Pottery Descriptions in the Excavation Reports of Timnah

Probably the most detailed accounts of the process of making a typology are included in the reports of Qasile (Mazar 1985) and Timnah II (Mazar & Panitz-Cohen 2001). Shorter descrip-tions are included in the excavation reports of Timnah III (Panitz-Cohen 2006), Tel Beth-Shean (Mazar 2001, Maier 2007, Mullins 2007 and Panitz-Cohen 2009), Tel Yin’am (Liebowitz 2003), and Megiddo (Arie 2006; Arie 2013a, 2013b). The report on Timnah II is detailed in its account of type formation, and thus more explicit than most reports in this regard. Therefore, I will consider it here in some detail. In general, its arrangement and style can be considered typical for the site reports.

The processing of the pottery at Timnah was based on a system earlier used at Tell Qasile, but modified (Mazar & Panitz-Cohen 2001: 10). Tell Qasile was excavated from 1971–1974, and

the publication of the stratigraphy and architecture was published in 1980 (Mazar 1980). The volume presenting the finds came out in 1985. Already at Tell Qasile, the process was to keep and analyze all rims and decorated shards, and a selection of handles and bases from all strat-ified loci (Mazar 1985: 21–22). At Tell Qasile, the pottery of strata XII –VIII (from 12^th to 10^th century, the focus of the 1970’s excavations) is presented in a type-series, where each vessel type and sub-type is discussed separately. Following this method, a detailed type series was built at Timnah. The type descriptions concentrate on form and surface treatment. The fabrics are discussed separately, and while the fabric groups are included in the section for pottery, it forms an isolated sub-chapter that is rarely referred to in the typological presentation in the following sub-chapter (Mazar & Panitz-Cohen 2001: 15–24; 30–144; 157; see also Panitz-Co-hen 2006; CoPanitz-Co-hen-Weinberger 2006).

The orientation of the artifact studies was essentially chronological and culture-historical (Mazar & Panitz-Cohen 2001: XI; 10; Panitz-Cohen 2006: XIII). In addition, the material was used for spatial-functional analyses of the architectural units (Mazar & Panitz-Cohen 2001:

10). The pottery was studied according to a typological classification in conjunction with stra-tigraphy (Mazar & Panitz-Cohen 2001: XI). Material in the plates is arranged according to ar-chitectural units. Pottery presented in the plates includes all the complete vessels and a “large and representative collection” of shards. A “maximum amount” of the homogeneous assem-blages’ content was presented as “objectively and comprehensively as possible”, reflecting the development of pottery. Shards that could be classified typologically and originated from stratified loci were also processed (Mazar & Panitz-Cohen 2001: 1). The selection of pottery to be drawn was explained as well: “the pottery selected for drawing is a representative sam-ple, depicting entire assemblages alongside examples of all types from every stratum. Repeti-tion in the graphic presentaRepeti-tion was often found necessary in order to emphasize small varia-tions” (Mazar & Panitz-Cohen 2001: 10–11). “The emphasis is on presenting all the complete vessels and representative shards from well-defined units belonging to each architectural and stratigraphic horizon” (Panitz-Cohen 2006: 3). It is noteworthy that the term “representative”

is used in a twofold meaning. First, it defines a collection of shards, giving the impression that the collection is a representative sample of all pottery shards; later, the term defines shards themselves as if it would refer to certain kinds of shards, like those with a larger part pre-served.

The Timnah type-series is taxonomy with aclass as the uppermost category, and the type with a narrower definition in the next step. Types are sometimes further divided into sub-types.

There are ten functionally determined classes in the discussion of strata IV–II (bowls, chalices, kraters, amphorae, cooking pots, storage jars, jugs, juglets, bottles, and lamps) and 21 classes in the discussion of strata XII–V (in addition to the classes above there were goblets, pithoi, biconical vessels, stirrup jars, flasks, a pyxis, a cup-and saucer, funnels, stands, imports, and amphoriskoi instead of amphorae). In the presentation, several types are grouped under com-mon morphological headings. The types were determined on several criteria. Preferably whole forms were used as prototypes, but several types were also defined on the basis of rim form.

A fundamental practical necessity for a type definition was the ability to attribute shard ma-terial to a certain type (Mazar & Panitz-Cohen 2001: 11; see Adams & Adams 1991: 77–89). It is stated that the authors aimed at type definitions that would be mutually exclusive. This was done by creating broad categories for types and small variations for differentiating sub-types.

