Defining Susquehannock An Attribute Analysis of Susquehannock

Document Sample
Defining Susquehannock An Attribute Analysis of Susquehannock Powered By Docstoc
					Defining Susquehannock: An Attribute Analysis of Susquehannock Pottery, AD 1575-1680

          A Report on the Execution and Results of a PHMC Scholar in Residence Grant


                                            Lisa M. Lauria
                                     PhD Candidate in Anthropology
                                         University of Virginia

                                 Do not cite without permission of author.

         The project outlined below was created as an attempt to refine the existing
Susquehannock pottery typology by looking at ceramic stylistic attributes on both complete pots
and ceramic sherds. 1 Ceramic analyses serve as one of the primary tools through which
archaeologists reconstruct the past. By comparing design elements of ceramics within and
between archaeological sites, scholars can make informed decisions about the relationships
between people utilizing these sites. Current Susquehannock pottery types were created without
the use of statistical methods. Moreover, the existing typology can only be applied to whole
vessels or pottery sherds that come from the rim of the pot. In designing my project, I looked for
a method that would apply to all ceramic sherds. The method I chose was an attribute analysis.
Rather than look at the decorative motif on a whole vessel, attribute analyses break the motifs
down into their constituent parts. The benefits of this method are that it allows for analysis of
sherds too small to display a complete decorative motif, and it eliminates the bias associated with
pre-grouping sherds and pots into types before the analysis has taken place. Ceramic attribute
analyses are not absent from Eastern North American research, yet this approach has not been
used for Susquehannock pottery. 2
        The enormity of the Pennsylvania Historical and Museum Commission’s Susquehannock
pottery collection necessitated a sampling of the collection. I selected six Susquehannock
artifact collections to analyze: 36LA1, 36LA3, 36LA7, 36LA8, 36LA54, 36YO170. These sites
represent five different phases in Susquehannock cultural history between AD 1575 and 1680
and include samples representative of the four identified Susquehannock pottery types. 3 In this
manner they are representative of the changes in Susquehannock pottery design and construction
between AD 1575 and 1680, when communities were thriving in the Lower Susquehanna Valley.
        I recorded ceramic attributes in two databases—one for ceramic sherds and one for whole
or nearly whole pots. I have processed the data with the assistance of computerized statistical

  The most recent definition of the Susquehannock typology appears in Barry Kent, Susquehanna’s Indians
(Harrisburg: The Pennsylvania Historical and Museum Commission, 1984).
  Good examples can be found in Jay F. Custer, “Late Woodland Ceramics and Social Boundaries in Southeastern
Pennsylvania and Northern Delaware,” Archaeology of Eastern North America 15 (1987), 13-28, and Custer,
Prehistoric Cultures of Eastern Pennsylvania (Harrisburg: Pennsylvania Historical and Museum Commission,
1996); Michael J. Klein, An Absolute Seriation Approach to Ceramic Chronology in the Roanoke, Potomac and
James River Valleys, Virginia and Maryland, (Ph.D. Diss, Department of Anthropology, University of Virginia,
1994); also eight of the seventeen essays in Charles F. Hayes III, ed., Proceedings of the 1979 Iroquois Pottery
Conference, Research Records No. 13, (Rochester: Rochester Museum and Science Center, 1980), 113-131.
  Kent, Susquehanna’s Indians.

                                                                                                         Lauria 1
software in an attempt to measure variation and identify relationships between individual sherds
and pots as well as within the defined Susquehannock pottery types. In addition to creating a
fully portable database for comparison of Susquehannock and non-Susquehannock pottery, I also
hoped to uncover relationships between ceramics that have been previously overlooked and
could lead to an adjustment of the current Susquehannock pottery typology.

