3.0 Introduction to the Terminal Archaic Period

This chapter serves to establish a baseline from which to compare the data generated from within the study area in the lowest Connecticut Valley. This initial section gives a broad perspective of the Terminal Archaic period throughout the Northeast.

The Terminal Archaic period from 3700 to 2700 BP as outlined by Snow (1980) can be characterized as exhibiting two specific cultural systems. One is the persisting and expanding Mast Forest tradition that was first seen during the Late Archaic period. The other cultural system witnessed within the Northeast is the "Susquehanna tradition" (Ritchie 1969a:150; 1969b:219, Snow 1980:235, Funk 1983:331). I (Pfeiffer 1984) have referred to this cultural manifestation as a complete cultural system with a generalized pattern of adaptation to the river plain. I have therefore used the term River Plain Adaptation. The basis for such a distinction will be fully explained within the framework of the research for the study area.

3.1 The Mast Forest Tradition and the Terminal Archaic Period

Contemporary hypotheses concerning the Terminal Archaic period suggest that in New York there is a complete replacement of both the Mast Forest and Lake

Forest traditions by the "Susquehanna tradition". Snow (1980:252) suggests that in fact a "complete culture system" is involved in this replacement and that there is the strong likelihood of migration--thus intrusion. "The data indicate that there were favorable conditions for migration. Independent invention is out of the question. Diffusion seems inadequate to explain such a sharp and pervasive change. Finally, to the degree that we are able to examine them, all four subsystems of the Frost Island complex (Susquehanna) appear to contrast with those of the Mast and Lake Forest systems. The criteria for migration are all met" (Snow 1980:252).

3.1a The Mast Forest Tradition during Terminal ArchaicPeriod in Northern New England

In northern New England, Snow (1980:244) states that there is a replacement and intrusion into the territory of the Maritime Archaic. The Maritime Archaic was pushed out as the ever expanding "Mast Forest-Susquehanna complex" from southern New England advanced northward. Bourque (1975:43) suggested "that the data for central Maine indicated the migration of a substantially new population." Dincauze (1975:27) attempts to simmer down the wholesale intrusion concept for northern New England to central Maine in suggesting that, "no mass migration was involved, but rather an infiltration of small groups of people."

3.1b The Terminal Archaic and the Mast Forest Traditionof Southern New England

In southern New England, the data from Martha's Vineyard caused Ritchie (1969b) and Dincauze (1975) to infer the continuity of Mast Forest from the Late Archaic through the Terminal Archaic. There was also the addition of the Susquehanna tradition. Dincauze (1972) in her discussion of the Atlantic phase, New England's initial phase of the Susquehanna tradition, suggested that this complex did not represent a lifestyle different from what was evidenced within the Late Archaic. From this, Snow (1980:247-248) suggested that there was no dramatic systematic change between the Late and Terminal Archaic periods in southeastern New England as had been evidenced elsewhere. Therefore, in the southern New England case, what was being seen in the Terminal Archaic was the embellishment of the technological subsystem, probably through a process of diffusion, of an already existing Late Archaic Mast Forest tradition. Similarly, in Snow's perception the mortuary system shows continuity from the Mast Forest Wapanucket burials. This may be a misidentification of Robbins' data by Dincauze (1975) and then Snow (1980).

A similar condition may exist in the interpretation of Dincauze's (1972) Atlantic Ledge Site. While the discourse pertaining to the Atlantic phenomenon is thorough with respect to typological analogs, many questions go unanswered. A striking point is that the site excavation yielded little. I excerpt from the report, "the stratigraphic position of the occupation floor was clearly established but objects could not be observed in association, nor features delineated--these disappointments prevented further detailed investigation of the sedimentary history of the site" (Dincauze 1972:46-47). What was analyzed by Dincauze were surface and beach finds. The data from this site are not at the level from which significant cultural information can be derived. There is clear evidence of a mixed and multi-component situation, based upon the assemblage of surface finds including Squibnocket, Atlantic, Brewerton eared notched, Brewerton side-notched, Brewerton eared triangle types, Stark and Neville points (Dincauze 1972:51).

Snow and Dincauze were unable to accurately assess the complex nature of southern New England culture history during the Terminal Archaic period since they lacked the necessary evidence. This was the consequence of the history of archaeology in Connecticut, and the lack of detailed local and regional data, or archaeology oriented toward specific questions. Until recently the Connecticut data base was just not available as it had been in New York and northern New England.

3.2 General Description of the Susquehanna Tradition:

Introductory Comments

The Susquehanna was first outlined as a tradition by Witthoft (1949) and has been further detailed by Dincauze (1968, 1972, 1975), Kinsey (1972), Kraft (1970), Ritchie (1969a&b), Funk (1976, 1983,), Snow (1980), McBride (1984a), Lavin (1988), and Pagolatos (1988) just to name a few. In the introductory sections of this paper I shall use the term Broad Spear. This term recognizes the technological uniformity of projectile points and the related ceremonial complex and is therefore suited to the initial depiction of the manifestation. In the later sections of this study the Broad Spear term is superseded by a less restrictive one that effectively characterizes the whole culture system.

I have not applied the term "Susquehanna" because I believe that the manifestation transcends regional boundaries. "Susquehanna" has regional connotations for New York, Pennsylvania and Maryland. However, there is substantial information that points to a range for the Broad Spear "tradition" from Georgia to Maine.

The term "tradition" in this case has been considered tentative by many archaeologists, because in southern New England it was believed that Broad Spear sites are ceremonial rather than habitational. Therefore, the artifactual material only relates directly to the ideological subsystem. The term "complex" used by Ritchie (1969a) is far more appropriate, at this stage of the discussion, because until recently we were unable to demonstrate that the Broad Spear phenomenon related to more than one cultural subsystem. The misuse of the term "tradition" is outlined by Snow (1980:190) and I offered my own reasons for it not being universally applied above in section 1.4b.

A major purpose of this paper is to show that there was a complete cultural system pertaining to this complex. The data presented in this chapter are a historical view of the cultural, temporal, and spatial aspects of the Broad Spear complex.

3.3 The Terminal Archaic Broad Spear Complex

There have been an increasing number of Broad Spear sites discovered on the eastern slopes of the Appalachians from Georgia to Maine. These date to the fourth millennium BP and have been detected along coastlines, rivers and streams.

Witthoft (1949,1953) and Ritchie (1969a), as well as others, viewed the Broad Spear (Susquehanna phenomenon) in a developmental-stage perspective and used the term "Transitional"to describe it. This term is not specific and therefore unsatisfactory. It has been used in various regions for many periods that represent a change from one lifestyle to another. With respect to this term as it was applied by Ritchie and Witthoft, the implication was that there was a transition from an archaic pre-ceramic stage to a woodland ceramic stage.

The "Susquehanna tradition is a sequence of historically related cultural units which share a complex set of behavioral patterns related to the manufacture and use of a few distinct styles of stone tools" (Dincauze 1975:23). I have attempted to demonstrate that a larger set of cultural units exist at the Griffin Site (Pfeiffer 1980 a and b), a single component Broad Spear mortuary site in southern Connecticut. Data generated from the excavation of this burial site related directly to the ideological subsystem. However, careful professional excavation including techniques integral to the recovery of both faunal and floral materials revealed far more than technological information. While this cemetery site was undoubtedly representative of the ideological subsystem, some data may have related to economic factors suggesting subsistence strategies and the potential of nearby habitations.

The Broad Spear complex of the Terminal Archaic period has been identified at many sites in southern New England. In this region the sites that have been reported are predominantly ceremonial rather than habitational. Only in a few cases are there any indications that the sites were anything more than special purpose sites that reflected the ideological subsystem. Dincauze (1968:61) suggests that part of the Mansion Inn site may have been related to habitation and interprets many functional utilitarian artifact types. This site, however, suffers from inadequate controls and methodology of excavation. As a result, the Mansion Inn site does not serve to clarify the nebulous connection between ceremonial-ideological subsystem sites and habitation sites which evidence technological, economical, and social aspects of the whole culture. The establishment of this relationship is necessary before the Broad Spear cultural system can be accurately interpreted.

It has been demonstrated in New York that Broad Spear complex sites have strong associations with watercourses (Ritchie 1969a; Ritchie and Funk 1973; Funk and Rippeteau 1977). In New England, Dincauze (1968), Basto (1937), Robbins (1968), Bourque (1975), Snow (1969) and Pfeiffer (1984,1990) have also discussed Broad Spear site locations within this physiographic setting. There is no proximity to waterfalls or rapids at any of these sites, rather there is a predisposition towards river sections with slow water velocities, relatively deep water, and minor gradient. Ritchie (1969a:157) points out that these sites, "occupy a riverine setting, never far from the main stream, usually upon the bank of the first terrace or higher portions of the flood plain."

