Gordion Excavations Project: Early Hellenistic Pithos

Pottery Processing at Gordion 1988-1997: A Brief Overview


Sherdyard with sherd boy	Sherdyard

INTRODUCTION

The recording system used for processing pottery excavated at Gordion since 1988 has been designed to emphasize speed, flexibility, and consistency. This system provides basic information on vessels/rims (types, shapes and sizes), relative proportions of local wares, minimum abundances of imports, and, indirectly, basic information on the potter's craft.

Based on an approach originally developed for use on the Royal Ontario Museum (Toronto) Mahidasht Project in Iran (1975, 1978), I adapted it in collaboration with my assistants in 1990 for the recording of the material from the 1988-1989 excavations. This data underpinned the basic ceramic chronology for the Yass höyük Stratigraphic Sequence (YHSS) (see Chronology). Most of the basic recording for the basic YHSS was done in 1990, with a team of five assistants (four pottery assistants, one draftsman). Pottery Staff 1988-99 This field experience, working up the data recorded in 1990, and my further work during the 1992 season and later has resulted in ongoing modifications which yielded the current system.

In order to maximize consistency, from 1992 on I have done all basic recording, using a simple ware and shape recording system in the sherdyard to process material as it was excavated, in so far as was possible. In the 1995-1997 seasons, experienced pottery assistants were given specific bodies of material to sort (Pottery Staff 1988-99), although I did the final recording.

RECORDING

Since pottery cannot be exported, except for limited samples for analyses such as neutron activation, field time must concentrate on the sherds and vessels themselves. As far as was possible, pottery was processed day by day:

DAY	ACTIVITIES
1	Material excavated, bagged and tagged by lot, sent to house, and checked in
2	Material washed; site supervisors do basic ware sorts and separate diagnostics for each lot in individual sherdyard bins; preliminary dates established
3	Ware counts done, body sherds (except joins) discarded, diagnostics recorded, material chosen for drawing or photography, special ceramic items (e.g., Greek and other imports, amphoras, and tile fragments) separated; remaining diagnostics bagged for storage
4+	Material drawn, photographed, or conserved

This schedule provided the quickest feedback to excavators for dates of contexts, possible intrusive material, and interconnections.

Ideally material would be sorted and recorded by provenience units, varied constraints, particularly of time and staff, did not allow this as a regular procedure. Excavation can stretch over days, weeks, or even seasons. Several specific deposits, such as a rich and closely dated trash dump in Operation 17 from the end of the Middle Phrygian period (ca 550 BC), were put aside and studied as a single corpus.

WARES AND COUNTS: Wares are the basic unit of recording, and incorporate technological as well as stylistic and descriptive attributes. A ware is defined as a recurring combination of distinctive attributes including color, temper, forming and finishing methods, characteristic vessel forms, and type of decoration. The hierarchy of sorting is:

Basic Steps in Sorting Sherds for Wares
First	Second	Third	Fourth
Color-->
	Temper (Size and Nature)-->
		Surface Finish (Basic)-->
			Decoration

Phrygian-era (YHSS 6-3) Ware Descriptions

Basic Bronze and Early Iron (YHSS 10-7) Ware Descriptions

Counts were recorded using strings of short codes. Typical entries would translate to "Grey coarse-grit with smoothed surfaces", "Black Fine Polished, with Incised Decoration", or "Buff Medium-fine grit with Self-slipped surfaces and Red-Painted decoration". Because wares are recorded with segmented definitions and independent codes, the data can regrouped according to different hierarchies of attributes (e.g., beginning with type of temper or decoration rather than color). Combined with vessel shape and size data, the basic technological data incorporated in ware definitions facilitates reconstruction of the potter's craft and the organization of production.

At Gordion, many imports are identifiable, some more readily than others. Whenever sherds are identifiable as some specific type, such as Greek Black Glazed or Lydian Streaky (or Marbled etc.), this is noted as part of the ware count recording. Since some are more identifiable than others -- Black Glazed is highly distinctive while much Lydianizing material is less so -- counts provide a basic measure of minimum abundances. Greek and East Greek material is passed on to K. DeVries for detailed study, and transport amphoras to M. Lawall.

Although further characterization of the sherds beyond counts (such as weight as an indicator of relative fragmentation) could be useful, limitations of time and staff have foreclosed such recording. Distinctive overall characteristics of material from individual lots were noted (e.g., "small and worn" suggesting decayed mudbrick or surfaces, or "green" staining for 'latrine' deposits).

DIAGNOSTICS: Diagnostics include all sherds which preserve features (e.g., rim, handle, base), decoration (e.g., painted, applique, or burnished), and technological information (e.g., forming joins). In addition to diagnostic attribute(s) of each sherd, recording included ware, vessel size (rim or base diameter where feasible), and extent of profile (e.g., "rim to shoulder" or "neck to base"). For rims, a gross "percent of circumference present" is recorded, as an indicator of fragmentation or for calculating minimum number of vessels. Since the emphasis was on speed and flexibility, shape information was usually recorded in a profile sketch done in 2-5 seconds. Personal experience has shown that this was considerably faster than using a coding system, which would have been difficult to use in the usual breeze/wind and dust of the sherdyard at any time, and preserved more detailed information for possible use in later analyses. Since the importance of many contexts only emerges during the later stratigraphic analyses, retaining as much information as possible is desirable. Sherds were chosen for drawing and passed to a draftsman.

DATA ENTRY: Data entry is done in the U.S., maximizing field time spent working with sherds and recording data.

CONSTRAINTS AND PROBLEMS

Several basic constraints have influenced the use and refinement of this pottery recording system since 1993.

Season Sherds Recorded

1988-89 [ ]

1993 28,000

1994 60,000

1995 80,000

1996 48,000

1997 28,000

Total 244,000
VOLUME: Large --very large -- numbers of sherds have been recovered. Backlogs from previous seasons were processed as time allowed. For each season, the cumulative numbers recorded are:

ONGOING EXCAVATION: Excavation was continuous from 1993-1997; there were no study seasons. Any pottery not processed immediately had to be stored for later analysis when time allowed.


More of the sherdyard work area	Bagged sherds in mudplaster bins

WEATHER: Since all sherd processing was done in the relative open, windstorms, rain, and hail all could and did slow or halt analysis, sometimes for several days at a time. A tarp shaded about half of the sherdyard at any given point in the day. Even on the best of days a moderate breeze blew through the sherdyard; at least a couple days a week, conditions (high wind and the resulting dust) would halt work. Dust, sand, and chaff from the mudplaster bins continues to sift out of notebooks. Some field seasons were plagued with repeated downpours (notably 1988 and 1997), while others were arid (1989). Rainstorms could, and did, leave the sherd bins filled with water. The tags survived, but sherds and the bins themselves would take several days to dry out. Rain, and wind, would dismantle the canopy over the sherdyard at least once a season (see Sherdyard after a storm).

STRATIGRAPHIC INFORMATION: As much as possible of the pottery recovered was recorded as it flooded through the sherdyard. Lots and/or loci identified as crucial at the time of excavation were given special attention. Material from the backlog was recorded as time allowed. A relatively small backlog remained at the end of the 1997 field season.

STAFFING: Each season was a battle to cope with the masses of sherds flooding through the sherdyard (see Pottery Staff 1988-99).

POTTERY WARES: The Middle-Late Bronze (YHSS 10-8) industries were relatively straightforward in terms of wares, and were stratigraphically sealed (see Basic Bronze and Early Iron (YHSS 10-7) Ware Descriptions). The Early Iron Age pottery (YHSS 7) presents some problems but were likewise sealed below Early Phrygian strata (YHSS 6) (see Basic Bronze and Early Iron (YHSS 10-7) Ware Descriptions). The later Phrygian/Hellenistic material (YHSS 5-3) have been the focus of the 1993-1997 excavations. These present some difficulties. Most of the local common wares, typically grit-tempered grey in the Phrygian era, were often not very distinctive in either fabric or shape (see Basic Phrygian-era (YHSS 6-3) Ware Descriptions). Grey wares and typical vessel types (e.g., ledge-rim pots and jars) had long currency and exhibit slow or infrequent episodic change.

Imported pottery presents varied problems, depending on the assemblage.

Pottery	PROBLEMS AND ACTIONS
Greek	Most Greek imports were readily recognized (Black Glazed, Red Figure, and Black Figure). Passed to K. DeVries for study.
East Greek	Some "East Greek" material is easily recognized, other less so. Some overlap with "Lydian/Lydianizing" material yields problems of identification. Relatively few comparanda are available. Passed to K. DeVries for study.
"LYDIAN" and "LYDIANIZING"	"Lydian / Lydianizing" imports, and probable local imitations and adaptations, are perhaps the largest body of imported material, but relatively few comparanda are available. Visitors from Sardis (especially E. McIntosh and A. Ramage) have helped greatly.
Transport Amphoras	Amphora diagnostics (toes, many rims, some handles) are usually readily recognized; increasing familiarity with the local fabrics has led to more extensive identification of potential body sherds. M. Lawall's work at Gordion beginning in 1996 has shed much new light on these problems.
Roman	Roman levels were not part of the original research design, but excavation of extensive deposits has led to the addition of Roman pottery to the analyses. Roman material is unrelated technologically or typopologically to that from previous periods. A. Goldman is working on the Roman period at Gordion for his dissertation.