Collections Utilized

         While in residence in the Section of Archaeology of the State Museum of Pennsylvania, I
utilized portions of five different data collections— artifact collections, computerized databases,
the pre-electronic card catalog collection, manuscript files, and library journals. As mentioned
above, the ceramic artifacts that I examined for this study came from six Susquehannock sites—
36LA1 (Roberts), 36LA3 (Strickler), 36LA7 (Schultz), 36LA8 (Washington Boro Village),
36LA54 (Ibaugh), and 36YO170 (Byrd Liebhart). In total I examined 12,264 pottery sherds and
206 whole or nearly whole pottery vessels from these six collections (see Figure 1). In an effort
to get a larger statistical sampling of attributes found in the Susquehannock pottery typology, I
also looked at pottery vessels from the following eight collections while in residence: 36LA4,
36LA6, 36LA9, 36LA12, 36LA956, 36BR3, 36BR147, and 36BR148. These specimens boosted
my total pottery vessel count from 206 to 352.
         In order to locate ceramic vessels and sherds in the Section of Archaeology’s storage
area, one must consult the computerized databases that the section maintains. The computerized
artifact collection drawer database and the artifact glass cabinet database were essential for
locating portions of the Susquehannock artifact collection by site and artifact type. During my
residency, these databases also served as sorting tools to help in the organization and collection
of my data. I also consulted three other computerized databases—the NAGPRA Associated
inventory, the NAGPRA Unassociated inventory, and the computerized manuscript file database.
The NAGPRA inventories provided burial data relevant to some of the ceramics I examined. I
used the electronic manuscript database to search for manuscripts related to the Susquehannocks.
I looked at each of these articles to assess their usefulness for my research. In addition to these
secondary manuscripts, I also examined the county file records relevant to the excavations of the
six sites in my proposed study and the excavation records of 36LA4, 36LA6, 36LA9, and
36LA12. These excavation records provided valuable information regarding the provenience of
the artifacts in my study. They also aided me in my sampling decisions. During breaks from
pottery measuring, I consulted the Section of Archaeology’s library resources. In addition to
several books that Janet Johnson suggested I examine, I also browsed through the complete
holdings of three journals—Pennsylvania Archaeology, Ethnohistory, and Northeast
Anthropology—for articles relevant to my research.


         Before the start of my residency, I created two Microsoft Excel spreadsheets in which to
record data on ceramic attributes, provenience, and design motifs. One spreadsheet was devoted
to ceramic sherds and one to ceramic vessels. For the purpose of my study, a ceramic vessel was
classified as a vessel if all or part of the body of the vessel was present or at least 50% of the
vessel rim was present. All other pottery fragments were classified as ceramic sherds. I
recorded 34 ceramic attributes for vessels and 31 ceramic attributes for sherds. My list of

                                                                                           Lauria 2
attributes was derived from written descriptions, line drawings, and photographs of
Susquehannock pottery.
        I began my residency by recording attributes for ceramic vessels from the six selected
study sites. By beginning with vessels, I was able to judge whether or not my spreadsheets were
capturing the variation in ceramics from these sites. It was at this time that I realized that my
vessel sample collections for the 1600 to 1625 AD time period and the Blue Rock Valanced
pottery type were quite small. By looking at the pottery vessels from several additional
Susquehannock burial sites, I was able to get a larger and more representative sampling of
Susquehannock pottery over time and space.
        Ceramic sherd data was collected for the six original study sites. Of the 12,264 ceramic
sherds examined, approximated 2000 of those were decorated. This is important because many
of the ceramic attributes I recorded were only present on decorated ceramics. All ceramic sherds
from 36LA1, 36LA3, 36LA54, and 36YO170 were examined. The collections from 36LA8 and
36LA7 were so large that I chose to sample these collections. To sample these collections, I ran
a simple random sample on an alphanumeric ascending list of drawer and box number as listed in
the computerized collection catalog. While I would have preferred to base my random sample
on provenience information, the storage processes for the pottery sherds from these sites did not
permit access to all provenience information for all sherds.
        Since completing my residency I have been slowly importing my ceramic data from
Excel spreadsheets into SPSS for computer aided analysis. Of the attributes I recorded, 27 of the
vessel attributes and 25 of the sherd attributes produced nominal data—a data type that has
limited statistical inference.