The Broad Spear complex appears to be a cultural manifestation of the eastern slopes of the Appalachian Mountains from Georgia to Maine. In the south, the Savannah River Focus (Claflin 1931) and North and South Carolina (Coe 1952,1964) show a Late Archaic complex that has a Broad Spear technology. Farther north, sites that belong to the Broad Spear complex have been identified in Virginia and the Delmarva Peninsula (Kraft 1970:55). The Broad Spear complex sites that are in New Jersey and Pennsylvania (Witthoft 1953) have been reported by Hawkes and Linton (1916) at the Koens-Crispin Site that was further reviewed by Cross (1941), at the Savich Farm Site excavated by Regensburg (1970), at the Miller Field Site (Kraft 1970), and at the Peters-Albrecht Site and Tocks Island (Kinsey 1968,1972). New York has produced analogous sites such as Snook Kill, Dennis, Shagabak, Sylvan Lake Rockshelter, and Lotus Point in the Hudson River drainage (Ritchie 1969a; Ritchie and Funk 1973). In the Susquehanna drainage the Susquehanna tradition has been identified at the sites of Camelot, Kuhr, Fortin, and Enck (Funk and Rippeteau 1977; Funk 1992 in press). The tradition is also represented in western New York sites of O'Neil and Hickory Hill on the Seneca River (Ritchie and Funk 1973), and Piffard and Claud in the Genesee drainage (Ritchie 1969a; Trubowitz 1977; Trubowitz and Snethcamp 1975).

These regions have produced habitation sites for the most part. However, Koens-Crispin and Piffard appear to have both habitation and burial site function. According to Kraft (1970), Koens-Crispin is a key site in that it ties the northern sites of New England to the western and southern sites. Kraft illustrates these similarities in terms of projectile point typologies. Funk (1992 in press) reports a possible ceremonial cremation burial and related grave offerings at Camelot #2.. Further investigation should concentrate on other cultural subsystems of the Broad Spear phenomenon, not only typological (technological) and ceremonial (ideological) considerations.

New England has produced several significant Broad Spear sites. In Connecticut, there are a series of unpublished Broad Spear sites that I have inspected. These are Schwartz, which has been mentioned but not analyzed (Dincauze 1975), Carrier, and the Toelle's Road sites. Published sites are the South Woodstock Site (Basto 1937; Praus 1945), the Griffin Site (Pfeiffer 1980 a & b), and the Woodbury Site (Sinnott 1965, Thompson 1989). Three sites of the Broad Spear complex have been reported in Rhode Island. These are Conanicut Burial Ground (Simmons 1970), Flat River (Fowler 1968), and Twin River (Fowler 1952). Massachusetts has produced Mansion Inn, Watertown Arsenal, Vincent, Boats, Coburn, Atlantic Ledges (Dincauze 1968/1972) and Swan Hold (Sautter 1967) . In New Hampshire and Vermont stray "diagnostic" Broad Spear projectile points are known. However, no sites are reported with the exception of the Litchfield site, that had a small feature and one typable Broad Spear projectile. To Dincauze (1972) this site represents a Broad Spear component and is further cited by Turnbaugh (1975) and Cook (1976) with its "appropriate" date. I suggest that the occurrence of one projectile point in a feature does not sufficiently define the diagnostic nature of the assemblage or indicate the cultural affinity of the site. The use of the diagnostic point and the date in the regional assessment for New Hampshire and Vermont should be considered tentative. Maine presents a good case for cultural intrusion (Bourque 1975:43; Snow 1980:245). Here we can talk about other attributes that went along with the complex. The sites of Eddington Bend (Snow 1980) and Turner Farm (Bourque 1975,1976) suggest the "intrusion of the 'Susquehanna tradition' into central Maine from southern New England" (Snow 1980:245). I question whether these investigators have traced the intrusive element to the homeland, as Snow and Rouse suggest, or whether they have traced it back to an area where there has been a gap in the information. There are strong similarities between the northern New England Terminal Archaic sites and those of southern New England. I have inspected a collection from the Muscongus Bay region of Maine that the excavator described as coming from a cremation burial. There was a surprising similarity in both style and composition of the assemblage between the Griffin site material in the study area and the collection of Florence Elliot of Round Pond, Maine. There is a necessity to correlate information from other cultural subsystems and not only visual similarities of grave goods.

Information concerning radiometric dating for the Broad Spear complex can be regionally synthesized. The Savannah River Focus, first outlined by Claflin (1931), has been subsequently dated in excess of 4200 BP--possibly as far back as 4500 BP (Bullen 1962, Stoltman 1972,1973). Within this focus, Claflin noted the production of Broad Spear projectiles and large ground stone tools at the sites of Stallings Island and Bilbo. Both these sites were within large shell deposits along watercourses. However, the Broad Spear component was at the bottom of the midden, beneath the shell association (Coe 1952:305). Coe has investigated sites of the Savannah River Focus in South and North Carolina. Radiocarbon dates are around 3950 to 4500 BP (Coe 1964:118). Coe(1964) points out that in the Carolinas, the Broad Spear complex is seen along watercourses but that there is no evidence for shellfishing in the archaeological record (Coe 1952:305).

