PARSONS SITE CERAMIC VESSELS Ronald F. Williamson and Terry G. Powis INTRODUCTION A total of 2,563 sherds, forming portions of vessel rims, necks, shoulders and bodies, individually or in various combinations, consti- tute the ceramic assemblage. Neck and shoul- der sherds number 384 and 103 respectively, while there were 32 neck/shoulder sherds. The assemblage also includes 1,874 analyzable body sherds of which 1,783 (95 percent) exhib- ited a plain surface appearance. These counts are based on sherds that are generally larger than 12 mm in diameter. Although there are a significant number of smaller fragments (634), they are typically too small to provide reliable surface treatment data. VESSEL RIMS Two hundred and one rim sherds were recovered, forming portions of 171 vessels. Rims were considered analyzable when they exhibited both exterior and interior rim sur- faces, the lip and a sufficient portion of the outer collar-neck surface to allow decorative styles and attributes to be ascertained. While an additional 493 rim fragments were recov- ered, they are too incomplete to make any reliable observations regarding descriptive attributes. Analytical data were entered directly into a computer database file, which was used to generate both an artifact catalogue and an attribute analysis. The database consists of seven attribute fields, recording the nature of the specimen, portion, shape or form of the portions, metrics, vessel zone decoration (area, motifs and techniques), and castellation decoration (area, motifs and techniques). Design Motifs The vessels in the assemblage exhibit five main design motifs from which all variations were derived. Simple. These are motifs that exhibit one or more horizontal bands of oblique or vertical lines executed in the same direction. Opposed. These motifs consist of two or more alternating bands of simples that repeat- edly change direction. Horizontals. These are lines or motifs that are predominantly characterized by horizontal elements. Hatched. These motifs are simples that are crossed by other simple elements in different directions. Plain. These lack any decorative motifs. Design Techniques Concomitantly, the vessels were decorated using five main design techniques from which all variations were derived. Incising. This is a technique accomplished by drawing a sharp pointed object across wet clay. Linear Stamp. Various impressions are created by pressing a linear object into the clay, leaving a distinguishable impression. Linear Punctate. Punctate impressions using a linear object, often used for notching a collar edge. Punctation. Punctation is accomplished by punching the clay with various tools at differing angles. While annular punctates are likely produced by hollow bone tubes, linear punc- tates are produced by hollow linear objects. Cord-wrapped stick. These are impressions produced by the application of a cord-wound stick or twig. The stick was presumably wrap- ped with a fibre of some type and was im- pressed upon the wet clay. WILLIAMSON AND POWIS ______ PARSONS SITE CERAMIC VESSELS ____________________________ 53
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
PARSONS SITE CERAMIC VESSELS
Ronald F. Williamson and Terry G. Powis
INTRODUCTION
A total of 2,563 sherds, forming portions ofvessel rims, necks, shoulders and bodies,individually or in various combinations, consti-tute the ceramic assemblage. Neck and shoul-der sherds number 384 and 103 respectively,while there were 32 neck/shoulder sherds. Theassemblage also includes 1,874 analyzablebody sherds of which 1,783 (95 percent) exhib-ited a plain surface appearance. These countsare based on sherds that are generally largerthan 12 mm in diameter. Although there are asignificant number of smaller fragments (634),they are typically too small to provide reliablesurface treatment data.
VESSEL RIMS
Two hundred and one rim sherds wererecovered, forming portions of 171 vessels.Rims were considered analyzable when theyexhibited both exterior and interior rim sur-faces, the lip and a sufficient portion of theouter collar-neck surface to allow decorativestyles and attributes to be ascertained. Whilean additional 493 rim fragments were recov-ered, they are too incomplete to make anyreliable observations regarding descriptiveattributes.
Analytical data were entered directly into acomputer database file, which was used togenerate both an artifact catalogue and anattribute analysis. The database consists ofseven attribute fields, recording the nature ofthe specimen, portion, shape or form of theportions, metrics, vessel zone decoration(area, motifs and techniques), and castellationdecoration (area, motifs and techniques).
Design Motifs
The vessels in the assemblage exhibit fivemain design motifs from which all variationswere derived.
Simple. These are motifs that exhibit one ormore horizontal bands of oblique or verticallines executed in the same direction.
Opposed. These motifs consist of two ormore alternating bands of simples that repeat-edly change direction.
Horizontals. These are lines or motifs thatare predominantly characterized by horizontalelements.
Hatched. These motifs are simples that are
crossed by other simple elements in differentdirections.
Plain. These lack any decorative motifs.
Design Techniques
Concomitantly, the vessels were decoratedusing five main design techniques from whichall variations were derived.
