BULLETIN OF THE MASSACHUSETTS ARCHAEOLOGICAL SOCIETY

BULLETIN OF THEMASSACHUSETTS ARCHAEOLOGICAL

SOCIETY

VOLUME 65 (1)

CONTENTS:

Editor's Note.

Spring 2004

1

An Early Woodland Period Ceramic Production Feature inBellingham, MA

A Late Woodland and Contact Period Ceramic AssemblageFrom the Hartford Avenue Rockshelter, Uxbridge, MA

Mary Lynne Rainey

Duncan Ritchie

2

9

Craig S. Chartier 25

The Tall Pines Rockshelter, Clinton, MA and Rockshelter Use Duringthe Late Woodland and Contact Periods. Martin G. Dudek and Craig S. Chartier 18

A Contact Period Fishing Point of Cow Bone From Grape Island,Boston Harbor, MA .

Native American Settlement in the Upper Housatonic During theWoodland Period

Betty Little: An Appreciation

Contributors.

Timothy Binzen 29

Dena F. Dincauze 39

40

THE MASSACHUSETTS ARCHAEOLOGICAL SOCIETY, Inc.P. O. Box 700, Middleborough, Massachusetts 02346-0700

THE MASSACHUSETIS ARCHAEOLOGICAL SOCIETY, Inc.Robbins Museum of Archaeology

http://webhost.bridgew.edu/masContact by phone: (508) 947-9005, or by email: [email protected]

Officers:Ronald Dalton, 100 Brookhaven Drive, Attleboro, MA 02703Tonya Largy, 59 Moore Road, Wayland, MA 01778Susan Jacobucci, 678 Chief Justice Cushing Highway, Scituate, MA 02066Edwin C. Ballard, 26 Heritage Road, Rehoboth, MA 02769Eugene Winter, 54 Trull Lane, Lowell, MA 01852James W. Bradley, 55 Park Street, Charlestown, MA 02129Curtiss Hoffman, 58 Hilldale Road, Ashland, MA 01721Darrell Pinckney, 2668 Edgewood Avenue, Schenectady, NY 12306

PresidentVice President

ClerkTreasurer

Museum CoordinatorBulletin Editor

Corresponding SecretaryPast President

Trustees:Elizabeth Chilton, Dept. Anthropology, Machmer Hall, UMass, Amherst, MA 01003Frederica Dimmick, 10 Sassamon Road, Natick, MA 01760Nicia Gruener, 30 Bancroft Road, Andover, MA 01810John F. Healey, 222 Purchase Street, Middleborough, MA 02346Lorraine Kerrigan, 96 Old Colony Avenue, U-554, East Taunton, MA 02718Gregory Lott, 142 Herring Pond Road, Plymouth MA 02360Thomas Lux, 38 Somerset Avenue, Riverside, RI 02915Richard Lynch, 12 Greenbrier Road, Greenville, RI 02828Maryanne MacLeod, Swett Hill Road, Sterling, MA 01564John Rempelakis, 7 Fairview Farm Road, Haverhill, MA 01832Alan F. Smith, 156 Ararat Street, Worcester, MA 01606John Thompson, 406 Main Street, Medfield, MA 02052

Term ExpiresOctober 2006October 2006October 2004October 2006October 2005October 2005October 2005October 2005October 2006October 2004October 2005October 2004

MAS Representative on the MHCNewsletter Editor

Membership DirectorLibrarian

Museum AdministratorMuseum Administrator

Judith Zeitlin, Dept. Anthropology, UMass, Boston, MA 02125Tom Lux, 38 Somerset Avenue, Riverside, RI 02915Curtiss Hoffman, 58 Hilldale Road, Ashland, MA 01721Kathryn M. Fairbanks, 145 Aldrich Street, Roslindale, MA 02131Heidi Savery, 180 Main Street #5204, Bridgewater, MA 02324Susan Jacobucci, 678 Chief Justice Cushing Highway, Scituate, MA 02066

The BULLETIN OF THE MASSACHUSETTS ARCHAEOLOGICAL SOCIETY is publishedsemi-annually, with a Spring issue, Number I, and a Fall issue, Number 2. Institutionalsubscriptions are $30; individual memberships in the Society that include receiving the Bulletinare $20. Information on special rates for membership without the Bulletin, family members,seniors and students, as well as requests for back issues of the Bulletin, should be addressed tothe Massachusetts Archaeological Society, P.O. Box 700, Middleborough, MA 02346 (508-9479005). Publication in the Bulletin is a privilege of membership. Manuscripts andcommunications may be sent to the editor, James W. Bradley, 55 Park Street, Charlestown, MA02129 or emailed to [email protected].

Printed by Kendall Press, 36 Charles Street, Cambridge, MA 02141

BULLETIN OF THE MASSACHUSETTS ARCHAEOLOGICAL SOCIETY, VOLUME 65(1) 2004

EDITOR'S NOTE

1

Once again, the articles in this issue of the Bulletin revolve around a particular theme - the sites andmaterial culture of the Woodland period, which extended from roughly 3000 B.P. to the time ofEuropean contact. It is also known as the Ceramic period since pottery vessels, or at least fragments ofthem, are a distinguishing characteristic on most sites.

The articles address this theme in several ways. One is technology. Rainey describes an unusualfeature and, after testin§ several hypotheses, concludes that it was used for ceramic production around2100 B.P. With firm 1C dates and associated wasters, this site provides an important example ofpottery making near the end of the Early Woodland period. The second approach presents specificexamples of ceramics recently recovered from Late Woodland to Contact period sites in centralMassachusetts. Ritchie describes two 14C dated vessels from a rockshelter in Uxbridge, while Dudekand Chartier present an example of a third ceramic style from a rockshelter in Clinton. Chartier'sarticle provides an interesting case study for how subtle the evidence for European contact can be. Inthis instance, the presence of European faunal remains indicate that an otherwise ordinary-looking LateWoodland site actually post-dates Contact. A final approach looks at the occurrence of Woodlandperiod sites in a broader regional context. Binzen compares the frequency of Woodland period sites tothose of the preceding Late Archaic and finds that, contrary to what has often been asserted, Woodlandperiod sites are well represented. As always, my thanks go to all the authors for contributing excellentarticles. Thanks also go to my faithful proofreaders Shirley Blancke and Kathy Fairbanks for catchingall the errors I missed.

Finally, Dena Dincauze provides a personal remembrance of Elizabeth A. Little, Past President of theMassachusetts Archaeological Society and Editor of the Bulletin, who passed away last summer. Anearlier version of these comments was read at a memorial service held in Betty's honor on September12, 2003 at the First Parish Church in Lincoln, MA. Plans for a more comprehensive volume celebratingBetty's life and accomplishments are currently underway. This is a joint project of the MassachusettsArchaeological Society and the Nantucket Historical Association with Mary Lynne Rainey ascoordinator. It will contain a comprehensive review of Betty's work as well as a series of scholarlyarticles by friends and colleagues. Current plans call for this volume to be completed by late 2005.

James W. Bradley

This journal and its contents may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling,loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. ©2011 Massachusetts Archaeological Society.

2 Rainey: Early Woodland Period Ceramic Production Feature

An Early Woodland Period Ceramic Production Feature in Bellingham, MA

Mary Lynne Rainey

Site Context

In 1999, the Public Archaeology Laboratory,Inc. (PAL) conducted a data recovery at theEast Terrace Locus 1 Site in Bellingham, MA(Figure 1; Rainey 2000). The site was one ofseveral discovered along the banks of the upperCharles River in 1997 during planning stagesfor an electric generating facility and associatednatural gas interconnect route. Prior to the datarecovery, the site was interpreted as a mediumsized Late Archaic to Middle Woodland perioddomestic base camp, where one or more familygroups resided (Leveillee and Waller 1998).These conclusions were based on a general lackof hunting gear, the presence of aboriginalpottery, calcined bone fragments and a featurebelieved to be a living surface, perhaps part ofan Early Woodland household structure. Thefeature was exposed in a 1 m2 unit set backfrom the river terrace. It consisted of a layered

sequence of disturbed soils containing charcoal,small pieces of fire-cracked rock, and a fewquartz flakes.

Although many other sites have been recordedin this interior setting, few have beensystematically studied. Virtually nothing isknown of Native ceramic traditions here, andthe potential for new information from the EastTerrace Site was promising. The only ceramicstudy specific to this area was based on a 1967Charles River survey by Dena Dincauze(Dincauze 1968). Although the ceramic samplewas very small, she observed, 'a markedpreference for smoothed vessel bodies in theCharles basin collections' (Dincauze 1975:14).Since that time, there has been little newinformation regarding development of ceramictechnologies in this region of Massachusetts.

Data recovery at East Terrace Locus 1 involved

NEW I1AMPSHIRE

20 30 40 nil!

Minor drainage basin

1\tajor drainage basin

10oi

HOOSIC

Nashwl

_ ...--r..<J'-

VERMONT

.r..-.. "---"~:-":::;r--- - .._..i HOOSI lXerfield ),..) Mille...

'vJ, .~,j '\..,'["~.cfNECTIcJT

{ ~ O,icopee

'. /\.1°,r~'b -if~J~-- c;J:._---,-_._..

CONNM:"nClJT

DRAINAGE BASINS OF MASSACHUSETTS

Figure 1. Location of the East Terrace Locus 1 site and its relationship to the drainage basins.

Copyright © 2004 Mary Lynne Rainey


BULLETIN OF THE MASSACHUSETTS ARCHAEOLOGICAL SOCIETY, VOLUME 65(1) 2004 3

excavation of 5% of the site within theconstruction easement, comprising 105 m2

• Thepipeline route parallels an existing power lineeasement, intersecting a section of the Charlesthat is broad and slow-moving. The Charles isthe longest river in Massachusetts, traversing 80miles on an irregular, winding course through22 communities (Figure 1). The river wasreportedly called the 'Quinobequin', or'meandering river' by local Native people(Bickford and Dymon 1990:62). The pipelineroute through the East Terrace Locus 1 siteintersects a prominent, rocky knoll 240 ft abovesea level composed of large biotite granitesurface boulders, outcrops and gravel as well asoccasional pockets of clay. Along itsnorthwestern perimeter, the rocky terrain dropssteeply in elevation (about 40 ft) to the marginsof the Charles River flood plain. To the southand east away from the river, the terrain isgently sloping to flat, where subsurfaceconditions change dramatically to silty sandwith intermittent boulders.

Feature 1

Prior to fieldwork, several research questionswere identified in the areas of ceramic studies,Native architecture and domestic households,and Late Archaic to Middle Woodland periodsettlement patterns. The data recoveryexcavation targeted the suspected domesticstructure (Feature 1) and other verifiedlocations of features or artifact scatters. A 2 x2 m unit placed adjacent to the suspecteddwelling exposed a circle of large graniteboulders 2 m in diameter with an approximate1 m gap directly on the east side (Figure 2).Homogenous charcoal-rich soils formed anearly continuous band along the northern,western and southern interior sides of theenclosure. The largest boulders were of thesame color, composition, general size andshape. Flat surfaces were tightly fitted in somesections and angled roughly 45 degrees outfrom the floor. The dense charcoal depositsfound within the enclosure at 30 cm belowsurface continued beneath and beyond the

Figure 2. Feature 1, East Terrace Locus 1 Site.

4

enclosure to about 70 cm below surface;evidence that the feature originally consisted ofa large, bowl-shaped pit. Concentrations ofhighly deteriorated granite high in crystallinecontent were found in the center. Throughoutthe initial excavation, only charcoal and rockswere recovered in the one-eighth inchscreening.

As excavations proceeded, the large size andformality of Feature l's construction led to aninitial hypothesis that this might be aTransitional Archaic Period mortuarycomponent, perhaps a primary cremationfacility. Charcoal samples from upper levelswere rushed out for radiocarbon dating andNative American representatives werecontacted. One-eighth inch screening was usedto search for small fragments of calcined bone,fish scales or charred botanical remains,however none were recovered. Once an EarlyWoodland period date of 2100 ± 70 B.P. (Beta133668) was received, the cremation facilityhypothesis was abandoned.

Elsewhere on the site, fieldwork revealed adispersed selection of seemingly unrelatedcultural deposits: a rhyolite workshop on theknoll near the river terrace; a quartz workshopat the opposite end of the site; a small MiddleArchaic period deposit; and an isolated ceramicdeposit. Seven projectile points wererecovered, including one Neville, four OrientFishtail variants, and two non-typed-triangularpoints. Test areas immediately surroundingfeature 1 were either sterile or contained verylow densities of cultural material. Faunalremains were absent and there was virtually noevidence of domestic activity.

Other Interpretations

At the conclusion of fieldwork, the meaningand purpose of Feature 1 remained a mysteryand became the focal point of further inquiry.Soil and charcoal samples underwent flotationanalyses, wood species identification and soilchemistry analyses. One additional charcoalsample from the bottom of the feature (65-70cm below surface) dated to 2050 ± 100 B.P.(Beta-142358), consistent with a date from theupper level of the feature and the associateddeposits found during the site examination.Calendar calibrations of the three dates from

Rainey: Early Woodland Period Ceramic Production Feature

Feature 1 contexts at 95 percent probabilityintersect for a 510-year span during which thefeature was created and used, modified andreused, and finally abandoned.

Flotation analyses of 12 soil samples resulted inrecovery of small amounts of macro-botanicalremains. These consisted of charcoal samples,16 unidentified charred and non-charred plantparts, two charred seeds belonging to the heathfamily, and 3 unidentified charred and noncharred seeds. Insect parts were also recovered.Wood species identification was completed for11 charcoal samples from Feature 1. Thesewere taken in 5 cm levels from a column thatextended from feature surface to its bottom.Lucinda McWeeney (Peabody Museum ofNatural History at Yale University) conductedthe analyses and concluded that all sampleswere derived from trees of the red and whiteoak groups (Quercus sp.). There was noevidence of charred, non-woody plant parts.According to McWeeney, 'the predominance ofoak charcoal recovered from Feature 1 atBellingham suggests that the species wasintentionally selected to produce a hot steady,fire' (Rainey 2000, Appendix E, p. 8).McWeeney and her associate Leo Hickey alsoexamined a few highly deteriorated rockspecimens. A preliminary wash andmicroscopic examination of residue did notprovide any evidence of phytoliths.

Soil chemistry analyses using Lamottes testswere carried out on a column of ten soilsamples and three non-feature control samples.The objective was to look for high calciumand/ or phosphorous readings that might havebeen caused by the burning of human or animalbone. Calcium and phosphorous readings fromthe feature contexts were then compared withsoil chemistry data gathered from human burialcontexts at two other sites using the sameprocedures (Leveillee 1998; Herbster andCherau 2001). The minor fluctuations incalcium were not sufficient to argue thatFeature 1 might be associated with animalprocessing or with the human cremationprocess.

During the data recovery fieldwork andsubsequent analyses, many ideas were posedregarding the purpose and meaning of Feature1. The initial mortuary concept was dismissedbased on the Early Woodland period


radiocarbon dates and the complete absence ofbone. It was also suggested that Feature 1might be the remains of a ceremonial sweatlodge. Based on cumulative lines of evidenceand strongly supported by charcoal analysesthat indicated an extremely hot, oxygen-poorenvironment for an extended period of time,the sweat lodge theory was also dismissed.

Another plausible idea was that Feature 1represented a type of cooking facility in whichlarge-scale plant and/ or animal processing wascarried out. Given the riverside setting, fishroasting or drying was an obvious possibility.Terrestrial species including deer or othermedium to small-sized mammals would havebeen another possible food resource. Elsewherethroughout the region, analyses of macrobotanical and faunal remains, tool assemblages,and associated features, supplemented byNative American ethno-historic accounts, haveprovided good evidence for the constructionand use of stone hearths in domestic contexts.A food-related interpretation for Feature 1seemed logical. Technical parallels existedbetween some of the feature's structuralelements and those of earth ovens used outsidethe region. The Archaic period earth ovens ofthe Southwest, specifically those at Hinds Cavein Texas and at other sites in the lower PecosRiver region, are a good example (Dering1999:659-674). At Hinds cave, one such featureconsisted of a ring of burned rock, 1.2 mdiameter, and radiocarbon dated to 1,820 ± 70B.P. (TX-2733). In addition to projectile points,large quantities of botanical remains wererecovered from two of the Hinds cave earthovens included 40 plant taxa represented by9,576 seeds and fruit fragments (Dering1999:662). Dering also built experimentalversions of these ovens to test models of caloricyield and concluded that these featuresprovided a low return at high cost (1999:673).

Here in the Northeast, evidence for theprocessing and consumption of marine andterrestrial resources in association with hearthfeatures is common. On Martha's Vineyard, forexample, Ritchie excavated hearths and/ orearth oven features at many sites. These heinterpreted as domestic cooking facilities due tothe presence of floral and faunal remains(Ritchie 1969). Some of these Martha'sVineyard features have morphological parallelswith Feature 1 at the East Terrace Locus 1 site.

These include circular plan outlines, burnedrock concentrations and deposits of charredwood. Many were constructed on the groundsurface in shallow, basin-shaped pits. Despitethe similarities, none of the hearths investigatedby Ritchie contained the volume of charredwood found in Feature I, and most wereassociated with extensive food remains. Incontrast, there was little evidence for foodrelated activities in Feature 1. Even with thesite's high soil acidity, some evidence of floralor faunal remains would be expected within thedense charcoal and rock fill if this feature hadbeen used to process bulk food items. Theabsence of food remains was verified throughflotation analyses, wood species identification,and an acid wash of rock samples.

Another factor in the interpretation of Feature 1was the overall cultural context of the site. Noevidence for any kind of aggregated settlementwas found nearby. Furthermore, all of the EarlyWoodland period sites in the immediate areawith evidence of domestic activity (smallhearths, calcined bone or charred botanicalremains, chipped stone tools, etc.) indicate apattern of limited duration campsites andsmall-scale consumption. Although a smallcollection of stone implements was recovered atthe East Terrace Locus 1 site, they represent thecumulative debris from multiple successive sitevisits that took place before, during and afterthe construction of Feature 1.

Evidence for Ceramic Production

After considering ceremonial and subsistencerelated explanations, the idea that Feature 1represented an experimental pottery kiln oropen firing pit was explored. The graniteenclosure, feature placement and nature of thefill are widely recognized aspects of outdoorpit-firing technology, prehistoric and modern.The site's location afforded the potter(s) withnecessary clay and tempering materials, whilemaintaining a safe distance from habitationareas. The feature's placement, set back fromthe exposed and elevated river terrace, wouldserve to minimize wind flow. Reduction orelimination of airflow is required for thesurvival of vessels fired in an outdoor facility(Rice 1987:15). Charred wood samples wereconsistently identified as oak, a preferredspecies commonly referred to in contemporary

6

descriptions of outdoor pottery firingtechniques in other regions. Although conjectural, anomalous high aluminum readingsfrom feature soils may have been caused bydeterioration of clay vessel fragments.

I believe that Feature 1 evolved in two, shortlived experimental stages. Initially, this was alarge fire pit using oak as a primary fuel.Secondary modifications designed to raise andmaintain temperatures included the addition ofa well-constructed granite enclosure andinternal use of smaller rocks high in mineralcontent. When these technical improvementswere made, some of the debris from inside the

o 2 San

~i~r==-o 2..

&s~ Terrace v..."U~ (Sl~ C>..t.l R.rco\~~P·\l

B

o 1 2 3 4 San

I~o I 2in

East Terrace Locus I SilO Oara RecoveryPAL

c

o San

Io I 2in

Eut "reo,,","~ I Slle [)a.L) Reco,"et'),P'l

Figure 3. Sample rim sherds from'dump' fromthe exterior (A), interior (B), and top (C).

Rainey: Early Woodland Period Ceramic Production Feature

pit may have been removed and deposited onthe exterior perimeter of the new structure.Layered deposits found outside the enclosureduring the site examination, and interpreted asa habitation floor, may be explained in thisway. Small charcoal and granite anomaliesfound south of the enclosure during the datarecovery could represent the former locations ofindividual vessels removed from the kiln aspart of a gradual cooling process.

In addition to the feature itself, the discovery ofa discrete vessel fragment dump supports theargument for ceramic production taking placeat the site. Of the 310 vessel fragmentsrecovered during the data recovery, 298 camefrom this one deposit. None of these ceramicsshowed intentional surface decoration. Rather,larger vessel sherds exhibited shallow,patterned relief across one or both surfaces(Figure 3). An unusual serrated argillite toolrecovered from a nearby unit may have servedas a decorating implement (Figure 4).However, since no decorated sherds wererecovered, this is conjectural.

Many of the sherds were characterized bydiagonal, parallel ridges and valleys. These areinterpreted as coils and coil joints, roughlysmoothed by hand or possibly with a paddle orimplement of some kind.

The interior surfaces of some rim fragmentsappear to have been reinforced by handthrough pinching, thereby diminishing thevisibility of the coil joints. The clay used tomake this pottery was tempered with crushedfeldspars and other minerals readily availableon the site. Many of the vessel sherds showbreakage patterns perpendicular to theorientation of coils, and significant colorvariations in cross-section, characteristics thatmay be evidence of waster sherds (Kapches1994). The ceramic fragments were discoveredin a section of the site that was not intensivelyused, about 50 m south of Feature 1. There wasno associated evidence of domestic activity tosuggest that the vessel(s} broke while in use.The absence of additional ceramic deposits maybe attributed to a hypothetical brief period ofkiln operation, and the likelihood thatsuccessfully completed products were carriedaway from the immediate area for distributionand use (Bernardini 2000:366).


ultimate destiny of any wares produced inFeature 1 is equally enigmatic.

After considering all the lines of evidence, Ibelie.ve that the pottery kiln hypothesisprovIdes the most logical explanation for theinternal structure and contents of Feature 1. Itwas built and modified during a relatively briefperiod during the Early Woodland periodabout 2100 B.P. By that time, the concept ofclay vessels had evolved from a simpleimprovement in basket technology to a newand pragmatic household object. Possibly thegeneral. area was known by local potters as aconvement source of the clay, minerals, andwater essential for pottery production. In thisscenario, activities such as raw clay acquisitionfrom the river margins, mineral extraction fromoutcrops for temper, and vessel construction onthe smooth, flat surfaces of outwash boulderswould leave little or no archaeologicalevidence.

The identification of a pottery productionfeature at East Terrace Locus 1 site does not~mply a 'prolif~c and widespread potterymd~stry m Bellmgham or within the regiondunng the Early Woodland period. Feature 1was not designed for large scale production andmay have been a short-lived experiment.Research c?nd.ucted as part of the data recoveryprogram mdIcated that other sites in thegeneral upper Charles River basin may containevidence of similar features and processes. Inthe case of the East Terrace Locus 1 site, thechoice of location reinforces the idea that thissite was part of a larger social and culturalle:md.s~ape dUri~g the Early Woodland period,slgm~cant for Its geological heterogeneity, yetmargmal to any associated domestic center.Whether the individual or group responsiblefor creating and using Feature 1 lived 100 yardsor several miles away cannot be determined.With the exception of the discarded wasters, the

Once Feature 1 was abandoned near the end ofth~ Early Woodland period, there is very littleeVIdence of any subsequent visits to the site.Why this was the case is not known. Elsewherein the upper Charles River drainage, Middleand Late Woodland period activity has beendocumented at many sites including short-termcamps and more complex habitation sites. Bythe Late Woodland period, settlement appearsto be concentrated on the margins of theCharles River estuary (Dincauze 1973).

The recognition of an Early Woodland periodkiln may have limited applicability to the firingtechniques used during the Late Woodland andContact periods. There are no ethno-historicaccount.s that describe specific potteryproduction techniques or firing in southernl'!ew England, a.nd no reported archaeologicalSItes that contaIn evidence of pottery vesselconstruction and firing. As Chilton hasobserved, it is likely that most pottery was firedin multipurpose hearths (1996:40). She citesSagard's well-known 17th century account ofHuron pottery making in which women firedtheir pots in cooking hearths. On Cape Cod,Dunford has argued that pottery firing during~e Late Woo?1and and Contact periods wasmcorporated mto the larger cycle of maizehorticulture, with green pots placed in surfacefires as part of annual field clearing andassociated ceremonial events (Dunfordpersonal communication 3/01).

In the realm of ceramic technology, innovationand craftsman behavior, Feature 1 stands as atechnical and contextual model for futurestudies, especially where similar features might~e expected ~r discovered. As a feature typehkely to be SItuated away from a communitycenter, the chances of discovering comparativedata on other sites may be low, since most CRMmethods are structured around known ore.xp~cted dense cultural deposits found throughhmIted shovel testing. As a result it isimportant to consider the potentialarchaeological sensitivity of areas even whenthey are not obviously conducive to habitation.The social and cultural significance of EarlyWoodland period ceramic production sites suchas Feature 1 will only be possible as additionalinformation comes to light.

Scm4321o

Figure 4. Serrated argillite tool.

Conclusion

8 Rainey: Early Woodland Period Ceramic Production Feature

References Cited

Bernardini, Wesley2000 Kiln Firing Groups: Inter-Household Economic Collaboration and Social Organization in the Northern

American Southwest. American Antiquity 65(2):365-377.

Bickford, Walter E., and Ute Janik Dymon, editors1990 An Atlas ofMassachusetts River Systems, Environmental Designs for the Future. Massachusetts Department of

Fisheries, Wildlife and Environmental Law Enforcement, University of Massachusetts Press, Amherst,MA.

Chilton, Elizabeth S.1996 Embodiments of Choice: Native American Ceramic Diversity in the New England Interior. Unpublished

Ph. D. dissertation, Department of Anthropology, University of Massachusetts, Amherst, MA.

Dering, Phil1999 Earth-Oven Plant Processing in Archaic Period Economies: An example from a Semi-Arid Savannah in

South-Central North America. American Antiquity 64(4):659-674.

Dincauze, Dena F.1968 A Preliminary Report on the Charles River Archaeological Survey. Peabody Museum, Cambridge,

Massachusetts. Manuscript submitted to the National Park Service, Washington, D.C.

1973 Prehistoric Occupation of the Charles River Estuary: A Paleographic Study. Bulletin of the ArchaeologicalSociety ofConnecticut 38:25-29.

1975 Ceramic Shards from the Charles River Drainage. Bulletin of the Archaeological Society ofConnecticut39:5-18.

Herbster, Holly and Suzanne Cherau2001 Partial Relocation Program, Lot 13 Native American Burial Ground (Site 19-DK-144). Squibnocket Ridge,

Chilmark, Massachusetts. PAL Report No. 901, Submitted to Mr. David Wetherell, North Andover, MA.

Kapches, Mirna1994 The Hill Site: A Possible Late Early Iroquoian Ceramic Firing Site in South-Central Ontario. Northeast

Anthropologtj 48:91-102.

Leveillee, Alan1998 "An Old Place, Safe and Quiet": Program of Archaeological Data Recovery, Millbury III Cremation

Complex, Millbury, Massachusetts, Vol. 1. The Public Archaeology Laboratory, Inc., Report No. 396.Submitted to New England Power Service Company, Westborough, MA.

Leveillee, Alan, and Joseph Waller1998 Intensive (Locational) Archaeological Survey of the Proposed Algonquin Gas Transmission Bellingham

Interconnect Pipeline Project Area and Archaeological Site Examination of the West Terrace Site and theEast Terrace Site, Bellingham, Massachusetts. The Public Archaeology Laboratory, Inc., Report No. 890.Submitted to Algonquin Gas Transmission Company, A Duke Energy Company, Boston, MA.

Rainey, Mary Lynne2000 A Program of Archaeological Data Recovery at the East Terrace Locus 1 Site. Algonquin Gas Pipeline

Interconnect, Bellingham, Massachusetts. FERC Docket No.CP98-1000. PAL Report No. 890. Submittedto Algonquin Gas Transmission Company, A Duke Energy Company, Boston, MA.

Rice, Prudence M.1987 Pottery Analysis, A Source Book. The University of Chicago Press, Chicago, IL.

Ritchie, William A.1969 The Archaeology ofMartha's Vineyard. Natural History Press, Garden City, NY.


A Late Woodland and Contact Period Ceramic AssemblageFrom the Hartford Avenue Rockshelter, Uxbridge, MA

Duncan Ritchie

9

Introduction

Rockshelters fonned in either bedrock outcropsor glacial erratic boulders have been a small butvaluable source of information about ancientNative American activities in the elevateduplands of central Massachusetts and RhodeIsland. Many of these sites had multipleepisodes of use from the Middle Archaic to LateWoodland or even Contact period, and wereimportant elements in the cultural landscape ofNative American peoples who used them overmany generations (Arnold 1969; Fowler 1962;Lemire 1975; Waddicor 1969). Over the lastseveral decades, cultural resource managementinvestigations have found and studiedimportant rockshelter sites in central andeastern Massachusetts. Among these are theFlagg Swamp Rockshelter in Marlborough(Huntington 1982), the Tall Pines Rockshelter inClinton (Dudek and Chartier, this volume) andthe recently excavated Den Rock in Lawrence(Carovillano 2003).

Many of these rockshelters had conditions thatpreserved fragile items such as ceramic sherds,faunal and floral remains better than open sitesin the same environment. The HartfordAvenue Rockshelter in Uxbridge yielded adiverse assemblage of well preserved ceramicvessel sherds and faunal remains that illustratesthis pattern. The assemblage of sherds fromthis site is significant for its information aboutLate Woodland to Contact period ceramictechnology, vessel form and function/ use in theinterior uplands of the Blackstone Riverdrainage basin. While this site also contained arelatively large and informative faunalassemblage, the ceramics are the primary focusof this article.

Site Discovery

Archaeological investigations leading to thediscovery of the Hartford Avenue Rockshelterbegan in the spring of 1978 during an intensivePhase I survey of a 13.6 mile long section of theRoute 146 highway corridor. This work was

Copyright © 2004 Duncan Ritchie

conducted by the Public ArchaeologyLaboratory (Brown University) for theMassachusetts Department of Public Works,now the Massachusetts Highway Department.