Significant Findings/Tentative Conclusions

        The primary goal of this project was to statistically define Susquehannock pottery
attributes for all Susquehannock pottery and for each Susquehannock pottery type. The details
of these definitions will be published in my dissertation which I will place on file with the
Pennsylvania and Historical Museum Commission when it is complete. By defining
Susquehannock ceramic attributes, I can begin to asses and refine the existing Susquehannock
pottery typology. The findings presented here are based only on the pottery vessel data that I
collected while in residence. I have yet to integrate the vessel and sherd databases into a single
working database. Statistical analyses conducted to this point suggest that the existing
Susquehannock pottery typology does capture the most significant relationships between cases as
well as the majority of variation between them. At this point in my analysis I do not suggest that
a refinement of the existing Susquehannock pottery typology is necessary.
        Analyses of ceramic attributes from the Susquehannock pottery vessel collections do
show clear shifts in attribute use over time. While some of these shifts are suggested by the
relative shift in Susquehannock pottery type prevalence over time 4 , these changes are more
noticeable when measured in terms of attributes. Figures 2 through 4 illustrate the shifts in
frequencies of four ceramic attributes—incised line decorations, parallel line decorations, cord-
impressed decorations, and punctate decorations. Written descriptions of the defined
Susquehannock types do indicate extensive use of incised lines, parallel lines, and punctation in

 For a diagram of this shift see W. Fred Kinsey, III, “Historic Susquehannock Pottery,” Susquehannock Miscellany,
John Witthoft and W. Fred Kinsey, eds., (Harrisburg: The Pennsylvania Historical and Museum Commision, 1959),

                                                                                                        Lauria 3
Schultz Incised and no not note these attributes for Strickler Cord-Marked pottery so one might
predict an outcome as seen in figures 2 and 3. The frequency pattern for cord-impressed
decoration (figure 4) is less likely to be noted on the basis of the written typology descriptions.
In these descriptions, cord-impressed decoration tends to get lumped with incised line
decoration. The decrease in frequency of these four attributes over time are clear, but because
my sample size for the period 1625 to 1645 AD is so small (n=6 vessels), I must be tentative
when drawing statistical conclusions. Excluding this time period from my analysis, hypothesis
testing using the chi-square statistic indicates that time period and cord-impressed decorations
are not independent variables. 5
         Another interesting attribute frequency pattern emerges when examining different
archaeological site collections from contemporary time periods. These patterns can be examined
for the two earliest Susquehannock time periods in my study. Vessels from two archaeological
sites dating between 1575 and 1600 AD are included in my study—36LA7 and 36LA9. When
examining attribute frequencies from these two sites, I identified four points of possible variation
in vessel decoration: treatment of vessel lip, presence of molded decorations, presence of cord-
impressed decoration, and presence of punctate decoration. Pottery vessel lips at these two sites
were most commonly either rounded or flattened in profile. At 36LA7, 76.9% of vessels lips
were flattened and 15.4% were rounded (n=13). At 36LA9, 49.4% were flattened and 43.2%
were rounded (n=81). Only one of the thirteen pottery vessels at 36LA7 had a molded
decoration (7.7%), while 21% of the 81 vessels at 36LA9 displayed molded decorations. The
difference in molded decoration is not surprising since molded decorations are a common
attribute of Blue Rock Valanced pottery type vessels and 22 vessels or 16% of the 36LA9 sample
is of this particular Susquehannock type. Cord-impressed decorations were more frequent on the
pottery from 36LA9 where 60.5% of the sample displayed this attribute. Only 38.5% of the
36LA7 sample displayed cord-impressed decorations. Punctate decoration was more frequent at
36LA7 (69.2%) than at 36LA9 (49.4%).
         Since cord-impressed and punctate decorations are not exclusive to any one
Susquehannock pottery type but are both more common in the earlier time periods when the
Schultz Incised pottery type was most common, I decided to look for patterns within the Schultz
Incised pottery type. The Schultz Incised sample for 36LA7 was very small (n=5) so I looked
for patterns of variation between 36LA9 and 36LA4. The more striking of the two attribute
frequency patterns was in the presence of cord-impressed decorations. While 43.5% of the
Schultz Incised pottery at 36LA4 had cord-impressed decoration, 70.8% of the same pottery type
at 36LA9 had cord-impressed decoration. I calculated the chi-square statistic to test whether or
not there is a relationship between cord-impressed decoration and site location and found that
there is a weak relationship between these two variables. 6 These findings suggest that pottery
type designations may mask some of the variation in Susquehannock pottery decoration.
         Vessels from five archaeological sites dated between 1600 and 1625 AD were included
in my study. Three of these sites had very small vessels counts (36LA12, 36LA8, and 36LA6)

