Corpus of Attic Vase Inscriptions
Home | History of the Corpus

History of the Corpus

By Henry R. Immerwahr, Prof. Emeritus, University of North Carolina at Chapel Hill (March 12, 2009)

The idea of a corpus of Attic vase inscriptions arose from a conversation I had in the 1940s with my friend Henry Hoenigswald (later professor of linguistics at the University of Pennsylvania), when I told him I was looking for a dissertation topic. I had spent some time at the American School of Classical Studies and had become fascinated by Greek inscriptions. He suggested a new edition of Kretschmer's book Die griechischen Vaseninschriften ihrer Sprache nach Untersucht (1894). It would have been an immense project, and the first step was to restrict it to Attic inscriptions. The next step was to focus on any weaknesses of what is still a very excellent and fundamental work, with a view toward possible improvement.

It was clear that nothing could be done unless a new collection of inscriptions could be used as the basis. Kretschmer had done a skilful job of collecting inscriptions, mostly from publications and from correspondence, but he was hampered by the fact that no systematic collection had been made since the fourth volume of Boeckh's great Corpus Inscriptionum Graecarum around the middle of the nineteenth century. In due course it became clear that this project (namely to replace the part of volume four that dealt with the Attic inscriptions) would be a work of many years and the whole project was not suitable for a dissertation. The idea of a corpus—and it was so stated by the editors of the Berlin corpus—was to admit nothing the editor had not seen or had had at least a mechanical copy (squeeze) of. This did not seem feasible, for it required lengthy visits to all the museums housing Attic vases and getting permission to handle all their vases to see whether they had any inscriptions, which in many museums would not be allowed. As for copies, they would of course be photographs, but many vases do not photograph well, while some museums allow only their own photographs (for which they charge a fortune). This was at a time before the invention of the digital camera which has made at least the technical side much more feasible. Another kind of copy, more subjective but potentially more accurate, was tracings, which would be very time consuming. I decided therefore to use the same methods Kretschmer had used: publications, pictures, inquiries; these to be supplemented by museum visits whenever possible. The only museum in which I have seen every Attic vase is the Museum of Fine Arts in Boston, through the great courtesy of Professor George Chase, then curator of the Classical collection (in the 1940s). Another great help was the permission granted by Professor Beazley at Oxford to let me see all his photographs in 1947. In other museums I was told I had to ask specifically which vases I wanted to see, which made a complete record impossible. The result of this methodology is of course not a true corpus but a collection of information about mostly published vases.

I should add at this point that we are dealing with a total of over 100,000 known Attic vases (a rough guess), of which (again roughly) one in ten has some inscription. There are certainly over 10,000 inscribed Attic vases. Rudolf Wachter, who is working on a project based on this corpus, conjectures that the number is about 4,000; my guess is that he did not include nonsense inscriptions and perhaps not graffiti made by users (as against inscriptions written in the workshop). Four thousand is probably about right for meaningful decorative inscriptions. In fact, Sir John Beazley, in 1947, advised me not to undertake this collection because I would not finish it. He was right, in the sense that my collection is not complete, although with well over 8,000 items it is extensive.

The next consideration was what to include.* It was clear to me that the inscriptions, being ornamental adjuncts to the vases, had to be presented in a context that included some information on each vase as a whole, e.g., provenance, date, and attribution and especially the subjects represented. As for the inscriptions, my original idea was to include all kinds of writing, whether painted or scratched, on an Attic pot, but I soon had to restrict myself. The graffito letters on the bottom of pots, most of which are testimony to ancient trade, were largely omitted if the pot had no other inscriptions. A substantial collection of them (although also not complete) has been published by A. Johnston (Trademarks on Greek Vases in 1979; Addenda in 2006). Then there are the Western Greek and other non-Greek (e.g., Etruscan) graffiti, which have their own literature. Finally, I have not fully excerpted a number of collections of vase inscriptions that can be easily consulted, such as M. Langdon's A Sanctuary of Zeus on Mt. Hymettos (1976), or the collections of ostraca and graffiti from the Athenian Agora published by M. Lang.

Once, however, a vase was to be included, all writing, including merchants' marks, etc., was to be part of its record. I also took a large view of what inscriptions were to be included, first of all the inscriptions produced in the pottery workshop, whether they made sense or not. I made the decision that the so-called nonsense inscriptions (of which there are thousands) had to be included, for the corpus was to be of service not only to linguists but also to archaeologists, and the nonsense inscriptions are a part of the ornamentation of the vase, which varies from workshop to workshop. They also raise the important question of the state of literacy in ancient Athens. Likewise to be included were the ubiquitous formulas καλός and ὁ/ἡ παῖς καλός/καλή, again for their different use in different workshops. There is an interest in the style as well as the meaning of these inscriptions. The basic premise is that the inscriptions are usually put on by the vase painters, although there are exceptions to that rule in that sometimes certain inscriptions (such as signatures) may have been put on by the potters. For example, the signatures of the potter Hieron are mostly found incised on the handles of his cups and do not always exhibit the same letter forms as the painted inscriptions in the scenes. In general, these inscriptions, although written in simple capital letters, are autographs, and it is possible in many cases to study aspects of handwriting.

