Ancient prescriptions in tablet form

Twentieth-century research on the inscriptions and writing systems of the prehistoric Aegean has been marked by two big bangs and, since the late 1940s, by steady and well-

organised international collaboration. The first big bang occurred in March/April 1900, when in the first season of excavation of "the Palace at Knossos" Sir Arthur Evans and his team discovered evidence "sufficient to establish the certain existence of Mycenaean writings on clay" ( Annual of the British School at Athens [ ABSA ] 6, 1899-1900). Evans described the first document found as "part of an elongated clay tablet with a chisel-like end, engraved with what appeared to be signs and numbers". It is fitting, in reviewing a corpus edition of epigraphical frustula , that the first trace of writing in "linear script B" unearthed by Evans was so fragmentary that he copied it and set it aside in what he called "a suspense account". The second big bang, of course, was the decipherment of Linear B which Michael Ventris announced on the BBC on July 1 1952. Ventris himself declared that careful editorial and script-analytical work by the American scholar Alice Kober in the late 1940s was an absolute prerequisite for his decipherment.

Both big bangs sent shock waves radiating outward well beyond the domain of specialist scholars towards a general public interested in Homer. The quality of Evans's and Ventris's scholarly "teams" and methods also set high standards for subsequent archaeological field work and for the continuing study and publication of Aegean Bronze Age texts.

Ventris's decipherment "working group" met for the first international Mycenological colloquium at Gif-sur-Yvette outside Paris in April 1956. By then Linear B inscriptions on clay had appeared at Mycenae, Pylos and Knossos, and there was a now-fulfilled promise of more.

Would-be editors of these documents faced major challenges. The script makes use of about 90 phonetic and 200 ideographic signs, some with values still unidentified. The scribes use peculiar book-keeping formats and distinctive textual features, such as varying the height of signs for semantic stress. The first official resolution of the scholars at Gif laid out the requirements for corpus publications. Ideally, each document is to be presented in three ways: a photograph at 1:1 scale, a drawing from autopsy of the actual document, and a transcription. The Corpus of Mycenaean Inscriptions from Knossos ( CoMIK ) volumes adhere to these standards.

Each document is to be preceded by a lemma that describes precisely its physical characteristics. Already participants at Gif were aware that these records must be interpreted as archaeological data within archaeological contexts. The CoMIK volumes do not give this information.

Each corpus volume is also to contain relevant archaeological and palaeographical information: dates of discovery, find contexts and assignments to scribal hands and palaeographical schools. The CoMIK volumes give most of this. The Gif resolution further prescribed presenting the documents according to their inventory numbers. The CoMIK volumes follow this procedure, and CoMIK 4 provides the recommended double table identifying the tablets according to classification by series and sets.

Within the last quarter-century, definitive corpus volumes of superior quality have appeared for documents in all three of the Minoan-Mycenaean scripts originally identified by Evans: Cretan hieroglyphic (1996), Minoan Linear A (five volumes 1976-1985), and now Mycenaean Linear B from Knossos (four volumes 1986-1999).

That nearly a full century of work on the Linear B texts from Knossos ends with a whimper, relatively speaking, has to do mainly with practical editorial decisions. As a result of culling for readable and substantial texts - recall what Evans did with the first fragmentary tablet he found - the smallest fragments have the highest inventory numbers and were left to be edited last. Over 1,300 of the 1,363 inscriptions in CoMIK 4 are small fragments that cannot be joined with fuller extant documents despite painstaking work, beginning with Emmett L. Bennett, Jr in the 1950s and culminating with the masterly joins and quasi-joins identified by Jose L. Melena and his collaborators in the 1990s. A volume of epigraphical fragments in a script readable by, generously estimated, 100 scholars worldwide might attract little attention from a general public no longer educated within a classical tradition, but it caps a century of extraordinary achievements in understanding all facets of the earliest intelligible records in western civilisation.