Kraft (1970:72) noted that there was a close relationship of several sites in the Virginia-Maryland-New Jersey area. These sites were associated with the Broad Spear complex that he had dated between 3590 and 3670 BP from radiocarbon samples from the Miller Field site (Kraft 1972:11). The Marcey Creek site in the Potomac drainage (Manson 1948:223) has many indicators of the Broad Spear phenomenon. Kraft (1970:55) has also suggested a close relationship between Koens-Crispin and Marcey Creek. Kinsey (1972) has seen a late transitional stage in the upper Delaware River valley at Tocks Island and describes projectile points from habitation sites that are typologically similar to eastern and southern New York "Orient" points. Fragments of steatite and coarse grit tempered pottery were also considered to be associated. Dates derived from the Upper Delaware samples of charcoal spanned the period 3120-3230 (Kinsey 1972:395). This kind of assemblage is considered to be late Terminal Archaic in age (Dincauze 1968, 1972, 1975).

Pennsylvania and New York have yielded Lehigh (Witthoft 1953) and Snook Kill (Ritchie 1969a) remains, which are Broad Spear complex analogs to the Koens-Crispin-Savannah River manifestations. Lehigh has radiometric dates of 3670 + 100 BP (Kinsey 1972:427) at the Peters-Albrecht Site in Bushkill, Pennsylvania. A date from Sheep Rock Site zone three, level five, dates a Frost Island component hearth at 3220 + 160 BP. A second series of dates was produced from a sample in Level Six in a hearth which was associated with "Savannah" Broad points and a thick living floor. A date of 3800 + 180 BP was generated from a charcoal sample collected within the hearth feature (Michels and Smith 1967:63).

Along the Hudson in New York, Ritchie has a date of 3420 + 100 BP [Y-1170] for the Snook Kill site (Ritchie 1969a:157). The Coffin and Dennis sites range from 2820 to 3040 BP (Ritchie and Funk 1973:72), and these investigators believe these components represent a "late" stage of the Broad Spear complex in New York. These investigators consider these components to be similar to the "Orient" phase, which was identified first in Long Island and dated to a range of 2713 to 2993 BP (Ritchie and Funk 1973:73).

An impressive sequence of dates and phases for the Broad Spear complex has been posted by Funk (1992 in press) and Funk and Rippeteau (1977:31). Here they suggest that the first evidence of the "Broad Spear complex" in the upper Susquehanna River Valley was the Genesee point that dated to 3780 BP in southwestern Ontario. A subsequent phase was demonstrated at the Kuhr site and was considered equivalent to Ritchie's Snook Kill. The dates for this were 3620 + 130 BP [I-6733]. The next phase evidenced was the Frost Island phase. It was seen at Camelot I and Camelot II, Johnson I, Kuhr I, Fortin, and Enck I. Camelot I dated to 3425 + 95 BP [I-6641]. Camelot II dated to 3240 + 95 BP [I-6744]. Kuhr I dated to 3485 + 90 BP [I-2094] and 3545 + 100 BP [I-7096], while Fortin dated to 3280 + 90 BP [I-7097], and Enck dated to 3250 + 110 BP [Dic-192].

The final phase was represented in the upper Susquehanna valley at the Rose site locus 2, locus 2 at Camelot I, and at the Davenport Creamery site. Features that contained these diagnostic assemblages did not have sufficient charcoal for dating. The exception was Camelot I, where a date of 1820 + 95 BP [I-6745]) was generated (Funk and Rippeteau 1977:32). This date is rejected by Funk as being too late. I have inspected the artifactual material from Kuhr, Camelot, and Fortin of the Upper Susquehanna and I see strong similarities to the southern New England material.

Farther west in the Seneca drainage, two sites attest to Broad Spear habitation at 3400 to 3200 BP (Ritchie and Funk 1973:71)--the O'Neil and Hickory Hill Marsh sites. These sites, along the margins of Howland Island, (Ritchie 1969a:xxii) have duplicated features of the upper Susquehanna sites but have also produced large tool types, steatite vessels, and bone tools. As in the Susquehanna drainage, "back country" sites are rarely encountered. From this region comes the first evidence of bone and antler objects, a bone awl and two bone barbed points, and an antler point with a line hole (Ritchie and Funk 1973:72). Non-local rhyolites were used to make projectile points as well as knives and scrapers. Also, a chipped chert adze was recovered which was very similar to examples found within cremation sites in New England.

The Piffard site is located in the Genesee Valley (Ritchie 1969a:154; Trubowitz 1977:119). Some of this large multi-component site was of the Broad Spear complex and appeared to have associated Broad Spear habitation and burial components. Ritchie (1969a:154) discussed the cremation burials and the cache of about fifty "killed" Perkiomen blades. He suggested that this was the first evidence for the burial practices of the Perkiomen associated culture. Trubowitz (1977:119) suggested that soapstone and Vinette I pottery also occurred within the "Frost Island" component of the site. At Claud I, also in the Genesee drainage, Trubowitz (1977:124) obtained a date of 3490 from samples of burned nutshells. Hickory, walnut, butternut, acorn, and red mulberry were identified along with Broad Spear projectiles which had similar ties to specimens from the Hudson, Susquehanna, and Seneca drainages.

In Connecticut there are now several radiocarbon samples that were obtained from cremation features. The dated charcoal samples from Connecticut ranged chronologically between 3500 and 3000 BP (Pfeiffer 1984:78; 1990:101). The sites and information will be discussed in Sections 5.3 and 8.4. Significant here is that the Connecticut sites are quite similar in feature and artifact form to the mid-Atlantic and New York sites.

Massachusetts Broad Spear sites are loosely dated by a sample from Vincent's, which yielded an age of 3470 + 125 BP [GX0568] (Dincauze 1968:45). This date has been cross-associated to the Watertown Arsenal and Mansion Inn sites. The Atlantic Ledges site (Dincauze 1972) dates prior to 3600 BP. This assessment is based upon sea-level rise calculations and the date when the site was first submerged. Atlantic Ledges is interpreted by Dincauze as a likely habitation site (personal communication, 1981).

In Rhode Island the important burial ground on Conanicut Island is well known as a historic cemetery. This site in the Narragansett Bay region has two prehistoric cremation burials that are associated with the Broad Spear complex. Simmons (1970:16-21) reports a date of 3380 BP [GX1113] for feature A-33 and 3280 BP [GX0735] from burial G-1.

Farther north in Maine, Eddington Bend is dated at 3430 + 145 BP [SI-789] (Snow 1975:53). Occupation Three of the Turner Farm Site has yielded three dates which pertain to the Broad Spear complex: 3650 + 75 BP [SI-1922], 3515 + 80 BP [SI-1924], and 3630 BP [SI-1919](Bourque 1975:38). Bourque notes that one component of the Hathaway Site (Snow 1969) has been dated at 3355 + 125 BP [SI-887] and further suggests that this date represents the most recent evidence of the Broad Spear complex in Maine. Most dates cluster around the middle of the fourth millennium BP (Bourque 1975:38).

There is a second group of sites in New England which Dincauze (1972) has suggested constitutes an early phase of the Broad Spear "tradition". She has called this the "Atlantic phase" and it is represented by the Atlantic Ledges Site (Dincauze 1972), the Call Site (Brennon 1960), Swan Hold Site (Sautter 1967), Curtis Site and Hoffman Sites (Dincauze 1972), which are all in eastern Massachusetts.

Similar sites are suggested by Swigert (1973) in western Connecticut. The LaRose site in South Windsor, Connecticut was discovered to be a multi-component Broad Spear complex site (Todzia and Pfeiffer M.S. given 1986 ASC annual meeting). This site exhibited horizontal separation of several components of Broad Spear, along with a distinct locus with Atlantic blades made of exotic black chert. These specimens were typologically consistent with those described by Dincauze (1972). The distribution of these projectile points extends into coastal Maine and most recently they have been discovered by Hoffman at Charlestown Meadows in eastern Massachusetts.

The dating of the Atlantic phase is not precise. Its cultural distinctiveness is not apparent on any other basis than projectile point typology. There is so little evidence from properly excavated sites that at this point the phase must be considered problematic.

Throughout the eastern slopes from Maine to Georgia there are several traits which are consistently witnessed as being within the Broad Spear complex. Projectile points as defined by Witthoft (1953:16-17) or knives (Kinsey 1972:426) are only one element of the associated materials of the complex. Other elements of the Broad Spear manifestation are iron pyrite and/or nuggets of other metals (Dincauze 1968; Pfeiffer 1984), cruciform drills (Claflin 1931; Bourque 1975; Dincauze 1968; Ritchie 1969a), soapstone vessels and/or plain thick coarse-tempered and steatite tempered pottery (Kraft 1970:108), the use of rhyolites or felsites for chipping material (Ritchie 1969a; Turnbaugh 1975), the lack of fishing equipment or shellfish (Coe 1952), or an apparent decrease from earlier components in the use of marine resources (Bourque personal communication 1981 and 1992; Ritchie 1969a; Pfeiffer 1984), scrapers based on broken projectile point forms (Kraft 1990:68), and the steatite perforated disc (Claflin 1931; Dincauze 1968; Manson 1948; Ritchie 1969a; Pfeiffer 1980 a & b; Snow 1980).

The sites are located in direct association with watercourses. Ritchie and Witthoft agree that in the Susquehanna drainage, Broad Spear sites are oriented along streams and rivers on the floodplain or on islands. "Apparently these folk were canoe wanderers who visited the back country only to replenish their supplies of steatite and rhyolite" (Ritchie 1969a:152). This comment may be a slight overstatement of the situation. However, it reflects Ritchie's view, and links several of the traits which I mentioned as being found together in Broad Spear contexts.

McBride (1984a&b) likewise sees evidence of riverine rather than interior sites for a "Susquehanna" period in the lower Connecticut River valley. The "Susquehanna period" a term which he uses is borrowed from Ritchie and Funk's (1973) conception of New York Frost Island and Orient phases. The archaeological sampling along transects which McBride employed in the town of Glastonbury, Connecticut has shown a very high percent of lowland site location and only a slight use of upland areas. Three percent of the total number of sites were situated in this back country zone (McBride 1984a&b). The methods used by McBride to generate this kind of information are the exception rather than the rule in New England archaeological studies. The laborious but unbiased task of test pit transects has successfully established a clear pattern. The comparison of the previous Ritchie statement to McBride's information illustrates the verification of previous impressions through a controlled sampling approach.

Predominant features in Broad Spear sites in the Northeast have been secondary burials of cremated bone, ash, and artifacts. These ceremonial sites are generally small, compact, and stratigraphically complex. Features are usually pits containing variable amounts of "greasy ash", heat spalled artifacts, bone fragments, burned nut caches, burned berries, and seed remains (Pfeiffer 1984).

In New York within the Hudson, Susquehanna, Seneca, and Genesee drainages there is clear evidence of the Broad Spear phenomenon stratified within multi-component sites. Within these drainage systems, sites of the Broad Spear complex are habitational and give us some insight into the subsistence base of the group. Ground stone tools that are quite prevalent in New England ceremonial contexts are not well represented in these occupation assemblages (Ritchie and Funk 1973:73). The ground stone tool associations between the ceremonial and habitational sites are not well established. However, projectile point typologies indicate a strong connection.

When I inspected several of the assemblages from the upper Susquehanna drainage, I was surprised to find that there was a striking resemblance between the lower Connecticut Valley cremation offerings and the materials from the upper Susquehanna drainage. Radiocarbon dates from the two regions' Broad Spear components are very close. The Geochron dates from the Griffin Site were 3535 BP and 3495 BP and the date for Camelot I and the "Perkiomen-Wayland" projectiles was 3470 BP.

Small winter-fall sites appear to be numerous in the Susquehanna drainage for the Frost Island phase. These investigators see evidence for a change in the seasonal settlement pattern for Broad Spear populations during the Terminal Archaic (Funk and Rippeteau 1977:21). Middle Archaic and Late Archaic sites were occupied during the spring-summer and functioned as fishing stations. Later on in the Terminal Archaic, Broad Spear populations used these same sites as collecting stations for seeds, nuts, and possibly berries. Components associated with the Broad Spear complex have quantities of bone, living floors covered with chipping debris, hearth features, and artifacts which are in numbers sufficient to allow the definition of types (Funk and Rippeteau 1977:21-29).

Figure 3.3.1 Northeastern Broad Spear Sites of the Terminal Archaic Dates and Calibrations
 
 

