) period in the Northeast is a bit Michigan, and most other nearby areas (e.g., confused by available dates for fluted points Curran 1996:3–4; Deller and Ellis 1992; Ellis and the succeeding unfluted lanceolate points and Deller 1990, 1997, 1998; Lepper 1999:370, of the Late Paleoindian period. However, the Table 3; Simons 1998; Storck 1998). The three last fluted points were seemingly made by ca. phases are tentatively dated: Gainey, ca. 10,200 B.P.–10,000 B.P., and this seems a rea- 11,000–10,700 B.P.; Parkhill, ca. 10,700–10,500 sonable estimate for the termination of the B.P.; and Crowfield, ca. 10,500–10,400 B.P., or Early Paleoindian period in general (Bradley later. Still other partially fluted and unfluted 1998; Curran 1996; Custer 2001:71; Petersen points (e.g., Holcombe) seemingly followed 1995:208; Spiess, Wilson, and Bradley 1998; cf. Crowfield, after ca. 10,400–10,200 B.P., until ca. Tankersley 1996:22, Figure 2.2). 10,000 B.P. (Curran 1996; Deller and Ellis Not all of the available dates for the Early 1988:255-258; Tankersley, et al. 1997). Paleoindian period in the Northeast are equal- Notably, a selected average of three dates ly reliable. Many individual dates and even from the Paleo Crossing site in northeastern suites of dates are potentially rejected by at Ohio places the earliest style, Gainey, at ca. least some analysts under careful scrutiny 10,990 + 75 B.P., but this is somewhat equivo- (e.g., Brose 1994:64–65; Curran 1996:4–7; cal due to other, considerably older dates for Grimes 1979:113; Levine 1990). Thus, we must the same context there (Brose 1994:64–65; be ever vigilant about the temptation toward Lepper 1999:382). At the Arc site in western “radiocarbon literalism,” and erroneous and New York, Gainey/Clovis lithics have been suspect absolute dates must be separated from broadly dated older than 10,370 + 108 B.P. reliable ones in building local and regional (average) and younger than 11,700 + 110 B.P., chronologies for the Paleoindian period, as but this is an indirect and imprecise associa- well as others. tion (Tankersley, et al. 1997). In any case, the For example, this situation pertains to a three–four part sequence has direct relevance well-known date of 12,530 + 370 B.P. obtained for West Athens Hill, as discussed below for caribou remains from the same stratum as where the first and perhaps the second subdi- an Early Paleoindian fluted point and other vision of the Early Paleoindian period, with late Pleistocene fauna at Dutchess Quarry Gainey and Barnes (Cumberland) types recog- Cave no. 1. Located near the Walkill River in nized by Bob Funk in the present publication. southernmost New York State, this putative Looking more broadly at the earliest dates association was identified during the 1960s for the Early Paleoindian period, it minimally (Funk, Fisher, and Reilly 1970; Funk, et al.

xvi 1969; Kopper, Funk, and Dumont 1980). At recognize the earliest northeastern fluted one time, it was used to tentatively suggest points as being quite close to western Clovis that the Early Paleoindian period began earli- points on the basis of technology and stylistic er in the Northeast than in other areas of form. Clovis points are sometimes even seen North America (e.g., Funk 1976:210–211, pan-continentally (Gardner and Verrey 1979; 1978:16; Snow 1980:121,135). Justice 1987:21, Map 1; Tankersley 1994:95–96, The Dutchess Quarry Cave association has 1998a, 1998b:10; Tankersley, et al. 1997). For been (and sometimes still is) erroneously used example, the Gainey phase and related point to help establish the beginnings of early fluted type, ca. 11,000–10,700 B.P., is a Clovis-related point manufacture across the continent. To his development in southern Ontario and nearby credit, Bob Funk was long suspicious of this areas, as is the Bull Brook phase and type in association and his later reanalysis, along with New England, though perhaps both are slight- paleontologist David Steadman, clearly estab- ly later (Bradley 1998; Curran 1999; Deller and lished that the Dutchess Quarry association Ellis 1988; Spiess, Wilson, and Bradley 1998). was not a reliable one. Other caribou, extinct Unless there was some sort of “Pre- peccary, and extinct giant beaver bones from Clovis” occupation in the Western Dutchess Quarry caves no. 1 and no. 8 were Hemisphere, the development of the ultimately assayed with 11 AMS dates between Clovis–type fluted points must have occurred 13,180 + 70 B.P. and 11,670 + 70 B.P. during something like the “next year” after the ances- Steadman and Funk’s reanalysis. It now seems tral Natives arrived in North America (as I obvious that all of these, and perhaps others, jokingly tell my students). Clovis points do were deposited well before the fluted point, not reliably occur on the Siberian side of and the original date has been rejected for the Beringia, although lanceolate bifaces and core Barnes-like (or “Cumberland”) fluted point and blade industries are known in eastern (Funk and Steadman 1994:53–55, 73). Siberia at some earlier time (Derev’anko Regardless of the precise details in the 1998:221–286, Figures 180–191; Mochanov and Northeast, most analysts in the West agree Fedoseeva 1996:Figures 3.1–3.6; West that the Early Paleoindian Clovis type was 1996:553–556, Figure 12.2). first made as early as ca. 11,500 B.P. and it last- For some analysts these Siberian bifaces ed until ca. 10,900 B.P. in the southern Plains and core and blade forms are linked with early and Southwest. After ca. 10,900 B.P., the remains in the Western Hemisphere, as at “Folsom” type replaced the Clovis type in the Meadowcroft Rockshelter. An unfluted lance- West. Folsom is a smaller, gracile, and more olate biface and blades date between and ear- fully fluted type that never occurred in the lier than, respectively, 12,800 + 870 B.P. and Northeast, or only on its far-western margins. 11,300 + 700 B.P. at Meadowcroft (Adovasio Parenthetically, the Barnes point type of the 1998; Adovasio, Boldurian, and Carlisle Parkhill phase in Ontario and nearby, includ- 1988:50–52, Figures 1–6; Adovasio with Page ing New York State, may be broadly related to 2002:154–160; Carr and Adovasio 2002:7–13). the Folsom type. In any case, most reliable Surely, some amount of time passed between dates related to Clovis–type fluted points clus- the time Native Americans first set foot on ter in the period ca. 11,200–10,900 B.P. on the modern–day North America and the inven- southern Plains (Boldurian and Cotter tion of Clovis–type fluted points and a broad- 1999:118; Collins 1999:40; Haynes, et al. 1984; er technological inventory. The question is just Holliday 1997:177; Justice 1987:17–21, 27–30; how much time passed; was it effectively only Stanford 1991:2, 1999:289-295, Table 2; a short time, or was there a substantial “Pre- Tankersley 1998a). Clovis” period, some longer time between the For some analysts, “Clovis”–type points arrival of the first people and the appearance never occurred in the Northeast, but others of Clovis–type fluted points?

xvii Where bona fide, the presence of “Pre- scene of both contemporaneous Late Clovis” sites at the extremities of the Paleoindian developments, toward the Lake Northeast and just beyond suggests that Ontario–St. Lawrence drainage in the north, Early–Paleoindian–period fluted point mak- and very earliest Early Archaic evidence to the ers were not the first colonists of the whole south in the Susquehanna Valley and near the region, at least not in the middle and souther- coast. ly, non-glaciated portions. Obviously, howev- Distinctive elongated lanceolate, parallel- er, early “Pre-Clovis” people could not have flaked, and unfluted Late Paleoindian projec- lived in the far northern and eastern portions tile points and related remains only rarely of the Northeast such as much of New York occur in northern New York State. They are State until after the late Pleistocene ice vari- known from at least the middle Hudson ably ablated and water barriers came and Valley northward and across the state to the went, generally from south to north. Local west and northwest, including at least one ablation of late Pleistocene ice apparently such isolated find rather close to West Athens occurred after ca. 15,400 B.P. at Newburgh, Hill (Funk and Schambach 1964:90). They also New York, in the lower Hudson valley, ca. occur in northern New Jersey and elsewhere 12,600 B.P.–12,500 B.P. near Albany, and by nearby (e.g., Kraft 1973). Early Archaic 11,700–11,400 B.P. near the international bor- remains were seemingly directly coeval with der between Vermont and Quebec (Ridge Late Paleoindian remains before 9000 B.P., 2003:33–36, Figure 3.8; Snow 1980:102–109). occurring near the coast at Staten Island, in the The first recognizably widespread occu- lower Hudson, and in the New York portions pants of much of the region, including New of the Susquehanna and Allegheny valleys York State, only occurred with the advent of and southward (Funk 1991:Table 1, 1993:175, the Early Paleoindian period by about 11,000 180–188; Ritchie and Funk 1973:38–39). B.P. Extensive areas of southern Canada were Bill Ritchie (1953, 1957, 1969:17–19) tenta- still uninhabited by people during the Early tively identified Late Paleoindian remains Paleoindian period, however, apparently early during his Paleoindian research. By the because of continuing environmental restric- 1950s and 1960s, Ritchie had recognized scant tions. Earlier “Pre-Clovis” occupation was Late Paleoindian evidence in mixed contexts physically possible in the more southerly, non- at the Brewerton site in upstate New York, glaciated portions of the Northeast and other among multiple Paleoindian components at areas that were deglaciated early on, depend- the Reagen site in Vermont, and elsewhere ing on local environments and potential food nearby, as in Ontario and New Hampshire. sources. Such evidence has yet to be reliably Additionally, Funk, along with a few others recognized anywhere to the north and east of (e.g., Kraft 1973; Mason 1962; Prufer and Baby southwestern Pennsylvania where Meadow- 1963), long ago recognized Late Paleoindian croft is located. remains. These investigators based this attri- The poorly known Late Paleoindian peri- bution on a small number of generally isolat- od clearly developed out of and followed the ed finds of distinctive projectile points region- Early Paleoindian period upon abandonment ally. Like others, Bob suggestively postulated of fluted point manufacture, beginning some- a northern orientation toward the Great Lakes time around 10,200–10,000 B.P., but only in the for the Late Paleoindian finds (Funk 1976: northernmost portions of the Northeast. 228–229, 1978; Funk and Schambach 1964:90; Elsewhere to the south, the Early Paleoindian Ritchie and Funk 1973:7). Available samples period was directly followed by a brief Late were so insubstantial, however, that they were Paleoindian period or it went directly into the afraid to generalize about their discoveries. Early Archaic period, as described further The Late Paleoindian period has been below. New York State seemingly was the dated at just a few sites in the Northeast. In its

xviii briefest span, where it is represented at all, as by bifurcate points later on. The Early Archaic probably pertains in New York State, the Late also saw the onset of “ground stone” tools and Paleoindian period spanned a few hundred other new technological and subsistence inno- years just before or after 10,000 B.P., as noted vations in some cases. Some spatial overlap above. However, in more northerly areas the likely occurred between Late Paleoindian and Late Paleoindian period surely extended from very early Early Archaic remains within a about 10,000 B.P. to 9000 B.P. or perhaps later broad east–west zone across the Northeast. (Doyle, et al. 1985; Ellis and Deller 1990:54–55; This zone of overlap seemingly included Jones 2003; Petersen, Bartone, and Cox northern Ohio, southernmost Ontario, por- 2000:115–119; Stanford 1991; Wright 1995). tions of New York State, northern New Jersey, Some have argued that it lasted as late as 8000 Pennsylvania, and elsewhere (Ellis and Deller B.P., at least in the most northern and eastern 1990; Funk 1977, 1978, 1991, 1993; Lantz 1984; portions of its distribution, as on the Gaspe in Lepper 1999:377–380; Morse 1998; Petersen, Quebec and later still to the north (Chalifoux Bartone, and Cox 2000:Figure 1; Tankersley 1999; Chapdelaine 1994; Dumais 2000; 1996). Dumais, La Roucher, and Poirier 1996; Wright The Early Archaic period per se endured 1995). from ca. 10,000 B.P. (or 9000 B.P.) to ca. 8000 Regardless of when the Late Paleoindians B.P. (or 7500 B.P), with dating variation ultimately disappeared and were replaced depending on regional (and investigator) dif- by/developed into something else, they were ferences. The subsequent Middle Archaic (ca. ultimately followed everywhere south of the 7500–6000 B.P.) and Late Archaic (ca. Subarctic (Canadian Shield) by one Archaic 6000–3000 B.P.) periods constitute the balance manifestation or another, using different proof the preceramic period in the Northeast jectile point forms. Most typically, bifurcate (Funk 1993:175–198; Lepper 1999; Petersen base points or an equivalent point type pro- 1991, 1995:214–222; Petersen and Putnam vide the earliest unequivocal evidence of the 1992; Ritchie and Funk 1973:37–51; Robinson Early Archaic period, dated ca. 8500–8000 B.P., 1992, 2001). I should reiterate that the “Pre- in the northern areas where the Late Clovis” and Late Paleoindian periods, as con- Paleoindian period pertained (e.g., Thomas ceived of here, were at best enigmatic and lit- 1992). Within the northern areas of the tle known or completely unknown in the Northeast, the Early Archaic period only Northeast as recently as 25–30 years ago, emerged roughly within the time span from depending on whom one checks. Thus, these 9000 to 8000 B.P. In contrast, the Early Archaic chronological units (and most others) are at period clearly emerged much earlier, by ca. best “works in progress.” 10,000 B.P., in more southerly and westerly portions of the Northeast such as in the Mid- Paleoindian Lithic Technology, Lithic Atlantic, Pennsylvania, and the Midwest. Very Analysis, and West Athens Hill little, if any, evidence of the Late Paleoindian period, as defined here, is known in the latter Not surprisingly, we are on more secure locales. ground when examining Paleoindian technol- As was apparently the case in New York ogy, specifically lithic tool technology, given State, fluted points in more southerly areas of the durability of these non-perishable the Northeast and farther south were followed remains. However, what about the perishable directly by what is generally recognized as the industries related to wood and other plant earliest of the Early Archaic evidence. These fibers, skins and furs, and bone and antler? included Dalton (alternatively said to be Late Beyond the lithic tools, what was the full Paleoindian or Early Archaic), Kirk, Palmer, nature of Paleoindian hunting gear? Did the and other point types, which were followed Paleoindians have any accessory travel equip-

xix ment such as snowshoes, sleds, or toboggans? product of inference and conjecture. Some What did their houses consist of, how big critics would have a field day with the under- were they, and were they variable across dif- lying assumptions represented in this won- ferent environments of the Northeast, from derful diorama (e.g., male tool knapper and the southern areas such as the Ohio Valley and hunter, woman processing the caribou, small the Mid-Atlantic to northern ones such as size of the group, etc.). However, no matter New York State, New England, and around the coarse, inferential nature of our recon- the Great Lakes? We have few answers yet, structions, we should not be stultified by the but ethnographic analogy can be judiciously lack of direct evidence in making such public employed here, as for interpreting West interpretations, and our models obviously Athens Hill. need to include both evidence and imagina- I remember having my breath taken away tion. Recognizing this, we still need to assess when I first saw the newly installed diorama and reassess our maximization of the evi- of the West Athens Hill Paleoindians at the dence, mainly lithic tools and the debris of NYSM during the late 1980s, I believe. There it lithic tool manufacture. was for the first time for me: a comprehensive Obviously, all the lithic artifacts, including picture of the Early Paleoindian period in 3-D tools (sometimes the only specimens called and full size, as if I was there at West Athens “artifacts” by Funk and others) and Hill too, and not just lithic tools! flakes/debitage, are an important foundation I am still drawn to the stone toolmaker sit- in Paleoindian research. Likewise, perhaps the ting near the West Athens Hill chert outcrops most important part of Funk’s work at West (on the left side of the diorama) and a woman Athens Hill is his extensive discussion of the and several children processing a dead cari- lithics, in no small measure because West bou (in the right foreground). The composi- Athens Hill was a lithic material acquisition, tion is balanced by the central hunter who or “quarry-workshop,” site, where points up the valley toward an approaching high–quality Normanskill chert was quarried. caribou herd, with a light scatter of trees The combined archaeological samples surely across the expansive landscape and the represent an incomplete but broadly represen- Helderberg Escarpment and the Catskill tative set of the inorganic artifacts employed Mountains in the far distance. For me, this and/or produced by the Paleoindians at West was the very first time that I could fully visu- Athens Hill. Parenthetically, Funk does us a alize Paleoindian lifeways in such detail. The great service by systematically comparing the West Athens Hill scene has surely already West Athens samples with various other rele- affected multiple generations of visitors to the vant Early Paleoindian ones from local and NYSM, much like the diorama of the regional contexts in the present publication. Paleoindian fire maker at the Royal Ontario Unlike some analysts of his generation Museum in Toronto and the newer, larger and others too, Bob usefully recognized that Paleoindian caribou hunting and butchering the West Athens Hill lithic samples represent a diorama at the Pequot Museum in continuum of production stages. In other Connecticut. words, the archaeological specimens represent Unfortunately, in this case we have little different points in a cyclical, not simply linear, else besides the lithic artifacts, limited local process of acquisition, production, use, reuse, paleoenvironmental data, and ethnographic and discard, rather than an “assemblage” of analogy, borrowing from historically known all equally finished tools in the sense of a hunter-gatherers, to inform us about this fixed, final “tool kit.” However, as we see in scene at West Athens Hill around 10,800 years Funk’s data and the reconstructed diorama at ago. Therefore, much of the close detail in the the NYSM, a host of other domestic activities West Athens Hill diorama is by necessity the were also undertaken at West Athens Hill,

xx representing more than just stone procure- Storck 1983, 1997; Tankersley 1995). Quite sig- ment and tool production. For the most part nificantly, continuities can be demonstrated the lithic “reduction” activities can be sorted between the Early and Late Paleoindian peri- out from these other activities, as Funk has ods on the basis of lithic technology alone. At profitably done in the present case. Lithic the same time that projectile point and other technology and assemblage composition are tool forms changed morphologically and obviously of primary importance at many sometimes technologically within each of Paleoindian sites, even where site contexts these two major subperiods, a remarkable have been extensively disturbed by natural degree of broad continuity is demonstrable and cultural post-depositional factors and not within each subperiod and also between them preserved in a pristine state, nor directly (e.g., Doyle, et al. 1985; Jones 1997; Petersen, dated reliably, as seen at West Athens Hill and Bartone, and Cox 2000:133). most other Paleoindian sites. Thus, the analysis of lithic technology has Paleoindian Regional Distributions, Lithics, been one of the most productive aspects of and Mobility recent Paleoindian research in the Northeast, yet it cannot be exhaustively reviewed here. Moving on beyond Paleoindian chronology Even this aspect of recent research is rather and technology, issues of broad regional distri- complicated, however. The analyst of butions and regional differences are likewise Paleoindian (and other) lithic technologies important, along with settlement pattern issues faces problems of preservation, raw material per se. First, a distinct boundary seemingly per- source identification (often only done macro- tained across the more northerly portions of the scopically and therefore tentative), adequate Northeast during the Early Paleoindian period, sample sizes (in light of typically small to beyond which to the north few, if any, fluted modest–sized samples), and reconstruction of points have ever been discovered, most likely the reduction processes. Paleoindian tool sam- marking the approximate northernmost limit ples consist of almost inevitably mixed, reached by fluted-point makers during the late incomplete, broken artifacts and partial rather Pleistocene epoch. This boundary must have than whole assemblages (Gramly and Funk been environmentally circumscribed and does 1990:19–24). not logically seem to have been sociocultural in Regional analysts have recently made very origin. significant strides in the area of Paleoindian For example, until last year absolutely no and other early lithic analysis and tool kit def- fluted points had ever been found in the inition. Numerous examples can be cited, but province of Quebec, whether in southerly sample variability is still enigmatic (e.g., areas near the St. Lawrence River, the modern Adovasio, Boldurian, and Carlisle 1988; Subarctic, or the Arctic portions. This absence Boisvert 1998; Carr 1989; Carty and Spiess occurs in spite of a very intensive and exten- 1992; Cavallo 1981; Cox 1986; Curran 1984; sive research program across much of Quebec Deller and Ellis 1988, 1992; Ellis and Deller over the past 30–40 years. It is clear that fluted 1988, 1990, 1997; Funk and Wellman 1984; points are largely absent in Quebec, probably Goodyear and Steffy 2003; Gramly 1982, as the result of widespread marine inundation 1988a, 1998, 1999; Gramly and Funk 1990; in the St. Lawrence Valley, ice to the north, and Gramly and Summers 1986; Grimes 1979; Hill other localized environmental restrictions at 2002; Jackson 1998; Jones 1997, 2003; Julig the time of fluted point manufacture. In 2003, 1994; Lothrop 1989; Lowery 1989; McAvoy the first notable exception to this generaliza- and McAvoy 1997; Moeller 1980; Sanders tion about Quebec, a single fluted point site, 1990; Shott 1990, 1993, 1997; Simons, Shott, was discovered in the extreme southeastern and Wright 1984; Spiess and Wilson 1987; uplands near Lac Megantic. This new discov-

xxi ery is very close to the border between Quebec southern Quebec, northern Ontario and and Maine (and not far from the Vail and Michigan, and beyond. Once glacial ice had Adkins Early Paleoindian sites) (C. finally retreated and food resources allowed Chapdelaine, personal communication 2003; habitation, subsequent Late Paleoindian occu- Chapdelaine and LaSalle 1995). Fluted points pation extended into portions of the modern do not occur anywhere else in Quebec as far as Subarctic and the southern margins of the we know, not farther east in Labrador and Arctic, to the west of Hudson’s Bay, for exam- Newfoundland, nor anywhere farther north ple (Wright 1995:96–99, Map 1). Broadly or to the east (into the High Arctic) (Wright speaking, Late Paleoindian remains thus 1995:24–29, Map 1). reached a more northerly distribution than Of comparable significance, a directly did fluted points in substantial portions of the analogous northern boundary for the distribu- continent. However, as noted above, Late tion of fluted points (and evidence of extinct Paleoindian developments never extended megafauna) was long ago defined in the very far (perhaps 300–400 km) to the south of Midwest–Great Lakes region, the so-called the “Mason-Quimby line” in the Northeast, at “Mason-Quimby line,” named for those who least not to the east of Lake Michigan, proba- first defined it, Ronald Mason and George bly because of contemporaneous Early Quimby. This boundary runs roughly Archaic populations in many places after ca. east–west across the upper Great Lakes of 10,000 B.P. northern Ontario, Michigan, and Wisconsin. Thus, Early and Late Paleoindian remains As in Quebec and southern Ontario to the do not have the same geographic distributions east, the equivalent of the “Mason-Quimby in the Northeast, and they spatially overlap line” extends westward into the Prairie across a relatively narrow zone in northerly Provinces of Canada, marking the northern- portions. This particular area of overlap most limit of known Early Paleoindian includes an east-west zone of southern remains (and megafauna again). This bound- Ontario and northern Ohio, much of upstate ary extends far westward close to but not fully New York, a bit of Pennsylvania, New Jersey to the Canadian Rockies, where fluted points and all of New England (Petersen, Bartone, are found much farther north within the pre- and Cox 2000; Wright 1995). Curiously, the sumed north–south “ice-free corridor.” This southern edge of both the Late Paleoindian hypothetical corridor is modeled by some as zone of distribution and the zone of overlap the late Pleistocene route for the first human closely parallel the farthest extent of migrants into the continent south of Alaska Pleistocene glaciation in the Northeast and the Yukon. In fact, small numbers of (Gardner 1989:11). Again, it would be difficult Clovis-like fluted points occur all along the to eliminate the likelihood of environmental eastern side of the Canadian Rockies and into influence on Late Paleoindian developments Alaska (Lepper 2002:81–83; Mason 1981:98; in this case, but seemingly distinctive socio- Storck 1988b:Figure 1; Wilson and Burns 1999; cultural factors may well have been relevant Wright 1995:Map 1). too, given the apparent contemporaneity of Glacial ice, fresh water, marine water, and Early Archaic populations in the region. other peri-glacial ecological conditions may Another broadly recognizable Paleoindian have restricted the regular presence of Early distributional pattern in the Northeast per- Paleoindians in northern areas over much of tains to differences between Early Paleoindian this huge region. Notably, Late Paleoindian sites, site sizes, lithic usage, and presumed developments, as defined here, ultimately mobility to the north and south of the extent of occurred well north of the “Mason-Quimby maximum Pleistocene glaciation regionally. line,” or the equivalent northern limit of flut- This hypothetical boundary, an organizational ed points elsewhere, including portions of distinction variably defined by David Meltzer,

xxii William Gardner, and others, has been called fewer site types are seemingly represented, by some the “biotic Mason-Dixon line.” It and often times known sites include multiple divides Early Paleoindian sites found in the components, with Archaic and Woodland north within the extent of maximum glacia- occupations mixed in, as presently under- tion from those to the south and beyond the stood. These distinctions between the north reach of glaciation (e.g., Gardner 1983:53, and the south suggest that one or more factors 1989, 2002:98–100; Goodyear 1999; Lepper differentiated these two zones across the 1999, 2002:81–83; Meltzer 2002:163–164; Northeast, from the Great Lakes eastward to Meltzer and Smith 1986). This second the present Atlantic coast, during the Early east–west boundary is situated well to the Paleoindian period. south of the “Mason-Quimby line” and it Logically, analysts like Gardner, Meltzer, marks regional distinctions between northern and others have suggested that late and southern portions of the Northeast (and Pleistocene paleoenvironments differed farther beyond to the south), correlating with between these broad zones, producing differ- paleoenvironmental differences during the ent subsistence practices and settlement pat- Early Paleoindian period. terns from north to south. Large Early This rough boundary was also at least par- Paleoindian sites, sometimes consisting of tially environmentally determined and may contemporaneous aggregation, occur in the have marked broad scale differences in the northern glaciated areas where the landscape sociocultural foundations for some of the earli- was initially open, with tundra or more likely est Paleoindians during the Early Paleoindian tundra parkland present at the time of occu- period based on differences in subsistence pation, and caribou hunting pertained, at least adaptations and resultant settlement patterns when and where “pioneer” occupation was from north to south. In the glaciated area to the characteristic. In this view, migratory caribou north, including southern Michigan, southern and perhaps other large game helped deter- Ontario, New York State, New England, and mine subsistence and settlement in the the Maritime Provinces, a greater range of northerly occupied portions of the Northeast Paleoindian site types and certainly site sizes for the Early Paleoindians, at least initially. has been recognized, for example. In marked contrast, more closed forests In this northern, glaciated zone, Early were present during all of the late Pleistocene Paleoindian sites are sometimes quite large, epoch in the non-glaciated southern portions though small sites seem to occur more com- of the Northeast and farther south into the monly. This pattern pertained at least for the Southeast per se, albeit the forests changed earliest portion of the Early Paleoindian peri- under different paleoenvironmental condi- od before the widespread establishment of tions and different plant and animal resources closed forests. Unfortunately, we know too lit- would have been available. For example, tle about the Late Paleoindian period to char- species not found together today sometimes acterize and compare it in this way. Also of co-occurred in the non-glaciated areas due to note, many sites of Early (and Late, as far as ecological compression, as we know from we can tell) Paleoindian vintage in the glaciat- paleontological analysis of late Pleistocene ed northern zone are single–component, deposits in caves and natural traps (Adovasio short-term occupations, or substantially so, with Page 2002:50; Funk and Steadman 1994; with occupation redundancy, if present, most- Lepper 1999:366). More importantly, large, ly confined to other Paleoindian components, gregarious herd animals were not common rather than later ones. within the continuously forested areas, and In the non-glaciated portions of the this circumstance had direct effects on the Northeast to the south, however, virtually all nature of Early Paleoindian subsistence in this Early Paleoindian sites are regularly small, generalized model.

xxiii Instead, in non-glaciated areas more typi- identifications. Regardless, most analysts of cally solitary game species such as deer, elk, Paleoindian lithic distributions assume that and moose would have been emphasized by virtually all lithics were obtained through the Early Paleoindians, and other foods such direct acquisition for one reason or another, as small game, plants, and fish, would have except perhaps in the most extreme cases. The required different landscape utilization pat- assumption that macroscopic identifications terns than those seen in the north. Such differ- are reliable may be most problematic. Thus, ences would have directly affected Early analysts typically reconstruct far-reaching, Paleoindian mobility and potentially other long-distance group mobility to explain the factors as well, including patterns of lithic raw presence of non-local/exotic lithics at Early material acquisition, for example (Carr and Paleoindian sites, whether coupled with (or Adovasio 2002:Table 7; Custer 1984:48–60, “embedded” in) subsistence or not. These lith- Figures 6 and 7; Dincauze 1993a:283–285, ic distributions serve as the foundation for 1993b; Ellis et al. 1998:152–154, Gardner 1983, establishing the territorial range of individual 1989:24–34, Figure 5, 2002; Goodyear groups (Curran and Grimes 1989; Deller 1999:433–435, 441–444; Lepper 1988, 1989:Table 8.4; Ellis 1989:146; Lothrop 1989; 1999:374–376; Meltzer and Smith 1986; Spiess, Meltzer 1989:12–13, 31, Table 2.2; Seeman 1994; Wilson, and Bradley 1998). This latter point, Spiess and Wilson 1989:95,Table 4.1; Storck concerning lithic raw material acquisition, is 1988a; Storck and von Bitter 1989). returned to below, especially as it pertains to A minority of analysts have casually allud- West Athens Hill. ed to or explicitly recognized the likelihood of Distribution issues related to smaller- some, perhaps typical, exchange of “exotic” scale, intraregional spatial differentiation and lithics during the Early Paleoindian period in tentative recognition of distinctive popula- the Northeast and elsewhere (e.g., Frison tions during the Early Paleoindian period 1988:94; Gramly and Summers 1986:98; have been addressed by various researchers in Hayden 1982; Petersen 1995; Snow 1980: the Northeast with variable success. These dif- 139–141, 152; Tankersley 1989:Table 11.3, ferences have been primarily identified using 1990:292, 1995:45, 1998b:16). This topic typological, tool sample composition, and deserves detailed consideration beyond what it especially lithic raw material variation. For can be given here because the practice of example, Paleoindian analysts typically differ- Paleoindian lithic exchange would necessarily entiate between “local” and “non-local” (or confound models of group mobility and group “exotic”) lithic raw materials in some compar- territories, but few analysts are so daunted. ative fashion (for example, those found farther In fact, more or less precise band territo- than 40 km away from a given site are non- ries for the Early Paleoindian period have local, “exotics” [Gould and Saggers 1985:119; been suggested for New York State and near- Meltzer 1989:31]). Regardless of the precise by areas, as well as New England and the definition of local vs. non-local, there are gen- Maritime Provinces, among others. In some erally two ways to broadly explain the pres- cases, these are more akin to what are called ence of exotic lithics within a given sample: “style zones” based on points and specific “direct” acquisition, in conjunction with some individual settlement patterns based on lithic degree of group mobility, and “indirect” raw materials. For example, Gramly (1988c) acquisition, or some form of transfer from one and Spiess and others (Spiess and Wilson group to another. 1987, 1989; Spiess, Wilson, and Bradley 1998) Macroscopically distinctive quarry sources have recognized important spatial subdivi- are generally assumed in these discussions, sions in New York State, New England, and leaving aside the issue of secondary sources for adjacent areas. In the most concrete local the most part, as well as the reliability of source example, Gramly (1988c:269–270, Figure 1)

xxiv has posited the existence of three Paleoindian Atlantic region of Delaware, Pennsylvania, “culture areas, or at the least, regions” in New and Virginia (e.g., Custer and Stewart York State. 1990:Figure 9; Lowery 1989; Seeman Gramly’s three territories are correlated 1994:Figure 2). with three “seasonally transhumant bands” of In Ontario, similar evidence of differen- Early Paleoindians in New York State (Gramly tially represented lithic materials has been 1988c:265, Figure 1). The first of these Early comparably used to suggest different band Paleoindian “bands,” the westernmost, had a preferences and even the use of some lithics as range that extended from western New York conscious (ethnic-like) group markers. These State into Ohio and Pennsylvania, based on patterns also apparently changed during the visually distinctive Flint Ridge and other Ohio span of the Early Paleoindian period and later lithic materials. The central “band” employed in Ontario, with a reduction in scale and thus, primarily Onondaga cherts from upstate New presumed mobility over time (e.g., Deller York, among others, but with southward links 1989; Ellis 1989;Ellis and Deller 1997:12, 14-15; to the Susquehanna Valley in Pennsylvania. Storck 1984, 1988a; Storck and von Bitter The third Early Paleoindian “band” in 1989). The possibility of limited exchange is Gramly’s model was situated within the also suggested in Ontario (e.g., Ellis and Hudson and Mohawk valleys, as marked by Deller 1990:54). Is it possible that these common usage of Normanskill and other changes over time represent transformations Hudson Valley cherts. in intergroup exchange for some reason, This latter territory included the West rather than a reduction in mobility? Athens Hill chert source and occupation site These and various other analysts have within a broader territory centered on the generally assumed that lithic raw materials Hudson valley (Gramly 1988b:Figure 1). Of and tools were transported by Early note, the vast majority of the lithic tools (not Paleoindians largely, if not solely, through debitage) at West Athens Hill, roughly 98 per- direct acquisition of some sort. Individual cent, were seemingly made from the immedi- Paleoindian groups (or representatives there- ately available (and thus very local!) of) directly acquired their needed lithic mate- Normanskill chert. Non-local exotics at West rials themselves, either at a primary source Athens Hill include distinctive Pennsylvania such as West Athens Hill or from a secondary jasper from the southeastern part of the state, source such as cobbles found in drift or till. In western Onondaga chert from western other words, this model assumes that the upstate, and Upper Mercer chert from eastern Paleoindians who directly acquired various Ohio. Not all of these exotics are easily lithic raw materials were also the same ones accounted for within a direct acquisition that ultimately used them to fashion needed model, as previously discussed by Funk lithic tools. This model is also commonly (1976:205, 223–226). applied to Early Paleoindians in other areas Spiess and others have comparably sug- well beyond the Northeast (e.g., Goodyear gested the existence of “a New England– 1989; Meltzer 1989). This model of direct Maritimes Paleoindian region” used by “two acquisition of lithics is typically related to and or more distinct bands” on the basis of fluted underpins overall reconstruction of point stylistic variations and lithic raw materi- wide–ranging mobility among Paleoindian als there (Spiess and Wilson 1989:82). These hunter-gatherer groups. and other data beyond New York State and Many researchers acknowledge that Maine have been used to model an overall exchange (or trade) of lithic raw materials series of “lithic supply zones,” or band territo- may be relevant in the most extreme cases ries, from Nova Scotia to Indiana and others where exotic lithics occur far from their geo- southward. The latter include the Mid- logical sources. This explains, for example, the

xxv presence of Knife River chalcedony, or chert, (e.g., Ellis 1989; Meltzer 1989). However, it is from North Dakota at the Lamb site in New reasonable to question what effect the small York State (Gramly 1988b, 1999), or Alibates number of identified sites and the much chert from Texas in western Pennsylvania smaller number of studied sites with adequate (Lantz 1984; Tankersley 1990). However, more samples would have. local lithic materials, recovered say 200–300 Leaving aside the reliability of source attri- km from their expected sources, are typically butions in the first place, a sharp falloff might related to direct acquisition (e.g., Curran and be the only pattern discernible without an Grimes 1989; Deller 1989; Deller and Ellis adequate set of lithic samples drawn from a 1988; Ellis 1989; Ellis and Deller 1997; Gramly geographically extensive and statistically sig- 1982; Spiess and Wilson 1989). nificant number of sites. This and other Many analysts thereby relate Paleoindian aspects of the argument suggest that the lithic acquisition to hunter-gatherer annual predilection to interpret exotic lithics as evi- subsistence rounds and thus, lithic acquisition dence of far-ranging mobility may be akin to was embedded within subsistence mobility, environmental determinism, perhaps in reali- following the ideas of Lewis Binford (1979) ty “lithic determinism,” among regional (e.g., Gramly 1980:828–829, 1988c:267–270). researchers. Some of these readily propose the However, not all see lithic raw material acqui- centrality of lithics for the Paleoindians and sition as part and parcel of the subsistence even explicitly call it just that (e.g., Carr round, but instead it may have been “disem- 1989:21; Ellis and Deller 1997:12; Gardner bedded,” or “uncoupled,” with a separate, 1983:57, 1989:24–34, 2002:97–100). Lepper specialized direct acquisition strategy or (2002:85) correctly notes, however, that: “lithic strategies at work (e.g., Seeman 1994; Spiess determinism is illusory. It is a function of how and Wilson 1989:89, 97). Of related interest, the archaeological record formed, not a reflec- the pattern of long-distance lithic acquisition tion of Paleoindians having tethered their for- and transport in the northern portions of the aging radius to a chert [or another] quarry.” Northeast, whatever the mechanism, appar- A predilection toward lithic determinism ently contrasted with more southerly portions among Paleoindian analysts is understand- of (and farther southward beyond) the able because there is little else in the archaeo- Northeast in correlation with greater and less- logical record besides lithics to work with in er degrees of mobility, as noted above. So, if the formulation and resolution of research we were to model the presence of non-local, questions. Most analysts apparently believe exotic lithics as at least the partial product of that Early Paleoindians would not have been Paleoindian exchange, rather than mobility present in sufficient numbers to have alone, we would need to account for rather exchange partners in the Northeast (and else- different exchange patterns between northerly where). Very limited, widely spread groups and southerly areas. may well have been characteristic among the In more northerly areas of the Northeast, very first Paleoindian pioneers in the support for direct acquisition of lithic materi- Northeast. Pioneering must have come to an als over long distances is drawn from the form end in most areas relatively quickly, perhaps of the distribution (or shape of the “fall-off only a few generations or less than 100 years, curve”) of Paleoindian lithic materials over a brief instant when compared to the full distance, where it is said to be quite steep or duration of the Early Paleoindian period, sharp, rather than gradual. Gradual fall-off spanning 1,000 (uncalibrated) years or so. lithic frequency distributions are believed to Once the earliest Paleoindian pioneers filled pertain archaeologically where “down-the- much or most (even all) of the extent of their line” exchange is represented and a sharp fall- ultimate distribution, exchange and other off supposedly represents direct acquisition forms of interaction would be logically expect-

xxvi ed as a means of maintaining social connec- Ritchie and Funk 1973:336; Seeman tions, exchanging marriage partners, etc., 1994:283–284; Shott 1997:215). some or all of which must have been critical Hunter-gatherer evidence drawn from for survival. around the globe may be relevant for Suffice it to say here, a process of informa- Paleoindian lithic acquisition studies in spite tion exchange was relevant over immense dis- of cautionary tales to the contrary. tances during the Early Paleoindian period Ethnographically and archaeologically record- and later as well. Such interaction is obvious- ed Arctic hunter-gatherers and Aboriginals ly manifested by a strong degree of common from Australia seem appropriate here. This is projectile point stylistic change over time (e.g., the case based on severe environments, result- Clovis/Gainey to Folsom/Cumberland/ ant far-ranging mobility (for reasons other Barnes types, etc.) and other innovations than lithic acquisition), a relatively limited across much of eastern North America, along tool kit (albeit no fluted points!), and ethno- with similar coeval developments during both graphic records of lithic usage, including the Early and Late Paleoindian periods. extensive exchange (Hayden 1979; Loring Although still incompletely understood, these 2002:166; Shott 1997:215; see Funk 1976:226). more or less synchronous changes may have Among both Arctic and Australian hunter- been linked to intergroup exchange of lithics gatherers, the propensity for information and perhaps perishable items as well. sharing as a mechanism of risk minimization This exchange of information and tangible fostered interaction networks intraregionally objects apparently diminished progressively and interregionally long ago, cumulatively over time, especially with the onset of the Late spanning Australia and large areas within the Paleoindian period and thereafter, with the Arctic and beyond prehistorically (Loring ultimate reduction of point style distributions 2002; Lourandos 1997:40–43). across broad regions, but it nonetheless persisted with a strong degree of synchronicity. Paleoindian Settlement Patterns Lithic raw material utilization became progressively more localized as well during this Recognizably bounded, if sometimes same temporal span. By the onset of the Early huge, lithic utilization territories for Early Archaic period, lithics were mostly local, at Paleoindians have enabled many researchers least in the far Northeast, even while synchro- to refine their sense of Early Paleoindian set- nous stylistic changes continued to a large tlement patterns per se in the Northeast. In degree across the region (Petersen 1995). fact, it is no longer fashionable to call Ethnographic analogy is variably appro- Paleoindians “free wandering” in terms of priate and useful within the context of broad their settlement patterns and instead, they are Paleoindian research, but here is a case when more often designated as “restricted wander- it seems potentially quite useful and even crit- ing,” although whether loosely or tightly ically important. Apparently, differences in restricted is an open question (cf. Ritchie geography and time have led most 1983:30; Ritchie and Funk 1973:336; Snow Paleoindian researchers to dismiss ethno- 1980:128–129, 150; Storck 1988a:247). graphic analogy as irrelevant in modeling lith- Suggestion of “central-based wandering,” ic raw material distributions in the Northeast. where territories are limited in size and a cen- Nonetheless, they are willing to employ anal- tral base serves as a focal point, has been ogy for modeling other factors such as tool briefly suggested but quickly rejected for production and use, subsistence, general Paleoindians (Funk 1976:224–227). Regardless mobility, and social structure, among others of the precise terminology, at present Early (e.g., Custer and Stewart 1990:310–315; Funk Paleoindians are more or less uniformly rec- 1976:226–228; MacDonald 1968:129–134; ognized as having lived within large but ulti-

xxvii mately fixed geographic territories, at least for Artifact (and rare feature) distribution those in “second generation,” non-pioneer data provide a means for comparison within mode. This minimally pertained in the north- and between sites, as in the present case at ernmost areas of the Northeast inhabited at West Athens Hill, among others. Nonetheless, the time, including New York State (e.g., the shallow, non-stratified, near surface condi- Dincauze 1993a; Ellis and Deller 1990, 1997). tion of nearly all such sites means that cultur- In light of the uncertainty about al and natural post-depositional disturbance Paleoindian group mobility and/or exchange has oftentimes partially or extensively com- on multiple levels, it should come as no sur- promised the archaeological contexts, prise that local aspects of settlement patterns accounting for the paucity of reliable radiocar- are poorly known in the Northeast for both bon dates. For example, I strongly suspect that the Early and Late Paleoindian periods, a significant degree of post-depositional dis- including both northern and southern areas. turbance, both natural and cultural, pertains This observation pertains to most intrasite and at West Athens Hill. Even though West Athens intersite analyses. Contemporaneity is a cen- Hill has never been plowed, the shallow, tral problem both within and between sites, in largely rocky context must have enabled vari- other words. The degree of coeval residential ous forms of natural disturbance to alter the aggregation versus long-term composite occu- artifact distribution to some (large?) degree. pation at any given site, the degree of local The putative artifact “clusters” at West Athens and regional mobility, and other variables Hill may be more fortuitous than real, espe- remain poorly known unless evidence of con- cially between the two major strata in Area B. temporaneity can be established. Similarities Funk clearly recognized this during his in point forms, broader tool sample composi- reanalysis of “clusters” at West Athens Hill tion, and raw materials are all used, however, between the 1970s and the present publica- to help establish (roughly) contemporaneous tion. In fact, the case might be made that sup- components at some Paleoindian sites. Many posedly discrete artifact clusters overlapping available reconstructions of Paleoindian set- spatially between the strata at West Athens tlement patterns are still likely to be illusory. Hill are, in fact, precisely the same clusters As Ritchie and Funk (1984:3) aptly pointed out across strata, with artifacts differentially and in their review of an important collection of variably transported by natural agents Paleoindian studies 20 years ago, “in our between them. opinion the search for the single occupation Likewise, the still enigmatic nature of remains an elusive chimera.” Paleoindian site location parameters across Significant effort has been expended over the landscape means that most sites likely the past few decades in delineation of intrasite remain to be discovered and/or have been settlement patterning through careful excava- permanently lost through disturbance. The tion and documentation at particular sites. very antiquity of Paleoindian sites and known This is certainly very important and com- environmental transformations over a long mendable work, and definition of internal site span suggest that many have been destroyed settlement patterns has proven to be highly by contemporaneous and later natural land- rewarding. Discrete Paleoindian activity areas scape alterations such as rising lake and ocean are demonstrable in some cases (Carty and levels, river channel migration, and other Spiess 1992; Deller and Ellis 1984, 1992; Ellis forms of erosion. Still other sites certainly lie and Deller 2000; Funk 1976:222–223; Gramly deeply buried, especially in the non-glaciated 1982, 1984, 1988a; Gramly and Lothrop 1984; portions of the region (e.g., Goodyear and Grimes 1979:Figure 4; Grimes, et al. 1984; Steffy 2003; McNett 1985). It seems unlikely Seeman 1994; Spiess and Wilson 1987; Spiess, that we will ever fully understand the com- Wilson, and Bradley 1998:Figure 13). plete range of Paleoindian site types and inter-

xxviii site settlement patterns, given the modest ros- are known from Udora (Curran 1984; Spiess, ter of sites currently known across the region Curran, and Grimes 1985; Storck and Spiess for all major period subdivisions (Early and 1994). As noted above, the very early associa- Late) and subperiods thereof. tion of caribou and Early Paleoindian remains Yet, 30 years ago Ritchie and Funk (1973:5, can now be dismissed for Dutchess Quarry 333–334) were able to define at least four Cave (Funk and Steadman 1994). Paleoindian site types in New York State and To the south, the deeply stratified (and the broader Northeast, including “quarry- thus very rare) Early Paleoindian levels at the workshops,” “small camps,” “major, recur- Shawnee Minisink site on the Delaware River rently inhabited camps,” and “rockshelter or in Pennsylvania preserve evidence of fish and cave camps.” Obviously, the “quarry-work- various potential plant foods. The plants shop” Paleoindian site type was directly mod- include acalypha (Acalypha virginica), ama- eled after West Athens Hill (Ritchie and Funk ranth (Amaranthus sp.), “blackberry” (Rubus 1973:5, 333–334). Still other Paleoindian site sp.), buckbean (Menyanthes triofoliata), cheno- types have been variably recognized in the pod (Chenopodium sp.), hawthorn (Craetagus Northeast since 1973. Supplemental site types sp.), smartweed (Polygonum sp.), and winter include, but are not limited to, “kill-butchery,” cress (Barbarea orthoceras). At least the fish and “burial sites and caches,” and “isolated finds,” hawthorn plum remains originated in a hearth if one chooses to call the latter category sites dated to 10,590 + 500 B.P. (Dent and per se (Gramly and Funk 1990:13–19). Various Kaufmann 1985:72–73, Table 5.2; McNett other permutations exist as well (e.g., see Carr 1985:17–19, Figure 2.3). and Adovasio 2002:29–34; Custer At the stratified Hedden site in southern 1996:107–110; Ellis and Deller 1990:51–52, Maine, a set of Early Paleoindian floral sam- 60–61; Gardner 1983:53–56; 1989:24–34; ples has been recently identified from a con- Lepper 1988:39–41; Tankersley 1996:26–27, text buried under aeolian sands. The Hedden 32–33, 35). More work remains to be done to floral sample, directly dated to 10,580 + 60 B.P. fully explore Paleoindian distribution evi- and 10,500 + 60 B.P., includes seeds of bristly dence and settlement patterns on multiple lev- sarsparilla (Aralia hispidia), bunchberry els, both within and between sites. (Cornus canadensis), bramble (Rubus sp.), and grape (Vitis sp.). Only the grape fragment was unequivocally associated with cultural Paleoindian Subsistence Patterns remains, but all were likely associated with Some researchers think that they are on the Early Paleoindian occupation (Asch Sidell firmer ground when discussing Paleoindian 1999:197, Tables 12.1 and 12.5; Spiess, et al. subsistence patterns in the Northeast. As with 1995). absolute dating in the different areas, several On the western margin of the Northeast as precious discoveries regarding Paleoindian broadly defined here, white-tailed deer (for- subsistence have been recently made, but this merly reported as caribou) was identified, topic is still hampered by poor preservation of along with more tentative caribou, elk, or nearly all subsistence remains in nearly all moose at the Cummins site near Lake cases. In each case from Early Paleoindian Superior in northern Ontario. Blood protein contexts of varying sorts, caribou and possible residue on 12 lithic artifacts and soil samples caribou remains are now known from the also tentatively demonstrated bison, and deer, Whipple site in New Hampshire and Bull rabbit, and rodent families (Julig 1994:214). In Brook in Massachusetts, as well as Udora in northern Wisconsin toward the Great Lakes, Ontario and supposedly Holcombe in the Late Paleoindian (to Early Archaic) Michigan. Beaver remains have been identi- Deadman Slough and Sucices sites produced fied from Bull Brook, and Arctic fox and hare white-tailed deer, black bear, beaver, porcu-

xxix pine, several turtle species, fish, and other and Funk 1973:335; Spiess, Wilson, and mammal remains, as well as mussel shells. A Bradley 1998:224–226; Storck 1984). For exam- broad generalist adaptation is obviously ple, witness the West Athens Hill diorama at reflected, at least in the case of the Wisconsin the NYSM, with caribou centrally featured in evidence (Kuehn 1998). spite of the complete absence of subsistence Beyond these few relatively recent identifi- remains from the site. In non-glaciated areas cations, there are no well-grounded reconstruc- to the south, various other large game species tions for Paleoindian subsistence in the such as deer and elk have been mentioned as Northeast, although many researchers have possible but unproven focal resources within guessed at relevant subsistence patterns in dif- an overall more generalized range of subsis- ferent portions of the region, as briefly tence resources (Dincauze 1993a:281; Lepper described above. It is certainly tempting to 1999). Meadowcroft preserves evidence of model Paleoindian subsistence based on evolv- white-tailed deer in “Pre-Clovis” contexts, ing knowledge of late Pleistocene and early however (Adovasio, Boldurian, and Carlisle Holocene vegetation and other regional biota, 1988; Carr and Adovasio 2002:8). but this is inevitably speculative, since knowl- Likewise, concerted efforts have been edge of the biota is imperfect and factors of cul- made in the Northeast to link at least the Early tural selectivity must pertain as well. Paleoindians to predation on “megafauna” In one very notable case, Gramly (1982, such as mastodons, mammoths, and others 1984, 1998) has reconstructed a possible “kill” that went extinct at the Pleistocene-Holocene site, presumably for caribou, at the Early transition, but virtually no one has been suc- Paleoindian Vail site in western Maine. Very cessful to date due to contextual and interpre- close nearby, Gramly (1988a) has further iden- tive problems (e.g., Dincauze 1993a:281; tified a “meat cache” at the contemporaneous Lepper 1999:374; Mason 1981:69–70, 99–101; Adkins site, presumably for storage of caribou McAndrews and Jackson 1988; McDonald meat. These are plausible but by no means air- 1994; Snow 1980:117–122). The Kimmswick tight suggestions. Nonetheless, the “kill” site site in Missouri is the closest unequivocal hypothesis for Vail seems virtually certain Paleoindian association with “megafauna” to based on conjoinable fluted point fragments the Northeast, where Early Paleoindian (true) found widely separated between the pre- Clovis points were intimately associated with sumed killing and habitation areas, and spe- mastodons (Graham and Kay 1988; Kay 1998). cific details of local geomorphology. We can- However, the presence of Paleoindian and not be absolutely sure what animals were presumed Paleoindian artifacts and extinct being dispatched at Vail, but some sort of herd fauna in New York State, Ohio, Kentucky, and animal seems most likely, given the local Wisconsin, for example, is suggestive (Ellis, et topography and paleoenvironments at the al. 1998:158–159; Tankersley 1996:27–28). The time. Caribou logically seem to be the best fit. Martins Creek site in Ohio has mastodon Gramly is not the only analyst of Early remains associated with lithic flakes, while the Paleoindians and Paleoindians generally that Schaefer mammoth in Wisconsin was associat- has postulated an emphasis on caribou. ed with a fragmentary biface, a flake, and a Various others have previously suggested a date of 10,960 + 100 B.P. (Brush and Smith Paleoindian caribou focus, both Early and 1994; Overstreet, Joyce, and Wasion 1995). Late, especially in the glaciated northern por- The Burning Tree mastodon in Ohio, dated ca. tions of the Northeast. There is a paucity of 11,390 + 80 B.P to 11,660 + 120 B.P., is inter- direct evidence, however (e.g., Chapdelaine preted as having been butchered, but there are 1994; Davis 1998:200; Doyle, et al. 1985; no associated artifacts and this interpretation MacDonald 1968:140; Mason 1981:99; is quite conjectural (Fisher, Lepper, and Hooge Petersen, Bartone, and Cox 2000:131; Ritchie 1994), among various others.

xxx Probably one of the best candidates for a game emphasis to the exclusion of other food bona fide association of extinct fauna and sources, even in the far northern portions of Early Paleoindian remains in the Northeast is the Northeast where various analysts have the Hiscock site, a spring side association of come to suspect it. Likewise, there is precious cultural artifacts and primarily (solely?) natu- little, if any, evidence to support an emphasis ral accumulation of faunal and botanical on now extinct “megafauna” anywhere in the remains in western New York State. More than broad Northeast. Thus, in reality we cannot 50 vertebrate species have been identified say conclusively if the Early Paleoindians from deposits of late Pleistocene and were “focal” or “diffuse,” emphasizing large Holocene age at Hiscock, including at least game or a broad spectrum of food resources. eight mastodons thus far, along with caribou, We are only a little better off for the Late elk, moose of some sort, California condor, Paleoindians on and near the Upper Great and others in the Pleistocene deposits. Dates Lakes where generalist patterns are quite between ca. 11,200 B.P. and 10,200 B.P., among obvious. others, were roughly associated with the mastodons, other fauna, and various artifacts. Paleoindian Mortuary/Ritual Patterns A minimum total of seven lithic tools at Hiscock includes six Early Paleoindian fluted Mortuary and other ritual aspects of the points and a scraper. Another 13 (+) bone arti- Early and Late Paleoindian periods in the facts are also generally associated, of which Northeast are also poorly known. Discrete several worked mastodon bones, a rib, and a lithic artifact caches are sometimes interpreted vertebrae have been AMS radiocarbon dated as mortuary goods related to burials where to 10,990 + 100 B.P. and 10,810 + 50 B.P. These the human remains and all other organics intriguing discoveries are problematic have been destroyed by the ravages of time. because of the complex deposition, erosion, Some sites preserve cremated human bones and redeposition at Hiscock, and thus, none of with lithics, as at the Late Paleoindian Renier these associations are completely unequivo- site in Wisconsin, and others preserve human cal. For example, it is possible that Early bones without lithics, as at the Late Paleoindians were acquiring bones of natural- Paleoindian Cummins site in northern ly deceased animals and not killing them Ontario. Still others preserve heat fractured or there. Continued research hopefully will pro- otherwise broken lithics but no human vide unequivocal evidence of Early remains, with and without red ocher (e.g., Paleoindian predation of megafauna and/or Julig 1994; Kuehn 1998:464; Mason other species at Hiscock, potentially adding 1981:117–120; Walthall 1999). Rather cautious- significant new data about Paleoindian sub- ly, lithic tool caches without human remains sistence (Ellis et al. 1998:158; Laub 1995:27, have been sometimes attributed to “sacred rit- 2000; Laub and Haynes 1998:32–34; Laub, et ual” behavior but not necessarily mortuary al. 1988:76, Table 1; Steadman 1988:Table 1; ritual for the later portions of the Early Tankersley 1998b:11–12). Paleoindian period. This pertains at the Evidence from sites like Bull Brook, Crowfield and Caradoc sites in southern Whipple, Udora, Shawnee Minisink, and even Ontario (Deller and Ellis 1984, 2001:279–281; Meadowcroft seem to support the idea of a Wright 1995:56). generalist subsistence adaptation for Early Neither Early nor Late Paleoindian burials Paleoindians in different portions of the with preserved human remains are known Northeast, but the relevant data are very slim, from New York State, the New England– to say the least. There seems to be no empiri- Maritimes area, or the Mid-Atlantic area, but cal evidence to support the idea of a large at least one lithic cache like those in southern

xxxi Ontario is known in New York State. The THE SIGNIFICANCE OF WEST Lamb site in the western upstate area pro- ATHENS HILL duced a cache of 18 lithic artifacts, including eight fluted Gainey points and 10 other lithic bifaces (“knives” and “preforms”), which are Bob Funk’s research at West Athens Hill assigned to the earliest Early Paleoindian, and other Paleoindian sites largely speaks for Clovis-related occupation. No bones or red itself. The significance of West Athens Hill ocher survived at Lamb, but the cache was should be quite obvious to the reader by this interpreted as a disturbed single human burial juncture, given how it fits into a broader (Gramly 1988b, 1999:35–37, 94; Gramly and analysis of the Early Paleoindian period in the Funk 1990:16). Northeast. Reiteration of all of this is unneces- If correctly identified, the Lamb site cache sary, but it may be worthwhile to summarize is the oldest known burial in much of the this topic before the reader moves on to the Northeast. Another comparably small cache, substance of Bob’s report proper. related to the Parkhill phase based on Barnes First of all, the present publication on West points, is known from the Thedford II site in Athens Hill is important for providing the sci- southern Ontario. The excavators did not des- entific underpinning of the influential ignate it as a burial, but it seems similar to, if Paleoindian diorama at the NYSM, as slightly younger than, the cache at the Lamb described above. The powerful vignette repre- site. Similar caches may have been present at sented in the NYSM diorama was obviously Bull Brook in eastern Massachusetts, but these directly centered on Bob’s Paleoindian are even less certain than Thedford II (Deller research at West Athens Hill, as well as his and Ellis 1992:99–100; Gramly and Funk broader anthropological vision, and the efforts 1990:16, 19). of the talented NYSM exhibition staff. Now, From the standpoint of anthropological we have a more complete idea of what evi- archaeology, it would be very useful to learn dence Bob and others had to work with in more about the caching, “sacred ritual,” constructing the diorama. and/or possible mortuary behavior of the Second, it should be obvious that there are Early Paleoindian period in the Northeast. very few comprehensive and detailed reports From the typical Native American standpoint, for sites of Early Paleoindian vintage anywhere however, this will be very difficult, if not across the Northeast. Thus, the details of Bob’s impossible, under liberal interpretation of the work, along with others, at West Athens Hill Native Graves Protection and Repatriation are significant in and of themselves, regardless Act, where all prehistoric human remains are of his precise interpretations or anyone else’s off limits for analysis. Regardless of this issue, for that matter. Archaeological research in the in undertaking any sort of future Paleoindian region will not progress without substantive research, whether habitation or otherwise, we data for this or any other period of prehistory should strive to work as collaboratively as or history. Various analysts have drawn on possible. As seen at Hiscock, Dutchess Quarry, Bob’s earlier summaries of his West Athens Hill Meadowcroft, and elsewhere, collaborative research in their own analyses (e.g., Eisenberg and interdisciplinary research teams, involv- 1978:136–137, Figures 21–25; Ellis 1989:Table ing earth scientists, natural historians, and 6.1; Gramly 1980:828–830; Meltzer 1989:34–35, anthropological archaeologists, seem to be the Table 2.2; Moeller 1980:Table 20; Snow best hope we have for advancing northeastern 1980:138–140, Table 3.2). Now, we have a more Paleoindian research. complete site report that can be used in future studies. Comprehensive published site reports have begun to appear for a small number of

xxxii Early Paleoindian sites in the Northeast. rather different (Boisvert 1998; Bonnichsen, Although of different scope, focus, and com- Kennlyside, and Turnmire 1991; Eley and von pleteness, these include reports for Early Bitter 1989; Gardner 1974, 1989; Hill 1997; Paleoindian sites in Connecticut, Kentucky, Pollock, Hamilton, and Bonnichsen 1999; Maine, Michigan, New Jersey, New York State, Spiess, Wilson, and Bradley 1998; Storck and Nova Scotia, Pennsylvania, and Ontario, for von Bitter 1989). example, but these number less than 20. Still another aspect of the importance of Parenthetically, comprehensive published site Bob Funk’s work at West Athens Hill stems reports for Late Paleoindian sites are even from his recognition that this site logically more rare regionally. Taken together with preserves evidence of the direct acquisition those previously published by the “pioneer” and reduction of a notable lithic material, regional analysts, we can begin to sense some Normanskill chert, as well as inferred domes- of the exciting research topics inherent in tic activities. Bob here largely follows the lith- Paleoindian research generally. Bob’s recent ic analysis format of his previous description reanalysis of West Athens Hill is thus highly of West Athens Hill (Ritchie and Funk significant in building on his previous work 1973:9–36; see also Funk 1976:205–207, there, his other valuable publications, and also 212–228, Table 31), among others. Again, in the work of many other colleagues. the present volume Bob broadly describes the Third, the West Athens Hill research is sig- process of lithic reduction and resulting “arti- nificant in terms of its examination of what facts” (read tools) and debitage (flakes, shat- Bob, alone and with others, called a “quarry- ter, etc.) at West Athens Hill in terms of a workshop” in an overall classification of dynamic set of reduction processes, or Paleoindian site types. Bob explicitly argued “stages,” as noted above. that West Athens Hill was also a habitation More might have been done with this site, or “camp,” in contrast to other research topic, but Funk’s processual model Paleoindian site types, as noted above. exists in contrast to other archaeological clas- However, his contribution here partially stems sification approaches that have been applied from the site type classification itself. It also to descriptions of Paleoindian (and other) lith- partially stems from a more basic recognition ic samples, generally interpreting them in stat- that not all Paleoindian sites were equivalent ic fashion, much like a comprehensive, con- in setting, size, and function, thereby helping temporaneous took kit or assemblage. In the Bob (and Bill Ritchie) to prompt a positive and static case, lithic samples are seen (implicitly ongoing trend of defining Paleoindian vari- or explicitly) as consisting of more or less ability in the Northeast. fixed sets of all equally “finished” or other- Moreover, “quarry-workshop” sites have wise equivalent items (e.g., Ritchie 1953, 1957; been subsequently recognized elsewhere in Shott 1997; Wilmsen and Roberts 1978; see different areas of the Northeast, in part given Funk 1976:Table 31). As noted above, such the importance of the West Athens Hill artifacts are either “knives,” “scrapers,” research. Northeastern Paleoindian lithic raw “points,” and “drills,” among many other cat- material acquisition of different sorts has egories. Static lithic analysis models are useful recently been studied at various places in their own right and sometimes said to be besides West Athens Hill. These other more “objective” than processual taxonomic Paleoindian “quarry-workshop” locales ones, but it is not necessarily so, as often wit- include Munsungan Lake in Maine, Israel nessed by their own conjectural taxa (Shott River in New Hampshire, and Fossil Hill in 1997:203–204). Ontario, among others. The Flint Run and Visualizing lithic reduction, use, and dis- Williamson locales in Virginia are also ger- card as a set of processes, representing a series mane here, but all of these lithic sources are of reduction “stages,” was still innovative in

xxxiii the 1970s (e.g., Callahan 1979; Collins 1975) ingly frequent long–distance lithic transport, when Bob employed it at West Athens Hill as well as diverse site types and sizes, pre- and it continues to be so today, even though it sumably pertained during the Early has a much older heritage. Funk has again Paleoindian period, ca. 11,000 B.P. to 10,000 made this a useful part of the present volume B.P. These conditions were apparently differ- on West Athens Hill, even if we might dispute ent than those in other portions of eastern a few of the reduction stage assignments. North America, such as the Southeast and Some analysts have recently argued that lithic beyond. reduction is, in reality, more of a continuum, No evidence of a “Pre-Clovis” occupation rather than a set of “stages” (Shott 1996). The has been identified anywhere close to West “stage” approach thus simplifies reality some- Athens Hill, although several other Early what arbitrarily but systematically, as is typi- Paleoindian sites are known nearby, such as cal in other taxonomies. Kings Road and another site only a few kilo- Nonetheless, Bob’s analysis generally meters away. Human occupation of all of New demonstrates that reduction stage models York State may have only begun with the help the analyst identify artifacts that entered Early Paleoindian period, as late as 11,000 B.P., the archaeological record at different points in as represented at West Athens Hill. Human the complicated nexus of reduction, use, colonization of this and other parts of the far reworking, recycling, discard, etc. Thus, they northern and eastern portions of the help to define just which of these and other Northeast was precluded before this time, or activities took place at a given site. At West less likely, such evidence has yet to be recog- Athens Hill, this general approach (and differ- nized. The inverse correlation between entiation of local from non-local, “exotic” glaciated areas and an absence of “Pre-Clovis” chert) enables recognition of the fact that it occupation is probably not coincidental. Since was more than a lithic workshop, however. the whole of New York State was glaciated, it The Early Paleoindian occupation of West must have been available for human coloniza- Athens Hill was also the scene of other tion only rather late during the Pleistocene domestic habitation activities and potentially epoch. Obviously, this matter remains to be also used as a lookout (both nicely portrayed tested through future research. in the diorama at the NYSM). Bob Funk’s Paleoindian research at West In summary, Bob Funk’s Paleoindian Athens Hill helps us to define and evaluate research at West Athens Hill is already a land- the earliest extensive human colonization of mark in northeastern and broader archaeolog- the Northeast as it occurred during the Early ical contexts on multiple levels. The present Paleoindian period. This period effectively volume further demonstrates this point and began instantaneously and synchronously, if more thoroughly presents the salient details. we are reading the limited data correctly, just Funk’s work helps us to better understand the as early Native people spread across nearly Early Paleoindian period generally and it the whole Northeast, the exception being the sheds light on the role of lithic raw materials extreme northern portions toward the in technological, settlement, and perhaps even Subarctic and beyond. Areas roughly north of subsistence systems, even if we don’t fully the “Mason-Quimby line” were apparently understand the meaning of lithic distributions uninhabitable until conditions changed dur- regionally. West Athens Hill is situated in the ing the Late Paleoindian period, at which time northern, glaciated portion of the Northeast ancestral Natives moved into the margins of and as such it provides a point of comparison the Subarctic and later even into the Arctic, for other sites both in northerly and southerly but only in limited areas. The southern non- areas. This is a region where a great degree of glaciated portions of the Northeast were mobility, long–distance movements and seem- seemingly colonized first during the “Pre-

xxxiv Clovis” period, as in the Southeast, followed wonderful man. later in the northern glaciated areas during the Beyond the details per se, Bob Funk and Early Paleoindian period. his boundless enthusiasm come through in By this time, sufficient biotic resources had this publication on West Athens Hill, as it been reestablished across the southern and should be. Though Bob was a careful and mid-northern portions of the Northeast to the exacting scholar in his archaeological analyses point that people could survive and prosper and publications, he was much less restrained there, including the mid-Hudson Valley and (unrestrained?) in his personal interactions. West Athens Hill. West Athens Hill provided Generosity and enthusiasm loomed large in the Early Paleoindians and later people with a the man and he was informal with most peo- relatively superb lithic resource, Normanskill ple after about the first 5 or 10 seconds, treat- chert (or “flint”) in this case. In fact, West ing the newcomer, student, or visiting scholar Athens Hill was precisely one of the local alike with boundless enthusiasm and help, resources that made local human residence in when it was possible for him to provide it. the Hudson Valley possible under late Like Bill Ritchie, Bob was willing to send a Pleistocene conditions. It also had broader “babe-in-the-woods” researcher in western regional significance since this material has Vermont (and reportedly elsewhere) any and been identified at Early Paleoindian sites else- all reprints of his archaeological publications where in New York State, and New Jersey, that he could provide and I remember that he Connecticut, Massachusetts, Vermont, and kindly gave me various publications while I Pennsylvania, among potentially others. was still an undergraduate student. However, we need to recognize that some I had the privilege of meeting Bob very other lithic material might have alternatively early in my involvement in archaeology, given served instead, had Normanskill chert not the proximity of my ancestral home and some- been available, and we should not be lithic times research area in the Champlain Valley of determinists here. In this case, Normanskill western Vermont. As many may know, the chert may be the most obvious “critical” Hudson and the Champlain valleys constitute resource to survive in the archaeological a single interlinked and more or less continu- record, but it is unlikely that this or any other ous north–south lowland area, or travel corri- particular lithic material was absolutely criti- dor, from the Atlantic coast near New York cal to the Paleoindians. Chert and other City to the St. Lawrence Valley near Montreal. high–quality lithics were likely related to a For those of us who work in the area, we small but significant part of Paleoindian life- know that archaeological manifestations ways, as they are ethnographically known in sometimes changed significantly over the dis- various contexts, but the jury is still out on this tance of this north–south corridor, some 600 matter. In any case, let’s celebrate Bob’s con- kilometers overall. However, a variable syn- tributions to all of these fascinating topics! chronicity pertained over much of the region, especially in the northern two thirds or so, from the mid-Hudson Valley northward A CLOSING (BRIEF) TRIBUTE TO through the entire Champlain Valley to the St. ROBERT E. FUNK Lawrence River, during the entire span of Native American prehistory and early history. Coastal New York, around the lower Hudson A lot has been already said about Bob and Long Island Sound, for example, was typ- Funk in tribute, both before and after his ically another world for the Native people, at death. Nonetheless, I will close this foreword, least as it is seen archaeologically. perhaps long overdue by this point, with a Bob Funk and Bill Ritchie before him rec- few more personal comments in tribute to this ognized the interconnections between the

xxxv Hudson Valley and the Lake Champlain and issues, anything Native American and archae- Richilieu drainage to the north. None of the ological in other words (e.g., Funk 1996; local and regional Native people lived in Ritchie and Funk 1984). Bob was also strongly social isolation, and extensive east–west con- enough a positivist and a “true” scientist to tacts along the Great Lakes–St. Lawrence recognize where Ritchie had incorrectly inter- River drainage were matched by comparable preted one aspect or another of regional pre- contacts north–south between the Hudson history, as I heard him say. Ritchie was pre- Valley and Lake Champlain and the St. scient on innumerable topics (as was Bob) Lawrence Valley. Perhaps that’s why Bob and regarding regional prehistory, nearly all of the Bill looked to the north and east to Vermont major issues that we still face. However, the (and many other places) for comparative relative dating of Lamoka vis-à-vis the archaeological information, and perhaps that Laurentian tradition, and the Adena “migra- explains their generosity to Vermont tion” are good examples of Bob’s dissension, researchers, but I don’t think that’s all that for example, with quite good reason. there was to it. Instead, this was the manner of I am not precisely sure when I first met both of these men, quite clearly Bob and Bob, but I think Louise Basa introduced us at Ritchie too. Both of them helped anyone who a New York State Archaeological Association could establish an earnest interest in the meeting in 1976, but one slightly later and details and/or broad patterns of regional pre- characteristic vignette will nicely sum up Bob history. From personal contact and reports, I as a human being, or so I hope. Bob was to be know that Ritchie was a bit autocratic, but Bob the principal (“banquet”) speaker at the was not in the slightest. The exchange spring meeting of the Vermont Archaeological between them was humorous on the few occa- Society (VAS) in April 1979, and I was sions when I had the privilege of seeing them assigned to pick him up and bring him to the interact. VAS meeting. This was because we had met It would be hard to say who “won” in several years before and we were correspon- some of the exchanges that occurred between ding by that time, and he was already sending Ritchie and Funk. Bill Ritchie would have me reprints of his articles and offering useful been dominant and sometimes dismissive, advice. whereas Bob would have let Bill get his way as When I phoned Bob at his hotel room a general survival strategy, sometimes chal- ahead of time, he asked me to delay a little bit lenging and nibbling around the edges of so that he could take an invigorating morning Ritchie’s ideas, but in the end having the con- jog around Burlington and I was happy to viction to express his differences, sometimes accommodate him, our important guest in print. In reality, these two people did quite speaker, in most any way that I could. Well, I well together over more than a 20-year period, finally arrived to pick him up at the appoint- both in person and in melding differences in ed time. Bob was freshly showered, smiling their writings, but preserving some degree of from ear to ear, and ready to enthusiastically (mostly Bob’s) dissension, as occasionally talk about (as it happened to be this time) seen in the 1973 settlement pattern volume. Middle Woodland archaeology. Once again he From my own observations and col- gave me another stack of fascinating leagues’ accounts, Bob Funk and Bill Ritchie reprints–wow, what I guy, I remember think- were clearly friends. In Ritchie’s absence, I ing–reprints for a lowly undergraduate stu- have heard Bob defend Ritchie’s views and his dent like myself, but that was Bob. And his own personal concordance on various things, VAS talk was great too. We will miss him for establishing that they had a lot invested in a many reasons, but I think his generosity, wide array of research topics together. These boundless enthusiasm, and good humor loom ranged from Paleoindian to protohistoric largest for me.

xxxvi REFERENCES CITED Boisvert, R. A. 1998. The Israel River Complex: A Paleoindian Manifestation in Jefferson, Adovasio, J. M. New Hampshire. Archaeology of Eastern 1998. Miller Complex. In Archaeology of North America 26:97–106. Prehistoric North America: An Encyclopedia, Boldurian, A. T., and J. L. Cotter edited by G. Gibbon, pp. 524–527. Garland Publishing, New York. 1999. Clovis Revisited: New Perspectives on Paleoindian Adaptations from Blackwater Adovasio, J. M., A. T. Boldurian, and R. C. Draw, New Mexico. University Museum, Carlisle University of Pennsylvania, Phildelphia. 1988. Who Were Those Guys? Some Bonnichsen, R., G. Jacobson, R. B. Davis, and Biased Thoughts on the Initial Peopling H. W. Borns of the New World. In Americans Before Columbus: Ice-Age Origins, edited by R. C. 1985. The Environmental Setting for Carlisle, pp. 45–61. Ethnology Human Colonization in Northern New Monographs 12. University of Pittsburgh, England and Adjacent Canada in Late Pittsburgh. Pleistocene Time. In Late Pleistocene History of Northwestern New England and Adovasio, J. M., with J. Page Adjacent Quebec, edited by H. Borns, P. 2002. The First Americans: In Pursuit of LaSalle, and W. Thompson, pp.151–159. Archaeology’s Greatest Mystery. Random Geological Society of American Special House, New York. Paper 197. Boulder, Colorado. Adovasio, J. M., D. R. Pedler, J. Donahue, and Bonnichsen, R., D. Keenlyside, and K. R. Stuckenrath Turnmire 1998. Two Decades of Debate on 1991. Paleoindian Patterns in Maine and Meadowcroft Rockshelter. North American the Maritimes. In Prehistoric Archaeology in Archaeologist 19(4):317–341. the Maritime Provinces: Past and Present Anderson, D. G., and J. C. Gillam Research, edited by M. Deal and S. Blair, 2000. Paleoindian Colonization of the pp. 9–36. Council of Maritime Premiers Americas: Implications from an Reports in Archaeology 8. Fredericton. Examination of Physiography, Bradley, J. W. Demography, and Artifact Distribution. 1998. Origins and Ancestors: Investigating American Antiquity 65(1):43–66. New England’s Paleo Indians. Robert S. Asch Sidell, N. Peabody Museum of Archaeology, 1999. Prehistoric Plant Use in Maine: Phillips Academy, Andover, Paleoindian to Contact Period. In Current Massachusetts. Northeast Paleoethnobotany, edited by J. P. Brennan, L. A. Hart, pp. 191–223. New York State 1982. A Compilation of Fluted Points of Museum Bulletin 494. New York State Eastern North America by Count and Education Department, Albany. Distribution: An AENA Project. Binford, L. R. Archaeology of Eastern North America 1979. Organization and Formation 10:27–46. Processes: Looking at Curated Brose, D. S. Technologies. Journal of Anthropological 1994. Archaeological Investigations at the Research 35:255–273. Paleo Crossing Site, a Paleoindian Occupation in Medina County, Ohio. In

xxxvii The First Discovery of America: Archéologie d’ un Site de la Culture Plano. Archaeological Evidence of the Early Pàléo-Quebec 22. Recherches Inhabitants of the Ohio Area, edited by W. Amerindiennes au Québec, Montreal. S. Dancey, pp. 61–76. Ohio Archaeological Chapdelaine, C., and P. LaSalle Council, Columbus. 1995. Physical Environments and Cultural Brush, N., and F. Smith Systems in the Saint Lawrence Valley, 1994. The Martins Creek Mastodon: A 8000–3000 B.P.: A Multidisciplinary Paleoindian Butchery Site in Holmes Framework. In Archaeological Geology of County, Ohio. Current Research in the the Archaic Period in North America, edited Pleistocene 11:14–15. by E. A. Bettis, pp. 115–129. Geological Callahan, E. Society of America Special Paper 297. Boulder, Colorado. 1979. The Basics of Biface Knapping in the Eastern Fluted Point Tradition: A Collins, M. B. Manual for Flintknappers and Lithic 1975. Lithic Technology as a Means of Analysts. Archaeology of Eastern North Processual Inference. In Lithic Technology: America 7:1–180. Making and Using Stone Tools, edited by E. Carr, K. W. Swanson, pp. 15–34. Mouton, The Hague. 1989. The Shoop Site: Thirty-Five Years 1999. Clovis Blade Technology: A After. In New Approaches to Other Pasts, Comparative Study of the Kevin Davis Cache, edited by W. F. Kinsey and R. W. Moeller, Texas. University of Texas Press, Austin. pp. 5–28. Archaeological Services, Cox, S. L. Bethlehem, Connecticut. 1986. A Re-Analysis of the Shoop Site. Carr, K. W., and J. M. Adovasio Archaeology of Eastern North America 2002. Paleoindians in Pennsylvania. In Ice 14:101–170. Age Peoples of Pennsylvania, edited by K. Curran, M. L. Carr and J. Adovasio, pp. 1–50. 1984. The Whipple Site and Paleoindian Pennsylvania Historical and Museum Tool Assembalge Variation: A Commission, Harrisburg. Comparison of Intrasite Structuring. Carty, Frederick M., and Arthur E. Spiess Archaeology of Eastern North America 1992. The Neponset Paleoindian Site in 12:5–40. Massachusetts. Archaeology of Eastern 1996. Paleoindians in the Northeast: The North America 20:19–38. Problem of Dating Fluted Point Sites. The Cavallo, J. Review of Archaeology 17 (1):2–11. 1981. Turkey Swamp: A Late Paleo-Indian 1999. Exploration, Colonization, and Site in New Jersey’s Coastal Plain. Settling In: The Bull Brook Phase, Archaeology of Eastern North America Antecedents, and Descendants. In The 9:1–18. Archaeological Northeast, edited by M. A. Levine, K. E. Sassaman, and M. S. Chalifoux, E. Nassaney, pp. 3–24. Bergin and Garvey, 1999. Late Paleoindian Occupation in a Westport, Connecticut. Coastal Environment: A Perspective from Curran, M. L., and J. R. Grimes La Martre, Gaspe Peninsula, Quebec. Northeast Anthropology 57:69–79. 1989. Ecological Implications for Paleoindian Lithic Procurement Economy Chapdelaine, C. (editor) in New England. In Paleoindian Lithic 1994. Il y a 8000 à Rimouski: Paléoecologie et Resource Use, edited by C. J. Ellis and J. C.

xxxviii Lothrop, pp. 41–74. Westview Press, Region, edited by R. S. Laub, N. G. Miller, Boulder, Colorado. and D. W. Steadman, pp. 251–263. Custer, J. F. Bulletin of the Buffalo Society of Natural Sciences 33. Buffalo, New York. 1984. Delaware Prehistoric Archaeology: An Ecological Approach. University of 1992. Thedford, a Paleo-Indian Site in the Delaware Press, Newark, Delaware. Ausable River Watershed of Southwestern Ontario. University of Michigan Museum 1996. Prehistoric Cultures of Eastern of Anthropology Memoirs 24. Ann Arbor. Pennsylvania. Anthropological Series 7. Pennsylvania Historical and Museum 2001. Evidence for Late Paleoindian Commission, Harrisburg. Ritual from the Caradoc Site (AfHj-104), Southwestern Ontario, Canada. American 2001. Classification Guide for Arrowheads Antiquity 66 (2):267–284. and Spearheads of Eastern Pennsylvania and the Central Middle Atlantic. Pennsylvania Dent, R. J. Historical and Museum Commission, 2002. Paleoindian Occupation of the Harrisburg. Upper Delaware Valley: Revisiting Custer, J. F., and R. M. Stewart Shawnee Minisink and Nearby Sites. In Ice Age Peoples of Pennsylvania, edited by 1990. Environment, Analogy, and Early K. Carr and J. Adovasio, pp. 51–78. Paleoindian Economies in Northeastern Pennsylvania Historical and Museum North America. In Early Paleoindian Commission, Harrisburg. Economies of Eastern North America, edited by K. B. Tankersley and B. L. Isaac, pp. Dent, R. J., and B. E. Kauffman 303–322. Research in Economic 1985. Aboriginal Subsistence and Site Anthropology Supplement 5. Greenwich, Ecology as Interpreted from Microfloral Connecticut. and Microfaunal Remains. In Shawnee Davis, S. A. Minisink: A Stratified Paleoindian-Archaic Site in the Upper Delaware Valley of 1998. Debert/Belmont Complex. In Pennsylvania, edited by C. W. McNett, pp. Archaeology of Prehistoric North America: 55–79. Academic Press, Orlando, Florida. An Encyclopedia, edited by G. Gibbon, pp. 199–200. Garland Publishing, New York. Derev’anko, A. (editor and compiler) Deller, D. B. 1998. The Paleolithic of Siberia: New Discoveries and Interpretations. University 1989. Interpretation of Chert Type of Illinois Press, Urbana. Variation in Paleoindian Industries, Southwestern Ontario. In Paleoindian Dincauze, D. F. Lithic Resource Use, edited by C. J. Ellis 1993a. Fluted Points in the Eastern and J. C. Lothrop, pp. 191–200. Westview Forests. In From Kostienki to Clovis: Upper Press, Boulder, Colorado. Paleolithic-Paleo-Indian Adaptation, edited Deller, D. B., and C. J. Ellis by O. Soffer and N. D. Praslov, pp. 279–292. Plenum Press, New York. 1984. Crowfield: A Preliminary Report on a Probable Paleo-Indian Cremation in 1993b. Pioneering in the Pleistocene: Southwestern Ontario. Archaeology of Large Paleoindian Sites in the Northeast. Eastern North America 12:5–40. In Archaeology of Eastern North America, Papers in Honor of Stephen Williams, edited 1988. Early Paleo-Indian Complexes in by J. B. Stoltman, pp. 43–60. Mississippi Southwestern Ontario. In Late Pleistocene Department of Archives and History and Early Holocene Paleoecology and Archaeological Report 25. Jackson. Archeology of the Eastern Great Lakes

xxxix Dineen, R. J. 1990. Paleo-Indians. In The Archaeology of 1996. Holocene Environments in the Southern Ontario to A.D. 1650, edited by Hudson Valley. In A Golden Chronograph C. J. Ellis and N. Ferris, pp. 37-63. for Robert E. Funk, volume 1, edited by C. Occasional Publications of the London Lindner and E. V. Curtin, pp. 57–72. Chapter. Ontario Archaeological Society, Occasional Publications in Northeastern London, Ontario. Anthropology 15. Bethlehem, 1997. Variability in the Archaeological Connecticut. Record of Northeastern Early Doyle, R. A., N. D. Hamilton, J. B. Petersen, Paleoindians: A View from Southern and D. Sanger Ontario. Archaeology of Eastern North America 25:1–30. 1985. Late Paleo-Indian Remains from Maine and Their Correlations in 1998. Crowfield Cultural Complex. In Northeastern Prehistory. Archaeology of Archaeology of Prehistoric North America: Eastern North America 13:1–34. An Encyclopedia, edited by G. Gibbon, pp. 183–185. Garland Publishing, New York. Dumais, P. 2000. An Early Paleo-Indian Site Near 2000. The La Martre and Mitis Late Parkhill, Ontario. Archaeological Survey of Paleoindian Sites: A Reflection on the Canada Mercury Series Paper 159. Peopling of Southeastern Quebec. Canadian Museum of Civilization, Hull. Archaeology of Eastern North America 28:81–112. Ellis, C. J., A. C. Goodyear, D. F. Morse, and K. B. Tankersley Dumais, P., C. Larouche, and J. Poirier 1998. Archaeology of the Pleistocene- 1996. Les Sites Paléoindiens Recents de Holocene Transition in Eastern North La Martre et de Mitis: Reflexions sur le America. Quaternary International Peuplement Plano de l’Est du Québec. 49/50:151–166. Archéologiques 10:54–76. Fiedel, S. J. Eley, B. E., and P. H. von Bitter 1999. Older Than We Thought: 1989. Cherts of Southern Ontario. Royal Implications of Corrected Dates for Ontario Museum, Toronto. Paleoindians. American Antiquity Ellis, Christopher J. 64(1):95–115. 1989. The Explanation of Northeastern Fisher, D. C., B. T. Lepper, and P. E. Hooge Paleoindian Lithic Procurement Patterns. 1994. Evidence for the Butchery of the In Paleoindian Lithic Resource Use, edited Burning Tree Mastodon. In The First by C. J. Ellis and J. C. Lothrop, pp. Discovery of America: Archaeological 139–164. Westview Press, Boulder, Evidence of the Early Inhabitants of the Ohio Colorado. Area, edited by W. S. Dancey, pp. 43–57. 1998. Thedford II. In Archaeology of Ohio Archaeological Council, Columbus. Prehistoric North America: An Encyclopedia, Frison, G. C. edited by G. Gibbon, pp. 833–834. Garland Publishing, New York. 1988. Paleoindian Subsistence and Settlement During the Post-Clovis Times Ellis, C. J., and D. B. Deller on the Northwestern Plains, the Adjacent 1988. Some Distinctive Paleo-Indian Tool Mountain Ranges, and Intermontane Types from the Lower Great Lakes. Basins. In Americans Before Columbus: Ice- Midcontinental Journal of Archaeology Age Origins, edited by R. C. Carlisle, pp. 13:111–158. 83–106. Ethnology Monographs 12. University of Pittsburgh, Pittsburgh.

xl Funk, R. E. Funk, R. E., G. R. Walters, W. F. Ehlers, J. E. 1976. Recent Contributions to Hudson Valley Guilday, and G. G. Connally Prehistory. New York State Museum and 1969. The Archaeology of Dutchess Science Service Memoir 22. Albany. Quarry Cave, Orange County, New York. 1977. Early Cultures in the Hudson Pennsylvania Archaeologist 39 (1–4):7–22. Drainage Basin. In Amerinds and Their Funk, R. E., and B. Wellman Paleoenvironments in Northeastern North 1984. The Corditaipe Site: A Small America, edited by W. S. Newman and B. Isolated Paleo-Indian Camp in the Upper Salwen, pp. 316–332. Annals of the New Mohawk Valley. Archaeology of Eastern York Academy of Sciences 288. New York. North America 12:72–80. 1978. Post-Pleistocene Adaptations. In Gardner, W. M. Handbook of North American Indians, vol- 1983. Stop Me If You Have Heard This ume 15, The Northeast, edited by B. G. One Before: The Flint Run Paleoindian Trigger, pp. 16–27. Smithsonian Complex Revisited. Archaeology of Eastern Institution, Washington, D.C. North America 11:49–64. 1991. Late Pleistocene and Early Holocene 1989. An Examination of Cultural Change Human Adaptations in the Lower in the Late Pleistocene and Early Hudson Valley. In The Archaeology and Holocene (ca. 9200 to 6800 B.C.). In Ethnohistory of the Lower Hudson Valley and Paleoindian Research in Virginia: A Neighboring Regions: Essays in Honor of Synthesis, edited by J. M. Wittkofski and Louis A. Brennan, edited by H. C. Kraft, T. R. Reinhart, pp. 5–51. Archeological pp. 49–67. Occasional Publications in Society of Virginia, Richmond. Northeastern Anthropology 11. Bethlehem, Connecticut. 2002. The Paleoindian Problem Revisited: Observation on Paleoindian in 1993. Archaeological Investigations in the Pennsylvania (A Slightly Southern Slant). Upper Susquehanna Valley, New York State, In Ice Age Peoples of Pennsylvania, edited volume 1. Persimmon Press, Buffalo, New by K. Carr and J. Adovasio, pp. 97–103. York. Pennsylvania Historical and Museum 1996. Holocene or Hollow Scene? The Commission, Harrisburg. Search for the Earliest Archaic Cultures in Gardner, W. M. (editor) New York State. The Review of Archaeology 17 (1):11–25. 1974. The Flint Run Paleo-Indian Complex: A Preliminary Report, 1971–73 Seasons. Funk, R. E., D. W. Fisher, and E. M. Reilly Archeology Laboratory, Catholic 1970. Caribou and Paleo-Indian in New University, Washington, D.C. York State: A Presumed Association. Gardner, W. M., and R. A. Verrey American Journal of Science 268:181–186. 1979. Typology and Chronology of Fluted Funk, R. E., and F. F. Schambach Points from the Flint Run Area. 1964. Probable Plano Points in New York Pennsylvania Archaeologist 49 (1):13–46. State. Pennsylvania Archaeologist 34 Goodyear, A. C. (2):90–93. 1989. A Hypothesis for the Use of Funk, R. E., and D. W. Steadman Crytocrystalline Raw Materials Among 1994. Archaeological and Paleoenvironmental Paleoindian Groups of North America. In Investigations in the Dutchess Quarry Caves, Paleoindian Lithic Resource Use, edited by Orange County, New York. Persimmon C. J. Ellis and J. C. Lothrop, pp. 1–9. Press, Buffalo, New York. Westview Press, Boulder, Colorado.

xli 1999. The Early Holocene Occupation of 1988c. Paleo-Indian Sites South of Lake the Southeastern United States: A Ontario, Western and Central New York Geoarchaeological Summary. In Ice Age State. In Late Pleistocene and Early Holocene People of North America: Environments, Paleoecology and Archeology of the Eastern Origins, and Adaptations, edited by R. Great Lakes Region, edited by R. S. Laub, Bonnichsen and K. L. Turnmire, pp. N. G. Miller, and D. W. Steadman, pp. 432–481. Oregon State University Press, 265–280. Bulletin of the Buffalo Society of Corvallis. Natural Sciences 33. Buffalo, New York. Goodyear, A. C., and K. Steffy 1998. The Sugarloaf Site: Palaeo-Americans 2003. Evidence of a Clovis Occupation at on the Connecticut River. Persimmon Press, the Topper Site, 38AL23, Allendale Buffalo, New York. County, South Carolina. Current Research 1999. The Lamb Site: A Pioneering Clovis in the Pleistocene 20:23–25. Encampment. Persimmon Press, Buffalo, Gould, R. A., and S. Saggers New York. 1985. Lithic Procurement in Central Gramly, R. M., and R. E. Funk Australia: A Closer Look at Binford’s Idea 1990. What Is Known and Not Known of Embeddedness in Archaeology. about the Human Occupation of the American Antiquity 50 (1):117–136. Northeastern United States until 10,000 Graham, R. W., and M. Kay B.P. Archaeology of Eastern North America 18:5–31. 1988. Taphonomic Comparisons of Cultural and Noncultural Faunal Gramly, R. M., and J. Lothrop Deposits at the Kimmswick and Barnhart 1984. Archaeological Investigations of the Sites, Jefferson County, Missouri. In Late Potts Site, Oswego County, New York, Pleistocene and Early Holocene Paleoecology 1982 and 1983. Archaeology of Eastern and Archeology of the Eastern Great Lakes North America 12:122–158. Region, edited by R. S. Laub, N. G. Miller, Gramly, R. M., and G. L. Summers and D. W. Steadman, pp. 227–240. 1986. Nobles Pond: A Fluted Point Site in Bulletin of the Buffalo Society of Natural Northeastern Ohio. Midcontinental Journal Sciences 33. Buffalo, New York. of Archaeology 11 (1):97–123. Gramly, R. M. Grimes, J. R. 1980. Raw Materials Source Areas and 1979. A New Look at Bull Brook. “Curated” Tool Assemblages. American Anthropology 3 (1–2):109–130. Antiquity 45 (4):823–833. Grimes, John R., W. Eldridge, B. G. Grimes, 1982. The Vail Site: A Palaeo-Indian A. Vacaro, F. Vacaro, J. Vacaro, N. Vacaro, Encampment in Maine. Bulletin of the and A. Orsini Buffalo Society of Natural Sciences 30. Buffalo, New York. 1984. Bull Brook II. Archaeology of Eastern North America 12:159–183. 1984. Kill Sites, Killing Ground and Fluted Points at the Vail Site. Archaeology Hayden, B. of Eastern North America 12:110–121. 1979. Palaeolithic Reflections: Lithic 1988a. The Adkins Site: A Palaeo-Indian Technology and Ethnographic Excavation Habitation and Associated Stone Structure. Among Australian Aborigines. Australian Persimmon Press, Buffalo, New York. Institute of Aboriginal Studies, Canberra. 1988b. Discoveries at the Lamb Site, 1982. Interaction Parameters and the Genesee County, New York, 1986–7. Ohio Demise of Paleo-Indian Craftmanship. Archaeologist 38 (1):4–9. Plains Anthropologist 27 (96):109–123.

xlii Haynes, C. V., D. J. Donahue, A. J. T. Jull, and Julig, P. J. T. H. Zabel 1994. The Cummins Site Complex and 1984. Application of Accelerator Dating to Paleoindian Occupations in the Northwestern Fluted Point Paleoindian Sites. Lake Superior Region. Ontario Archaeology of Eastern North America Archaeological Reports 2. Ontario 12:184–191. Heritage Foundation, Toronto. Hill, P. J. Justice, N. D. 1997. A Re-examination of the Williamson 1987. Stone Age Spear and Arrow Points of Site in Dinwiddie County, Virginia: An the Midcontinental and Eastern United Interpretation of Intrasite Variation. States. Indiana University Press, Archaeology of Eastern North America Bloomington. 25:159–173. Karrow, P. F., and B. G. Warner 2002. A Study of Biface Technology at the 1990. The Geological and Biological Williamson Paleoindian Site (44DW1) in Environment for Human Occupation in Southeastern Virginia: An Examination of Southern Ontario. In The Archaeology of Biface Reduction Using the “Callahan Southern Ontario to A.D. 1650, edited by Model.” Journal of Middle Atlantic C. J. Ellis and N. Ferris, pp. 5-35. Archaeology 18:57–76. Occasional Publications of the London Holliday, V. T. Chapter 5. Ontario Archaeological 1997. Paleoindian Geoarchaeology of the Society, London, Ontario. Southern High Plains. University of Texas Kay, M. Press, Austin. 1998. Kimmswick. In Archaeology of Jackson, L. J. Prehistoric North America: An Encyclopedia, 1998. The Sandy Ridge and Halstead Paleo- edited by Guy Gibbon, pp. 419–420. Indian Sites: Unifacial Tool Use and Gainey Garland Publishing, New York. Phase Definition in South-Central Ontario. Kelly, R. L., and L. C. Todd University of Michigan Museum of 1988. Coming into the Country: Early Anthropology Memoirs 32. Ann Arbor. Paleoindian Hunting and Mobility. Jackson, L., and H. McKillop American Antiquity 53 (2):231–244. 1991. Approaches to Palaeo-Indian Keenlyside, D. L. Economy: An Ontario and Great Lakes 1991. Paleoindian Occupations of the Perspective. Midcontinental Journal of Maritimes Region of Canada. In Clovis Archaeology 16 (1):34–68. Origins and Adaptations, edited by R. Jones, B. D. Bonnichsen and K. L. Turnmire, pp. 1997. The Late Paleoindian Hidden Creek 163–173. Center for the Study of the First Site in Southeastern Connecticut. Americans, Oregon State University, Archaeology of Eastern North America Corvallis. 25:45–80. Kopper, J. S., R. E. Funk, and L. Dumont 2003. Human Adaptation to the Changing 1980. Additional Paleo-Indian and Northeastern Environment at the End of the Archaic Materials from the Dutchess Pleistocene: Implications for the Quarry Cave Area, Orange County, New Archaeological Record. Unpublished Ph.D. York. Archaeology of Eastern North America dissertation. Department of 8:125–137. Anthropology, University of Connecticut, Storrs.

xliii Kraft, H. C. 1999. Pleistocene Peoples of 1973. The Plenge Site: A Paleo-Indian Midcontinental North America. In Ice Age Occupation Site in New Jersey. People of North America: Environments, Archaeology of Eastern North America 1 Origins, and Adaptations, edited by R. (1):56–117. Bonnichsen and K. L. Turnmire, pp. 362–394. Oregon State University Press, Kuehn, S. R. Corvallis. 1998. New Evidence for Late 2002. Reading Between the Lines: Paleoindian–Early Archaic Subsistence Drawing the Boundaries of Paleoindian Behavior in the Western Great Lakes. Culture Areas in Midcontinental North American Antiquity 63 (3):457–476. America. In Ice Age Peoples of Lantz, S. W. Pennsylvania, edited by K. Carr and J. 1984. Distribution of Paleo-Indian Adovasio, pp. 79–89. Pennsylvania Projectile Points and Tools from Western Historical and Museum Commission, Pennsylvania: Implications for Regional Harrisburg. Differences. Archaeology of Eastern North Levine, M. A. America 12:210–230. 1990. Accommodating Age: Radiocarbon Laub, R. S. Results and Fluted Point Sites in 1995. The Hiscock Site (Western New Northeastern North America. Archaeology York): Recent Developments in the Study of Eastern North America 18:33–63. of the Late-Pleistocene Component. Loring, S. Current Research in the Pleistocene 12:26–29. 2002. “And They Took Away the Stones 2000. A Second Dated Mastodon Bone from Ramah:” Lithic Raw Material Artifact from Pleistocene Deposits at the Sourcing and Eastern Arctic Archaeology. Hiscock Site (Western New York State). In Honoring Our Elders: A History of Archaeology of Eastern North America Eastern Arctic Archaeology, edited by W. 28:141–154. W. Fitzhugh, S. Loring, and D. Odess, pp. Laub, R. S., M. F. DeRemer, C. A, Dufort, and 163–185. Contributions to Circumpolar W. L. Parsons Anthropology 2. National Museum of 1988. The Hiscock Site: A Rich Late Natural History, Smithsonian Institution, Quaternary Locality in Western New York Washington, D.C. State. In Late Pleistocene and Early Holocene Lothrop, J. C. Paleoecology and Archeology of the Eastern 1989. The Organization of Paleoindian Great Lakes Region, edited by R. S. Laub, Lithic Technology at the Potts Site. In N. G. Miller, and D. W. Steadman, pp. Paleoindian Lithic Resource Use, edited by 67–81. Bulletin of the Buffalo Society of C. J. Ellis and J. C. Lothrop, pp. 99–137. Natural Sciences 33. Buffalo, New York. Westview Press, Boulder, Colorado. Laub, R. S., and G. Haynes Lourandos, H. 1998. Fluted Points, Mastodons, and 1997. Continent of Hunter–Gatherers: New Evidence of Late Pleistocene Drought at Perspectives in Australian Prehistory. the Hiscock Site, Western New York State. Cambridge University Press, Cambridge. Current Research in the Pleistocene 15:32–34. Lowery, D. Lepper, B. T. 1989. The Paw Paw Cove Paleoindian Site 1988. Early Paleo-Indian Foragers of Complex, Talbot County, Maryland. Midcontinental North America. North Archaeology of Eastern North America American Archaeologist 9(1):31–51. 17:143–163.

xliv MacDonald, G. F. 2002. Thomas Wolfe and Eastern North 1968. Debert: A Paleo-Indian Site in Central American Paleoindians. In Ice Age Peoples Nova Scotia. National Museums of of Pennsylvania, edited by K. Carr and J. Canada Anthropology Papers 16. Ottawa. Adovasio, pp.159–166. Pennsylvania Historical and Museum Commission, Mason, R. J. Harrisburg. 1962. The Paleo-Indian Tradition in Meltzer, D. J., and B. D. Smith Eastern North America. Current Anthropology 3 (3):227–278. 1986. Paleoindian and Early Archaic Subsistence Strategies in Eastern North 1981. Great Lakes Archaeology. Academic America. In Foraging, Collecting, and Press, New York. Harvesting: Archaic Period Subsistence and McAndrews, J. H., and L. J. Jackson Settlement in the Eastern Woodlands, edited 1988. Age and Environment of Late by S. W. Neusius, pp. 3–31. Center for Pleistocene Mastodont and Mammoth in Archaeological Investigations Occasional Southern Ontario. In Late Pleistocene and Paper 6. Southern Illinois University at Early Holocene Paleoecology and Archeology Carbondale, Carbondale. of the Eastern Great Lakes Region, edited by Mochanov, Y. A., and S. A. Fedoseeva R. S. Laub, N. G. Miller, and D. W. 1996. Dyuktai Cave. In American Steadman, pp. 161–172. Bulletin of the Beginnings: The Prehistory and Buffalo Society of Natural Sciences 33. Palaeoecology of Beringia, edited by F. H. Buffalo, New York. West, pp. 164–174. University of Chicago McAvoy, J. M., and L. D. McAvoy Press, Chicago. 1997. Archaeological Investigations of Site Moeller, R. W. 44SX202, Cactus Hill, Sussex County, 1980. 6LF21: A Paleo-Indian Site in Western Virginia. Virginia Department of Historic Connecticut. Occasional Paper 2. American Resources Research Report Series 8. Indian Archaeological Institute, Richmond. Washington, Connecticut. McDonald, H. G. Morse, D. F. 1994. The Late Pleistocene Vertebrate 1998. Dalton. In Archaeology of Prehistoric Fauna in Ohio: Coinhabitants with Ohio’s Native America: An Encyclopedia, edited by Paleoindians. In The First Discovery of G. Gibbon, pp. 194–196. Garland America: Archaeological Evidence of the Publishing, New York. Early Inhabitants of the Ohio Area, edited by W. S. Dancey, pp. 23–41. Ohio Overstreet, D. F., D. J. Joyce, and D. Wasion Archaeological Council, Columbus. 1995. More on Cultural Contexts of McNett, C. W. (editor) Mammoth and Mastodon in the Southwestern Lake Michigan Basin. 1985. Shawnee Minisink: A Stratified Current Research in the Pleistocene 12:40–42. Paleoindian-Archaic Site in the Upper Delaware Valley of Pennsylvania. Academic Petersen, J. B. Press, Orlando, Florida. 1991. Archaeological Testing at the Sharrow Meltzer, D. J. Site: A Deeply Stratified Early to Late Holocene Cultural Sequence in Central 1989. Was Stone Exchanged Among Maine. Occasional Publications in Maine Eastern North American Paleoindians? Archaeology 8. Maine Historic In Paleoindian Lithic Resource Use, edited Preservation Commission, Augusta. by C. J. Ellis and J. C. Lothrop, pp. 11–39. Westview Press, Boulder, Colorado. 1995. Preceramic Archaeological Manifestations in the Far Northeast: A

xlv Review of Recent Research. Archaeology of 1969. The Archaeology of New York State. Eastern North America 23:207–230. Natural History Press, New York. Petersen, J. B., R. N. Bartone, and B. J. Cox 1983. The Mystery of Things Paleo- 2000. The Varney Farm Site and the Late Indian. Archaeology of Eastern North Paleoindian Period in Northeastern North America 11:30–33. America. Archaeology of Eastern North Ritchie, W. A., and R. E. Funk America 28:113–140. 1973. Aboriginal Settlement Patterns in the Petersen, J. B., and D. E. Putnam Northeast. New York State Museum and 1992. Early Holocene Occupation in the Science Service Memoir 20. The Central Gulf of Maine. In Early Holocene University of the State of New York. Occupation in Northern New England, edit- Albany. ed by B. S. Robinson, J. B. Petersen, and 1984. Paleo-Indians in New Perspective: A. K. Robinson, pp. 13–61. Occasional Comments on the Assembled Papers. Publications in Maine Archaeology 9. Archaeology of Eastern North America Maine Historic Preservation Commission, 12:1–4. Augusta. Robinson, B. S. Pollock, S. G., N. D. Hamilton, and R. 1992. Early and Middle Archaic Period in Bonnichsen the Gulf of Maine Region: Mortuary and 1999. Chert from the Munsungun Lake Technological Patterning. In Early Formation (Maine) in Palaeoamerican Holocene Occupation in Northern New Archaeological Sites in Northeastern England, edited by B. S. Robinson, J. B. North America: Recognition and Petersen, and A. K. Robinson, pp. 63–116. Occurrence. Journal of Archaeological Occasional Publications in Maine Science 26:269–293. Archaeology 9. Maine Historic Prufer, O. H., and R. S. Baby Preservation Commission, Augusta. 1963. Palaeo-Indians of Ohio. Ohio 2001. Burial Ritual, Groups, and Boundaries Historical Society, Columbus. on the Gulf of Maine: 8600-3800 B.P. Unpublished Ph.D. dissertation. Ridge, J. C. Department of Anthropology, Brown 2003. The Last Deglaciation of the University, Providence, Rhode Island. Northeastern United States: A Combined Sanders, T. N. Varve, Paleomagnetic, and Calibrated 14C Chronology. In Geoarchaeology of 1990. Adams: The Manufacturing of Flaked Landscapes in the Glaciated Northeast, edit- Stone Tools at a Paleoindian Site in Western ed by D. L. Cremeens and J. P. Hart, pp. Kentucky. Persimmon Press, Buffalo, New 15–45. New York State Museum Bulletin York. 497. New York State Education Seeman, M. F. Department, Albany. 1994. Intercluster Lithic Patterning at Ritchie, W. A. Nobles Pond: A Case for “Disembedded” 1953. A Probable Paleo-Indian Site in Procurement Among Early Paleoindian Vermont. American Antiquity 18 Societies. American Antiquity 59 (3):249–258. (2):273–288. 1957. Traces of Early Man in the Northeast. Shott, M. J. New York State Museum and Science 1990. Stone Tools and Economics: Great Service Bulletin 358. The University of the Lakes Paleoindian Examples. In Early State of New York. Albany. Paleoindian Economies of Eastern North

xlvi America, edited by K. B. Tankersley and B. 1989. Paleoindian Lithic Distribution in L. Isaac, pp. 3–43. Research in Economic the New England–Maritimes Region. In Anthropology Supplement 5. Greenwich, Paleoindian Lithic Resource Use, edited by Connecticut. C. J. Ellis and J. C. Lothrop, pp. 75–97. 1993. The Leavitt Site: A Parkhill Phase Westview Press, Boulder, Colorado. Paleo-Indian Occupation in Central Spiess, A. E., D. Wilson, and J. Bradley Michigan. University of Michigan 1998. Paleoindian Occupation in New Museum of Anthropology Memoirs 25. England—Maritimes Region: Beyond Ann Arbor. Cultural Ecology. Archaeology of Eastern 1996. Stage versus Continuum in the North America 26:201–264. Debris Assemblage from Production of a Stanford, D. Fluted Biface. Lithic Technology 21 (1):6–22. 1991. Clovis Origins and Adaptations: An 1997. Activity and Formation as Sources Introductory Perspective. In Clovis: of Variation in Great Lakes Paleoindian Origins and Adaptations, edited by R. Assemblages. Midcontinental Journal of Bonnichsen and K. L. Turnmire, pp. 1–13. Archaeology 22 (2):197–236. Center for the Study of the First Simons, D. B. Americans, Oregon State University, 1998. Gainey Complex. In Archaeology of Corvallis. Prehistoric Native America: An Encyclopedia, 1999. Paleoindian Archaeology and Late edited by G. Gibbon, pp. 194–196. Pleistocene Environments in the Plains Garland Publishing, New York. and Southwestern United States. In Ice Simons, D. B., M. J. Shott, and H. T. Wright Age People of North America: Environments, Origins, and Adaptations, edited by R. 1984. The Gainey Site: Variability in a Bonnichsen and K. L. Turnmire, pp. Great Lakes Paleo-Indian Assemblage. 281–339. Oregon State University Press, Archaeology of Eastern North America Corvallis. 12:266–279. Steadman, D. W. Snow, D. R. 1988. Vetebrates from the Late 1980. The Archaeology of New England. Quaternary Hiscock Site, Genesee Academic Press, New York. County, New York. In Late Pleistocene and Spiess, A. E., M. L. Curran, and J. R. Grimes Early Holocene Paleoecology and Archeology 1985. Caribou (Rangifer tarandus L) Bones of the Eastern Great Lakes Region, edited by from New England Paleoindian Sites. R. S. Laub, N. G. Miller, and D. W. North American Archaeologist 6 (2):145–159. Steadman, pp. 95–113. Bulletin of the Spiess, A. E., J. Mosher, K. Callum, and N. Buffalo Society of Natural Sciences 33. Asch Sidell Buffalo, New York. 1995. Fire on the Plains: Storck, P. L. Paleoenvironmental Data from the 1983. The Fisher Site, Fluting Techniques, Hedden Site. Maine Archaeological Society and Early Paleo-Indian Cultural Bulletin 34 (2):25–54. Relationships. Archaeology of Eastern North Spiess, A. E., and D. Wilson America 11:80–97. 1987. Michaud: A Paleoindian Site in the 1984. Glacial Lake Algonquin and Early New England–Maritimes Region. Palaeo-Indian Settlement Patterns in Occasional Publications in Maine Southcentral Ontario. Archaeology of Archaeology 6. Maine Historic Eastern North America 12:286–298. Preservation Commission, Augusta.

xlvii 1988a. The Early Palaeo-Indian 1990. Late Pleistocene Lithic Exploitation Occupation of Ontario: Colonization or in the Midwest and Midsouth: Indiana, Diffusion? In Late Pleistocene and Early Ohio, and Kentucky. In Early Paleoindian Holocene Paleoecology and Archeology of the Economies of Eastern North America, edited Eastern Great Lakes Region, edited by R. S. by K. B. Tankersley and B. L. Isaac, pp. Laub, N. G. Miller, and D. W. Steadman, 259–299. Research in Economic pp. 243–250. Bulletin of the Buffalo Anthropology Supplement 5. Greenwich, Society of Natural Sciences 33. Buffalo, Connecticut. New York. 1994. Devil’s Nose: A Parkhill Complex 1988b. The Late Wisconsinan Ice Margin Site in Western New York. Current and Early Paleo-Indian Occupation in the Research in the Pleistocene 11:51–53. Mid-Continent Region. Journal of 1995. Paleoindian Contexts and Artifact Midcontinental Archaeology 13 (2):259–272. Distribution Patterns at the Bostrom Site, 1997. The Fisher Site: Archaeological, St. Clair County, Illinois. Midcontinental Geological and Paleobotanical Studies at an Journal of Archaeology 20:40–61. Early Paleo-Indian Site in Southern Ontario, 1996. Ice Age Hunters and Gatherers. In Canada. University of Michigan Museum Kentucky Archaeology, edited by R. B. of Anthropology Memoirs 30. Ann Arbor. Lewis, pp. 21–38. University Press of 1998. Parkhill Cultural Complex. In Kentucky, Lexington. Archaeology of Prehistoric Native America: 1998a. Clovis Cultural Context. In An Encyclopedia, edited by G. Gibbon, pp. Archaeology of Prehistoric Native America: 624–627. Garland Publishing, New York. An Encyclopedia, edited by G. Gibbon, pp. Storck, P. L., and A. E. Spiess 161–163. Garland Publishing, New York. 1994. The Significance of New Faunal 1998b. Variation in the Early Paleoindian Identifications Attributed to an Early Economies of Late Pleistocene Eastern Paleoindian (Gainey Complex) North America. American Antiquity 63 Occupation of the Udora Site, Ontario, (1):7–20. Canada. American Antiquity 59 Tankersley, K. B., S. Vanderlaan, J. D. (1):121–142. Holland, and S. Bland Storck, P. L., and P. H. von Bitter 1997. Geochronology of the Arc Site: A 1989. The Geological Age and Occurrence Paleoindian Habitation in the Great Lakes of Fossil Hill Formation Chert: Region. Archaeology of Eastern North Implications for Early Paleoindian America 25:31–44. Settlement Patterns. In Paleoindian Lithic Thomas, P. A. Resource Use, edited by C. J. Ellis and J. C. 1992. The Early and Middle Archaic Lothrop, pp. 165–189. Westview Press, Periods as Represented in Western Boulder, Colorado. Vermont. In Early Holocene Occupation in Tankersley, K. B. Northern New England, edited by B. S. 1989. A Close Look at the Big Picture: Robinson, J. B. Petersen, and A. K. Early Paleoindian Lithic Resource Robinson, pp. 187–203. Occasional Procurement in the Midwestern United Publications in Maine Archaeology 9. States. In Paleoindian Lithic Resource Use, Maine Historic Preservation Commission, edited by C. J. Ellis and J. C. Lothrop, pp. Augusta. 259–292. Westview Press, Boulder, Trigger, B. G. (editor) Colorado. 1978. Northeast. Handbook of North American Indians, volume 15, William G.

xlviii Sturtevant, general editor. Smithsonian Smithsonian Institution, Washington, Institution, Washington, D.C. D.C. Walthall, J. A. Wilson, M. C., and J. A. Burns 1999. Mortuary Behavior and Early 1999. Searching for the Earliest Holocene Land Use in the North Canadians: Wide Corridors, Narrow American Midcontinent. North American Doorways, Small Windows. In Ice Age Archaeologist 20 (1):1–30. People of North America: Environments, West, F. H. Origins, and Adaptations, edited by R. Bonnichsen and K. L. Turnmire, pp. 1996. The Archaeological Evidence. In 213–248. Oregon State University Press, American Beginnings: The Prehistory and Corvallis. Palaeoecology of Beringia, edited by F. H. West, pp. 537–559. University of Chicago Wright, J. V. Press, Chicago. 1995. A History of the Native People of Wilmsen, E. N., and F. H. H. Roberts Canada, Volume 1 (10,000–1000 B.C.). Archaeological Survey of Canada 152. 1978. Lindenmeir, 1934–1974: Concluding Canadian Museum of Civilization, Hull, Report on Investigations. Smithsonian Quebec. Contributions to Anthropology 24.

xlix PROLOGUE

The little band was tired after hours of ther up the narrow valley bottom, the other up climbing over ridges, fording streams full of the slope of the high hill to their right, oppo- high, rushing water, and padding quietly site the escarpment. The sun rose above the through mist-shrouded forest groves in search hill as the hunters disappeared into the of new hunting grounds. The sun was setting woods. Just then, squealing erupted as behind the high, gray stone escarpment that women trapped and killed two small, furry loomed not far to their left, and lengthening animals that lived along the creek. They had shadows threw a chill into their very bones. also caught some fish and began preparing There were only about 25 people, including their catches for a mid-morning meal. women and children as well as men. Some The second party of hunters returned were elders. A few very small children were hours later, calling out gleefully. They had not carried by their mothers, while older children found any game, but proudly showed off a trudged along with the adults. They closed up skin pouch full of green and gray colored their caribou hide garments to keep out the stone that was ideal for the manufacture of breeze and then their leader, motioning for their weapon points and other stone tools. them to stop walking, began looking for a They had discovered a large, rich source of the place to camp for the night. They finally material on top of the hill. After the other decided on a dry, level site near a small brook. hunters returned, the fittest members of the Fires were started and some meat from the last band ascended the hill to stand on the sum- successful hunt was cooked and eaten. After mit, gazing out at the stone scarp and the building hasty lean-tos beneath the pale cloud-wreathed mountains far beyond it. moon, they lay on the hard ground and went They could even look down into the valley to sleep, only occasionally stirring as the where they were camped. howls of wolves and the cries of other crea- Men and women began filling bags, sewn tures echoed through the darkness. from small animal skins, with fragments quar- The next morning, the people were awak- ried from the high-quality stone that lay just ened by a familiar sound. Some distance away under the carpet of leaves. Using cobble ham- an enormous, long-nosed animal was making mers and antler flakers, others began reducing snuffling noises as it chewed on tree branches cores to the proper size and shape for weapon and twigs. While the band prepared for the points and knives. Two broken points, still on day’s tasks, some of the male hunters checked their wooden shafts, were discarded and fresh the contents of skin pouches. They were wor- new ones fastened in their place. One of the ried about their dwindling supplies of women, who had wandered off, called out in weapon points and roughed-out blanks of a puzzled tone, and people walked over to hard, shiny stone that was pried from a join her. She showed them some stone artifacts bedrock exposure farther down the great river that were unlike any made by the band. There they were following. After some debate, two was a brief flurry of excitement; had other parties of hunters picked up their spears and beings, perhaps very different from the people set out from the camp, one party headed far- of the band, visited the hill long before and

l made tools and weapons from its outcropping cliffs. The following morning, the people stone? broke camp and went on their way. But in At dusk, the group descended the hill and later years, and on many occasions, others like returned to their camp. There two adventur- them climbed to the hilltop to collect tool- ous boys reported finding different kinds of stone, manufacture stone artifacts, and admire toolstone, gray and black, at the base of the the scenic view.

li INTRODUCTION

At this writing (August 2002) it has been Kopper, Funk, and Dumont 1980; Funk and 32 years since the writer completed field work Steadman 1994; Steadman, Stafford, and Funk at the West Athens Hill quarry-workshop site 1997); the “Hallock” site1 in Orange County; in Greene County, New York (Funk 1972, 1976; the Zappavigna site, also in Orange County Ritchie and Funk 1973: 9–36). It was the first (Funk, et al. 2003); the Davis site on Lake Paleoindian site I excavated and by far the Champlain (Ritchie 1965: 19–22); the Twin most productive, even though I had the good Fields site in Ulster County (Eisenberg 1978); fortune of investigating several other sites of the Port Mobil site on Staten Island (Kraft the period during my service at the New York 1977); the Potts site near Oswego (Ritchie State Museum. 1965: 22–30; Gramly and Lothrop 1984); the At the time West Athens Hill was discov- Corditaipe site near Utica (Funk and Wellman ered by avocational archaeologist R. Arthur 1984); the Arc site near Oakfield, Genesee Johnson in 1962, only two small Paleoindian County (Tankersley, et al. 1997); the Lamb site sites (not counting isolated finds of diagnostic near Darien (Gramly 1999); and the Hiscock fluted points) were on record in New York site near Byron, Genesee County (Laub, State, the Davis and Potts sites (Ritchie 1965: Miller, and Steadman 1988) (Figure 1). 19–30). Ritchie (1953) had previously reported With so much new information, our on the Late Paleoindian Reagen site in north- understanding of Paleoindian regional stylis- eastern Vermont, and Witthoft (1952) had tic variation, settlement distribution, tool described investigations at the Shoop site on function, and lithic technology has improved the Susquehanna River drainage in central considerably. Unfortunately, our understand- Pennsylvania. The Bull Brook site in eastern ing of chronology and developmental patterns Massachusetts (Byers 1954, 1955) was the leaves much to be desired, knowledge of largest and best-known site of the period in material culture is confined almost entirely to New England. Since then many more stone tools, and subsistence remains are lack- Paleoindian sites have been discovered and ing on nearly all northeastern Paleoindian excavated in New York, New England, south- sites. Adverse environmental conditions over ern Ontario, and other parts of northeastern the last 10,000 years have caused the loss of North America. In New York the notable sites food remains and artifacts made of organic now include: Kings Road, located a few miles substances (wood, bone, antler, hides). There from West Athens Hill (Funk, Weinman, and are few published studies of chert quarries Weinman 1969; Weinman and Weinman 1978); used by the Paleoindians, as well as the organ- the nearby and as yet unpublished Swale site; ization of work parties and the methods they the Dutchess Quarry Caves 1 and 8 in Orange developed for extracting and processing the County (Funk, Fisher, and Reilly 1970; raw material. We are still forced to rely on

______1 The name “Hallock” refers to the discoverer, the late Russell Hallock, who found several fluted points on the surface of the site. The present owner wishes to avoid publicity, therefore his name is not used here.

1 Figure 1. Map of New York State showing location of West Athens Hill and other Paleoindian sites mentioned in the text. 1, Port Mobil; 2, 3 Dutchess Quarry Caves 1 and 8; 4, “Hallock”; 5, Zappavigna; 6, Twin Fields; 7, West Athens Hill; 8, Kings Road; 9, Swale; 10, Davis; 11, Corditaipe; 12, Potts; 13, Arc; 14, Hiscock; 15, Lamb.

ethnographic analogy and outright guess- descriptive statistics for the various artifact work concerning social structure, ideology, classes, omitted from the original report at artistic activity, and religious behavior, among editorial request. I have also conducted more other topics. detailed analysis, including size and weight This seemed like a good time to take measurements, of selected quarry and work- another look at the West Athens Hill site and shop debris (blocks, cores and flakes) collected to place it in the perspective of our rapidly from Area B. But the most important change growing knowledge of the earliest occupa- here is the addition of a detailed report on tions of the Northeast. My reevaluation Area C, in place of the very brief summary in includes a second presentation of the report Ritchie and Funk (1973: 15, 35). At the time of published in Ritchie and Funk (1973), which that publication, the recovered data and mate- dealt with the findings at two major loci, rials had only been studied in cursorily fash- Areas A and B, but with a number of stylistic, ion. I have completed a detailed analysis of grammatical, factual, and organizational the artifacts from Area C, plotted their loca- changes. Also included are tables presenting tions on the grid map, and revised the overall

2 site map, accurately showing the location of are sections on the sedimentary contexts, a Area C and its contour lines in relation to summary of geological history, descriptions of Areas A and B. Also included for the first time artifacts from Areas A, B, C, D, E, and F, an are formal and metrical analyses of quarrying analysis of stratigraphy, typology, and settle- byproducts and debitage from Area C. ment patterns, a comparison with other sites, Following this introduction, I describe the a summary of the site’s position in regional location of the site, the history of discovery, Paleoindian culture, a speculative model of the regional geography and geology, the site quarry-workshop activity, and finally a sec- description and setting, the research goals of tion on general conclusions. the project, and the details of excavation. Next

3 LOCATION AND DISCOVERY

West Athens Hill is a rocky ridge located The results of these preliminary examinations in Greene County, New York, north of the were published soon thereafter (Funk and town of Catskill and about two miles west of Johnson 1964). the Hudson River. In the spring of 1962, R. Test explorations continued sporadically Arthur Johnson, a New York Telephone in 1963–1965. During most of this time, a wide Company engineer and member of the Van assortment of scapers, knives, bifaces, ham- Epps–Hartley Chapter, New York State merstones, and utilized flakes, plus several Archaeological Association, learned that his Archaic projectile points, were collected with- company planned to construct a mobile tele- out finding more fluted points. In 1965, how- phone relay tower atop West Athens Hill. ever, two fluted points were recovered in the Suspecting the archaeological potential of the hollow between the main summit knoll and ridge, Johnson visited the site several times two smaller adjoining knolls. The deposits in during the clearing of trees from the summit the hollow seemed only partially disturbed by using a bulldozer. It was evident from the the bulldozing operations, and furthermore large quantity of debitage littering the surface intact areas there were stratified. Therefore, that the ridge was a chert quarry and work- the Anthropological Survey of the New York shop. Finally, in April 1963, Johnson was State Museum commenced preparations for rewarded with the discovery of two fluted full-scale systematic investigations at the site, projectile points, one complete, the other bro- which took place in the summers of 1966, ken in process, that clearly demonstrated the 1967, 1969, and 1970. Fortunately, by this time existence of a Paleoindian component on the the New York Telephone Company had decid- ridge. He also found several end and side ed, for the time being, not to install the relay scrapers of Paleoindian form. tower, the concrete foundation for which had William A. Ritchie (then New York State been constructed on the main knoll in 1963. Archaeologist) and the writer accompanied At this writing, after some 40 years, the Johnson on a visit to the site a few days after relay tower still has not been installed. Little his discovery, during which a surface collec- has changed since the excavations, except for tion was made and several test pits excavated. regrowth of trees and brush and occasional A few more scrapers, utilized flakes, and sev- digging by artifact looters. eral cobble hammerstones were recovered.

4 REGIONAL GEOGRAPHY AND GEOLOGY

The West Athens Hill site lies atop a and Valley Province is a belt of much-folded north–south trending ridge, one of a series of and faulted Paleozoic sediments, reaching linked parallel outcrops five miles long and from Tennessee on the south to Canada on the varying in height, that reaches its maximum north, for a total length of 1,200 miles. Its elevation of about 410 feet above sea level at width varies from 14 to 35 miles in New York the top of the hill. On the west it is separated State. The New York segment is 270 miles from the limestone face of the Helderberg long. Scarp by the narrow but fairly level valley of The Hudson valley, the Lake George the Hans Vosen Kill, which flows south to the Trough, and the Lake Champlain basin parent Catskill. On the east it overlooks the together make up the Hudson–Champlain mile-wide Athens Flat, through which flows Lowland, which constitutes the New York sec- the Corlaer Kill, another tributary of the tion of the Ridge and Valley Province. The Catskill. Between Athens Flat and the river are region is characterized by alternating valley knobs and ridges alternating with level areas floors and narrow ridges, the latter varying and the 60-foot deep main gorge of south- chiefly between 200 and 800 feet above sea flowing Murderer’s Creek. From north to level. These features resulted primarily from south, the villages of Coxsackie, Athens, and differential erosion of rocks of varying resist- Catskill lie along the river’s west shore at the ance, including shales, sandstones, dolo- eastern boundary of Greene County. stones, and limestones. The topography was Here the main channel of the Hudson considerably modified by glaciation during River is generally oriented north–south, the Pleistocene epoch. almost precisely parallel to such major New In its middle section from Catskill to England rivers as the Housatonic and Albany the Hudson Lowland is bordered on Connecticut. In its southward flow from its the west by an escarpment known as the headwaters to the Atlantic Ocean, the Hudson Helderberg, which intervenes between the River traverses four major physiographic river and the Catskill Mountains. To the east, provinces: the Adirondack Mountains, the the Lowland is bordered by the Taconic Hudson–Champlain Lowland, the Hudson Range. Highlands, and the Piedmont. Its waters are Within the Hudson–Champlain Lowland, augmented along the way by runoff from sev- large areas are covered by glacial drift. Many eral tributary watersheds, both inside and hills and terraces, especially along the east outside New York State. side to the river, are composed of drift. The From West Point, 55 miles north of its Pleistocene glaciation considerably modified mouth, to Glens Falls farther north, where it preglacial drainage patterns. A proglacial lake debouches from the Adirondack Mountains, which formerly extended from the Hudson the Hudson valley is located within the great Highlands to Glens Falls, known as Lake Ridge and Valley province of the eastern Albany, left behind extensive sand and clay United States, also known as the Folded deposits that attain their greatest breadth in Appalachians (Fenneman 1938). The Ridge the area of the mouth of the Mohawk River.

5 Upland soils of the region are characterized as abounded in such fresh-water food fish as “brown gravelly and stony loams,” derived brook trout, small-mouth bass, and walleyed from glacial drift. Defined soil types in hilly pike, but the largest amounts of protein were terrain include the Lackawanna Stony Loam, provided by spring runs of shad, alewives, the Dutchess Stony Loam, and the Cossayuna herring, striped bass, and other anadromous Stony Loam. The soils atop West Athens Hill species in what must have been astronomical conform to the Cossayuna Stony Loam, which numbers. Popular in later periods was the was derived from Hudson valley sandstones, Atlantic sea sturgeon, which often exceeds 100 shales, slates, and limestones and occurs pounds in weight. This species was confined below an elevation of 1,000 feet. But the flat, largely to tidal portions of the Hudson low-lying areas between the river and the (Brumbach 1986; Smith 1985). Helderberg Escarpment are covered by com- As far upstream as Stony Point, where pact, clayey silts deposited by glacial Lake Hudson River salinity is fairly high, oysters Albany. The resultant soils are “light brown thrived and were consumed in vast quantities silty loams” and are classified as Vergennes by the prehistoric Indians, especially those of Clay (Smith 1954). the Archaic stage (Udell 1962; Weiss 1971). At Poughkeepsie in the mid-Hudson val- Farther upriver, under fresh–water condi- ley the average annual temperature is 50.9 tions, there were extensive beds of river mus- degrees Fahrenheit, and the average annual sels, also utilized by the Indians. Plant foods precipitation is 40.8 inches. Average annual were diverse and abundant, including mast runoff is 20 inches, and average annual dis- products such as acorns, beechnuts, hazel- charge along the Hudson River is 14,400 cubic nuts, chestnuts, walnuts, and butternuts. feet per second in this area (Carter 1966). The environmental setting in New York, as Native vegetation of the Hudson valley first encountered by Europeans, had persisted south of Glens Falls conforms to the chest- with relatively minor variations through the nut–oak-yellow poplar zone of the Southern long time span of the Holocene epoch. The sit- Hardwood Forest. The flora of immediately uation was rather different at the close of the surrounding uplands, including the Catskill Pleistocene epoch as the climate began to and Taconic Mountains, are classified with the warm up. During glacial retreat from the mid- Northern Hardwoods, characterized as Hudson region about 15,000 years ago, the birch–beech–maple–hemlock (Braun 1950). land was first colonized by tundra plants such At the time of local European intrusion as lichens, sedges, grasses, and dwarf willows. during the early 17th century, the regional The succeeding forest cover, first dominated hardwood forests supported a great variety of by spruce and fir around 12,000 years ago, fauna, including such familiar species as gave way to a pine-oak assemblage, which by white-tailed deer, black bear, elk, beaver, 7500 years ago was finally replaced by the woodchuck, raccoon, otter, bobcat, gray fox, mixed hardwoods dominant in the area today. timber wolf, squirrel, fisher, muskrat, turkey, This succession is indicated by pollen spectra ruffed grouse, and many others. Migratory (Connally and Sirkin 1970, 1971, 1986; Cox birds, including Canada goose, were plentiful 1959; Newman, et al. 1968; Sirkin 1965). in season. The deer, elk, and bear, as the During late phases of glacial retreat from largest animals, produced the most meat per the Hudson valley, between about 15,000 and individual and were therefore the mainstays 10,000 years ago, glacial Lake Albany attained among game, in all aboriginal periods except a maximum elevation of over 330 feet. The that of the Paleoindians, who occupied a summit of West Athens Hill was an island in rather different habitat, when caribou and the lake. The water levels began dropping extinct megafauna were locally present. after 14,000 years ago, and the modern The Hudson and its tributaries ultimately drainage system was established by 10,300

6 years ago (Connally and Sirkin 1986; Dineen Onondaga formations. The Onondaga forma- 1986, 1996). Paleoindians had probably tion is dated to around 390 million years ago. entered the valley by around 11,000 years The Glenerie and Onondaga are limestones, ago.2 the Esopus and Carlisle are shales, and the The fauna comprised many species no Schoharie is mixed limestone and shale. longer present in the state, including mam- Overlying the Tristates, but occuring moth, mastodon, caribou, giant beaver, giant entirely west of the Helderberg, are shales and sloth, dire wolf, moose-elk, horse, and bison sandstones of the Hamilton Group, major (Fisher 1955; Martin and Klein 1989; Ritchie bedrock formations of the Catskills. These are 1965: Figure 3). Most of these animals were not known to produce cherts, and Middle and extinct by 10,000 B.P. Others, such as the cari- Late Devonian lithologies will not be further bou and musk ox, were able to survive by fol- described here. Beneath the Helderberg lowing the retreating ice northward to their Group is an uncomformity, meaning there was present subarctic and arctic habitats in Alaska an erosional or depositional gap between the and Canada. Helderberg and much older underlying strata. The area of present-day Greene County Those strata directly beneath the Helderberg was of great interest to prehistoric Indians cuesta are Ordovician shales, graywackes, silt- because of the abundant sources of chert, stones, and limestones. The Early Ordovician some of high quality, exposed in bedrock out- Deepkill formation is about 500 million years crops. This helps to explain the unusually old and directly underlies the Normanskill high frequency of chert quarries and work- Group, which is comprised of the sequential shops and the considerable abundance of sites Indian River, Mount Merino and Austin Glen of all periods. West Athens Hill and Flint Mine formations. Thrust faulting prior to the depo- Hill are the best-known quarry-workshop sition of the Helderberg Group rendered the sites in the region (Brumbach and Weinstein structure of this period rather complex. Rocks 1999; Parker 1924; Ritchie and Funk 1973). originally deposited east of the Hudson At first glance, the bedrock geology in River’s present channel were pushed west- Greene County (Figure 2) appears to be rela- ward during the Taconic orogeny, intruding tively simple and straightforward (Goldring into local formations and producing a and Cook 1943; Isachsen, et al. 1991; melange of different units from several time Ruedemann, Cook, and Newland 1942). periods (Brumbach and Weinstein 1999; Indeed, the stratigraphy west of the Hans Isachsen, et al. 1991). The severe folding and Vosen Kill is fairly easy to read (Figure 3). At fracturing of units in the area of Greene the base of the Helderberg carbonate sequence County east of the scarp is partly obscured by is the Early Devonian Rondout Formation, outwash overlain by the planar deposits from deposited around 408 million years ago. glacial Lake Albany that lie over and around Stacked atop the Rondout, from older to the tilted and upthrust Ordovician rocks. younger, are the Coeymans/Manlius, Chert occurs in the Deepkill, Indian River, Kalkberg, New Scotland, Becraft, and Alsen and Mount Merino formations. The great bulk limestones. Capping this pile are Middle of cherts at West Athens Hill, Flint Mine Hill Devonian rocks, more discontinously repre- (Brumbach and Weinstein 1999; Parker 1924), sented west of the escarpment and east of the and other quarries of the region derive from Catskills. These comprise the Tristates Group. the Mount Merino. Because the variable, but From early to late this group consists of the usually gray, green, black, and red cherts of Glenerie, Esopus, Carlisle, Schoharie, and the Mount Merino have traditionally been

______2 All radiocarbon dates listed here are uncalibrated; in other words, they are presented in radiocarbon years.

7 Figure 2. Geologic map of the middle Hudson Valley, showing the major bedrock types including the Helderberg series and the Normanskill formation. Chert occurs in the Normanskill, Kalkberg, New Scotland, Becraft, Alsen, Glenerie, Esopus, Schoharie, and Onondaga formations.

8 Figure 3. Idealized west to east cross-section of bedrock formations in the vicinity of West Athens Hill, from the Mount Marion shales at Leeds, to the Tristates series, to the Helderberg escarpment, to the Normanskill formation underlying, and in places, projecting above, Lake Albany sediments, ending at the village of Athens on the Hudson River.

assigned to the Normanskill formation of Scotland, Becraft, and Alsen formations may Ruedemann, Cook, and Newland (1942), that also occur in the talus. These cherts tend to be usageof Normanskill is continued here, gray in color. Other chert sources lie at the top although some writers now prefer the term of and west of the escarpment, and these are “Mount Merino” (Brumbach and Weinstein largely Onondaga cherts, of the “eastern” 1999). As noted above, the term variety, that is, the material is relatively dull “Normanskill” is also inclusively used by and uniform in surface appearance, with sub- Isachsen et al. (1991) to denote a group that dued mottling confined chiefly to areas of begins with the Indian River Formation, fol- gray interspersed with brown. This variety lowed by the Mount Merino Formation, and contrasts vividly with Onondaga of the west- concludes with the Austin Glen Formation. ern New York or “Divers Lake” variety, which Not usually mentioned by researchers are is characterized by a lustrous surface, and extensive quantities of readily available, high- brighter colors occurring in mottled areas of quality cherts derived from the weathering of gray, brown, and blue. Native quarries and the Helderberg escarpment. These consist pre- workshops occur in areas west of the New dominantly of the dark blue to deep black, York State Thruway. Kalkberg, Onondaga, homogeneous, relatively dull surfaced and other locally available cherts only rarely Kalkberg chert, which occurs as innumerable occur in Paleoindian assemblages in the mid- blocks in the deep talus at the foot of the scarp. Hudson valley, because Normanskill/Mount This chert could easily be confused with the Merino chert was apparently preferred. The dark gray to black Normanskill variety. Cherts Appendix presents descriptions of the major occurring above the Kalkberg zone in the New cherts occurring on eastern New York sites.

9 SITE DESCRIPTION AND SETTING

The summit of West Athens Hill is reached State Thruway runs along the top of the by a gravel road that ascends the relatively escarpment, just behind its rim; and looming gentle south slope from U.S. Route 9W (Figure in the background is the skyline of the Catskill 4). From the highest point atop the main knoll Mountains (Figure 5). The eastward view is one commands a fine westward view of the partially blocked by remaining trees on the picturesque valley of the Hans Vosen Kill and hill, but on a clear day it is possible to catch the Helderberg Escarpment. The New York glimpses of the rolling terrain between the hill

Figure 4. Map of the West Athens Hill site, showing major loci (Areas A, B, and C) and excavated areas.

10 U.S. Route 9W. Significant quantities of chert wastage are to be found on all of the upper slopes and benches. Bulldozer disturbance was most extensive on the main knoll, where a considerable volume of topsoil had been pushed over the edge along with trees. Bedrock projects above topsoil here in many places, occasionally displaying veins of chert. The ledge also reaches the surface on the less disturbed, fairly steep east slope of the knoll, as well as on the slopes and tops of the two minor knolls and the bench. Most areas of the site were probably never Figure 5. View looking west from the summit covered by more than a thin, patchy topsoil (Area A) at West Athens Hill, showing the Hans mantle like that present in undisturbed loci Vosen Kill valley, the Helderberg escarpment, and today. The excavation locus named Area B the Catskill Mountains in the background. was the chief exception, due to somewhat thicker deposits overlying bedrock. Massive quantities of shale were removed and the Hudson River, which is 2.3 miles dis- by commercial quarrying on the east and tant from West Athens Hill. Trees also screen south sides of West Ahtens Hill. These activi- the north and south lines of sight. ties apparently did not damage prehistoric Most of our explorations took place on the cultural resources, since no chert is known to summit area, which was littered with chert exist in the rock units at those lower eleva- and some artifacts resulting from the bulldoz- tions. ing operation, and in an adjoining small hol- The forest cover of the ridge and its com- low. The elevation of the main knoll is around panions consists mainly of oak, maple, cedar, 410 feet. At the base of the main knoll’s eastern ash, and associated species of the Southern slope is the aforementioned hollow, on the Hardwoods (Braun 1950). Since the summit east side of which are two small wooded rises. was cleared in 1962, sumac and other shrubs The crest of the northeastern rise has an eleva- and grasses have thrived on the site. No tion of 392 feet. The lowest part of the hollow springs have been observed on the sides of the is 380 feet above sea level. On the north side West Athens Hill, and the nearest stream is the hill drops rather steeply to a low saddle hundreds of yards distant from its base. Water between it and the next prominent rise. East of may, however, have been available to the pre- the small rises is a gradually steepening slope historic Indians from springs which have not to a topographic bench, which at elevations of been active for centuries. 339 to 350 feet overlooks a steep decline to

11 RESEARCH GOALS

No research design was specifically for- 4. To gather information on quarry tech- mulated for West Athens Hill as reported in nology and other on-site activities; Ritchie and Funk (1973). The investigations 5. To find and collect organic material were subsumed under the overall research suitable for radiocarbon dating; design of the settlement pattern project. 6. To find and collect faunal and floral Nevertheless, an implicit set of research goals remains representing subsistence prac- was as follows: tices; 1. To determine the importance of the 7. To gather sedimentological, palyno- Paleoindian component(s) relative to logical, pedological, and other data later occupations on the site; that might assist in reconstruction of 2. To determine whether parts of the site the late Pleistocene environment; and were stratified, thus contributing to 8. To place the site in the context of the separation and delimitation of regional subsistence-settlement pat- components; terns for the Paleoindian period. 3. To acquire a representative sample of As shown on later pages, we were partial- artifacts associated with the Paleoindi- ly successful in meeting all of these goals. an component(s);

12 EXCAVATION

The total area of the West Athens Hill site capped by a partially perforated wooden comprises about two acres (about 8,100 square block, on which its number was painted. meters), not counting possible extensions on Elevations were read at the ground surface the steep slopes or immediately adjacent low adjoining all stakes plus a series of scattered ground and rises. Because our explorations points, using the transit, and a plane table were to be focussed on the hollow between map was made. The contour map of West knolls, our north–south baseline, or EO line, Athens Hill (Figure 4) is based on these data. was set up there (Figure 6). This line almost In Area B, 20 whole and 7 partial 10-foot precisely bisected the area to be excavated squares were excavated, for a total area of within the hollow, which has a similar orien- 2,300 square feet or 207 square meters (Figure tation. A grid of 10-foot squares was estab- 4). This area includes 2,250 square feet opened lished in the hollow, which was designated as by the close of work in late July 1966, plus 50 Area B. The grid system was extended to the square feet within a test area dug in the fall of main knoll, west of the concrete tower foun- 1967. Portions of sections E0S10, E0N10, dation. This area, which yielded most of the W10N10, and E0N0 had been disturbed by surface-collected objects on the site, was unauthorized amateur digging prior to the referred to as Area A. 1966 season. Areas A and B were excavated by the In Area A, 5 whole and 5 partial squares writer and a small State Museum crew in the were excavated for a total area of 700 square summer of 1966 and the fall of 1967. In August feet (63 square meters). In addition, a number 1969, and again in 1970, I returned for a few of small areas had been dug by R. Arthur days with another crew to open a number of Johnson during testing from 1963 to 1965. In squares on the narrow topographic bench that overlooks Route 9W. This bench at West Athens Hill is referred to as Area C. Small loci, designated Areas D, E, and F, were tested on top of, and to the north and east of, the main hill in 1970. Due to the shallowness of bedrock and the predominant stony composition of the surface deposits, the grid stakes employed were pointed steel rods 21 inches long and 3/8s of an inch in diameter. These rods could be driv- en vertically into the ground, including bedrock crevices, with some degree of accura- cy. A transit and tapeline were used to Figure 6. View of excavations in Area B, looking emplace the grid. The baseline was oriented north. one degree east of north. Each stake was

13 Area C, 13 whole and 9 partial squares were troweling. At the end of the 1966 and 1967 excavated for a total area of 1,735 square feet, field seasons, and again at the close of the 1969 or 156 square meters. and 1970 investigations, several hundred Field methods were as follows. Brush was pounds of workshop debris had been hauled first cleared from the gridded areas. Where the away to the State Museum laboratory, in addi- topsoil, Stratum 1, still existed it was removed tion to artifactual material and soil and char- by troweling down either to bedrock or to the coal samples. top of Stratum 2, the yellowish-brown layer. Soil samples and pollen samples were Where Stratum 2 existed, it was excavated to taken from exposed cross-sections at various culturally sterile levels or, in some instances, places in Areas A, B, and C. Charcoal of possi- to bedrock. With a few exceptions the exact ble Paleoindian origin was collected assidu- horizontal and vertical position of artifacts ously, but there were several burned tree was recorded, noting any unusual associa- stumps on the site with charred roots pene- tions. The exceptions include some surface trating deeply into the ground. Possible fea- finds, and occasional items which were inad- tures were carefully recorded. Balks were left vertently thrown into level bags along with between squares in Areas B and C until all debitage. profiles were drawn. Photographic records in Since Area A had been skimmed by bull- black and white and color were made wherev- dozer, and thin topsoil directly overlay er appropriate. bedrock in most places, little debitage besides Much to our disappointment, no hearths, utilized flakes and cores was saved, in the storage pits, refuse bone, plant food remains belief that distributional studies of workshop or post molds were found at the West Athens activity would not be feasible for Area A. Parts Hill site. All uncalcined bone material must of Area B had been bulldozed or dug by col- have long since perished in the strongly acidic lectors. This fact and the sheer difficulty of soil (pH 4.5 to 5.5), but we were surprised at saving and transporting the masses of chert the complete absence of calcined bone and wastage encountered in the hollow excava- charred vegetal matter apart from burned tree tions led us to select only a few squares from stumps, since the occupants presumably which all cultural material would be saved. camped and ate briefly on the hill during Screening was not a regular part of field tech- quarrying. The absence of confirmed hearth nique, but samples were collected by careful features poses a major interpretive problem.

14 SEDIMENTARY CONTEXTS

The term “context” is more applicable ken sandstone, some cobbles, and chert debris than “stratigraphy,” at West Athens Hill since in its upper five to eight inches (12.8 to 20 cm), the artifact-bearing deposits varied consider- but it graded into a homogeneous, culturally ably in thickness, color, texture, distribution, sterile silt and gravel component in its lower and amount of disturbance, and as will be levels. In the northern part of Area A some tree seen, in some places the vertical distribution root disturbance was evident, and possible of artifacts may be lacking in significance. rodent burrows were noted. Bedrock projections in Stratum 2 evinced some untouched AREA A nodules of high-grade chert. Stratum 2 displayed vague horizontal gradations in color As the site map (Figure 4) shows, the exca- and texture. vated squares of Area A were for the most part Some 30 feet north of the S20 line a five- arranged in a rectangular north–south orient- foot-by-five-foot test square produced a some- ed block. Between sections W180S70 and what different stratigraphic picture. This W180S40, the sandstone bedrock was very square was located on the steepening north- close to the surface in Area A, and in fact was ern slope of the main knoll, hence gravity may already exposed in some places. Here a thin, have assisted in the natural movement of brown-colored topsoil, Stratum 1, in which a loose material from the summit. In this square, few weeds had taken hold, unevenly mantled Stratum 1 and the yellowish-brown grainy the stone at thicknesses ranging from zero to deposit were separated by a layer of fine light four inches (0 to 10 cm). This apparent humic brown, powdery soil containing chert chips layer contained a high percentage of chert but little broken stone. The yellowish-brown debitage and crushed sandstone fragments. It layer was full of rock fragments, but it was was probably somewhat thicker prior to bull- culturally sterile. The light brown, powdery dozer disturbance. Occasionally, small pock- stratum is referred to as Stratum 2A, because ets of yellowish-brown soil invested shallow it seems to correlate with the upper portion of depressions in bedrock below Stratum 1. Stratum 2 in the nearby excavations. Artifacts were present in small quantities, In the main excavation, utilized flakes and occurring wholly on the surface or in the top- other artifacts in Strata 1 and 2 were observed soil. to cluster tightly around stake W170S30 when In sections W180S30, W170S30, and plotted on a map, suggesting a concentration adjoining squares the yellowish-brown zone, within an aboriginal feature. No pit outlines, Stratum 2, was thicker, covering the bedrock post molds, or other evidence of a feature to varying depths. Here Stratum 1 varied from were observed, however. three to four inches (7.5 to 10 cm) deep. It con- In sections W180S60 and W180S70 an tained fairly numerous chert chips, cores, a interesting feature (No. 1) was excavated few hammerstones, and other artifacts. (Figure 7). This was an elongated, bathtub-like Stratum 2, a yellowish-brown, silty deposit depression in the bedrock, filled with dark with some small gravel, contained much bro-

15 Figure 7. View of excavations in Area A, showing presumed quarry pit, looking north. J. Cynthia Weber holds stadia rod for scale. brown humic soil containing bits of sand- associated. It is possible, though very unlikely, stone, a few chert chips, and 16 cobble ham- that bones and offerings of organic materials merstones. It measured 9 feet long, 3 feet had been placed in the pit along with the ham- wide, and 3.5 feet deep (274 cm by 91 cm by merstones but have long since decayed away. 107 cm). Only a few small chert inclusions were observed in the walls of the trough. AREA B Possibly it was the site of a good chert vein The investigated part of the gully, or hol- quarried to exhaustion by the Indians. But one low, lay between the S20 and N90 lines. This might also consider another hypothesis, that area is contoured like a small trianguloid the feature was used for storage. It doesn’t basin, narrow at the south end and broaden- seem likely or practical, however, for the resi- ing toward the north. It is bounded on the dents to store the hammerstones in the pit in west by the steep eastern slope of the main anticipation of their later use since glacial cob- knoll, and on the east by one of the smaller bles were abundant locally. There is also no knolls. At its extreme south end the gully evidence it funtioned as a burial pit, since no swings south-southwest, funneling into a nar- human bones or mortuary offerings were

16 Figure 8. West profile of section W10N0. Visible Figure 9. West profile of section E20N0. Readily are the dark brown topsoil, stratum 1, and the apparent is the concentration of debitage in the underlying yellowish-brown zone, stratum 2. topsoil (stratum 1) and in the upper levels of the Debitage is visible in stratum 1 and at the top of yellowish-brown zone (stratum 2). The sterile lower stratum 2. portions of stratum 2 rest on bedrock. row defile. On the north, it extends to the stains exposed facing upward. Much of this brink of the steep slope of the hill. disturbance was essentially superficial, con- The floor of the excavated area was fairly fined to the upper inch or two of Stratum 1, level toward the center, along the north–south and can be attributed to the churning of bull- baseline, but sloped upward with increasingly dozer treads. In some squares, however, the steep gradient to the east and west, merging bulldozer blade seems to have scraped or with the sides of the adjacent rises. As previ- gouged into the soil, wreaking havoc with the ously mentioned, bedrock projections were stratification, and reaching even to the shal- fairly common on the shoulders of the knolls. low bedrock in the southern sections between Chert inclusions were visible in some of them. the S10 and S20 lines. Evidence that whole Before excavation, the hollow, under- trees were torn from the ground was present standably more moist than higher elevations, in a few places. The best-preserved deposits was filled by a heavy overgrowth of grasses, were in the central and northern parts of the berries, briars, and sumac, which had to be grid. Here machine damage seemed minimal. removed prior to emplacing the grid. The stratigraphic picture in Area B was It was obvious in scraping off the sod and basically the same as in Area A, with the impor- troweling into the topsoil that there was con- tant difference that the hollow was in an accu- siderable variation in the extent of bulldozer mulative, rather than erosive, geological situa- disturbance. In some places the topsoil tion (Figures 8, 9, 10). Hence the deposits were seemed untouched, firm and compact, occa- thicker than on the main knoll. sionally laced by the intact roots of trees. In Stratum 1, the dark brown humic topsoil, other areas this layer was loose, contained irregularly covered by a thin sod, varied con- twigs, broken roots, and other plant debris; siderably in thickness from one inch on the and fragments of sandstone and chert, once sloping edges of the hollow to 12 inches (30 half-buried so that their exposed surfaces cm) in its center. Over 50 per cent of its vol- were weathered and clean while their buried ume comprised chert debitage and sandstone portions were soil-stained, had been displaced fragments. This layer was easily distinguished and tumbled into new positions with the soil from the deposits on which it rested. Most of

17 Figure 10. Stratigraphic profiles in Area B.

stituent was a silt apparently derived from the regional bedrock by glacial action. In postglacial times some of it may have collected in the hollow through erosion and soil creep. Scattered through the silt in its lower portions were small angular fragments of sandstone, some pebbles and cobbles of exotic material, and a few natural spalls of chert. The upper few inches of the Stratum 2 generally contained a high percentage of such fragments, plus many larger slabs, and a significant quantity of debitage and artifacts. There were no sharp breaks between the upper levels and lower levels in terms of color or texture, but Figure 11. Fluted point uncovered in stratum 1, chert flakes, cores, artifacts, and slabs termi- section E10N20. nated abruptly when a certain plane was reached in excavation. Two squares (E10N20, E10N30) along the eastern side of the grid, the Paleoindian artifacts, including fluted located on the lower slope of the northeastern points (Figures 10, 11), were found in it. knoll, were not culturally productive. Here Stratum 2 showed a certain amount of hor- topsoil was thin, and only a few flakes izontal and vertical variation in color, texture, occurred in Stratum 2. and composition. In a majority of squares, the In squares W10N20, W10N30, E0N20, and undisturbed zone was yellowish-brown to E0N30, all located in the central, flattest part reddish-brown in color, and the primary con-

18 of the basin, the deposit became a bit more Other possible features consisted of tight complicated. Within a roughly circular area concentrations of chert wastage, largely con- about 20 feet (6 m) in diameter, Stratum 2 was fined to circular or oval areas about 12 inches separable into two easily distinguished sub- (30 cm) across and seeming to occur in shal- zones, 2A and 2B (Figure 10). Near the east low depressions within the top of Stratum 2; a edge of section E0N20, Stratum 2A appeared postmold-like feature 8 inches (20 cm) in as a thin lens between Stratum 1 and the yel- diameter first noted at the base of Stratum 1, lowish-brown, culturally sterile lower deposit, identified by its tan-colored fill (quite unlike thickening to the west. This subzone, light the usual postmold), extending 17 inches (43 brown in color, was a dry, powdery, loose, cm) into Stratum 2, nearly straight-sided in finely textured silt containing sandstone frag- cross-section and narrowing to a blunt point ments, debitage and artifacts. It was thickest— at the base; concentrations of charcoal, proba- about 6 inches (15 cm)—near the E0N30 stake. bly of modern origin considering the Within a radius of approximately 10 feet (3 m) unburned as well as burned wood they con- from the stake on the north, south and west tained; and some irregular depressions filled this enigmatic lens merged by subtle changes with brown humic soil identical with that of in color and texture into the upper, culturally Stratum 1. Only the debitage and artifact con- productive levels of the main Stratum 2 bed. centrations, and a handful of definitely fire- Thus, there is good reason to consider Stratum cracked cobble fragments (manifesting irregu- 2A as a special, localized facies of Stratum 2. It lar, angular fractures and all-over reddened may have been truncated by bulldozing oper- surfaces), were clearly the result of prehistoric ations in sections E10N20 and E10N30. The human activity. underlying yellowish-brown zone designated Stratum 2B, devoid of cultural remains, corre- AREA C sponded to the lower portion of Stratum 2 in The topographic bench locus is about 60 the areas where Stratum 2A was lacking. feet (28 m) wide and situated part way down Mineralogical examination of samples the eastern slope of the hill. The deposits that from Stratum 2 in Areas A and B was carried overlie bedrock there are thinner than in the out in 1970 by M. Raymond Buyce, then hollow (Figure 12). The dark, grayish-brown Curator of Mineralogy, New York State (10YR 4/2) topsoil, Stratum 1, was two to Museum and Science Service. His analysis eight inches (5 to 20 cm) thick, averaging confirmed the affinities of the samples to the about four (10 cm). Stratum 2, of pale-brown local bedrock. He described the sediment as color (10YR 7/4), varied in thickness but was fine-grained weathered sandstone and silt- rather shallow, since bedrock was encoun- stone. The constituents are largely quartz, tered in every excavated square, protruding with a minor component of limonite and up into Stratum 1 and sometimes reaching the traces of mica. surface. Stratum 2 contained much rubble A possible feature (No. 2) was noted in the derived from the bedrock. north half of section E0N20 and the south half Stratum 1, consisting mainly of silt and of section E0N30, in the central portion of gravel, but with some humus content, yielded Stratum 2A. Here was found a very high con- large quantities of debitage and over 150 arti- centration of chert flakes and cores, a few facts. Unlike the corresponding deposit in angular rock fragments, possibly fire-cracked, Areas A and B, Stratum 2 in Area C was large- and scattered flecks of charcoal, all of which ly devoid of cultural remains. In some excava- were carefully collected. The concentration tion units, however, artifacts and debitage appeared to be about five feet (1.5 m) in diam- were occasionally recovered from apparently eter. Again, no pit outlines or other traces of undisturbed upper portions of Stratum 2. deliberate aboriginal construction were seen.

19 Figure 12. View of excavations in Area C, showing cross-sectioned portion of dark brown colored stratum 1, resting on yellowish-brown stratum 2. The great majority of artifacts and debitage occurred in Figure 13. Exposed bedrock feature in Area C, stratum 1. showing irregular edges where chert was quarried off, detached fragments adjoining mass, and bat- These may have been introduced into Stratum tered areas (outlined in chalk) used as an anvil 2 from Stratum 1 by the action of tree roots, platform. tree falls, animal burrowing, frost heaving, and the activities of prehistoric inhabitants. broken off, presumably by vigorous hammer- The bedrock exposures were rich in chert ing, leaving irregular, sometimes scalloped veins. As in Area B, Stratum 2 occasionally edges instead of the otherwise straight, produced pebbles and cobbles of quartzite smooth erosionally rounded edges. Detached and gneiss, apparently imported by glacial angular fragments of chert lay around and action. between the slabs. The flat top surfaces of Certain areas where bedrock protruded these slabs were also generally smooth except above or lay just under modern ground sur- for several locations showing scarring or mul- face would have provided the Indians with tiple pitting (Figures 13–15). Evidently the sources of raw material for artifacts and also scarred areas were the result of hammering as platforms for stone working. In section with cobble tools during use of the rock mass- E260S20 a mass of bedrock projected upward es as anvils—perhaps not only for chert-knap- at an angle of about 75 degrees and contained ping but for breaking open animal bones for a six-inch-thick (15 cm) vein of high-quality, marrow or making bone tools. Although no greenish-gray Normanskill chert. Some of this hammerstones were directly associated, sever- chert had been intentionally removed, as evi- al were found within a few feet of the out- denced by loose blocky fragments surround- crops. ing the outcrop. Some fragments may have There were no clearly defined features been loosened by natural agents but others such as fire hearths, storage pits, or post could not have fallen directly onto the posi- molds, in the Area C deposits. There was some tions where they were found. Also, they dis- debate among the team concerning the possi- played irregular edges where they had been bility that certain angular fragments of rock forcibly detached from the bedrock. could be called fire-cracked, and also whether More impressive was the activity area in slightly reddened soil in some units was from section E300S30. Near-horizontal large exposure to prehistoric fires. But it proved bedrock segments, still in situ but partially impossible to confidently identify hearths, detached from the underlying matrix, dis- fire-cracked rocks, and other such phenomena played places where chert blocks had been at West Athens Hill. Questions were also

20 Figure 14. Map of the quarried chert outcrop in Area C.

21 were located several hundred feet north of Area C, and at elevations slightly below it. They were also well north and slightly east of the main knoll and Area B. Stratigraphy at these three loci was very similar to that at Area C. Areas E and F were quarry-workshop loci, undisturbed in historic times, and the depth of the artifact-bearing deposit varied, in some places up to 15 inches (38 cm) thick. The deposits were largely quarry debris. Refered to as Stratum 1, the cultural deposits overlay Stratum 2, the yellowish-brown sand and gravel of variable thickness, and overlying chert-rich bedrock. In Area F, there were sev- Figure 15. Close-up view of the areas on the bedrock mass showing anvil-pitting. eral large, moderately deep depressions that were interpreted as Indian quarry pits. They were not a result of tree falls, and were filled raised about possible cultural stratigraphy chiefly with chert quarrying debris at least 15 within the shallow Stratum 1 deposit in Area C, to 20 inches (38 to 50 cm) thick. The largest and one crew member attempted to demon- depressions were approximately 20 feet (6 m) strate such vertical patterning by piece-plot- in diameter (Figure 16). Our limited tests indi- ting all debitage in one unit (section E300S60). cated that Stratum 2 was lacking in artifacts My analysis of the data from that unit failed to and workshop debris, but this inference must convince me that there were real stratigraphic be confirmed by more extensive excavation. patterns within Area C, but one difference from other parts of the site consisted of an unusually high frequency of non-local rocks in that unit; there were boulders of fossiliferous Becraft limestone, and cobbles of Alsen and Kalkberg chert and Oriskany sandstone. G. Gordon Connally identified these and suggested they were glacial erratics, first washed into the main Hudson valley along energetic streams and then transported down the valley by glacial ice.

AREAS D, E, AND F Area D was a five-foot-by-five-foot test square located on the northern slope of the Figure 16. View of chert quarry pit in Area F. main knoll, north of Area A. Areas E and F

22 SUMMARY OF GEOLOGICAL HISTORY

The depositional history of the West was slowed by the heavy carpet of debitage, Athens Hill site after the final advance of the the compact silt till sediment, and outcrops of Wisconsinan ice is interpreted as follows.3 As bare rock. the glacier withdrew, it left behind an exten- In the hollow, Area B, the Paleoindian liv- sive mantle of pulverized shale and sandstone ing floor was probably first established at the drift, varying in thickness, that had been level of the base of Stratum 2A, or several derived largely from the Normanskill group inches below the present-day top of the yel- of rocks north of the site in Greene County lowish-brown portions of Stratum 2. Through and Albany County. The occasional exotic the years, the action of rainfall, seasonal vari- pebbles and cobbles of quartzite and gneiss ations in temperature, and gravity caused sur- found in Stratum 2 date to this period. ficial portions of the silt to move from the side Immediately thereafter, weathering and ero- slopes toward the middle of the depression, sion took effect, slowly removing some of the partially covering the concurrently building finer-grained sediment from the higher layer of debitage. At the same time, lighter ground and depositing this material a little at wind- or water-deposited dust and silt built a time in the hollow and on other areas of low, up the lens of Stratum 2A in the central area of level ground. During this period, Paleoindian the hollow. On its western periphery, this hunters arrived on the scene, and at various material mingled and merged with the yel- times quarried chert from bedrock exposures. lowish-brown silt and angular stone which They scattered chert tailings and chipping was moving downslope from the main knoll. debris around the site; made, lost, and dis- After the departure of Paleoindians, some carded artifacts; and carried out other tasks. debitage still overlay the silt deposit. Soil- Eventually they left, never to return, probably forming agents, including vegetative growth because their cultural pattern evolved into an and decay, bacterial action, and the move- as-yet-undetermined new pattern. Subse- ments of micro- and macro-fauna, began to quent visits were made by Archaic and bury the workshop detritus under a humus Transitional period Indians. Meanwhile, dur- layer, but this process was not yet fully coming the centuries after the disappearance of pleted when the telephone company erected Paleoindians a pine-oak forest, then a modern its relay station foundation. deciduous forest cover, replaced the spruce-fir The process of soil deposition inferred for parkland that had established a foothold after Areas A and B probably applied, in large part, glacial retreat (Connally and Sirkin 1971, to Area C. However, it is a mystery that 1986). This vegetation contributed to forma- Stratum 2 in Area C was lacking an occupation of a humus-rich topsoil, but the process tional zone, as contrasted with occasional

______3 For assistance in interpreting geological phenomena at the site, I am indebted to glacial geologist G. Gordon Connally, who examined the deposits on August 25, 1970. He referred to Stratum 2, the yellowish-brown deposit, as a “silt till.”

23 finds of artifacts and flakes probably intrusive expected the deposits at the juncture of slope from Stratum 1. Although there was less accu- and bench were thicker than elsewhere on the mulation of the bench deposits over bedrock bench, there was no rise to the east which than in Area B, Stratum 2 on the level top of could have contributed sediment to the the summit knoll contained some cultural deposits, in a manner analogous to that in the remains. Perhaps the local topography in Area hollow. Also, because there was a slight C accounts for this situation. The topographic depression between the gridded area and the bench is wider than the floor of the hollow; the base of the slope, some material from the side highest part of the bench is some 30 feet from of the hill may have washed or crept north- the base of the slope behind it, within our grid ward along the sloping bench, rather than fan- area; and the bench slopes gently downward ning out in the area of our excavations. to the north. Thus, although as might be

24 ARTIFACTS FROM AREAS A AND B

Descriptions of artifacts from the 1966 not appear in the following summaries and excavations in Areas A and B are presented tables, but are listed at the conclusion of the here as reported in Ritchie and Funk (1973) chipped stone descriptions. Of the total of but with minor changes. Artifacts from Area C 1,535 artifacts, 665 are deliberately chipped are described separately on later pages. The stone items. An additional 685 objects are uti- artifact illustrations originally published are lized flakes and cores (inclusive of 5 pieces included here, but they are complemented by esquillee), and there are 185 rough stone tools. new illustrations of representative artifacts No artifacts of other materials such as bone, from Area C. antler, wood, or shell were found. Most of the material to be described is in A small number of later prehistoric tools the collections of the New York State Museum. have been recovered from the West Athens Before his death in 1994, R. Arthur Johnson Hill site. A drill tip and 10 projectile points or donated his own collection to the museum. In point fragments—five of which were found on the 1960s small surface collections were also the surface in Area A, while the remainder donated by Paul and Thomas Weinman and were unearthed from Stratum 1 in Area B— by John Forstenzer. The late John McCashion represent relatively late occupations. They are also made part of his collection available for chiefly Late Archaic to Transitional types, and study. will not be further described here (Figure 17). The West Athens Hill site has been looted by various people since its first discovery and CHIPPED STONE subsequent publication. Some of these collectors have come from as far away as New Bifaces England. It is rumored that several fluted Bifaces were placed in four main groups, points have been found by these individuals relying on several criteria.4 These groups rep- in the course of surface hunting and digging resent successive stages in manufacture, mod- in various places on the site. It seems certain ified from chert nodules, vein plates, and that the nature and extent of these finds will blocks picked up or quarried by the Indians never be available to science, studied in detail, on the site. The first two groups, Stages 1 and and added to the published record. 2, are blanks or preforms which are relatively A total of 1,535 recognizable artifacts was crude, thick, percussion-flaked objects illus- recovered from Areas A and B at West Athens trating early stages in roughing out and shap- Hill in the five years of investigation from ing an artifact with a finished form in mind 1963 through 1967. This total differs from the (cf. Callahan 1979; Fitting, De Visscher, and figure of 1,493 published in 1973, because 42 Wahla 1966: 39–46). Stage 3 objects represent items donated by R. Arthur Johnson in 1979 the near-final form, but in the case of projectile are reported here for the first time. They do points and biface knives, some final modifica-

______4 In Ritchie and Funk (1973), these groups or “stages” were designated A, B, and C.

25 Figure 17. Archaic and Transitional projectile points and a drill from Areas A and B. No. 1, untyped expanded-stemmed broad-bladed point; 2, notched or stemmed, broad-bladed point lacking base; 3–6, points closely resembling the Susquehanna Broad type; 7, 8 bifurcated-base points; 9, drill tip fragment. Material: 1, 3, 5–9, Normanskill chert; 2, black chert; 4, speckled gray chert

26 tions may remain to be performed and the fin- mass. Many of these bifaces might be consid- ished product is referred to as Stage 4. ered “quarry blanks,” roughed out at the All Stage 1 and 2 bifaces, and most Stage 3 quarry and intended for transportation else- bifaces, are made of Normanskill chert. where prior to final shaping. Some represent Accompanying the descriptions of the rejects due to unreducible humps and others bifaces from Areas A, B, and C are scattergrams represent failures due to breakage during the showing the relation of the variables width vs. reduction process. thickness, and in some cases, length vs. width, Metrical data for Stage 1 bifaces are pre- for each group. In most cases there is a tenden- sented in Table 1. Only 15 of 53 pieces studied cy for the numerical data to cluster (i.e., they showed evidence of use, in the form of crush- are not distributed randomly, and there is a ten- ing or nibbling appearing discontinuously dency to slope upward from left to right). along the edges, usually on high points Sometimes the clustering is tight, but in other between flake scars (in rare cases the affected cases it appears very loose indeed. The slope areas may have been intentionally prepared indicates that as length increases, so does striking platforms for the removal of thinning width, and as width increases, so does thick- flakes). ness. But the width vs. thickness data on most The mean width/thickness ratio of Stage 1 charts suggest that the range of thickness stays bifaces is 2.68. Figures 19 and 20 show the fairly constant. These charts also appear to vin- clustering of Stage 1 bifaces in the attributes of dicate my original intuitive (pre-measurement) width vs. thickness, and length vs. width. sorting of all the recovered bifaces into the several stages of the reduction process. Stage 2 Bifaces Bifaces in this group have been further Stage 1 Bifaces reduced toward the intended end product In this group are 96 whole or fragmentary from Stage 1 bifaces. Flaking is still by percus- items (Figure 18). These objects, often irregu- sion, but applied more evenly, with less force, lar, asymmetrical, relatively thick, and and a symmetrical, more definite form is start- “crude,” show evidence of the first steps in ing to emerge. A critical factor is thickness; reduction from a core or spall. Most are rough- comparison cannot be made in terms of ly ovate. Percussion chipping has left deep, length, because only three Stage 2 preforms broad, rather uneven flake scars, and usually are whole (Table 2). Sixty-four objects are in there are untouched facets of the original chert the group (Figure 21). Eighteen of 52 meas- Table 1. Summary statistics for Stage 1 Bifaces from Areas A and B.

Attribute N Mean Range Standard deviation Length (mm) 30 74.7 52–120 17.02 Width (mm) 45 47.7 28–65 8.34 Thickness (mm) 53 17.8 9–32 4.71 Weight (g) 29 83.52 23.6–245 55.15

Table 2. Summary statistics for Stage 2 bifaces from Areas A and B. Attribute N Mean Range Standard deviation Length (mm) 3 73 42–107 — Width (mm) 45 44.4 30–-62 8.69 Thickness (mm) 52 13.1 8–20 2.73

27 Figure 18. Stage 1 bifaces from Areas A and B. Material: all Normanskill chert.

28 Figure 19. Scattergram of width vs. thickness for Stage 1 bifaces from Areas A and B.

Figure 20. Scattergram of length vs. width for Stage 1 bifaces from Areas A and B. ured bifaces of the group showed wear (dis- As width increases their range in thickness continuous edge-crushing or abrasion) on one remains constant at 5–10 mm. or both lateral edges. Because only three specimens were complete, data on the distribution Stage 3 Bifaces by weight are not presented. These “advanced” bifaces represent yet The mean width/thickness ratio of Stage 2 another step in manufacture. Items in this bifaces is 3.39. A comparison of Figure 19 with group are rather symmetrical, have attained a Figure 22 supports the original subjective sep- definite shape, and are relatively thin with aration of bifaces in process into stage 1 and evenly controlled chipping. Whereas some are stage 2; the chief difference is that stage 2 certainly preforms for projectile points or bifaces tend to be thinner than stage 1 bifaces. knives, others appear to be finished knives. In

29 Figure 21. Stage 2 bifaces from Areas A and B. Material: all Normanskill chert

30 Figure 22. Scattergram of width vs. thickness for Stage 2 bifaces from Areas A and B. an attempt to distinguish them statistically, nos. 14, 15 broke during removal of the chan- two primary subdivisions have been made. nel flake, which hinged through the blade. The first subcategory involves 24 bifaces An odd specimen, shown as no. 8, is which, on the basis of size (Table 3), geometric included with the group because it strongly outline, and the frequent presence of channel suggests an unsuccessful attempt by a novice flake thinning scars on one or both faces, are to produce a fluted point. Two channel flakes considered to be preforms for fluted points were removed from one face, and three from (Figure 29, nos. 9–28). The whole specimens the obverse. The flutes run the full length of vary in outline from ovate to lanceolate. The the point. One edge is bifacially chipped, the most obvious examples of fluted points in other unifacially worked. Part of the formerly process are illustrated in Figure 29, nos. 11–16, straight base is missing; the remaining portion 23–38. In most cases the base is straight, or as displays delicate bifacial nibbling. The oppo- in nos. 11 and 26, slightly concave. All have site end, or “tip,” is narrow and straight, with had channel (end-thinning) flakes struck from delicate unifacial nibbling. This object has the base on one or both faces. Three examples slight wear on the edges, apparently from use (nos. 13–15) display traces of grinding on the as a knife. The weights of five whole preforms base, probably to prepare the striking plat- are 11.5, 23, 28, 44.5, and 63.5 g. form. The specimen shown as no. 13 has had The mean width/thickness ratio for fluted two adjacent flutes removed from one face, point preforms is 3.67. Figures 24 and 25 and a ground nubbin for the third or central demonstrate the relatively high clustering by flake is clearly visible. The two examples in the metrical attributes for Stage 3 bifaces. The

Table 3. Summary statistics for fluted point preforms from Areas A and B.

Attribute N Mean Range Standard deviation Length (mm) 10 71.1 37–108 23.53 Width (mm) 22 36.6 25–53 7.54 Thickness (mm) 22 10.0 6–14 2.28

31 key factor is the reduced thickness, and small- using (Deetz 1967:45–49). There are two possi- er range in thickness, of the preforms com- ble formal goals here: fluted points and biface pared to Stage 1 and Stage 2 bifaces. knives. Only three bifaces have been definite- The second subcategory of Stage 3 bifaces ly classified as deliberately fashioned knives, comprises broad, ovate bifaces, some of which of probable broad ovate or lanceolate form. It are very likely knives. Others, as previously is evident that preforms at almost any stage of indicated, are probably preforms for either manufacture could have served as cutting knives or projectile points. Unfortunately, all tools, if one or more edges were sharp enough. of these specimens are fragmentary, only two In point of fact, a significant proportion of small ones being nearly complete (Table 4). bifaces in each stage actually were used as Channel (end-thinning) flakes are almost knives, no matter what their ultimate form entirely lacking on these items. Thirty-five was to have been. objects are included in this subcategory, most of which are illustrated in Figure 23. Probable Stage 4 Bifaces: Fluted Points finished knives are represented by three There is considerable variation in the size broad, thin tip fragments, all evincing consid- and some other attributes of Stage 4 bifaces, or erable rounding and gloss on the edges (nos. fluted points, as can be seen in Figure 27. 2–4). Although suggesting finished objects Thirteen definite examples have been recov- assignable to Stage 4, they are lumped with ered at the West Athens Hill, 10 from the Area Stage 3 bifaces due to their fragmentary con- B excavations and three from the surface in dition. Fifteen of the remaining bifaces in the Area A.5 There are both whole and fragmen- group display from slight to considerable tary points, all of which are finished except for rounding/gloss on the lateral edges. The two the large one in Figure 26, no. 13, which is well small ovate examples (nos. 14, 15), a broad, advanced but has an unsharpened tip. The lobate-stemmed specimen (no. 22), and the point shown as no. 2 was the first recovered at basal fragment of a large, lanceolate biface the site by R. Arthur Johnson, and is a partic- with a large channel flake scar on one face (no. ularly fine specimen. 20) are also worthy of special note. Measurements, materials, and other attrib- The mean width/thickness ratio of Stage 3 utes of the fluted points are given in Table 5. bifaces is 4.03. The stage 3 sample tends to be Summary statistics are provided in Table 6. slightly thicker and wider than fluted point Among general comments, the lack of edge or preforms (cf. Figure 24 and 26). basal grinding on three apparently finished It seems necessary to make a distinction points (Figure 27, nos. 3, 10, 11) should be between functional and technological classes noted. The presence or absence of this trait of artifacts. In the case of bifaces, Stage 1, 2, or cannot be determined for two broken points even 3, it is often impossible to know precise- missing the base (nos. 5, 6) and there are no ly just what end product the maker had in signs of rubbing on the not-quite-finished mind—what mental template he (or she) was specimen, (no. 13). The midsection in no. 6 has Table 4. Summary statistics for Stage 3 bifaces other than fluted point preforms from Areas A and B.

Attribute N Mean Range Standard deviation Length (mm) —— — — Width (mm) 24 47.9 32–63 8.58 Thickness (mm) 30 11.9 9–15 1.64

______5 Twelve more fluted points or fluted point preforms were found in the 1969 and 1970 excavations at Area C.

32 Figure 23. Stage 3 bifaces from Areas A and B. Nos. 2–4, large finished knives; all others probable preforms for knives. Material: all Normanskill chert.

33 no. 11 No. 10 No. 12 Figure 26 reference No. 4 No. 3 No. 9 No. 7 No. 5 No. 6 No. 13 No. 1 1 wide, single flute each face. flake scar present. overlies trace of preliminary flute;2 other flutes; face base has indented 4 mm. flutes visible each face. lapping flutes. Base indented 3 mm; Weight 5.7 g. other face shows 2 overlapping flutes. Slightly eared base; base indented 3 mm 3 mm. Weight 5.5 g. visible on each face. visible one face. flute hinged through blade. skill chert each face, 1 lateral, one central; weight 17 g. ““ Midsection frag.; Upper portions single “ Mid-section frag.; upper end of a flute Unfinished tip frag.; upper part attempted chert faced. One ear; base indented 3 mm. Onondagachert Point was broken at base, and reworked. NoneNone Norman- SlightlyLower eared, edges base idented 4 mm; “ 2 flutes “ removedBase andlower edges Figure 26, One “ slight ear, Mid-section, base upper indented end 3 of mm; one channel Reworked, used as knife; Central flute one face, No. 8 Base andlower edges “Lower edges “None One short flute one face; other face shows One 2 face over- shows oneBase lateral, and one central flute; lower Smoky edges No. Norman- skill 2 chert Triple fluted oneLower face, edges single Tip flute one re-sharpened; other face, fully 1 fluted, on 2 other. “ flutes Base indented Weight 6 g. ? ? Midsection frag.; upper parts single flutes Lower edges Western Slightly convex base. Both faces fluted. (mm) (mm) (mm) Surface Catalog No. Provenience44042-4, 13 Length Width Sec.44066-18 E10N0, str. 1 Thickness Grinding? 6844007-1 Sec. E0N20, Material str. 1 Remarks 44041-11 26 — Sec. E0N0, str. 1 Sec. E10N0 36RAJ 8 — collection Area A surfaceJHM collection 30 7 — Sec. E0S10, str. 1 4144042-14 24 8 37 2044067-17 Sec. E10N0, str. 17 1 6 6 45 Sec. E0N20, str.44034-18 1 6 40 2244014-15 Sec. E10S10, str. 2 2344080-7 — Sec. 7 W180-S50,44069-9 27 — 5 Sec. W10N-40, str. 1 —JHM collection Sec. 22 E10N20, str. 8 1 Area A surface 34 — 5 51 40 7 27 10 8 ? Table 5. Metric and non-metric data for finished Stage 4 fluted points from Areas A and B .

34 Figure 24. Scattergram of width vs. thickness for Stage 3 fluted point preforms from Areas A and B.

Figure 25. Scattergram of length vs. width for Stage 3 fluted point preforms from Areas A and B. the blade form and flaking characteristics of other face (Figure 27, nos. 3, 4, 11), and one of fluted points and the upper end of a channel which had two channel flakes removed from flake scar is visible on the illustrated face. The one face, and three from the other (no. 8). point in no. 8 seems to have been slightly These are the sole examples of possible reworked after the tip was broken off, and “Enterline” fluting (Witthoft 1952), but it is of used as a knife. The point in no. 1 was origi- course possible that the technique was prac- nally somewhat longer, but after some mishap ticed on other points where the final fluting the base was reworked for further use. has obliterated all traces of the preliminary Multiple fluting is evident on only four flutes. points, three of which have traces of two chan- All of the points except three (Figure 27, nel flakes on one face and one flake on the nos. 6, 12, 13) are fluted on both faces. The

35 Figure 26. Scattergram of width vs. thickness for Stage 3 non-fluted preforms from Areas A and B. three exceptions are broken, lacking the base, New York Onondaga chert. The fine speci- and it is possible that there were short channel men, no. 3, is of translucent smoky chert. scars on the missing portions of the apparent- The mean width/thickness ratio of Stage 4 ly blank faces. fluted points is 3.87, and the range is 2.83 to Nearly all of the points conform generally 5.14. Compared to graphs of previous stages to the Eastern Fluted, or Gainey style, consid- in biface production, Figure 28 shows that fin- ered to represent the oldest period of occu- ished fluted points are both narrower and pancy by Paleoindians (Gramly and Funk thinner, and the range of thickness is more 1990). All are lanceolate, and with one excep- restricted then on the other bifaces. The tion the greatest breadth is at or just above weights of four whole points are 17, 5.7, 6, and midpoint. The exception (no. 3) is widest at 5.5 g. the base, narrowing gradually toward the tip. The point in no. 4 is the best example of the Projectile Point Tips and Midsections relatively late Cumberland-Barnes type. A Twenty-one fragmentary points are in this majority of the points had straight to very group (Figure 29, nos. 1–7) and are classified slightly constricted sides between base and as Stage 4 bifaces. All seem to fall within the midpoint. The base, where intact, is indented. size and shape range of Eastern Fluted points, In every case but two the raw material is but without exception the basal section is local Normanskill chert. The point with a missing. Three pieces (nos. 3, 6, 7) display reworked base (Figure 27, no. 1) is western what may be the extreme upper ends of chan-

Table 6. Summary statistics for Stage 4 fluted points from Areas A and B.

Attribute N Mean Range Standard deviation Length (mm) 6 47.0 37–68 10.37 Width (mm) 12 27.1 17–40 6.50 Thickness (mm) 13 7.0 5–10 1.35

36 Figure 27. Fluted points from Areas A and B. Materials: no. 1, western Onondaga chert; 3, smoky gray chert; all others Normanskill chert (no. 4 is plastic replica of specimen in private collection).

37 Figure 28. Scattergram of width vs. thickness for Stage 4 fluted points from Areas A and B. nel flake scars immediately adjacent to the line (Figure 31, nos. 10–17). In nearly every case, the of fracture. As shown in Table 7, the group thickest part of the tool is opposite the working seems to resemble the fluted point sample in edge, frequently displaying the weathered thickness. original surface of the nodule, and provides a The scattergram for width vs. thickness convenient surface for the hand to grip. These (Figure 29) conforms to expectations from the tools uniformly possess edge wear, generally fluted point data (i.e., these bifaces are rela- edge-crushing and nibbling, and might be con- tively narrow and thin and the range in thick- sidered a form of backed knife. Despite the ness is ca. 2-3 mm). presence of retouched or steeply beveled work- Most of the fluted point fragments are of ing edges, these items are clearly not unifaces Normanskill chert; one tip is clear crystal and deserve a separate classification. In length quartz, another is yellow Pennsylvania jasper, they range from 48 to 94 mm, in width from 35 and a third is black chert containing tan to 74 mm, and in thickness from 10 to 29 mm. specks, possibly a Normanskill variety. Another group of 10 small, ovoid thick bifaces (Figure 31, nos. 1–3, 5–9) are more dif- Miscellaneous Bifaces ficult to assign as to function. In six cases (nos. This category is a catch-all for artifacts 5, 7–9) at least one, sometimes two, steeply which cannot be placed satisfactorily in the bevelled edges appear to have been used for other groupings. Fifteen examples can be scraping. Battering along the edges of four of described as core tools, based on spalls, nod- these may evince unsuccessful attempts at ules, or slabs of chert, which have been modi- thinning. Another small ovoid (no. 6) appears fied in such a way as to produce a single bifa- not to have been used, and may be a spent cially sharpened cutting and/or scraping edge core. The example in no. 1 may be an ovate

Table 7. Summary statistics for probable fluted point fragments from Areas A and B.

Attribute N Mean Range Standard deviation Width (mm) 16 33.9 25–40 5.64 Thickness (mm) 21 8.2 4–12 1.8

38 Figure 29. Fluted point fragments and preforms from Areas A and B. Nos. 1–7, fragmentary fluted points; 8, unique bifacial object with multiple fluting on both faces; 9–28, fluted point preforms. Material: No. 1, brown Pennsylvania jasper; all others Normanskill chert.

39 Figure 30. Scattergram of width vs. thickness for probable fluted point fragments from Areas A and B. knife, but the edges are sharp and unworn. tion of chert slabs, cores, and flakes. This mod- The broken biface in no. 2 and the oval piece ification consisted of retouching, generally by in no. 3 seem to have been knives. This group percussion flaking, along one or more discrete of artifacts range from 51 to 71 mm long, 30 to edges. In the majority of cases, the chipping 45 mm wide, and 13 to 19 mm thick. A small, tool was directed against the bulbar face, broad, tapering-stemmed object lacking signs detaching flakes from the other (dorsal) face of of use (no. 4) may not be of Paleoindian origin. the tool. A minority of smaller tools display It bears some resemblance to Susquehanna delicate flaking or nibbling, which appears to knives. Seven other “nondescript” pieces are be the result of pressure flaking. also in this miscellaneous biface group. My reexamination of these objects indicated that End Scrapers some, at least, could more properly be classi- The great majority of these artifacts are of fied with Stage 1 or 2 preforms, and this classic trianguloid or trapezoidal form, with impression is reinforced by the graph of width the bulbar face unmodified. All end scrapers vs. length (Figure 31) when compared to the are based on medium-sized flakes struck from Stage 1 bifaces (Figure 19). cores, steeply beveled by retouching the broad A total of 33 items are included in this cat- end and in many cases trimmed to form along egory. All of them are of Normanskill chert. the sides. The flakes used were sometimes flat, Statistics for the weight in grams of 12 whole but more often hump-backed or ridged. Over specimens are as follows: 30 percent of the end scrapers still retain the Mean 52.2 striking platform at the narrow end. Median 37.8 Many of the 92 end scrapers from Areas A Standard dev. 50.5 and B show wear on the bit ends, and often Range 18.5–209 along the sides as well, whether the sides were Count 12 retouched or not. The wear is usually in the form of edge-crushing. A small number display wear or battering on the butts. Unifaces Commonly, the working edge on the broad Objects in this large class were intentional- end has been rechipped to an extremely steep ly produced by the simple unifacial modifica- angle.

40 Figure 31. Miscellaneous bifaces from Areas A and B. Nos. 1-3, 5-9, small ovoid bifaces; 4, tapered-stem bifaces; 10-17, biface-edged cores and flakes. Material: all Normanskill chert.

41 Figure 32. Scattergram width vs. length for miscellaneous bifaces from Areas A and B.

The end scrapers have been sorted, more working edge; two (no. 1) are double-ended. or less arbitrarily, into several morphological An additional seven end scrapers can be varieties. Of the total number, 85 are of trian- described as irregular, with one main scraping guloid or trapezoidal form; simple examples of edge perpendicular to the longest axis. the type are shown in Figure 33, nos. 2, 8–42, Measurements were taken of edge angles 44, 45. Within this group, 15 display weak to on the bit portions (“working ends”) of end prominent graving spurs (nos. 8, 9, 12, 22, 28, scrapers (a sample of 26). The results indicate a 33, 40, 43), which are characteristic of range of 35–75 degrees, with the majority (17) Paleoindian implements. The spurs are usual- falling between 40 and 60 degrees, only 3 ly to be found on one or both of the front cor- between 35 and 40 degrees, and 6 between 60 ners, but in a few instances they are located and 75 degrees. along the sides. Twenty-nine other tools Most end scrapers were made from local among the 85 triangular specimens have Normanskill chert. Seven are of red or yellow straight to slightly rounded main working Pennsylvania jasper (Figure 33, nos. 31–35, edges with sharp or right-angled corners. 39–41), 1 is of an unidentified cream-colored, These corners, like spurs, could have been tan-speckled exotic jasper (no. 42), and 4 are of used for graving bone or wood. The remainder Western New York Onondaga chert (nos. in the class have rounded or irregular ends. 36–38). Summary statistics for end scrapers are One scraper (no. 40) is notched just behind the presented in Table 8.

Table 8. Summary statistics for end scrapers from Areas A and B. Attribute N Mean Range Standard deviation Length (mm) 91 32.16 21–75 9.47 Width (mm) 92 26.97 13–43 9.95 Thickness (mm) 92 8.78 4–19 3.13 Weight (g) 39 7.03 2.5–14.4 3.1

42 Side Scrapers Twenty-one, or 11 percent, of the side The 185 items in this class, like end scrap- scrapers are directly based on natural chert ers, can be broken down into rough sub- blocks. The remaining pieces were nearly all groups. Metrical data are presented in Table 9. made from cores or flakes struck from cores. These are the most numerous of the intention- Of this group of 164, only 48, or 29 percent, ally modified tools on the site, and they were still retain the striking platform, which for one made from almost any convenient flake, reason or another was removed from the other block, or core, whatever its size or shape. The 116. On four side scrapers, the striking plat- bulk of these tools conform in a general way to forms can be described as faceted butts. Byers’ (1954) “ear-shaped” category for scrap- The weight of the “pulping plane” is 564 g; ers at the Bull Brook site (Figure 34, nos. 9–12, this weight is not included in the calculations 14, 15, 17, 21, 24; Figure 35, nos. 4, 6, 7, 10, 13, above. 14). Approximately 16 specimens, mostly Local Normanskill chert used for over 96 based on the terminal portions of nodules, percent of the side scrapers. Exotic stones which in some cases had initially served as comprised 2 pieces of Upper Mercer, Ohio cores, are placed in a “turtleback” group chert (Figure 34, nos. 6, 10), 1 of yellow (Figure 35, nos. 3–6, 9–16). Twelve specimens Pennsylvania jasper, 2 of Western Onondaga can be referred to as convergent scrapers, fea- chert, 1 of Oriskany chert (no. 5), and 2 of turing two retouched long edges converging locally available Kalkberg chert (nos. 4, 17). to a tip. The tip has in some cases been broken in use (Figure 34, nos. 4–6). One of the conver- Flake Knives gent scrapers, plus several others of varying The separation of uniface knives from form (Figure 35, nos. 3, 5), possess spurs prob- side scrapers is essentially arbitrary and intu- ably used in graving. Very few of the scrapers itive. The two tool types are alike in almost can be considered to have true spokeshave every way, except that the retouched edges of edges, however. Other specimens of particular knives are shallow rather than steep and that interest are a massive fist-sized “pulping knives tend to be thinner than scrapers. plane” (Figure 35, no. 15) and a heavy tool Thirty-nine items have been classified as with roughly retouched semicircular working knives (Figure 36). All but one appear to be edge (no. 16). The rest of the side scrapers run based on flakes, about half of which still pos- the full gamut of forms through “ear-shaped,” sess striking platforms. In almost every case oval, trianguloid, and simply irregular. there are signs of wear (chiefly edge-crush- Over 75 percent of the side scrapers evince ing) along the retouched edges, and on heavy wear, usually in the form of edge-crush- unmodified edges as well. Metrical data are ing, on the retouched edges. In the majority of presented in Table 10. cases, unmodified edges also show signs of The great majority of flake knives are of use. On one example the striking platform has Normanskill chert, but a small number were been heavily battered; on two other scrapers it fashioned from non-local stones. Two examples is much worn. are of Fort Ann chert (Figure 36, nos. 18, 21), 1

Table 9. Summary statistics for side scrapers from Areas A and B.

Attribute N Mean Range Standard deviation Length (mm) 168 58.9 20–120 17.62 Width (mm) 172 39.41 15–81 9.43 Thickness (mm) 174 15.93 4–59 7.53 Weight (g) 125 47.5 4.5–377 43.1

43 Figure 33. End scrapers from Areas A and B. Nos. 8, 9, 12, 22, 28, 33, 40, 43, spurred variety. Material: all Normanskill chert except nos. 31–35, 39-41, Pennsylvania jasper; 36–38, western Onondaga chert; 42, creamy speckled chert of unknown origin

44 Figure 34. Side scrapers from Areas A and B. Material: all Normanskill chert except nos. 4, 17, of Kalkberg chert; 6, 10, Upper Mercer, Ohio chert; 5, grainy black (Oriskany?) chert.

45 Figure 35. Large side scrapers from Areas A and B. Nos. 3 and 5 have graving spurs. Material: all Normanskill chert

46 Table 10. Summary statistics for flake knives from Areas A and B. Attribute N Mean Range Standard deviation Length (mm) 34 49.86 23–95 17.29 Width (mm) 34 33.15 17–51 10.80 Thickness (mm) 34 9.0 4–18 3.25 Weight (g) 33 20.5 2.5–58 16.6

is of crystal quartz (no. 8), and 1 is of quartzite ing but not identical to intentional retouch. (no. 15). The modifications are almost entirely on the thinnest, sharpest edges of the flakes or cores. Retouched Flakes All of these soecimens are Normanskill chert. This designation applies to 30 tools that cannot be conveniently assigned to the other Pieces Esquillees uniface categories. Generally, they are based Despite careful inspection of over 12,000 on oval or irregular flakes that have either waste flakes from Areas A and B, only five slightly or irregularly retouched edges. Wear examples of this class of tool were found in the (step-flaking or crushing) is present in a collection. All are Normanskill chert. Three are majority of cases. Four specimens were made flakes that display battering on opposite, from naturally occurring chert chunks. Seven paired edges. Bifacial chipping, sometimes of the artifical flakes still have striking plat- resulting in the removal of long, ribbon-like forms. The retouched flakes range in length flakes, is a result of such battering (Figure 37, from 20 to 88 mm, in width from 20 to 55 mm, no. 2). Two objects assigned to this class have and in thickness from 6 to 24 mm. All but one the appearance of small biface preforms (no. 3). are of Normanskill chert, the exception being Pieces esquillees were numerous at the of Western Onondaga chert. Debert site, Nova Scotia, where they were first recognized by MacDonald (1968: 85–90). This Graver tool type, common in the Old World Upper This specimen of Normanskill chert is a Paleolithic, seems to have been used for spall bearing a single narrow projection, pro- grooving and splitting bone or antler duced by deliberate retouch on two edges. (Semenov 1964:149–150). Utilization is evident on the tip and on the Though the years since discovering the unretouched edges of the spall. site, R. Arthur Johnson had collected artifacts from various loci, principally Areas A and B. Utilized Flakes The following (Table 11) is a summary listing A total of 660 cores, spalls, and flakes lack- of 42 chipped stone items donated to the New ing deliberate, systematic retouching display York State Museum in 1979 and also following evidence of use on edges, ends, or corners. Johnson’s death. These artifacts had not been This evidence was produced mainly by edge- previously studied, nor have they been sub- nibbling, or removal of tiny flakes by pressure jected to metrical and wear pattern analysis, against another solid object. In some cases the except for the three fluted point preforms. edge is so rough and jagged as to suggest bat- They are all Normanskill chert. The listing is tering or chopping. On a few pieces utilization not exhaustive, since some items in Johnson’s has pressed off an even row of tiny chips, cre- donation have not been examined in detail. ating a steeply angled working edge suggest-

47 Table 11. Trait list for chipped stone tools in the R. Table 12. Weights of a sample of hammerstones Arthur Johnson collection. from Areas A and B (in g). Bifaces, stage 1: 8 Mean 500.6 Bifaces, stage 2: Median 3,84.5 Fluted point preforms: 3. Mode 148 Standard dev. 475.13 One whole biface measures 56 mm long, Range 2,864 36 mm wide, and 13 mm thick. It does not pos- Minimum 55 sess end thinning flake scars. A second speci- Maximum 2,919 men is a basal fragment, measuring 43 mm wide, 12 mm thick, broken during end thin- Count 66 ning of one face. The third item is also a basal fragment, measuring 35 mm in width, and 10 mm in thickness, and is fluted on one face. Anvil-hammerstones None of these objects show basal or edge rub- Six objects are similar to simple cobble bing. hammerstones, but in addition to battered Other biface fragment: 1. ends or edges they display the unmistakable Side scrapers: 7 scarification that results from use as anvils on Retouched flake toools: 3 one or two facets or flat surfaces. Utilized flakes: 20 Abradingstones These artifacts are included in the trait list This section represents a modification of below (Table 13), and has been integrated into the typology used in Ritchie and Funk (1973), Tables 35 and 36 in final sections of this report. in which anvilstones and abradingstones were listed separately. Reanalysis indicates that the Rough Stone items in the site sample are not easily subdi- Cobble Hammerstones vided in this fashion. Glacially derived cobbles were used by the This interesting group of 11 artifacts were occupants of Areas A and B at West Athens apparently primarily abraders, also often Hill to work the chert veins and nodules called whetstones. Eight can be typed as exposed in the sandstone outcrops, and prob- grooved abraders (Figure 38, nos. 1–3). But ably to make lithic tools (Figure 37, nos. 1, 4–6, three tools, not assigned to the group of 8, 9). Hammerstones (n = 167) occurred in grooved stones, have one broad, shallow, many sizes, but only a few are so large as to smoothed depression each bearing fine stria- have required the use of two hands rather tions or scratches, generally on the naturally than one to hold them. Most of the cobbles are most even and flat surface that faced upward of ovate shape, the rest being spheroidal or when the tool lay on the ground. In each case, discoidal, and almost invariably display bat- the smoothing and concavity may have been tering on one or both ends of the long axis. A produced by rubbing another flat, solid object few examples are battered around the entire of stone back and forth across the stationary periphery. slab. One of these tools was also used as a Of 66 hammerstones examined from Areas heavy hammerstone because both ends show A and B, 24 are quartzite, 38 are sandstone, 1 removal of large fragments from forceful bat- isgneiss, 1 is granite, and 2 are conglomerate. tering. Perhaps it could be described as a ham- The weights of the hammerstones are present- mer-milling stone, or alternatively, as a ham- ed in Table 12. mer-abrader. The other two closely resemble

48 Figure 36. Flake knives from Areas A and B. Material: all Normanskill chert except no. 8, quartz; 15, quartzite; 18, 21, Fort Ann chert.

49 Figure 37. Chipped and rough stone tools from Areas A and B. No. 1, chert pebble hammerstone; 2, 3, battered objects identified as pieces esquillees; 4–6, 8, 9, cobble hammerstones; 7, possible petroglyph. Material: 1–3, Normanskill chert; 4–9, quartzite.

50 Figure 38. Rough stone tools from Areas A and B. Nos. 1–3 grooved abradingstones. Material: all sandstone

51 Table 13. Trait list for Areas A and B. the millingstones commonly found on later sites of the Archaic and Woodland periods. It is conceivable, although difficult to prove, that Chipped Stone the smoothed surfaces on these objects were Bifaces produced by milling or mealing of nuts and other plant foods. No tools resembling Projectile Points, Fluted mullers or handstones were found to support Finished 13 In process 27 this hypothesis, and this was confirmed by Possible 21 reexamining the rough stone tools from all loci at West Athens Hill. Projectile Points, It should be noted that some of the tools Archaic 10 display small scarred areas indicative of use as Other Bifaces anvilstones as well as abraders. These 11 tools, Stage 1 104 whether anvilstones, millingstones, abraders, Stage 2 64 or combinations of these functional types, Stage 3 35 range in weight from 230 to 5,422 g. They Miscellaneous 34 range in length from 95 to 215 mm, in width Total Bifaces 308 from 71 to 155 mm, and in thickness from 23 to Unifaces 55 mm. Sandstone slabs were used for all of Scrapers, end 92 these tools. The questions raised by the wear Scrapers, side 192 patterns on the “abraders” are considered in Knives, flake 39 later pages. Flakes, retouched 33 Graver 1 Possible Petroglyph Pieces esquillee 5 One small quartzite cobble (Figure 37, no. Flakes, utilized 680 7) bears what appear to be crudely scratched Total Unifaces 1,042 lines arranged in a rather enigmatic pattern. The main design is in the form of a ladder, the Rough Stone sides of which diverge at one end. Fainter lines can be seen on both sides of it. Hammerstones 167 It is difficult to suggest what the scratches Anvil-hammerstones 6 may represent—they are definitely not ran- Abradingstones/ dom. There is some resemblance to a butterfly anvilstones 11 or other arthropod; the ladder motif could Petroglyph (?) 1 stand for a segmented body, and two small Total Rough Stone 185 projections at one end could be antennae. The faint crescentic scratches on both sides of the Total Assemblage 1,535 main motif, with some stretch of the imagina- tion, can be visualized as wings. Other interpretations are possible. For example, it has object such as a stone flake. The quartz grains been suggested that the central figure is an appeared to have been broken by the applica- elephant’s head and trunk, the crescentic lines tion of strong force. being tusks. During the reexamination of the collection I examined this specimen with a stereomi- in 1999, I experienced some doubt about the croscope under 5 to 25 magnification. It original interpretation of the object as an arti- seemed to me that the main part of the design fact and wondered whether it was a fossil (it exhibited grooves that had been incised or had been examined by a geologist circa 1966 engraved into the stone using a hard, sharp but he was unable to verify the lines repre-

52 sented a fossil). I submitted it to Dr. Ed in Ohio and Pennsylvania. Western Onondaga Landing, New York State Paleontologist, New chert and brown or red Pennsylvania jasper York State Museum, for his examination. He were the two most important materials was unable to conclusively determine besides Normanskill chert. whether it was an invertebrate fossil or an It is of further interest to note that more artifact. Therefore, for this report I have mod- non-local stones were used for end scrapers ified my description to designate it a possible than for any other artifact class. Over 60 per- petroglyph. Whatever one sees in this object, it cent of such materials present in Areas A and may well be of Paleoindian origin, coming as B are to be found in end scrapers, and nearly it did from Stratum 2A. Its weight is 479 g. It 90 percent are accounted for by uniface arti- measures 101 mm long, 74 mm wide, and 43 facts. A further breakdown shows that 7 of the mm thick. 9 Pennsylvania jasper tools and 4 of the 9 items of Western Onondaga chert are end scrapers. Summary of Lithic Materials Used in Chipped Stone Tools Lithic Debitage Analysis The data in Table 14 are self-explanatory. Only 33 artifacts from Areas A and B, 2.5 per- Over 12,000 pieces of chert debitage were cent of the total, are made of stones other than recovered in the State Museum explorations of Normanskill chert. Of this group, only a few West Athens Hill in 1966. The total does not items are of other locally obtainable materials include utilized flakes, utilized cores, or (viz., quartz and quartzite from glacial cobbles pieces esquillees, all of which are classed as or rock outcrops) and Kalkberg chert (from the artifacts (total number 665). With one excep- Kalkberg or Helderberg escarpment). The tion, a flake of red Pennsylvania jasper, the remainder are exotic to the area, and in 13 source was the Normanskill chert veins on or cases the sources are outside New York State, near the site. There was not a single identified

Table 14. Frequencies of lithic materials used for artifacts other than Normanskill chert, from Areas A and B (1973 analysis).

Type No. Percent of Percent of Deliberately Made All Chipped Stone Artifacts (623) Artifacts (1,308) Western Onondaga Chert 9 1.4 .7 Fort Ann Chert 4 .6 .3 Upper Mercer Chert 3 .5 .2 Kalkberg Chert 1 .2 .08 Oriskany Chert 1 .2 .08 Pennsylvania Jasper 9 1.4 .7 Smoky Quartz 1 .2 .08 Clear Quartz 3 .5 .2 Quartzite 1 .2 .08 Speckled Creamy Chert 1 .2 .08 Totals 33 5.4 2.5

53 flake of Western Onondaga chert or any of the intentionally split to inspect the qualities of other exotic stones used for some lithic tools. the material. These might be called block frag- Occasional chips of quartzite were probably ments. Some of the unworked pieces are of derived from the manufacture of a small num- tabular form, and can be referred to as vein ber of uniface tools from local cobbles of that plates. material. A small sample of five blocks showing no In Area B of the excavations, varying modification range from 94 to 150 mm in quantities of debitage were saved from all length, 75 to 110 mm in width, and 51 to 100 squares except those disturbed by collector mm in thickness. They range in weight from digging. However, every chip and core was 431 to 1,029 gm. This sample provides a gen- collected from only six squares. As might be eral idea of the size and weight of these expected, the highest counts were from these objects, but a few visually inspected blocks in squares (total 10,408). In this group, the pro- the total sample from the locus are even larg- portion of debitage to artifacts was 11 to 1. er than those measured. Summary statistics The major task of analyzing the numerous for a limited sample of block fragments, not pieces of debitage was carried out by Beth qualifying as cores, are presented in Table 15. Wellman, then Laboratory Technician, in the 2. Cores. In this category are blocks, block Anthropological Survey, New York State fragments, or vein plates bearing at least one Museum. The lithic refuse was broken down flake scar denoting the intentional removal of into eight morphologic categories. flakes, for some purpose. Some of these items display flake scars over most of their surface. 1. Blocks. These are angular fragments of They vary from rectilinear blocks or block chert from the hilltop veins, apparently picked fragments to roughly spheroidal shape. up on the surface, or pried and hammered Summary statistics for a small sample of cores from their matrix, by the Indians. Some are are presented in Table 16. These data give weathered on all facets to a yellowish-brown some idea of the size and weight attributes of or white color. There are also some pieces these objects in Area B and permit comparison bearing a rind from the plane of contact with with the data from Area C. sandstone. Other pieces showing one or more fairly fresh-appearing chert surfaces were 3. Core Fragments. These are spalls that are apparently separated from parent exposures usually thick and semi-pyramidal, or keeled, by rough hammer blows, or were loose blocks in form, too small to have been useful as cores

Table 15. Summary statistics for block fragments from Area B.

Statistic Length (mm) Width (mm) Thickness (mm) Weight (g) Mean 55.3 41.9 27.7 91.3 Median 52 39 26 59.5 Mode 78 32 26 20 Standard deviation 20.0 12.0 13.6 88.7 Range 74 41 49 298 Minimum 32 26 11 14 Maximum 106 67 60 312 Count 20 20 19 20

54 Table 16. Summary statistics for a sample of cores from Area B.

Statistic Length (mm) Width (mm) Thickness (mm) Weight (g) Mean 86.8 65 41.5 281.3 Median 86 62 41 221 Mode 90 62 42 456 Standard deviation 21.4 13.4 9.6 155.6 Range 105 49 43 578 Minimum 55 46 24 105 Maximum 160 95 67 683 Count 39 39 39 39

Table 17. Summary statistics for core fragments from Area B.