Incising. This is a technique accomplishedby drawing a sharp pointed object across wetclay.
Linear Stamp. Various impressions arecreated by pressing a linear object into theclay, leaving a distinguishable impression.
Linear Punctate. Punctate impressions usinga linear object, often used for notching a collaredge.
Punctation. Punctation is accomplished bypunching the clay with various tools at differingangles. While annular punctates are likelyproduced by hollow bone tubes, linear punc-tates are produced by hollow linear objects.
Cord-wrapped stick. These are impressions
produced by the application of a cord-woundstick or twig. The stick was presumably wrap-ped with a fibre of some type and was im-pressed upon the wet clay.
WILLIAMSON AND POWIS ______ PARSONS SITE CERAMIC VESSELS ____________________________53
VESSEL RIM ANALYSIS
The ceramic vessels were analyzed using
attribute, linked-attribute (motif) and traditional
typological approaches in order to facilitate
future inter- and intra-site comparative studies.
Attribute Analysis
Table 28 provides the metric and non-metric
data for individual attributes. Differences in the
observation and description of certain attrib-
utes between this and previous analyses of
Parsons vessels are also noted and discussed
in detail below.While most of the rims in the assemblage
have well-defined collars (96.5 percent), a feware either collarless or are characterized byincipient collars. Three-quarters of the collaredvessels have angular bases, while the remain-der are rounded. Collar heights vary, although75 percent of the vessels have heights under 21mm (Figure 13). The mean is 17.25 mm, with astandard deviation of 9.6 and a coefficient ofvariation of 92.2. Basal collar widths wererelatively thin, with a mean of 9.9 mm, a stan
profiles are concave, while the remainder arestraight, convex, concave/convex, and con-vex/concave. The exterior profiles are predomi-nantly straight (40.4 percent), while theremain-der are concave or convex.
Tables 29 and 30 provide comparable inte-rior and exterior profile data drawn fromRamsden's (1977) attribute study of Huronceramics. Ramsden's analysis was of theUniversity of Toronto's assemblage from thesite and while one would not necessarily ex-pect the two analyses to produce identicalresults, they are so different as to suggestconsiderable inter-observer inconsistency inthe identification of this attribute. On the otherhand, our observations concerning this attrib-ute are consistent with those of Emerson, asreflected in his relative frequencies for Lawsonand Huron Incised types, the former beingdifferentiated from the latter on the basis oftheir characteristic concave interior profiles(Emerson 1968:37). It should be noted that it isconceivable that potters intentionally formedthis area of the pot, as is reflected on thevessel illustrated in Figure 14, where anappliqué had been added to transform the
Figure 13. Parsons Site Ceramic Vessel Collars: Distribution by Height Increments.
dard deviation of 2.57 and a coefficient of
variation of 6.6.All lip forms, with the exception of one
rounded specimen, are flat and they occurmostly at a right angle (83.3 percent) to thevessel interior. No other lip forms (concave,pointed) were discernible. Lip widths have amean of 5.61 mm, a standard deviation of 1.87and a coefficient of variation of 3.51.
Rim orientation is primarily outflaring, while
vertical forms comprise the remainder of the
sample. Approximately half of the interior rim
interior profile of a Black Necked sherd fromconcave to convex. The appliqué is not consis-tent with the rather more common practice ofclay having been rolled over the interior sur-face of the vessel from the lip.
The primary techniques used to decorate thecollars were incising and linear stamping.Combinations of both, some of which includepunctates, comprise the remaining vessels.While necks were most often undecorated (69percent), other motifs included incised horizon-tal elements and incised obliques. Interior rimdecoration is almost absent in the assemblage,although a few vessels have linear punctates,punctates, linear stamped obliques and incisedhorizontal elements.
Up decoration was primarily absent (80.7percent) although linear stamped and incisedobliques constituted a significant portion of theassemblage. There were also two examples ofcord-wrapped stick impressions.
' Table 31 provides comparative frequencydata for the attributes employed by Ramsdenin his 1977 ceramic analysis of a number ofsouthcentral Ontario Iroquoian sites.Consider-able differences ( > 10 percent) werenoted between his Parsons assemblage andthat recovered during the 1989-1990 ASIexcavations with respect to the frequencies forsimple and opposed collar motifs, as well asfor interior and exterior collar profiles. Therewas also a greater than eight percentdifference in the presence of lip decoration. Asit was not possible to combine Ramsden's rimobservations with the ASI vessel analysis, noeffort was made to incorporate the ASI resultswith those of Ramsden (1977). Acknowledgingthe obvious observer inconsistency in thecollar profiles, the other attribute datanevertheless suggest that caution must beexercised in seriation exercises based onincomplete site assemblages and may haveimplications for Rams-den's placement ofParsons in his site clusters.
Linked Attribute Analysis
General design attribute data were alsocross-tabulated with predominant exteriorcollar motifs. A number of variants, defined onthe basis of neck, lip and interior decoration,were noted for each major exterior collardesign sequence. Those vessels with onlypartial necks were not included in the calcula-tion of frequencies.
These data were collected as a preliminarystep towards an understanding of the correla-tion of various collar motif and other areadesign attributes in this particular sample, andhas no particular bearing on the definition of
Table 31. Comparison of Ramsden and ASIAssemblages with Respect to Single AttributeFrequencies.
preconceived types.The most common (69 percent) exterior
collar motif in the assemblage includes thosevessels with obliques on the collar to the exclu-sion of any other primary motif in the collararea. Variations include those with basal collarnotching or punctates as well as various lip,interior and neck designs. Sixty-seven vessels(39.2 percent) have obliques on the collar andno other secondary collar, interior, lip or neckdecoration. Table 32 lists the varieties andfrequencies for various combinations of neck,interior and lip designs for the remainingvessels with collar obliques.
The second most common collar motif in-cludes all the vessels with an opposed motif onthe collar. Some examples include basal collarnotching or punctates as secondary decora-tion, with or without lip, interior and neckdecoration. Three vessels have some form ofsecondary collar decoration. Table 33 lists thevarieties and frequencies for various combina-tions of neck, interior and lip designs for thosevessels with opposed designs on the collars.
The third most common collar motif has themost complex decorative motif in the assem-
WILLIAMSON AND POWIS PARSONS SITE CERAMIC VESSELS 57
blage. The basic design sequence consists ofhorizontal over opposed over horizontal witheither basal collar notching or punctates assecondary decoration. One vessel has a bor-dered oblique design, rather than the typicalbordered opposed motif with lip decoration, al-though this piece may be a castellation frag-ment. One other vessel lacks the basal notch-ing. While nine of the vessels have the basiccollar design in addition to lip decoration,including one with shoulder and body dec-oration, two vessels lack any other decoration.
Eleven vessels have obliques crossed byhorizontal line(s) on the collar. Four of thesehave no other form of decoration, while the re-mainder have various combinations of lip,interior and neck decoration (Table 35).
Seven vessels have a predominantly hori-zontal motif, consisting of either continuous orbroken incised line(s) on the collar. While all ofthese vessels lack lip decoration, one has onlylinear punctates on the interior and one has noother decoration whatsoever. The remainderall have neck decoration only consisting of twohorizontals, two obliques and one opposed.
The other collar design sequences includedtwo vessels with plain collars and no otherdecoration, one vessel with a hatched motif onthe collar and no other decoration, and threeother miscellaneous vessels with combinationsof horizontal, opposed and punctate elements.
The various collar designs were also cross-tabulated with collar height categories (Table35). While most of the designs were applied tovessels with collar heights under 21 mm, op-posed designs and horizontal over opposedover horizontal sequences increase in fre-quency with collar height. This is likely a reflec-tion of the presence of St. Lawrence Iroquoianvessel types in the assemblage.
Table 32. Obliques on Collar by Neck by Lip by InteriorDecoration
The typological approachused in this analysis wasbased on MacNeish (1952),Wright (1966) and Emerson(1968). While Figures 15-20illustrate a representativeselection of the vessels re-covered, Table 36 provides alisting by feature number oftraditional ceramic typeswithin the assemblage, Ta-
bles 37 and 38 summarize the frequencies byhouse/midden and exterior area, respectively.It is immediately evident from these tables that47% of the vessels from the site are derivedfrom midden contexts, including those featuresunderlying and adjacent to Midden 4 (Features240 and 245). Unfortunately, there are too fewvessels from the houses to draw any meaning-ful distinctions among them or between mosthouses and the middens.
There were, however, enough vessels fromHouse 8 and Midden 4 (including Feature 245)to allow for a more detailed analysis of thesimilarity between these two proveniencecontexts. André Bekerman (1994) concluded,on the basis of a comparison of traditionaltypes from the two contexts, that there musthave been other houses in addition to House 8contributing to Midden 4, given a broaderrange of vessel types in Midden 4 than House8. This is consistent with differences in deposi-tion between individual houses and middens.He also noted, however, that there were morevessels with neck decoration in Feature 245than in either House 8 or the rest of Midden 4.He attributed this to the fact that Feature 245underlay and therefore predated the majorityof the Midden 4 deposit.