Site	Laboratory Lab # Date	Cal BP	Cal BC	Max/Min BP	Max/Min BC
Peters A         Y1826	3670+100	4063,4035,3988	2114,2086,2034	4148-3869	2199-1920
Sheep Rk      M2085	3220+160	3463	1514	3639-3270	1690-1321
Sheep Rk      M1907	3800+180	4227,4183,4158	2278,2234,2209	4503-3928	2554-1979
Snook Kl       Y1170	3420+100	3689	1740	3829-3569	1880-1620
Kuhr                I6733	3620+130	3967,3951,3929	2018,2002,1980	4141-3731	2192-1782
Camelot 1        I6641	3425+95	3690	1741	3832-4576	1883-1627
Camelot 2 I6744	3240+95	3468	1519	3584-3374	1635-1425
Kuhr 1 I2094	3485+90	3822,3789,3765, 3752,3729	1873,1840,1816,1803,1780	3881-3640	1932-1691
Kuhr 1 I7096	3545+100	3839	1890	3974-3699	2030-1750
Fortin I7097	3280+90	3519,3477	1570,1528	3631-3399	1682-1450
Enck Dic192	3250+110	3470	3629-3369	1521	1680-1420
Vincent GX0568	3470+125	3815,3795,3721	1866,1846,1772	3899-3619	1950-1670
Conanct GX1113	3380+160	3634	1685	3680-3629	1731-1680
Conanct GX0735	3280+90	3519,3477	1570,1528	3551-3473	1602-1524
Edd Bnd SI789	3430+145	3692	1743	3879-3478	1930-1529
Turn Fm SI1922	3650+75	3983	2034	4088-3879	2139-1930
Turn Fm SI1924	3515+80	3831	1882	3903-3693	1954-1744
Turn Fm SI1919	3630+70	3976,3943,3936	2027,1994,1987	4082-3855	2133-1906
Hathwy SI887	3355+125	3624	1675	3821-346	1872-1520

 3.4 The Broad Spear Complex as a Cultural System in the East

There is an interesting relationship which is apparent from the information in the previous section. The regions outside of southern New England have evidence from the Broad Spear complex in habitation contexts. Burials have been discovered in this region but they are rare and and when they have been identified are not elaborate. In southern New England Broad Spear complex habitations are incompletely documented. However, there are many burial sites that exhibit elaborate ceremonial activity.

There is clear evidence that the Broad Spear complex has direct habitation associations in the areas outside of southern New England. For regions outside of southern New England a good case can be made for a complete culture system to exist. This cultural system can be defined by the archaeological reconstruction of its cultural subsystems.

3.4a Economic Subsystem

The settlement pattern over the Broad Spear region, outlined in the introduction, suggests that the sites of this complex were predominantly located along river banks and coastlines. Even the more interior sites such as South Woodstock in Connecticut, Snook Kill in New York, and Gaston in North Carolina have been located along watercourses.

Turnbaugh (1975) has suggested that this settlement pattern was the consequence of the "Broad Spear tradition's" reliance upon shad (Alosa sapidissima) and other anadromous fish. He has superimposed the distribution map of the shad yearly migration on the settlement pattern map of the Broad Spear "tradition" and derived a close match. He stated that this match was a good indication of the economic base for the Broad Spear tradition. Turnbaugh further outlined the apparent northward spread of the Broad Spear "tradition" during the first half of the fourth millennium BP and paralleled this with the amelioration of the climate during the same period. He said that this climatic change enabled shad to extend their northward range and that the Broad Spear tradition followed the resource into new regions.

This hypothesis is countered by evidence derived from the archaeological record. The portion of a stream where prehistoric aboriginal fishing was most heavily undertaken was at fall lines where the fish had to follow steep constrained channels as they passed upstream to spawn. Fishing in areas such as these was done by employing weirs, nets, gaffs, and barbed spears. Sites of the Broad Spear complex do not occur in proximity to this portion of a river system. Instead, Broad Spear sites show up in slow water sections. Additionally, the archaeological record shows no artifactual material that is even remotely considered to be fishing equipment. Furthermore, no suggestion of line and bait fishing equipment has been found. Of course, anadromous fish do not feed as they go upstream to spawn. Nevertheless, it is doubtful that fishing activity was the cause of riverine site location as Turnbaugh suggests.

At the Griffin Site, many food offering remains were found in funerary pits (Pfeiffer 1980). Edible plant materials such as goosefoot and pigweed, hickory nut, hazel nut, and acorns, as well as faunal materials such as deer, dog, rabbit, and birds were incorporated in the pit. Neither shellfish nor fish bone were found, even with the use of flotation methods. Thus, evidence for the incorporation of fish and shellfish as an important constituent to the economic base in the Broad Spear complex is weak. This pattern has been recognized and published from the Carolinas to southern New England(Coe 1952; Funk and Rippeteau 1977; Ritchie 1969a). In Maine, Bourque has noted at Turner Farm that in this multi-component site the "Susquehanna occupations showed a marked decrease in the usage of marine resources and an increase in the plant materials (Bourque personal communications 1992).

Site location alone is a weak indicator of specific activity and economic base. There is an important difference between the total environment and the effective environment. Not every source of subsistence that exists within an environmental niche was necessarily used by aboriginal peoples (Jones 1978). In my assessment of the literature pertaining to the Broad Spear complex, I cannot find definite evidence for the use of either fish or shellfish resources (Pfeiffer 1984). The one exception to this is the Turner Farm site on the central Maine coast (Bourque personal communications 1992). Along the Southeast coast at the Bilbo and Stallings Island sites the Broad Spear components were beneath and unassociated with midden materials.

Why, then, do Broad Spear complex sites occur along watercourses? I have suggested that the Broad Spear complex was dependent on a specialized gathering strategy with supplemental hunting in much the same way as Funk and Rippeteau (1977) envision the adaptation of Broad Spear components in the upper Susquehanna drainage basin (Pfeiffer 1984, 1990). The incorporation of nuts and chenopodium within the Griffin Site may be a substantial indicator of economic base (Pfeiffer 1980). Chenopodium grows along lowlands and floodplains. Its prehistoric northern range is not well known. However, it certainly extended to north central Connecticut (McBride 1978). The present range extends up to northern Maine. The processing of chenopodium required blanching and finally grinding into flour. Flat mortars and grinding pestles are an important category of tools at Broad Spear complex sites (Kraft 1970:85-90). Thus, in the archaeological record, one finds chenopodium as well as other floral remains, tools to process the plant material, and local availability that corresponds to the distribution of the Broad Spear sites. At the same time, such sites located along waterbodies may have provided good avenues for travel and efficient routes for exchange.

One of the most striking elements by which Mast Forest sites and Broad Spear sites can be differentiated is the source of lithic materials used in tool manufacture. The Mast Forest adaptation generally concentrated on local pebbles or cobbles as chipping material (Ritchie 1969a&b; Dincauze 1975; Snow 1980), while the Broad Spear complex employed predominantly quarried cherts and rhyolites as the preferred material (Dincauze 1975; McBride 1984; Pagolatos 1988). In regions where these were not locally available, trade or foray was employed to attain the material.

One of the most efficient means for transporting material during prehistoric periods would have been either travelling on foot along banks of rivers or paddling along watercourses. This has long been recognized throughout the Northeast (Driver 1961:235-238; Willoughby 1935:135). Therefore, there are advantages to site locations along water courses that are unrelated to fish resources.

3.4b Social Subsystem

Ritchie (1969a:158) inferred from the O'Neil site that there was band level social organization. He states that, "sites are small, indicative of mere camps, but they may overlap extensively" (Ritchie 1969a:157). At the O'Neil site, the Frost Island component dated to 3200 BP [Y-1274]. The camps were superimposed and Ritchie interpreted this to indicate repeated returns by the same group. No house patterns have been noted in New York. However, small spiral circular house patterns are reported in southern Maine. These are thought to be large enough to house a family or extended family (Yesner personal communications 1983).

3.4c Technological Subsystem

Tool types from Broad Spear sites are axes, adzes, gouges, hones, whetstones, bevelled cobbles, mullers, flat mortars, steatite bowls, cylindrical hammerstones, punches, large knives, spear and dart points, drills, scrapers, shaves, and ceremonial blades. The point type that is the dominant form for the Broad Spear complex is a bifacially flaked stone cutting and piercing tool. Specimens of this type exhibit remarkable thinness and fine craftsmanship. These artifacts are less than 1 cm thick, can range up to 20 cm in length, and are as much as 6 cm wide (Dincauze 1968: 21-26) (Pfeiffer 1980b). Manufacture of these implements has been described by Dincauze (1972) as involving percussion techniques using cylindrical hammerstones and punches. I have suggested (Pfeiffer 1984a:84) that the industry exhibits standardized results in terms of the form of the biface and that this is the result of a specialized blade-flake technology. Punches were requisite production tools that yielded standardized results. (Note the punch in figure 8.4.2a.)

3.4d Ideological Subsystem

Burial practice is one of the few surviving representatives of the ideological subsystem that can be discovered archaeologically. Cremation burial is indicated for the Broad Spear complex. In New York the Piffard site showed what appeared to be a cremation feature. Many of the Perkiomen points were heat spalled in the classic cremation mode (Pfeiffer 1980; Dincauze 1968). At the Frontenac Island site, Burial 142 was a cremation with killed rhyolite blades. Ritchie (1969a) attributed this burial to the Frost Island phase. Camelot II in the upper Susquehanna valley also had a possible cremation feature with a Broad Spear association. In New Jersey, Koens-Crispin was a multi-purpose site with both habitation and cremation loci. The New England series of cremation burials has already been mentioned.

The interpretation of the significant aspects of the burial practices that indicate at least some of the ideological subsystem is presented in sections 8.1d and 8.4d.