The four CoMIK volumes together constitute the official corpus of the 4,147 or so Linear-B-inscribed clay tablets, labels, sealings and unjoined fragments from final-palatial Knossos (c. 1490-1200 BCE). The editors have followed the Gif resolution and published the inscriptions consecutively according to the numbers assigned to the individual tablets or fragments during excavation or museum inventorying; thus Fp 1 opens volume 1 and Xf 9947 closes volume 4. The letter prefixing comes from Bennett's pre-decipherment system of tablet classification, used first in his edition of the Pylos tablets (1951). Then in his Minoan Linear B Index (1953) he applied it to the Knossos tablets as edited in 1952 by Sir John L. Myres.

Myres's Scripta Minoa II finally made available more Knossos material than the small selection offered by Evans in the fourth and final volume of his monumental The Palace of Minos (1921-1935). Scholars who decry how long it took to produce the first corpus edition of the Knossos Linear B tablets generally overlook the fact that publication of even a representative sample of Knossos tablets in Palace of Minos IV took more than three decades. It barely preceded the fateful day in 1936 when Michael Ventris and his classmates at Stowe School heard about Knossos and Cretan writing from the aged

Sir Arthur in person at a Minoan exhibition at Burlington House. It is no accident that Ventris's decipherment came during a flurry of publication work on tablets from Pylos and Knossos.

All Linear B tablets are now identified by series classification. Discrete "sets" of records within the Knossos material may be further specified by Arabic numerals in parentheses. Thus an informed reader knows that Fp(1) 1 and the ten other tablets of this set by scribal hand 138 are of the page-shaped type ("p") and refer ideographically to a commodity such as olive oil ("F"). The Fp(1) tablets record for lexically specified months individual allocations of 1.6 to 28.8 litres of olive oil to deities like Diktaean Zeus, Erinys, "all the gods", and the apparently non-Greek pa-de and pi-pi-tu-na , to a "priestess of the winds", and to specific locales with religious associations (Amnisos, Dikte, u-ta-no , the Daedaleion).

This set can be contrasted with four other Knossian oil tablets: Fp(2) 354, Fp(2) 363, Fp(2) 5472 and Fp(2) 5504 - all possibly by scribal hand 222. The Fp(2) tablets also record olive oil, but they specify in one instance that the oil is "to be paid as a religious fine". Three of the four texts lack month names, and the reused and cut tablet Fp(2) 5504 records a single entry of possibly 288 litres. A final fragmentary tablet of the Fp class is a small singleton Fp 7707. It preserves two entries of the ideogram OIL in much larger quantities still: 3,504 and at least 4,809.6 litres.

The CoMIK system leaves the four Fp(2) tablets spread out over two volumes and the 11 Fp(1) tablets occur on nine different, mostly non-consecutive pages of volume 1. This is a serious drawback to the utility of the corpus that cannot be remedied completely by using the volumes simultaneously with the edition of transliterated texts according to series classification by John Killen and Jean-Pierre Olivier, The Knossos Tablets (1989), in which tablet Fp(1) 13 was accidentally omitted. The CoMIK volumes also do nothing to correct the seemingly erroneous ascription of tablet Fp(1) 48 to the Room of the Chariot Tablets. They also leave out the lemmata mandated by Gif, a lacuna that cannot be filled by reference to the separate palaeographical studies of Olivier and Jan Driessen.

We have made great advances in our understanding of Mycenaean record-keeping procedures. Palaeographical assignments and the archival grouping of tablets by find-spots have been ably exploited in the study of Mycenaean dialects and scribal idiolects and in debating the relative chronology of different sites and different areas of the palatial complex of Knossos itself. By adhering to the Gif resolution about inventory numbers and thereby failing to incorporate fully the results of two generations of fundamental work with the archival aspects of the Knossos tablets, CoMIK falls short of achieving the status of an independent reference tool.

Still, CoMIK 1-4 enables Mycenologists worldwide to look at Evans's first published photograph of 13 "Tablets with Linear Script" (ABSA 6, 1899-1900, pl. 1), read each tablet, supplement several with joining fragments, study the other work of the scribes who wrote them, and place the left-hand tablets together in a single area of the Palace of Minos dealing with religious offerings of agricultural products (olive oil, honey, wheat, spelt). In CoMIK 4 they can also find additional fragments of related texts, such as Gg 8053. The epigoni of Ventris and Evans have done well by their inheritance.

Thomas G. Palaima is director, programme in Aegean scripts and prehistory, University of Texas at Austin, United States.