The 1989-1990 excavation produced asignificant number of St. Lawrence Iroquoianceramic vessels. Compared to other roughlycontemporaneous sites in the Toronto region(Draper- 5.0 percent [Pearce 1978]; Keffer- 2.0percent [Smith 1991]), the frequency of St.Lawrence vessels (9.9 percent) at Parsons isconsiderably higher. The Parsons assemblagealso contains more St. Lawrence Iroquoiantypes than fifteenth and sixteenth century sitesin the Balsam Lake region such as Jamieson(3.4 percent), Hardrock (1.2 percent), Kirche(1.5 percent), and Coulter (2.8) percent and isactually more comparable to later sixtenthcentury sites such as Benson (9.8 percent),Dawson (15.6 percent and Trent (13.4 percent),when interaction with St. Lawrence popula-tions is thought to have been more extensive(Djamkar 1990:Table 7; Nasmith Ramsden1989:64; Ramsden 1990c:Table 1).
According to James Pendergast, who hasexamined the Parsons assemblage, the St.Lawrence vessels are poorly executed in bothmotif and technique. Pendergast attributesthese qualitative differences to the possibilitythat these vessels were made by either local
potters or descendants of St. LawrenceIroquoian peoples attempting to reproduceceramics from their ancestral territory (Pender-gast, personal communication 1992).
Several distinctive characteristics define St.Lawrence ceramics. These attributes includebasal collar notching, annular punctate faces,corn-ear, ladder platts and undulating castel-lations. Some of these attributes, such aspunctate faces, corn-ear and ladder platts, arenot represented in the ASI assemblage but arefound in the University of Toronto and JohnMorrison collections (Figure 21). In the ASIassemblage, basal collar notches are mostprevalent. In terms of traditional types, half ofthe vessels are Durfee Underlined with theremainder being either Roebuck Low Collar,Salem Horizontal or vessels that cannot beidentified to traditional types.
Of those St. Lawrence type vessels fromknown specific provenience contexts, whichincludes all but three vessels, 77 percent arefrom the extreme east side of the site (Houses8 and 9; Midden 4, Features 240, 245), withmost from Midden 4 and Feature 240. TheDutch Hollow Notched and Lalonde HighCollar vessels were also recovered from House8. These data suggest that the individualsmanufacturing these vessels were resident inHouses 8 and 9 and were contributors to Mid-den 4 and the associated refuse features.Finally, the vessel found in close associationwith the crania in Feature 245 was the DurfeeUnderlined pot illustrated in Figure 20.
Table 39 provides a comparison of thefrequencies of traditional types between Emer-son's 1968 analysis and the 1990 assemblage.If one assumes that the description and identi-fications of vessels from the two assemblageswere made in a similar fashion, there areseveral notable differences between the twosub-assemblages. The most obvious differenceis the wide range of vessel types present in the1990 assemblage, compared to the 1968 analy-sis, probably due to the recognition of St.Lawrence Iroquoian types as different fromOnondaga Triangular and the presence ofSidey Notched and Sidey Crossed varieties.The remaining type frequencies, however, areremarkably similar .
Using the 1989-1990 data, Black Necked,Huron Incised, Lawson Incised and LawsonOpposed types account for approximately 62percent of the vessels from the site. Using
WILLIAMSON AND POWIS _______ PARSONS SITE CERAMIC VESSELS _____________________________ 59
60 ONTARIO ARCHAEOLOGY No. 65/66, 1998
Figure 15. A Selection of Lawson Incised Type Vessels.
WILLIAMSON AND POWIS PARSONS SITE CERAMIC VESSELS 61
Figure 16. A Selection of Huron Incised Type Vessels.
Figure 17. A Variety of Vessel Types: Black Necked (a-f), Pound Necked (g-i), Sidey Notched (j, k) and SideyCrossed (I, m).
62 ONTARIO ARCHAEOLOGY No. 65/66, 1998
Figure 18. A Variety of "Exotic" Vessel Types: Salem Horizontal (a), Roebuck Low Collar (b), St. Lawrence or"Eastern" (c, d), and variants of Durfee Underlined (e-j).
WILLIAMSON AND POWIS PARSONS SITE CERAMIC VESSELS 63
Emerson's data, they account for 70 percent ofthe sample, as they also do for the Keffer site(Smith 1991:18). Since the Black Creek,Wood-bridge and Seed sites have 67 percent,50 percent, and 43 percent cumulativefrequencies, respectively, all of these sitescan be placed comfortably in the late pre-contact Iroquoian period in the north LakeOntario shore area (Wright 1966:70-71). If oneaccepts that the decreasing presence of neckdecoration is one of the more reliablechronological indices (Wright 1966:71;Ramsden 1977:184), these sites shouldseriate with Black Creek early in the sequencefollowed by Parsons and then McKenzie-Woodbridge and Seed-Barker. Keffer would fitbetween Parsons and the later sites. As notedpreviously, however, seriation exercisesshould not be based on incompleteassemblages. Any final resolution of the tem-poral placement of this site, or any other in theregion, must await either supportive absolutedates or further detailed analyses of additionalassemblages that are as complete as possible.
