EXCAVATIONS AT THE AMITY SITE: FINAL REPORT
THE AMITY SITE
After the discovery of the Amity Site in 1985, there was considerable optimism among both the local and professional communities that John White’s Pomeiooc had been located. The location of the site (Figure 10) was entirely consistent with Pomeiooc’s placement on the sixteenth century maps, and the artifact assemblage of simple-stamped pottery and rouletted pipes was suggestive of a historic period occupation. Archaeological excavation was organized, therefore, to ascertain if this optimism was justified. The primary goals of the fieldwork were the recovery of temporally diagnostic artifacts and the exposure of sufficient area to determine if the site matched the presumed archaeological signature of Pomeiooc.
Fieldwork at the Amity Site spanned five years, four field seasons and two principal investigators (Figure 11). Paul Green served as principal investigator in 1985 and 1986, assisted by Jay Holley. The author served as principal investigator in 1988 and 1989, assisted by David Jones. All of these personnel were affiliated with the Department of Sociology and Anthropology, East Carolina University. Discussion of the 1985 and 1986 seasons is based largely on Green (1987). Description of features follows that of the fieldwork.
1985 Fieldwork: The Amity site spans two separate property tracts farmed by different tenants. When first located in the summer of 1985 the northern portion of the site was in winter wheat and the southern portion in corn. The initial tests of the site focused on the northern portion following the harvest of the wheat (Figure 12). This portion of the site proved to have much lower density of artifacts than did the cornfield, and produced a substantial amount of Middle Woodland period Mount Pleasant pottery. One arc of widely spaced postmolds was located, but Contact period material was sparse. In addition, a shovel-dug dog burial (Feature 10) and a fowl burial (Feature 11) revealed that the area had served some time in the not-too-distant past as a pet cemetery. This fact, along with the paucity of Contact period material, induced Green to shift the excavations to the southern portion of the site.
During the period of July 8 to 24, 1985, Paul Green, Jay Holley and several volunteers from the Office of State Archaeology and Friends of North Carolina Archaeology, Inc. conducted an excavation in the southern portion of the site. During the first week, eight two-meter squares were dug and plowzone screened through 1/4 inch mesh. A substantial amount of Colington pottery and pipe fragments was recovered and three features -- a midden lens, a ditch and a third animal burial (a cat) -- were revealed.
To expedite the excavation, a tractor with a back-blade was used to clear the plowzone from three trenches approximately 60 meters long and two meters wide (Figure 13). Trench I, located on the highest part of the lake ridge, uncovered numerous postmolds including structural patterns. Trench II, located about 10 meters west of Trench I, uncovered another structure pattern and a large area of preserved midden (Feature 8). Trench III, located about 10 meters west of Trench II, was largely devoid of cultural features except for a concentration of Colington Simplestamped pottery (Feature 7) near its southern terminus. The remainder of the 1985 season, July 16 to 24, was spent flatshoveling and mapping these trenches.
[Page 27] 1986 Fieldwork: The 1986 field season spanned March 15 to July 18. The field crew consisted of Paul Green and one to two assistants. In order to achieve results of unmistakable interest to the public, it was decided to remove plowzone from a large portion of the site in expectation of revealing a close match between structural patterns and the John White painting. Again a tractor with back-blade was employed to remove plowzone. This time an area of about 30 meters by 24 meters overlying the two structure patterns was opened (Figure 14). Plowzone was not screened.
During the 1986 season the primary emphasis was on flatshoveling, photographing and plotting this area, but some features were excavated. In particular, Features 6 and 16 were excavated and determined to be agricultural drainage ditches not associated with the Colington occupation.
Quite importantly, the blading of the site revealed not only the two anticipated longhouse patterns, but also a series of postmolds defining a segment of palisade (Feature 25). Furthermore, radiocarbon dating of two midden lenses, Features 2 and 8, produced ages of 160 ± 50: uncalibrated A.D. 1790 (Beta-17508) on oyster shell and 450 ± 100: uncalibrated A.D.1500 (Beta-17507) on wood charcoal. Thus at the end of the 1986 field season, Green had produced evidence of a palisaded village of longhouses associated with a radiocarbon date that incorporated the Fort Raleigh period within the one sigma range. Obviously, the Amity Site was a promising candidate for Pomieooc. All the necessary-but-not-sufficient criteria had been met: the general location, date, and site structure of longhouses and palisade were correct. On the other hand, one radiocarbon date was clearly too young for Pomeiooc, artifact density was unexpectedly low compared to other Late Woodland period villages (Green 1987:22), no clearly sixteenth century artifacts were recovered, and some artifacts, particularly glass projectile points, seemed more suggestive of a seventeenth century occupation (Green 1987:25). Nevertheless, enthusiasm among the local people and the involved archaeologists remained high.
[Page 30] 1988 Fieldwork: Funding constraints precluded fieldwork in 1987. In 1988, however, the support of the local public, the Friends of North Carolina Archaeology, Inc. and the public relations efforts of the Office of State Archaeology resulted in a special appropriation from the North Carolina General Assembly funding further archaeology at the Amity Site. This appropriation was made possible by State Representative Howard Chapin. Paul Gardner was hired by East Carolina University to serve as Principal Investigator for the project, Paul Green having accepted employment elsewhere.
Meanwhile the Office of State Archaeology had entered into an agreement with the Operation Raleigh American Expedition, an international organization providing scientific and community service opportunities for young people from around the world. As a result of this agreement, Operation Raleigh "venturers" provided a labor force for the Amity Site excavation from July 25 to September 30,1988. Crew size varied from four to 19 people. During the field season of 48 working days, 52 venturers participated in the excavation. Turnover of personnel was exceptionally high. Individuals’ tenures on the crew ranged from one to 20 days. Ages of the venturers ranged from 17 to mid-twenties, and a great range of personal attitudes and abilities were encountered. The crew included at various times people with auditory, emotional and learning impairments, as well as people with markedly disparate degrees of English competency. Overall they performed well under adverse conditions (Figure 15).
The goals of the 1988 fieldwork were straightforward: to establish the temporal context of the site and to determine its extent. Establishing temporal context was to be undertaken by excavation of features and postmolds with the expectation of recovering temporally diagnostic artifacts and dateable charcoal. This proved to be easier in concept than in practice. Only one additional Colington feature was uncovered during the 1988 season, although a further extent of the palisade, and an additional expanse of Feature 8 were revealed (Figure 16). These produced no artifact classes not already recovered during the previous seasons except for glass seed beads from Feature 8. In addition to the Colington features, one Mount Pleasant sherd concentration was recovered and seven tree stains were partially excavated. Except for stains that were quickly determined to be noncultural, i.e. trees, all feature fill was waterscreened in the [Page 32] field and a soil sample saved for laboratory flotation. Waterscreening was made possible by the laying of a waterline from the county water system to the site.
By both design and default, considerable attention was paid to the postmolds within the 1986 tractor-cleared area, (hereafter referred to as the main excavation unit [MEU]). For each square containing part of a structure pattern, each postmold was individually dug with spoons and its fill waterscreened through a kitchen strainer with mesh size of approximately 1.3 millimeters. This facilitated the recovery of small items, particularly trade beads, and with the fine silt soils of the site proceeded rather rapidly. All artifacts retained in the strainer were curated by individual postmold number, while charcoal and bone were aggregated from all postmolds within a square.
Unfortunately the site was not backfilled after the 1986 season, and winds stripped it of the weighted black plastic sheeting meant to protect the excavated units. Thus, when the 1988 season began in midsummer, the site had a two-years growth of weed cover growing from the unprotected units and had experienced two winters of frost-heaving. As a result many of the postmolds recorded by Green in 1986 could no longer be located in 1988 (cf. Figure 14 and Figure 16). In addition, Green’s site datum was plowed away during the hiatus and could not be relocated with precision. These factors necessitated a complete re-photographing and re-plotting of the MEU concomitant with the excavation of postmolds.
Determination of the size of the site and its internal organization was addressed by completing excavation, photography and plotting of squares within the MEU not undertaken by Green, by extending the excavation of the MEU to the east, and by excavating additional units within the area of surface artifactual scatter. These squares failed to locate any cultural features or postmold patterns, and it fact did not produce many convincing postmolds. Many of the outlying postmolds indicated on the site maps are likely to be rodent or root disturbances, but time constraints precluded their sectioning and excavation.
At the end of the 1988 season a sufficient extent of the palisade had been uncovered to determine that, if circular, its diameter would be only about 16 meters. Not only is this insufficient to enclose 18 longhouses but is, in fact, too small to enclose even the two longhouses uncovered in 1986. Obviously Feature 25 could not be a circular palisade seen by John White. Furthermore, although an additional amorphous expanse of postmolds was revealed to the east of Structure 2, excavation away from the MEU produced no evidence of additional structures. The Amity Site thus appears to represent a small hamlet or farmstead rather than a substantial village. Finally, the Amity Site artifact assemblage contained no jetons or "fancy" beads, but did include glass projectile points, rouletted pipes, and seed beads. These artifacts compare most favorably to seventeenth century assemblages (See Chapter 4 below). In short, by the end of 1988 there seemed little basis for considering the Amity Site to be Pomeiooc.
1989 Season: From March 13 to May 17, 1989 Gardner and Jones, assisted occasionally by one or two volunteers, conducted limited excavations south of the MEU and in the vicinity of Feature 8 to recover more artifactual and ecofactual material. Also, two narrow trenches east of Trench I and west of Trench III were opened by mechanical stripping of plowzone. These trenches essentially "boxed in" the excavation area of 1986-88 and provided further evidence as to the spatial extent of the Site.
No features or postmold patterns were encountered, and the two bladed trenches were devoid of evidence of cultural activity except for some possible postmolds in square 116R180. Excavation of contiguous squares failed to reveal any convincing evidence of cultural activity. These results strengthened the opinion among professionals that Amity was not Pomeiooc, but in order to make the inference undeniable, a tractor with back-blade was used to clear plowzone [Page 34] from an area approximately 80 meters X 7 meters in size west of the MEU. Thirty-eight two-meter squares in this area were flatshoveled, towelled, photographed and plotted (Figure 17). Again, the work revealed no postmold patterns (and few convincing postmolds) and no cultural features except for an extension of the drainage ditch, Feature 6. Rather than continue work in this area, excavation shifted to the area east and southeast of the MEU. Squares 176R220 to 176R236 were excavated with shovels and screens, as were 18 squares south of 160R228. Results were no better: artifactual recovery was very low (c. Appendix B) and no cultural features were found. In spite of assiduous attempts to determine otherwise, no evidence supporting the inference that the Amity site was a substantial village could be found. The excavations ceased on June 30, 1989.
The fine silt soils of the Amity Site were greatly mottled and displayed numerous large stains beneath the plowzone (Figure 18). Thirty-three stains were assigned feature numbers. These include 13 tree stains, three modem animal burials, two drainage ditches, tour sherd concentrations, four empty pits, four midden lenses, and three structure patterns. Feature numbers 1 to 25 were assigned in 1985 and 1986; feature numbers 40 to 48 were assigned in 1988 and 1989. Numbers 5 and 26 to 39 were not assigned.
Tree Stains: (Features 9,12-15, 41-45, 47) Mottling of subsoil by vegetation was very pronounced at Amity with virtually every square displaying one or more dark amorphous stains. In the initial excavations in 1985 and 1986 these were assigned feature numbers, but not excavated. (These are not indicated on the site maps.) In 1988 and 1989, only the stains with regular outlines suggestive of cultural origins were assigned numbers. These features were excavated by removing with small tools the fill from one-half of the feature. If the walls of the feature were sloping and the floor highly irregular, sharply tapering, or deeper than the water-table, the feature was pronounced a tree stain and its excavation terminated (Figure 19).
Animal Burials: (Features 3,10,11). Three features encountered during the 1985 and 1986 seasons proved to be recent animal burials. Feature 3 is a shovel-dug dog burial. Feature 10 was a burial of a fowl, and Feature 11, that of a cat (Figure 20). The latter two were accompanied by lead shot. All three are considered unassociated with the aboriginal occupation.
Empty Pits: (Features 17, 18, 19, 40). Four pits have regular, essentially circular outlines and approximately flat bottoms (Figure 21); thus they mimic humanly dug pits. However, waterscreening of their fill revealed that they contain no artifacts. Underground storage would be impractical in the poorly drained soils of the site, so it is unlikely they served as storage pits. With depths of approximately two feet below present ground surface, they seem too shallow to be either wells or human burial pits but could be animal burials from which all bone has decayed. These pits remain enigmatic and may well be natural phenomena.
Ditches: (Features 6, 16). Feature 6 is by far the most prominent feature at the site, extending completely across the main excavation unit and into the western trench as well. When first uncovered, it was anticipated to be an shallow trash-filled aboriginal borrow pit such as those identified at the Patawomeke (Schmitt 1965), Warren Wilson (Dickens 1976), and Bessemer (Whyte and Thompson 1989) sites. Excavation, however, revealed that it extended about one meter beneath the present ground surface and to have had a sharply trigonous profile (Figure 22). Its fill contained a sparse mixture of aboriginal artifacts along with historic wbbish. Thus it seems too deep, too regular in cross section and too sparsely filled with aboriginal artifacts to be a borrow pit. On the other hand, the feature is too narrow and shallow to be a defensive moat such as occurred at the King site (Hally 1988). Furthermore, rather than paralleling the palisade line as a borrow pit or moat would be expected to do, Feature 6 intrudes and obliterates it. Feature 6 [Page 37] seems most likely to be a agricultural drainage ditch (Green 1987), although no local informant can recall a ditch in this location.
Feature 16 is likewise interpreted as a recent drainage ditch. It extends only a few centimeters into subsoil and clearly displays the imprints of shovel tips along its floor (Figure 23).
Potsherd Concentrations (Features 7, 20, 21, 48): Three concentrations of potsherds were recovered during the 1985 and 1986 seasons and one during the 1989 season. Feature 7, consisting of 141 Colington Simple-stamped sherds is the largest of these and the only one associated with the Colington occupation of the site. Feature 7 is located in Green’s tractor trench III at approximate co-ordinates of 132R188. No pit outlines or other midden debris were observed. Excavation of nearby squares in 1988 produced a sizable number of very small Colington ware sherds from the plowzone, but revealed no evidence of cultural features. Feature 20, located in the MEU, consisted of 24 Mount Pleasant ware sherds from at least two vessels, one net-impressed and one cord-marked. Feature 21, also located in the MEU, consisted of five Mount Pleasant ware sherds and nine Colington sherds. Feature 48, (Figure 24) a cluster of 113 Mount Pleasant sherds, is located in 140R200 and underlies Feature 8, a midden lens containing primarily Colington Ware pottery. Feature 48 includes both Mount Pleasant Net-impressed and Mount Pleasant Cord-marked sherds that are for the most part very poorly fired. They possess prominent reduction cores and are quite soft and friable.
Midden lenses (Features 1, 2, 4, 8, 46): Although the stratification of the Amity Site was almost exclusively plowzone over sterile subsoil, five remnants of cultural midden were encountered. Midden was recognizable by dark brown to black soil containing bone, shell or artifactual debris. Feature 1 contained only one fragment of unidentifiable bone and two fragments of oyster shell with no associated artifacts. Feature 2 was an compact area of darker midden soil that intruded Feature 4, a more diffuse area of midden or old humus. Feature 2 contained 34 Colington Ware sherds and two of Mount Pleasant Ware. Faunal remains recovered included oyster shell, and fish, turtle, opossum, bear and deer bone. A radiocarbon date on oyster shell from this feature indicates an age of 160 ± 100 years: uncalibrated A.D. 1790 ± 100 (Beta-17508). This date is uncorrected for carbon isotope fractionation or reservoir effect. Feature 4 contained no bone or shell and only 12 potsherds, only one of which was Colington Ware, the remainder being Middle Woodland Mount Pleasant and Hanover Wares. The fill from these three features, excavated in 1985 was removed with small tools and screened with 1/4 inch mesh. Flotation and waterscreening were not performed.
Feature 8, a dark stain containing bone, shell and pottery was revealed by the mechanical removal of plowzone from Trench II in 1985. Following tlatshoveling of the trench, the portion of the feature exposed in square 140R200 was removed with small tools and screened through 1/4 inch mesh. In 1988 contiguous squares 142R202, 140R202 and 138R202 were opened. An examination of the subsoil revealed no pit outline, nor was there any discernable change in soil coloration as one moved away from the Feature 8 area. However, a sparse scatter of potsherds, oyster shell, and bone fragments extended across the southern half of 140R202 (Figure 25) and into the northwestern quadrant of 138R202. After removal of plowzone and plowscars from 140R202 and 138R202, a 10 centimeter level of zone 2 was removed with trowels and waterscreened through windowscreen. Fifteen-liter flotation samples were taken from the southwestern quadrant of 140R202 and from the northwestern quadrant of 138R202. In addition, a dark stain remaining in 140R200 and presumed to be the base of the feature excavated by Green produced 10 liters of soil which were floated as well.
The pottery from Feature 8 includes 69 Colington Ware sherds and three sherds of Mount Pleasant Ware. Feature 8 and the surrounding area produced most of the botanical and faunal remains recovered from the site. Food plant remains include corn, beans, [Page 40] squash/pumpkin, hickory nut, walnut, acorns, little barley, maygrass, grape, and sumac. Faunal remains identified include deer, rabbit, opossum, turtle, and fish.
Three radiocarbon ages have been obtained from charcoal from this feature. Green obtained a radiocarbon age of 450 ± 100 years: uncalibrated A.D. 1500 ± 100 (Beta-17507) from charcoal from the 140R200 portion of the feature. This date is not corrected for carbon isotope fractionation. Two ages are available from the 140R202 portion of the feature: 450 ± 80 years: uncalibrated A.D. 1500 ± 80 (Beta-34062) and 210 ± 50 years: uncalibrated A.D. 1740 ± 50 (Beta-31110). These last two dates are corrected for carbon isotope fractionation (see Table 1 for uncorrected ages).
The two A.D. 1500 radiocarbon dates support a Fort Raleigh period temporal placement for the site, but seem too early for a coastal ceramic assemblage lacking fabric-impressed pottery (Egloff and Potter 1982; David Phelps, personal communication 1989). As three Middle Woodland period sherds were recovered from this zone, the possibility of contamination of the charcoal sample from the earlier component cannot be ruled out. In contrast, the eighteenth century radiocarbon date, although agreeing quite closely with the Feature 2 radiocarbon date, is relatively late in the period of the Mattamuskeet Reservation (A.D. 171 5-1755) and is too recent for an occupation displaying so few signs of acculturation. Calibration of the isotopically corrected dates suggests that the mid-seventeenth century is an appropriate date for the feature but does not preclude a sixteenth century placement. (see pp. 42-44 for discussion of radiocarbon dates).
Feature 46, located in square 192R190, was a small concentration of oyster shell, animal bone and a single Colington Simple-stamped sherd. Identified bone specimens included one of fish, nine of turtle, five of bird, and one of rabbit. No pit outline was visible, but the three liters of soil filling the interstices of the bone and shell were floated. To augment this small sample, another 15 liters of soil surrounding the feature were floated as well. Together the float samples produced only wood charcoal and a single grain of little barley.
Structures (Features 23, 24, 25): Two longhouse patterns and a portion of a palisade are present at the Amity Site. All three structures were located during the mechanical removal of plowzone in 1985 and exposed by further plowzone blading in 1986 except for the easternmost five meters of palisade which were uncovered by shoveling in 1988. The structures were mapped in 1986 but not excavated. In 1988 the structures were relocated and the postmolds defining them excavated and waterscreened. Unfortunately, a number of postmolds were lost due to the effects of weather on the unprotected site.
Structure 1 (Figure 26) was clearly defined in 1986 as a longhouse structure approximately 14 m long by 6.5 m wide (46 ft by 21 ft). No convincing interior support posts could be isolated, but lines of postmolds paralleling the longer walls of the structure may represent supports for benches or sleeping platforms (Green 1987).
Structure 2 (Figure 26) was less clearly defined even in 1986. Green (1987) suggests reasonably that it represents two overlapping building episodes of a house approximately 9 m by 6 m (30 ft by 20 ft) in size. Having a length - width ratio of only 1.5 to 1, the house seems more subrectangular than a true "longhouse".
Longhouses are quite prominently mentioned in the historical accounts concerning the North Carolina and Virginia Algonquians (Quinn 1955:370; Barbour 1986:170,173), but are surprisingly uncommon archaeologically. In North Carolina only two longhouses, both from the central coastal region, were known prior to the Amity excavations. At the Uniflite Site (310n33) a longhouse pattern measuring 13 m by 6 m (43 ft by 20 ft) was found (Loftfield 1979). At Permuda Island (310n196) a longhouse pattern 8 m by 4 m (26 ft by 13 ft) was uncovered (Loftfield 1985). The structures at neither the Uniflite or Permuda Island Site are convincingly dated, but are likely [Page 42] assignable to the earlier portion of the Colington Phase, circa A.D. 800 to A.D. 1400. In Virginia, two longhouse patterns were revealed at the Riding Ring Site (44Vb7 and 44Vb9). One measures 9 m by 5 m (30 ft by 16 ft) (Egghart 1986); the other, 12.5 m by 6.5 m (41 ft by 21 ft) (Egloff and Turner 1984). The houses are assigned to the pre-contact Late Woodland period (Keith Egloff, personal communication 1989). Finally, from the Jordan's Point Site (44Pg303) nine Late Woodland house patterns have been found, including longhouses measuring approximately 9 m by 5.5 m (30 ft by 18 ft) . These are associated with Gaston (viz. Cashie) Simple-stamped pottery and are thought to date, at least in part, to the early seventeenth century (Keith Egloff, personal communication 1989).
Feature 25 is the site palisade. It is divided into two approximately seven meter segments by Feature 6, a historic agricultural ditch, and disappears to the south and west. If circular its radius would only be about eight meters (26 ft), insufficient to enclose the two longhouses. Presumably the palisade, if ever completed, was not circular.
The postmolds comprising the palisade pattern average 8 cm to 10 cm in diameter and about 8 cm to 12 cm in depth below plowzone. In these respects they do not differ from the postmolds defining the longhouses, nor does the depth of plowing differ between the palisade and longhouse areas. Furthermore, the postmolds near the ends of the palisade segments are as deep as those from the middle of the segment; that is, the postmold pattern does not 'feather- out' but disappears abruptly. These facts, combined with the failure of the excavations to intercept any other palisade sections, suggest that the palisade may never have been completed.
All postmolds in the palisade were excavated with spoons, and their fill waterscreened through kitchen strainers except for those in square 186R214 which were flotated. Palisade postmolds produced small amounts of Colington Ware pottery, fragments of deer and unidentifiable bone, and carbonized remains of walnut, acorn, and corn and a grain of little barley.
Three grams of charcoal aggregated from all the palisade postmolds yielded a radiocarbon age of 630 ± 90 B. P.: uncalibrated A.D. 1320 ± 80 (Beta-30866). This date is corrected for carbon isotope fractionation, and the sample received four times the normal counting time due to its small size. The date is clearly too eady for the Post-Contact period component, and likewise too late for the Middle Woodland one. One or more postholes may have incorporated some charcoal from the earlier Middle Woodland period occupation into their fill, or it may be that some charcoal occurs naturally in the silty soils of the site. Certainly ash is a frequent component of soil profiles in the area (John Gagnon, Soil Conservation Service, Edenton Office, personal communication 1989). More speculatively, some of the palisade charcoal might have dehved from some ancient wood scavenged as fuel from Lake Mattamuskeet or a peat bog. For whatever reason, the date is not credible for either Amity component.
A palisade is, of course, a most prominent part of John White's painting of Pomeiooc, but prior to the Amity excavation such features were unknown archaeologically from coastal North Carolina. In coastal Virginia the terminal Late Woodland/early Contact period sites of Accokeek Creek (Stephenson et al. 1963) and Patawomeke (Schmidt 1965) produced evidence of multiple palisades, and the Great Neck site (44Vb7) (Egloff and Turner 1984) and the Native American component at Flowerdew Hundred (Keith Egloff, personal communication 1990) revealed a single palisade each.