Six potential rockshelters were identified underoverhangs along the western face of a largebedrock outcrop located about 700 ft north ofthe original Hartford Avenue overpass overRoute 146 in the town of Uxbridge. Only thelargest overhang of the six, designated as theHartford Avenue Rockshelter, appeared tohave been used as a temporary shelter byancient Native Americans. A single test pit inthe floor of this rockshelter produced a bifacefragment of quartzite, a small ceramic sherd,pieces of quartz chipping debris and calcinedbone. The thin, grit-tempered ceramic sherdsuggested that it had been used during the LateWoodland period (Public ArchaeologyLaboratory 1978). During a second phase ofinvestigations in August 1981, several test pitsexcavated into the floor of the rockshelteryielded more quartz and quartzite chippingdebris, ceramic sherds and a variety of wellpreserved faunal remains including almost 100pieces of bone, turtle carapace and freshwatermussel shell. This material was recovered froman organic, midden-like deposit with a maximum thickness of about 25 em (Cox et al.198l).

Burnt rock and calcined bone fragments inoxidized subsoil below this midden-like depositwere recognized as probable feature fill from adeeper occupation zone of unknown age. Itwas apparent that the rockshelter contained atleast two roughly stratified depositions. Thehigh density of ceramic sherds and faunalremains under the rockshelter floor suggested itwas occupied primarily during the LateWoodland period. From this sample of sherdsand bone fragments it was clear that intensiveprocessing of animals hunted for food, theirpreparation and cooking as well as thesubsequent discarding of food remains werethe major activities during the Woodlandperiod occupation of the site.


10 Ritchie: Late Woodland and Contact Period Ceramic Assemblage

The Hartford Avenue Rockshelter wasconsidered to be significant and eligible forlisting on the National and State Register ofHistoric Places given its potential to contributeto under~tanding of ancient Native Americanuse of upland forest environments in thesouthern Worcester Plateau area. A datarecovery plan to mitigate the adverse effect ofconstruction activities, including removal of thebedrock outcrops containing the site, wascompleted in October, 1983 by the PublicArchaeology Laboratory, Inc (PAL) undercontract with the Massachusetts Department ofPublic Works

The research design for the data recoveryprogram posed questions about the types ofactivities that took place at the HartfordAvenue Rockshelter through the prehistoricperiod, and what the depositional history of thesite might reveal about Native American use ofupland interior environments. The site wasexpected to contain evidence of bothabandonment and episodes of intensive useduring the Archaic and Woodland periods. Therockshelter appeared to contain archaeologicaldeposits created during short-term occupationsby groups of hunter / gatherers based in a corearea such as upper Narragansett Bay or thelower Blackstone River while they hunted,foraged or traveled through interior uplands.The presence of shell-tempered ceramic sherdssuggested that Late Woodland people using the

shelter may have been based in a coastal zoneterritory such as Narragansett Bay (Ritchie1985: 12,13).

Physical Setting

The Hartford Avenue Rockshelter was situatedin the middle Blackstone River drainage withinan area of rocky ground moraine depositsmarked by numerous large glacial boulders andbedrock outcrops. Hillsides in this area reachmaximum elevations of about 500 ft above sealevel. The confluence of the Blackstone andMumford rivers is about 1000 ft (300 m)northwest of the site location. The MumfordRiver is one of the primary tributaries in thispart of the Blackstone drainage.

The rockshelter was created by a tilted ridgelike outcrop of Scituate granite gneiss (Zen1983). At one point along a roughly 200 ftsection of this exposure, a large opening hadbeen formed by weathering and faulting in thewest face of the outcrop (Figure 1). A shorttalus slope below the opening was covered withboulders that probably had their origin in thefaulted outcrop. A nearly vertical face about16.5 ft (5 m) in height formed one wall of therockshelter. The rear wall was a low overhangcaused by the upward tilt of the parent outcrop.

The open floor area and talus slope below ithad a westerly aspect and were somewhat

Figure 1. The Hartford Avenue Rockshelter.


exposed to the northwest, however, the outcropprovided good protection from north/ northeastwinds. Beyond the talus slope the terraingradually sloped to the west/ northwesttowards a small area of wooded wetlands.Drainage from these wetlands was orientedwest to Dunleavey Brook, a small tributarystream to the Mumford River located about 1500ft (457 m) from the rockshelter site. Within aone mile (1.6 km) radius of the rockshelter thereis a range of upland forest and wooded wetlandenvironments representing potential habitat fora diverse mix of plant and animal species.Wetlands along the Mumford River also appearto have been potentially important to thevarious prehistoric hunter / gatherer groups thatused the Hartford Avenue Rockshelter. Theassemblage of faunal remains found within therockshelter illustrated the wide range ofmammal and bird species used by ancientNative American occupants of this site.

The historic period construction ofimpoundments such as Lackey Dam about 2.5miles (4.0 km) downstream from therockshelter, have changed the originalappearance of the Mumford River and createdmore open water and marsh than would havebeen present in prehistory. Other historicperiod alterations in the general vicinity appearto have been minimal, limited to forest clearingand the construction of stonewalls. HartfordAvenue, from which the rockshelter derives itsname, was located about 1500 ft (457 m) southof the site. The original Route 146, a two-laneroad, was built in the 1940s when the removal ofbedrock and grading disturbed a narrow stripwithin about 80 ft of the rockshelter. Still, untilits discovery, the Hartford Avenue Rockshelterhad remained in untouched condition despiteits near proximity to Route 146, the mostactively used highway linking the urban centersof Worcester and Providence.

Site Structure and Depositional Patterns

The Hartford Avenue Rockshelter site wasfound to include a total area of about 20 m 2

,

which was substantiallr less than the originalestimate of ± 100 m. The floor of therockshelter was confined to a small triangulararea of about 8 m2 by the two bedrock outcropsforming the back wall of the shelter and severallarge boulders on the floor. This limited floor

space appeared to have resulted in theformation of exceptionally dense amounts ofcultural material, features and faunal remains.The talus slope outside the floor, whichcontained much lower densities of culturalmaterial, was a narrow, rectangular strip about12 m2 bordered by a bedrock outcrop extendingsouth from the rockshelter floor onto the talusslope itself. The talus slope was expected tohave evidence of activities different from thoserepresented in the rockshelter floor.

The floor area and upper talus slope containedconcentrations of cultural material, faunal andfloral remains associated with Terminal Archaicto Early Woodland, Late Woodland andContact period depositions. One distinctconcentration of chipping debris, calcined boneand floral remains occurred within theTerminal Archaic/Early Woodland level. Thebone fragments were from medium and smallmammals such as muskrat, and snake(Colubridae sp.). The floral remains were mostlycarbonized hickory nutshell fragments. Thecentral portion of a Late Woodland middencontained a dense deposit of ceramic sherds,faunal and floral remains and chipping debris.On the upper talus slope a concentration ofceramic sherds around a small hearth featurewas associated with a Contact periodoccupation of the site.

Intensive use of the site during the LateWoodland period, especially the formation of athick midden (20 to 25 em) on the floor of therockshelter, had affected the underlyingTerminal Archaic/Early Woodland deposits.Identified faunal remains from the LateWoodland midden included deer, bear, beaver,muskrat, skunk, and turtle. A small amount offish bone from an unknown species andfreshwater mussel shell were also present. Thedeepest sections of this midden intruded intothe Terminal Archaic/Early Woodlanddeposits, and some of the older chipped stonetools and debitage from these deposits wereincorporated into the midden. A MiddleArchaic Stark projectile point and two TerminalArchaic Orient Fishtail points were found outof their original context in the lower portion ofthe Late Woodland midden. Pieces of bouldersized rock fall had covered and protected thesemidden deposits and their high densities ofceramic sherds, bone fragments and carbonized


floral remains (nutshells) by providing a sealeddepositional context for them. Radiocarbondates placed the formation of the midden atbetween 820 and 540 years ago (1130 to 1410A.D.) with the rock fall having occurred sometime after that.

Ceramic Assemblage

The data recovery program at the HartfordAvenue Rockshelter produced a large and wellpreserved assemblage of prehistoric ceramicvessel sherds from the Late Woodland andContact/Historic periods. The majority of thetotal sample (290 sherds) were reliablyassociated with one of the two features thatrepresent the focal points of activity. Feature 6was the deepest and densest concentration offaunal remains and ceramic sherds within theLate Woodland midden deposit. A radiocarbondate of 740 ± 80 B.P. (Beta 8929) was obtainedon wood charcoal from this feature. Feature 8was a small hearth feature built on the talusslope during a Contact/ Historic periodoccupation. It contained the highest density ofceramic sherds found on the site. Charcoalfrom this hearth was radiocarbon dated to 290 +100 years B.P. (Beta 9255). -

These radiocarbon dated features and sherdconcentrations provided an excellentopportunity to document ceramic vessel form,manufacturing techniques and stylisticattributes from securely dated cultural contexts.Compared with the coastal lowlands ofsoutheastern New England such as BostonHarbor, Cape Cod and Narragansett Bay,Woodland period ceramics from centralMassachusetts are virtually unknown.Therefore, one objective of analysis was tomake comparisons between the ceramic vesselsused at this interior rockshelter site and theLate Woodland and Contact period types fromthe better known coastal sites.

Ceramic sherds were found in all but four ofthe 18 m 2 excavation units. Three of the unitsnot containing sherds were on the lower talusslope outside the rockshelter floor area. Thehighest densities of ceramic sherds occurred inthe two excavation units that contained thefeatures described above. In general, theceramic sherds were tightly clustered aroundthe estimated center of the Late Woodland

midden deposit and several hearth features.With only two exceptions, all the larger (3 to 7em) rim, neck and body sherds were around thecenter of the midden or the two clearlyidentified hearth features (Features 5, 8).Smaller (0 to 3 em) body sherds were morewidely dispersed across the site area, includingthe lower talus slope and edges of therockshelter floor.

The two basic ceramic wares, shell-temperedand grit-tempered, had somewhat differentdistributions. The shell-tempered sherdsassociated with the Late Woodland middendeposit were found mostly in the rockshelterfloor. The grit-tempered sherds from theContact/ Historic period occupation weremostly found in or near Feature 8 on the talusslope, although some were found in the floorarea. Ceramic sherds were distributed throughmost of the vertical levels of the site from 5 to30 em below surface. Most of the sherds wererecovered from the upper portion of the site 5to 20 em below surface. The deepest sherds inLevels 5 and 6 were in feature fill that hadobviously intruded into the underlying archaicdeposition. These basal portions of LateWoodland and Contact period featuresextended to depths of 25 to 30 em belowsurface. There was no evidence to suggest thatany ceramic sherds were associated with theTerminal Archaic/Early Woodland occupation.

The analysis of ceramic attributes was based ona format designed by Dincauze (1975) todescribe prehistoric sherds from sites in theCharles River drainage. This approach wasused to group small samples of sherds intoattribute clusters that, in turn, were assigned togeneral temporal divisions (Early, Middle, Late)within the Woodland period. Although thetemporal affiliation was well established byradiocarbon dates, this format provided a wayto describe sherd attributes in a consistentframework. It also provided comparability at alarger scale since the same attribute format hadbeen used to analyze ceramic sherds from sitesin the Taunton basin (Thorbahn, Cox andRitchie 1983; Thorbahn et al.1982).

The six attributes recorded from ceramic sherdswere: temper (crushed shell, burnt rock/ grit),color (Munsell color codes), thickness (inmillimeters), surface treatment (cord marked,smoothed over cording, plain, etc.),


The decorative elements on this vessel closelyresemble the motifs used on Chance Incisedand Oak Hill Corded, two Late Woodlandtypes common in eastern New York (Funk

lines arranged on the collar in parallel anddiagonal rows. Small notches or tick marksoccurred on the interior and exterior edges ofthe vessel lip. The upper surface of the lip wasplain. On the most intact rim and collar sherds,short diagonal notches along the outside of therim were underlined by three rows of incisedlines. Beneath this, sets of six diagonal lineswere placed in triangular zones defined by linespunctuated with dots (Figure 2A).

These grit-tempered sherds appear to representone vessel associated with Feature 8. The formof this vessel is difficult to reconstruct since fewlarge sherds were recovered. It did have a widedecorated collar below a non-castellated,beveled rim, a slightly constricted neck with nodecoration. The body was probably globular orsemi-globular in form. This vessel was quitesmall with an estimated oral diameter of 5.5inches (13.9 em).

B

A

Figure 2. Rim sherds and reconstructed profilesfor ceramic vessels from the Hartford AvenueRockshelter. Rim sherd from a grit-temperedvessel with incised line decorations (A).Reconstructed rim, neck and shoulder from ashell-tempered vessel with stamped decorationand a cord-marked body (B).

construction mode (coiling, slab construction),and decoration technique (stamping, incisedline, etc.). The attribute analysis isolated twodistinct ceramic wares and vessel formsassociated with the radiocarbon dated LateWoodland and Contact/ Historic features.These two ceramic types are described below.

Vessel construction by some kind of slabtechnique was indicated by grit-temperedsherds with split, eroded surfaces. Thin slabswere probably molded together andcompressed using a paddle and anvil.Decoration was confined to the rim and collarand consisted of fairly wide, shallow incised

Grit-Tempered Ceramics

A total of 176 sherds of grit-tempered warewere found in association with or in closeproximity to Feature 8, the Contact periodhearth on the talus slope. The temper used wasa fine to medium grit or sand ranging in sizefrom 2 mm to small flakes of mica. A fewangular fragments appeared to be very finelycrushed burnt rock. Some of the larger piecesof temper were exposed on the less carefullysmoothed interior surfaces of the sherds. Muchof the mica was probably a natural inclusion inthe clay since very small flecks of it weredispersed throughout the sherds.

Color ranged from pale to brown (lOYR 5/3,6/3) on exterior surfaces. Interior surfacestended to be somewhat darker grey / brown(10YR 5/2, 4/2), mottled grey or dark grey(10YR 4/1) particularly on the body sherds.Sherd thickness varied depending on the partof the vessel from which it derived. Rim, collarand neck sherds were about 6 to 7 mm inthickness. Lower neck sherds tapered veryrapidly from 7 to 3 mm while body sherds wereuniformly thin, averaging about 3 to 4 mm inmaximum thickness. Both interior and exteriorsurfaces were smooth and the neck and bodyareas were undecorated. Faint, parallelstriations are visible on the wiped surfaces,particularly on neck sherds. Wiping of theexterior surfaces seems to have brought the tinymica flakes to the surface giving this ceramicware a shiny, micaceous appearance. The rimand collar areas were carefully smoothed to aslightly burnished surface before anydecoration was applied.


1976). Similar motifs also occur on NewEngland ceramics. Both ceramic types havelinear decoration arranged within triangularzones although on the Oak Hill ware the linesare cord marks rather than incised lines.Chance Incised vessels have three rows ofhorizontal lines below lips, which are decoratedwith small notches like the Hartford AvenueRockshelter vessel. The rockshelter vessel doesnot have basal collar notches, a stylistic traittypical of later New York types such as Garogalike ceramics. It also shares few if anydecorative or stylistic attributes like thecastellations or nodes that appear on Shantoktype ceramic vessels found in Late Woodland toContact period contexts across southeasternNew England (Goodby 2002). Perhaps thegreatest contrast is with the Bear Hollow site innearby Sutton, a Late Woodland to Contactperiod site, which produced a fine sandtempered ware with a plain flat rim and nodecoration of any kind.

Shell-Tempered Ceramics

A large sample of 114 shell-tempered ceramicsherds was found in the Late Woodlandmidden deposit covering the rockshelter floor.Most were from Features 3 and 6 in the centralportion of the midden. The temper used in thisware was medium to fine fragments of shellranging in size from 3 mm to less than 1 mm.The shell temper was evenly distributedthroughout the sherds in fairly dense amounts.The less acidic soil conditions within themidden deposit preserved the shell temper inmost of the sherds and some large fragmentswere occasionally visible on exterior surfaces.Some leaching of temper did occur on the shelltempered sherds deposited near the less denseperimeter of the midden. A few small, roundedsand grains ~ 2 mm in diameter were visible inseveral sherds and may have been naturalinclusions in the clay used to make this ware.

Color on both exterior and interior surfaces wasvery uniform, although organic staining fromthe midden soil matrix probably accounted forsome of this. Exterior surfaces were darkgrayish brown (Munsell1OYR 4/2). Some of theless stained surfaces were yellowish brown(10YR 5/4, 5/6). The leached sherds from theedge of the midden had similar exterior surfacecolors in the brown range (1OYR 4/4, 4/3). The

interior surfaces were generally darker grey toblack (10YR 3/2, 2/1) with visible evidence ofcarbonized organic material on some sherds.

The thickness of this shell-tempered ware didnot vary much over different vessel parts. Rimand neck sherds were all 6 to 7 mm inthickness. Body sherds were slightly thinnerranging from 5 to 4 mm in section. The exteriorsurfaces of all the body sherds were coveredwith partially smoothed over cord marks. Neckand rim areas were left plain to provide spacefor decorative elements and had striated, wipedsurfaces. Some secondary smoothing of theneck area also appears to have taken place afterthe application of decoration.

In comparison with the grit-tempered ware, theshell-tempered sherds showed fewer tendenciesto split and the vessel may have beenconstructed with a coiling technique. Nodefinite evidence of breakage along coil lineswas seen but sherd cross sections did not showthe compressed, laminated structure visible inthe grit-tempered ware.

A single decorative technique, stamping, wasused on the lip, rim and lower neck sections ofthe vessel. The tool used, possibly a smallcomb, had five closely spaced teeth andproduced a series of short diagonal rows. Thelip was covered with even shorter rows slantedin the opposite direction from those on theadjacent surface of the upper rim. Each rowconsisted of small triangular impressions fromthe individual teeth of the stamping tool. Thelongest rows on the lower neck sherdscontained seven triangular impressions, theshort ones on the lip consisted of only two orthree. The impressions on the lip weresmoothed over during final straightening andfinishing of the rim (Figure 2B).

Most, if not all, of the shell-tempered sherdsbelong to one vessel. The large rim and bodysherds were found in Feature 6. This vessel hada narrow, decorated collar above a shallowneck and shoulders that expanded out from theneck into a jar-like body that probably taperedto a narrow, slightly pointed base. The oraldiameter of the vessel was ~ 5 inches (12.7 cm)and it was ~ 10 inches (15 em) in height.

This shell-tempered vessel shares some basicattributes with other Late Woodland ceramics


from coastal Massachusetts. Here a series ofshell-tempered wares with 'cord malleated'bodies, plain necks and collars or rims withcorded, stamped or incised decoration were theprimary Late Woodland types. The closestanalog to the rockshelter vessel may be shelltempered ceramics from Late Woodland siteson Calf Island in Boston Harbor. The CalfIsland vessels had smoothed over cord markedbodies and rims stamped with a toothed objectapplied at an angle leaving rows of triangularimpressions. These rows were sometimesoriented obliquely and the same triangularstamping was used on the vessel lip (Luedtke1980:48, personal communication 4/84). OnMartha's Vineyard shell-tempered vessels withcord malleated exteriors and short oblique orparallel lines of 'corded stick' impressions onthe rim were made by Late Woodland groups.A radiocarbon date of 720 ± 100 years B.P. (Y1852) associated with these ceramics on theVincent site (Ritchie 1969) is very close to one ofthe dates (740 ± 80 B.P.) from the HartfordAvenue Rockshelter.

Summary

Both the incised grit-tempered and stampedshell-tempered vessels recovered from theHartford Avenue Rockshel ter differsignificantly from the other Late Woodlandceramics found in the Taunton basin. Themajority of ceramic sherds from Late Woodlandsites in the 1-495 highway corridor are from finesand and/ or shell-tempered vessels with littleor no decoration. A large sample of severalhundred sherds, all belonging to one fine grittempered vessel, was found in a LateWoodland/ Contact period (435 B.P.) hearth atthe Newcomb Street site in Norton (Thorbahn1982). However, this vessel was undecoratedwith the possible exception of a few faint,impressed lines just below the rim. Dincauze,noting a preference for smooth, surfacetreatment on Late Woodland ceramics from theCharles River drainage, has suggested that thistradition might be centered in the easternMassachusetts/Rhode Island region (1975:14).In contrast, Late Woodland/ Contact period

(550 B.P.) ceramics from the Campbell site inNarragansett, Rhode Island were thick, coarselyshell-tempered wares with channeled, scrapedinterior and exterior surfaces and wide, incisedline decoration (Cox and Thorbahn 1982:92).Clearly, there is a great deal still to beunderstood in terms of sub-regional variationin Late Woodland ceramic technology.

In conclusion, an important feature of therockshelter's ceramic assemblage was itsrelative uniformity. All of the sherds recoveredcould be assigned to one of two basic ceramictypes and these vessels belong to two discreteoccupational episodes. From the availableevidence, it appears that groups visiting therockshelter only carried a limited number ofceramic vessels with them or did not stay at thesite long enough to break more than one or twovessels. It seems likely that the Late Woodlandand Contact period people who used the sitedid so only on a temporary basis and were notyear-round residents. There is evidence thatrockshelters served as temporary camps alongimportant trail routes (Dincauze and Gramly1973). The Hartford Avenue Rockshelter islocated a short distance from Route 16 whichfollows the approximate alignment of the OldConnecticut Path. This was one of the majortrail routes through southeastern New Englandpassing through the present towns of Mendonand Uxbridge. Ultimately, this trail connectedthe Massachusetts Bay (Boston) area with theConnecticut River Valley near Windsor Locks,Connecticut (Ayres 1940). During the LateWoodland and Contact periods, a number ofNative American settlements were located onor near the Path. The final occupation (Feature8) at the Hartford Avenue Rockshelter mayhave resulted from the brief stay of a smallgroup traveling along this trail.

Acknowledgements

Tonya Largy identified the faunal and paleobotanical remains from the Hartford AvenueRockshelter. Jean Blackburn drew the originalillustrations of ceramic sherds and vesselprofiles. Gail Van Dyke of PAL prepared thegraphics for this article.


References Cited

Arnold, Henry N.1969 Arnold Spring Rock Shelter. Bulletin of the Massachusetts Archaeological Society 30(2):1-4.

Ayres, Harral1940 The Great Trail ofNew England. Meador Publishing Co. Inc., Boston, MA.

Carovillano, Jeffrey2003 Adaptation and Resistance: A Contact Period Component at Den Rock, Lawrence, Massachusetts. Bulletin

of the Massachusetts Archaeological Society 64 (1):28-36.

Cox, Deborah c., Joan Gallagher, Alan Leveillee, and Duncan Ritchie1981 Site Examination of the Bear Hollow, Black Bear, Cracker, Dead Dog, and Hartford Avenue Rockshelter

Sites: State Highway Route 146, Sutton to Uxbridge, Massachusetts. The Public Archaeology Laboratory,Inc. Department of Anthropology, Brown University Report. Submitted to Massachusetts Department ofPublic Works, Boston, MA.

Cox, Deborah C. and Peter Thorbahn1982 Prehistoric Archaeological Investigations at Narragansett, Rhode Island: Campbell and Sprague I Sites.

The Public Archaeology Laboratory, Department of Anthropology, Brown University, Providence, RI.Submitted to Lee Pare & Associates, Providence, RI.

Dincauze, Dena F.1975 Ceramic Sherds From the Charles River Drainage. Bulletin of the Archaeological Society ofConnecticut

39:5-18.

Dincauze, Dena and R. M. Grarnly1973 The Powissett Rockshelter, Alternative Behavior Patterns in a Simple Situation. Pennsylvania Archaeologist

43(1):43-61.

Dudek, Martin G. and Craig S. Chartier2004 The Tall Pines Rockshelter, Clinton, MA and Rockshelter Use During the Late Woodland and Contact

Periods. Bulletin of the Massachusetts Archaeology Society 65(1):18-24.

Fowler, William S.1962 Rattlesnake Rock Shelter. Bulletin of the Massachusetts Archaeological Society 26(6&7):49-52.

Funk, Robert E.1976 Recent Contributions to Hudson Valley Prehistory. New York State Museum Memoir 22, Albany, NY.

Goodby, Robert2002 Reconsidering the Shantok Tradition. In: A Lasting Impression, Coastal, Lithic and Ceramic Research in New

England Archaeology. Edited by Jordan Kerber. pp. 141 -154. Praeger, Westport, CT.

Huntington, Frederick W. et a1.1982 Preliminary Report on the Excavation ofFlagg Swamp Rockshelter. Institute for Conservation Archaeology,

Peabody Museum, Harvard University, Cambridge.

Lemire, Raymond1975 The Cracked Rock Shelter. Bulletin ofthe Massachusetts Archaeological Society 37(1&2):20-22.

Luedtke, Barbara1980 The Calf Island Site and the Late Prehistoric Period in Boston Harbor. Man in the Northeast 20:25-76.

Public Archaeology Laboratory1978 Phase I Reconnaissance Survey of Route 146. Report on file, Massachusetts Department of Public Works,

Boston.

BULLETIN OF THE MASSACHUSETIS ARCHAEOLOGICAL SOCIETY, VOLUME 65(1) 2004 17

Ritchie, Duncan1985 Archaeological Investigations at the Hartford Avenue Rockshelter, Uxbridge, Massachusetts. PAL report

no 40-2. Submitted to Massachusetts Department of Public Works, Boston, MA.

Ritchie, William A.1969 The Archaeology ofMartha's Vineyard. Natural History Press, New York.

Thorbahn, Peter F.1982 The Prehistoric Settlement Systems in Southern New England. Final Report of the Interstate Highway

495 Archaeological Data Recovery Program. Volume I. Submitted to the Massachusetts Department ofPublic Works (revised 1984).

Thorbahn, Peter F., Deborah C. Cox and Duncan Ritchie1983 Archaeological Investigations at the G. B. Crane Site, Norton, Massachusetts: The Data Recovery

Program. PAL, Inc. Report No. 13-1. Submitted to Camp, Dresser and McKee, Inc., Boston,Massachusetts.

Waddicor, Arthur1969 Church Brook Rock Shelter No.1. Bulletin ofthe Massachusetts Archaeological Society 30(2):10-15.

Zen, E.1983 A Bedrock Geologic Map ofMassachusetts. U.S. Geological Survey, Commonwealth of Massachusetts,

Department of Public Works, Boston, MA.

18 Dudek and Chartier: Tall Pines Rockshelter and Use

The Tall Pines Rockshelter, Clinton, MA and Rockshelter Use During the LateWoodland and Contact Periods

Martin G. Dudek and Craig S. Chartier

Abstract

Rockshelters have often been important locations forNative American activities. How, when and whyrockshelters came to be used by Native Americanswill be discussed with a focus on one smallrockshelter in Clinton, MA, where part of a brokenpot from the Late Woodland or Contact period wasrecently recovered.

The Tall Pines Rockshelter

The Tall Pines Rockshelter site consists of anexposed outcrop of gneissic schist on a lowerterrace on the east side of Rubens Hill inClinton, MA. Glacial erratics, some withexposed quartz veins, are present on the terraceand nearby slopes. The site was found duringan intensive survey and investigated furtherthrough a site examination. The rockshelter

formation consists of two natural overhangsthat protected two gaps or splits in the rock(Figure 1). A grid of 61 test pits defined thelimits of the site across the terrace, while ten 1m2 units were excavated within and around therockshelter. Both chambers of the rockshelterwere excavated completely. On the terraceadjacent to the rockshelter, shattered quartzwas investigated revealing both naturalweathering and cultural fracturing of quartzprobably for the acquisition of usable chunks.

The north chamber of the rockshelter producedquartz shatter with very few flakes and aconcentration of sherds from a collared NativeAmerican vessel. Several wall niches werepresent in the north chamber, the largest ofwhich could have held a small pot (Figure 2,see next page). Several pieces of the samevessel were also recovered from the south

Figure 1. The Tall Pines Rockshelter.

Copyright © 2004 Martin G. Dudek and Craig S. Chartier



Figure 2. Rockshelter niches.

chamber and beyond it. Southwest of therockshelter was a small area with a few 18thcentury artifacts including burned scratch bluestoneware, redware, flint and wrought ironassociated with calcined bone. This occupationdated from the third quarter of the 18th centuryand does not appear to be associated with theNative American pot.

The Tall Pines Rockshelter Pot

The Tall Pines pot is represented by 109 piecesof which only four are rim sherds. Of this total,95 sherds were recovered from the northchamber, 10 carne from the south chamber andfour fragments were found outside of the rockshelter. All of the sherds have the samecharacteristics of paste, temper, surfacetreatment, thickness and color, suggesting thatonly one vessel lot, probably a single pot, isrepresented. About 20 to 30 percent of the potis present.

A partial reconstruction of the non-castellatedrim, collar and neck indicate a vessel about 15cm in diameter (Figure 3). Vessel thickness

varies from 5 to 7 mm at the rim and from 5 to8.5 mm on the body. All sherds were temperedwith fine grit and occasional larger pieces. Thepot exhibits smoothing from scraping, wipingor brushing on the interior and exterior. Belowthe rim, a horizontal line of small punctationsform an upper border below which are sets ofvertically and diagonally incised lines. Thecollar base is defined by a second, morepronounced set of notches that marks theboundary between the collar and neck.Stylistically, the Tall Pines pot can be dated tothe Terminal Woodland and Contact periods, orbetween the 15th and 17th centuries A.D. Theuse of punctations below the rim and along thecollar base has been reported elsewhere in NewEngland but is unusual in the Worcesterhighlands.

Technologically, the Clinton Tall Pines pot issimilar to other Terminal Woodland andContact period ceramic vessels from theHudson, Housatonic and Connecticut Rivervalleys (Pretola 2000). The pot was constructedby the coiling method, as confirmed throughthin-section analysis by John Pretola (2002).The vessel was also well-fired in an oxidizing

20

atmosphere of sufficient duration to leave onlya narrow black core in the middle of the section.Analysis of the tempering materials indicatedmedium to fine rounded sand grains and arange of minerals including orthoclase feldspar,quartz, clinopyroxene, orthopyroxene, andbiotite mica, probably from a local source.

During the Woodland and Contact periods, thetown of Clinton was well within the homelandof the Nipmuck. Therefore, this pot may be anexample of local Nipmuck ware. A comparisonof Mohawk pottery styles and those of possibleNipmuck origin is informative. Of the Mohawktypes defined by Fowler (1946:3-4), all havethree line linear horizontals, either over incisedvertical and diagonal lines or over chevrons.The Tall Pines pot also has incised markings ona pronounced collar above a constricted neck.However, unlike Mohawk pottery, a line ofpunctations mark the top of the collar instead ofhorizontal lines, while vertical and diagonallines occur in sets below.

Nipmuck pottery may also be represented at

Figure 3. Rim sherd from the Tall Pines pot.

Dudek and Chartier: Tall Pines Rockshelter and Use

the Hartford Avenue Rockshelter in Uxbridge,Massachusetts where incised, grit-temperedpottery was recovered in association with a 290± 100 BP dated feature (Ritchie 1985; 2004).This vessel exhibited several characteristicssimilar to the Tall Pines pot. These include awide collar with incised decoration below anon-castellated rim, a slightly constricted neckwith no decoration, and a body that wasprobably globular or semi-globular, and fine tomedium grit or sand temper. A characteristicthat differs between Tall Pines and HartfordAvenue vessels is the lack of a notched collarbase on the latter. However, it is not clear ifthis portion of the vessel was represented in thesherds recovered.

Comparative Data on Rockshelter Use

A number of previously reported rockshelterswere reviewed for the present work (Table 1).From a survey of 12 rockshelters, it appears thatonly two were definitely used during theMiddle Archaic period, while people usingBrewerton or Squibnocket style points duringthe Late Archaic period occupied eight of thetwelve. All 12 rockshelters were subsequentlyre-used during the Woodland period. Of these,two had Early Woodland components; threehad Middle Woodland occupations while atleast eight were used during the LateWoodland. Of the 12 rockshelters, all but onewere used during the Middle Woodland, theLate Woodland or both.

Late Archaic period rockshelters werecharacterized by lithic production and evidenceof habitation. These sites frequently produceda number of projectile points. In contrast, theMiddle to Late Woodland occupations appearto have been more transient. Diagnostic pointsare correspondingly scarce. At three of therockshelters, pottery was recovered with noassociated lithics. At two of the sites, completetobacco pipes were recovered. Two of therockshelters had natural internal cavities withinwhich 'special' artifacts were recovered. Theseincluded two pipes in one and a pendant in theother.

Three of the rockshelters contained pottery thatappears to have come from individual Middleto Late Woodland pots. The House Rockrockshelter in Millbury contained possible fire-


Table 1. Rockshelter sites with the number of projectile points recovered and pottery occurrences.

Archaic Archaic/ Woodland

Site Name Woodland

Early Middle Late Early Middle Late

House Rock

(Bullen 1948) Pottery

Indian Rock House 3

(Powell 1981) Pottery

Scituate

(Fowler 1965) Pottery

Hartford Avenue 2 2 7* 2,Uxbridge

Pottery(Ritchie 1985)

Flagg Swamp 23 38* 4, 2

(Huntington 1982)Pottery

Bitter Rockshelter 13

(Powell 1965) Pottery(?)

Arnold Spring 1(?) 23

(Arnold 1969) Pottery

Church Brook 12 1

(Waddicor 1969)

Aircraft Road 4 2,

(Zariphes 1970) Pottery

Wilbraham 3,

(Mohnnan 1946) Pottery

Cracked-Rock 3 6

(Lemire 1975) Pottery

Hemlock Boulder(Macpherson and

PotteryRitchie 1998)

*Small-stemmed and Fishtail points are included and may date from the Early Woodland period.

cracked rock, a piece of calcined bone, one pieceof quartz and 46 grit-tempered pottery sherdsthat were probably Late Woodland (Bullen

1948:16). The Hartford Avenue rockshelter inUxbridge contained shell-tempered potteryfrom the Late Woodland period, radiocarbon

22

dated to 740 ± 80 B.P. as well as fine to mediumgrit-tempered pottery from the Contact orHistoric period, radiocarbon dated to 290 ± 100B.P. As noted, the latter pottery vesselexhibited several characteristics similar to theClinton Tall Pines pot (Ritchie 2004). A rockshelter in Scituate investigated by WilliamFowler contained no lithics, but producedconsiderable shell and a pit feature withnumerous pottery sherds from two collaredpots (Fowler 1965:50). The sherds included acastellation and collar design elements withlinear dentate and incised markings stylisticallyLate Woodland. In the same rock shelter, therewas a crevice with an adjoining shelf on theinside of the shelter. Fowler noted that it was anatural 'layaway shelf for objects worthpreserving' (1965:51). On this shelf, he found along stem ceramic pipe and a second shelltempered elbow pipe.

Several other rockshelters show evidence forsimilar caching or ritual-related behavior.While excavating a Late Archaic and LateWoodland rockshelter, Bernard Powell alsofound an artifact cached on 'a shelf', in this casea sandstone celt (Powell 1965:56). A possiblehuman cremation burial was also uncoveredalong with a notched pendent and three roundcobbles found in a deep recess in the wall of theshelter (Powell 1965:60). The Wilbrahamrockshelter, excavated in 1946, included ahuman burial (Mohrman 1946). Pottery andtriangular points found within the rocksheltersuggested a late Middle Woodland to LateWoodland date. The Aircraft Road rockshelterin Middleton, Connecticut contained a middenwith a complete clay elbow pipe, a completesteatite pipe, fragments of a third clay pipe anda Genesee point (Zariphes 1970:18).

As can be seen from this limited survey, Nativepeople in southern New England consideredrock shelters as special places as well ashabitation sites. Several of these rockshelterscontained burials or objects of ceremonialsignificance such as the smoking pipes.

Ethno-historical accounts of NativeAmerican use of rock formations

It is not known exactly what role caves androckshelters played in the seasonal rounds andlandscape narratives of the Native people. It is

Dudek and Chartier: Tall Pines Rockshelter and Use

obvious from the artifacts recovered that Nativepeople recognized and used these locationsboth as sacred places and fortuitous spots tocamp. Some of the ethno-historical accountsthat have been preserved help to shedadditional light on this subject.

The mental as well as physical landscape of theNative world was marked with stones. Thesewere locations where, as Edward Winslowstated in 1624, 'any remarkable act is done'.The acts remembered or commemorated atthese sites may be of a corporal or spiritualnature. The rocks may represent a place suchas the Devil's Den, Rhode Island where a groupof Niantic fooled a band of Mohawks and thussaved themselves (Simmons 1986:273). Anotherexample is Witch Rock in Rochester,Massachusetts, a place where Native 'PauWaus' are said to have sat and watched themist rise from the cracks in this large erratic.In Mashpee, 'sacrifice rocks', as the Englishcalled them, are still recognized. During the19th century, Native people who passed bythese rocks would always cast a stone or stickupon the rock. It was believed that this wasdone in

,acknowledgement of an invisible agent, atoken of the gratitude of the passenger on hisjourney for the good hand of providence overhim thus far, and may imply a mental prayerfor its continuance' (Simmons 1986:253).

A similar story relates to two stones located onthe side of the road from Plymouth toSandwich. In 1807, Kendall reported that

,one of them may be six feet high, and theother four; and both are of ten or twelve ft inlength; and they differ in nothing as to theirfigure, from the masses of granite and otherrock which are scattered over the surface ofall the adjacent country'.

He noted that Indians cast stones and sticks onthese stones because they were told to andbecause they expected blessings from theobservance of the practice, and evils from theneglect. When asked to whom this worshipwas offered? [they replied] 'To a Manitou... ' orspirit (Simmons 1986:253).

In Mashpee, sacrifice rocks eventually evolvedinto what were called 'taverns'. These were


stick and brush structures located along theside of the road where food, property or whiskylibations were offered to the ghosts who werebelieved to wander the roads (Simmons1986:254, 255). The practice of putting whiskeyor alcohol at these sites led to them being calledtaverns. As can be seen, there is a Nativeprecedent for recognizing large stones in thelandscape and acknowledging them as specialsites. Simmons characterized so-called'sacrifice rocks', 'wishing rocks' and 'taverns' asan attempt by the living to keep in touch withthe dead through memorials and shrines(Simmons 1986:251). These observations

suggest that the pots occasionally found inrockshelters, especially when no other evidenceof occupation is present, may indicate that thesevessels were left as offerings, not because theywere broken and discarded.

Acknowledgments

This paper is based on the Tall PinesDevelopment site examination report byTimelines, Inc. (Dudek and Chartier 2002).Information on the pottery analysis comes fromJohn Pretola, Appendix C of the same report.

References Cited

Arnold, Henry N.1969 Arnold Spring Rock Shelter. Bulletin of the Massachusetts Archaeological Society 30(2):1-4.

Bullen, Ripley P.1948 A Rock Shelter near Worcester. Bulletin ofthe Massachusetts Archaeological Sociehj 10(1):16.

Dudek, Martin G. and Craig S. Chartier2002 Site Examination at the Tall Pines Development, Clinton, Massachusetts. Timelines, Inc. report on file

Massachusetts Historical Commission, Boston, MA.

Fowler, William S.1946 Ceramic Design Elements of Massachusetts. Bulletin of the Massachusetts Archaeological Society 8(1): 1-5

1965 Significant Pipe Recoveries. Bulletin of the Massachusetts Archaeological Society 26(3 & 4): 49-52

Huntington, Fred1982 Preliminary Report on the Excavation of Flagg Swamp Rockshelter. Institute for Conservation

Archaeology Report No. 214. Peabody Museum, Harvard University, Cambridge, MA.

Lemire, Raymond1975 The Cracked-Rock Shelter. Bulletin of the Massachusetts Archaeological Society 37(1 & 2):20-22

Macpherson, Jennifer and Duncan Ritchie1998 Intensive Archaeological Survey, Treetops Subdivision, 641 Boston post Road and Adjacent Parcels,

Sudbury, Massachusetts. Public Archaeology Lab, Inc. report on file at the Massachusetts HistoricalCommission, Boston, MA.

Mohrman, Harold W.1946 An Indian Rock Shelter and Burial in Wilbraham, Massachusetts. Bulletin of the Massachusetts

Archaeological Society 8(1):6-7.

Powell, Bernard W.1965 Bitter Rock Shelter: a stratified Connecticut site. Bulletin of the Massachusetts Archaeological Society 26(3 &

4):53-63.

1981 Salvage Archaeology at Indian Rock House. Bulletin of the Massachusetts Archaeological Society 42(2):40-49.

Pretola, John P.2000 Northeastern Ceramic Diversity: An Optical Mineralogy Approach. Ph. D. Dissertation, University of

Massachusetts, Amherst, MA.

24 Dudek and Chartier: Tall Pines Rockshelter and Use

2002 Petrographic Analysis of a Terminal Woodland Ceramic Vessel from the Clinton Rock Shelter. AppendixC in report: Site Examination at the Tall Pines Development, Clinton, MA, by Martin G. Dudek and CraigS. Chartier. Timelines, Inc. report on file at the Massachusetts Historical Commission, Boston, MA

Ritchie, Duncan1985 Archaeological Investigations at the Hartford Avenue Rockshelter, Uxbridge, Massachusetts: The Data

Recovery Program. Public Archaeology Lab, Inc. report on file at the Massachusetts HistoricalCommission, Boston, MA.

2004 A Late Woodland and Contact Period Ceramic Assemblage From the Hartford Avenue Rockshelter,Uxbridge, MA. Bulletin of the Massachusetts Archaeological Society 65(1):9-17.

Simmons, William S.1986 Spirit of the New England Tribes: Indian History and Folklore, 1620-1984. University Press of New England,

Hanover, NH.

Waddicor, Arthur1969 Church Brook Rock Shelter No. 1. Bulletin of the Massachusetts Archaeological Society 30(2):10-15.

Zariphes, Constantine1970 Aircraft Road Rock Shelter. Bulletin of the Massachusetts Archaeological Society 31(3 & 4):16-19.


A Contact Period Fishing Point of Cow BoneFrom Grape Island, Boston Harbor, MA

Craig S. Chartier

Abstract

25

This paper examines a Native-made bone fishingpoint recovered during intensive survey testing onGrape Island. This point, made from a piece of cowmetacarpus, was recovered from a midden contextthat produced Native pottery and lithics as well asEuropean-related artifacts. The point providesevidence for the incorporation of European faunainto the 17th century Native culture. It alsoemphasizes the need to look closely at the faunalremains recovered from 'Late Woodland' sites. Twocase studies demonstrate how faunal remains can beused to distinguish Contact period componentshidden within larger multi-component sites.

Introduction

In May 1997, Timelines, Inc. conductedintensive archaeological survey testing of aportion of Grape Island, one of the islands inthe Boston Harbor Islands State Park. Testingwas required prior to the installation of newbenches near a stone house foundation. As aresult, several features were identified andnumerous artifacts recovered. The majority ofthe artifacts came from a series of shell middensencountered in the three test trenches. Thesemiddens are believed to date from the LateArchaic period to the early 18th century andrepresent the use of this area by Native people,both pre- and post-Contact, as well asEuropeans. Among the artifacts recovered wasa single-barbed bone fishing point. While thisartifact is not unusual in and of itself, the factthat it was made from a fragment of Europeancow bone and associated with Native lithicsand ceramic artifacts is important. The pointwas not identified as made of cow bone wheninitially found and catalogued, but wasrecognized during subsequent analysis of allthe faunal remains from the site. This reporthas three main objectives: to place this unusualartifact within its archaeological context, todiscuss its possible use in terms of Nativefishing technology, and to encourage morecareful examination of other potential Contactperiod faunal assemblages.Copyright © 2004 Craig S. Chartier

Figure 1. Location of the project site onGrape Island.

Site Context

The bone point was recovered from Trench 3,which was located approximately 30 m north ofthe seawall on the south side of Grape Island(Figure 1). An exposed house foundationdating from the 17th to 19th century waslocated roughly 25 m northwest. The trenchwas 1.75 m long and 0.75 m wide, andexcavation was done at 10 cm levels followingnatural stratigraphy. All soil was screenedthrough 1/8" inch hardware cloth (Dudek 1997:54). Trench 3 revealed the followingstratigraphy (see Figure 2, Levels 1-5):

Level 1 at 0 to 10 cm was a disturbed surfacelayer of dark grayish-brown silty sand withmodern shell and recent artifacts, such asmachine-made glass, as well as older faunalremains and prehistoric lithics.

Level 2 at 10 to 20 em was a less disturbed layerof similar soil with increasing evidence ofshell midden. Artifacts recovered included


26 Chartier: Contact Period Fishing Point of Cow Bone

12

3

4

5

Unexcavated

50.

eo

70

80

ClTI. 0 ..f'!N....-.-----,,.,---'--O__-J14......0-:--_--'-!60 8_0__......ii:-.100__--'-__---l.__---.:1....16_0_...;NE Levels

10_.t:sJ~~~~~~.::..:~:;:;::.:~:<.:;;~:;.:t:Tf7T>~~~2:2:.~~~20 -r--=---=-----:~-=----=-::::----=-~::---.,-...,,;......,:.....,,:......;....;.,,:...,..,......,:......:...............:.~ ................:....~~

Figure 2. North wall profile from Grape Island Area A, Trench 3.

faunal remains and lithics like those fromLevell, as well as two fragments of grittempered pottery.

Level 3 at 20 to 30 cm was an undisturbed,moderately dense shell midden composed ofdark grayish-brown silty sand. Numerousartifacts were recovered. These included onehornfels scraper, one quartz uniface, severalutilized flakes and 84 pieces of debitage,representing a wide range of lithic materialbut primarily rhyolite. Four pieces of Nativepottery were found. Two were shelltempered ware; one of these was a rimfragment with a dentate stamped exterior (8mm thick). Other ceramics included onepiece of grit-tempered body (8 mm thick) andone piece of grit and shell-tempered pottery(9 mm thick) with a cord marked exterior.Indigenous fauna included shellfish (mussels,periwinkles and one other gastropod) and afragment of sturgeon scute. Europeanmaterials from Level 3 included a single pieceof lead-glazed redware (with only the interiorsurface present) and seven pieces of cow (Bastaurus) metapodial bone, one of which hadbeen made into a barbed point. The cow boneis described in greater detail below.

Level 4 at 30 to 50 cm was mottled-brownsandy silt with no shell. It did contain lithicdebitage associated with Feature 8, a LateArchaic pit. This feature extended to a depthof 64 cm below surface and contained onepiece of calcined bone and a slight amount of

shell as well as debitage (Dudek 2000:56-57).

Level 5 at 50 to 64 cm was yellow-brown sandysilt that yielded no artifacts.

This stratigraphy indicated three episodes ofuse. Most recent was the disturbed upper layerwith its mix of recent and ancient artifacts.Below this was an undisturbed shell middendating from Late Woodland to 17th century.Portions of this midden post date Europeancontact although it is unclear whether thismeans initial settlement in the area ca. 1628, orthe period ca. 1675-1730, represented elsewhereon the site. The other trenches near thefoundation yielded Native pottery as well asEuropean fauna and ceramics dating to thislater period. Beneath the shell midden was aLate Archaic, or possibly, earlier level. The lateDr. Barbara Luedtke tested this area in the1970s and found a stemmed rhyolite point,possibly a Middle Archaic type, in the lowerlayers of the shell midden she encountered(Luedtke 1975:67).

The Bone Point

Seven pieces of cow bone were recovered fromLevel 3. When comparisons were made withmoose and deer metapodial bones, the formerwere larger and more robust than the excavatedexamples while the latter were smaller andmore gracile. All seven fragments appear tohave come from the mid-section of a singleright metacarpus. Metapodial bones occur in a

BULLEfIN OF THE MASSACHUSEfTS ARCHAEOLOGICAL SOCIEfY, VOLUME 65(1) 2004 27

cow's lower front leg and are analogous tothose that make up the palm of a human hand.Metapodial bones, the metacarpus andmetatarsus, are dense and thick, ideally suitedfor bone tools. Six of the seven pieces werecross-mended forming a section 6.5 cm inlength or roughly one third the length of acomplete metacarpus (as based on acomparative example in the author's collection).

The seventh piece of cow bone may have comefrom the same metacarpus (Figure 3). Thispiece was 5.5 cm long, 1.6 em thick and had aslight degree of curvature that matched theother cow bone fragments recovered from thesame level. It also showed evidence of humanmodification into a barbed point through thegrooving, splintering and scraping techniquesdescribed by Will (2002). Similar single-barbedbone points are common in Late Woodlandshell middens. Several morphologicallyidentical examples were recovered from nearbySpectacle Island (Simon 2002:5-6). This pointwas broken at the barb. Since it is fragmentary,it is not possible to tell whether it wasdiscarded after use or was never completed.

17th Century Native Fishing

This point may have been intended to procuresturgeon. A fragment of sturgeon scute, thebony plates on the sturgeon's back and head,was also recovered from Level 3. Both Nativepeople and English settlers appreciated thesefish that could attain lengths up to 18 ft. TheEnglish so favored this fish that by 1634 thesettlers in Massachusetts Bay pickled much oftheir catch and shipped it back to England(Wood 1977: 55). In England, sturgeon wasused both for food and the production ofisinglass, glue made from their swim bladders(Josselyn 1672: 32). Roger Williams, thereligious dissenter who founded ProvidencePlantation, noted that sturgeon was calledkauposh (singular) or kauposhshauog (plural)by the Narragansett, which he translated as 'hewho is shut up or protected'. Williams alsoobserved that sturgeon 'abounded in diverseparts of this country' and that Natives prized itso highly, that they would not furnish it to theEnglish 'for such a cheap rate that it would beprofitable' for trade (Williams 1973: 100).

Native people developed two special

techniques to catch sturgeon. The first involvedthe use of gill nets that were strung up acrossthe mouth of rivers or sand bars. These strong

Figure 3. The Grape Island bone point.

nets were made from native fibers such asmilkweed, dogbane, false nettle, and possiblybasswood bark. The second technique wasnight fishing from a canoe. Once sand barswere reached, a birch bark torch was ignitedand waved over the surface of the water. It wasbelieved that sturgeon would swim up to thelight and 'tumble and play, turning up hiswhite belly' into which a lance or spear wasthrust (Wood 1977: 107). The preferred lance orspear was described as a 'sharp bearded dart'and was fastened to a long line. Originallythese lance points were made of bone, however,after European contact they were made of ironas well (Wood 1977: 107). Fish spears, oranneganuhtuk, literally 'long spears', were alsoidentified for sturgeon fishing by RogerWilliams (1643) and John Josselyn (1672).Josselyn stated that they would hunt forsturgeon at night

'striking them with a fishgig, a kind of dart orstaff, to the lower end whereof they fasten asharp jagged bone with a string attached to it,[and] as soon as the fish is struck they pullaway the staff, leaving the bony headfastened in the fishes body and the string tothe canoe' Gosselyn 1672: 100).

Implications for Contact period Studies

The bone point from Grape Island illustratesthe need for careful analysis of the faunalassemblages from Late Woodland sites in NewEngland. Grape Island is the second site theauthor has encountered where the faunalremains have helped identify Contact or laterHistoric period components on a site otherwise

28

considered Late Woodland. The secondexample is the Tura site in Kingston,Massachusetts (Chartier 2001). Here thefragmentary remains of a complete Europeanrooster were recovered from a small pit withina larger pit feature. The only other artifactsrecovered from this feature were a few pieces ofshell-tempered Native pottery and some lithicchipping debris. The rooster remains wereinitially identified as a probable duck species.It was not until a complete analysis of thefaunal assemblage from this site had been donethat the skeleton was identified as a rooster(Gallus gallus). As at Grape Island, the featurethat contained the rooster was spatiallyseparate from the other Contact/Plantationperiod materials recovered from the site.Without the rooster's identification, this featurewould have continued to be considered asdating from the Late Woodland.

Chartier: Contact Period Fishing Point of Cow Bone

Conclusion

The occurrence of a probable fishing pointmade from cow bone found on the Grape islandsite, as well as the rooster recovered at the Turasite, serve as cautionary notes to archaeologistsworking on Late Woodland sites in NewEngland. It is important that the faunalassemblages from these sites be evaluatedcarefully as Contact or Historic periodcomponents may be hidden within typical 'LateWoodland' assemblages. Although theevidence for Native use of European livestockis often subtle, this largely unexplored subjecthas great potential for teaching us more aboutthe early stages of acculturation and howNative people responded to and used the novelanimals Europeans brought with them.

References Cited

Chartier, Craig S.2001 Faunal Remains from the Contact Period Tura site, Kingston, Massachusetts, with Special

Emphasis on the Remains of Gallus gallus, the European Domestic Chicken. Paper presented atNEAA conference Bridgewater State College.

Dudek, Martin1997 Report on Intensive (Locational) Archaeological Survey of Bumpkin, Gallops and Grape Islands

Boston Harbor Islands State Park Boston, Massachusetts. Prepared for Massachusetts Departmentof Environmental Management. On file at the Massachusetts Historical Commission, Boston, MA.

Josselyn, John1672 New England's Rarities Discovered. Applewood Books, Distributed by the Globe Pequot Press, Chester,

Connecticut.

Luedtke, Barbara1975 Final Report on the Archaeological and Paleobotanical Resources of Twelve Islands in Boston

Harbor. On file at the Massachusetts Historical Commission, Boston, MA.

Simon, Brona2002 Boston Harbor: The Shape of Things Past and Present. Bulletin of the Massachusetts Archaeological

Society 63(1,2):2-10.

Will, Richard2002 Bone Artifacts and Technological Continuity in Pre-European Archaeological Sites along the

Maine Coast. Paper presented at the Society for American Archaeology Annual Meeting,Denver, CO.

Williams, Roger1973 [1643] A Key Into the Language ofAmerica. Wayne State University Press, Detroit, MI.

Wood, William1977 [1634] New England's Prospect. University of Massachusetts Press, Amherst, MA.

BULLETIN OF THE MASSACHUSEITS ARCHAEOLOGICAL SOCIETY, VOLUME 65(1) 2004

Native American Settlement in the Upper HousatonicDuring the Woodland Period

Timothy Binzen

29

Abstract

Archileological site data from the upper Housatonicwatershed in western Massachusetts andnorthwestern Connecticut indicate that NativeAmerican settlement occurred during the greaterWoodland period (500 to 3,000 years ago) at a rategenerally similar to that seen during the precedingLate Archaic period (3,000 to 6,000 years ago).However, evidence for settlement during the LateWoodland period (after about A.D. 1000) isproportionally much more common in the southernpart of the study area (Connecticut) than in thenorthern part (Massachusetts). This paperconsiders whether this difference is due to the natureof the data set or an actual shift in the settlementsystem in the six centuries before Contact.

IntroductionRecent research in archaeology and ethnohistory has demonstrated that Mohicansettlements were widespread in the HudsonRiver valley of New York State during theContact period, circa A.D. 1500-1600 (Dunn1994; Lesniak 2001). In significant respects,the Mohican settlement system encounteredby the first European explorers likelyembodied patterns that had developed bythe Late Woodland period, which beganshortly before A.D. 1000. Less is currentlyknown about the Native Americansettlements that were located to the east ofthe Hudson, in the upper Housatonic Rivervalley of Massachusetts and Connecticut,during the Late Woodland and Contactperiods. The notion that the upperHousatonic was a cultural "backwater"during the pre-Contact period has beenrefuted (Johnson et al. 1994). However, thebelief persists that the upper Housatonicarea served primarily as a seasonal huntingground that witnessed only intermittentoccupation by Native people in the centuriesprior to the early colonial period.

Recent examination of archaeological sitedata from the Housatonic watershed in

Copyright © 2004 Timothy Binzen

Massachusetts and Connecticut suggests thatWoodland period occupations in theHousatonic Valley were more Widespread thanpreviously has been thought, a patternsupported by research regarding the Mohicansettlements of the early Historic period in theupper Hudson and Housatonic valleys (Dunn2000). This paper reviews archaeologicalevidence from the Housatonic, in order tosuggest patterns related to the settlements ofthe ancestral Mohicans there.

Geographical and Historical Background

The Housatonic watershed is the largest riverdrainage between the Hudson on the west, andthe Connecticut River on the east. TheHousatonic River arises from three ponds, thelargest of which is Onota Lake in Pittsfield,Massachusetts, forming fast-flowing streamsthat unite in the Berkshire Valley. From there,

tI

Figure 1. Study area in the Upper Housatonicwatershed


30 Binzen: Native American Settlement in the Upper Housatonic

the Housatonic meanders through extensivefloodplains, passes through westernConnecticut, and empties into Long IslandSound at Stratford, Connecticut. TheHousatonic watershed occupies nearly twothousand square miles, of which approximatelyone quarter is in Berkshire County,Massachusetts. A small part of the watershed islocated in eastern New York State (Bickfordand Dymon 1990). The study area for thispaper (henceforth "the upper Housatonic") isshown in Figure 1 and consists of theMassachusetts portion of the watershed locatedin Berkshire County, combined with thenorthern half of the Connecticut portion inLitchfield County.

Because the central and northern reaches of theHousatonic were not navigable for Europeanvessels during the early historic period, theriver valley was less well known than theHudson and the Connecticut river valleys.Surprisingly, the upper Housatonic was notexplored by European colonists until the late17th century, and was not extensively settled bythem until the mid-18th century (Binzen 1997).In 1694, the Reverend Benjamin Wadsworthvisited "a place called Ousetonuck formerlyinhabited by Indians" (Smith 1946). Thislocation, believed to be a fording point in GreatBarrington, Massachusetts, is rememberedtoday as the Great Wigwam Site. Wadsworthstated that

"Thro' this place runs a very curious river, thesame which some say runs thro' Stratford,[Connecticut,] and it has on each side someparcels of pleasant, fertile intervale land".

However, he went on to describe the area ingeneral as "a hideous, howling wilderness"(Smith 1946).

Because the upper Housatonic was beyond thefrontier of early New England, thedocumentary record concerning the NativeAmerican communities and villages locatedthere in the Contact period is virtually nonexistent, and even during the early colonialperiod is sparse compared to the Hudson andConnecticut River valleys. This contributed tothe misperception that the Housatonic Valleyhad long been devoid of Native Americaninhabitants, a misperception that wasreinforced in the nineteenth century by

DeForest (1852). In his history of the Indians ofConnecticut, DeForest wrote that in the earlyhistoric period "the whole country now knownas Litchfield County [that is, the Connecticutpart of the upper Housatonic] ...presented anuninhabited wilderness. The birds built theirnests in its forests, without being disturbed bythe smoke of a single wigwam; and the wildbeasts, who made it their home, were startledby nO fires save those of a transient war-party,or a wandering hunter".

Europeans defined the Housatonic from theperspective of their regional settlement centersat Albany, Hartford and Springfield. Separatedfrom the Hudson Valley by the steepescarpment of the Berkshires, the Housatonicvalley also formed the last frontier ofMassachusetts and Connecticut, forming awedge of land unfamiliar to the colonialgovernments. During the early 18th century,this sense of remoteness and distance from theadministrative reach of the colonialgovernments may have appealed to thesachems of the Mohicans (Binzen 1999); in the1730s the Mohican sachems Konkapot andUmpachenee established a new political centerin the upper Housatonic that attracted Nativepeople from the Mohican diaspora and beyond(Frazier 1992).

Despite the frequent discovery of NativeAmerican artifacts in plowed agricultural fields,19th-century historians tended to downplay oreven ignore the Native American heritage ofthe region. By depicting the Native Americansas being primitive, few in number, and animprovident, "vanished" race, histories of thattime helped to rationalize the confiscation ofNative lands that had occurred during theColonial period (Handsman and Richmond1992).