 For all hypothesis testing I selected a significance level of α=.05. In this case χ²=96.22 with n=328 and df=3. The
Phi measurement of association calculates a value of .542 for this relationship—a moderate strength of relation.
  I selected α=.05 as my significance level. With n=71 (# of cases) and df=1 (degrees of freedom), χ²=4.937. The
significance of this test is .026 so I reject the null hypothesis and conclude that cord-impressed decoration on
Schultz Incised pottery and site location are not independent. The Phi measurement of association calculates a value
of .264 for this relationship—not a very strong relationship.

                                                                                                           Lauria 4
and so were not used when looking for general patterns. Two interesting differences emerged in
attribute patterns at 36LA4 and 36LA54. At 36LA4, 51.6% (n=91) of vessel lips were rounded
and 46.2% were flattened while only 13.3% (n=15) of vessel lips were rounded and 66.7% were
flattened at 36LA54. The treatment of the vessel lip may seem like a minor difference, but this
pattern does suggest that the potter populations at these two sites were doing things differently.
The second interesting difference is the types of effigies found on vessels. The presence of an
effigy on a Susquehannock pot is one of the attributes that is unique to the Washington Boro
Incised pottery type. At 36LA4, 20 of 91 pots (22%) had effigies present and all of these effigies
were anthropomorphic faces. At 36LA54, 4 of 13 (30.8%) pots had effigies present. Two of
these effigies were anthropomorphic faces, one was an anthropomorphic body, and one was an
unidentifiable figure. While the sample size for 36LA54 is small, there is a great deal of
variation within this small sample. Looking at a third site from this time period, 36LA6, I again
see greater variation than at 36LA4. Here 4 of 7 pots had effigies present. Three of the four
effigies were anthropomorphic faces but the fourth was an anthropomorphic body. Again this
pattern suggests that at least some of the potters at 36LA54 and 36LA6 were doing things a bit
differently from their 36LA4 neighbors. It is certainly possible that the variation is the result of a
single potter since the sample is so small. Further comparison of these vessels would be
necessary to draw such a conclusion. The uniformity of the anthropomorphic face effigy in the
much larger sample at 36LA4 becomes more striking in comparison.
         I do believe that it is significant that certain attributes do appear to be more common or
even distinctive to particular Susquehannock pottery types. These findings suggest that the
existing Susquehannock pottery typology is a useful way to organize pottery and look for
patterns. I would conclude that the existing typology does capture the general patterns of
variation in Susquehannock pottery. More interestingly, my findings suggest that an attribute
analysis of Susquehannock pottery does suggest greater diversity across space and time than the
typology readily suggests. Differences in attribute frequencies within Susquehannock pottery
types are evident at different archaeological sites. This suggests that the populations creating
these pots may have been drawing on different sets of artistic elements and that some of these
elements may be site specific. These findings suggest that different populations of potters were
operating at different contemporary sites. While they were certainly communicating and sharing
many elements of pot design, individuals were creating these pots and were drawing on different
sets of common elements.

