Introduction to the Early Iron Age Tel Kinrot Typology

Chapter 5 Presenting the Pottery – Description and Analysis

5.1 Introduction to the Early Iron Age Tel Kinrot Typology

Typology as a method of classifying artifacts in archaeology was first applied on a large scale in Scandinavia; it was used for primarily chronological purposes (Montelius 1884; 1899; Müller 1899; Furumark 1941: 3–6). It was based on evolutionary thinking (Åberg 1928; Furumark 1941: 3). Typological analysis, working on the assumption that similar forms are approximately contemporary, has been a central tenet of the archaeology of Israel-Palestine from the late 19^th century on, beginning with Sir Finders Petrie (Petrie 1904: 125; Davis 2004: 29–30) and continuing with William F. Albright (Albright 1934; Davis 2004: 66–76) and Kathleen Kenyon (Kenyon 1961: 153; 1979: 15), and is still a strong force in the field today (e.g. Mazar 2005).

In the following typological description of the Tel Kinrot pottery, I have aimed at a structured depiction of the excavated material: its date, nature, and position in the geographic and cul-tural setting of Israel-Palestine. In order to make comparative work easier for readers, the report follows the pattern of most pottery reports. It presents groups proceeding from open to closed forms, and ends with a chapter for various vessels that do not fit any of the defined classes. A list of distributions at Tel Kinrot and a list of parallels from literature follow the type descriptions. The typology (chapter 5.2) is planned to be a part of the site report, while the contexts will appear as separate chapters. Such fragmentation is typical for archaeological re-ports (Jones 2002: 46).

Constructing a typology requires that the researcher take on an active role in the process – the artifacts themselves are passive objects, and the types are based on the experience and evaluation of the scholar touching and looking at them physically, and visually as drawings and photographs, and reading their descriptions in find cards. The researcher also needs to decide what will be sorted: the boundaries of the material that will be included. Typologies have been criticized of artificial rigidity of classes and their presentation as independent of the observer (Shanks & Tilley 1987: 117; Pfälzner 1995: 10; Langin-Hooper 2011: 40–59; 2013; Gnecco &

Langebaeck 2014: v–vi). However, archaeology as a study of the material remains of past so-cieties cannot do without classifications. Human perception is profoundly dependent on grouping and creating types, and archaeology is not an exception (Bowker & Star 1999; see also Rice 1987: 274–288; Sinopoli 1991: 49–56).

A typology is a scholarly construction, and should not be regarded as are-construction of the ancient potter’s or consumer’s categories. My aim has been to create types that are as homo-geneous as possible within the types, with explicit and clear boundaries between the types.

Sharp boundaries enable an easy distinction of one type from another while sorting. The clar-ity aspired to for the typology stands in opposition to the nuances of the studied material. In order to ‘do justice’ to these nuances, I have also included a description of the variety within at least some features of the vessels that I grouped together in a type. I began the Tel Kinrot

typology building as anintuitive grouping of objects. By intuition I mean here a quick assessing of patterning in the material. I do not mean intuitive as something opposed to the rational, but rather as a thinking mode that precedes the rational, rule governed, and explicit argumen-tation (see e.g. Witteman et al. 2009: 39). After the initial groupings, I made the explicit type definitions based on the items that I had grouped together. I then used these definitions, to-gether with the earlier grouped items, as guidelines for sorting new material (for sorting and classification, see e.g. Adams & Adams 1991; Sinopoli 1991: 49–50). I do not believe in grasp-ing for the mental template orideal pot of the ancient potter (following Maier 2007: 242, Mullins 2007: 391; contra Panitz-Cohen 2009: 219). Consistent patterns in material culture that coincide with chronology or geographic regions may well have been irrelevant for the contemporary people, yet they are still informative for archaeologists.

As types are a means of communication, I did not consider it a drawback that I built the typol-ogy of the Tel Kinrot pottery on the well-established tradition of archaeoltypol-ogy in Israel-Pales-tine. On the contrary, using established terminology increases the understandability of the presentation. According to Mazar and Panitz-Cohen, establishing “a type series of some sort enables more meaningful study of the diachronic and synchronic aspects of pottery develop-ment, so that the advantages outweigh the disadvantages of this system” (2001: 12). In the beginning of the work I did not consider typology a problematic enterprise at all – it was the normal way of working with ceramics. I saw explicit types as an improvement over the artifact studies of the old tradition and their implicit types, where vessels were grouped together and labelled and descriptions were written for individual vessels. The most important goal was to create a typology that ‘works’ with the material from Tel Kinrot. A practical measure for the success of a typology is the (desirably small) amount of items that fall between the defined types. I have kept this number as small as possible without creating many subtypes, which easily make a typology tangled – in practice, it may often lead to describing a single vessel as a type (see section 2.5 above).

The types are different in their easiness of recognition. There are some types that are easy to distinguish (e.g. the carinated Krater KR04, or the Phoenician style jug JG04, see below). These types have distinctive morphology, and the vessels show a great similarity to each other. How-ever, in many cases it is difficult to differentiate between two (or more) close types. The groups differ from each other in their inner consistency: cooking pots and storage vessels seemed to have a few types only, and these types were uniform and easy to classify. This probably relates to their functions. On the other hand, kraters (a form combining features of storage jars, cooking pots, and bowls), bowls, and small containers were far more heteroge-neous groups, and I had much more difficulty creating clear and distinguishable types. Com-parable difficulties have been expressed by authors of pottery typologies (e.g. Mazar 1985;

Mazar & Panitz-Cohen 2001).

My starting point at Tel Kinrot

When I started the typological work for the KRP in 2003, I used earlier typologies as my guide in order to identify the vessel type for shards. I used Ruth Amiran’s (1969) standard typology

of pottery in Israel-Palestine as a basic reference, and I was familiar with the published pottery from Tel Kinrot (Fritz 1990) and two major sites in the Northern Jordan Valley, Tel Hazor (e.g.

Yadin 1958; 1960; ed. Ben-Tor 1989 and 1997) and Tel Dan (e.g. Biran 1989a, 1989b; 1994).

For Tel Dan, I also used the unpublished dissertation of David Ilan (1999). The classifications I ended up with have thus been influenced by other excavation reports, and their pottery presentations as well. The typologies that I used most often were those of Tel Qasile (Mazar 1985), the renewed excavations of Tel Beth Shean (Mazar 2006, Maier 2007, Mullins 2007, Panitz-Cohen 2009), Tel Yoqneʿam (Zarzecki-Peleg et al. 2005), Megiddo (Arie 2006), Timnah (Mazar & Panitz-Cohen 2001), and Tel Dan (Ilan 1999). The most helpful of these have been the reports from Tel Beth-Shean, since they have avoided using an overwhelming amount of types and sub-types, and were in many cases explicit in their type definitions. The report on the Iron Age pottery from Tel Beth Shean was especially useful, because the material is often similar to that of Tel Kinrot (Panitz-Cohen 2009).

The typology for the pottery from the 1980’s Tel Kinrot excavations (Hübner 1990; Fritz 1990;

section 3.4 above) was implicit, and followed in this sense the example of the large-scale ex-cavations at Megiddo by the University of Chicago and at Hazor by Yadin. The finds were iden-tified as belonging to broad groups of bowls, chalices, cooking pots, storage jars, pithoi, jugs, pyxides, flasks, or lamps. Further sub-divisions were avoided, and parallels were given for sin-gle vessels described in detail. The groups were based on vessel forms, with a certain associ-ation with assumed functions such as cooking, storage, and other household activities.

During the excavations by Fritz in 1994–2001, some distinctive wares were classified on find cards. These wares were distinctive ceramics, considered to be imported items produced somewhere else, such as Tell el-Yahud ware (from the Middle Bronze Age), shards of Myce-naean, Cypriot (so called milk-bowls or chocolate-on-white-ware, both Late Bronze Age), or Phoenician (decorated jug, Early Iron Age) imported vessels, and Abydos-ware and grain-washed ware (from the Early Bronze Age). There was no petrographic study used for such inferences, but the interpretations rest on the macroscopic evaluation of an experienced ar-chaeologist, Volkmar Fritz himself.

In the earlier report, as well as in the find cards of the 1990’s, the pottery that had painted decoration in red and black (or dark brown) was sometimes labelled bichrome ware, but this designation in the find cards was used somewhat inconsistently. The bichrome style was iden-tified only once in the preliminary plates for Kinneret II for a shard (6119/2), and only once for a jug in an article published in 1998 (Fritz 1998: 438, Fig. 11). The style was defined by Ruth Amiran in 1969, and further discussed by Amihai Mazar (1985), Ayelet Gilboa (1998, 1999), and Gilboa & Ilan Sharon (2003). Bichrome decoration is a style considered of Phoenician origin, typical for Iron Age I (Gilboa 1998; 1999). The style comprises rounded jugs with a high neck and rounded base (sometimes a ring base), and one bowl form, decorated in red and black with rather settled patterns of mainly concentric circles, and in jugs also other patterns.

Thus, not all pottery decorated in black and red was included, but the borders of the definition were not detailed by Gilboa. In the find cards from Tel Kinrot, several vessels that could have

been defined as bichrome style, according to the widely used definition by Amiran, were not identified as such. However, the term was not defined, and it remains unclear if this was on purpose or a sign of the publication process still being under construction. As there is no com-monly accepted definition for the term ‘bichrome,’ I decided to avoid the term and describe the decoration colors and patterns separately, as they would turn up during sorting. When writing the descriptions of the pottery types I also decided to refrain from the use of the term bichrome, as the patterns painted in red and black did not seem to follow such fixed patterning according to the arrangement and width of the lines as suggested by Gilboa.

When I started as the registrar in the KRP I had experience of two archaeological digs. I com-pleted my first field school as a volunteer in 1999 at Tel Kinrot, as an undergraduate student.

These excavations already provided me with an impression that ceramics played an important role during the excavations. During the summers of 2000 and 2001 I worked at Tall Mozan in North-Eastern Syria in a team analyzing pottery. These seasons were formative for my ap-proach in pottery studies. At Tall Mozan, we counted all shards and analyzed all rim parts and decorated shards in detail, with over 20 features including forming technique, colors, decora-tion, surface treatment, use wear, and rim type (Schmidt 2011; 2013; for a closely similar ob-servation set, see Pfälzner 1995: 10–12). The pottery at Tall Mozan was mainly from the Mid-dle Bronze Age (Dohmann-Pfälzner & Pfälzner 2002: 153–156). When I joined the Tel Kinrot excavations in 2002, I wished to continue working with analyzing all rims, which I considered would make the assemblage representative of all excavated ceramics. I was convinced that intuitive selection would lead to a biased collection that would not enable trustworthy results (similarly Pfälzner 1995: 5–7). The features discussed in the literature, as well as my experi-ence at Tall Mozan, affected my decisions about the details that I chose to record. I focused on details of the rim part, and the vessel form as far as it could be recognized. In addition, the reports I had used included descriptions of the clay material, the colors of the surface and possible decoration, and sometimes a description of the walls as thin or thick (e.g. Mazar 1985:

42; Epstein & Dothan 1989: 231, 249), so that I also thought that thickness was an important feature. However, I wanted to have a clearer and more ‘objective’ measure for the thickness, and decided to measure it with a scale of 0.1 millimeters.

Creating types

Guided by the literature, I set the emphasis for the type definitions of the Tel Kinrot ceramics on morphology, and in the case of cooking pots also on ware (see below). The system is hier-archic, with functional categories (groups such as bowls, jars or jugs) further divided into types (different “kinds” of bowls), and in many cases further into sub-types. I defined the major categories and types in ways comparable with other reports, such as those of Megiddo IV (Arie 2006) or Tel Beth Shean III (Panitz-Cohen 2009). In addition, I needed a class for the objects not further definable. Feeling uneasy with creating a complex taxonomy with many sub-types, I decided to create a typology for rims separately, so that the variations of the rim form (rim types) could be recorded without introducing sub-types. I assumed that similar rim forms could occur on different vessel groups, and that this might be chronologically significant.

I wished to create homogeneous types with clear boundaries between them. The clarity, how-ever, was artificial to some extent. The real classified items were always more or less well apt to the description that was made on the basis of the first artefacts studied. Thereal material to be sorted is always more fluid, and the types that serve as mental categories for the sorting mind are artificial constructions imposed on an unwilling object. The process is recursive: 1) inspecting the material and creating the first, tentative groups; 2) describing the groups formed; 3) sorting new material according to the descriptions; 4) creating additional types when needed; 5) trying to evaluate the uniformity of each type and re-considering some des-ignations in order to achieve as uniform groups as possible; and 6) editing the descriptions according to the material sorted. Thus the process goes back and forth between the real items and the abstract types.

In practice, the types were constructed in several stages. I formed the first, tentative groups with the help of line-drawings I had at my disposal from the excavation seasons of 1994–2001.

These drawings presented mainly well-preserved vessels. They were a helpful tool for identi-fying fragmentary material (the majority of the pottery from areas U and W excavated 2003–

2008 consisted of shards). The drawn material from the excavation seasons of 1994–2001 consisted mainly of well-preserved profiles, supplemented by selected rim shards and deco-rated shards. The first step of the work can be described as grouping together objects, and this is to a large extent an intuitive process. After the first groups were created in this way, I labelled and defined the groups according to their joint features. Thus the groups became types. I supplemented this preliminary typology during the field season while sorting the ce-ramic material – the real vessels and fragments – in August 2003. The objects found during the recording were then assigned to a type, or when they did not fit any existing type, I created a new type. Thus, some vessel and rim types were added during the sorting process. After the first season in 2003, I had the opportunity to study the material excavated during 1994–2001, over two weeks in December 2003, making notes about the forming technique, inclusions of the clay, size related attributes (rim diameter, full height, base width), and the state of preser-vation of the items, which is not always evident from the drawings. During April 2004 I stayed three weeks in Jerusalem going through the material at that time stored in the German Insti-tute of the Archaeology of the Holy Land. This time I added details to my notes from the first visit, and packed the material in order for it to be sent to the Israel Antiquities Authority (IAA).

During the excavation seasons of 2004–2008 I still added types if an item did not fit the for-mulated categories, but as the process progressed it became rarer to face such a need.

During the sorting process I was inclined to create new types rather quickly if a shard did not fit any of the existing types. This was for the practical reason that it is easier to combine groups later during the analyses than to separate them. Thus, the type forming process was flexible.

I regarded this as desirable, since more material was still going to be accumulated. However, flexibility creates a problem of ‘moving’ types. By this I mean a situation where the first, ten-tative types were created from the material that was at my disposal already in 2002, but the material that was later accumulated changed the types: both as to what could be identified as

belonging to the original types, and the amount of types alltogether. The set of defined pot-tery types in the beginning was not the same as the set of types defined at the end of the process. As new material was sorted and items were found that did not fit the first types, I created new types. The decision to create a new type was purely intuitive, as I had no criteria for that – which appears to be common (Sorensen 2015: 90). The items that were sorted at the end of the process thus had a different set of types with different type-definitions to be assigned upon than the items registered during the first seasons of my work. The items that were not close to ‘the ideal specimen’ of any of the types created in the beginning of the type formation process might have been classified to another type if they were sorted later in the process. In order to diminish this drawback of ‘moving types’, I checked a selection of items sorted during the first excavation season after the excavations were closed. In June 2010 I checked and photographed ceramics from 38 loci of areas U and W, altogether including 1247 pottery items. I selected the loci so that they were of stratigraphic relevance, and therefore more important for interpreting the site’s history. I made some corrections, but for me the amount of errant shards was surprisingly low. However, I at this point I did not document the corrections, and now cannot recall the exact number of those corrections. In October 2013 I still checked five loci (altogether 149 shards). This time I did not change any of the classifica-tions.

I started the typology building for the Tel Kinrot pottery material within the culture-historical setting. Even though I think the culture-historical framework has been too dominant in Israel-Palestine, I classify myself into the category of culture-historians. My initial interest was in looking for minor changes reflecting chronology, and in connections with similar artifacts from other sites in the region. I wished to define time dependent changeswithin vessel types and between the frequencies of the types, not only presence/absence data. Tracking differences that one could attribute to chronological factors meant that one should compare the frag-ments of the same vessel types in different strata to each other. In practice, one has to divide the material into groups according to both strata and vessel type for such analyses. Therefore, it is possible to reach statistically significant results for frequently occurring vessel groups, but not for types that include few items, as the differences can be the result of random differ-ences. In addition, the contexts varied as to their stratigraphic clarity, and to what degree the ceramics from each context could be considered to be contemporaneous with each other. The

In document From the Field to the Publication : The Retrieval and Presentation of Pottery : a Case study from Early Iron Age Tel Kinrot, Israel (sivua 129-147)