Archival research has provided new insightsinto the lives of the Native people of the upperHousatonic during the early Historic period(Dunn 1994, 2000). The archaeological recordalso provides a unique link to their way of lifeprior to the Contact period, indicating whereand how they lived, and perhaps offering acloser sense of who they were.

"Where are the village sites?"

BULLETIN OF THE MASSACHUSETIS ARCHAEOLOGICAL SOCIETY, VOLUME 65(1) 2004 31

Among the open questions challengingarchaeological inquiry today are the following:Where are the Native American village sites inthe upper Housatonic? If there was a sizeableNative population there during the Woodlandperiod (500 to 3,000 years ago), where is thearchaeological evidence of those communities?

Archaeologists have defined the greaterWoodland period as beginning about threethousand years ago and ending with Europeancontact. The period is divided into the EarlyWoodland (2,000 to 3,000 years ago), MiddleWoodland (1,000 to 2,000 years ago) and LateWoodland (500 to 1,000 years ago) on the basisof changes in Native settlement systems andtechnologies. Archaeology indicates thatduring the greater Woodland period, theNative people of the Northeast manufacturedpottery and adopted maize horticulture to adegree. The use of the bow and arrow (inaddition to spears) began during the Woodlandperiod, and the projectile point types indicativeof occupations during the period includeLevanna, Jack's Reef, Greene, Fox Creek, andRossville points. Across the region, systems oftrade intensified and social relations betweenthe main tribal groups were formalized(Dincauze 1990). People lived in nucleatedvillages, practicing an annual round ofsubsistence that included fishing and hunting,and they favored river valleys and coastal areasfor their major settlements (Lavin and Mozzi1996). However, in most respects the culturalpractices of the Woodland period were thecontinuation of cultural trajectories that hadoriginated much earlier (Feder 1999).

There is a discrepancy, as yet unexplained,between early European explorers' descriptionsof the Algonquian people they encountered,and the archaeological record from the LateWoodland and Contact periods. Although theexplorers described well-populated N ativecommunities, where people cleared andcultivated extensive fields and maintained greatstores of maize, beans and squash (Dunn 1994),maize may not have attained its historicallydocumented importance in Native diet,economy and spirituality until shortly beforethe Contact period (McBride and Dewar 1987).The archaeological evidence for maizecultivation in southern New England hasturned out to be uncommon, and the centralityof maize cultivation in Native subsistence has

been questioned. In western Massachusetts, noarchaeological evidence for large, year-roundhorticultural villages has yet been obtained(Chilton et al. 2000).

Questions have been raised for some timeconcerning this lack of archaeological signs ofLate Woodland village sites in New England(Thorbahn 1988). An interesting set ofexplanations has emerged for this absence. Oneexplanation has to do with the nature ofEuropean settlement. During the Contact andearly Colonial periods, locations of the largestNative settlements typically became centers oftrade between Native Americans andEuropeans. In many parts of southern NewEngland, European settlement followed theAlgonquian pattern, and colonists tookadvantage of prime farmland that had beencleared and prepared by Native people. As aresult, many of the largest Late Woodlandvillages may now be underneath the streets ofcities like Albany and Hartford, unavailable forarchaeological excavation (Snow 1980).

Another possible reason for the lack of evidenceis that in places like the upper Housatonic, themain villages of the Algonquian people werenot the large, palisaded Iroquoian towns oftendepicted in the movies. More likely, theHousatonic villages were smaller clusters ofwigwams (Handsman 1989), and people movedregularly between summer and wintersettlements, using small satellite camps forseasonal subsistence activities (Binzen 1997).Specific main village sites may not have beenoccupied for more than one or two generationsbefore other locations were used nearby, in aform of rotation that precluded the outstrippingof natural resources. Many horticulturalsettlements of the Woodland period may bedeeply buried in floodplain areas, where theyare beyond the access of conventionalarchaeological testing methods (Hasenstab1999). It should be expected that thearchaeological record resulting from a seasonalsettlement system will be subtle, and achallenge to recognize today.

A third factor has to do with archaeologicalpreservation. Four centuries of architecturaldevelopment, intensive farming and collectionof artifacts has resulted in the depletion of thearchaeological record in the Northeast region(Hasenstab 1999). As it has been said, however,


"the absence of evidence is not evidence ofabsence" (Thorbahn 1988). And indeed, recentresearch into the archaeological files fromMassachusetts and northwestern Connecticuthas offered evidence of widespread occupationsin the Housatonic during the Woodland period.

Archaeological Evidence: The WoodlandPeriod in the Upper Housatonic

The study area considered for this paperconsists of the Housatonic River watershed inwestern Massachusetts and northwesternConnecticut, an area referred to as "the upperHousatonic." In his 1980 synthesis of NewEngland archaeology, Snow proposed awatershed-based model for understanding thecultural dynamics of pre-Contact Nativepopulations. It was presumed that thewatersheds occupied by those groups definedthe territories of tribal groups. However, thedistribution of lithic materials and potterystyles in the lower Housatonic suggests that amechanism of cultural interaction overrodethese environmental parameters (Cassedy1996), and it makes intuitive sense that theancestral Mohicans would have used parts ofboth the Hudson and Housatonic river systems.Nonetheless, reference to the Housatonicwatershed transcends the modern politicalboundaries between Massachusetts,Connecticut and New York, and creates a frameof reference that would have been meaningfulto the ancient Native societies.

For this study, the state archaeological site fileswere consulted for 25 Massachusetts towns andeight Connecticut towns in the upperHousatonic. To date, the majority ofarchaeological sites known in the study areawere recorded on the basis of informationobtained from local collectors of NativeAmerican artifacts, and were not initiallyidentified through systematic testing, althoughseveral cultural resource management projectshave provided important overviews of preContact archaeology in the area. These projectshave included archaeological surveys in theMassachusetts towns of Lee (Macomber 1992),Pittsfield (Shaw et a1. 1987), and Sheffield(Nicholas and Mulholland 1987). The mostcomprehensive analysis of pre-Contact NativeAmerican settlement and land use yetproduced in the study area resulted from data

recovery excavations at the Chassell 2 Site (19BK-141) and Kampoosa Bog Site (19-BK-143) inStockbridge, MA Gohnson et al. 1994).Often, the site forms lack detailed informationabout site dimensions and artifact assemblages,and rely heavily on projectile point types todate sites. Five of the Massachusetts towns inthe study area contain no recorded NativeAmerican sites at this time, and three of thetowns contain only one known site. It is notedthat many of the Connecticut sites wererecorded as a result of public outreach effortsby staff of the former American IndianArchaeological Institute in Washington,Connecticut, who were trained in therecognition of Woodland period culturalmaterials. A comparable level of publicoutreach has not yet been attained in BerkshireCounty, Massachusetts, although efforts to thisend would likely have favorable results.

It is probable that the pre-Contact NativeAmerican sites recorded to date in the studyarea represent just a fraction of those thatactually exist. Given the increasing pressures ofresidential and commercial development in theregion, however, there is now an urgent need torecognize and record as many additional sitesas possible, in order to ensure that the culturalresources of the Mohicans and other Nativepeople can be protected and, if necessary,properly investigated (Hasenstab 1999; Binzen2001). While the site files of Massachusetts andConnecticut do provide important locationaldata, there is a great deal of research and siterecording yet to be done to confirm some of thepatterns that are suggested by this preliminaryreview.

The Massachusetts Sites

As of 2001, 112 pre-Contact Native Americanarchaeological sites had been recorded in theMassachusetts portion of the study area (Table1). Of these sites, 32% (T=36) containedevidence of occupation during the greaterWoodland period. Among these Woodlandsites, about one in six (16.6%, T=6) providedevidence of occupation during the LateWoodland period (after A.D. 1000). Overall,however, only 5.4% of all the Massachusettssites currently offered evidence of LateWoodland occupations occurring after A.D.1000.


Table 1. Information concerning pre-Contact Native American archaeological sites in theupper Housatonic study area.

Number of sites Percentage of All SitesInclude

pre-Contact Woodland Late Woodland Late Woodland

Site Files sites in Upper period Woodland period periodHousatonic component( period component( component(s)study area s) component(s) s)

Massachusetts 112 36 6 32.1 5.4(as of 2001)Connecticut 85 29 21 34.1 24.7(as of 1995)

RecentNW 17 5 3 29.4 17.6Connecticut

Outreach(Binzen 2002)

Totals 214 70 30 32.7 14.0

33

The Connecticut Sites

As of 1995, 85 pre-Contact Native Americansites had been recorded in the Connecticutportion of the study area (Table 1). Of thesesites, 34% (T=29) contained evidence ofoccupation during the greater Woodlandperiod. This is virtually the same proportionseen in Massachusetts. Among theseWoodland sites in Connecticut, however, nearlythree quarters (72%, T=21) provided evidenceof occupation during the Late Woodlandperiod, that is, after A.D. 1000. This is afrequency five times greater than that which isseen for the Massachusetts sites of the greaterWoodland period. Overall, one quarter (24.7%)of all the recorded Connecticut sites hadevidence of Late Woodland occupation.

Recent Outreach Sites in Connecticut

Recently, a public outreach event was held inthe town of Salisbury in Litchfield County,Connecticut (Binzen 2002). Members of thepublic were invited to bring in NativeAmerican artifacts that they had found foridentification (typically projectile points fromagricultural fields), and plot the find-spots ontopographic maps. Seventeen previouslyunrecorded pre-Contact Native American sitesin the towns of Salisbury, Canaan and NorthCanaan were recorded. (Updates were

obtained for three previously recorded sites.)Evidence for Woodland period occupation wasreported from 5 of the sites. Of theseWoodland sites, 3 had evidence of LateWoodland occupation.

Summary of Pre-Contact Site Information

In the overall upper Housatonic study area of33 towns, a total of 214 pre-Contact NativeAmerican archaeological sites has beenrecorded. Of these sites, one third containedevidence of occupation during the greaterWoodland period, which began about 3,000years ago. Among the sites of the greaterWoodland period, close to half providedevidence of occupation during the LateWoodland period, after about A.D. 1000.Overall, 14% of all the recorded sites in thestudy area have provided evidence of NativeAmerican occupation(s) during the LateWoodland period, or the six centuries leadingup to first contact between the NativeAmericans and the Europeans.

Patterns from the Site Data in the UpperHousatonic

The archaeological evidence indicates that thefrequency of occupations during the greaterWoodland period (500 to 3,000 years ago) isvirtually identical among sites in the


Massachusetts and Connecticut portions of thestudy area (Table 1). Proportionally, however,evidence for Late Woodland occupations(occurring after A.D. 1000) has been reported atfive times more of the archaeological sites innorthwestern Connecticut than to the north inMassachusetts. This would seem to suggestthat Native American settlement in thenorthernmost quarter of the Housatonicwatershed was comparatively sparse after AD.1000. Johnson (1994) suggests that with theadoption of a seasonal round of horticultureand hunting during the Late Woodland period,the Native people of the Housatonic movedseasonally between separate settlements in theriver valley and upland areas. It is possible thatthe Late Woodland sites in westernMassachusetts, though fewer in number, werelarger, more centralized settlements than thecontemporary sites in western Connecticut.Alternative explanations involve the possibilityof a proportionally smaller population in thenorthern part of the watershed; the apparentconcentration of Native populations near thecoast and in the valleys of the Hudson andConnecticut rivers during the Woodlandperiod; and the possibility that archaeologicalevidence for Late Woodland occupationssimply has had greater visibility in the lowerpart of the watershed.

The Late Woodland sites of the Housatonic inMassachusetts and extreme northwesternConnecticut may have been occupied by Nativepeople who were affiliated with the Mohicansociety, while the sites in the Connecticutportion of the study area may have been small,seasonal, short-term habitations, used bypeople who were affiliated more closely withthe Native communities of the lowerHousatonic valley and the Connecticut coast.This possibility is supported by a previousstudy of the distribution of lithic materials andpottery styles in the lower Housatonic, whichsuggested that the Native people of the upperpart of the watershed interacted more closelywith the Mohicans of the Hudson (Cassedy1996). In Stockbridge, Johnson (1994) reportedthe presence of a variety of lithic material(chalcedony) traded or transported from theKent, Connecticut area on the middleHousatonic. This suggests cultural connectionsbetween the Native people of westernMassachusetts and those living downriver, tothe south. Archaeological evidence supports

"historically documented, traditionally recalledties" between the Native people of Stockbridgeand the Native communities of northwesternConnecticut and western Massachusetts(Johnson 1994, citing Brasser 1974, 1978; Frazier1992; Handsman and Lamb Richmond 1992).

Is there a natural landmark that symbolized apoint of transition between the Native groupsof the northern and middle Housatonic?Pawachtuek, the Great Falls on the Housatonicin Canaan, Connecticut (Dunn 1994), has thegreatest drop in elevation on any major river inNew England. It can be speculated that thislandscape feature represented a gateway toMohican country for Native people whotraveled up the river from the south. Evidenceof Mohican influence in the upper Housatonicnorth of Pawachtuek is provided by earlydocuments from Albany, which demonstratethat the series of riverside flats upstream fromthe Great Falls all had distinct Mohican placenames at least by the late 17th century:Kenachkehantick, Achneganick, Awaankaniss,and Taashammik (Dunn 1994). It seems likelythat these places along the Housatonic had beennamed by the ancestral Mohicans manygenerations earlier. The Mohican presenceupriver in the Massachusetts towns of Sheffield,Great Barrington, Stockbridge and adjacentparts of New York between 1675 and 1750 hasalso been demonstrated (Dunn 1994, 2000;Binzen 1997). As Johnson (1994) observed,archaeological evidence from the upperHousatonic supports the tradition that theNative people of western Massachusetts hadstronger cultural ties to the Hudson Valley thanto the Connecticut Valley during the historicalperiod, "ties that extend deep into the remotepast."

Al though evidence for Late Archaicoccupations that occurred three to six thousandyears ago is very common in the study area,there is no indication that the rate of occupationsignificantly increased or decreased during thesubsequent Woodland period. The soleexception to this observation is the relativescarcity of Late Woodland sites in the northern,Massachusetts quarter of the watershed. Thiswas clearly a time when Native settlementintensified in the lower Housatonic and incoastal Connecticut. Perhaps a re-orientation ofNative settlement towards the lowerHousatonic, combined with an intensification of


horticulture in the Hudson Valley, attractedNative people from the northern Housatonicand resulted in the partial depopulation of thestudy area by the ancestral Mohicans after AD.1000. During the Colonial period, however, thestrategic advantages of settlement in the remote"hunting grounds" of the Housatonic wereonce again recognized by the Mohican people(Binzen 1999). With the ascent of the fur tradein the 17th century, moreover, control ofheadwater areas had become a new priority forthe Native people of southern New England(McBride and Soulsby 1989). The upperHousatonic area may have regained logisticalsignificance for this reason also.

While it is possible that the Native populationin the northern part of the Housatonicwatershed decreased after AD. 1000, peoplecertainly did not disappear. In Massachusetts,several towns have sites with evidence ofNative occupation during the Late Woodlandperiod. These towns are Great Barrington,which includes the Skatekook Site (19-BK-28),the Great Wigwam Site (19-BK-25), and the Mt.Peter Site (l9-BK-108); Sheffield, with theClark's Field Site (19-BK-101) and the ChapinFarm Site (19-BK-103); and near the headwatersof the Housatonic in Pittsfield, with theCaldwell Site (19-BK-137), the Village Site (19BK-5) and the Canoe Meadows Site (19-BK-13).

In northwestern Connecticut, the places thatwere favored for habitation during the LateWoodland are distributed along the floodplainsand terraces of the Housatonic and itstributaries in Cornwall, Canaan, North Canaan,Salisbury and Sharon; the vicinity of the TwinLakes and Lake Wononscopomuc in Salisbury;Lake Waramaug in Warren; and Bantam Lakein Litchfield.

Artifacts other than projectile points canprovide insights into the Woodland period.Native American pottery is a well-knownindicator of Woodland period occupations inthe region. It is noteworthy that pottery hasbeen reported from only four sites in theMassachusetts portion of the study area. It maybe that this type of artifact has goneunrecognized or unreported at other Woodlandperiod sites. Constituting one of the fewsources of information about stylistic trendsand ethnic affiliations, Native pottery meritsfurther investigation in the upper Housatonic.

One of the most interesting secondary patternsto emerge from the Housatonic study involvesthe frequent occurrence of pestles. Thesetapered, cylindrical implements of workedstone were used to grind food materials, andare often associated with societies that practicehorticulture. Frequently seen at sites of theWoodland period, pestles have sometimes beenfound in association with women in funerarycontexts (Gibson 1980). Pestles have beenreported from one quarter (25%, T=9) of thesites that contain Woodland period componentsin the Massachusetts portion of the study area.Among related implements, stone hoes werereported from sites in Massachusetts andConnecticut, and one mortar stone wasreported.

Artifacts that may provide a glimpse into thesymbolic and ritual aspects of Mohican lifewaysin the upper Housatonic include a pestle withan animal head from a site in Great Barrington(l9-BK-112), and a set of bear teeth with drillholes that evidently formed a necklace, from asite in Pittsfield (l9-BK-171). Animal symbolswere associated with the bear, turkey, deer,wolf and turtle clans in Mohican society (Dunn2000).

Conclusions

In conclusion, archaeological site data indicatethat Native American occupations did occur inthe upper Housatonic study area during thegreater Woodland period (500 to 3,000 yearsago). However, evidence for occupations thatoccurred during the Late Woodland period(after AD. 1000) has been recorded in markedlyfewer locations in the northern part of thewatershed. As the regional trade andcommunication networks of the Woodlandperiod developed, the people of the upper partof the watershed appear to have had a closersocial affiliation with the ancestral Mohicans ofthe Hudson Valley to the west than with thepeople of the Connecticut River Valley to theeast. The people of the middle and lower partof the Housatonic watershed were probablyaffiliated with the large Native communities ofthe southern Housatonic Valley and theConnecticut coast. Native occupation of thelower Housatonic watershed in Connecticut


apparently continued at a steady rate duringthe Early and Middle Woodland periods, evenintensifying during the Late Woodland. In theupper part of the watershed in Massachusetts,however, the number of Native American sites(and presumably the amount of settlement)appears to have decreased during the LateWoodland period, or it became concentrated ata smaller number of main villages in the rivervalley.

same period, Mohican leaders recognized thestrategic benefits of re-settlement in the upperHousatonic. Although the Mohican town atStockbridge may have been newly establishedin the 1700s, the community made use of asystem of Native settlement, travel and landuse which in many respects had first emergedin the Housatonic during the Woodland period.

Acknowledgements

This paper was originally presented at theannual Mohican Seminar, New York StateMuseum, Albany, New York in March 2002. Amodified version may be included in a futureNew York State Museum Bulletin.

When colonists from New York, Massachusettsand Connecticut explored the upperHousatonic area in the late 17th and early 18thcenturies, they documented vast tracts offorested land, but also many open meadowsand the settlement areas of Native people whoidentified themselves as Mohicans. During the

References Cited

Binzen, T.1997 Mohican Lands and Colonial Corners: Weataug, Wechquadnach and the Connecticut Colony, 1675-1750.

M. A. Thesis, Department of Anthropology, University of Connecticut, Storrs.

1999 Living in the Borderlands: Native American Communities of the Upper Housatonic During the ColonialPeriod. Paper presented at the Annual Meeting of the American Society for Ethnohistory, MashantucketPequot Museum and Research Center, Mashantucket, Connecticut.

2001 Mohican Communities of the Housatonic: An Archaeological and Historical Perspective. Paper presentedat Many Trails of the Mohican Nation: A Conference on Mohican History and Culture. Bowler, WI.

2002 Results of an Archaeological Public Outreach Event in Salisbury, Litchfield County, Connecticut.UnpUblished manuscript in possession of the author.

Bickford, W. and U. Dymon1990 An Atlas ofMassachusetts River Systems: Environmental Designs for the Future. University of Massachusetts

Press, Amherst.

Brasser, T.1974 Riding on the Frontier's Crest: Mahican Indian Culture and Culture Change. National Museum of Man

Mercury Series, Canadian Ethnology Service Paper No. 13, Ottawa.

1978 Mahican. In Northeast, edited by B. Trigger, pp. 198-212. Handbook of North American Indians, vol. 15,W. G. Sturtevant, general editor. Smithsonian Institution, Washington, D.C.

Cassedy, D.1996 Prehistoric Interaction Between Eastern New York and Southern New England. Paper presented at the

Annual Meeting of the Society for American Archaeology, New Orleans, Louisiana.

Chilton, E., T. Largy and K. Curran2000 Evidence for Prehistoric Maize Horticulture at the Pine Hill Site, Deerfield, Massachusetts. Northeast

Anthropology 59.Connecticut Archaeological Site Files

Various dates. Archaeological site files of the State of Connecticut, located at the Office of the Connecticut StateArchaeologist, Storrs.


DeForest, J.1852 History of the Indians ofConnecticut from the Earliest Known Period to 1850. William Hamersley, Hartford,

Connecticut.

Dincauze, D.1990 A Capsule Prehistory of Southern New England. In: The Pequot: The Rise and Fall ofan American Indian

Nation. Ed. By L. Hauptman and J. Wherry, pp. 19-32. University of Oklahoma Press, Norman.

Dunn, S.1994 The Mohicans and Their Land, 1609-1730. Purple Mountain Press, Fleischmanns, New York.

2000 The Mohican World, 1680-1750. Purple Mountain Press, Fleischmanns, New York.

Feder, K.1999 The Late Woodland Revisited. Bulletin ofthe Archaeological Society ofConnecticut 62.

Frazier, P.1992 The Mohicans of Stockbridge. University of Nebraska Press, Lincoln and London, Nebraska.

Gibson, S., ed.1980 Burr's Hill: A 17th-Century Wampanoag Burial Ground in Warren, Rhode Island. Haffenreffer Museum of

Anthropology, Brown University, Providence, Rhode Island.

Handsman, R.1989 The Fort Hill Project: Native Americans in Western Connecticut and an Archaeology of Living Traditions.

American Indian Archaeological Institute, Washington, Connecticut.

37

Handsman, R. and T. Lamb Richmond1992 Confronting Colonialism: The Mahican and Schaghticoke Peoples and Us. Essay prepared for Making

Alternative Histories, an Advanced Seminar at the School of American Research, Santa Fe, New Mexico.

Hasenstab, R.1999 Fishing, Farming and Finding the Village Sites: Centering Late Woodland New England Algonquians. In:

The Archaeological Northeast, edited by M. Levine, K. Sassaman, and M. Nassaney, pp. 139-54. Bergin andGarvey, Westport, Connecticut.

Johnson, E., E. Chilton, C. Laing and T. Largy1994 Data Recovery Operations at the Chassell 2 (19-BK-141) and Kampoosa Bog (19-BK-143) Sites,

Stockbridge, Massachusetts. Copy on file at the Massachusetts Historical Commission, Boston.

Lavin, L. and M. Mozzi1996 Historic Preservation in Connecticut, Volume I, Western Coastal Slope: Overview of Prehistoric and

Historic Archaeology and Management Guide. Connecticut Historical Commission, Hartford.

Lesniak, M.2001 Contract Archaeology and Precontact Sites in the Mid-Hudson Valley. Paper presented at Many Trails of

the Mohican Nation: A Conference on Mohican History and Culture. Bowler, Wisconsin.

Macomber, G.1992 Archaeological Site Examination of Seven Prehistoric Sites in Proposed Segment 2X of Tennessee Gas

Company's Northeast Settlement Project-Phase II in Lee and Tyringham, Massachusetts. Copy on file atthe Massachusetts Historical Commission, Boston.

Massachusetts Archaeological Site FilesVarious dates. Archaeological site files of the Commonwealth of Massachusetts, located at the Massachusetts

Historical Commission, Boston.


McBride, K. and R. Dewar1987 Agriculture and Cultural Evolution: Causes and Effects in the Lower Connecticut Valley. In: Emergent

Horticultural Economies of the Eastern Woodlands, edited by W. F. Keegan, pp. 305-328. Southern lllinoisUniversity at Carbondale Center for Archaeological Investigations Occasional Paper No.7.

McBride, K. and M. Soulsby1989 Prehistory of Eastern Connecticut: Phase I, II and III Archaeological Surveys for Relocation of Route 6 / 1

84 Project, Vol. I. Submitted to the Connecticut Department of Transportation, Hartford, CT.

Nicholas, G. and M. Mulholland1987 Archaeological Locational Survey of the Housatunnuk at Skatehook in Sheffield, Massachusetts. Copy on

file at the Massachusetts Historical Commission, Boston.

Shaw, L., E. Savulis, M. Mulholland and G. Nicholas1987 Archaeological Locational Survey in the Central Berkshires, Pittsfield, Massachusetts. Copy on file at the

Massachusetts Historical Commission, Boston. MA.

Smith, C.1946 The Housatonic, Puritan River. J.J. Little and Ives Company, New York.

Snow, D.1980 The Archaeology ofNew England. Academic Press, New York.

Thorbahn, P.1988 Where Are the Late Woodland Villages in Southern New England? Bulletin of the Massachusetts

Archaeological Society 49(2):46-56.


Betty Little: An Appreciation

Dena F. Dincauze

39

How does one eulogize a physicist who cameinto the social sciences through the garden doorand rapidly became indispensable?Acquaintance with Betty was among the mostgratifying adventures of my life; for more than25 years she was student and mentor, colleagueand friend. Her passion for understanding theworld was fuelled by a childlike universalcuriosity, supported by formidable analyticalskills and nearly tireless energy. In short, shewas the ideal scientist: curious abouteverything, observant, alert, analytic, energetic,resourceful.

A born teacher, she volunteered in schoolswhere she effectively instilled questioninghabits in children, as well as in college studentsat UMass, where she did her best to demystifymathematics for anthropologists. She taughtherself the technicalities of radiocarbon datingand persevered to publish articles in journalsrespected by specialists in many fields. Hersuccessful private study of 17th century Englishhandwriting allowed her to make originalcontributions to New England history from offshore Nantucket, her major laboratory.

In honoring this unique human being, I setaside the temptation to merely list the titles ofher many and diverse publications, and instead,offer select anecdotes from a single week in1990 during which four women touredNewfoundland and Labrador, visitingarchaeological sites of Indians, Eskimos,Vikings, Basques, English and French explorers.As new to the place as we all were, Bettyprovided an informed and runningcommentary on the vegetation, wildlife, andcultural landscapes.

• Near a ferry slip in Quebec, she requested astop to investigate seal skins stretched on racksfor drying and curing. Despite the thick blackflies, Betty engaged the householder inconversation about the legalities of seal huntingand the process of curing skins. The man'syoung daughter was so interested that shestood around us in the cloud of flies, listening.

Copyright © 2004 Dena F. Dincauze

• Near the Puffin rocks, Betty was the onewho took off her shoes to wade briefly in thechill, exotic waters of the Gulf of St. Lawrence.

• At the tip of Newfoundland, Parks Canadabuilt a group of Viking turf-covered houses tocreate a successful tourist attraction. Minutelyscrutinizing the grass growing on them, Bettyasked the baffled guide why that particulargrass was used for the reconstruction. Sheexplained that, since Timothy grass was notbrought to North America until the 18thcentury, it could not be authentic for Vikingbuildings. She was the first visitor to notice thecritical anomaly.

• At an early Eskimo site on the coast, Bettyvolunteered to join the excavators in a middenthick with seal bones - the kind of dirty workthat she loved. Later, we all stood agape as aGulf of St. Lawrence sunset laid a broad,straight, brilliant orange carpet down the lowerGulf, reaching east from the river towards theshore of Newfoundland. As we watched, twowhales breached and leaped through theorange stripe, taking our breath away. 'Finwhales' remarked Betty quietly.

• Betty realized that the small square ' cellarholes' we observed in a coastal meadow mightbe the products of early post-Viking explorersin the region. Later inquiries proved herspeculation accurate. She also called myattention to stone alignments across a brook,relating them to my ongoing interest in thefunction of simple fish weirs.

• A few days after the trip, Betty sent eachvoyager pages of notes on the groundvegetation we had crushed underfoot, markingthe species that were found on Nantucket (theancient edge of the glaciers).

One couldn't ask for a better companion,whatever the activity. Betty enriched everyonewho knew her.


40

CONTRIBUTORS

MARY LYNNE RAINEY is a Project Archaeologist with the Public Archaeology Lab, Inc. and hasworked in cultural resource management since 1979. She received a B.A. in Anthropology from WestChester University, Pennsylvania in 1980, and is currently completing a graduate degree inAnthropology at the University of Connecticut.

DUNCAN RITCHIE is a Senior Archaeologist with the Public Archaeology Lab, Inc. He received anM.A. in Anthropology from Brown University and has been a long-time member of the MassachusettsArchaeological Society as well as a past contributor to the Bulletin.

MARTIN G. DUDEK is a Senior Archaeologist with Timelines, Inc. He received a B.A. inAnthropology at University of Maryland in 1985 and a M.A. in Anthropology from Brandeis Universityin 1992. He has worked on archaeological projects in the Mid-Atlantic states, New York, Alaska,California, Honduras and Mexico, with a focus in New England since 1992.

CRAIG S. CHARTIER is Principal Investigator with the Plymouth Archaeological Rediscovery Projectin New Bedford, M.A. He received a B.A. in Anthropology from the University of Rhode Island in 1993and an M.A. in Historical Archaeology from University of Massachusetts, Boston in 2000.

TIMOTHY BINZEN is a Staff Archaeologist at the University of Massachusetts Archaeological Services,Department of Anthropology, University of Massachusetts, Amherst. He received his M.A. inAnthropology from the University of Connecticut in 1997.

DENA F. DINCAUZE is Professor Emerita of Anthropology at the University of Massachusetts,Amherst. She is past Editor of the Bulletin and Past President of the Society for American Archaeology.


ISSN 0148 1886

Documents

BULLETIN OF THE MASSACHUSETTS ARCHAEOLOGICAL SOCIETY