OTHER COMPONENTS OF POTTERY ANALYSIS

The ongoing collaboration with M. J. Blackman (Conservation Analytical Laboratory, Smithsonian Institution) on neutron activation analysis of the medium grit-tempered wares has shed considerable light on pottery production at Gordion during YHSS 10-6 so far. Work on samples from YHSS 5-4 is in progress, while limited samples from YHSS 3-2 have been run (see Neutron Activation).

Selected Bibliography
Gordion Project Bibliography

DeVries, Keith
1996 "The Attic Pottery from Gordion." Pp. 437-445 in Athenian Potters and Painters, edited by O. Alexandri and J. Oakley. American School of Classical Studies [Athens].

Henrickson, Robert C.
1993 "Politics, Economics, and Ceramic Continuity at Gordion in the Late Second and First Millennia B.C." Social and Cultural Contexts of New Ceramic Technologies, edited by W. D. Kingery, pp. 89-176. Ceramics and Civilization VI. Westerville, OH: American Ceramic Society.

1994 "Continuity and Discontinuity in the Ceramic Tradition at Gordion during the Iron Age." Anatolian Iron Ages 3, Proceedings of the Third Anatolian Iron Ages Colloquium Held at Van, 6-12 August 1990, edited by D. French and A. Cilingiroglu, pp. 95-129. British Institute of Archaeology at Ankara, Monograph No. 16. Ankara.

1995 "A Comparison of Production of Large Storage Vessels in Two Ancient Ceramic Traditions." Materials Issues in Art and Archaeology IV, edited by P. B. Vandiver, J. Druzik, J. L. Galvan, I. A. Freestone, and G. S. Wheeler, pp. 553-572. Materials Research Society Symposium Proceedings, Volume 352. Pittsburgh: Materials Research Society.

In press "The Craft of the Early Phrygian Potter." Türk Arkeoloji Dergisi.

Henrickson, Robert C., and Blackman, M. James
1996 "Large-Scale Production of Pottery at Gordion: A Comparison of the Late Bronze and Early Phrygian Industries." Paléorient 22(1): 67-87.

1999 "Hellenistic Production of Terracotta Roof Tiles Among the Ceramic Industries at Gordion." Oxford Journal of Archaeology 18: 301-326.

Henrickson, Robert C., and Voigt, Mary M.
1998 "The Early Iron Age at Gordion: The Yassihöyük Stratigraphic Sequence." Pp. 79-107 in Thracians and Phrygians: Problems of Parallelism (Proceedings of an International Symposium on the Archaeology, History, and Ancient Languages of Thrace and Phrygia [Middle East Technical University, Ankara, Turkey, 2-6 June 1995]), edited by Numan Tuna, Zeynep Aktüre, and Maggie Lynch. Centre for the Research and Assessment of the Historic Environment. Ankara: Middle Eastern Technical University (METU), Faculty of Architecture Press.

Sams, G. Kenneth
1979a "Imports at Gordion." Expedition 21(4): 6-17.

1979b "Patterns of Trade in First Millennium B.C. Gordion." Archaeological News 8(2/3): 45-53.

1994 The Early Phrygian Pottery.The Gordion Excavations, 1950-1973: Final Reports IV. University Museum Monograph 79. Philadelphia.

Schaus, Gerald P.
1992 "Imported West Anatolian Pottery at Gordion." Anatolian Studies 42: 151-177.

Voigt, Mary M.
1994 "Excavations at Gordion 1988-89: The Yassihöyük Stratigraphic Sequence." Anatolian Iron Ages 3: Proceedings of the Third Anatolian Iron Ages Colloquium Held at Van, 6-12 August 1990, edited by D. French and A. Cilingiroglu, pp. 265-293. British Institute of Archaeology at Ankara, Monograph No. 16. Ankara.

Voigt, Mary M., and Henrickson, Robert C.
In press "The Early Iron Age at Gordion: The Evidence from the Yassihöyük Stratigraphic Sequence." To be published in The Sea Peoples, edited by E. Oren. Philadephia: University of Pennsylvania Museum of Archaeology and Anthropology.

Last Revised: 17 October 2000

Send mail to: R.C. Henrickson