Castellations
Castellations from 56 vessels were analyzed,of which 32 retained enough of the vessel rimto be included in the rim sherd analysis. Theremaining 24 castellations were not usedbecause of a lack of analyzable rim attributes.
Twenty-three castellation forms have adeveloped expression, while the remainderwere incomplete and their form could not beproperly defined. Castellation shapes varied
as follows: pointed (n=15), rounded (n=3),notched (n=3), angular (n= 1), crenellated(n=1) and indeterminate (n=33). There are twocastellations with interior punctate designsand six with lip decoration. The lip decorationconsists of three with a simple oblique motif,two with linear punctates and one with a faintchecked stamp treatment.
Fourteen castellations, including threenotched examples, have considerable lateralprotrusion, which involves a thickening of thecollar. Two castellations show slight thicken-ing, while seven, including one crenellatedexample, exhibit none at all. There are 14castellations that have an inverted chevronmotif, of which ten manifest lateral protrusionand two have notching at the base of thecastellation.
Four unusual castellation features werenoted: one castellation base with a roughlyformed annular punctate measuring 15 mm indiameter; three examples of basal collarnotching; one example of linear punctation atthe collar-lip interface; and one of collar-lippunctation plus a strap handle.
JUVENILE CERAMICS
The juvenile assemblage includes twenty rimsherds representing 19 vessels. These vesselsare characterized by their crudely formedinterior and exterior profiles, lip edges, as wellas distinctive size and decorative motifs (Fig-ure 22:a-h). Juvenile ceramics are admittedlyinappropriate for seriational purposes, butthey do have the potential to reveal importantinformation about the learning process amongyoung potters. A total of six juvenile bodysherds (including two bases) and sixneck/shoulder sherds were also identified.There are six juvenile collared vessels, ofwhich two are outflaring. One of these exhibitsa complete castellation. The remaining 13vessels are collarless to incipiently collared,being either vertical or slightly insloping inform. The collar heights range from 3 to 20mm, while vessel lips range from 3 to 9 mm inwidth.
Eight vessels have plain collars, while 11exhibit motifs consisting of either obliques,opposed obliques, horizontals(s) or obliquescrossed by horizontal(s) made by incising(n=7), fingernail impressions (n=3) or linearstamp (n=1). One of the vessels (Figure 22:h)
68 ONTARIO ARCHAEOLOGY No. 65/66, 1998
WILLIAMSON AND POWIS PARSONS SITE CERAMIC VESSELS 69
Figure 21. Selected St. Lawrence Iroquoian Type Castellations in the Morrison Collection.
70 ONTARIO ARCHAEOLOGY No. 65/66, 1998
Figure 22. Selected Juvenile Vessel Rims.
with an opposed collar motif has basal collarnotching, which might be interpreted as anattempt by a more experienced individual tocreate a decorative motif similar to the easternvariant (St. Lawrence) types found at the site.Similarly, one of the plain collared vessels hasa combination of attributes suggesting that thepotter was a relatively experienced artisan.Despite unskilled execution of the collar, collarbase and interior shoulder area, a certaindegree of experience in the ceramic art isdemonstrated in the interior profile, lip edge,neck/ shoulder area and castellation.
Five vessels have both collar and neckdecoration, while one has a plain collar withneck decoration. The neck decoration on thesevessels consists of either obliques, horizon-tal(s) or obliques crossed by a horizontal linemade by incising (n=4) or fingernail impres
sions (n=2). There is also one heavily deco-rated neck/shoulder/body sherd (with base),which exhibits fingernail incised horizontalscrossed by punctates on the neck, and twohorizontal rows of circular punctates over asingle row of linear punctates on the shoulderand body. This vessel is discussed in the con-sideration of ""exotica' recovered from the siteduring the course of previous excavations(Robertson, Monckton and Williamson, thisvolume).
There are two castellation appliqués repre-sented in this assemblage. Both exhibitobliques on each of their lateral edges, whichform an opposed motif. One has a deep verticalincised oblique in the centre, while the otherhas three deep horizontal notches in the centre.Their lengths and widths are 19 and 22 mmand 10 and 13 mm respectively.