Significance to Larger Research Project

         This PHMC scholar in residence project was to serve as a spring-board for my
dissertation research. It has served that purpose although not in the originally intended manner.
When proposing this project, I hoped to come away with a simplified ceramic attribute database
that I could easily transport to other Susquehannock and non-Susquehannock pottery collections.
While I have created a readily portable database, it has not been simplified in the ways I had
hoped it would. In the twelve weeks that I spent in residence, I collected far more data than I
believed was possible. In consultation with my dissertation advisor, I will now be using this data
as the bulk of my dissertation research.
         The original scope of my larger research project was to compare Susquehannock ceramic
attributes found at sites in the Lower Susquehanna River Valley with ceramic attributes found in
the Upper Potomac River Valley at sites now labeled Susquehannock. I still intend to undertake

                                                                                             Lauria 5
that project, but that project is now the focus of my archaeological research post-dissertation. I
hope to someday use my attribute database to look at Iroquoian collections from New York. The
goal of my larger project is to continue to explore connections between native peoples in the
Middle-Atlantic region of AD 1500 through 1800.


        While my conclusions at this point in time are tentative, I hope that the results of my
dissertation research might some day enhance the interpretive programs of the PHMC. The
museum exhibits showcasing the Susquehannocks could perhaps highlight Susquehannock
interactions with neighboring native peoples in addition to Europeans. An educational exhibit
could even be designed around the interpretive uses of ceramics in archaeology. Children would
probably enjoy some hands-on replicas of sherds or vessels. These replicas could help to
demonstrate some of the attributes that archaeologists look for when studying pottery.
        I do believe that I have made some valuable contributions to the Section of Archaeology
as a direct result of my residency. Very little of the Susquehannock artifact collection has been
systematically examined since the move from the museum building to the Keystone building.
Since I was working closely between the collections databases and the artifacts, I was able to
identify errors in the database and maintained lists of errors for future correction. I also re-
housed some artifacts in archival storage materials as I came across old storage materials. One
of the challenges I faced in analyzing the PHMC ceramic artifact collections was the lack of any
definitive count of individual ceramic sherds. While my research only sampled the collections, I
was able to provide ceramic artifact counts for some of the collection. All of these contributions
should aid future research scholars.

                                                                                          Lauria 6
Figure 1: Pottery Analyzed
TIME PERIOD        SITES STUDIED                                                          VESSELS   SHERDS
1575 to 1600 AD 36LA7 Schultz                                                             35        7586
                   36LA9 Funk                                                             81
1600 to 1625 AD 36LA8 Washington Boro Village                                             4         916
                   36LA54 Ibaugh                                                          15        588
                   36LA4 Keller                                                           91
                   36LA12 Eschelman                                                       2
                   36LA6 Frey-Haverstick                                                  7
1625 to 1645 AD 36LA1 Roberts                                                             6         199
1645 to 1665 AD 36LA3 Strickler                                                           116       2239
1665 to 1680 AD 36YO170 Byrd Liebhart                                                     30        736
                   Other                                                                  5
                   Total                                                                  352       12264

Figure 2:

                                  Frequencies of Parallel Line and Incised Line

   % of Vessels with Attribute

                                        1575 to 1600 to 1625 to 1645 to 1665 to
                                       1600 AD 1625 AD 1645 AD 1665 AD 1680 AD
                                                     Time Period

                                                                                                            Lauria 7
Figure 3: Frequencies of Punctate Decoration through Time

                                                                                     52.1        51.3
   % of Vessels with Punctate Decoration




                                           20                                                                                           17.2


                                                                                  1575 to       1600 to     1625 to      1645 to        1665 to
                                                                                 1600 AD       1625 AD     1645 AD      1665 AD        1680 AD
                                                                                                          Time Period

Figure 4: Frequencies of Cord-Impressed Decoration through Time


                                           % of Vessels with Cord Impressions





                                                                                20                  17.6


                                                                                                                    0            1.2           0
                                                                                       1575 to     1600 to      1625 to     1645 to        1665 to
                                                                                      1600 AD     1625 AD      1645 AD     1665 AD        1680 AD
                                                                                                             Time Period

                                                                                                                                                     Lauria 8

Shared By: