|
the Aborigines of Southwestern Connecticut By Bernard W. Powell * ABSTRACT This attempt to order eleven archaeological manifestations in southwestern Connecticut compares artifact types, excavated features, and stratigraphy. Certain relationships to a cultural chronology proposed earlier for nearby coastal New York, and to the McKern taxonomy for North American archaeological phenomena are suggested. The present study is a first approximation; future revisions should be anticipated. _____________________________ To provide the basis for a first regional site synthesis and overall chronology, eleven prehistoric sites in southern Connecticut are examined; an attempt is made to relate them to one another, and to arrange them within a time continuum. All sites, with one exception, have been described. For the sake of brevity, I have kept comparisons mainly to two artifact categories: ceramics and projectile points. These I feel are the most sensitive diagnostics we have to work with. Ancillary cultural traits such as inhumations, hearths, pits and other features are cited in several instances. At the risk of oversimplification, but again for purposes of brevity and in conformity with a literature too lengthy to cite here, I have assumed that there is a most probable temporal and/or stylistic placement (stratigraphic position) for certain cultural traits (Ritchie 1958). Thus, narrow-bladed, stemmed and notched projectile points usually precede others and are predominantly of the Archaic period (Fig. 2). In ceramics, grit temper generally takes temporal precedence over shell, while brushed and roughly corded wares precede incised, carefully cord-impressed and shell-stamped wares - ordinarily decorated around rims, collars, and lips. Figure 1 shows site locations as numbered in text and the cultural periods or patterns thought to be present. Figure 2 is a modification after Ritchie (1958: Fig. 5, end paper) and Smith (1950, 1955), of a proposed chronology for nearby coastal New York.
Fig.
1 Descriptions of each site, including pertinent artifact
distributions, follow: 1) Bitter Rock Shelter A probable human cremation, crude chopper, a chipped semi-lunar knife (similar to the Samp Mortar Quarry series cited below), exotic stone flakes, and Lamokoid style projectile points were interpreted by the excavator, Powell (1965a), as Archaic - perhaps Early Archaic - evidences at this stratified rockshelter three miles from Long Island Sound at Norwalk, Fairfield County. These materials were overlain by refuse of shellfish-consuming, pottery-making cultures, probably affiliates of the East River Tradition (Fig. 2). Ceramics included 231 sherds. Cord-marked, shell-tempered wares were dominant in this sample. The type Bowmans Brook Stamped (Late Woodland period) was probably present, along with an assortment of undescribed lip-notched, corded, and broadline-incised sherds, and one sherd showing a curvilinear motif.
Fig.
2 Powell (1963) suggested projectile point distributions similar to those noted by him at Mianus Gorge Rock Shelter 20 km. to the northwest, and by Ritchie (1958), in the lower Hudson Valley 60 km and more to the northwest. Of these points (total: 20), 55.0% were triangular and (with one exception) occurred stratigraphically within 45 cm from the surface. These were accompanied by one (5.0%) broad-bladed, corner-removed point. The remaining 40.0% of the points were narrow-bladed, Lamokoid stemmed and notched forms (with one exception) confined to positions below 45 cm. 2) Eagle Hill Sites An extensive, probably multi-component site that included evidences of village life was reported by Coffin (1946) at the mouth of the Housatonic River near Laurel Beach in Milford, New Haven County. A nearby and probably related site, the Eagle Hill Ridge Site, was subsequently described (Coffin, 1951). An assemblage of artifacts (sometimes amazing for sites of this region) was illustrated in Coffin's two reports. Pottery is said to have been decorated with incising, punctations and clam-shell crenelations, and with herringbone, chevron, and crenelated plat decorations. One illustrated vessel (Coffin 1951: Pl. 1) is unclassified but thought by Coffin to show "Hopewellian influence." It is unfamiliar to me. Another unidentified vessel (Coffin 1946: Fig. 26) appears to me to be a Clasons Point Stamped piece of pottery (Late Woodland, Protohistoric period). This, plus "Contact" goods (glass, metal knives) and a well-developed bone industry (which is illustrated), may in this context suggest a synchrony with the Clasons Point Focus of the East River Tradition. Rouse (1947), in a separate study that included the ceramics from this site, suggested the presence of a complete sequence of four Windsor stages, with North Beach types at the bottom of the deposit, Clearview types stratigraphically above them, and Sebonac and Niantic pottery with an admixed minority of East River ceramics at the top. Typological criteria apparently guided Coffin's thinking in the separation of projectile points, but diagnostics are not clearly stated and references and documentation in general are wanting. Allowing for any inherent bias, insofar as I can determine from the points illustrated in both papers, 28.20% (30 specimens) were triangular, 46.06% (49 specimens) were Lamokoid, and 25.39% (27 specimens) were broad-bladed notched and stemmed forms. There was a scattering of pentagonal, lozenge, and aberrant forms, and bone points (late) were present as well. Extended burials, bundle burials, cremations, and dog burials were reported. "Mounds" and "parapets" were cited. These are much too loosely described to support much conjecture but, taken with the rich and varied assemblage of artifacts, may hint tenuously at the presence of an exotic manifestation in the region - perhaps even an Adena (or derivative Middlesex) outpost - as now noted from several locales remote from the Ohio homeland (Ritchie 1960; Keith n.d.). 3) Eckart Site This site, farthest inland of all the sites considered here, was reported by Jostrand (n.d.). It is near a bend in the Housatonic River in New Haven County, 6 km. north of Newtown. Lack of pottery and minimal evidence of domestic activity (although some hearths were noted) supported the inference that this was an Archaic chipping station. The inventory of 427 projectile points included many notched and stemmed forms. No percentage breakdown accompanied the report, but a majority of the forms I was able to observe personally were definitely Lamokoid. On a visit to the site itself, I noted depressions near rock outcrops reminiscent of those I cite as being present at the probably Archaic Samp Mortar Quarry (below). The depressions are also similar to features of this kind which occur at aboriginal quarries I have seen at Flint Ridge in Licking County, Ohio, but they were not mentioned in the report. The excavator stated that the site had been plowed, so stratigraphy was largely ignored. 4) Hunting Ridge This inland "site" was really a findspot for 66 potsherds and a small chipped quartz knife. A few evidences of fire also occurred at the base of a rock outcrop in the northern residential section of Stamford, Fairfield County. The finder, Powell (1959), thought the pottery bore resemblances to ceramics of the Owasco Tradition in New York, dated there between 600 and 900 B. P. 5) Indian Field Site This coastal site, reported by Powell (1958), was several acres in extent and lay along the west bank of Cos Cob Harbor in Greenwich, Fairfield County. Physical manifestations included a thin marine-shell midden, and numerous isolated shell pits. Pottery characteristic of the Windsor and East River traditions was recovered. Within a sample of 526 sherds, 15.20% were classed as Windsor and 84.74% were classed as East River. A relative lack of early Windsor types supported the inference that a terminal Windsor component was present. The East River ceramics included: Eastern Incised, Van Cortlandt Stamped, Bowmans Brook Stamped, and East River Cord Marked. Since post-Contact items were not present, the bone industry was weak, and wampum-making not in evidence, a pre-Contact placement for the East River occupation was suggested. (The author, however, cited literary references suggesting indigenes frequented the site in Contact times). It is proposed that the site harbored a component of the Bowmans Brook Focus of the East River Tradition, plus an earlier (unspecified) Windsor manifestation. Projectile points were mixed; a sample of 44 extracted from Table 2 contains 54.48% which were triangular and 45.52% (the balance) that included stemmed, side-notched and other forms. Narrow-bladed points were not distinguished from broad-bladed ones. 6) Laddins Rock Burials Three aboriginal burials in a 1.5 km. radius at a location near the Stamford-Greenwich town line in Fairfield County were analyzed by Powell (1962: 31) for traits and similarities. Burial 1, purported to have been a child's burial, was said to have been accompanied by cord-marked sherds and marine shells; fragments of a kaolin trade pipe were also associated with this burial. This last circumstance permitted its placement within the post-Contact period. Burial 2 was an adult human cranium and miscellaneous postcranial bones recovered with quartz chips and marine shells in a gravel bank off Laddins Rock Road. Powell tentatively assigned these human remains a Late Woodland provenience on the basis of his rather tenuous assumption that certain traits were shared with Burial 1. Burial 3, the only one recovered by Powell personally, was the most complete as to the human skeleton involved, but it was not accompanied by either grave goods or cultural refuse. A shell lens 27 m. southeast of this burial and thought to be in the same horizon yielded a broken bone awl and one cord-marked sherd. Powell felt that this burial, too, probably was related to the Late Woodland period. 7) Manakaway Site This site, a coastal shell midden on Greenwich Point (Todd's Point) in the Sound off Old Greenwich, Fairfield County, was reported by Suggs (1958). It was interpreted by him as a component of the Bowmans Brook Focus of the East River Aspect. A weak projectile point assemblage (14 points) was cited (Suggs 1958: 29, 34), with a shift to triangular points characteristic of Woodland times; the ceramic inventory (988 sherds) contained pottery types: Van Cortlandt Stamped, Bowmans Brook Incised, and East River Cord Marked of the East River Tradition; Vinette Interior Cord Marked and Modified Vinette of the Windsor Tradition; one Owasco trade sherd was also present. Stratigraphic interpretations were difficult. A radiocarbon—supported date of 1200 A.D. (750 B.P.) was advanced by the author. Duration of occupation was said to be narrow, and no clear evidence of earlier or later occupations was adduced. 8) Mianus Gorge Rock Shelter An absence of pottery in the lowest levels together with the typology of the projectile points present led Powell (1963: 152) to postulate a primary Archaic occupancy followed by a diffuse Woodland scatter in the upper levels. This site is an inland rockshelter located north of Stamford, Fairfield County. Lowest layers yielded problematical chipped forms and exotic stone flakes suggestive of very Early Archaic (or marginally-lagging Archaic) materials observed in other collections; perhaps also there might be similarities to the "Unspecialized Lithic" cited by Byers (1959) at an upstate Connecticut site. Inventory included 23 projectile points, of which 26.04% appeared to be Lamokoid; 56.42% were broad-bladed stemmed and side-notched; 13.02% were triangular; and one specimen (4.34%) was pentagonal. In context, the author suggested that their chronological distributions were broadly analogous to those reported by Ritchie (1958) for sites in the nearby lower Hudson Valley of New York. A small sherd inventory did not permit accurate typing, but 95% of the sherds were said to have been grit tempered and 5% possibly fiber tempered. None were shell tempered. Fifty-seven percent had cord-marked exteriors. In context, this suggested that potters of the Windsor Culture had briefly frequented the area. Later East River pottery was also represented by a sample of three small rimsherds showing cord impressions on and near the lips; a transient visitation during an early phase of this period is possible. 9) Samp Mortar Quarry This aboriginal quartz quarry north of Fairfield, Fairfield County (Powell n.d.) yielded a profusion of flakes and rough cores - plus some partly blanked artifacts and a series of semi-lunar knives. These knives were cross-related to a similar form found deep in the Bitter Rock Shelter (Powell 1965a). That this quarry manifestation is Mid-to-Late Archaic period in age was suggested by virtue of provisionally identified artifact types. 10) Sasqua Hill Site This site, still under study by Powell, is located on the inland side of an extensive salt marsh near East Norwalk, Fairfield County. Primary manifestation was a densely packed midden of marine shells along with some isolated shell pits and human burials.1 A rich ceramic inventory, not yet described, includes Mid-to-Late Woodland types (Powell n.d.) Two burials with ceramic affiliations, one flexed (Powell 1965a), are reported.2 The site had been extensively pot-hunted prior to our excavations, but I secured good stratigraphic profiles from undisturbed portions that were representative. Multicomponentcy is stratigraphically suggested from Archaic times (ground siltstone ulu in lowest levels) through Transitional times (Orient Fishtail points) to what is probably a Late Woodland/Protohistoric times representation (Clasons Point pottery and kaolin pipe fragment) near the top. 1Future radiocarbon dates on the soil and shell fractions are anticipated through an independent study by the Connecticut Agricultural Experiment Station. 2Radiocarbon date for the flexed burial is being derived through assistances rendered by the Heye Foundation, New York. 11) Spruce Swamp Site The semi-drowned midden at this coastal site in East Norwalk, Fairfield County, yielded evidences of Late Archaic or Transitional period Indian presence and Woodland occupations prior to Contact. Powell (1965b) reported that 17.1% of the projectile point inventory total of 111 were Lamokoid, 35.9% broad-bladed stemmed and notched forms, and 45.9% triangular forms. Identification within the sample of the following Archaic point types based on published (Ritchie 1961) criteria for New York State was suggested: Lamoka, Brewerton Corner Notched, Brewerton Eared Notched, Brewerton Side Notched, Orient Fishtail, and Steubenville Stemmed. Additionally, one specimen each of the Middle Woodland forms, Jack's Reef Pentagonal, and Levanna were reported. Pottery was relatively scarce, but what was present was likened to that known for both Windsor (Sebonac Focus) and East River (Bowmans Brook and Clasons Point foci) times. Of 52 specimens available for study, the following types were identified: Bowmans Brook Stamped, Van Cortlandt Stamped, Windsor Fabric Marked, and Windsor Brushed. The dense shell layers were quantitatively analyzed and a negative correlation was observed between all species of clam and one species of oyster (Crassostrea virginica). A curious scratched design on a hematite paintstone was compared (Powell 1964) to motifs of the Southern Cult. In conjunction with this, the presence at the site of an artifact, made from green plasma (var. quartz) "either from Alabama or Georgia" was noted. This anomalous presence may relate to 1) a purported "burial mound" destroyed before these investigations began, 2) "mounds" and "Hopewellian pottery" cited by Coffin (1946) at the Eagle Hill Site 28 km. to the east, or 3) curvilinear ceramic motifs reported from the Manakaway Site (Suggs 1958), the Bitter Rock Shelter (Powell 1965a), and at Pelham Boulder (Lopez 1958) - a New York coastal site 45 km. to the west. It was Lopez' suggestion that the curvilinear motifs found at Pelham Boulder might relate to Hopewellian and exotic influences in the region. Suggs (1958), however, felt that motifs of this kind could be explained on the basis of indigenous development. An adult, probably male, burial accidentally uncovered by workmen near the east side of the site after field work was completed, apparently lay flexed in a shallow shell pit and showed an unusual healed trauma of the cranium.3 3 0n the basis of expert opinion, ante-mortem trephination is suspected here; a thorough investigation is under way. SUMMARY The sites described above contained cultural diagnostics of both the Woodland and preceding Archaic periods. This assessment is partly made in terms of the McKern (Midwest) taxonomy as it has been used by some analysts in systemizing various and diverse archaeological materials (McKern 1939). Multi-componentcy at several sites suggests a temporal arrangement with similarities to Smith's chronology (Fig. 2). A fluted point of the Paleo-Indian tradition is recorded as having been found on the west bank of the Housatonic River near its mouth (Ritchie 1957). The earliest human occupancy of sites in Fairfield and New Haven counties after this enigmatic trace would appear to be represented in the lowest levels of Bitter Rock Shelter and Mianus Gorge Rock Shelter, in the open sites at Spruce Swamp and Eagle Hill, and in the stoneworking stations at Samp Mortar Quarry and the inland Eckart Site. These phenomena may extend from the very earliest Archaic - guessdated maximally to 5000 B.P. and thought to be present on the basis of such indicators as ulu, semi-lunar knives, Lamokoid points and "unspecialized" forms - to later Archaic times (Spruce Swamp) perhaps 2000 years later. These predominantly hunting-gathering cultures were replaced in time by pottery makers of the Windsor culture, represented at Indian Field and Eagle Hill. Ceramics of the East River tradition, probably of the Bowmans Brook focus, sometimes coeval with Windsor pottery, reflect later Woodland manifestations at Indian Field, Mianus Gorge, Hunting Ridge, Eagle Hill, Sasqua Hill, Bitter Rock Shelter, and Spruce Swamp. A still later manifestation of East River times is the Clasons Point focus, probably present at Manakaway (radiocarbon dated there at 750 B.P.), Sasqua Hill, and Eagle Hill. Aboriginal occupancy of the region was apparently in a state of decline by Contact times, but Contact materials may be present at Eagle Hill, Sasqua Hill, and in association with one of the burials of the Laddins Rock area. *Bernard W. Powell, Stonebridge Road, Wilton, Connecticut 06897. The paper was written in 1965 and accepted for publication in 1967. Citations were au courant as of those dates. REFERENCES Byers, D. S. 1951 The Eagle Hill Ridge Site -
Laurel Beach, Milford, Connecticut. Bulletin of
the Jostrand, T. Keith, B. Lopez, J. McKern, W. C. Powell, B. W. n.d.b An Aboriginal
Quartz Quarry at Samp Mortar Reservoir, Fairfield,
Connecticut. 1958 Preliminary Report on a
Southwestern Connecticut Site. Bulletin of the
Archeological 1959 A Ceramic Find at
Hunting Ridge, Connecticut. Bulletin of the
Massachusetts 1962 Some Connecticut
Burials. Bulletin of the Massachusetts Archaeological
Society. 1963 The Mianus Gorge
Rockshelter. Pennsylvania Archaeologist,Vol. 33, No.
3, 1964 A Curious Scratched
Design on a Connecticut Paintstone. American
Antiquity, 1965a Bitter Rock Shelter: A
Stratified Connecticut Site. Bulletin of the
Massachusetts 1965b Spruce Swamp: A
Partially Drowned Coastal Midden in Connecticut.
American 1965c On the Recovery of
Burial Number Two at the Sasqua Hill Site, East
Norwalk, Ritchie, W. A. 1958 An Introduction to
Hudson Valley Prehistory. NewYork State Museum
and 1960 The Eastern Dispersal
of Adena. New York State Museum and Science
Service, 1961 A Typology and
Nomenclature for New York Projectile Points. New York
State Rouse, I. Smith, C.S. 1955 Revised Chronology for
the Archeology of Coastal New York. Bulletin of
the Suggs, R.C. End |
