This Page has been updated! If you are a return visitor you will need to reload/refresh your browser in order to see any new content.
THE AMATEUR ARCHAEOLOGIST ONLINE
The Vail Habitation and Kill Site: Implications for Palaeo-American Behavior and Band Size
By Richard Michael Gramly PhD
Seven conjoined fluted points linking the Vail Palaeo-American encampment and presumed kill site in the Magalloway Valley, northwestern Maine highlands are evidence that six tent-sites were occupied seasonally by a band of hunters who likely pursued caribou. After nine consecutive occupations the herd was nearly annihilated and the Vail band departed from the Magalloway Valley, perhaps with the expectation of returning after the herd had regenerated.
Thirteen flaked stone artifacts have been recovered from Vail kill site #1. All are projectile points, as follows: (see accompanying photographs and line drawings)
Artifacts a-l pertain to the early Palaeo-American Vail site and its habitation loci; while, artifact m, perhaps 2,000 years later in age (Petersen et al. 2000), was left by a Palaeo-American hunter of the Lanceolate Point Tradition who briefly occupied a knoll immediately north of the Vail encampment (aerial photo, Figure 2).
The number of repeatedly occupied loci at the Vail site is six, which are arranged along an arc with a cross measurement of 100-110 m. The question may be posed: Were all six inhabited during every visit by Palaeo-Americans to the Vail site? Early in our analysis of the Vail site assemblage, when conjoined fragments of fluted drills were found to link widely separated habitation loci, we argued that at least two tents stood during visits (Gramly 1982, 1985). Not until 2001, however, did evidence come to hand supporting the idea that all six loci were occupied simultaneously.
In that year two medial fragments of fluted projectile points (V.13026 and V. 13027) were discovered at Vail kill site #1 by Maine resident, John Halunen (photograph, Figure 20) who boldly ventured out upon the exposed and unusually snow-free bed of Aziscohos Lake during mid-winter. It was possible to conjoin these medial sections to basal fragments that we had recovered during 1980 excavations at the habitation site. A basal fragment from Locus F fitted V.13026; while, V.13027 fitted an "ear" that had been catalogued from Locus E (Figure 6J and L) and another ear from the site's surface found by a sportsman.
These two refitted fluted points were welcome additions to a group of five others linking kill site #1 with the Vail habitations. Found in the 1970s and 1980s, these other five fragments from the kill site (V.5583, V.5582, V.5585, V.9034, and V.5586) matched pieces from Loci A, D, B, B, and C respectively (see Figures 5 and 6). Few Palaeo-American sites have been cross-linked by conjoined artifacts (but see the Murray Springs Clovis locality, Haynes and Huckell 2007), and no pair of sites has more conjoined artifacts linking them than the Vail encampment and its kill site.
All six repeatedly occupied tent loci are represented among the group of seven conjoined points linking the habitations and kill site. Tent locus B is represented twice. Such a complete spatial linkage for so small an artifact sample implies parity of opportunity among the various households of Vail hunters. More specifically, we surmise that 1) households of hunters damaged or wore out their equipment at the same rate, 2) the skill levels of the Vail hunters were compatible, and 3) every household had equal and unrestricted access to the killing ground upwind of the encampment. Going another step, we might argue that if the number of active hunters living in each tent were nearly the same, then all six tents were occupied at the same time.
The size of the band of fluted point-using hunters at the Vail site may be reckoned as the number of people who could be accommodated within six tents. Our excavation of a well-preserved tent locus at the Adkins fluted point site (Gramly 1988), just a kilometer south of Vail, provides an estimate of the number of Palaeo-Americans who might have been accommodated within a tent, which is 6-10 (depending upon their age). Thus, the Vail site's seasonal population might have been 36-60 persons. To this figure, one presumes, must be added an unknown number of dogs.
A patri-lineal and patri-local group of 36-60 individuals with a modal size of 48 conforms very well to Williams' hypothesized exogamous lineage-band (1974: 29). The Vail band would have been large enough to furnish marriage partners to neighboring Palaeo-American bands and, predictably, to have sustained new and old generations. In a group of this size cultural and technological traditions could have been perpetuated seamlessly.
Evidence of Band Size at Other Northeastern Palaeo-American Sites
Important to note, the number of habitation loci at Wheeler Dam, like Vail, is six.
Farther afield in eastern Massachusetts, near Ipswich, we note the Bull Brook II encampment, which was occupied by users of fluted projectile points. Its six (possibly seven) loci each produced roughly equal numbers of tools /tool fragments representing a wide range of types (Grimes et al. 1984). The companion, but much larger, Bull Brook I site also had six habitation loci, which were positioned along an arc. The tool inventories of these six loci are dominated by bifaces. This core group of habitations was surrounded by 30 loci where endscrapers were more abundant (Robinson et al. 2009). This hunters' encampment or village appears to have been neatly laid out. Spatial organization was abetted by a wide expanse of well-drained ground free of physical obstructions. The Vail encampment, by contrast, was constrained by boulders, bedrock, a small stream channel on its northern margin, and the channel of the ancient Magalloway River itself. Nonetheless, it is remarkable that the cross-measurements of the arc of Vail habitation loci and those of Bull Brook I are nearly identical (see Figure 21) — a clear indicator of nearly identical populations.
An equivalent size for the Vail and the core Bull Brook I encampments is expected given what we know about the habits of congregated hunters-and-gatherers in the modern day. Among the Tiwi of northern Australia, for example, households of hunters keep their distance (20-30 yards apart) when camped together (Hart and Pilling 1960: 36). The distance is not too great among households, however, to prevent cross-camp conversations at night. Other spatially well-defined encampments are on record for South African and Arctic hunters.
Another Palaeo-American encampment in New England that may have been inhabited by six families of fluted point-users is the Turners Falls Airport site (also known as the Hannemann site), Connecticut River valley, Massachusetts, Test-trenching revealed four regularly spaced loci, and according to excavators it is likely that additional loci will be found (Binzen 2005).
Palaeo-American encampments where all families of a band convened in order to exploit abundant, seasonally available resources appear to be more rare than small sites with evidence of only one or two tents. Much of the year it may have been necessary for a band to remain dispersed across the region — reduced to single families or pairs of families — re-uniting only when animals aggregated. Where caribou was the quarry, their movement towards calving grounds and winter quarters would have offered hunters the best chance for a high return (Gordon 1975).
The Caribou Model and Overkill
Dividing by six — the number of tents that we believe stood at the Vail site during every seasonal occupation — we arrive at nine seasons of occupation. This figure is also corroborated by the sum of occupational episodes for each locus, calculated separately. It is interesting to note that a timespan of nine years is nearly identical to the length of stay (10 years) within a residential core area, as reported for Nunamiut hunters of Alaska (Binford 1983: 114-115). According to Binford, both firewood and animals become depleted after 10 years. The band must move its residence to a district where resources have been allowed to re-generate for 30 years or more.
Hunting must have been very good and the bag of animals dependable for the Vail band to have returned nine times to the Upper Magalloway River valley. The organization of the encampment into an arc of well-spaced tent-loci with the boundary of each habitation area sharply defined (see Gramly 1982 for plots of artifacts within individual loci) indicates that all visits must have occurred within a brief period — likely nine consecutive years.
The question is why this Palaeo-American band abandoned the Vail site, never to return?
Assuming that caribou was the intended quarry of the aggregated Vail band, the valley herd must have been depleted severely by well-armed, skilled hunters during nine consecutive seasons. Overkill, that is to say, hunting in excess of a species' ability to regenerate, may have become a problem. The idea of "Pleistocene overkill" has pervaded archaeological literature since the 1960s when it was presented by Dr. Paul Martin (see, for example, Martin 1963: 70; Martin 1967). His hypothesis has been difficult to test; however, in the case of the Vail site with its linked kill and habitation sites we have evidence that bears upon the issue.
Ethnologists have argued that Eskimo and Indian families with dogs who are heavily dependant upon caribou require an animal per day - 365 per year (e.g., Mallet 1930, Harrington 1952: 125, Burch 1972: 362). If fewer animals are bagged, the deficit must be made up by other game and fish. In addition to meat and fat, caribou in their prime furnish hides, antler and sinews. Without a steady supply of these raw materials, life in an arctic world would be impractical and risky.
For Vail site families lying in wait to ambush passing game, a minimum of six caribou per day was needed for basic nourishment — but setting aside nothing against times of shortage. If the Vail encampment were occupied when caribou were on the move but prior to the cold season and the rut, occupation could have lasted 2-3 months. The duration of the hunt depended, in part, upon constancy of a prevailing wind out of the northwest. Given the uncertainties of wind, weather, and caribou behavior itself, each family at the Vail site may have expected to consume at least 75 animals while waiting in camp. It is reasonable to believe that not fewer than 450 animals were required to sustain the entire band during its annual harvest of the Magalloway herd.
If the opportunity presented itself, the Vail band might have killed enough caribou to satisfy their needs for an entire year. Transporting such a large bag would have posed a problem; some surplus might have been deposited within cache-pits that we have observed at the Vail and Adkins sites (Gramly 2009: Chapter 4 photos). Additional butchered remains might have been left at depots en route to the band's winter quarters. Given the challenges of life in a New World, hunters may have been compelled to maximize their harvest as insurance against unforeseen want. Seen in this light, overkill was a sound strategy as long as new lands with fresh herds remained to be exploited. With no need to conserve, the Vail band might have killed as many as 2,200 animals during each visit to the site, netting them a staggering 200-250 tons of flesh and bone.