The method proved difficult, and it is admitted that types differ in their homogeneity, which is discussed for each type separately. It is also expressed that the typology is subjective and has no intrinsic value, but is a tool for organizing and discussing the material in a meaningful way (Mazar & Panitz-Cohen 2001: 11–12; see also Adams & Adams 1991: 48–49). An idea of

“true” types is related to the variability of the classified material: a type can be considered true if the vessels are homogenous in form (Mazar & Panitz-Cohen 2001: 12).

From strata IV–II, 15 002 vessel items were analyzed, including shards and complete vessels (Mazar & Panitz-Cohen 2001: 12). The amount of data was reduced in quantitative analyses to include only complete vessels andindicative shards, with a total of 7027 pottery finds: 1408 well preserved vessels and 5619 shards (Mazar & Panitz-Cohen 2001: 24). The material ana-lyzed from the earlier strata XII–V was less, 2806 finds all together, where 2251 were shards (Panitz-Cohen 2006: 10). This material includes a time sequence from Middle Bronze Age IIB to Iron Age I/IIA (Panitz-Cohen 2006: XIV).

In the presentation of the smaller corpus of material from a longer time sequence (Panitz-Cohen 2006: 27–120), the amount of types and sub-types is especially large. For example, the Bowls(1041 items) were divided into 28 types and 15 vessels were illustrated as varia. There are eleven types with five or less finds, and four types with one or two finds.Kraters (a deep vessel form between bowls and jars) included 252 items all together, and were divided into 11 types with 38 vessels being described as varia. Of these types, six include two or three items only.Storage jars (515 finds) were divided into 11 types (and 51 in varia). Two types included three or four finds. Both of these types were defined based on a feature that could not be observed on most shards (painted decoration and four handles on body).Juglets (50 finds) were divided into three types, two of which included two finds and the third type eight finds.

Most of the juglets were confined to the broader definition of the class only (38 finds). This class well exemplifies what happens when types are formed based on complete vessels, but the sorted material mainly consists of shards: many types do not really materialize. Flasks numbered all together 14, divided into three types, two of which only included one find. Type FL3 (flask with a spoon) is defined according to a feature that can only be observed if the rim part is preserved (3 of the 7 illustrated examples were not). Thus the types are not mutually exclusive. Amphoriskoi (a specific jar type with handles at the neck or rim) number only three, and still they are divided into two types, one of which is still divided into two sub-types – each vessel is thus described individually. The typology of the pottery from Yoqneᶜam is in many respects similar to that of Timnah in the structuring and division of types and sub-types (Zar-zecki-Peleg et al. 2005).

Types that only include one vessel from the site are a relatively common phenomenon in the pottery typologies. There were more than ten single-item-types or sub-types in the report on

Qasile (Mazar 1985) and the Iron Age I typology from Megiddo (Arie 2006), and seven in the report on the Late Bronze Age IIB pottery from Megiddo (Martin 2013). The resulting com-plexity of types and subtypes is not exceptional, as is demonstrated by the table in Fig. 2.2.

The sites included in the table include Iron Age pottery presentations, in order to have to some extent comparable assemblages. Some representations prefer many types defined in detail, and others tend to have broad groupings including variability. It is probable that the material and its heterogeneous or homogeneous nature, as well as the aims of the sorting researcher and his/her sensitivity towards the sorted material, all play a role.

Site report time sequence classes types in a class total of types

sub-types total of types

& sub-types Mode³ mean𝑥̅ max Mode⁴max total

Beth-Shean I Iron I 11 3 & 5 3 5 33 0 4 19 52

Beth-Shean I Iron II 13 1 4.8 11 62 2 2 6 68

Beth-Shean III Iron I 16 1 2.9 9 46 2 7 22 68

Dor I Iron II 11 2 12-13 40/47 96 - 104 2 4 52 148/156

Megiddo IV Iron I 16 2 4.6 11 69 2 4 27 96

Qasile Iron I 15 2 4.6 18 65 2 4 23 88

Tell Qiri Iron I 22 6 4 7 52 3 6 63 115

Timnah II Iron II 10 2 9 26 99 2 5 41 140

Timnah III MB IIB–Iron I 21 3 5.2 23 89 2 8 38 127

Yoqneᶜam II Iron I – II 12 3 6.2 11 74 5 24 66 140

Fig 2.2 Amount of presented classes, types and sub-types in the pottery presentations. References: Dor (Gilboa 1995), Qasile (Mazar 1985), Megiddo (Arie 2006), Timnah II (Mazar & Panitz-Cohen 2001), Timnah III (Panitz-Cohen 2006), Tell Qiri (Hunt 1987), Yoqneʿam (Zarzecki-Peleg 2005), Beth-Shean I (Mazar 2006a for Iron Age I, Mazar 2006b for Iron Age II) and Beth-Shean III (Panitz-Cohen 2009).