Stories & News

Archaeology, philology and science: a case of productive synergy from the Cretan Late Bronze Age

By John Bennet, Professor of Aegean Archaeology, University of Sheffield

It is much clearer in 1979 than it was in 1965 that no other Aegean pottery problem potentially capable of solution by analytical methods has so much promise as the ISJs. (Catling et al. 1980: 53)

Cover of my personal offprint from the 1980 article with signature ‘JFC’

These words appeared in the introduction to a substantial article published in the Annual of the British School at Athens: ‘The Linear B Inscribed Stirrup Jars and West Crete’, by Hector Catling, John Cherry, Richard Jones and John Killen. It presented one of the first major programmes of research carried out by the Fitch Laboratory in the early years of its existence. It also impacted me personally, since it appeared in the first year of my doctoral research, supervised by John Cherry, one of the authors.

As I came to appreciate, the article had considerable significance for my research, both specifically in relation to my investigation of Crete in the Late Minoan II-IIIB period, but more generally because it went to the heart of the effective combination of archaeology and Linear B, facilitated by archaeological science and statistical approaches.

In many ways the article was a vindication of a vision first articulated by two important figures in the BSA’s history. Sinclair Hood[1] suggested to Hector Catling[2] that elemental analysis – being pioneered at the time in the Oxford Research Laboratory for Archaeology and the History of Art (RLAHA) – be applied to Minoan and Mycenaean ceramics to determine their provenance.[3] But what exactly was the ‘Aegean pottery problem’ to which the authors refer?

The ‘Aegean pottery problem’

The fundamental question the 1980 article answered was: where were the stirrup jars inscribed with Linear B script, before firing, produced? A first ‘answer’—east Crete—had come in 1965 from early work by the RLAHA (Catling and Millett 1965), with compositions closely resembling those of Palaikastro and Zakros. As we’ll see this conclusion came increasingly to be considered the ‘wrong answer’ on both philological (Linear B) and scientific grounds.

So, what is an (inscribed) stirrup jar (henceforth ISJ)? These coarse ware vessels, typically about 40cm high with a capacity of 12-14 litres, have a characteristic pair of handles—shaped like a stirrup, giving them their English and German name (Bügelkanne)—linked to a central false spout (hence their Greek name ψευδόστομος); the real spout is offset and is occasionally found with a clay stopper. They originally contained liquids—oil or wine—and were used as storage and transport vessels, the amphora-equivalents of the Aegean Late Bronze Age. Inscribed examples were first discovered on the Greek mainland: a single complete jar at Orchomenos in 1903 and fragmentary examples in early excavations at Mycenae and Tiryns. A major find came in 1921, when Antonios Keramopoullos excavated 80 or so (inscribed and uninscribed) in space Δ of the so-called ‘Kadmeion’ in Thebes,[4] a discovery that informed his pioneering article on Late Bronze Age trade a decade later (Keramopoullos 1930). The decipherment of Linear B in 1952 meant that the inscriptions could be read, and Michael Ventris and John Chadwick (1956: 212) noted in the first edition of Documents in Mycenaean Greek that a place-name found in the Knossos documents (o-du-ru-wi-jo: KN C 902.2) also appeared on a Thebes ISJ (TH Z 839), a first inkling that their origin may have lain on Crete.

Image of the ISJ found in the Unexplored Mansion at Knossos (KN Z 1716); it bears the inscription wi-na-jo. Two other ISJs, bearing the same inscription are known from Armenoi on Crete and Midea on the mainland (source: Popham et al. 1984, pl. 110 a-c)

The origin of the ISJs was important because, until the discovery in 1939 of Linear B clay tablets—which must have been written in situ—at Pylos, the ISJs might have been imports from literate Crete to an illiterate mainland. Discovery of the Pylos tablets not only confirmed the active use of Linear B on the mainland, but also ignited a surprisingly bitter controversy, initiated in Oxford by Leonard Palmer, over the dating of the Knossos tablets that even reached the letter pages of national newspapers. Palmer argued that the Knossos and Pylos tablets were so similar that they must be of the same date; since the dating of Pylos was fixed in LH IIIB (c. 1200 BCE), then the Knossos tablets – whose precise chronology was less clear – must also date to the same period. His view was vigorously opposed by John Boardman (and others), resulting in the unusual publication of two books in one: On the Knossos Tablets (Palmer and Boardman 1963). By now, more place-names had been recognised on mainland ISJs, including an example from BSA–Archaeological Society of Athens excavations at Mycenae (Palmer 1959), further implicating the ISJs in the dating controversy: if the ISJs could be dated, then they would help confirm the date of the Knossos archive.

East or West?

Within the fast-developing community of Mycenologists, the Cretan origin of the ISJs was fairly clear, but, in the same year as the first provenance study appeared, Gillian Hart pointed out that the associations of some of the place-names on the ISJs were with western, not eastern Crete (Hart 1965: 15-19). Palmer developed this strand in a series of articles entitled ‘Mycenaean Inscribed Vases I-IV’ (Palmer 1971; 1972; 1973; 1978). Grist to Palmer’s mill in seeking to demonstrate a late date for the Knossos archive was the discovery, in 1968, in the Unexplored Mansion at Knossos, in BSA excavations by Mervyn Popham and Hugh Sackett, of an ISJ in a LM IIIB context, as well as other examples in Chania from 1964 onwards, further strengthening the ISJs’ west-Cretan associations. These arguments undermined the analytical conclusions, but further criticism had also emerged of the science, particularly in relation to how associations between ISJ compositions and those of ‘local’ samples were arrived at (e.g., McArthur and McArthur 1974; Wilson 1976; McArthur 1978).

With the establishment of the Fitch Laboratory in 1974, elemental analysis of Aegean ceramics moved to Athens, led by its Research Officer Richard Jones. The Lab initiated a substantial analytical programme involving the original ISJ material plus new samples, including the Chania material. It is testament to the good will of the Greek Archaeological Service and individual excavators that such a programme was possible. The analytical results were published in Archaeometry (Catling and Jones 1977) and now confirmed the west Cretan origin of many of the ISJs, a conclusion celebrated in the last of Palmer’s series of articles: ‘Mycenaean Inscribed Vases IV: Final Agreement’ (Palmer 1978).

But more was yet to come. Tony Wilson (1978) had advocated the use of multivariate statistics in discriminating more effectively between compositional profiles and these were applied to the new analytical data. John Cherry had already used the approach in an earlier, high-profile provenance study of Marine Style ceramics from the mainland and Crete (Mountjoy, Jones and Cherry 1978). The last element required to offer a definitive answer to the questions posed was to involve a Linear B scholar and John Killen, John Cherry’s Cambridge colleague and one of my mentors during my doctoral research. Killen explored patterns in the provenance data against the Linear B inscriptions.

Resolution (for now)?

The 1980 article provided a definitive answer to the question of the origin of the ISJs, but the Fitch’s involvement with ISJs did not end there. With the establishment of ceramic petrology at the Lab in the early 1980s, leading to the development of the Fitch’s trademark integrated ceramic analysis, my former University of Sheffield colleague, Peter Day[5] enriched the study with petrographic analysis. For this new programme, yet more vessels were analysed elementally and petrographically, including many non-inscribed examples. Halford Haskell, who had been working on refining the typology of these vessels, joined the team, and the vessels came to be known as Transport Stirrup Jars (TSJs) to reflect the inclusion of non-inscribed examples. The result was a book-length presentation (Haskell et al. 2011) that can probably be considered the final word, for now at any rate. In reviewing it, I concluded:

This volume is a thoroughly documented case-study showing how a question of ceramic provenance linked to the economic operation of complex entities can be tackled through a combination of approaches (Bennet 2012: 577).

Bibliography

Footnotes

[1] BSA Director, 1954-1962.

[2] Then a Keeper at the Ashmolean Museum, later BSA Director, 1971-1989.

[3] See Fitch Blog 7 December 2023 on the Oxford Laboratory’s role in the establishment of the Fitch.

[4] Pictured here on display in the Thebes Museum

[5] Williams Fellow in Ceramic Petrology 1984-1986.