Obviously such a large bag could have been attained only during the first years of preying upon the Magalloway Valley caribou herd. After initial heavy slaughter, caribou would have become wary. More critical was the loss of breeding stock. Inevitably, returns to hunters diminished, and during the final (ninth) year of hunting, the harvest might have been small. Such a "boom and bust" cycle must have been familiar to Palaeo-Americans, who anticipated future shortages by reconnoitering another killing ground well in advance of the Magalloway valley herd's (inevitable) annihilation. At the Vail site we may have evidence of searching for a new place to hunt in the form of a few flaked tools made of raw materials that outcrop 200 km away in north-central Maine. Apparently, during a pioneering foray into uninhabited lands, a new lithic source was found.
The number of caribou killed annually by the Vail band lay within the range 450-2,250 animals, averaging 1,350 of any sex or age. In order to lay out so many butchered carcasses a large piece of ground along the Magalloway River would have been required. The nearly culturally barren swath of sandy soil lying shoreward of the arc of Palaeo-American tent-sites A-F (see Vail site map, Figure 3) would have been convenient for this purpose. Observers of caribou-hunting Indians and Eskimo in Labrador during the nineteenth century reported that well over 1,200 animals were harvested each year at lake and river crossings. The shorelines for hundreds of feet were covered with butchered remains (Turner 1894; Grenfell 1909: 210-11; Cabot 1920: 242-3).
At the averaged rate of 1,350 caribou per year, a total of over 12,000 animals would have been slain and processed at the Vail site during the nine years it was occupied. Although estimates of the number of head in a caribou herd vary widely for different regions, herds of 10,000-15,000 seem to be a norm — at least for Alaska (Spiess 1979). The Vail hunters undoubtedly were capable of destroying a herd of that size and of exacting a terrible a toll from larger herds. If slaughter were followed by a time of profound environmental change, the Magalloway Valley herd, whatever its size, may have needed decades, perhaps centuries, to recover. In this light it is worth noting that the Vail site was occupied at most 100-200 years after the onset of the Younger Dryas (reckoned as 12,900 calendar years BP), when vegetation was still responding to a cooling climate (Newby et al. 2005: 142). Although good fodder for caribou may have been freely available in northern New England and the Canadian Maritimes — as a whole — during the Younger Dryas period (12,900-11,600 calendar years BP), localized, temporary disruptions might have occurred.
Years after the occupation of the Vail site, when a band of fluted point-using Palaeo-American hunters re-entered the upper Magalloway Valley, they must have expected to find a re-generated caribou herd and good hunting. The fact that the six families occupying tents at the Wheeler Dam site north of Vail stayed for only one season is proof of their unfulfilled expectations. It seems to indicate that the Magalloway herd had failed to regenerate in the aftermath of the heavy slaughter at the Vail site — perhaps as a result of a diminished food supply due to a period of unfavorable climate?
An idea of what might be found within the ancient channel of the Magalloway River is provided by the Meiendorf and Stellmoor sites of northern Germany, which belong to Late Palaeolithic or epi-Pleistocene Ahrensburgian and Hamburgian archaeological cultures (Rust 1937, 1943; Clark 1975: 87-93). There, as at Vail, reindeer were ambushed during their seasonal moves, and bone and antler representing hundreds of individuals were preserved because of deep burial under waterlogged conditions. Among the jumbles of reindeer remains were perishable artifacts, flaked stone tools (some still in hafts), and art objects. In light of these remarkable discoveries and the likelihood that similar finds might one day be made at Vail, future excavators of that site are reminded of Sir Mortimer Wheeler's (Euripedes') admonishment to: "Leave no stone unturned."
The author is indebted to several colleagues who have commented upon early versions of this paper. Also, I wish to acknowledge four anonymous reviewers for the journal American Antiquity — some of whom made constructive criticisms that were incorporated here. The basic interpretation of a train of discoveries made during 31 years, however, rests solely with me.
Binford, Lewis R.
Binzen, Timothy L.
Burch, Ernest S., Jr.
Cabot, William Brooks
Gordon, Bryan H. C.
Gramly, Richard Michael
Grenfell, Wilfred T.
Grimes, John R., William Eldridge, Beth G. Grimes, Antonio Vaccaro, Frank Vaccaro, Joseph Vaccaro, Nicolas Vaccaro, and Antonio Orsini
Hart, C. W. M. And Arnold R. Pilling
Haynes, C. Vance, Jr. and Bruce B. Huckell (editors)
Martin, Paul S.
Newby, Paige, James Bradley, Arthur Spiess, Bryan Shurman and Phillip Leduc
Petersen, James B., Robert N. Bartone, and Belinda J. Cox
Robinson, Brian S., Jennifer C. Ort, William A. Eldridge, Arthur I. Burke, and Bertrand G. Pelletier
Spiess, Arthur E.
Williams, B. J.
Captions and Figures
The Amateur Archaeologist welcomes articles and communications from members; please note that authors waive any claim to copyright if they submit materials for publishing in the Society's journal.
Advertisers are welcome, and they need not be members of the ASAA. Please write for advertising rate cards and specifications. The Society does not furnish its membership list to anyone or any organization for promotional mailings.