Five radiocarbon dates are available for the Colington occupation of the Amity site (`Table 1). They range uncalibrated from A.D. 1320 to A.D. 1790. The most recent date (from Feature 2) is best ignored, as it is based on estuarine oyster shell that contains an unknown amount of carbon derived from marine rather than atmospheric sources. I suggest, as well, that [Page 43] the oldest date (from Feature 25, the palisade) is not associated with either the Postcontact or the Middle Woodland occupation of the site. The charcoal that generated it may include a mixture of charcoal from both components, be contaminated by naturally occurring charcoal or be derived in part from ancient firewood. As this was an very small sample (about three grams), it would have been especially susceptible to contamination of this sort. The three dates that remain are all derived from Feature 8, the midden lens in the southern excavation area.
It is now well established that contrary to the initial assumption underlying radiocarbon dating, the C-14 activity of the atmosphere has varied through time (Stuiver and Reimer 1986). Therefore, to convert radiocarbon dates to calender years a calibration curve must be used. Subtraction of a radiocarbon age from A.D. 1950, while cognitively simple, does not produce a valid calender year date (Long 1990:6). Recalibration requires dates corrected for isotopic fractionation, and unfortunately, the dates acquired in 1986 were not so corrected. The CALIB program (Stuiver and Reimer 1986) was used to calibrate the remaining three Amity radiocarbon dates (Table 2).
Examination of the calibrated dates shows that even at the two sigma range, the palisade date remains inexplicably early. At the one sigma range, the two dates from Feature 8 do not overlap, one indicating a seventeenth century or later occupation, the other a fifteen century occupation. At the two sigma ranges, the dates overlap from A.D. 1528 to A.D. 1640. This (Page 44) includes both the Fort Raleigh period and the period of early exploration of the Carolina Sounds by Virginians.
Averaging the two Feature 8 dates yields a mean date of A. D. 1643 with a one sigma range of A.D. 1525 to A.D. 1654. By including the suspect palisade date in the average, alternative mean dates of A.D. 1516, 1599 and 1617 with a one sigma range of A.D. 1476 to A.D. 1636 are indicated. The A.D. 1599 date approximates the Fort Raleigh period closely, but again a considerable portion of the seventeenth century lies within the one sigma range.
The Amity Site radiocarbon dates do not in and of themselves allow a conclusive assignment to either the late sixteenth or the seventeenth century. At the one sigma range both averaged dates include both periods. All but the suspect palisade date includes both periods at the two sigma range. Obviously, calibration is no panacea for problems engendered by widely varying dates. It can be seen though that only one of the calibrated dates includes any portion of the post-contact period within its one sigma range, and that is the Feature 8 date with a mean of A.D. 1662. Furthermore the average of the two Feature 8 dates is A.D. 1643. Thus the dates lend some tenuous support to a mid-seventeenth century date for the Amity Site.1
1. If the Feature 8 date from 1986 (Beta-17507) is calibrated in spite of its not being corrected for isotopic fractionation, the results do not change appreciably. A mean date for the sample of A.D. 1437 is indicated with a one sigma range of A.D. 1407 to A.D. 1609 and atwo sigma range of A.D. 1280 to A.D 1650. When averaged with the other Feature 8 dates, alternative mean dates of A.D. 1531, A.D. 1544, and A.D. 1636 are indicated with a one sigma range of A.D. 1515 to A.D. 1646 and a two sigma range of A.D. 1472 to A. D. 1658. When averaged with the Feature 8 and Feature 25 dates, a mean date of A.D. 1492 is indicated with a one sigma range of A.D. 1453 to A.D. 1631 and atwo sigma range of A.D. 1440 to A.D. 1643. Although slightly older than the dates presented above, these dates, even if valid, do not compel the acceptance or rejection of either the sixteenth or seventeenth century.
Carolina Algonkian Project, All Rights Reserved