Statistic Length (mm) Width (mm) Thickness (mm) Weight (g) Mean 66.0 42.8 28.5 89.1 Median 61.5 43 28.5 81.5 Mode 59 45 40 100 Standard deviation 15.0 7.7 6.8 38.0 Range 63 29 26 139 Minimum 46 27 14 27 Maximum 109 56 40 166 Count 24 23 24 24

in themselves. They appear to have been cortex or patinated surface on their dorsal side struck from cores as an attempt to prepare and show fresh chert on their ventral face. new striking platforms. Technically, they They are sometimes removed by hard ham- equate with both primary and secondary mer percussion. They are sometimes difficult flakes. It is frequently difficult to identify to distinguish from block fragments. either the striking platform or the bulbar face. In fact, it is easy to confuse some of these 5. Secondary Flakes. These are flakes struck items with block fragments. Summary statis- from cores by the percussion method. The tics for a small sample of core fragments from scars of flakes previously removed from the Area B are presented in Table 17. Again, these cores are visible on the dorsal surfaces. The figures provide some notion of the range in striking platforms are clearly evident on these such items from the locus, and permit com- flakes, which are generally of expanding ovate parison with similar materials from Area C. or trianguloid form. Some are irregular in outline. 4. Primary Flakes. These are first flakes to 6. Retouch Flakes. This category of debitage be struck from vein plates or blocks that have is considerably smaller than the mean dimen- weathered surfaces. Such flakes preserve the sions for secondary flakes. Striking platforms

55 and bulbs of percussion are weakly devel- broken off. In size and shape these flakes oth- oped. Soft hammer percussion is indicated. erwise correspond to the classes of secondary Some of these flakes may be from the retouch flakes and retouch flakes. of end and side scraper working edges. No polyhedral or prepared cores occurred in the debitage. True blades are extremely rare 7. Biface Thinning Flakes. These are similar among secondary flakes, although a small to retouch flakes, except that they are pro- number of elongate, blade-like flakes are pres- duced by thinning a biface; the striking plat- ent. Some cores bear the scars of narrow, rib- forms are at an acute angle to the surface of bon-like flakes that were detached from them. the face from which the flake was removed The same cores also display the scars of the and retain part of the face from which the more typical expanding ovate or trianguloid blow was struck. flakes, removed by means of striking platforms on various edges or corners. 8. Broken Flakes. No striking platform or The counts and percentage frequencies of bulbs of percussion can be discerned on these the eight debitage categories from six squares flakes. The portions with striking platforms in Area B are presented in Table 18. may have been deliberately or accidentally

Table 18. Frequencies of debitage classes in complete samples saved from six excavated squares in Area B.

Class No. Percent Blocks & Block Fragments 1,723 16.6 Cores 302 2.9 Core Fragments 3,209 30.8 Primary Flakes 466 4.5 Secondary Flakes 3,030 29.1 Retouch Flakes 842 8.1 Bifacial Thinning Flakes 12 0.1 Broken Flakes 824 7.9 Totals 10,408 100.0

56 ARTIFACTS FROM AREA C

The artifacts from this locus (Figure 39) are will not be presented for them). These objects generally similar in typology, morphological conform to the description of Stage A bifaces in range, size, and lithology to those from Areas the original report (Ritchie and Funk 1973:16), A and B at West Athens Hill but with some and as presented above. They are relatively interesting differences. The distribution of crude, thick, roughly ovate, showing evidence artifacts in Area C is plotted in Figure 40. of the initial steps in reduction from a core or spall. Both faces and both edges show at least CHIPPED STONE partial flaking. Percussion flaking resulted in deep, broad, flake scars, and unmodified areas Bifaces of cortex are frequently observed. The mean ratio of width over thickness for Bifaces in Process Stage 1 Bifaces is 2.79. The range is 1.06 to 4.87. Stage 1 (n = 103) Nearly all of these items are of Normanskill chert, the single exception being of a black It should be recognized that the attempt to chert reminiscent of Kalkberg chert. create fine subdivisions of Stage 1, 2, or 3 Scattergrams for width vs. thickness and bifaces is a subjective process. Although I ini- length vs. width for State 1 bifaces are pre- tially divided the sample into Stage 1A, Stage sented as Figures 41 and 42, respectively. 1B, and generalized Stage 1, all of these have This clustering for State 1 bifaces in Area C been combined as simply Stage 1 for this analy- compares well to the graph of width vs. thick- sis. Metrical data are presented in Table 19 (24 ness for Stage 1 bifaces from Areas A and B items assigned to this stage were found in the (see Figure 17). debitage during the analysis, and were chiefly very small fragments, therefore metrical data Table 19. Summary statistics for Stage 1 bifaces from Area C.

Statistic Length (mm) Width (mm) Thickness (mm) Weight (g) Mean 74.1 50.8 18.2 79.5 Median 72 49 17 59.5 Mode 73 56 17 N/A Standard deviation 16.5 12 5.3 49.7 Range 76 62 26 170.4 Minimum 51 28 8 35.6 Maximum 127 90 34 206 Count 20 55 77 12

57 Figure 39. Chipped and rough stone tools from Area C at West Athens Hill. Nos. 1–2, 4 basal fragments of finished (Stage 4) fluted points; 3, 5, 6, fluted points in process, broken during end-thinning; 7–16, fragmentary bifaces (7, 11, 12 are probably Stage 3 fluted points, while the others are larger and probably functioned as knives); 17–26, trianguloid or trapezoidal end scrapers (note probable graving spurs on 17, 18, 20, 22, 23, 24, 26); 27, elongate limace or side scraper with parallel steeply retouched edges and pointed ends; 28, steeply retouched side scraper on “turtleback” core; 29, side scraper on thick, flat flake with one long retouched working edge; 30, 31, cobble hammerstones. Material: Chipped stone items are of Normanskill chert except nos. 25, 29, of Western New York Onondaga chert, and no. 27, of Pennsylvania jasper. The hammerstones are of quartzite.

58 Figure 40. Map showing distribution of artifacts in Area C.

59 Figure 41. Scattergram of width vs. thickness for Stage 1 bifaces from Area C.

Figure 42. Scattergram of length vs. width for Stage 1 bifaces from Area C.

Stage 2 Bifaces (n = 40) not presented for them). Again, the classifica- Bifaces in this group from Area C were tion into Stages 2A, 2B, and generalized Stage further reduced toward the final product. 2 was abandoned and all of these bifaces are Percussion flaking was more evenly applied, combined as Stage 2. All of these specimens and the resulting form is again smaller and are fragmentary, therefore length measure- thinner than Stage 1 bifaces and more sym- ments are not given. It is also worth noting metrical. Descriptive statistics are shown in that one of the Stage 2 bifaces is clearly in the Table 20 (as with Stage 1 bifaces, study of the form of a Susquehanna knife, not pertaining debitage turned up additional specimens, to the Paleoindian occupation. It is not includ- chiefly small fragments, so metrical data are ed in Table 20.

60 Table 20. Summary statistics for Stage 2 bifaces from Area C.

Statistic Length (mm) Width (mm) Thickness (mm) Mean N/A 50.2 12.1 Median N/A 49 12 Mode N/A 60 10 Standard deviation N/A 12.1 2.3 Range N/A 55 8 Minimum N/A 35 9 Maximum N/A 90 17 Count N/A 27 31

The mean ratio of width over thickness for The material is Normanskill chert. All are Stage 2 bifaces from Area C is 4.15. The range fragmentary. They range in width from 19 to is 2.57 to 5.22. All of these objects are 64 mm, and in thickness from 7 to 14 mm. Normanskill chert. A scattergram for width There are some interesting morphological vs. thickness of State 2 bifaces is presented as differences within this small group, represent- Figure 43. This graph compares well with the ed by three tip-to-midsection fragments of graph for Stage 1 bifaces from Area C (Figure what may have been narrow stemmed or side- 41) as well as that for Stage 2 from Areas A and notched projectile points. They range from 19 B (Figure 21). to 25 mm in width of blade and 7 to 8 mm thick. The other five bifaces may be knives, Stage 3 Bifaces (n = 8) tending to be wider and thicker, ranging from These eight objects are the only items in 39 to 64 mm wide and 8 to 14 mm thick, com- this advanced biface category from Area C, mensurate with the majority of other bifaces apart from the possible fluted point preforms. in the assemblage.

Figure 43. Scattergram of width vs. thickness for Stage 2 bifaces from Area C.

61 The narrow-bladed bifaces may well have indented to 6 mm. Slight fire-spalling is pres- pertained to a Late Archaic component at ent. The third finished point (see Figure 39, no. West Athens Hill, but the absence of the diag- 4) is a basal fragment, 26 mm wide and 6 mm nostic basal portions precludes confident thick, with a straight base. Both faces are flut- identification. Their presence is not surprised. There is no basal or edge grinding. It too is ing, considering the discovery of two broad- Normanskill chert. bladed contracting-stemmed Susquehanna All of the other bifaces included in Table knives in Area C, a Susquehanna knife and 21 are assigned to Stage 3. They are small stemmed point in Area E, and the rela- Normanskill chert in gray and grayish-green tively late point types in Area B. varieties. There are seven basal fragments. Two of these bear single flutes or end-thinning Fragmentary Bifaces, Not Classified (n = 2) flake scars on one face, and the other face is Both are of Normanskill chert. basally thinned. Two others show fluting on Fluted Points (n = 12) one face, but the attempt at fluting the oppo- Twelve specimens from Area C were con- site face hinged through the blade. One frag- fidently identified as fluted points. Because so ment bears a single flute in one face, and two many bifaces in process are represented in the overlapping channel flake scars on the other assemblage, it is difficult to decide which ones face. The base is straight, and only slightly are fluted point preforms. Apart from size and rubbed. This point was very nearly finished shape, a chief criterion is the presence of end- before it was broken. thinning flake scars on the base of a preform The basal section of a reconstructed biface assigned to Stage 1 or 2. Items at Stage 3 are (see Figure 39, no. 12) was broken during end- easier to classify because they have been fur- thinning of one face; no channel flake scars are ther thinned, have a symmetrical outline, and on the other face. The slightly excurvate base display one or two scars from removal of end- as well as the edge lack grinding. It is thinning flakes or even final channel flake Normanskill chert. Two distal sections proba- scars. bly from fluted points (Figure 39, nos. 11–12) Table 21 presents summary data for the 12 are Normanskill chert. Four other Stage 3 bifaces from Area C classified as fluted points. biface fragments, 2 basal, 1 tip, and 1 midsec- Only three, however, are considered to be fin- tion are possibly from fluted points. Three are ished specimens. Only one is nearly whole. Normanskill chert, 1 is a dark gray chert with Summary statistics for fluted points are pre- small white specks. sented in Table 22. The mean ratio of width to thickness is The first finished point (see Figure 39, no. 5.01 mm, and the range is 4.22 to 6.0 mm. A 1) is nearly whole, lacking just the tip. It is scattergram of width vs. thickness for these ground on the base and also on the sides (19 artifacts is presented as Figure 44. mm and 15 mm, respectively). There is one This scattergram compares well with the channel flake scar on one face, two on the graph for fluted points from Areas A and B other face. The point was reworked and now (Figure 27). It shows a similar range of width the blade has a pentagonoid outline. The base and thickness. is indented 6 mm. The material is Normanskill There are four possible fluted points in chert. The second finished specimen is process, none of which is whole (Table 23). Normanskill chert (see Figure 39, no. 2) and The mean width of possible fluted points consists of the basal half. Both lateral and in process is 33.5 mm, the mean thickness 8.3 basal grinding are present. Two channel flake mm. The ratio of width over thickness ranges scars are evident on the face, and one channel from 2.8 to 5.29. flake scar on the other face. The base is deeply

62 Figure 39 reference No. 3 No. 1 No. 11 No. 4 Fluted one face, thinned other face. One tang missing. 2 flute scars one face, 1 on other face. Base deeply indented (6 mm).One face fluted to break,concave other No. (2 face 2 mm). basally thinned. Base slightlyOne flute scar one face,pentagonoid 2 outline. scars Base other indented face. 6 Point mm. reworked to No. 6 Flute scar on one facefluting extended attempt past other break, face was hinged 21–24 through mm blade. wide, Base straight. Broken No. in 5 half by attemptedface. end-thinning. Rounded No base. flute scar on otherBroken in half by attemptedRounded thinning. base. No flute scars. No. 12 One flute one face, 2 flutes other face. Straight base. Broken during end-thinning on onelength. face, No hinged flute through other at face. 40 mm Both faces fluted, straight base Remarks Normans-kill Slight fire-spalling. Normans-kill Normans-kill Normans-kill Normans-kill Normans-kill Normans-kill Normans-kill basal (19 mm and l15 mm up from base) rubbing rubbing (mm) (mm) (mm) (Stage 3A) finished (Stage 4) (Stage 3A) minus tip, finished (Stage 4) (Stage 3A) (Stage 3A) 3A Preform(Stage 3A) . (Stage 3A) (Stage 3A) (Stage 3A) finished (Stage 4) Catalog45693-1 Description Length45007-1 Basal frag. Width Basal frag.,45679-2 — Thickness Grinding? —45006-1 Basal frag. 34 Material Nearly whole, 29 —45679-1 7 — 33 645679-3 Basal frag. None 3645693-2,5 Upper half Lateral — and 6 Whole Stage45679-4 6 — Normans-kill 38 None 7845677-1 Basal frag. Basal and latera 3945680-7 Normans-kill Upper half 38 7 —45006-8 Basal frag. 8 — None 9 3845680-1 Basal frag. — None 40 None Basal frag., 7.5 — 26 8 Slight — basal 49 5 Normans-kill None 26 9.5 Slight basal 6 None Normans-kill None Number Table 21. Summary of data for finished (stage 4) fluted points and Stage 3 preforms from Area C.

63 Table 22. Summary statistics for Stage 4 fluted points and Stage 3 preforms from Area C. Unifaces End Scrapers (n = 37) Statistic Length Width These tools are broken down as follows; 24 (mm) (mm) trianguloid, 3 trapezoidal, 1 massive, 3 Mean 35.5 7.08 oblong, 5 fragmentary, and 1 in process. Size Median 37 7 and weight measurements are presented in Table 24. Most of these tools were made of Mode 38 6 Normanskill chert, but 4 are of Western Standard deviation 6.5 1.36 Onondaga chert and 1 of Eastern Onondaga chert. Range 23 4.5 Edge angles were measured on the bits of Minimum 26 5 end scrapers from Area C. All (n = 14) the Maximum 49 9.5 measured scrapers from Area C showed angles of 40–60 degrees, and half (7) fell with- Count 12 12 in 40–50 degrees. This result indicates a nar-

Table 23. Metrical data for possible fluted points in process from Area C.

Catalog Description Length Width Thickness Material Figure 39 No. (mm) (mm) (mm) reference 45678-1 Basal frag. — 37 7 Normanskill No. 7 (Stage 3) 45676-2 Basal frag. — 47 11 Normanskill No. 10 (Stage 3) 45695-1 Tip frag. — 22 5 Dark gray speckled (Stage 3B) chert 45695-2 Midsect-ion (Stage 3) — 28 10 Normanskill

Figure 44. Scattergram of width vs. thickness for State 4 fluted points and Stage 3 preforms from Area C.

64 Table 24. Summary statistics for end scrapers from Area C.

Statistic Length (mm) Width (mm) Thickness (mm) Weight (g) Mean 35.1 23.7 8.6 8.8 Median 34 23.8 87 Mode 35 25 77 Standard deviation 12.3 6.0 2.6 11.4 Range 65.5 35 11 63.5 Minimum 19.5 15 62 Maximum 85 50 17 65.5 Count 29 30 32 28

rower range of edge angles than represented on small flake, with two retouched converging in Areas A and B, but it should be noted that edges from same face, 1; on large flake, with less than half of the scrapers from Area C were one retouched edge, other edge utilized, 1: on measured for that attribute. small flakes, with one retouched edge, 9; with At first glance the mean weight of end two converging edges, from retouch on dorsal scrapers from Areas A and B is significantly and ventral faces, 2; on a rectangular flake, less than that from Area C, despite the simi- retouched on entire periphery, 1; on a flake larity in size. But if one unusually massive end with single retouched semicircular edge, scraper from Area C is removed from the cal- showing small “spurs” or possible graving culation, the mean weight of end scrapers points between retouch flake scars, 1; and a from both loci is approximately 7.0 g. fine limace or slug-shaped side scraper, 1. Summary statistics for these tools are present- Side Scrapers (n = 19) ed in Table 25. The objects classified as side scrapers have All but 5 of the side scrapers from Area C been subdivided as follows: massive, with were fashioned from Normanskill chert. One two retouched working edges, 2; hump- (the “limace”) is brown Pennsylvania jasper, 1 backed, on core, with two retouched edges, 1; is Western Onondaga chert, 1 is a dark bluish

Table 25. Summary statistics for side scrapers from Area C.

Statistic Length (mm) Width (mm) Thickness (mm) Weight (g) Mean 62.6 43.5 14.4 39.8 Median 63 40 14 33.5 Mode 58 35 11 14 Standard deviation 15.9 14.9 5.1 31.5 Range 66 50 18 120 Minimum 28 20 64 Maximum 94 70 24 124 Count 17 17 17 14

65 gray chert with quartz streaks, and 2 are gray mm long, 32 mm wide, and 5 mm thick. chert, possibly an Eastern Onondaga variety. Other Chipped Stone: Retouched Flakes (n = 19) Pieces Esquillees (n = 2) These tools from Area C are not to be con- One object from Area C is an elongated fused with “biface thinning flakes,” also often blade core of Normanskill chert, both ends of called “biface retouch flakes,” or with “retouch which show heavy step-flaking or battering. It flakes.” Some of the specimens in this class is 73.5 mm long, 38 mm wide, and 21 mm could have readily been placed in the “side thick and weighs 70 g. Possibly it actually scraper” category. They generally consist of served as an unusual chipping hammer, but it secondary flakes bearing some retouch on a is strongly reminiscent of pieces esquillees single edge, but one is on a nodule, another from other site contexts. with retouch on one ventral, one dorsal edge. The second object is a secondary flake Some of these tools show only rude flaking pieces esquillees showing bipolar battering, (that is, uneven retouch flaking or very short and weighs 13 g. retouch), and one is a biface with a steeply retouched scraping edge. Thirteen of these Utilized Flakes (n = 903) objects range in length from 32 to 72 mm, in This category includes 31 possible utilized width from 17 to 52 mm, and in thickness from flakes and consists largely of primary and sec- 5 to 17 mm. The range in weight is 6 to 53 g. All ondary flakes, but also includes biface thin- of these tools are Normanskill chert. ning flakes (7), as well as snapped flakes. By far the predominant material was the local Other Unifaces (n = 4), all of Normanskill chert, but there was 1 utilized flake of Western chert, as follows: Onondaga chert and 5 of a mysterious, lus- Scraper or Core, Humpbacked (n = 1). trous blue-black chert that may be a This specimen, measuring 79 mm in Normanskill variety. length, 42 mm in width, and 23 mm thick, One large utilized secondary flake also weighs 77 g and shows crushing and round- apparently served as a burin, because a longi- ing on the rudely retouched parts of the edges. tudinal flake was removed along one edge from the end opposite the striking platform. Chopper-Scraper? (n = 1) The sturdy graving corner thus created shows This large spall or core has a retouched moderate rounding and abrasion. It measures edge, and a prominent beak or graving point 79 mm in length, 38 mm in width, and 15 mm adjacent to the retouch. It is 106 mm long, 87 in thickness, and weighs 47 g. mm wide, and 26 mm thick. Its weight is 267 The counts for utilized flakes here differ g. It has no visible wear. from those in the Area C catalogue. I ultimately rejected as unmodified many pieces identi- Gravers? (n = 2) fied as “utilized” in the catalogue. One of these is on a secondary flake, with a stubby point or tip created by removal of Debitage single flakes from each edge adjoining the tip. Due to time constraints and lack of techni- It measures 42 mm long, 30 mm wide, and 10 cal assistance, the writer selectively examined mm thick and weighs 10 g. There is no visible the large collection of chert debitage from wear. The other tool is a biface retouch flake Area C, concentrating on identifying, measur- modified to a graver. One long edge shows ing, and weighing quarried blocks, block frag- crushing and light rounding/gloss, and the ments, cores and core fragments. All other cat- tip is beveled with moderately heavy round- egories of debitage were examined piece by ing/gloss and nibbling. This tool measures 34 piece, but no attempt was made to count,

66 measure, and weigh the numerous primary, produced by the splitting of blocks in order secondary, uniface retouch, and broken flakes. to ascertain the amount and quality of chert Blocks, unmodified, chert visible on surface. underneath the weathered or encrusted sur- These represent the original quarried chert as face. They are distinct from cores in that they it occurred in the rock strata. In general, they do not show any scars of deliberately have not been further reduced and no flakes removed flakes. Summary statistics for a were removed. A sample of 11 was examined sample of these artifacts are presented in and measured. A different but overlapping Table 27. sample of 18 of the largest blocks in the collec- Cores. I counted, measured, and weighed a tion was weighed in order to provide a rough sample of 158 cores identified in the samples comparison to the size of block fragments and from Area C. Cores are relatively large objects cores. The block weighing 4,000 g was the of chert bearing one or more flake scars. Some largest in the collection, measuring over 200 are blocks or block fragments that are partial- mm long, but was not part of the first sample. ly flaked, while others are blocky but display Summary statistics for these artifacts are pre- evidence that numerous flakes have been sented in Table 26. taken off, and still others are roughly spherical Block fragments. These are simply pieces in shape. The cores were repeatedly turned by of the original quarried blocks, presumably knappers to find satisfactory striking plat-

Table 26. Summary statistics for samples of quarried chert blocks from Area C.

Statistic Length (mm) Width (mm) Thickness (mm) Weight (g) Mean 93.8 68.1 56 935.5 Median 87 69 59 534.5 Mode 84 73 N/A N/A Standard deviation 18.5 11.5 10.2 918.2 Range 61 47 37 3,771 Minimum 65 45 32 229 Maximum 126 92 69 4,000 Count 11 11 11 18

Table 27. Summary statistics for a sample of block fragments from Area C.

Statistic Length (mm) Width (mm) Thickness (mm) Weight (g) Mean 64.5 46.1 28.5 40.3 Median 64.5 47 29.5 21 Mode 77 50 17 1 Standard deviation 17.8 10.1 12.2 56.5 Range 77 36 43 275 Minimum 37 27 71 Maximum 114 63 50 276 Count 28 28 28 66

67 forms and flakes, usually of expanded shape, weighed, but a sample of only 30 core frag- were knocked off by hard hammer percussion. ments were measured to provide some indica- No polyhedral or prismatic cores were seen in tion of their sizes in comparison to cores. the collections from either Areas A and B or Primary flakes. These were examined indi- from Area C, but a rare few appeared to have vidually but not counted, weighed, or meas- been used for the production of blades. ured. Summary statistics are given in Table 28. Secondary flakes. These were also examined Core fragments. The 114 core fragments individually but not counted, weighed, or examined from Area C by no means represent measured. This is certainly the numerically the whole sample from the locus. All of these predominant category from Area C, number- artifacts are Normanskill chert, except for one ing in the thousands, and mostly Normanskill of blue-gray chert with numerous lighter-col- chert. Other materials include Western ored veins. This material vaguely resembles Onondaga chert (n = 8), gray quartzite (n = 1), both Fort Ann and Eastern Onondaga chert, milky quartz (n = 1), reddish-brown chert (n = but could not be conclusively identified. 4), and a lustrous blue-black chert (n = 32). Summary statistics for core fragments are pre- In one excavation unit, the debitage sented in Table 29. One hundred thirteen were included over 100 secondary flakes of fire-

Table 28. Summary statistics for a sample of cores from Area C..

Statistic Length (mm) Width (mm) Thickness (mm) Weight (g) Mean 94.9 69.2 45 332.9 Median 93 67 45 273 Mode 95 60 49 214 Standard deviation 18.9 14.6 11.4 195.9 Range 100 71 51 1071 Maximum 150 112 72 1167 Minimum 50 41 21 96 Count 156 154 154 152

Table 29. Summary statistics for a sample of core fragments from Area C.

Statistic Length (mm) Width (mm) Thickness (mm) Weight (g) Mean 68.0 45.3 22.2 33 Median 64 44 20 22 Mode 62 50 17 5 Standard deviation 15.0 11.1 8.1 33.3 Range 59 48 30 148 Minimum 50 25 10 1 Maximum 109 73 40 149 Count 30 30 30 113

68 spalled Normanskill chert. Given the absence a choice of raw materials quite similar to that of identified features on site, it is possible that at Areas A and B. Summary statistics for ham- these flakes were affected by a forest fire or merstone weights are presented in Table 30. historic burning of tree stumps rather than exposed to fire in a prehistoric hearth. Anvil-Hammerstone (n = 1) Retouch flakes, small, non-bifacial. Few of This tool is of sandstone and weighs 858 g. these were seen in the collection, but their low numbers may reflect the fact that the excavat- Anvilstone (n = 1). ed sediments were not screened. Despite care- This specimen is quartzite and weighs 259 g. ful trowelling by the crews, it is likely that some small flakes were not observed and Milling-anvilstone (n = 1). therefore were not collected. This is a massive combination tool of Biface thinning flakes (n = 414). All of these sandstone, weighing 3,769 g. It measures 245 flakes are Normanskill chert, with the excep- mm long, 200 mm wide, and 64 mm thick. tion of 1 possibly of Knauderack chert, and 1 of a reddish-brown chert. Abraders or Grooved Stones (n = 2) Channel flakes (n = 3). These relatively long, These are massive tools, on sandstone narrow, straight-sided flakes represent the best slabs, weighing 3,051 and 5,424 g. candidates for this by-product of end-thinning Respectively, they measure 225 mm and 230 fluted points. Several other flakes may pertain mm long, 150 mm and 175 mm wide, 61 mm to this group, but using caution they were and 83 mm thick. The questions raised before instead lumped with the secondary flakes. concerning the functions of similar objects from Areas A and B, apply to these tools also. Rough Stone Tools (n = 75) Hammerstones (n = 69) Teshoa (n = 1) As in the assemblages from Areas A and B, This item is a split quartzite cobble and hammerstones from Area C were based may have served as some sort of chipping tool. chiefly on glacial cobbles, but a small number were based on slabs from the local bedrock. Of A trait list for Area C is presented in Table 67 hammerstones examined, 38 are quartzite, 31. 23 are sandstone, 5 are granite, and 1 is gneiss;

Table 30. Weights of hammerstones from Area C.

Statistic Weight (g) Mean 437 Median 301 Mode 553 Standard deviation 574.5 Range 3,892 Minimum 46 Maximum 3,938 Count 67

69 Table 31. Trait List for Area C.

Chipped Stone Bifaces Fluted Points Finished 3 In Process 9 Possible 4

Other Bifaces Stage 1 103 Stage 2 40 (includes 2 Sus- quehanna knives) Stage 3 (includes 3 possible narrow side-notched or stemmed points) 8 Fragments 2 Total Bifaces 169 Unifaces End Scrapers 37 Side Scrapers 19 Retouched Flakes 19 Other 6 Utilized Flakes 903

Total Unifaces 984

Rough Stone Hammerstones, cobble 69 Anvilstones, cobble 1 Anvil-hammerstones, cobble 1 Grooved abraders 2 Milling-anvilstones 1 Teshoa 1

Total Rough Stone 75

Total Assemblage 1,228

70 USE-WEAR PATTERNS ON ARTIFACTS

Use-wear patterns on select tools from dif- on many utilized flakes and on tools such as ferent areas of West Athens Hill were ana- end scrapers and side scrapers. lyzed as part of the overall research. A number My experiments with wood were less con- of researchers have systematically studied clusive. Chert flakes were used to scrape the patterns of edge-wear on chipped stone arti- bark off white pine branches, and to shave the facts, with a view to determining how and on underlying wood as in shaping weapon what materials the tools were used. shafts. Both thick and thin flakes accom- Experimental studies have been of consider- plished the job beautifully, but without visible able help in narrowing down the functional signs of wear. A thin flake with a concave edge possibilities. A pioneering study by Semenov was used like a knife to slice deeply into the (1964) was followed by the work of Keeley wood, and as a saw to cut through it. Those (1980), Tringham (Tringham, et al. 1974), tasks were easily managed, but despite the Wilmsen (1968), and many others. Some of application of considerable force, the thin, these studies have employed high-powered sharp working edge remained without modi- microscopes at varying magnifications (gener- fication. Doubtless hard wood should also be ally from 100 to 500 power) in order to observe tried, but it seems likely that most trees avail- edge damage not ordinarily visible to the able to the Paleoindians were relatively soft, naked eye. Others have relied on stereomicro- including spruce, fir, willow, and alder. There scopes at lower power (5 to 50 magnification), is, however, evidence that some oak trees were which some workers (including me) feel are present in the Spruce–Fir forests around adequate for identifying the principal types of 11,000 B.P. (Moeller 1980). wear, although some micro-wear may be Edge rounding, gloss/polish, and occa- missed. sionally striations, like the attributes observed In the 1970s, the writer undertook a series on unifaces from Area C, were not produced of experiments in which unmodified chert in my brief experiments. These types of wear flakes of varying thickness and shape were are often assumed to result from the scraping employed in scraping bone and wood. Leg of hides, caused not by the material of the parts of freshly butchered deer, consisting of hides but by soil and grit that adheres to the the metapodial and foot with skin and fur still surface of the hides and eventually produces on, were obtained. After removal of the skin some attrition of the tool’s edge. But the often- and feet, the tendons and adhering membra- reported presence of phytolith smears on the nous tissue were cut and scraped from the edges of end scrapers and other tools suggests metapodials. The resulting wear on flakes some use on plants. Experiments by Keeley consisted of small chips pressed off the work- (1980) and others also showed that the working edge. The longer the flakes were used, and ing of bone produced rounding and gloss on the more force used in pressing them against unifaces. the bones, the more extensive the wear Chipped stone tools from Areas A and B became. This pattern of wear is identical with were initially examined for possible use-wear the edge crushing and edge nibbling observed patterns using a stereomicroscope at 5 to 25

71 magnification. Little or no edge-wear was Nibbling on right side of bit, observed on fluted points or bifaces in process, near corner: 1 and what was present on the latter could be Slight crushing and gloss on bit: 1 attributed to deliberate edge-abrasion to pre- Moderate rounding/gloss whole bit: 1 pare striking platforms for removal of flakes in the thinning of edges. Use-wear in the form of Heavy rounding/gloss on side spur: 1 rounding and gloss was observed on the edges Resharpening or crushing on bit: 11 of some Stage 3 bifaces that had been tentative- No visible wear: 21 ly identified as biface knives. This type of wear suggests they were used as meat cutters. End This study confirmed the impression scrapers, side scrapers, and retouched flakes gained from the original analysis, although a showed definite indications of wear, generally much higher percentage of end scrapers in the form of edge-crushing, and various lacked any other signs of wear than stated in degrees of rounding/gloss (but ordinarily only Ritchie and Funk (1973). As in previous analy- one kind of wear appeared on any particular ses, I used a Wild Stereomicroscope at 5 to 25 tool). These tools were apparently employed in power. This discrepancy is difficult to explain. various tasks, not necessarily always the kind The summary for 32 Area C end scrapers that would produce discernible modifications is as follows: of the edges or adjoining surfaces. But contact Crushing on bit only: 6 with bone, hardwoods, and hides may be indicated for the tools that showed evidence of use- Crushing on bit, front corners, and/or wear. spurs: 3 A reexamination of the wear patterns on Crushing on bit, rounding/gloss on 51 of the 92 end scrapers from Areas A and B bit, lateral edges, and rounding was undertaken in October 2000 because the on dorsal surface: 1 pattern noted for the original report (i.e., a Rounding/gloss on bit, front corners predominance of crushing) seemed inconsis- or spurs: 4 tent with the wear patterns observed on scrap- Rounding/gloss on bit, crushing on ers from Area C. The summary follows: spurs: 1 Rounding/gloss on bit: 5 Light crushing on one front corner: 1 Rounding/gloss on lateral edge only: 2 Light crushing on left spur: 1 Rounding/gloss on edge and bit: 1 Light crushing on right side bit: 1 No wear observed: 9 Light crushing on left side bit: 2 Light crushing on central bit: 3 The summary for 19 Area C side scrapers Light crushing on: whole bit is as follows: and corners: 2 Crushing on a retouched edge: 4 Heavy crushing left and right Crushing and rounding/gloss on sides bit: 1 retouched or non-retouched edge: 2 Moderate crushing on front Rounding/gloss on one or more corners: 1 retouched edges, plus some Heavy crushing on left edge, dorsal smoothing: 2 some on corners: 1 Rounding/gloss on main retouched Mild crushing/nibbling on corners: 1 edge, heavily ground spur: 1 Mild crushing on bit, nibbling Rounding/gloss on retouched edges: 3 on left corner: 1

72 Rounding/gloss on retouched edge, ascertain the types of use-wear, including 14 opposite edge battered items from Areas A and B and 40 from Area C. (step-flaked): 1 A majority of these ad hoc tools showed nib- Rounding/gloss on striking platform: 1 bling (rarely crushing) along the longest and thinnest edges. It is usually assumed that uti- No wear observed: 5 lized flakes were used to cut meat. Small flakes were pressed off when the edges struck The summary for 13 retouched flakes is as underlying tendons and bone. But a curious follows: aspect of the wear identified as edge-nibbling Crushing on non-retouched edges: 5 is the predominance of flakes removed from Rounding/gloss on retouched edges, just one side (face) of an edge, as if pressure crushing on non-retouched edge: 1 was applied with a sideways scraping or Rounding/gloss on retouched and dragging motion, the flake held perpendicular non-retouched edge, crushing on to and angled away from the direction of non-retouched edge: 1 action, rather than a cutting motion with the No wear observed: 6 flake held more or less vertically and moved back and forth in a direction parallel to the Thus, more rounding/gloss was evident cutting edge. It is difficult to determine how on the unifaces from Area C than those from these wear patterns were made on the West Areas A and B, suggesting a contrast between Athens Hill tools; one can only guess the kind the tasks performed and materials used in one of material that created the form of edge-nib- locus vs. the others. bling observed. What was scraped with these The frequently sharp or right angle front tools? Bone is a possibility, though the associ- corners of end scrapers, and the somewhat ated polish seen in experimental studies is less frequent presence of spurs on one or both lacking on most of these tools (Keeley 1980). corners, probably indicates an additional use But the thinnnest, sharpest edges on 10 of for these tools, such as the piercing and cut- the 40 flakes inspected from Area C displayed ting of hides for garment making or the fash- both nibbling and rounding/gloss suggesting ioning of bone and antler tools and weapons. use in working bone, hides, or other soft mate- Similar functions are usually assumed for rials. Wear was usually not evident on thick gravers and denticulates. portions of the flakes, except occasionally on It seems clear that some of the patterns of the striking platforms or, where the platforms wear on uniface tools from northeastern were absent, on the broad ends of the flakes. Paleoindian sites could easily have been pro- As previously noted, grooved or striated duced by using them to clean and work bone stone slabs from Areas A, B, and C pose anoth- and antler. Heavy and continued use on er enigma. The grooved abraders are tabular wood, especially hardwood, might also even- rock slabs upon which multiple parallel or tually cause similar signs of wear. It seems radial grooves and also fine linear scratches rather unlikely that the narrow, rounded bits have been worn. On two specimens the over- of end scrapers would be useful in working all surface containing the grooves has been wooden shafts for weapons, given the difficul- smoothed (see Figure 38, no. 3). The grooves ty of maintaining contact between two small, and striations vary in depth and width. One concave surfaces. On the other hand, concave exceptionally fine implement (see Figure 38, scrapers and side scrapers may have been well no. 2) bears relatively deep grooves on both suited for the task. As stated above, another primary faces. The crucial question is how the function of end scrapers was probably for pro- grooves were produced. They had to be processing hides. duced with materials as hard as the surface of Samples of utilized flakes from the main the slabs, in other words as hard as the silica loci at West Athens Hill were examined to composing them. This amounts to a hardness

73 of 7 in the Mohs scale. Only objects of stone dominance of small, shallow scratches and that were available to the prehistoric Indians grooves with V-shaped cross-sections? could have created the grooves and scratches. The writer previously suggested that the The writer’s experiments using unprove- principal function of the grooved abraders nienced “junk” bifaces from surface collec- was to sharpen bone awls or needles (Ritchie tions on sandstone slabs had suggested that and Funk 1973:27), but on reanalysis it seems the grooves could not have been produced by obvious that at best it was a secondary func- the chert knapper either: (1) hammering on tion. On several specimens the relatively cores, bifaces, or flakes placed edge perpendi- large, deep grooves occur on the same tool cular to the flat surface, or (2) grinding the with shallower grooves and fine scratches or edges of bifaces using a prolonged back-and- striations. This suggests that the larger forth motion (Funk 1976:217). grooves represent prolonged repetition of the The first action would have produced same action that produced the small, shallow irregular, short linear scars at odd angles to grooves and striations. If the tools with deep each other that, with continued use, tended to grooves were used primarily to sharpen the spread horizontally, creating a larger, oval tips of bone tools, why are there so many scarred area. This is not the case. The second small, shallow scratches on the same tools that action would have produced long, straight would have been useless for such a task? It grooves with polished bottoms and sides and seems certain that the grooves could only U-shaped cross-sections. These types of have been produced by the back-and-forth grooves were produced by the writer on the motion of stone objects such as chert flakes or unmodified bottom surface of a slab from perhaps bifacial and unifacial tools. My exper- Area B. The grooves were quite unlike those of iments with bone tools made from leg bones prehistoric origin on the upper surface, how- of freshly killed deer failed to produce any ever. The observed grooves tend to be of var- visible effect on the much harder surfaces of ied length, width and depth, ranging from a the stone slabs. It remains possible that the few millimeters to 10 cm long. Many are V- larger, deeper grooves, once made using sharp shaped in cross-section, and some of the stone tools, were useful for sharpening the grooves have subsidiary lineaments or tips of bone tools such as awls or needles. grooves at the bottom. Contrary to the Recent experiments (in May 2002) yielded writer’s earlier published opinion (Funk more definitive results. Using an unmodified 1976), however, some of the grooves do have sandstone slab from a roadside outcrop, upon relatively broad, U-shaped cross-sections. which I placed chert blocks collected from the None of those grooves show smoothed or pol- talus at the base of the Helderberg escarp- ished edges and bottoms, perhaps because ment, I first “quartered” or broke up the 10,000 years of weathering has affected the blocks with repeated blows from a granite surfaces within the grooves, although this cobble. After vigorously striking the blocks seems unlikely. After all, grinding and use- and pieces of blocks several dozen times and wear polish have survived on bifaces for the scattering fragments over a radius of three same length of time. The abraders bearing meters, I had produced at least 15 short, irreg- grooves with U-shaped cross-sections could ular indentations or scars on the surface of the have been used for grinding the lateral edges slab. These scars were unlike those on the of fluted points, but could not have produced abraders that had not also been used as anvil- the grinding on the indented bases of the stones, but identical to those occurring along points. Small stone hones would presumably with scratches and grooves on a few abraders have better served this task, but none has been and on anvilstones per se. Next, however, I identified in the collection. But what was the applied the thinnest, sharpest edges of block function of the abraders that display a pre- fragments and flakes to the slab with a vigor-

74 ous forward-and-backward motion, pressing process, including “nubbins,” preparatory to down at the same time. As few as two or three the removal of channel flakes. But my exami- such actions produced permanent, shallow nation of all the fluted points in process from scratches in the surface of the slab. The more the site showed only two possible examples of this action was repeated, the broader and grinding on the central projecting nubbin or deeper the grooves. These indentations were striking platform. In other cases there was initially V-shaped in cross-section, but with some rounding and gloss on parts of the base heavy, continued effort became more U- near the lower corners of the preform, but not shaped as thicker portions of the chert frag- in the central portion of the base. ments adjacent to the edges intruded slightly To some extent, the deeper, broader into the slab. The edges of the chert fragments scratches and grooves on abraders could have showed abrasion damage. been created by lightly grinding the lower lat- The same results were obtained with the eral edges of fluted points. But repeated and thin, sharp edges of a milky quartz bifacial heavy application would have produced not projectile point from a disturbed surface con- only heavy grinding on the points—which we text. Just one vigorous, unidirectional applica- do not see— but also the deep, U-shaped tion to the sandstone surface created a narrow, grooves with polished sides and bottoms that linear scratch. Repeated efforts resulted in resulted from my original experiments in the slightly broader, deeper grooves. Needless to 1970s. Those types of grooves are rare on the say, the edges of the quartz point became slabs from West Athens Hill. rounded and smoothed, losing their former It is also important to note that among the sharpness. hammerstones listed for Areas A and B are My working hypothesis from these exper- two objects with flat surfaces that bear stria- iments is that the scratches and grooves on the tions and scratches identical to those on the abraders from West Athens Hill were, in large grooved stones described above. One object is part, produced by deliberate abrasion of the 114 mm long, 90 mm wide, and 46 mm thick edges, including bases, on biface preforms to and weighs 718 g. The broadest, flattest face produce striking platforms with a better shows multiple striations varying in depth, “bite” for flaking tools, such as soft hammer- length, and width that are predominantly nar- stones and antler billets. But the problem with row, shallow, and relatively faint. The other this interpretation is that I observed only rare tool is 78 mm long, 76 mm wide, and 46 mm instances of possible platform preparation thick, weighs 383 g, and has striations on both after examining a majority of the bifaces, at all flat faces, also generally faint. reduction stages, from the site. At 5 to 25 mag- The grooved abraders and hammers nification, only seven specimens displayed appear to represent an important activity on slight to moderate rounding at high points on the site, and it is difficult to escape the conclu- the edges. sion that it was associated with stone-work- Another possibility was that the abraders ing. But the specific tasks that produced the were employed to produce better striking distinctive grooves remain a mystery. platforms on the bases of fluted points in

75 ARTIFACTS FROM AREAS D, E, AND F

Area D produced 4 Stage 1 bifaces and 1 these loci are multicomponent, and like Areas Stage 3 biface, the latter a possible fluted point A and B, used during Archaic and Transitional in process, and 4 utilized flakes. Area E yield- period occupations. Debitage but no diagnos- ed 1 small, broad-bladed, narrow stemmed tic pieces were recovered at Area F. All items projectile point, 4 Stage 1 bifaces, 1 Stage 3 are Normanskill chert. Paleoindian occupancy biface that is clearly a Susquehanna knife, 4 has not been established for these loci. utilized flakes and 1 utilized core. Therefore,

76 PALYNOLOGY

Pollen samples were collected from pro- by Donald M. Lewis, Scientist, Botany, New files in three different parts of the West York State Museum and Science Service. Anthens Hill site: section W10N20, Stratum 1, The results of the study were inconclusive. Stratum 2 top, Stratum 2 bottom; section Most of the pollen grains in the samples were E20N0, Stratum 1, Stratum 2 top, Stratum 2 badly damaged through mechanical breakage. middle, and Stratum 2 base near bedrock; and The small number of identifiable grains repre- section W180S30, Stratum 1, Stratum 2 top, sented floral species that are found in the area Stratum 2 bottom. The samples were analyzed today.

77 RADIOCARBON DATING

As previously mentioned, a concentration The dispersed bits of charcoal were col- of debitage, artifacts, possible fire-cracked lected as one combined sample. It was sub- sandstone fragments, and flecks of charcoal mitted to Isotopes, Inc. for radiocarbon assay occurred in the southwest quadrant of section in March 1968 in the hope that it represented E0N30. This concentration was tentatively the remains of a Paleoindian hearth. The designated Feature 2. resultant determination, received in April The charcoal appeared to be randomly dis- 1968, was disappointingly late; the uncali- persed within the concentration, but a linear brated date was A.D. 1650±115 years (I-3443). mass on the periphery was the same size and The charcoal probably originated in a recent shape as the charred roots of tree stumps forest fire. encountered elsewhere on the site.

78 STRATIGRAPHY,TYPOLOGY, AND INTRASITE SETTLEMENT PATTERNS

West Athens Hill is unique among New characteristically possess squarish or sharply York Paleoindian sites in several respects: (1) right angled front corners and also often grav- it was not only a chert quarry, but a workshop ing spurs, traits rarely found in Laurentian and possibly also a short-term camp site; (2) it ones). contained undisturbed, stratified artifact- Some of the bifaces in process could have bearing deposits; and (3) it produced evi- originated after the period of Susquehanna dence relating to possible family dwelling occupation at West Athens Hill. The difficulty areas or chert-knapping stations, to be is even greater with rough stone tools other described below. than hammerstones because such traits as Considering the varying extent of bulldoz- anvilstones and abradingstones are extremely er disturbance, some evaluation of the contex- rare in Paleoindian contexts, being rather tual situation is in order. Most of the cultural- characteristic of Archaic complexes. ly relevant deposits in Area B, and in the Anvilstones and hammer-anvils have, howev- northern part of Area A, seemed to be intact er, been reported for the Paleoindian manifes- and can probably be used for valid studies of tations at the Debert site, Nova Scotia artifact distribution. Area C was never dis- (MacDonald 1968), Bull Brook, Massachusetts turbed except by naturally occurring tree falls. (Byers 1954, 1955, 1966), and the Vail site, Crucial to determining the composition Maine (Gramly 1982). But “millingstones” and and integrity of the Paleoindian assemblage is “grooved stones” like those at West Athens the stratigraphic position of Archaic and later Hill are lacking on those stations, except per- artifact styles, as contrasted with Paleoindian haps for a striated stone tool found at Vail. An items. There is little difficulty in assigning the abrader and cobble choppers were part of the great majority of the uniface tools to the Debert assemblage. Paleoindian component, since most of the Despite the fact that Stratum 1 in Area B non-fluted points belong to Susquehanna tra- may be to a large degree a product of humus dition complexes in which such traits as end formation on a base originally similar in com- scrapers are rare to absent. The presence of position to Stratum 2, there is much justifica- several Laurentian point types might have tion in regarding the two deposits as physical- confounded such an interpretation, since uni- ly discrete zones with temporal significance. face scrapers are common in Laurentian Thus, the stratigraphic distribution of artifact assemblages. However, there is the further traits within these layers in Area B should consideration that the unifaces from West reflect a generalized depositional sequence, Athens Hill are much more typical of even though partly obscured by weathering Paleoindian assemblages in the Northeast, and soil formation processes during the being not only larger in general, but often dif- Holocene epoch. ferent in form from characteristic Laurentian The presence or absence of fluted and non- implements (that is, Paleoindian end scrapers fluted points in undisturbed portions of Strata

79 1, 2, or 2A would seem to be a useful aid to the clusters. Also given are totals for the sur- interpretation. Indeed, one finished fluted face and excavation units in Area A, as well as point and six fluted points in process occurred from Area C. The totals for each category or in Strata 2 or 2A, whereas the non-fluted type on the entire site appear in the last col- points were all found either in Stratum 1 or on umn. the surface. Several abradingstones, anvil- A high percentage of utilized flakes and stones, and the possible petroglyph were cores occur in all three loci. This is to be recovered from Stratum 2 or its cognates, as expected on a quarry-workshop–camp sta- was a full complement of end scrapers, side tion, where chert rejectage littered the ground, scrapers, and bifaces. ready to hand whenever an Indian had sud- There seems to be little reason to doubt den need of a cutting or scraping edge. The that except for a small percentage of the col- relatively high frequency of hammerstones is lection, artifacts from West Athens Hill pertain also explained by the nature of the site. These to the Paleoindian occupation. The exceptions tools, based on cobbles imported from nearby include the late point styles of stemmed and stream beds or till exposures, must have notched form, the Susquehanna knives, the served many functions, including the extrac- drill tip, and possibly the small tapering- tion of chert from its matrix, percussion chip- stemmed biface (see Figure 31, no. 4). ping of artifacts, breaking of animal bones to The stratigraphic distribution of all arti- obtain marrow or make bone tools, and so on. facts recovered from Areas A and B, in terms The many utilized flakes, scrapers, knives, of general categories, is presented in Table 32. and other tools reflect considerable domestic Regarding the vertical spread of excavated and industrial activity at West Athens Hill. artifacts, Stratum 1 was slightly more produc- Chert, obviously, was quarried and made into tive than Stratum 2. About 77 percent of arti- various objects. Bifaces were roughed out, facts with provenience occurred within Area then made into points and knives. Fluted B. Even if the 348 items marked “general sur- point preforms were often broken during end- face,” nearly all of which came from Area A, thinning. Many of the finished (Stage 4) fluted are considered in the calculations, at least 58 points seem to have been broken in use, either percent of the entire collection was found in as knives or as weapon tips. The bifacial the hollow. industry on the site was directed mainly to the Frequency distributions of individual hunting and butchering of game. traits in Area B are summarized by stratum in Uniface scrapers have often been regarded Table 33. Frequencies are given for artifact as hide-working tools, but William A. Ritchie types from all excavated contexts, including (personal communication 1971) once suggested

Table 32. Depth distribution of general artifact categories in Areas A and B (all provenience units combined).

Chipped and Rough Utilized Cores and All Artifacts Stone Artifacts Flakes Combined Provenience No. Pct. No. Pct. No. Pct. Surface 392 44.7 173 23.0 565 35.0 Stratum 1 265 33.8 304 45.7 569 38.0 Stratum 2 193 21.5 208 31.3 401 27.0 Totals 850 100.0 685 100.0 1,535 100.0

80 5 9 1 1 1 2 7 4 7 5 10 11 1 1 3 16 3813 104 59 36 24 129 39 50 207 1313 36 25 13 42 16 56 39 236 Area A Totals 2 2 1 1 2 3 9 4 8 1 1 40 37 19 100 211 23 69 103 903 171 1583 1,228 565 2,763 Area B 3 100.0 42 100.0 100.0 7 100.0 5 100.0 8 100.0 1 100.0 5426 100.0 100.0 2133 100.0 92 100.0 15 100.0 100.0 1014 100.0 100.0 21 100.0 17 100.0 970 Totals (in context), Area C Surface and Site 40.7 46.2 47.6 45.4 38.0 53.3 66.7 25.0 50.0 28.6 40.5 509 100.0 10.0 43.0 25.0 28.6 17.6 53.2 128 100.0 100.0 Area B, Stratum 2 8 2 1 1 2 1 6 2 6 3 1 22 12 10 15 35 68 206 401 Area B, Stratum 1 5 100.0 7 46.7 1 33.3 31 75.0 50.0 5 71.4 9 90.0 11 52.4 321415 59.3 53.8 18 71.4 57 54.6 14 62.0 82.5 No. Percent No. Percent No. Percent 303 59.5 569 Types Archaic points Fluted Points Fluted points in processFluted point fragmentsStage 1 bifaces 8Stage 2 bifaces Stage 6 3 bifaces Misc. bifaces 57.0 End scrapers 75.0 Side scrapers Flake knives Retouched Flakes and Misc. Unifaces Drills Pieces esqillees Gravers Utilized Flakes Cobble Hammer-stonesAnvilstones Anvil-hammer-stones 60Milling-hammer-stone Teshoa Abrading-stones 46.8 Petroglyph (?) Totals Table 33. Depth distribution andsurface. frequencies The of numbers artifact are types for with all provenience excavated data units from in Area Area B, B plus including artifact the lists from artifact Area clusters. C, Area A and the

81 that on sites in the Northeast, they were often concentrations that were separated by three to used as wood-working implements, an impor- eight feet of open space. Some of the concen- tant end product being dart or spear shafts. trations were of oval shape, others were rather This notion would seem to have some support amorphous, while the remainder were more in the frequent pattern of wear on scrapers, or less well-defined arcs or semicircles. consisting of edge-crushing or the detachment In places undisturbed by bulldozing, it of small step flakes by pressure against some seems likely that the artifacts occurred in their hard object. But some scrapers also display original positions, since they were within a edge-rounding and polish or gloss, occasional- “silt till” that tends to be compact and resist- ly even manifested as heavy grinding. Tools ant to slope wash, solifluction, and other used in working the softer, more pliant materi- processes that might transport items from one al of hides should logically be expected to dis- location to another within the hollow (G. play smoothed, polished edges (Keeley 1980; Gordon Connally, personal communication, Semenov 1964: 85–93). Another problem is that 1970). it is hard to visualize the use of the small, con- An attempt was made to distinguish and vex bits of end scrapers on rounded wooden designate by number the various clusters. spear and dart shafts. It would be difficult to They are presented as published in the origi- maintain firm contact between the scrapers and nal 1973 report, but reanalysis throws consid- the shafts. As also noted, experiments by erable doubt on my original interpretations. Keeley (1980) and others indicate that crushing and rounding/gloss may result from working Stratum 1 bone and antler. (Solid Symbols on Map, Figure 45) The Paleoindians at West Athens Hill must have possessed a varied inventory of tools in Cluster 1. In sections W10N0, W10S10, bone, antler, and perhaps even ivory, long since E0N0. A rather neat arc about eight feet in disintegrated by high soil acidity. Some of these diameter, which forms about one-half a circle. implements would have served the chert-knap- It intersects at one end with Cluster 2. It con- per’s art. Bone awls and needles were doubt- tains at least 17 items: 1 fluted point plus 3 in less important traits, essential to the skin and process, 1 Stage 1 biface, 3 Stage 3 bifaces, 6 hide industry. They may have been sharpened side scrapers, and 3 end scrapers. No corre- on the grooved abradingstones. Other tools of sponding exists in Stratum 2, which produced bone and wood would have been well-suited 3 pieces positioned horizontally along the arc. for preparing hides in the form of beamers, Cluster 2. In sections W10N0, W10N10. scrapers, knives, and so on. Another good arc about eight feet across, also Our decision to record the precise loca- forming one-half circle. It contains at least 10 tions of all artifacts recognized as such during items: 3 fluted points in process, 1 Stage 2 excavation may have yielded data relevant to biface, 1 side scraper, 1 end scraper, 1 internal settlement patterns. The clustering of retouched flake, and 3 cobble hammerstones. artifacts and debitage around stake W170S30 It seems to match very well with Cluster 10 in in Area A has been previously mentioned, but Stratum 2. the extensive disturbance to the area pre- Cluster 3. In section E0N10. A small, fairly cludes any effort to discern patterning in the tight group roughly in the form of an arc. At plotted positions of artifacts. For Area B the least 11 items: 1 possible fluted point tip, 1 horizontal distribution of artifacts from both Stage 2 biface, 5 side scrapers, 1 flake knife, 3 Stratum 1 and Stratum 2 is shown in Figure cobble hammerstones. If superimposed on 45. Study of this map shows that artifacts are Cluster 13 in Stratum 2, a convincing pattern not randomly dispersed through the hollow. emerges. Many items occurred in reasonably compact

82 Figure 45. Artifact clusters possibly representing localized activity concentrations in Area B.

83 Cluster 4. In sections E0N0, E10N0, lection, and are only slightly trimmed in each E10S10. A large concentration of 32 items: 3 case along one short edge. fluted points, 1 fluted point tip (fits to the base Cluster 9. Confined almost entirely to sec- of one point), 5 Stage 1 bifaces, 2 Stage 2 tion E0N20. A semicircle about 7 feet across. bifaces, 1 Stage 3 biface, 1 miscellaneous May include 11 artifacts: 2 fluted points, 2 flut- biface, 3 end scrapers, 5 side scrapers, 1 flake ed points in process, 3 side scrapers, 1 abrad- knife, and 10 cobble hammerstones. ingstone, and 3 hammerstones. May corre- Cluster 4 as a whole is roughly oval, about spond to a small group of pieces in the same 12 feet by 10 feet. With the possible exceptions part of Stratum 2, attributed to amorphous of 1 side scraper and 1 fluted point, this area Cluster 11. produced no artifacts in Stratum 2 that might be attributed to the cluster. Stratum 2. Cluster 5. In sections E10N10, E20N10. A (Open Symbols on Map, Figure 45) vaguely defined short arc which seems to comprise 1 bifurcated-base Archaic point, 1 Cluster 10. In sections W10N0, W10N10. A fluted point in process, 2 Stage 1 bifaces, 2 side well-defined arc about 8 feet in diameter, scrapers, 1 retouched flake, and 1 hammer- which is congruent with Cluster 2 in Stratum stone. This area is nearly blank in Stratum 2. 1. It is made up of 17 apparently associated Aside from the fluted point preform, which is pieces: 1 Stage 1 biface, 2 Stage 2 bifaces, 1 located on the periphery of the main group, miscellaneous biface, 3 side scrapers, 2 end none of the other artifacts is a particularly scrapers, 1 flake knife, 1 retouched flake, and good example of a Paleoindian type, and 6 cobble hammerstones. could have been associated with the bifurcate Cluster 11. Almost entirely within sections point. W10N20 and E0N20. Tentatively designated Cluster 6. In section E0N40. A small isolat- since the artifacts are not tightly concentrated, ed oval group of artifacts measuring 3 feet by but rather diffusely spread. Types: 1 possible 6 feet. Includes 1 Stage 3 biface, 1 miscella- fluted point tip, 2 fluted points in process, 1 neous biface, 1 side scraper, 1 end scraper, 1 Stage 1 biface, 2 Stage 2 bifaces, 3 Stage 3 flake knife, and 3 cobble hammerstones. This bifaces, 4 side scrapers, 3 end scrapers, 1 flake spot is devoid of artifacts in Stratum 2. knife, and 15 cobble hammerstones. Cluster 7. In section W10N40. Another iso- In contrast to Cluster 8, overlying Cluster lated group of oval shape, about 7 feet by 3 11 in Stratum 1, no Archaic points are present. feet. This time with 9 items: 1 fluted point, 2 Furthermore, the two end scrapers are of clas- side scrapers, 1 flake knife, 3 retouched flakes, sic Paleoindian type, and one is of exotic chert. and 2 cobble hammerstones. Stratum 2 here is The side scrapers are good “Enterline” forms; empty except for 1 side scraper. one is of Upper Mercer, Ohio chert, and anoth- Cluster 8. In section W10N20. A poorly er is Oriskany chert. defined linear cluster comprising 2 Late Cluster 12. In sections E0N0, E0N10. A neat Archaic points (1 Orient Fishtail, 1 untyped arc about 6 feet wide which seems to match stemmed), 1 Stage 2 biface, 1 side scraper, 2 nicely with Cluster 3 in Stratum 1. Artifacts retouched flakes, 1 anvilstone, and 1 abrad- are: 3 Stage 1 bifaces, 1 miscellaneous biface, 3 ingstone. This cluster cannot convincingly be side scrapers, 2 end scrapers, 1 flake knife, and related to the more numerous and scattered 1 cobble hammerstone. objects from Stratum 2 in the same area. It is also important to note that none of the uniface tools in the cluster area, can by themselves by Strata 1 and 2. demonstrated to have Paleoindian affinities. Cluster 13. In sections E20N0, E30N0. A They are smaller than most unifaces in the col- roughly curvilinear arrangement of artifacts,

84 identified as 2 end scrapers, 3 side scrapers, 1 Clusters 1, 4, 5, 6, 7, 8, and 9. Clusters 5 and 8 retouched flake, 1 Stage 1 biface, and 3 cobble may be partly or wholly of Archaic origin. hammerstones. Comparisons of the frequencies of artifact types in the presumed stratigraphic sets of DISCUSSION clusters (Table 34) reveal minor differences between sets. The relative proportions of such The great majority of artifacts from the traits as fluted points (8.1–14 percent), Stage 1 Stratum 2 clusters conform well to ideal bifaces (3.1–8.1 percent), Stage 2 bifaces Paleoindian forms, in contrast to the items (2.9–8.1 percent), end scrapers (9.4–10.2 per- described for Clusters 5 and 8 in Stratum 1. cent), side scrapers (12.5–24.5 percent), and If these clusters are hypothesized to repre- flake knives (2.9–6.1 percent) show relatively sent structural features resulting from pat- small fluctuations from set to set. terned human behavior, it remains to be seen Hammerstones are most common in Cluster whether they have chrono-stratigraphic sig- 11 (Set A) and are more common in Set B than nificance. in Set C. These differences are fairly small, Clusters 1, 4, 5, 6, and 7 appear at first within 3 to 5 percent in most cases, varying by glance to be reasonably discrete entities, all of 12 percent in the case of side scrapers and 20 which pertain to Stratum 1. In each case only percent in the case of hammerstones, and a few items, or none at all, were found in probably not significant, considering the small Stratum 2 in the same locality, and these could samples involved. have been pressed, trampled, or otherwise One important qualitative difference intruded into that deposit from the higher emerges: three Archaic points are confined to level containing the cluster. Set C, believed to be the youngest of the clus- Certain clusters of objects in Stratum 1 ters. seem to be closely matched in location and Further comparisons have been made of outline with other clusters in Stratum 2. These the type frequencies in the hollow in Area B, congruent pairs, therefore, may be actual fea- employing cluster sets A, B, and C, separately tures representing single archaeological or combined, and the type frequencies for the episodes. These paired clusters are nos. 2 and site as a whole (Table 34). Whereas the fre- 10, and 3 and 12. Clusters 8 and 11 do not form quency of projectile points remains fairly close a congruent pair; separately Cluster 8, per- from cluster to cluster, it is rather higher in haps of Archaic origin, is better defined than is Area B than for the site as a whole. The varia- the combination of it with Cluster 11 and, as tion in percentages of other biface categories already mentioned, it lacks artifacts that are between the clusters and the total collection is typologically convincing Paleoindian forms. It about 3 to 4 percent. For unifaces, the corre- is noteworthy that no clearly patterned cluster spondences are very close, generally within 2 occurred in Stratum 2 which was not a mem- percent. The greatest range in percentage ber of a paired cluster. Cluster 11, amorphous applies to cobble hammerstones. as it is, may entirely precede Stratum 1 in age. The total number of items other than uti- It may therefore be suggested that Cluster lized flakes in the clusters is 183, or slightly 11 was possibly the first of the whole group to under one-fourth of the total collection from be deposited, and associated exclusively with Areas A and B; yet Area B seems to be a micro- Stratum 2. Next in age would be the paired cosm of the whole site with respect to type fre- clusters, nos. 2 and 10, and 3 and 12, which quencies. This implies that Area A had a pat- overlap from one stratigraphic zone into the tern of type distribution similar to that of Area other. Cluster 13, composed about equally of B, so that basically the same range of human artifacts from both strata, may belong to this activities is represented in each locus (but see group. Last to be deposited, and youngest in the discussion of Area C below). age because limited to Stratum 1, would be

85 Presence–absence analysis of each individ- that there was no significant division of activ- ual cluster or pair, and comparisons between ities within the horizontal space of the hollow. clusters, show that some traits occurred in all This holds true even when the cluster sets are clusters, some in most, and others in only one studied as possible discrete episodes in time. or two. Thus, side scrapers appeared in every The clusters—no matter what their reality, cluster; cobble hammerstones in all but two; or lack of it, as loci of tasks performed inde- end scrapers in all but five; and fluted points, pendently of other loci—were products of blanks, or fragments thereof in 7 out of the 13 human behavior. Singly or in groups of small clusters. Other items were more evenly dis- size, it would seem that people remained in tributed, but with one exception, every cluster each spot long enough to carry out certain had at least one example of a biface belonging activities: chert-knapping, weapon-making, to one of the Stages 1, 2, or 3. possibly butchering and hide-working. The similarity in range of artifact cate- Despite the absence of hearths or post molds, gories from one cluster to another, no matter it is possible that each cluster represents a what the variation in frequency, may indicate nuclear family domicile. The artifacts were

Table 34. Comparisons of frequencies of artifact classes in Area B between cluster groups and between the combined clusters, Area C, and the site as a whole. Set A cluster 11; Set B clusters 2, 3, 10, 12; Set C clusters 1, 4-9. Utilized flakes not included.

Set A Set B Set C Clusters Area C Total Site Combined Artifact Traits No. Pct. No. Pct. No. Pct. No. Pct. No. Pct. No. Pct. Fluted points, whole, fragmentary, in process 3 9.4 4 8.1 14 13.7 21 11.5 16 4.9 77 6.5 Archaic points 3 2.9 3 1.6 10 0.8 Stage 1 bifaces 1 3.1 4 8.1 9 8.8 14 7.6 103 31.7 207 17.5 Stage 2 bifaces 2 6.3 4 8.1 3 2.9 9 4.9 40 12.3 104 8.8 Stage 3 bifaces 3 9.4 5 4.9 8 4.4 8 2.5 42 3.6 Misc. bifaces 2 4.1 2 2.0 4 2.2 2 0.6 36 3.1 End scrapers 3 9.4 5 10.2 7 6.9 15 8.2 37 11.4 129 10.9 Side scrapers 4 12.5 12 24.5 24 23.5 40 21.9 19 5.9 211 17.9 Flake knives 1 3.1 3 6.1 3 2.9 7 3.8 39 3.3 Pieces esquillees 2 0.6 7 0.6 Retouched flakes and misc. unifaces 2 4.1 7 6.9 9 4.9 23 7.1 56 4.8 Cobble hammerstones 15 46.8 13 26.5 22 21.6 50 27.3 69 21.2 236 20.0 Other rough stone tools 3 2.9 3 1.6 6 1.9 24 2.0 Other chipped stone 2 0.2 Totals 32 100.0 49 100.0 102 100.0 183 100.0 325 100.1 1,180 100.0

86 perhaps discarded or lost within or around Areas A and B. I suspected that this may have small huts or lean-tos that left no permanent resulted from the application of classificatory imprint on the underlying soil. Thus, each criteria that differed between the 1973 analysis cluster would represent at most occupation by and the present one, since 842 “retouch flakes” one small family or work party within a single are listed for Area B in Table 18. I wondered season. But the lack of hearths or other prehis- whether some of those could actually be toric features poses a serious problem for the biface thinning flakes. However, my reexami- notion that people actually camped on the nation (in May 2002) of nearly all of the small site. flakes from Areas A and B failed to find any Furthermore, as a result of my reanalysis biface thinning flakes in addition to those pre- of the reality of clusters 1, 2, 3, 10, 12, and 14 is viously identified. This is quite enigmatic highly questionable. They adjoin the exten- because on loci where so many bifaces were sively disturbed area at the center of the exca- produced, one would expect quite a large vations, and the few artifacts found in those number of such flakes. Also, Beth Wellman units were no longer in their original posi- and I failed to find a single example of a chan- tions. If the original distribution of those arti- nel flake from Areas A and B, and found just facts was known, plotting them on the map three in Area C, although the manufacture of would have changed the configuration of fluted points was an important activity there. some clusters and perhaps even rendered Three distal fragments of thick, narrow- them more amorphous. The other clusters bladed projectile points, probably of either the may still be valid as independent activity Bare Island or Normanskill type (Ritchie 1971) areas. and therefore of Late Archaic age, were In Area C, artifacts did not occur in readi- encountered at Area C, and at least two bifaces ly delineated clusters. Instead, an irregular appear to be Susquehanna knives, indicating distribution is evident within the grid excava- that Area C, like the other loci, was occasion- tion (see Figure 40). Most of the artifacts in ally occupied by Archaic and Transitional Area C were hammerstones and bifaces in (Terminal Archaic) people. process, a fact that is consistent with the chert It would be difficult, if not impossible, to exposures and the large quantities of debitage. separate all the debitage of the Archaic and The locus was, however, more than just a chert Transitional components from debitage of the quarry and workshop for production of early Paleoindian component(s). But the production stage bifaces, since fluted points, a number of of broad-bladed Susquehanna tradition end and side scrapers showing edge wear, and bifaces results in distinctive biface thinning utilized flakes were also present, suggesting flakes that might well be picked out of the such activities as hunting, hide-working, and larger sample. wood-working. Once again, no convincing The total span of Paleoindian occupancy features such as hearths, pits, or house pat- cannot be determined without datable organ- terns were encountered in the excavations. ic material. It is unlikely that only one short Analysis of artifact type frequencies in period of repeated occupations over, say, 50 Area C suggests that there are some differ- years or, alternatively, centuries of continuous ences between Areas A and B. Side scrapers occupation are represented. Sporadic, brief and end scrapers were relatively less common visits by bands or work parties over a longer in Area C than in the other loci. No flake period are more likely, and West Athens Hill knives, as such, were identified there. Bifaces may have been abandoned for years at a time. in process were slightly more common. But the total period of on-and-off occupancy Surprisingly, the Area C debitage produced by Paleoindians may have amounted to cen- 414 examples of biface thinning flakes, in con- turies. From time to time, work parties were trast to only 12 identified in the debitage from almost certainly attracted to other equally fine

87 sources of chert within a few miles of the hill. small locus. We also succeeded in locating and The range in fluted point size and morpholo- excavating stratified artifact-bearing sedi- gy may be a clue to the total period of ments in Area B (the second goal). Concerning Paleoindian visits to West Athens Hill, since the third goal, we recovered 2,763 artifacts and large and small examples are present, with over 20,000 pieces of quarry debris and deb- some variation in fluting technique, depth of itage from the three principal loci. This is sure- basal indentation, and other attributes. But the ly a representative sample since major por- relatively large, straight-sided Gainey-Debert tions of partially disturbed Area B and undis- style predominates. There only one or two turbed Area C were excavated. examples of a later stage, such as that exem- In terms of the fourth goal, we acquired plified by relatively small, thin, often “fish- some data, however incomplete, on quarry tail” shaped Cumberland-Barnes points (see activity. We are able to delineate stages in Figure 27, no. 4; Ellis and Deller 1990; Gramly biface production, the most important non- and Funk 1990). quarry activity on the site, and can state with How successful were we in attaining our conviction that other tasks such as wood-, research goals? Regarding the first goal, we hide-, and bone-working also occurred on the established that the Paleoindian component site. We failed, however, to meet the fifth goal was by far the most important in terms of the by finding charcoal or other organic material sheer number of diagnostic artifacts, and fur- that could date the Paleoindian occupation, thermore they were distributed over Areas A, and subsistence remains of any kind were B, and C, rather than concentrated in one absent.

88 COMPARISONS

Numerous Paleoindian sites have been Road, Swale, Twin Fields, Dutchess Quarry found and excavated throughout North Caves 1 and 8, Zappavigna, “Hallock,” Port America and parts of Latin America. Mobil, Davis, Potts, and Corditaipe. Sites are Therefore a great deal of information is now defined in contrast to isolated finds of fluted available on these oldest known inhabitants of points, and include plowed sites as well as the New World.6 More and more sites are undisturbed ones, where the projectile points being discovered, excavated, and reported in are associated with other tools in assemblages. the Northeast, but much remains to be learned Nevertheless, isolated finds are important for about all aspects of Paleoindian cultural tradi- a full picture of Paleoindian adaptations in the tions. Despite many similarities in artifact Northeast. Occasional reference will also be assemblages, there is increasing evidence of made, where appropriate, to sites located in regional and temporal variation (Gramly and other parts of northeastern North America. In Funk 1990). It is inaccurate to speak of a western New York those include the Lamb site Paleoindian “culture” in the sense of a nar- (Gramly 1999), the Hiscock site (Laub, et al. rowly defined set of traits and adaptive strate- 1988), and the Arc site (Gramly 1999; gies, replicated over and over again wherever Tankersley, et al. 1997). found. Nearly nothing is known about non- This analysis considers aspects of cultural lithic artifacts that must have comprised major geography that guided and constrained the portions of material culture. In addition, there modes of aboriginal adaptation to the middle is an almost complete lack of data regarding Hudson valley environment. Topography and skeletal biology, burial procedures, burial cer- drainage patterns would have tended to con- emonialism, social organization, and so on. trol Paleoindian movements across the land- Depending on the challenges of local and scape. People on foot would have followed regional environments and the technological routes of least effort in search of desirable means of dealing with them, subsistence and campgrounds and special-purpose stations settlement patterns, economic organization, located within convenient distances of ideology, ritual, trade and exchange, and other resources critical to their physical survival, traits must have varied in time and space. This such as fresh water, food, and raw materials variability may, however, prove difficult to for garments, shelters, tools, and weapons. discern in the archaeological record. Ancient peoples also needed to find sources of A comparative analysis is needed in order chert and other materials for stone artifacts. to place West Athens Hill in the larger context Local resources of food animals and plants of northeastern Paleoindian manifestations. had to be adequate to their needs. Another Detailed comparisons will be primarily con- factor might have included territorial behav- fined to the other 11 sites on record for eastern ior, involving an awareness of the territorial and central New York state, namely: Kings boundaries of other bands. Although there

______6 Admittedly, there is continuing controversy over the possibility of pre-fluted point occupations.

89 might have been competition between bands sources of the stone, and whether the group for important resources, occasional peaceful represented a pioneering population bringing interaction would be crucial in the search for exotic materials from other regions or were mates, social reinforcement, shared ritual long-established residents who had located behavior, trade partnerships, and so on. And and were exploiting locally available sources. while we lack sufficient information to define The amount and areal extent of chipping “catchments,” some very preliminary and debris was determined by the reasons that a general models of subsistence-settlement pat- site was occupied or utilized, the length of terns may be proposed. time it was occupied, the number of repeated Factors considered in the intersite compar- occupations, the size of the occupying isons include: elevation above sea level; relief; group(s), and the status of chert-knapping in local topographic setting; structural attributes; relation to all other activities on site. relation to bodies of water; relation to sources The amount of exotic stone in assemblages of raw material such as chert; relation to food was determined by distance from the source, resources; chronology; place in regional devel- the size of Paleoindian territories (larger terri- opmental patterns; characteristics of artifact tories would have familarized a group with assemblages; within-site activities and organi- resources occurring over a broad area), trade zation; and position in regional settlement relations with other bands living at or near systems. Thhe interrelationship of these vari- outcrops of the stone, and whether the occu- ables can be summarized as follows: pants of the sites under study were pioneers Topography and associated resources on carrying material from other regions or long- the landscape that satisfy basic human needs established residents who had found and (food, water, raw materials, stable living sur- come to rely on local raw materials. faces, shelter) are primary determinants of settle- The ratio of the quantities of debitage to number ment patterns. Most important, particular geo- of artifacts was determined by the kind of site graphic regions must have sufficient food and (the purposes for its occupation and the kinds water to support people bearing a hunting/ of associated activities), the types of artifacts fishing/gathering technology. manufactured, the average size of artifacts, and Site locations were chosen that were within the relation to chert outcrops (primary produc- convenient (time and energy saving) distances tion vs. retouch or resharpening). of resources. Multiple sites may have been Wear patterns on artifacts were determined occupied/used by a band if needed resources by the nature of tasks performed and the were not located within convenient distances kinds of materials to which tools were applied of each other, seasonally or otherwise. in those tasks. Living surfaces had to be level to only mod- Productivity was determined by the kinds erately sloping, as well as dry, in order for of activities on a site (e.g., chert-knapping, people to comfortably and efficiently carry out hunting, butchering, food preparation, cook- the various tasks and activities of daily life. ing, wood- or hide-working) as well as the Raw materials, especially chert, and pre- length of occupation, number of occupations, sumably also wood, bone, and antler were size of occupying groups, and other factors needed to make the weapons and tools that leading to increased accumulation of refuse. enabled such bands to obtain and process In the case of Paleoindians, we are considering their food as well as obtain animal hides for only the quantities of stone tools and debitage clothing, build shelters, and so on. since most other residues including features, The kinds of chert-knapping debris occurring organic remains, burials, and so forth are on archaeological sites were determined by absent from the sites. the types of artifacts manufactured, the reduc- The size and weight of particular artifact types tion procedures followed by particular were determined by the kinds of tasks or groups, the distances of campsites from activities in which they were employed (func-

90 tional demands), the strength of humans and slopes down to the east on the edge of a using them, and the size of chert blocks, vein clay terrace that lies some 150 feet above the plates, or nodules in available chert sources. waters of nearby Lake Champlain. Artifact type frequencies were determined by the kinds and relative importance of tasks Local topographic setting: West Athens and activities that involved the items (e.g., Hill is part of a system of rocky ridges, many hunting/butchering, fishing, plant food pro- chert-bearing, running north–south in Greene cessing, hide-working, working of bone and County. They are adjoined on east and west by wood, biface manufacture). low-lying flats and creek valleys that are also The size of occupied areas was determined oriented generally north–south. Kings Road by structure and local topography (caves/ and Swale are on low-lying clay flats, adjoin- rockshelters vs. open-air sites, relief, slope, ing a small creek east of the above-mentioned amount of living surface available), the nature ridges and the Helderbergs. The Dutchess of activities, the size of occupying groups, the Quarry Caves, however, are in hilly terrain number of repeated occupations, and the that interfingers with parts of the Black Dirt length of occupation. The larger the group, the Area, a vast former marsh that was the bed of longer or the more frequently repeated the a proglacial lake within the present Wallkill occupation, the more refuse would be pro- River Valley watershed. Port Mobil is a series duced, the more widely dispersed it would be of loci situated on low terraces and knolls as a result of shifting loci of activity. adjoining the Arthur Kill, presently a salt water channel but in late Pleistocene times a Specific intersite comparisons follow: fresh water stream. Zappavigna and “Hallock” are on low rolling terrain outside Elevation: West Athens Hill reaches a the Black Dirt Area and within a fairly short maximum elevation of just over 400 feet. The distance of the Wallkill River. Twin Fields is on Dutchess Quarry Caves are at about 580 feet. a late-glacial terrace along the Dwar Kill, a The Corditaipe site sits at about 500 feet, Potts Wallkill tributary. Davis is on a similar terrace at 410 feet, Port Mobil at 10 to 50 feet, adjacent to Lake Champlain. The Corditaipe Zappavigna at 400 feet, “Hallock” at about the site lies on an outwash terrace next to a same elevation, Kings Road and Swale at 140 Mohawk River tributary, and the Potts site sits feet, Davis at 200 feet, and Twin Fields at 300 atop a drumlin in generally flat terrain on the feet. In Maine, the Vail site is situated at 1,870 old bed of glacial Lake Iroquois. The Arc site feet and Michaud at 250 feet. But except when occupies a flat, gently sloping field close to sites are close to tidewater, elevation by itself Oak Orchard Swamp and the Lamb site is on is not particularly meaningful; relief, however, an outwash plain next to a kettle hole bog. informs us about the relationship of a site to In contrast, the Templeton site in north- its immediate surroundings in the continental west Connecticut is on the floodplain of the interior. Shepaug River (Moeller 1980), and so far is the only reported instance of this kind. It discov- Relief: West Athens Hill is the highest hill ery suggests that many more Paleoindian sites in Greene County east of the Helderbergs, ris- remain to be found deeply buried on the ing over 250 feet above the adjoining terrain. floodplains of northeastern rivers. The Dutchess Quarry Caves on Mount Lookout are also some 200 feet above the local Structure: All of the sites except the terrain. All of the other sites are at lower relief, Dutchess Quarry Caves are open-air stations. generally less than 30 feet above adjacent Those stations vary in associated physical landforms. An exception is Davis, which occu- characteristics, including bedrock lithology, pies terrain that rises much higher to the west soils, slope of occupied surface, area of occu-

91 pation, nature of cultural deposits, presence or Swale sites are on nearly level clay flats, the absence of features, the impact of modern cul- Davis site is on land sloping gently toward a tivation, and other forms of disturbance. West bluff overlooking Lake Champlain, the Potts Athens Hill is the only Paleoindian station in site lies atop a drumlin sloping down from the New York State that is a quarry-workshop, occupied area on two sides, and the directly associated with chert-bearing out- Corditaipe and Lamb sites are on outwash crops (Flint Mine Hill, a much more heavily plains. Looking out of state, the Whipple site utilized quarry and workshop site, has pro- is on a high late-glacial terrace, the Vail site duced two fluted points from the top of the occupies a 5 percent slope down to the ridge, but the site is multicomponent and arti- Magalloway River, the Michaud site is on an facts of Archaic and Woodland periods are outwash plain with about a 2 percent slope, very abundant on and around the hill). West and the Templeton site is on a flood plain. Athens Hill is also the only local Paleoindian Soils on the open-air sites are generally site with sharply contrasting intrasite microto- gravel, sand, and silt loams based on glacial pography, namely, artifact-bearing loci on drift deposits. Exceptions include the “cave knolls and a hollow within the summit area as earths” that contained cultural material with- well as on a bench on the east side of the hill, in the Dutchess Quarry Caves, the Lake well below the elevations of the other loci. Albany clay-silts at Kings Road and Swale, All of the other open-air sites are on flat and the fine sand loam at the Templeton site. glacial terraces, lake beds, or the crests of low The agricultural potential of these soils is not ridges and knolls, and in one case on a flood relevant to interpreting Paleoindian subsis- plain terrace. Except for West Athens Hill and tence and settlement, in the absence of culti- the caves, the New York sites have been gens. But differences in the fertility of soils on plowed for many years, disturbing the for- or near sites could have affected the abun- merly intact occupational remains, destroying dance and distribution of wild plants that features and middens, and dislodging arti- were some part of the Paleoindian diet. facts from their original positions. Apart from There is variable information on the occu- the Dutchess Quarry Caves, there are no other pied areas of sites. Most of the open-air sites confirmed instances of fluted points found in have been plowed for many years. Not only New York caves and rockshelters. does this disperse artifacts horizontally from Precise information on the slope of occu- their original positions, thus artificially pied surfaces is lacking for some sites. For enlarging sites, but it provides ready access example, at Port Mobil, artifacts occurred on for surface collectors whose activities can both low, level terraces and on knolls, but the reduce assemblage size and differentially data on slope have not been recorded. At remove certain artifact types. For example, Zappavigna, slope varied from level to about one collector of my aquaintance collected only 2 percent. Although the “Hallock” site has not complete projectile points, disregarding the been mapped, it is on a ground surface slop- numerous fragmentary points in the plowed ing at about 2 percent toward the Wallkill fields; yet his collection contained over 4,000 River. points. Another collector ignored scrapers, Occupied surfaces within the Dutchess focusing his attention on projectile points. Quarry Caves were nearly level despite the Some individuals are interested only in objects steep talus slope outside their mouths. The of historic period origin, such as coins, gun deglacial terrace at Twin Fields is level. Slope parts, bullets, and ceramics. This “thinning varies at West Athens Hill, but is nearly level out” may diminish the visibility of a site in a on Area A, level to gently sloping at Area B, cultivated field but also modify the original level to gently sloping at Area C, and steeper frequencies of artifact types in an assemblage. on the upper hillsides. The Kings Road and Sites may even be literally erased by intensive

92 collecting of waste flakes and cores as well as presently visible on the summit or slopes of artifacts, as reported in the Unadilla and West Athens Hill or Mount Lookout, although Susquehanna basins (Whitney 1974). running water is within one-half mile (800 m) In other cases, severe disturbance pre- in each case. The Davis site is within 150 yards cludes accurate estimation of site area. Also, (135 m) of the shore of Lake Champlain and occupation by multiple aboriginal groups may possibly have been closer to water during over millennia may obscure the distribution of the period of the brackish Champlain Sea. items from particular periods, complicating Ritchie (1965: 19–22) did not report the exis- the task of determining the areal extent of tence of creeks near the fluted point find- Paleoindian components on a site. spots, but aerial photos of the local topogra- The smallest Paleoindian sites in the study phy indicate a small stream was close by the area are: Dutchess Quarry Cave No. 1, area 45 site. “Hallock” is on high ground gently slop- square meters; Dutchess Quarry Cave No. 8, ing down to the Wallkill River and there is a area 30 square meters; Twin Fields, area 230 spring nearby. The salt water branch of the square meters; Kings Road and Swale, each Hudson River estuary that presently adjoins with an area of about 400 square meters; and Port Mobil was a fresh water river in late-gla- Zappavigna, area 1,800 square meters. The cial times, but there is also a small creek and Lamb site in western New York was small, at former wetland near the site. about 300 square meters. The Vail site in Maine was estimated at 5,600 square meters, Relation to sources of lithic raw material: and West Athens Hill at 8,100 square meters, Cherts of good quality for stone tool manufac- making it the largest of these sites. ture were available on, or very near, the West Two sites that might be considered “dif- Athens Hill, Dutchess Quarry Caves 1 and 8, fuse” also have large estimated areas: Potts Kings Road, Swale, “Hallock,” Zappavigna, occupied roughly 5,500 square meters, and Davis, and Twin Fields sites. A considerable Corditaipe roughly 6730 square meters. These variety of cherts was to be found in the out- sites consisted of some eight and four different crops of Cambrian and Ordovician rocks of the loci, respectively, based on the presence of Wallkill Valley (La Porta 1996) and these pre- artifact and debitage clusters, separated by dominated at the Zappavigna, Dutchess relatively barren areas within the large farm Quarry Caves 1 and 8, and “Hallock” sites, all fields where they occurr. in Orange County. Normanskill and The Port Mobil site was undoubtedly rela- Martinsburg cherts, as well as the Wallkill tively large, but accurate estimates of its occu- Valley cherts, were available to the occupants pied area were precluded by disturbance by of Twin Fields. West Athens Hill itself is a modern industrial activity. Ritchie (1965: 21) Normanskill chert source, and both Kings estimated the area where Paleoindian diag- Road and Swale are located within a half mile nostic items were found at the Davis site to be (800 m) of excellent sources of the same materi- one-third of an acre, or about 1,350 square al. High-quality chert occurs only a few kilometers, but the picture was confused by the meters south of the Corditaipe site, in the multicomponent nature of the site.No infor- Onondaga carbonate belt. The occupants of the mation is presently on record concerning the Potts site had to go farther south (30 to 40 km) size of occupied area at the “Hallock” site in to obtain Onondaga chert for their chipped Orange County, New York. stone industry. Some local pebble jaspers were available to the residents of the Port Mobil site, Relation to potable water: All of the sites, but they also obtained Hardyston chert from except West Athens Hill and the Dutchess eastern Pennsylvania at a much greater dis- Quarry Caves, are close to fresh water (i.e., tance. Glacially imported cobbles of chert and within 330 feet [100 m] or less). No springs are other stones were also available prehistorically

93 in areas south of bedrock sources like the County. Few such finds were located between Onondaga carbonates. Albany and Syracuse, but are fairly numerous from the Finger Lakes to Buffalo. With the Relation to sources of plant and animal exception of the Byron Mastodont, now food (subsistence): All of the sites are located known as the Hiscock site (Laub, et al. 1988), within the deciduous forest environment that no Paleoindian artifacts are yet reported in has been in place since the close of the association with the bones of Pleistocene crea- Oak–Pine period some 7,500 years ago. But tures anywhere in the state or in adjoining presumably the plant and animal species at parts of North America. However, it is quite the time of occupation, during the terminal possible, even likely, that the inhabitants of Pleistocene, offered rather different subsis- West Athens Hill, Kings Road, and Swale tence resources to Paleoindians. Fossil preyed upon mastodont, caribou, and some remains of caribou, mastodont, moose-elk, other mammals whose bones are found in dire wolf, ground sloth, cave bear, peccary, late-glacial sediments. giant beaver, and other mammal species occur The Dutchess Quarry Caves, “Hallock,” in Pleistocene deposits across the country and and Zappavigna were all near the wetland presumably the living animals were available that became the Black Dirt Area, and therefore to the Indians prior to the extinctions that their inhabitants probably took advantage of occurred some time preceding 10,000 B.P. its bounty in plant and animal foods (Funk (Martin and Klein 1989). Elk, white-tailed 1992; Funk and Steadman 1994). The caves deer, black bear, and other mammals charac- contained varied faunal remains characteristic teristic of the Holocene occurred farther south of the Holocene, but there were also bones of but were probably gradually infiltrating the caribou, giant beaver, and flat-headed peccary Hudson drainage as far back as 11,000 years that pertained to the Pleistocene. Unfortunate- ago. Numerous varieties of fish swam in the ly, direct dating of the bones of those three streams and lakes (Smith 1985), and both species by tandem accelerator mass spectrom- migratory and non-migratory birds were eter showed that they are much older than the abundant. Some seeds and other edible plant known dates for Paleoindians, and it is now parts must have been available, but the oaks assumed they were introduced into the caves predominated among nut trees and other by animal predators and scavengers (Funk mast-producing trees were few and far and Steadman 1994; Steadman, Stafford, and between in the dominant Spruce–Fir forest of Funk 1997). Therefore, the caves failed to pro- the period. Undoubtedly, however, berries vide reliable evidence of the subsistence were varied and abundant in “edge” situa- habits of their Paleoindian residents. tions. None of the other sites in the study area Skeletal remains of Pleistocene mammals produced food remains. The Zappavigna site have been found in sediments ranging from is located on the crest of a ridge adjoining a southeastern and eastern New York to the large, amphitheater-like depression on the state’s western border. Relatively few sites are headwaters of a creek. It is speculated that on record for the far northern parts of New Paleoindians camped there one winter to prey York including the Adirondack Mountains. upon caribou and other game, taking shelter Most of the specimens are mastodont bones, at against the foot of the ridge from cold wind, least 35 of which have been reported for snow, and rain blowing from the north and Orange County alone (Ritchie 1965: Figure 3). west (Funk, et al. 2003). Three mastodonts, a horse, and a deer were Sites near wetlands, large creeks, rivers, or found in Greene County, two mastodonts, a lakes, such as Port Mobil, Twin Fields, deer, and a bison in Albany County, and a Corditaipe, and Davis afforded their occu- mastodont and a caribou in Schenectady pants access to fish, shellfish, aquatic mam-

94 mals and plants, and in some cases migratory Hampshire (Curran 1984) and at the Bull ducks and geese. The Potts site is located near Brook site, in Massachusetts (Spiess, Curran, a small stream where some fish may have and Grimes 1985). Calcined mammal bones been available, but Gramly and Lothrop were also found at the Sugarloaf site in west- (1984) speculate that a caribou killing ground ern Massachusetts (Gramly 1998). The was close by. Kings Road and Swale, situated Michaud site in Maine (Spiess and Wilson on low, very slight rises in the middle of the 1987) produced some fragmentary, calcined Athens Flat, are a half mile from the nearest mammal bones that could not be identified as source of Normanskill chert, “in the middle of to species. On comparative, topographic and nowhere,” so I suggest that the sites were cho- ecological grounds, Gramly (1982, 1988) posit- sen for proximity to important food resources. ed that caribou were intercepted and killed by Fish and mussels were probably rare in the people camped at the locations of the Vail and small brook close by the sites, though larger Adkins sites on the Magalloway River in creeks and the Hudson River were not far Maine. Foods consumed by the groups who away. But the north–south oriented escarp- visited West Athens Hill and camped either on ment, ridges, flats, and streams in the general the summit or on the nearby lowlands remain area would have restricted the lateral unknown. Perhaps they hunted and feasted (east–west) movement of caribou and other on the same migrating animals envisioned for game animals which presumably therefore the Kings Road and Swale sites. tended to follow a north-south route passing The limited evidence suggests that near the sites. The choices available to the ani- Paleoindians subsisted on a moderately mals were limited, funneling them either diverse range of animals and plants provided along the Flat or along the Hans Vosen Kill by the terminal Pleistocene landscape. valley. The occupants of Kings Road and Caribou and fish may have provided their Swale may have taken advantage of that con- “staple” foods in some parts of the Northeast. figuration, lying in wait for game to appear in Migratory fowl may have contributed large their vicinity. But there is no proof that people amounts of protein to bands living along were actually taking advantage of those par- major flyways (Dincauze 2001). We might ticular resources. speculate that Paleoindians living along the In western New York, caribou and seacoast were more oriented toward marine mastodont bones occurred in the basal, resources, at a time when sea level was much Pleistocene zone of the Hiscock site, in appar- lower than at present. Therefore, if any of their ent contemporaneity with fluted points, chert sites survived wave erosion as the sea once flakes, and scrapers of Paleoindian origin again encroached on land, important informa- (Laub, et al. 1988). It is uncertain whether the tion on their adaptations lies at considerable Indians scavenged or killed the mastodont depth below modern sea level. and caribou. The Templeton site in western Connecticut produced evidence that the Chronology and place of sites in regional Paleoindians living there were collecting developmental stages: The only sites in the acorns, reflecting the presence, though proba- study area that have produced organic materi- bly not abundance, of oak trees in the terminal als for potentially radiocarbon dating Pleistocene environment (Moeller 1980). Fish Paleoindian assemblages are West Athens Hill bones and hackberry seeds were recovered and the Dutchess Quarry Caves. But as stated from the Paleoindian zone at the Shawnee- above, the date for West Athens Hill is on Minisink site on the Delaware River in eastern charcoal of modern origin, the cave dates on Pennsylvania (McNett and McMillan 1974). charcoal pertain to Archaic components, and Caribou bones were identified at the Whipple the bone dates to Pleistocene animals appar- site on the Connecticut River in New ently dragged into the caves by scavengers or

95 predators before Paleoindians arrived. In temporal overlap with “Parkhill” stage mani- western New York, excavations at the Arc site festations. by Tankersley, et al. (1997) yielded a date of Cumberland-Barnes: Corditaipe, Potts, ca. 10,400 B.P. for the Paleoindian occupation Port Mobil, Zappavigna, “Hallock” (?), Davis, zone (as noted above dates are presented here Dutchess Quarry Caves 1 and 8, Twin Fields, in radiocarbon years, ignoring standard devi- Swale, Kings Road. ations). The stratum at the Hiscock site con- Crowfield: Some Crowfield points taining mastodont and caribou bones and flut- occurred with other later Paleoindian styles at ed points was dated at about 11,000 B.P. The the Reagen site, Vermont (Ritchie 1953, 1957). Templeton site provided one date of 10,190 B.P. The Whipple site is dated between 11,400 Artifact assemblages: Northeastern and 9540 B.P., the Debert site, Nova Scotia at Paleoindian assemblages consist almost an average date of 10,600 B.P., the Vail site in entirely of artifacts made of lithic materials Maine at 11,120 to 10,040 B.P., the Bull Brook that, unlike objects of bone, hides, flesh, and site in Massachusetts at ca. 9380 to 6940 B.P., wood have not perished from the agents of Shawnee-Minisink in Pennsylvania at 10,750 decay. Stone artifacts were especially impor- to 10,590 B.P., the Michaud site in Maine at tant to the survival of those ancient people. 10,200 to 9010 B.P., and the Neponset site in They had to locate and exploit sources of the Massachusetts at 10,120 B.P. Although many raw materials for artifacts, principally chert. more dates are needed, it has been suggested The particular form of artifacts had to con- that there are two populations of dates: an form to their function, otherwise they would early one of about 11,000 to 10,500 B.P., associ- be useless. To some degree, function may ated with so-called “Gainey-Debert” style determine style, but stylistic differences often points (most similar to western Clovis points) consist of an idiosyncratic constellation of and a later series of around 10,500 to 10,120 attributes permitted by the “slack” in purely B.P. that applies to assemblages containing functional attributes. And some writers have “Cumberland-Barnes” style points (Curran theorized that style has an important function 1996). These age estimates leave aside the puz- apart from efficiency in hunting or other tech- zling “late” dates of ca. 9,540 to 9010 B.P. for no-economic activities, namely, it may be sym- Whipple, Bull Brook, and Michaud. bolic of group identity (Weissner 1983). The early Paleoindian sequence of Gainey- Almost without exception, the artifacts in Debert points, Cumberland-Barnes points, the assemblages relate to the Paleoindians’ and Crowfield points has been proposed by livelihood in other words, they represent the various writers (Curran 1996; Ellis and Deller technology that enabled them to extract food 1990, 1997; Gramly and Funk 1990), and to and other goods from the environment. Given date it has not faced any serious challenges. the absence of subsistence remains on most There is no doubt that the different styles New York sites, we are forced to rely on lithic overlapped in time. Any two of them not only items to interpret Paleoindian economic intergraded in size and form, but were proba- behavior, so that a great deal of time and effort bly associated at times of change from Gainey- is devoted to analyzing the assemblages. Non- Debert to Cumberland-Barnes, and from economic aspects of their culture are less evi- Cumberland-Barnes to Crowfield in particular dently observed in the weapons, tools and assemblages. byproducts of manufacture and must be The assemblages within the study area are inferred largely on a comparative basis using therefore assigned to these stages as follows: data from ethnographically known cultures. Gainey-Debert: West Athens Hill. The assemblages are broken down into However, at least one point is of the two major categories; weapons and tools. Cumberland-Barnes type, indicating some Weapons (projectile points) were used chiefly

96 for hunting, though at times they must have may also reflect the distance of sites from aided in defending bands against hostiles bedrock sources of the lithic raw materials (i.e., from other bands. Other tools were employed at greater distances the groups may tend to in the diverse tasks of domestic life. These conserve items made of their preferred materi- functional types are postulated on the basis of als, hence those items may be reworked and data from ethnographic analogy and from rehabilitated, reducing their dimensions), or experimental studies testing hypotheses gen- the particular outcrops may offer blocks and erated to account for use-wear patterns nodules of either relatively small or relatively (Keeley 1980; Semenov 1964). large size. Differences in use-wear patterns Artifact assemblages reflect the tasks and may reflect differences in the types of tasks per- activities performed on a site. Therefore eluci- formed on given sites; for instance, a predomi- dating the differences and similarities nance of edge-crushing on scrapers indicating between sites may contribute to understand- use on bone and antler, a predominance of ing the position of sites in regional settlement edge-rounding indicating use on hides. systems. Such comparisons can contribute to The ratio of artifacts to chert-knapping the definition of site types, each type illustrat- debitage is one clue to on-site activities, ing one aspect of adaptation to the late-glacial including the relative importance of chert landscape. For example, a heavy predomi- quarrying and knapping. As might be expect- nance of projectile points and biface knives in ed, only rarely do sites completely lack deb- contrast to other artifact types means the site itage. Aboriginal peoples had to rely on work- was primarily a hunting and butchering sta- able lithic material for tasks like hunting and tion; conversely, a predominance of killing game, cutting up meat, processing netsinkers, bone harpoons, fishhooks, and hides, and so on. Indices of site productivity related items on an Archaic site (those traits (e.g., the quantities of artifacts and debitage are not known for eastern Paleoindian sites) per area excavated) probably reflect several demonstrates the relative importance of fish- things including the nature and relative ing. Theoretically, the more diverse an assem- importance of activities, the length of occupa- blage (the greater the number of functional tion, intensity of occupation, and size of occu- types) the more varied the activities and the pying groups. more the site conforms to the definition of a Finally, the types of lithic material used for central base-camp, rather than a limited-pur- an assemblage obviously provide clues to pose station, defined from a much more limit- where the material outcrops, and materials ed array of functional types. Artifacts exhibit- exotic to a given region represent extra-local ing the most complex morphology, such as or extra-regional travel and trade relations of projectile points, are more sensitive indicators the people who occupied the site. of regional and temporal stylistic variation Sample Size than unifaces. In this regard, projectile points are the Paleoindian equivalent of Woodland The recovered artifacts vary in number ceramics. from 110 to 1,525 per site. These figures are not Another aspect of interassemblage variabil- particularly informative because, short of total ity consists of differences in the size and weight excavation, all of the assemblages are actually of bifaces and unifaces and the types of use- samples, incomplete portions of all the artifacts wear observed on them. Size and weight dif- deposited prehistorically on the sites. More ferences may provide clues to the types of useful would be relating the size of assem- activities performed (e.g., large scrapers used blages and the size of areas investigated, to for processing the hides of large animals such the total areas of particular sites. Sampling as elk, small scrapers used for processing the designs are important, since the problem is to hides of smaller animals such as foxes). They evaluate the relationship of samples obtained

97 to the total area and internal structure on a distance, the greater the tendency to conserve, site. Data are fortunately available concerning resharpen, or rehabilitate artifacts). the relation of investigated area (excavated Ten of the numerous fluted points found area, excavated area plus surface collected at the Lamb site, Genesee County, New York, area, or simply surface collected area) to total ranged from 92 to 140 mm long and 30 to 39 site area for the Zappavigna, Kings Road, mm wide; summary statistics were also pub- Swale, Twin Fields, Potts, Corditaipe, and lished for a selected group of bifacial pre- West Athens Hill sites. forms (Gramly 1999). Stage 1 and 2 bifaces represent early phases of the reduction trajectory from core to fin- Frequencies of Artifact Types ished projectile point or knife. Therefore, Although items such as pieces esquillees, objects in these categories from the various drills, spokeshaves, and denticulates are sites should be consistently larger and heavier found on some sites but not on others, giving than Stage 3 or finished (Stage 4) bifaces. This an impression of greater diversity in some is borne out by comparing the data in Table 36 assemblages than in others, they are few in with the data in Table 37. number where they do occur. They are far out- Although Table 38 shows some variation weighed by the objects in the basic tool kit in end scraper size and weight, there is a high (i.e., fluted points, other bifaces, end scrapers, degree of uniformity among most of the side scrapers and utilized flakes). Those assemblages. The major disparity is in the objects were universal on the sites in Table 35, rather smaller and lighter scrapers from the and bifaces in process occurred on all but one Zappavigna and Corditaipe sites, and the site. The ratios of the quantities of some types much larger and heavier scrapers from the to others do, however, show contrasts from Potts site. This Potts data are surprising, since site to site. For example, the proportion of its distance to Onondaga chert outcrops is the bifaces to unifaces varies, as do finished greatest of all the sites. bifaces to those in process, biface knives to Although the end scrapers from Areas A other bifaces, “made” artifacts to utilized and B at West Athens Hill appear significantly flakes, and the amounts of exotic vs. local lith- lighter than those from Area C despite a close ic materials. Rough stone tools, primarily cob- similarity in size the mean weight for both ble hammerstones, occurred on most sites. groups is about 7 g, if one massive end scraper Some of the differences probably relate to dif- from Area C is removed from the calculation. ferences in the relative importance of certain Comparisons with New England sites, tasks and activities on the sites. Another con- despite the unevenness of published data, sideration is the different types of wear pat- show much similarity in the sizes of end terns observed on artifacts in the assemblages. scrapers to the samples from central and eastern New York. For example, the six end scrap- Size and Weight of Certain ArtifactTtraits ers of chert in the assemblage from the Adkins site, Maine, range from 27 to 32 mm in length; Tendencies in the size and weight of arti- the 12 scrapers of crystal quartz tend to be fact categories from the sites under study may smaller, averaging 22 mm long (Gramly 1988). reflect several determining factors, such as: End scrapers found at the Bull Brook II site in the varying requirements (“heavy” vs. Massachusetts range in length from 13 to 50 “light”) of different tasks; the multi-function mm, the mode tending from 25 to 35 mm utility of certain types; the stage reached in the (Grimes, et al. 1984). reduction process; the extent to which some A high degree of correspondence is also items were reworked or resharpened; the size seen in the size and weight of items classed as of raw material blocks and cores; and distance side scrapers (Table 39). This is surprising in from sources of lithic material (the greater the view of the considerable formal variability in

98 2 .2 Areas A, B Area C 5 .3 2 .2 34 2.2 2 .2 3933 2.6 2.2 19 1.6 185 12.1 75 6.1 168 11.0 151 12.3 6 4.2 212 1.4 1.4 3 4 2.1 .3 4 2.8 11 .1 .7 2 .2 15 10.4 3 1.8 10 6.9 27 1.8 9 .7 1 0.6 3 2.1 35 2.3 1 0.6 1 .7 5 .7 2 .3 3 .4 6 3.6 22 3.1 2 1.2 400 56.3 1 .9 52 4.4 1.8 33 4.6 7 2.8 3 1.2 16 6.5 5 1.3 12 4.9 3 .8 1 .3 4 1.6 29 7.4 4 1.641 1 10.5 .9 5 2.0 6 5.3 2 .3 2 1.2 3 2.73 3 2.7 .8 2 6 .5 2.4 13 2 .4 2.7 1.8 67 6 17.23 2.83 36 2.7 2 14.6 2.7 1.2 371 18 32.5 .97 12.5 46 1 6.4 13 6.5 .3 6 14 .9 1.6 8 9.4 3 3.2 22 15.3 .2 192 12.6 19 1.6 11 10.0 102 26.2 28 11.3 16 14.0 114 16.0 26 15.6 12 8.3 680 44.6 903 73.5 17 15.5 6547 16.7 42.7 23 55 9.3 14.1 89 212 36.0 10.9 1.8 10 42 12 36.8 2.6 1.7 66 5 17 9.3 2.0 10.2 93 55.7 47 37.6 92 6.0 37 3.0 No. Pct. No. Pct. No. Pct. No. Pct. No. Pct. No. Pct. No. Pct. No. Pct. No. Pct. 110 100.0 390 100.0 247 100.0 114 100.0 711 100.0 167 100.0 144 100.0 1525 100.0 1228 100.0 Zappavigna Kings Road Swale Twin Fields Potts Corditaipe Port Mobil West Athens Hill West Athens Hill Artifact Types Fluted points (finished) Fluted points in process Possible Fluted points Biface knives Bifaces in Process Misc. bifaces Drills/awls End scrapers End-side scrapers Side scrapers Spokeshaves Flake knives Retouched flakes Gravers Denticulates Pieces esquillees Tool fragments Other unifaces Utilized flakes Rough stone tools Totals for all Artifacts Table 35. Comparison ofand artifact 8, types the in Davis assemblagespoints, site, from out and Paleoindian of the sites context,points “Hallock” in and are site central the excluded. are and handful not eastern of included widely New because York. scattered the The artifacts collections Dutchess from from Quarry the the Caves Davis caves 1 site and was “Hallock” mixed are with confined those to of a later few cultures. fluted Archaic projectile

99 Table 36. Comparisons of fluted point size and weight between sites, in terms of means and standard deviations. (N/A means not applicable).

Site Length (mm) Width (mm) Thickness (mm) Weight (g) Port Mobil 38±8.6 21±3.4 5.6±1.3 No data (N=15) (N=15) (N=15) Zappavigna 43±6.8 21.7±3.3 5.1±1.2 No data available (N=3) (N=7) (N=9) for 3 whole points in private collection West Athens Hill 47.0 ±10.4 27.1±6.5 7.0±1.4 8.6±5.6 Areas A, B (N=6) (N=12) (N=13) (N=4) West Athens Hill, N/A; none 35.5±6.5 7.1±1.4 N/A; none Area C complete (N=12) (N=12) Complete Kings Road N/A; only one 28.3±1.5 6±1 N/A; only one whole specimen (N=3) (N=3) whole specimen Swale N/A; none 36.3±10.8 8.6±2.4 N/A; none complete (N=17) (N=16) complete Corditaipe N/A; only one 34.8±10.5 7.8±1.6 N/A; only one whole specimen (N=5) (N=5) whole specimen Michaud 57.3 24.2 5.2 No data (N=3) (N=4) (N=6) Vail 62.4±17.9 28.7±3.3 7.4±1.5 13±5.6 (N=29) (N=29) (N=29) (N=29) these tools, which are less easily defined than and so on. end scrapers. The Corditaipe side scrapers, Perusal of Table 40 shows that the artifact however, are rather smaller than those from to debitage ratios vary considerably. They the other sites. should be most reliable for excavated sites, such as Zappavigna, West Athens Hill, Vail, Lamb, and Templeton. Comparisons of Artifacts to Debitage The ratio of counts of debitage items to Measures of Site Productivity counts of artifacts should theoretically indicate the status of chert-knapping relative to As noted above, the counts of artifacts col- other activities at a site. High counts of deb- lected from particular sites mean little unless itage could signify a quarry-workshop, or at certain conditions are specified. Clearly, many least proximity to a chert source. Low quanti- factors can account for collection size, includ- ties could signify considerable distance from ing the total number of artifacts originally such a source, or perhaps little need for biface deposited, whether or not a site has been manufacture on briefly occupied limited-pur- plowed, whether or not parts of it have been pose sites. The quantities of recovered deb- severely disturbed or destroyed, how much of itage would be affected by disturbance of a the site lies below the reach of the plow, site, for instance by plowing or leveling, whether or not it is stratified, whether or not it whether or not screens are used, whether or has been excavated rather than surface collect- not the site is excavated or surface collected, ed, how much of it has been excavated, the

100 Table 37. Comparisons of the size and weight of Stage 1 and 2 bifaces, in terms of means and standard deviations (N/A means not applicable).

Stage 1 Bifaces Site Length (mm) Width (g) Thickness (mm) Weight (g) Zappavigna 57.3+/-18.7 30.3+/-12 12.3±4.5 36.7±43 (N = 4) (N = 9) (N = 13) (N = 4) West Athens Hill, 74.7±17.02 47.7±8.34 17.8±4.71 83.5±55.15 Areas A and B (N = 30) (N = 45) (N = 53) (N = 29) West Athens Hill, 74.1±16.5 50.8±12 18.2±5.3 79.5±49.7 Area C (N = 20) (N = 55) (N = 77) (N = 12) Kings Road 77.7 49.1 19.1 No data (N = 9) (N = 36) (N = 36) Swale 74.5±9.7 46.7±11.2 13.1±3.9 61.2±43.4 (N = 6) (N = 19) (N = 19) (N = 6) Corditaipe 56.4±14.5 42.6±9.9 16.3±5.7 51.2+/-31. (mixed Stages 1, 2) (N = 10) (N = 18) 9(N = 23) (N = 10) Potts (Stages 2, 3, 4; N/A N/A N/A N/A only 12 total)

Stage 2 Bifaces Site Length Width (g) Thickness (mm) Weight (g) Zappavigna N/A, no stage 2 bifaces identified West Athens Hill, N/A 44.4±8.69 13.1±2.73 N/A Areas A and B (N = 45) (N = 52) West Athens Hill, N/A (none 50.2+±12.1 12.1±2.3 N/A (none Area C complete) (N = 27) (N = 31) complete) Kings Road 58( 43.5 13.2 No data N = 3) (N = 14) (N = 12)

extent of collecting activity, whether some sole occurrence of fluted points without asso- items are preferentially selected over others, ciated tools. the intensity of occupation, the length of occu- In Table 41, Vail and West Athens Hill are pation, and the nature of activities on the site. the most productive of all the sites, ranging The most reliable productivity measure is the from 4.7 to 9.1 artifacts per square meter of number of artifacts per square meter, based on excavation (not counting debitage). Next data from excavated sites. It has less meaning highest is Templeton, at 1.8 items per square for surface collected sites, no matter how meter. All the others produced less than 1 intense or prolonged the collecting activity. object per square meter. Table 41 presents the relevant data for such a The potential maximum number of arti- measure. The Dutchess Quarry Caves are not facts is calculated at an astronomical 50,000 included due to the mixed deposits and the plus at the Vail site, in contrast to 5,850 at West

101 Table 38. Comparisons of end scraper size and weight between sites, in terms of means and standard deviations.

Site Length (mm) Width (mm) Thickness (mm) Weight (g) Port Mobil No data No data No data No data Zappavigna 26.5±5.9 23.1±4.4 8.1±2.1 5.5±2.7 (N = 36) (N = 36) (N = 42) (N = 35) Twin Fields 29.7 24.3 6.7±3.1 8.1 (N = 42) (N = 42) (N = 42) (N = 33) West Athens Hill, 32.2±9.47 27.0±10 8.8±3.1 7.1±3.1 Areas A, B (N = 91) (N = 92) (N = 92) (N = 37) West Athens Hill, 35.1±12.3 23.7±6.0 8.6±2.6 8.8±11.5 Area C (N = 29) (N = 30) (N = 32) (N = 28) Kings Road 33.0±8.1 24.3±3.8 8.4±2.3 7.3±4.0 (N = 53) (N = 54) (N = 55) (N = 52) Swale 33.5±11.2 25.6±5.4 8.8±3.4 9.6±12.2 (N = 66) (N = 83) (N = 86) (N = 66) Corditaipe 27.6±5.0 21.7±3.2 7.8±2.6 4.9±1.8 (N = 36) (N = 36) (N = 36) (N = 89) Potts 41.5±8.3 31.0±4.6 9.2±2.4 11.8±4.6 (N = 54) (N = 54) (N = 54) (N = 31) Vail 33.3±6.6 25.1±4.7 8.0±2.0 No data (N = 154) (N = 154) (N = 154) Michaud 30.3±8.1 22.5±4.33 5.8±7.0 No data (N = 10) (N = 10) (N = 11)

Athens Hill Areas A and B and 5,060 at Potts. Frequencies of Non-local Cherts in Chipped Next largest is Zappavigna at 1,278. By con- Stone Assemblages trast, the lowest totals are for Twin Fields and Non-local stones used for the manufacture Templeton, both 100 percent excavated, and of artifacts are, in general, far more common Lamb, nearly 100 percent excavated. There is in Paleoindian components than in most later an apparent correlation with size, the smallest periods. Such “exotics” do, however, occur to artifact totals coming from the smallest sites a minor degree in Early Archaic complexes of (Twin Fields, Templeton, and Lamb) and the New York State, as well as in Susquehanna largest totals from the largest sites (Vail, Potts, tradition assemblages (specifically, small Zappavigna, and West Athens Hill). quantities of Pennsylvania jasper and South The quantities of debitage items per Mountain, Pennsylvania rhyolite) and Middle square meter of excavation were highest at the Woodland complexes (including Vanport Templeton site, next highest at West Athens chert from Ohio). The higher incidence in Hill Area B, next at the Vail site, and lowest at New York Paleoindian contexts is usually the Potts and Lamb sites. It seems no coinci- assumed to reflect either wide trade relations dence that multiple loci of activity were evi- with other Paleoindian bands living closer to denced at West Athens Hill, Potts, and Vail. such resources, or actual visits by the local No such internal patterning was convincingly bands to the bedrock outcrops. Both scenarios demonstrated at Zappavigna.

102 Table 39. Comparisons of side scraper size and weight between sites, in terms of means and standard deviations (N/A means not applicable).

Site Length (mm) Width (mm) Thickness (mm) Weight (g) Zappavigna ———-N/A (only 3 items in this category)———- West Athens Hill, 58.9±17.62 39.41±9.43 15.93±7.53 47.5±43.1 Areas A and B (N = 168) (N = 172) (N = 174) (N = 125) West Athens Hill, 62.6±15.9 43.5±14.9 14.4±5.1 39.8±31.5 Area C (N = 17) (N = 17) (N = 17) (N = 14) Kings Road 62.6 (note: 42.1 15.8 No data identical to the (N = 65) (N = 65) (N = 65) above mean for Area C at West Athens Hill) Swale 59.9±24.2 39.1±14.4 12.3±7.2 44.3±56.03 (N = 37) (N = 37) (N = 37) (N = 37) Corditaipe 43.6±15.5 26.1±5.8 6.6±2.1 15.2±13.03 (N = 14) (N = 14) (N = 14) (N = 14) Potts 65.1±16.9 42.2±10.8 9.9±2.5 28.6±16.9 (N = 29) (N = 29) (N = 29) (N = 29) Michaud 66.9±15.59 37.7±15.7 8.6±2.78 No data (N = 9) (N = 9) (N = 10) are probably valid. A major assumption is that from those quarries with miminal difficulty. the wide geographical reach of lithic resource Those living farther away would have found procurement resulted from the large territori- it more convenient (less demanding of time al ranges of the thinly distributed and popu- and energy) to trade with the bands living in lated groups. A corollary assumption is that proximity to those quarries for the needed the search for raw materials was embedded supplies of toolstone. within the subsistence rounds of Paleoindian It is tempting to speculate whether other bands. commodities were exchanged between bands Groups moving around within their own, during the late Pleistocene epoch; for exam- perhaps loosely bounded, territories must ple, food items such as dried fish, shellfish, or have occasionally met and interacted peace- mammal meat (possibly even something like fully with other groups bearing similar cultur- pemmican), edible plant foods such as berries, al traditions and having the same need for plant parts useful for tools and weapons, bird high-quality cherts for their weapons and feathers, animal hide clothing, ornamental implements. Almost certainly, exchange for items of bone or wood, and so on. Such items desired but locally unavailable raw materials could also have been used to barter for high- was a major activity shared by and helping to quality exotic cherts. link up different bands. Those living within Data on the frequency of non-local cherts relatively short distances of important sources occurring in chipped stone assemblages from such as the jasper quarries in present-day east- Paleoindian sites in central and eastern new ern Pennsylvania could have traveled to and York are presented in Table 42. The Port Mobil

103 Table 40. Artifact to debitage ratios and percentages for Paleoindian assemblages (N/A means not applicable).

Site Number Number Debitage/ Percentage Remarks of of cores artifact artifacts/ artifacts and flakes ratio debitage Port Mobil No data 144 N/A N/A Much disturbed; also, no debitage reported Zappavigna 110 1,550 14/1 .07 Sample from grid excavation Twin Fields 114 No data N/A N/A Multicomponent; despite complete excavation of plow zone and underlying zone, Paleoindian debitage could not be distinguished from later material West Athens Hill 970 10,408 11/1 .09 Sample from grid excavation Area B West Athens Hill 1,228 Unknown N/A N/A Sample from grid excavation. Area C Debitage incompletely studied Kings Road 390 6,570 17/1 .06 Debitage surface collected Swale 241 914 3.8/1 .26 Debitage surface collected Templeton 75 7,360 98/1 .01 100% excavation Corditaipe 167 1,400 8.4/1 .12 Debitage surface collected Potts 711 2,930 4.1/1 .24 Excavation within grid plus surface collection over whole site Lamb 82 494 6/1 .17 Nearly 100% excavation Vail 2,732 4,378 1.6/1 .62 Sample mostly from grid excavation

site is not included in Table 42 because local It is worth noting that the identification of jaspers were difficult to distinguish from non- “exotic cherts” is neither simple nor easy. Some local jaspers (Kraft 1977). The definition of are readily distinguished from local cherts, oth- “non-local” refers to a source or sources out- ers are not. This is especially true of items made side the immediate drainage system where a of materials from sources located at large dis- particular site is located. On most of the sites tances from areas familiar to archaeologists. the non-local material is chiefly Pennsylvania Even local stones may need petrographic or jasper (Hardyston chert). Minority exotics trace element examination for confident identi- include Upper Mercer, Ohio chert and the fication. Some assemblages described here pro- Western New York Onondaga chert type duced artifacts manufactured from cherts lack- Divers Lake. ing clear-cut attributes of color and texture that would enable their classification by simple

104 10 3.8 2.8 N/A N/A N/A area) total site puted using square meter 914 494 6570 7360 171 (com- 1400 111 N/A N/A 324 2,560 No data N/A 700400 3,290400 10,408 50.2 300 1800 1,278 1550 1,3506,730 N/A N/A No5,600 data 50,960 N/A 4,378 14.5 5,500 5,060 2,930 Unknown N/A No data N/A 7.9 9.1 .71 4.7 .92 .47 N/A area) site area) using total excavation) per sq. m. (sq. m.) of artifacts flakes items per using total site excavation) 93 1.8 (computed 43 (100% 75 N/A 500 .23 (computed154 207 230 (100% 114 No data N/A 773 174 area) collected) collected) N/A collected) collected) (disturbed 75 10 N/A (surface82 N/A 114 110 711 144 390241 N/A (surface N/A (surface N/A N/A 167 N/A (surface facts 2,732 300+ of arti- area (sq. m.) artifacts of site total number cores and debitage Number Excavated Number of Total area Estimated Number of Number of Site Port Mobil Twin Fields Zappavigna West Athens Hill Area BWest Athens Hill Area CKings 970 Road 1,228Swale Templeton 156 Davis Corditaipe Potts Lamb Vail Table 41. Measurements of Paleoindianber site of artifacts productivity: per comparisons square(N/A of meter, means lithic to not assemblage number applicable). of size debitage in items relation per to square size meter, of to total excavated areas area, of to sites, num- and estimated total numbers of artifacts

105 Table 42. Frequencies of artifacts made of non-local cherts in chipped stone assemblages from Paleoindian sites in central and eastern New York.

Site Total Sample Number of Percentage of of Chipped Artifacts of Non- Items made of Stone Artifacts Local Cherts Non-Local Cherts in the Sample Zappavigna 103 5 4.9 Twin Fields 121 16 13.2 West Athens Hill, Areas A and B 1,308 22 1.7 West Athens Hill, Area C 1,153 7 0.6 Kings Road 384 35 9.1 Swale 247 122 49.4 Davis 10 3 30 Corditaipe 161 34 21.1 Potts 708 8 1.1 visual inspection, even after years of experi- from the Delaware Valley, but Swale is almost ence. These are often “gray” cherts resembling certainly part of the same encampment as the materials from several different bedrock adjoining Kings Road site, which features a sources. This is the case, for example, with much lower incidence of red cherts (including many items from the Corditaipe site. At least one piece of Munsungan chert). The Port two dozen resisted confident assignment to Mobil site is not included because local jaspers Western Onondaga chert and may have been were difficult to distinguish from non-local cherts from local Eastern Onondaga outcrops, jaspers (Kraft 1977). and therefore are not included in the “exotic” As shown in Table 44, the highest percent- categories in the tables. ages of non-local cherts occur in two artifact As shown in Table 43, there are striking categories, end and side scrapers, but in all differences in the percentages of exotics in the but two cases (Twin Fields, Davis) the end assemblages. The largest frequencies are scrapers predominate. Scrapers were often shown for the Twin Fields, Swale, Davis, and reworked or resharpened, resulting in modifi- Corditaipe sites. The smallest are indicated for cation from their original sizes and shapes. the Potts and West Athens Hill sites. Potts is a These data indicate that scrapers were surprise because due to its central location one retained by their owners for longer periods might expect significant amounts of Vanport than other tools; in other words, they were chert, Upper Mercer chert, and other lithics “curated” to a greater extent than other arti- from the Ohio region, some jasper from the fact categories. Either the exotic materials Hardyston quarries, and larger quantities of were considered to be superior to other mate- Normanskill and other cherts from the rials for such implements, they were prized Hudson valley. Most impressive is the very because the materials were of unusually high high amount of Pennsylvania jasper in the aesthetic appeal, or the tools were so broadly Swale collection, plus other red cherts perhaps useful that they were kept close to and trav- obtained from the Munsungan quarries in elled with their owners, perhaps in animal Maine. One is tempted to speculate that the skin bags. One thing is sure: the scrapers were Swale material represents a pioneering group far easier to make than bifaces, in particular

106 Table 43. Types of exotic lithic material in the chipped stone assemblages in central and eastern New York.

Site Pennsyl- Maroon Vanport Western Norm- Other Total vania jasper chert New anskill jasper, (including (Flint York chert all Munsun- Ridge, Onon- varieties gan jasper Ohio daga from Maine) “chalcedony”) chert Zappavigna 1? 45 Twin Fields 15 1 16 West Athens Hill Areas A, B 99422 West Athens Hill, Area C 167 Kings Road 31 21 135 Swale 109 11 2 122 Davis 213 Corditaipe 5 18 11 34 Potts 11 417 fluted points. On the other hand, points were Onondaga chert: 4.8±1.7 to 4.9±1.8 g, respec- an essential part of hunting weaponry, often tively. broken or lost on the hunt and replaced, Examination of 13 end scrapers of exotic whereas the scrapers were less susceptible to cherts from West Athens Hill does, however, breakage or loss and had a longer service life. show some support for the above hypothesis. In support of the curation/reworking/ This combined group from Areas A, B, and C rehabilitation hypothesis, the mean weight of has a mean weight of 6.5±2.4 g. This figure 30 end scrapers from the Swale site, made of contrasts with the average of 7.1±3.1 g for the red and brown jasper, is 6.7 g (range 1.7 to 14.5 total sample of end scrapers of all materials g), considerably less than the mean weight of from Areas A and B, and 8.9±11.7 g for the 12.6 g (range 4 to 87.5 g) for 32 end scrapers of sample of local materials from Area C. local material. Also, the mean weight of 23 Swale site side scrapers of exotic material is 25.5 Patterns of Use-Wear Observed on Lithic g (range 1.5 to 243 g), much less than the mean Assemblages from Paleoindian Sites of 90.7 g (range 7.3 to 216.4 g) for 15 side It should be obvious that the principal scrapers of local material. forms of use-wear observed on stone tools On the other hand, the figures for the from West Athens Hill and several other sites Kings Road site show a nearly identical aver- used in the analysis are edge-crushing, round- age weight for all collected end scrapers, ing/gloss, and edge-nibbling (see Table 45). including those of local Normanskill chert, as Other variations include striations associated compared with the subsample (n = 24) made of with rounding/gloss, or differential occur- exotic stones, 7.3±4.0 vs. 7.5±4.7 g, respective- rence of the various types on different parts of ly. End scrapers of exotic material in the the tools; for example, rounding and blunting Corditaipe site assemblage have the same on spurs or front corners of end scrapers, asso- average weight as the scrapers of local ciated with crushing on the main bit or work-

107 Table 44. Frequencies of exotic lithic material by artifact types in the chipped stone assemblages in central and eastern New York.

Bifaces End Side Other Totals of Scrapers Scrapers Exotic Items Sites No. Pct. No. Pct. No. Pct. No. Pct. No. Pct. Zappavigna 1 20.0 4 80.0 5 100.0 Twin Fields 1 6.3 2 12.5 8 50.0 5 31.3 16 100.0 West Athens Hill Areas A and B 2 10.0 12 60.0 5 25.0 1 5.0 20 100.0 West Athens Hill Area C 4 57.1 2 28.6 1 14.3 7 100.0 Kings Road 2 5.7 25 71.4 5 14.3 3 8.6 35 100.0 Swale 5 4.1 56 46.3 21 17.4 39 32.2 121 100.0 Davis 1 33.3 1 33.3 1 33.3 3 100.0 Corditaipe 2 5.9 24 70.6 6 17.6 2 5.9 34 100.0 Potts 5 62.5 1 12.5 2 25.0 8 100.0

ing edge. Sometimes the alterations are seen and a common mode of interaction with the on lateral edges of end scrapers, or on the end-of-Pleistocene environment. But in most unretouched edges of side scrapers. Occasion- cases, specific aspects of adaptation, such as ally, polish is evident on dorsal surfaces of end animals hunted and trapped or plants collect- scrapers, or on the faces of bifaces near lateral ed, remain obscure given the spotty and vari- edges and tips. On utilized flakes, the primary able evidence. form of alteration is nibbling on the longest, It is hard to suggest reasons for differences thinnest edges, with crushing or round- in the size and weight of fluted points and end ing/gloss less frequent. scrapers from one site to another. Fluted As might be expected, the chief problem is points from the Lamb site in western New interpreting the signs of use in terms of actual York and from the sites in Maine tend to be materials on which tools were applied, as well rather larger than those from the other sites in as the particular tasks for which they were New York State. Possibly the game hunted at made. Experimental approaches by various the Lamb site and the Maine sites was larger authors are the basis for our own sometimes than that hunted at the Hudson valley sites, tentative identifications (Keeley 1980; but in the absence of refuse animal bone this Semenov 1964; Wilmsen 1968). remains pure speculation. Similarly, it is interesting that end scrapers from most of the list- Interpretation of Similarities and Differences ed sites tend to be very similar in measured Between Assemblages size and weight (except for scrapers from the Paleoindian lithic artifacts and their Potts site that lie statistically outside the means of production exhibit general similarity means from those from the other sites). The across the Northeast and other parts of North similarities must represent a range of optimal America. There is, however, considerable vari- sizes for the tasks to which they were applied. ation on the major themes. Different assem- This implies they were used in similar tasks, blages seem to reflect a common technology such as working hides, bone, or antler. The

108 Continued. Other unifaces Retouched flakes: rounding/gloss 2, nibbling 1.Denticulates; rounding/gloss 2 No data Crushing on many flake knives and retouched flakes. Nibbling on utilized flakes. flakes, rounding/gloss on 2. on 2 items; rounding/gloss on 2; nibbling on 1;ing crush- and nibbling on 1 rounding/gloss on 8. (on retouched edges) Rounding/gloss: 2 No data Crushing on most rounding/gloss on 2, Predominantly crushing Predominantly crushing crushing (5) End scrapers (on bits) Side scrapers Crushing on several striations, dorsal polish, andnibbling rounding/ gloss without crushing and rounding/gloss lower edge grinding. No wearon present stage 1 bifaces. Rounding/gloss 38 bifaces show edge crushing and nibbling; 18 of 64 Stageshow 2 edge bifaces crushing; 15 ofknives 35 show bifacial rounding/gloss on edges; and most miscellaneous bifaces have crushing or nibbling. (nibbling, rounding/gloss) seen onother bifaces. rounding/gloss on 13. on 3 items Bifaces Fluted points show the usuallower basal edge and grinding.Variable rounding/gloss on 9 bifaces ofcrushing Stages and 1, Crushing; rounding/gloss 2, on 3; 75 them scrapers; also 24 have of rounding/gloss; Various combinations crushing of and plus rounding/gloss various combinations of (flake knives, retouched Crushing predominant on (21); 13 also objects some flakes, end-side scrapers); lower edge grinding. Other bifaces manifest frequent edge crushing and rounding/gloss. Hill, Areas Aand B lower edge grinding. 15 of 53 Stage 1 Site Zappavigna Fluted points show the usual basal andTwin FieldsWest Crushing Athens 1 No data. Fluted points show the usual basal and Crushing on most West AthensHill, Area C Fluted points show the usual basal and lower edgeKings grinding. Road Very little wear Crushing on 9, crushing and Fluted points show the usual rounding/gloss basal on and 1, Swale Crushing on 4, crushing and Predominantly crushing Crushing on 5 retouched Table 45. Wear patterns on chipped stone artifacts from Paleoindian sites in New York and New England.

109 Table 45. continued.

Site Bifaces End scrapers (on bits) Side scrapers Other unifaces (on retouched edges)

Corditaipe Fluted points show the usual basal and Crushing on most but some 7 show rounding/ gloss on Rounding/gloss on 2 flake lower edge grinding. 1 of 18 bifaces in show edge-nibbling, rounding/ edges; 2 show nibbling; knives; rounding/gloss and process showed rounding/gloss on gloss, and striations others show crushing nibbling on a denticulate 1 edge. tool Potts Fluted points show the usual basal Crushing on most Crushing on most No data and edge grinding. Vail Fluted points show the usual basal and No scraper study except 27 No data No data 110 edge grinding. showing phytoliths on working edges Lamb Fluted points (not fluted knives) show No data No data No data the usual basal and edge grinding. Templeton The fluted point and “miniatures” were No data Nibbling and polish on No data not ground on basal or lower lateral gravers and “spurs” edges. Michaud Fluted points show the usual basal and No data No data No data edge grinding. size was determined less by the original size Vail, and West Athens Hill sites. Possible of cores from bedrock sources than by the explanations relate to the length of occupa- nature of the jobs as well as the mechanical fit tion, the size of occupying groups, the number to human hands and strength. Scrapers were of separate occupations, and the nature and almost certainly tied onto or inserted into relative importance of on-site activities. It hafts of bone, antler, or wood, since they seems a daunting task to untangle these would have been difficult to hold in the hand causative factors in order to explain the simi- for prolonged use. larities and differences. The percentage of artifacts in relation to Wear patterns in bifaces and unifaces are debitage for excavated sites ranged from .01 at remarkably consistent from site to site. Apart the Templeton site to .62 at the Vail site, with from the preparation of fluted points for haft- West Athens Hill toward the lower range at ing by grinding the bases and lower lateral 0.09. Despite the considerable quarrying and edges, not really a product of use-wear, bifaces reduction of chert on the hill, the ratio of deb- show a preponderance of rounding/gloss on itage to artifacts of 11 to 1 is lower than at all lateral (working) edges, and less crushing the other sites except Lamb at 6 to 1, Potts at (step-flaking) or nibbling. End and side scrap- 4.1 to 1, and Vail at 1.6 to 1. One might specu- ers, on the other hand, display a predomi- late that the difference is explained by recov- nance of edge-crushing, in some cases com- ery methods: soils at West Athens Hill were bined with rounding/gloss and striations. troweled rather than screened. Perhaps many Rounding/gloss prevailed only at West small flakes were missed that might have Athens Hill Area C. increased the sample, resulting in a higher These patterns reflect a uniformity in the ratio. Screens were employed in excavations materials to which the tools were applied, at the Lamb, Potts, and Vail sites, where nev- across a wide geographic area. Again, despite ertheless the ratios were smaller than at West the complexity of possible alterations due to Athens Hill. Possibly, the specific kinds of the effects of different materials, it is suggest- reduction processes differed from site to site. ed that rounding/gloss on end scrapers is pri- An emphasis on thinning and retouch as marily due to hide-working or currying, on opposed to production of preforms might bifaces due to meat-cutting. Edge-crushing, result in the deposition of many small flakes, on the other hand, must be the result of pres- as exemplified at Lamb, Potts, and Templeton. sure exerted on hard substances such as A greater emphasis on primary reduction and wood, bone, and antler. Presumably, therefore, manufacture of preforms would presumably biface knives and scrapers were essential produce fewer and larger flakes, relatively equipment for the survival of Paleoindian speaking. This seems to be the case at West bands, part of the technology developed to Athens Hill. Other factors such as length of deal with resources available in the late-gla- occupation, size of occupying groups, number cial environment. of occupations, and nature of activities must Inferred Activities: No subsistence remains, also be considered. nor indeed have any organic substances sur- Differences in site productivity (i.e., the vived the agents of decay on the Paleoindian number of artifacts per square meter of exca- components in the local study area. Therefore, vation) are also subject to several variables. As we are forced to depend on the lithic assem- noted before, West Athens Hill Areas B and C blages for inferences concerning the various and Vail show the highest values at 4.7, 7.9, tasks and activities performed at these sites. and 9.1 items per square meter, Templeton is Functional aspects, rather than culture-histor- at 1.8, and the rest are under 1. On the other ical attributes of artifacts, are essential to such hand, the number of debitage items per analyses. The following combinations of func- square meter was highest at the Templeton, tional types (see Table 46 and Figure 46) are

111 linked with high probability to specific activi- horizontally discrete loci, not necessarily cre- ties: ated at the same time. Examples include the 4 Hunting and butchering: finished projec- to 5 loci at the Corditaipe site, the 7 or 8 at the tile points, other finished bifaces such Potts site, the 42 at Bull Brook, 3 or 4 at as knives, and utilized flakes. Whipple, and the 18 to36 at Vail. Concerning Fishing: netsinkers. not applicable in this the Vail site, Gramly (personal communica- case. tion 2002), now believes that there were at Plant food processing: pitted stones, most 6 or 7 occupations, on the basis of his milling stones, mortars, pestles, analysis of conjoined fluted points from the mullers. primary campsite and the killing ground. Biface production: bifaces in process, ham- Such non-random distributions can be merstones, anvilstones, hammer-anvil- interpreted in several ways. One is that larger stones, grooved abraders. clusters like those at Bull Brook, Potts, Hide-, bone-, and wood-working: end and Corditaipe, and Vail represent occupations by side scrapers, other unifaces. individual bands. It may prove difficult to establish whether those bands were all present Internal Patterning: As noted above the on site at the same time, or came and went at apparent clustering of artifacts in Area B at different times. If from simultaneous occupa- West Athens Hill was hypothesized to repre- tions, the clusters may represent subgroups, sent family dwelling or workshop areas, but perhaps extended families, within large my latest analysis inclines me to doubt that bands. more than perhaps 5 or 6 of the 13 originally Place in Regional Settlement Systems: proposed clusters may have had reality in pre- Data concerning settlement aspects of the historic time. Any patterning that may have above-listed habitation and quarry-workshop existed in Area A was destroyed by the bull- sites are augmented by analysis of the distri- dozing for the telephone company microwave bution of isolated finds of fluted points, as tower foundation. No obvious patterning was well as characteristic unifaces made of jasper observed in Area C. Some other northeastern and other exotics that characteristically repre- Paleoindian sites have produced evidence for sent Paleoindian occupancy (Funk 1993;

Figure 46. Bar graph comparing the relative importance of the hypothesized activities at Paleoindian sites in central and eastern New York. Series (activity) and site numbers correspond to those in Table 46.

112 (8) A, B (9) (7) Hill Areas Hill, Area C (5) (6) (4) 0 0 0 0 0 0 0 0 0 3? 0.2 0 0 (2) (3) 00000000000000001 0.3 23 9.3 (1) 110 100.0 390 100.0 247 100.0 114 100.0 711 100.0 167 100.0 144 100.0 1,525 100.0 1,228 100.0 Zappavigna Kings Road Swale Twin Fields Potts Corditaipe Port Mobil West Athens West Athens Activity1. Hunting andButchering 17 No.Fishing 15.5 Pct. 1072. Plant food No.Processing 27.4 Pct. 423. 0 Biface No. 2?Production 17.0 Pct.4. 1.8 18 Hide and/or No.Woodworking 22 15.8 Pct. 0 69 120 20.05. Other No. 62.7 16.9 105 0 Totals Pct. 177 29 26.9 No. 45.4 39 17.4 143 Pct. 0 15.8 39 57.9 No. 27.1 Pct. 3 93 749 81.6 No. 2.6 49.1 576 Pct. 910 15 81.0 No. 74.1 112 Pct. 2.1 67.1 26 95 15.6 66.0 10 359 23.5 6.9 81 414 27.2 6.6 237 19.3 Table 46. Relative importance ofresenting activities at particular Paleoindian tasks.* sites in central and eastern New York inferred from the frequencies of stone tools rep- *Activity and site numbers correspond to numbers in Figure 46.

113 Gramly and Funk 1990;Ritchie 1957; Wellman face site on a hill in Saratoga County pro- 1982). A settlement pattern much like that of duced four fluted points of relatively late later cultures is indicated. Isolated finds and styles (Funk and Walsh 1988). Also, three flut- defined sites occupy the same landforms, in ed points were recovered by Brennan (1977) relation to many of the same resources, as Late and his associates from the Piping Rock site in Archaic/Terminal Archaic groups. Westchester County. Because these locales are To summarize the data for isolated finds of considered sites, as opposed to isolated finds, fluted points in the Hudson valley (Ritchie those points are not listed in the above county 1957; Wellman 1982): inventory. Although this report is concerned Albany County: 2, from the Pine Bush on mainly with early Paleoindian manifestations, the delta of glacial Lake Albany. we might also note that late Paleoindian arti- Greene County: 3, from high fields along facts bearing typological resemblances to the the river. assemblage from the Reagen site in Vermont Orange County: 1, near Allard’s Corners. (Ritchie 1953, 1957) have been recovered from Richmond County (Staten Island): 1. some parts of the study area (Funk 1976; Funk Saratoga County: 4, including one from a and Schambach 1964). Among these is a group high field along the river, another from of parallel-flaked lanceolate projectile points the upland headwaters of the Snook of “Plano” form found in a feature at the Kill, a tributary of the Hudson. Lower Saranac River site near Lake Schenectady County: 1. Champlain and radiocarbon dated ca. 7000 Ulster County: 5 (3 from the glacial terrace B.P. (Karen S. Hartgen, personal communica- at or near the Twin Fields site). tion 1993). Westchester County: 1 from the hilly inte- Thus, both sites and isolated finds in rior. Orange County exemplify the same relation to These counts differ from a previous report the Black Dirt Area as later groups. Finds also (Wellman 1982) in that the points from exca- tend to be located along major river systems vated sites or systematically collected surface and lakes elsewhere in New York (Ritchie sites such as West Athens Hill, Kings Road, 1957, 1965). Most of the sites, even the Port Mobil, and Twin Fields are not included. Dutchess Quarry Caves, contained artifacts of The major stages of fluted point develop- later cultures and demonstrate the overlap in ment are represented in the group illustrated activities of these very different periods across by Ritchie (1957), including the Gainey-Debert the landscape. In other words, Paleoindian (Ritchie 1957: Plates 3A and 3B, Figures C and sites and find-spots are not necessarily to be D; Plates 2A and 2B, Figures F and G), found at higher elevations well above, or far Cumberland-Barnes (Plates 1A and 1B, Figure from, present-day drainage patterns. There is A; Plates 10A and 10B, Figure B); and no evidence that these people required special Crowfield (Plate 6, Figure B; Plates 10A and environments. Paleoindians entered deglaciat- 10B, Figure C). ed terrain some time after the retreat of the Since and overview in Archaeology of last Wisconsinan ice, a terrain already evolved Eastern North America (Wellman 1982), a few well along in the direction of its modern con- additional discoveries of fluted points have figuration. What was different in the terminal been reported for the Hudson valley. A com- Pleistocene milieu were the regimens of tem- plete re-inventory must await future study perature and precipitation, and the species of and publication. Two were found on the low- plants and animals living within the spruce-fir lands of Greene County during contract forests that were rapidly changing toward a archaeology investigations by Hartgen pine-oak forest. Undoubtedly, however, some Archaeological Associates (Karen S. Hartgen, geomorphic change took place, for example, personal communications 1993, 1997). A sur- in meandering of rivers that would have

114 destroyed some Paleoindian sites, or in the Dincauze further proposed that Paleoindians form of colluvial action that would have entered New England by way of a northern buried some sites under sediment. route south of the retreating ice front, presum- There have been few reports of ably crossing the narrower, shallower, less Paleoindian isolated finds in uplands and intimidating stretch of the river just south of mountainous areas, and no reports of sites per Lake George and Lake Champlain. I am skep- se. Perhaps 3 fluted points have been reported tical of this model, not only because some flut- in the Adirondack Mountains; this count does ed points have been found in Westchester not include 11 points found along or near the County, or because both points and sites are west shore of Lake Champlain including 5 on record for immediately adjoining from the Davis site (Wellman 1982). Small Connecticut, but because there has been rela- encampments of Archaic and Woodland peri- tively meager professional and amateur sur- ods do occur with some frequency on vey activity in those New York counties situ- Adirondack lakes and streams, but few ated along the river’s east side. I suspect that Paleoindian sites are on record for upland more fluted points and even substantial regions within the Glaciated Appalachian Paleoindian sites will eventually be found Plateau. Several retouched flake tools of there. Furthermore, it is entirely possible that brown Pennsylvania jasper, Paleoindian in Paleoindians had crude but serviceable water- style but without associated fluted points, craft and would not have been fazed by the were found on a farm in the uplands of the prospect of crossing the Hudson fjord Delaware Valley near Delhi (reported by (Engelbrecht and Seyfert 1994). Frank Schambach and seen by the writer, ca. With so few sites currently known and 1972). As archaeological survey and excava- almost no direct evidence of subsistence traits, tion projects continue and uplands become it is premature to offer detailed models of better known, with less bias toward lowland New York Paleoindian settlement systems— and major riverine locations, sites ranging in placing sites in seasonal rounds, for example, age from Early Archaic to Late Woodland and linking them up with inferences about the have been turning up with regularity, and it is movements of bands. The following prelimi- only a matter of time before Paleoindian sites, nary scheme of functional site types is offered: perhaps large and productive ones, are dis- Quarry-workshops located some distance covered in upland regions. from lakes, ponds, or running water: West Fluted points are rare in collections from Athens Hill. the east side of the Hudson valley, and no sites Open-air workshops/camps on low flat per se have been reported there unless the ground near quarries and small streams: the multicomponent Piping Rock site is accepted flats around Flint Mine Hill, the Railroad site as such. The situation in the mid-twentieth near West Athens Hill, the Kings Road and century prompted Ritchie (1957: 11) to suggest Swale sites near the Scott Farm quarries. that during warm-weather excursions of Open-air habitation sites near relatively Paleoindians into the valley from southern large streams and good fishing resources: Port regions, their “cultural equipment was Mobil, Twin Fields, Corditaipe, “Hallock.” unequal to the crossing of this wide and deep Open-air habitation sites near small creeks river.” Similarly, Dincauze (2001) proposed and brooks: Zappavigna, Potts, Davis. that because of lower sea level during the Late Sheltered sites (caves, rockshelters) locat- Pleistocene, the southern reaches of the ed at some distance from running water: Hudson River flowed through a deep gorge Dutchess Quarry Caves 1 and 8. that prevented Paleoindians from crossing from the west shore to the east shore.

115 Summary of Intersite Comparisons: size and weight of bifaces and unifaces, but artifacts in certain assemblages tend to be Most local Paleoindian sites lie at low smaller than those in the majority. The Vail, relief. The exceptions are West Athens Hill Potts, and Swale sites show the highest ratios and the Dutchess Quarry Caves. The sites are, of artifacts to debitage. Productivity in terms however, in varied topographic settings. The of artifacts and debitage per square meter of occupations were on ground that was level to excavation was highest at Templeton, Vail, gently sloping, except for the relatively steep and West Athens Hill. All the New York sites slopes adjoining Areas B and C at West Athens display exotic cherts among chipped stone Hill, on which some quarrying took place. items. The percentages vary, however, as do Most sites are on glacially derived soils the types of exotic stones represented. The including the Lake Albany clays (exception: assemblages with the highest percentages of the Templeton site on a flood plain). Most are exotic stones were at Twin Fields; Kings located in cultivated fields (again, the excep- Road/Swale, and Corditaipe (the very small tions are West Athens Hill and the Dutchess Davis site collection is not considered here), Quarry Caves). Occupied areas range from but Pennsylvania jasper was most heavily rep- small (the caves, 30 to 45 sq. m.) to fairly large resented at Kings Road and Swale. On most (open-air sites up to 8,100 sq. m.). Most are sites, exotic materials are seen primarily in located within 100 m of fresh water (the excep- end scrapers, secondarily in side scrapers. tions are the caves and West Athens Hill). Patterns of wear on unifaces consist chiefly of Most are near (within 800 m or less) bedrock edge-crushing on the bits or working edges, sources of good quality chert (the exceptions with rounding/gloss secondary. Wear on are Potts and Corditaipe). Most New York bifaces consists mainly of rounding/gloss. sites lack evidence of Paleoindian subsistence The primary activities represented are hunt- practices (the Hiscock site and just possibly ing/butchering, biface manufacture, and the Dutchess Quarry Caves may be the excep- hide-, bone-, or wood-working. There is little tions). Most pertain to the Gainey-Debert peri- to no evidence for plant food processing. od of early Paleoindian, but there is some evi- Internal patterning, in the form of discrete loci dence of occupation during the Cumberland- or clusters of artifacts—and sometimes fea- Barnes period. Radiocarbon dates from about tures—was observed in just a few sites; West 11,000 to 10,400 B.P. are confined to the Athens Hill Area B, Potts, and Corditaipe in Hiscock and Arc sites in western New York New York, and Bull Brook, Vail, and Whipple State. Artifacts in the assemblages are general- in New England. The eastern New York sites ly similar, the basic bifacial and unifacial types are too few and too scattered, and subsistence are repeated over and over. However, the data are too sparse to model subsistence-set- assemblages show considerable variation in tlement systems in any detail. Nevertheless, type frequencies; hence perhaps, in the relative known sites and isolated finds of fluted points importance of tasks and activities. Most of the are located in the same places on the land- assemblages show strong similarities in the scape as the traces of later peoples.

116 THE POSITION OF WEST ATHENS HILL AMONG THE PALEOINDIAN SITES OF NEW YORK STATE

West Athens Hill is the only substantial Quarry Caves 1 and 8 (Funk and Steadman quarry-workshop (with a possibly associated 1994), West Athens Hill lacked subsistence encampment) on record in New York State. remains. Also like the others, it failed to pro- Rare finds of fluted points on Flint Mine Hill duce datable organic substances pertaining to do, however, indicate a Paleoindian presence the Paleoindian occupation; only the Hiscock there. It is also possible that sizable fluted site (Laub, et al. 1988) and the Arc site point components await systematic excava- (Tankersley, et al. 1997) in western New York tion at the Mountain Top Quarries, Scott Farm have been successfully radiocarbon dated. quarries,7 and other chert-bearing stations in The relation of occupying bands to local Greene County and the rest of the Hudson food resources is not known, but I suspect that valley. they hunted the caribou and other late Apart from the caves on Mount Lookout, Pleistocene fauna, as hypothesized for the West Athens Hill is the only site elevated well Swale and King Road sites. It is even possible above the surrounding terrain. It is an open that the people who occupied West Athens site, not a cave or rockshelter, and is the Hill were the same people who occupied the largest of the studied sites, exceeding the Potts Kings Road and Swale sites. The fluted point and Corditaipe sites in area. As is the case styles indicate West Athens Hill was, like the with Potts, Vail, and Corditaipe, it produced other sites, occupied mainly in the Early evidence of intrasite activity loci. Paleoindian Gainey-Debert period, but there Slope on West Athens Hill is variable, but is reason to believe it was also occupied in the ground surfaces at the occupied loci were Cumberland-Barnes period. either level or gently sloping. Soils atop the The artifact assemblage from West Athens hill are derived from glacial till, and this is Hill is typologically very similar to assem- typical of most of the other sites. The site is blages from other northeastern sites. The prin- roughly 800 m (one-half mile) from the nearest cipal contrasts are in the large quantities of source of drinking water. Nearly all the other quarry-workshop debris and hammerstones sites are within 100 m of fresh water. from West Athens Hill. Oddly, the ratio of deb- Like all the other central and eastern New itage to artifacts at the site is not unusually York sites, possibly excepting the Dutchess high, compared to the others (Table 40). Like

______7 Both the Mountain Top and Scott Farm quarries have been extensively disturbed, the former by landscaping for a private home and construction of an apartment complex, the latter by removal of the most extensive chert-bearing deposits for “road metal” early in the 20th century. Nevertheless, substantial areas of both locations remain intact and could contain important workshops and campsites of the Paleoindian period.

117 those sites, no objects of bone, antler, wood or of the other assemblages, suggesting function- other organic materials survived alongside al and mechanical restrictions on those attrib- the stone tools.8 Along with the Vail site, West utes. Another shared trait is the presence of Athens Hill sports the highest artifact per non-local cherts. But the ratio of exotics to square foot figures in the comparisons of pro- local material on the site is among the smallest ductivity (Table 41). of any of the sites. Measurements of the size and weight of Table 47 summarizes the various compar- bifaces and unifaces from West Athens Hill are ative factors discussed in previous pages. surprisingly close to measurements for most

______8 Some of the bone tools from Dutchess Quarry Dave No. 1 (Funk and Steadman 1994) could have been associated with the Paleoindian occupation.

118 Table 47. Summary of data for Paleoindian sites mentioned in the text. N/A means not applicable.

119 Table 47. Part One, continued.

120 Table 47. Part One, continued.

121 Table 47. Part Two, continued.

122 ASPECULATIVE RECONSTRUCTION OF PALEOINDIAN QUARRY AND WORKSHOP ACTIVITY

On the basis of his detailed study of chert West Athens Hill was only one of many excel- quarries and workshops in the Wallkill Valley, lent sources of chert that was available to Phillip La Porta (personal communication Paleoindians and their successors in the mid- 1999) has proposed a “folk geology” that dle valley. enabled prehistoric denizens of the area to Having located promising bedrock locate, test, and extract chert from regional sources of chert, the Indians would set about outcrops. Millennia of experience in the Old evaluating the suitability of the outcrops for World, and later in the New World, created a their purposes. One consideration would be large and eminently practical body of tradi- the relative ease of access to the chert: whether tional knowledge that, passed orally from one too deeply buried, too high on a mountain or generation to another, would have enabled cliff face, or requiring too much time and hunter-gatherers entering new territory to dis- effort to mine. Others would include the cern promising rock formations from the “lay amount of chert within an outcrop, the thick- of the land.” ness of nodules or veins, the degree of weath- Paleoindians were in unfamiliar territory ering, and its general appearance. After upon first entering what is now the northeast- extracting chert from the mother rock, the ern United States, yet they had traversed very Indians would examine such attributes as the similar terrain on their way through parts of size of tabular pieces, vein plates, or nodules North America lying to the west and south. in terms of the size of artifacts to be produced; They had perfected their survival skills in late- the quality of material in terms of homogene- glacial environments, even before their ances- ity and the presence or absence of impurities; tors crossed the Bering Strait land bridge from the presence or absence of cortex, seams or Siberia some 12,000–13,000 years ago. Upon vesicles that might interfere with the flaking arriving in the Hudson valley, Paleoindians process and lead to premature breakage; and were already familiar with the animals and less crucial “aesthetic” factors such as color, plants needed for sustenance, and they would texture, and luster. have been able to evaluate the potential of Once selected, a bedrock source would be local bedrock exposures for cherts instead of exploited by attacking the veins or nodules wandering haphazardly around the land- using cobble hammers, as at West Athens Hill, scape. They quickly located and exploited out- or large, discoidal quartzite hammers or crops of the stones needed to make weapons wedges such as those reported by La Porta on and tools. the Wallkill Valley quarries. Fortunately, the Hudson valley abounds in The organization of people for quarrying a variety of high-quality cherts as well as would not necessarily be highly formalized. other rocks and minerals useful for tools, also Like the great majority of ethnographically perhaps for pigments used in body paint, dec- known hunters and gatherers, Paleoindian orating bone and antler objects, and so on. bands were undoubtedly egalitarian, organ-

123 ized by age and sex statuses, with very little ally assumed that men did most of the hunt- specialization of any kind, and it seems likely ing, and therefore were most proficient at that men, women, and children all participat- making fluted points, bifacial knives, and ed in different activities. Men, with their larg- other tools, it is possible that some women er muscles, would doubtless have handled the were equal, or even superior, to some men in tasks requiring the most strength, such as chert-knapping skill (some women may also extraction from the rock matrix, moving large have been superior to some men in their hunt- rocks, and the lifting and carrying of heavy ing skills). A sexual division of labor would chert samples. Women and children may have not necessarily preclude the occasional partic- done some of the extraction less demanding of ipation of women in “male” activities (or vice strength and taken part in most other quarry- versa). ing activities, including: raking away tailings We might also consider the possible role of from the quarry face, selecting blocks and ritual accompanying the process of quarrying, plates that would best serve as cores, rejecting selecting, and knapping chert. La Porta (per- less suitable materials, and packing and trans- sonal communication 1999) has observed porting of lighter chert samples. elaborate rituals taking place in concert with At ateliers, whether on top of West Athens the quarrying of limestone in India and con- Hill or down on the flats, a comfortably level siders this activity to be analogous to the min- ground surface would be used for artifact pro- ing of chert in the New World. Historically, duction as well as other tasks. Perhaps men, chert (flint) has had symbolic meaning in the women, and even older children were profi- myths and rituals of people in various parts of cient at all stages of the reduction process. the world, including the Iroquois (George These skills were crucial to the survival of Hamell 1982, personal communication, 1990). hunters and gatherers. Even though it is usu-

124 GENERAL CONCLUSIONS

One might speculate about the possibility piercing, and scraping tools. I doubt, however, of pre-fluted point occupations in the that pre-fluted point people could get along Northeast. It has proved very difficult to without bifaces for hunting, butchering, and establish the reality of occupations older than other tasks. Further, it seems unlikely that the about. 11,500 years (uncalibrated) throughout first New World occupations would have the Americas. No one has yet succeeded in given up the bifacial technology inherited finding and describing “Pre-Clovis” manifes- from their Siberian predecessors, who have tations to the satisfaction of the entire archae- been dated as far back as 25,000 years, only to ological profession. The debate goes on (Bryan resume making bifaces by 11,000 years ago 1978; Dillehay and Meltzer 1991; Shutler1983; (West 1996). West 1983, 1996). But if remains of such cul- The “Early Beringian” trait-complex of tures exist within the glaciated Northeast, Alaska and the Yukon described by West they could not be older than the beginning of (1996) represents a territorial expansion of the deglaciation in any particular location, unless Dyuktai complex from northeastern Siberia Pre-Wisconsin in age and confined to caves or and its principal manifestation is known as deep fissures overridden by the ice. In the Denali. These traits were carried by people middle Hudson valley, deglaciation occurred moving across the Bering Strait Land Bridge, around 15,000 radiocarbon years ago (Dineen an area of ocean bottom that was exposed to 1996). Like later cultures “Early Lithic” com- the sky during the drop in sea level caused by plexes would have needed to acquire chert for the accretion of the Wisconsin ice sheets that the manufacture of their weapons and tools. covered large areas of the globe. The lithic Therefore, chert-rich locations like West types included: bifaces of bipointed, ovate, or Athens Hill, Flint Mine Hill, Scott Farm straight-based forms; end scrapers; side scrap- Quarry, the Mountain Top Quarries, and others; limaces; burins; pieces esquillees; and ers, including the Munsungan sites in Maine, wedge-shaped and conical blade cores. These might be good places to look for traces of Pre- complexes were ultimately derived from the Clovis assemblages. Unfortunately, no evi- Eurasian Upper Paleolithic (close relatives dence of such assemblages has been reported consisted of the Aurignician, Magdalenian, from these sites or other chert quarries in the and Solutrean cultures). Radiocarbon dates Northeast. strongly indicate that the initial occupation of Some writers have speculated that “Early eastern Beringia (Alaska) was no older than Lithic” industries would look rather different about 12,000 years. The once-popular belief from the familiar fluted point complexes. For that the first people to venture into the New example, Alan L. Bryan (personal communica- World dated back 15,000–20,000 years or more tion 1980) proposed that bifaces would be has not been supported by recent research in much less important than in Paleoindian the Arctic (Hopkins 1996; West 1996). assemblages, even absent in most cases. Fluted points are rare in the early Alaskan Instead, the older assemblages would consist assemblages and where they do occur, appar- chiefly of various unifacial chopping, cutting, ently pertain to the initial Holocene epoch,

125 although an occupation very similar to the attributes, they would appear indistinguish- Agate Basin complex of the Great Plains, fea- able from the products of later industries. turing lanceolate, parallel flaked projectile Furthermore, “Early Lithic” unifaces might be points, has been dated ca. 10,000-11,000 B.P. hard to distinguish from Paleoindian unifaces. Geological and paleoecological data support We do know that no more than 3,000 or the existence of an ice-free corridor between 4,000 thousand years after the final recession the Cordilleran and Laurentide ice lobes in of the Wisconsinan glacial ice from the midwestern Canada from about 10,000 to 13,000 Hudson valley, but shortly before the final years ago. West (1996) postulates a rapid draining of glacial lake Fort Ann and its suc- migration of Early Beringians down the corri- cessors (ca. 10,300 radiocarbon years ago) dor into the area that is now the United States. (Dineen 1996), small bands of Paleoindians He further assumes that that vast territory had entered the area and discovered the excellent never before been occupied by humans. They chert resources in the rock exposures of pres- were probably hunting large and small mam- ent-day Greene County. They established a mals, including mammoth, caribou, red deer, succession of quarry-workshops and also, per- musk ox, and horse. At some time in their haps, short-term, limited-purpose camp sites travels they developed a fluted point technol- atop West Athens Hill. By far the main attrac- ogy that spread rapidly throughout the conti- tion of the hilltop was the high-quality nent. Normanskill chert in bedrock outcrops. Since It is possible that the earliest immigrants this occupation dates from the final into our area possessed the basic Denali tool Pleistocene, it seems likely that vegetation kit, including bifaces that had not yet been was rather sparse on the rocky hill, where the transformed by knappers into fluted points modern humic cover had not yet developed. (fluting was actually a logical application of At best, perhaps, scattered conifers, dwarf wil- the Early Beringian sophisticated blade tech- lows, sedges, and grasses had obtained a nology).9 This very early period of occupation foothold. Thus, the view in any direction may probably would not have lasted more than a have been almost entirely unobstructed by few decades, and sites would be few and far trees, and the summit of the hill would have between. Unless archaeologists are fortunate provided an ideal vantage point for observing enough to encounter assemblages with Early the movements of game in the surrounding Beringian non-fluted bifaces and unifaces terrain, including the adjoining Hans Vosen from the time of ca. 12,000 to 11,000 years ago, Kill Valley. it may prove very difficult to demonstrate the At one time or another, the whole summit reality of pre-fluted point occupations south area of West Athens Hill was exploited. The of the Canadian border. debris of quarrying and chert-knapping cov- It remains possible that hidden among the ers the top and upper slopes of the ridge. The numerous bifaces from West Athens Hill, Flint evidence from Area B was originally interpret- Mine Hill, and other investigated quarries are ed to mean that human activities were con- some bifaces derived from pre-fluted point centrated in rather small loci, either work- components. Ovate and lanceolate bifaces, shops or nuclear family domiciles. The artifact especially those in process, found on camps types in each locus seemed to reflect a wide and workshops located in surrounding low- range of behavior, relating to the hunt, the lands, would not attract much attention butchering of game, the mining and working because lacking fluting, parallel flaking, of chert, and possibly the processing of bone, stemmed basal portions, or other distinctive hides, and wood. But my reanalysis shows the

______9 It should be noted that true blades and polyhedral or prismatic cores are extremely rare in eastern Paleoindian assemblages, though they have been reported for Clovis in the western United States.

126 so-called “clusters” to be much less convinc- basal fragments discarded, and newly made ing and less well defined than I once believed, points attached. But if this was a regular in part because of prior disturbance in middle occurrence, then why are there also a fair units of the excavation grid. Apart from the number of tip and midsection portions of fin- problem of the clusters, our data do not indi- ished fluted points? Why would tips and mid- cate spatially separated zones of specialized sections of broken points be brought to the activities on the site, such as heat-treatment of quarry-workshop along with the bases, chert, butchering of game, processing of hides, instead of left in the field? The Stage 4 frag- manufacturing of dart shafts, and so forth. ments do not appear to have broken during It is quite certain that the West Athens Hill efforts at edge retouch or end thinning; the site was a prehistoric chert quarry. But the pre- fractures apparently resulted from force exert- cise nature, sequence, and scheduling of the ed against the faces and tips of the bifaces. extraction process and the organization of Perhaps some fluted points were broken on work parties remain unknown. The bathtub- site because they were secondarily used as shaped pit atop Area A, containing 16 cobble knives. hammerstones, suggests that promising veins Stage 3 and 4 bifaces, other than fluted were worked to exhaustion by digging down- points, were only moderately abundant in ward. In other areas, the Indians attacked the Areas A and B, and rare in Area C at West exposed strata still visible today on the side Athens Hill. These include finished biface slopes of the summit. Cobble hammerstones, knives. All these knives are fragmentary and most not of particularly large size, apparently some were possibly broken in use at campsites sufficed to separate the good-quality material and brought to the site, still in hafts, to be from the matrix. In Area C, chert-rich bedrock replaced. This suggests that many advanced sections projected up above the ground sur- bifaces were carried off site and finished at face or lay flat just under leaf litter. The pic- other localities. But on the other hand, most of tured outcrop (see Figures 13 and 15) had a the knife fragments are tips rather than bases. scalloped edge that resulted from knocking off It seems illogical for tips as well as hafted large chunks of chert. There were also small, bases to be brought to the site, rather than left oval-scarred areas on the top from use as a at the place where breakage occurred. It platform for quartering blocks. The quarry appears instead that most knives were manu- technology described for the Wallkill Valley factured, used, broken, and discarded on the chert resources by La Porta (personal commu- site. This inference, and the presence of the tip nications, 1990–1999), including large, dis- and mid-sections of finished, broken Stage 4 coidal quartzite “picks” or “wedges,” is not in fluted points, pose a problem unless the West evidence at West Athens Hill. However, one Athens Hill site was a multipurpose encamp- such object of unknown cultural and temporal ment as well as a quarry-workshop. provenience was found by Thomas Weinman More than quarrying and biface produc- on the surface at the Scott Farm Quarry. tion also seems indicated by the abundance of The West Athens Hill site was also a work- scrapers. Unifaces were probably also pro- shop, where bifaces were roughed out and in duced on site, although this has not been sys- some cases finished. Clearly, fluted points tematically investigated in the debitage. The were manufactured on site, as shown by the wear on scrapers from Areas A and B consists moderate number of fluted points in process mostly of edge-crushing, very possibly from in addition to finished fluted points. Along use on hardwood and bone in the manufac- with broken finished points, this implies that ture of dart shafts and bone tools. Scrapers weapons such as darts or spears with point from Area C show a different pattern, chiefly tips broken off during the hunt, but the bases rounding and gloss on the working edges; this still attached, were brought to the site, the may indicate use on softer materials such as

127 animal hides, soft wood, or bark. Further- absent. There should be numerous rock frag- more, the apparent artifact clusters in Area B ments displaying both reddening and angular (those that upon restudy appear to retain edges from exposure to high temperatures. some validity) may represent restricted activi- Again, a very small number of indisputable ty loci, in each of which the same basic activi- fire-cracked rocks were found. These could be ties were performed. Even in the absence of from Archaic and Transitional occupations features and post molds it is difficult to escape rather than Paleoindian, however. There the notion that the clusters were family should also be charcoal from fires in the dwelling and working areas. On the other undisturbed Area B deposits, but none was hand, both clusters and features were absent found apart from the historic charcoal in in Area C. Feature 1 and in other concentrations in Area If the lack of features means the Indians B believed to be from burned tree stumps. did not camp on site, we must envision the Utilization of West Athens Hill by possibility that they camped nearby some- Paleoindian bands or work parties was proba- where on the flats, scaling the hill to quarry bly confined to no more than a few days per chert, make bifaces, refit darts with new visit, perhaps occasionally only a few hours, chipped stone points, manufacture bone and and could have occurred at any time of the antler tools and weapon shafts, and work year. The notion of very brief and sporadic hides. Based on surviving evidence, no fea- visits seems to be supported by consideration tures were constructed, nor food cooked, and of the situation in Area B. There the artifact so on. Furthermore, we must assume that, clusters were limited in number and size, and when finished with their tasks, they walked if my original interpretation of the strati- back down the hill to their campsites, where graphic data is correct, only from three to five they built fires, cooked food including meat loci in the excavated area could have been from successful hunts, and carried out other occupied simultaneously. These figures tasks not necessarily accomplished on top of should probably be lower, since not all the the hill. But this scenario seems eminently clusters from one stratigraphic set were neces- impractical. If the camps were on the flats not sarily coeval. If the occupations had been long far from the hill, the trip up or down probably and intensive, and repeated over decades or took 15 minutes, a 30-minute round trip. Once even centuries, the locations of activities chert was acquired or bifaces roughed out, would have shifted around within the area, why stay on the hill for all the other inferred more objects would have been deposited, and tasks rather than return to camp? It would more overlapping would have occurred, with seem that heat treatment of chert, finishing of consequent “smearing” of the artifact distri- points and other tools, hunting, butchering, bution. A limited amount of such overlapping hide-working, wood-working and other activ- may have occurred in the case of Clusters 4 ities, would most efficiently be conducted on and 11. Individual bands did not necessarily or near a camp, not on a hilltop quarry visited return to West Athens Hill every time they for short periods. But if people were living, needed some chert, since there were equally quarrying, and knapping chert and perform- fine sources available nearby at places like ing other tasks on top of the hill, why aren’t Flint Mine Hill (Brumbach and Weinstein there any hearth features? I am unable to offer 1999; Parker 1924), the Mountain Top a convincing answer. Quarries, and the Scott Farm Quarries. The sediments on the hilltop should show The exotic materials used for some reddened patches from oxidation of iron salts chipped stone artifacts, including Pennsylva- by fires—but no convincing reddening was nia jasper, Vanport chert (Flint Ridge “chal- observed. Basin- or saucer-shaped features cedony”) Upper Mercer chert, and western indisputably of human construction were also Onondaga chert are a clue to the high mobili-

128 ty of the Paleoindians who used the site and which passed through Greene County on its they again suggest that hilltop activities were way to the Hudson River and Connecticut not limited to quarrying and lithic reduction. (Cassedy 1996), but no Paleoindian compo- This does not mean, however, that in all cases nents or stray fluted points were found there. visits were actually made by these people to Extensive workshops are to be found atop the sources of the stones in Pennsylvania or other chert-bearing ridges in Greene County, Ohio. It is equally reasonable to suppose that, including the Mountain Top Quarries, Scott for example, a band roving the lower Hudson Farm Quarry, and Flint Mine Hill. Early stage valley on its way to the middle valley would bifaces have been unearthed from Scott Farm occasionally encounter a band from eastern and Flint Mine Hill (Brumbach and Weinstein Pennsylvania, with the result that chert and 1999; Weinman and Weinman 1969, 1977). jasper specimens would be exchanged and Two fluted points in process have been recov- carried farther from their points of origin. ered from the highest portions of Flint Mine Where might other Paleoindian sites be Hill. But the extent to which Paleoindians found? Since only West Athens Hill, Kings lived as well as worked atop these ridges Road, and Swale are on record for the large remains to be established. As at West Athens area from the northern Hudson valley south Hill, no features were observed in the excava- to Twin Fields in Ulster County, and none are tions of the State University at Albany on Flint on record in counties east of the river, it is pre- Mine Hill (Brumbach and Weinstein 1999) or mature to construct a model of Paleoindian the Pleasantdale quarry near Troy (Brumbach subsistence-settlement systems and seasonal 1987). Certainly there is much to be learned rounds. It would be sheer fantasy to propose through research on chert quarries that were such functional types as fishing, fowling, exploited by Native American groups of all berry picking, and chert-knapping stations as prehistoric periods. integral parts of the system that comprised the At least two fluted points were found by West Athens Hill quarry-workshop and the collectors at multicomponent workshops on possible Kings Road/Swale caribou hunting the clay flats surrounding Flint Mine Hill. One encampment. Nevertheless, we should expect also turned up in testing an area at the nearby more variability in settlement types than indi- Coxsackie Correctional Facility, and another at cated by the list of known sites, most of which the location of a compressor station for the are assumed to be hunting/butchering/ Iroquois Natural Gas Pipeline (Karen S. biface-producing encampments. Such assump- Hartgen, personal communications 1993, tions should be provisional in view of the 1997). It seems probable that additional sites major gaps in our knowledge. comparable to Kings Road/Swale exist in But where might additional Paleoindian other places on the Athens Flat. Yet others sites be found in Greene County? The avail- may exist in the largely unsurveyed terrain of able data show that sites and isolated finds of low hills and occasional chert-bearing out- fluted points occur in much the same places as crops east of the flat and west of the river. So subsequent Archaic peoples. In other words, far neither fluted points nor fluted point sites they are all over the landscape. Some have been reported at low-lying terrain along Paleoindian sites including quarry-workshops the river, as for example at Four Mile Point might be found west of the Helderberg (Funk 1976; Schambach 1962). It seems clear Escarpment, where there are a number of that in the absence of an effective predictive small quarry-workshops of largely unknown model of site locations, new Paleoindian sites cultural provenience. Some of these sites were may only be discovered by extensive coverage systematically investigated during surveys of Greene County on foot. along the Iroquois Pipeline right-of-way,

129 What happened to the people who occu- points and other notched forms, dated as far pied West Athens Hilll? The same groups who back as 9,500 years (Broyles 1971; Funk 1976, exploited the chert outcrops there must also 1983, 1993; Ritchie and Funk 1971, 1973). But have visited other sources such as Flint Mine the interval between Paleoindian and Kirk Hill and Scott Farm. Also, different groups remains a mystery in the Northeast. In the must have visited the site at different times. Midwest, Southeast, and parts of Canada, the Presumably, the bands were not permanent transition from fluted points to later Archaic residents but usually moved on to new hunt- horizons is represented by late Paleoindian ing grounds, like the fictional people in the unfluted lanceolate projectile point styles, but Prologue to this report. Newcomers must there is meager evidence for a lanceolate point often have entered the area from adjoining horizon in New York and New England regions like the Delaware Valley. Eventually, (Benmouyal 1987; Funk 1978, 1983; Funk and the regional cultural configuration changed Schambach 1964; Gramly and Funk 1990; until its artifact traits were no longer identifi- Ritchie 1953, 1957; Wright 1972). Bridging the able as Paleoindian, a pattern repeated across gap after the lanceolates in the Southeast and North America by around 10,000 radiocarbon Midwest is the Dalton complex, defined years ago. This means that projectile points chiefly by a group of distinctive point styles were no longer of fluted lanceolate form, collectively known as “Daltons” and some although the other identifying traits such as associated traits. Daltons often display long spurred end scrapers and large retouched basal thinning flake scars similar to fluting, flake side scrapers probably persisted into and were probably derived from the fluted immediately following complexes. This point technological tradition, although the episode of cultural change may have been blade form varies. The culture is radiocarbon stimulated by the climatic and biotic changes dated ca. 9000–10,000 B.P. (Gramly and Funk that accompanied retreat of the Wisconsinan 1991; Gramly 2002; Goodyear 1982). ice sheets. The ultimate result was the series of But a Dalton horizon has not been identi- manifestations known as “Archaic” that rep- fied in the Northeast and no convincing occu- resented adaptations of eastern New York pational remains of comparable age have populations to early Holocene ecological con- come to light. Dalton points are rarely seen in ditions. northeastern surface collections. Possible Almost certainly, the area of Greene alternative candidates have been suggested County was never completely depopulated (Funk 1991). Until concrete evidence is avail- after the close of the Paleoindian era. The able for something like Dalton, I offer the same goes for the rest of the Northeast. But hypothesis that regional Paleoindian bifaces there are 2,000 “missing” years in the extant evolved first, into a lanceolate point tradition archaeological record for Greene County. The similar to Crowfield and Plano, then into a next-oldest occupation we can be sure about complex featuring as yet unidentified projec- consists of bifurcated-base projectile points tile points, perhaps triangular or lanceolate from Area B at West Athens Hill and in collec- triangular in form but not yet recognized in tions from other local sites. These types have surface collections because they resemble tri- been radiocarbon dated elsewhere at ca. 8,000 angular points of later Archaic and Woodland B.P. (Broyles 1971; Chapman 1977; Justice periods. Convincing evidence of such occupa- 1987). tions has yet to be found in context on single- There is little information on the pre-bifur- component or stratified multicomponent sites. cate occupations of the Hudson valley. In the But triangular points of variable form are Southeast, on Staten Island, and in the Upper occasionally recovered from the deep Archaic Susquehanna Valley, the preceding horizon is levels of stratified sites, including caves and known to comprise Kirk Corner-Notched rockshelters. Whatever their identity, the earli-

130 est northeastern Archaic occupations remain Despite the partial disturbance of West elusive. Perhaps Indian populations were rel- Athens Hill by construction of an access road atively small because there were environmen- and foundation for the mobile telephone relay tal constraints on the adaptations of hunters tower and the limited and sporadic but ongo- and gatherers in the early Holocene period; ing looting by collectors, a considerable area hence there would be relatively few sites dis- of the deposits remains intact. Unquestion- tributed over the landscape, and the existing ably, the site has the potential to yield addi- sites might be small, poor in occupational tional information on Paleoindian occupa- debris, and hard to find (Funk 1996). tions of the Hudson drainage basin. Therefore, It is assumed that the complexes that fol- it is hoped that ways will be found to protect lowed, and presumably evolved from, the the summit from development and prevent hypothetical Dalton-like horizon were repre- further mining of the hillsides. Several times sented by the Kirk, and possibly Kessel, St. in past years the writer and his colleagues Charles, and other Early Archaic projectile tried to arrange for the State of New York to point types first described in the Southeast purchase the property and provide some (Justice 1987). Unfortunately, despite the best measure of protection. Unfortunately, those efforts of archaeologists, key projectile point attempts failed. types are weakly represented in surface collec- Since West Athens Hill is listed on the tions, and no single-component or stratified National Register of Historic Places, it is subject stations containing occupational remains of to Section 106 Review if threatened by federal- these periods have been found and excavated ly funded or licensed construction projects. in the Hudson drainage. Kirk components This review would ensure that serious consid- were, however, excavated and radiocarbon eration be given to the site’s various qualities, dated ca. 9,000 B.P. in the Upper Susquehanna including but not necessarily limited to the sig- Valley. The somewhat better known bifurcat- nificant prehistoric components. The outcome ed-base complexes succeeded the Kirk hori- would be either a limited and systematic exca- zon (cf. Ferguson 1995; Funk 1993, 1996). The vation of areas to be adversely impacted, or following Middle Archaic Neville horizon of preferentially complete preservation of the site about 7,000 years ago is fairly well under- from any disturbance by federal government stood, largely as a result of investigations by actions. But private home building remains a Eisenberg (1991) at the Mohonk Rockshelter, possibility and looters continue to damage the situated atop the Shawangunk Mountain site. Unfortunately, we will never have com- Ridge near New Paltz, and by research in plete information on the types and quantities of other parts of the Hudson valley (cf. Dincauze artifacts removed from West Athens Hill by col- 1976; Funk 1989, 1991a, 1991b, 1996). lectors over the last 40 years.

131 ACKNOWLEDGMENTS

The writer was assisted in the 1966 exca- Wellman and Anthropology Curator Charles vations by four Science Research Aides: H. E. Gillette catalogued the collections. As pre- David Tuggle, J. Cynthia Weber, Daniel viously noted, Wellman carried out the Danker, and R. Michael Gramly. The crew in tedious chore of counting and classifying the August 1969 included Ralph M. Houck, J. thousands of cores and flakes recovered from William Bouchard, Neal Trubowitz, and Beth Areas A and B. Wellman. Wellman was a staff archaeologist at I am also indebted to officials of the New the State Museum and still is. In 1970, the York Telephone Company (the company has team consisted of Houck, Bouchard, Wellman, since changed its name, first to Bell Atlantic and Ronald C. Corbyn. and presently Verizon), who granted permis- Many volunteers from the New York State sion for the investigations in 1966, 1969, and Archaeological Association contributed their 1970. labor to the excavations on this important site. Comparative analysis was facilitated by Of this group, the dedication of R. Arthur access to important collections of Paleoindian Johnson was particularly outstanding. The artifacts from other sites. Thomas P. Weinman writer is grateful to all of these people for their loaned me his materials from the Kings Road valuable assistance. and Swale sites, and Joseph Diamond, Assistance in interpreting the sediments at Department of Anthropology, State University the site was provided by glacial geologist G. of New York at New Paltz, made available the Gordon Connally, then in the Department of collection recovered from the Twin Fields site Earth Sciences at the State University at by the late Leonard Eisenberg. R. Michael Buffalo. M. Raymond Buyce, Curator of Gramly, then at the Buffalo Museum of Mineralogy, Geological Survey, New York State Natural History, donated the materials he Museum, analyzed the collected soil samples. excavated at the Potts site to the New York Donald M. Lewis, Palynologist in the Biological State Museum. The collection from the Survey, processed the pollen samples, and Corditaipe site was donated to the Museum Edward Landing, State Paleontologist, also in by the discoverer, Noel Strobino. the Geological Survey, examined the possible Finally, I wish to express my appreciation petroglyph from Area B. to the New York State Museum for providing David R. Wilcox, Laboratory Worker (until budgetary support for the research, including 1968), and his successor, Beth Wellman, were field equipment, field vehicles, travel expens- of great assistance in analyzing and classify- es, and laboratory facilities. ing the materials collected at West Athens Hill.

132 APPENDIX DESCRIPTIONS OF CHERTS COMMONLY FOUND ON PREHISTORIC SITES IN EASTERN NEW YORK

The cherts of the middle Hudson valley Colors are creamy blue, dark pearly blue, dark have been characterized by several researchers cobalt blue, Munsell value 5. Slight vitreous (Hammer 1976; Kuhn and Lanford 1987;Lavin luster, weathers gray with yellow-brown and Prothero 1981; Wray 1948). Methods of stains. analysis vary, but the chief approaches are visu- Whitehall/Ticonderoga chert. Similar to al examination with a petrographic microscope Fort Ann chert. Also called Beekmantown and trace element analysis using x-ray fluores- chert or Mount Independence chert. Blue-black cence. Here the emphasis is on visual qualities. to black. Fine structure emerges under the microscope, Deepkill chert. Lower Ordovician in age; including crystals and particles of varying underlies the Normanskill formation. composition as well as fossils. The following Distinguished by its predominantly apple- characterizations are selective paraphrases of green color, and its tendency to weather those published in Wray (1948), Hammer brownish. There are also gray, blue-gray, and (1976), and Lavin and Prothero (1981). red varieties. The Deepkill formation is the Little Falls or Knauderack chert. Upper oldest chert-bearing shale in New York State. Cambrian in age. Colors are tan with milky Normanskill chert. As explained above, blue inclusions, grading into milky medium this term has been used by archaeologists as a gray, Munsell N4 to N6. May be mottled with general term for the material that character- limestone and quartz inclusions. Some vari- izes the Mount Merino subunit of the series eties tend to blue, gray and white, dark and predominates in the outcrops at Flint brown, and even red. Mine Hill, West Athens Hill, Scott Farm, and Petrographic analysis indicates a very other places in Greene County. Also Lower fine-grained, even textured chert with abun- Ordovician in age. The colors are primarily dant well-rounded clastic quartz grains, green, green and black, black, blueish green, sometimes showing undulose extinction. and dark olive green. It sometimes occurs in Muscovite flakes and small amounts of red, red with black, grayish green, and dark opaques—usually hematite—are present. gray. Munsell colors include: 5BG 4/1, 5BG Large euhedral dolomite rhombs and their 3/1, 5GY 3/1, 5GY 4/1, 5GY 5/1, 5B 4/1, 5G pseudomorphs, and ghosts of ooids and fos- 3/2, 5Y 3/1, and 2.5 YR 3/2. Texture is sils, are often present; all are commonly smooth, luster varies from dull to waxy. rimmed with hematite. Hematite also occurs Mostly weathers white or brown. in beds. Small patches of chalcedony are often The Indian River subunit of the present. Well-defined relict bedding is visible Normanskill series also produces gray, green, in some specimens. Intra-outcrop variation is and black cherts, and less often, red cherts. minimal. The Austin Glen member is not known to con- Fort Ann chert. Also of Cambrian age. tain cherts.

133 Snake Hill chert. Middle Ordovician. imens. Fossils (trilobites, brachiopods), frac- Restricted to Saratoga County. Actually an ture planes recrystallized with calcite, sparry indurated siliceous shale. Greenish gray in calcite pseudomorphs of fossils, dusty calcare- color, white weathering. It is splintery and not ous aggregates, and opaques (including iron of good quality for flaked stone tools, though oxide) may sometimes occur. There is only a it was sometimes quarried and used by minor development of fibrous chalcedony; Indians living in the area. strongly chalcedonized chert and chalcedony Helderberg cherts. Lower Devonian in age. veins are uncommon. Hammer (1976:50) lumps the four chert-bear- Glenerie chert. Lower Devonian. Colors ing members of this sequence (Kalkberg, New blue to shiny black, dull, Munsell N2. Vitreous Scotland, Becraft, and Alsen) together. “All of and granular. Weathers to drab gray, often these cherts are black or very dark blueish with reddish stains. black with a characteristic white dendritic Oriskany chert. Also Devonian in age. scaling which occurs parallel to the original Dull black with quartz grains included, bedding plane. The cherts are smooth and Munsell N2. waxy and may have pockmarks or pure inclu- Esopus chert. Lower Devonian. Color dull, sions scattered throughout; Munsell N2 to dark and muddy black or gray, Munsell N3. 5BG 2, 5BG 1.” They weather to a thin, gray- Weathers to a slight grayish cast. to-white coating. Eastern Onondaga chert. Middle Within this group, Lavin and Prothero Devonian. “Color ranges from dark brownish (1981) describe Alsen chert as having numer- blue to blueish black to black, Munsell N2 to ous well-preserved Lower Devonian fossils of N3, value 2 to 4. The chert is hard, smooth, corals, bryozoans and brachiopods, usually and waxy” (Hammer 1976:48). showing replacement to chalcedony. Dolomite The bedrock sources of central and west- is present, usually as numerous small rhombs, ern New York Onondaga cherts are outside but few to numerous large rhombs may also the Hudson-Mohawk drainage, but are occur. Pyrites and small amounts of carbona- described below because they often occur in ceous matter are also present. Intra-outcrop Paleoindian chipped stone assemblages of the variation is minimal. region. New Scotland chert is a strongly chal- Central Onondaga chert. Middle cedonized carbonaceous chert. Large angular, Devonian. A cloudy dark to brownish blue anhedral grains of clastic quartz occur. Ghosts chert. Smooth and waxy, Munsell value 4 to 5. of carbonate clasts, dolomite rhombs, pyrites, Western Onondaga chert. Middle and abundant fossils including trilobites, bra- Devonian. Color from tan, khaki, through chiopods, and corals are present. Iron oxides, light gray to milky blueish white, Munsell chalcedony spherules, opal and segregated 5YR4, N6, N5. Texture smooth and waxy, col- zones of clean chalcedony, and opaque-laden ors mottled and cloudy. carbonate-rich chert may sometimes occur. The Onondaga cherts are generally fine- Kalkberg chert is very fine-grained and grained. Dolomite is common as is fine patch- uniform in texture. It is predominantly a silici- es and/or rhombs. Both large and small fied limestone with carbonates comprising rhombs occur. Pyrites are always present. 30–55 % of each slide. Dolomite rhombs, when Fossils of brachiopods, corals, crinoids, and developed, are usually small. Pyrites and car- bryozoans are often well-preserved. Chert is bonaceous matter are present. The latter strongly chalcedonized. But fibrous chal- clouds calcite and forms ghosts, styloitic cedony frequently occurs as veinlets, seams, seams, and pelletoids. Anhedral, angular clas- patches, and spherules. tic quartz is often present; it is much finer- Kuhn and Lanford (1987) studied samples grained than quartz in the New Scotland spec- of Kalkberg, Normanskill, and Beekmantown

134 cherts using x-ray fluorescence. Relative pro- Helderberg cherts from Normanskill cherts portions of iron, rubidium, strontium, and zir- due to very similar trace element profiles. conium were examined, and with the aid of Cherts from Cambrian and Ordovician statistical programs, the investigators were formations in the Wallkill drainage of New able to differentiate the samples from the three York and New Jersey have been studied by La main bedrock sources. The technique showed Porta (1996). They are rarely seen in archaeo- promise as a means of identifying chert sam- logical collections from the middle Hudson ples from archaeological sites, but there were valley and will not be described here. some problems with correctly discriminating

135 REFERENCES CITED

Benmouyal, J. Brook Site. American Antiquity 20 1987. Des Paleoindians aux Iroquoiens en (3):274–276. Gaspe: Six Mille ans d’Histoire. Ministere 1966. The Debert Archaeological Project: des Affaires Culturelles Direction de l’Est The Position of Debert with Respect to du Quebec. Quebec, Canada. the Paleoindian Tradition. Quaternia Braun, E. L. 8:33–47. 1950. Deciduous Forests of Eastern North Callahan, E. America. Blakiston, Philadelphia. 1979. The Basics of Biface Knapping in Brennan, L. A. the Eastern Fluted Point Tradition: A Manual for Flintknappers and Lithic 1977. The Lower Hudson: The Archaic. In Analysis. Archaeology of Eastern North Amerinds and Their Paleoenvironments in America 7: 1–180. Northeastern North America, edited by W. S. Newman and B. Salwen. Annals of the Carter, D. B. New York Academy of Sciences 288:411–430. 1966. Climate. In Geography of New York Broyles, B. J. State, edited by J. H. Thompson, pp. 54–78. Syracuse University Press. 1971. Second Preliminary Report: The St. Syracuse, New York. Albans Site, Kanawha County, West Virginia. West Virginia Geologic and Economic Cassedy, D. F. Survey, Report of Investigations 3. 1996. From the Erie Canal to Long Island Morgantown. Sound; Technical Synthesis of the Iroquois Brumbach, H. J., and J. Weinstein Pipeline Project. Iroquois Gas Transmission System, Shelton, Connecticut. 1999. Material Selection, Rejection, and Failure at Flint Mine Hill: An Eastern Chapman, J. New York State Chert Quarry. Northeast 1977. Archaic Period Research in the Lower Anthropology 58:1–25. Little Tennessee Valley. The University of Bryan, A. L., editor Tennessee, Department of Anthropology, Report of Investigations 18. Tennessee 1978. Early Man in America From a Circum- Valley Authority. Knoxville. Pacific Perspective. Occasional Paper No. 1, Department of Anthropology.University Connally, G. G., and L. A. Sirkin of Alberta. Calgary. 1970. Late Glacial History of the Upper Byers, D. S. Wallkill Valley, New York. Geological Society of America Bulletin 81:3297–3306. 1954. Bull Brook—A Fluted Point Site in Ispwich, Massachusetts. American 1971. Luzerne Readvance Near Glens Antiquity 19 (4):343–351. Falls, New York. Geological Society of America Bulletin 82:989–1008. 1955. Additional Information on the Bull

136 1986. Woodfordian Ice Margins, York State Museum Bulletin 455. The Recessional Events, and Pollen University of the State of New York. Stratigraphy of the Mid-Hudson Valley. Albany. In The Wisconsinan Stage of the First 1996. Holocene Environments in the Geological District, Eastern New York, edit- Hudson Valley. In A Golden Chronograph ed by Donald Cadwell, pp. 50–72. New For Robert E. Funk, volume 1, edited by C. York State Museum Bulletin 455. The Lindner and E. Curtin, pp. 57–72. University of the State of New York. Occasional Publications in Northeastern Albany. Anthropology, No. 15. Bethlehem, Cox, D. Connecticut. 1959. Some Postglacial Forests in Central Eisenberg, L. and Eastern New York as Determined by the 1978. Paleo-Indian Settlement Pattern in the Method of Pollen Analysis. New York State Hudson and Delaware River Drainages. Museum and Science Service Bulletin 377. Occasional Publications in Northeastern The University of the State of New York. Anthropology 4. Rindge, New Albany. Hampshire. Curran, M. L. 1991. The Mohonk Rockshelter: A Major 1984. The Whipple Site and Paleoindian Neville Site in New York State. In The Tool Assemblage Variation: A Archaeology and Ethnohistory of the Lower Comparison of Intrasite Structuring. Hudson Valley and Neighboring Regions: Archaeology of Eastern North America Essays in Honor of Louis A. Brennan, edited 12:5–40. by H. C. Kraft, pp. 155–176. Occasional 1996. Palaeoindians in the Northeast: The Publications in Northeastern Problem of Dating Fluted Point Sites. The Anthropology 11. Bethlehem, Review of Archaeology 17 (1):2–11. Connecticut. Deetz, J. Ellis, C. J., and D. B. Deller 1967 Invitation to Archaeology. Natural 1990. Paleo-Indians. In The Archaeology of History Press, New York. Southern Ontario to A.D. 1650, edited by C. J. Ellis and N. Ferris, pp. 37–63. Dillehay, T. D., and D. J. Meltzer, editors Occasional Publication of the London 1991. The First Americans: Search and Chapter, Ontario Archaeological Society, Research. CRC Press, Boca Raton, Florida. London, Ontario. Dincauze, D. 1997. Variability in the Archaeological 1976. The Neville Site: 8000 Years at Record of Northeastern Paleoindians: A Amoskeag. Peabody Museum Monographs View from Southern Ontario. Archaeology 4. Cambridge, Massachusetts. of Eastern North America 25:1–30. 2001. The Birds of Summer: Lakeside Engelbrecht, W., and C. K. Seyfert Routes into Late Pleistocene New 1994. Paleoindian Watercraft: Evidence England. Canadian Journal of Archaeology and Implications. North American 25:121–126. Archaeologist 15 (3):221–234. Dineen, R. J. Fenneman, N. M. 1986. Deglaciation of the Hudson Valley 1938. Physiography of Eastern United States. Between Hyde Park and Albany, New McGraw-Hill, New York. York. In The Wisconsinan Stage of the First Ferguson, J. Geological District, Eastern New York, edited by D. H. Cadwell, pp. 89–108. New 1995. The Haviland Site: The Early

137 Archaic in Schoharie County. Bulletin and Journal of Middle Atlantic Archaeology Journal of the New York State Archaeological 7:7–18. Association 110:1–15. 1992. Some Major Wetlands in New York Fisher, D. State: A Preliminary Assessment of Their 1955. Prehistoric Mammals of New York. Biological and Cultural Potential. Man in New York State Conservationist the Northeast 43:25–41. February–March:18–22. 1993. Archaeological Investigations in the Fitting, J. E., J. De Visscher, and E.J. Wahla Upper Susquehanna Valley, New York State. vol. 1. Persimmon Press, Buffalo, New 1966. The Paleo-Indian Occupation of the York. Holcombe Beach. University of Michigan, Museum of Anthropology, 1996. Holocene or Hollow Scene? The Anthropological Papers No. 27. Ann Search for the Earliest Archaic Cultures in Arbor. New York State. The Review of Archaeology 17 (1):11–25. Funk, R. E. Funk, R. E., D. W. Fisher, and E. M. Reilly 1972. Early Man in the Northeast and the Late-Glacial Environment. Man in the 1970. Caribou and Paleo-Indian in New Northeast 4:7–39. York State: A Presumed Association. American Journal of Science 268:181–186. 1976. Recent Contributions to Hudson Valley Prehistory. New York State Museum and Funk, R. E., and R. A. Johnson Science Service, Memoir 22. The 1964. A Probable Paleo-Indian University of the State of New York. Component in Greene County, New York. Albany. Pennslyvania Archaeologist 34 (1):43–46. 1978. Post-Pleistocene Adaptations. In Funk, R. E. and F. Schambach Handbook of North American Indians, vol- 1964. Probable Plano Points in New York ume 15, The Northeast, edited by B. G. State. Pennsylvania Archaeologist 34 Trigger, pp. 16–27. Smithsonian (2):90–93. Institution, Washington, D.C. Funk, R. E., and D. W. Steadman 1983. The Northeastern United States. In 1994. Archaeological and Paleoenvironmental Ancient North Americans, edited by J. D. Investigations in the Dutchess Quarry Caves, Jennings, pp. 302–371. W. H. Freeman and Orange County, New York. Persimmon Company. San Francisco. Press, Buffalo, New York. 1989. Some Contributions of Archaeology Funk, R. E., and J. Walsh to the Study of Cave and Rockshelter Sediments: Examples from Eastern New 1988. Evidence of Late Paleo-Indian York. Man in the Northeast 37:35–112. Occupation at Saratoga Lake, New York. The Bulletin, Journal of the New York State 1991a. Late Pleistocene and Early Archaeological Association 96:1–4. Holocene Human Adaptations in the Lower Hudson Valley. In The Archaeology Funk, R. E., P. L. Weinman, and T. P. and Ethnohistory of the Lower Hudson Valley Weinman and Neighboring Regions: Essays in Honor of 1969. The Kings Road Site: A Recently Louis A. Brennan, edited by H. C. Kraft, Discovered Paleo-Indian Manifestation in pp. 49–67. Occasional Publications in Greene County, New York. New York State Northeastern Anthropology 11. Archeological Association Bulletin 45:1–23. Bethlehem, Connecticut. Funk, R. E., and B. Wellman 1991b. The Middle Archaic in New York. 1984. The Corditaipe Site: A Small

138 Isolated Paleo-Indian Camp in the Upper Period in the Mid-South, edited by C. H. Mohawk Valley. Archaeology of Eastern McNutt, pp. 23–33. Occasional Papers 16, North America 12:72–80. Anthropological Research Center. Funk, R. E., B. Wellman, H.R. Decker, and W. Memphis State University. F. Ehlers, Jr. Gramly, R. M., and J. Lothrop 2003. A Small Paleo-Indian Encampment 1984. Archaeological Investigations of the in Orange County, New York. The Bulletin Potts Site, Oswego County, New York. and Journal of the New York State Archaeology of Eastern North America Archaeological Association. Submitted to 12:122–158. editor. Grimes, J., W. Eldridge, B. G. Grimes, A. Goldring, W., and J. H. Cook Vaccaro, F. Vaccaro, J. Vaccaro, N. 1943. Geology of the Coxsackie Quadrangle, Vaccaro, and A. Orsini. New York. New York State Museum 1984. Bull Brook II. Archaeology of Eastern Bulletin 332. Albany. North America 12:159–183. Goodyear, A. Hamell, G. 1982. The Chronological Position of the 1982. The Crystal Man-Being, the Little Dalton Horizon in the Southeastern White Wolf Man-Being and Dog Sacrifice United States. American Antiquity 47 in the Great Lakes and Northeastern (2):382–395. Woodlands Region. Unpublished manu- Gramly, R. M. script. 1982. The Vail Site: A Palaeo-Indian Hopkins, D. M. Encampment in Maine. Bulletin of the 1996. Introduction: The Concept of Buffalo Society of Natural Sciences 30. Beringia. In American Beginnings: The Buffalo, New York. Prehistory and Paleoecology of Beringia, edit- 1988. The Adkins Site: A Palaeo-Indian ed by F. H. West, pp. xvii–xxi. University Habitation and Associated Structure. of Chicago Press, Chicago. Persimmon Press Monographs in Isachsen, Y., E. Landing, J. M. Lauber, L. V. Archaeology. Buffalo, New York. Rickard, and W. B. Rogers 1999. The Lamb Site: A Pioneering Clovis 1991. Geology of New York: A Simplified Encampment. Persimmon Press, Buffalo, Account. New York State Museum New York. Educational Leaflet 28. The University of 2002. Olive Branch: A Very Early Archaic the State of New York. Albany. Site on the Mississippi River. Special Justice, N. D. Publication of the American Society for 1987. Stone Age Spear and Arrow Points of American Archaeology. The Amateur the Midcontinental United States. Indiana Archaeologist 8 (1–2). University Press. Bloomington. Gramly, R. M., and R. E. Funk Keeley, L. H. 1990. What Is Known and Not Known 1980. Experimental Determination of Stone about the Human Occupation of the Tool Uses. University of Chicago Press, Northeastern United States until 10,000 Chicago. B.P. Archaeology of Eastern North America Kopper, J. S., R. E. Funk, and L. A. Dumont 18:5–32. 1980. Additional Paleo-Indian and 1991. Olive Branch: A Large Dalton and Archaic Materials from the Dutchess Pre-Dalton Encampment at Thebes Gap, Quarry Cave Area, Orange County, New Alexander County, Illinois. In The Archaic

139 York. Archaeology of Eastern North America Ottawa. 8:125–137. Martin, P. S., and R. G. Klein, editors Kraft, H. C. 1989. Quaternary Extinctions: A Prehistoric 1977. The Paleo-Indian Sites at Port Revolution. University of Arizona Press. Mobil, Staten Island. In Current Tucson. Perspectives in Northeastern Archeology: McNett, C., and B. A. McMillan Essays in Honor of William A. Ritchie, edit- 1974. The First Season of the Upper Delaware ed by R. E. Funk and C. F. Hayes III, pp. Valley Early Man Project. Department of 1–19. Researches and Transactions of the Anthropology, the American University. New York State Archaeological Washington, D.C.. Association 17 (1). Rochester and Albany. 1977. The Shawnee-Minisink Site. In Kuhn, R. D. and W. A. Lanford Amerinds and Their Paleoenvironments in 1987. Sourcing Hudson Valley Cherts Northeastern North America, edited by W. from Trace Element Analysis. Man in the S. Newman and B. Salwen. Annals of the Northeast 34:57–69. New York Academy of Sciences 288:282–296. La Fleur, R. Moeller, R. 1965. Glacial Geology of the Troy, New York 1980. 6LF21: A Paleo-Indian Site in Western Quadrangle. New York State Museum and Connecticut. Occasional Paper 2. American Science Service, Map and Chart Series 7. Indian Archaeological Institute. The University of the State of New York. Washington, Connecticut. Albany. Newman, W. S., D. H. Thurber, H. S. Zeiss, A. La Porta, P. Rokarch, and L. Musich 1996. Lithostratigraphy as a Predictive 1968. Late Quaternary Geology of the Tool for Prehistoric Quarry Sites, Orange Hudson River Estuary: A Preliminary County, New York. In A Golden Report. New York Academy of Sciences, Chronograph for Robert E. Funk, edited by Transactions, series 2, 31 (5):548–570. C. Lindner and E. V. Curtin, pp. 73–83. Parker, A. C. Occasional Publications in Northeastern Anthropology 15. Bethlehem, 1924. The Great Algonkin Flint Mines at Connecticut. Coxsackie. Researches and Transactions, New York State Archeological Laub, R. S., N. G. Miller, and D. W. Association. Rochester. Steadman, editors Ritchie, W. A. 1988. Late Pleistocene and Early Holocene Paleoecology and Archeology of the Eastern 1953. A Probable Paleo-Indian Site in Great Lakes Region. Bulletin of the Buffalo Vermont. American Antiquity 18 Society of Natural Sciences 33. (3):249–258. Lavin, L. and D. R. Prothero 1957. Traces of Early Man in the Northeast. New York State Museum and Science 1981. Microscopic Analysis of Cherts Bulletin 358. The University of the State within and Adjacent to the Delaware of New York. Albany. River Watershed. Man in the Northeast 21:3–17. 1965. The Archaeology of New York State. Natural History Press, New York. MacDonald, G. F. 1971. A Typology and Nomenclature for New 1968. Debert: A Paleo-Indian Site in Central York Projectile Points. Revised edition. Nova Scotia. National Museums of New York State Museum and Science Canada, Anthropological Papers 16.

140 Service Bulletin 384. The University of the from New England Paleoindian Sites. State of New York. Albany. North American Archaeologist 6 (2):145–159. Ritchie, W. A., and R. E. Funk Spiess, A. E., and D. B. Wilson 1971. Evidence for Early Archaic 1987. Michaud: A Paleoindian Site in the Occupations on Staten Island. New England-Maritimes Region. Occasional Pennsylvania Archaeologist 41 (3):45-59. Publications in Maine Archaeology, 1973. Aboriginal Settlement Patterns in the Number 6. The Maine Historic Northeast. New York State Museum and Preservation Commission and the Maine Science Service Memoir 20. The Archaeological Society, Inc. Augusta. University of the State of New York. Steadman, D. W., T. W. Stafford, and R. E. Albany. Funk Ruedemann, R., Cook, J.H., and D.H. 1997. Nonassociation of Paleoindians Newland with AMS-Dated Late Pleistocene 1942. Geology of the Catskill and Kaaterskill Mammals from the Dutchess Quarry Quadrangles. New York State Museum Caves, New York. Quaternary Research Bulletin 331. The University of the State 47:105–116. of New York. Albany Tankersley, K. B., S. Vanderlaan, J. D. Schambach, F. Holland, and S. Bland. 1962. An Archeological Survey of the 1997. Geochronology of the Arc Site: A Hudson North 7.5 Minute Quadrangle, Paleo-Indian Habitation in the Great New York. Senior Project, Bard College, Lakes Region. Archaeology of Eastern North Annandale-on-Hudson. Albany. America 25:31–44. Semenov, S. A. Tringham, R., G. Cooper, G. Odell, B. Voytek, and A. Whitman 1964. Prehistoric Technology. Barnes and Noble, New York. 1974. Experimentation in the Formation of Edge Damage: A New Approach to Shutler, R., Jr., editor Lithic Analysis. Journal of Field Archaeology 1983. Early Man in the New World. Sage 1:171–196. Press, Beverly Hills, California. Udell, H. F. Sirkin, L. A. 1962. Oysters—A Delicacy from the 1965. Late Pleistocene Pollen Stratigraphy of Cradle of the Sea. New York State Western Long Island and Staten Island, New Conservationist 17 (2):9–11. York. Journal of the Institute of Science Weinman, P. L., and T. P. Weinman and Mathematics, Adelphi University 15. 1969. A Snook Kill Workshop. Smith, R. H. Pennsylvania Archaeologist 39 (1–4):23–28. 1954. A History and Definition of Game 1977. The Bittersweet Sites I, II, and III Ranges in New York State. Division of Fish (Cox 41, 42, 44), Components of the and Game, The New York State Susquehanna Tradition. The Bulletin of the Conservation Department. Albany. New York State Archaeological Association Smith, L. 71:19–28. 1985. The Inland Fishes of New York State. 1978. Kings Road Site: Further Evidence New York State Department of from this Paleo-Indian Component. New Environmental Conservation. Albany. York State Archaeological Association Spiess, A. E., M. L. Curran, and J. R. Grimes Bulletin 72:1–5. 1985. Caribou (Rangifer tarandus) Bones

141 Weiss, D. Whitney, T. 1971. Late Pleistocene Stratigraphy and 1974. The Indian Occupations of Paleoecology of the Lower Hudson River Chenango County to 1975. Chenango Estuary. Geological Society of America Chapter, New York State Archeological Bulletin 58:1561–1570. Association Bulletin 15 (3). Wellman, B. Wilmsen, E. N. 1982. A Survey of New York Fluted 1968. Functional Analysis of Flaked Stone Points. Archaeology of Eastern North Artifacts. American Antiquity 32:383–388. America 10:39–40. Witthoft, J. West, F. H. 1952. A Paleo-Indian Site in Eastern 1983. The Antiquity of Man in America. Pennsylvania: An Early Hunting Culture. In Late-Quarternary Environments of the Proceedings of the American United States, edited by H. E. Wright, Jr.; Philosophical Society 96 (4):464-495. vol.1: The Late Pleistocene, edited by S. C. Wright, J. V. Porter, pp. 364–382. University of 1972. The Shield Archaic. National Minnesota Press, Minneapolis. Museum of Canada, Museum of Man, 1996. The Archaeological Evidence. In Publications in Archaeology No. 3. American Beginnings, edited by F. H. West, Ottawa. pp. 537-559. University of Chicago Press, Chicago.

93-102PM

142