In the museums, when I was fortunate enough to be permitted to have the vase moved out of its case so that I could handle it, I used very simple methods to establish a reading. I had with me a soft pencil and paper slips or a note book, tracing paper, a small six-power magnifying glass (which enlarged the individual letters to the size of letters on stone inscriptions) and a rectangular pocket mirror (to hold at right angles to the surface, which brings out its irregularities). A small pot of water was also sometimes useful. Alcohol, however, needed the presence of a curator, for it could delete modern restorations that needed to be recorded before being erased. The magnifying glass could tell you whether a graffito was inscribed in the soft clay (hence in the workshop before firing) or in the hard clay, which always raises the specter of forgeries, of which there are a good many. It is amusing to read in the Annual of the British School at Athens (BSA 50 [1955] 291ff.) about the activities of a late-nineteenth-century forger in Athens who produced, among many small inscribed fragments, what looked like a long political text on a black-glazed pottery fragment originally published as genuine in BSA 48 (1953) 190ff. Beware of early Attic graffiti that use a cross for punctuation! But the main problems for a reader are the painted inscriptions on red-figured vases, because they were done in applied paint which has often flaked off. However, the area covered by the paint did not acquire the famous shiny glaze that the rest of the pot received in the furnace and shows up as a matte area, which is frequently overlooked. At any rate, I did not usually ask for special technical help such as ultra-violet photographs, and my own photography was not always successful because of lighting conditions.

I started with paper slips and had help from a number of graduate students and some financial support, although it is difficult to obtain any long-term supports, especially in America where foundations like to see quick results. I received a further impetus from a letter of A.E. Raubitschek, in which he considered it a desideratum that each inscription should be illustrated. Since for reasons mentioned above I found this impossible, I decided to write a small companion volume with photographs, which I am afraid got rather more comprehensive, including as it does comparisons with inscriptions on other materials, by the time it appeared as Attic Script: a Survey (1990). This book makes no attempt at improving Kretschmer's book in any way and is purely epigraphical. I did not realize at the time that its main interest would be in the lists of inscribed vases or I would have given more lists and made a better Greek index.

While the book was still written on a typewriter (only the indices were made by computer), there had in fact been a crisis at an earlier time when I realized that the amount of paper involved was overwhelming and my files were hard to handle. At that point the computer intervened and saved the project. This was not an unmixed blessing, for the tendency of the computer is to schematize. I devised a format of fifteen fields with a view of constructing a database: corpus number; collection; provenance; shape; attribution; date; bibliography; subjects; inscriptions; letter forms; comments; footnotes; tags (key words for searching); personal notes; and an 'end' field. In looking for a suitable Greek font I chose Ibycus, a small computer system developed by David Packard for use in Greek, Coptic, and Hebrew studies. I decided to use Greek lower case letters and the diacritical signs available, while handling letter forms in a chart and a special descriptive field for them (this has not been made public). I have never thought much of attempts to render epichoric alphabets with printed Greek letters. On paper one is freer to add small observations and make drawings. But the greatest advantage of the computer is that no entry has to be permanent and that in fact a corpus as a database is infinitely expandable. I believe corpora today of whatever kind should no longer principally be published in print, for they are out of date as soon as they are published. Hard copy is useful only for archiving.

Coming to the end of my academic life I was anxious to make available the imperfect results of my collecting to whoever could draw a profit from it. With this end in mind, I transferred the data from Ibycus to Macintosh in a somewhat simplified form, reducing the fifteen fields to seven and printing a number of hard copies for the libraries of the American School of Classical Studies, The Beazley Archives at Oxford, The Getty Museum, and some others. I have also distributed the computer version to a number of interested parties. In no case have I insisted on licensing or any reliance on my copyright because the work is not a finished product and I want it to be used freely by all interested parties.

The next form this project takes will be Wachter's AVI (note that it is no longer called a corpus), and it will also be in electronic form which will make it possible to update it continuously. This is especially important for vases, for new material is constantly produced by excavation and tomb robbery. In addition, some vases, especially those in private collections, travel around, disappear, and reappear again. A prime example of a lost vase is a bell krater by the Dinos Painter with the inscription Ψόλων καλός ('Dusty' is fair) which Beazley had known only from an eighteenth-century publication. Recently it found itself in a private collection in Canada and was broken by movers, which caused the fragments to be given to the Royal Ontario Museum in Toronto. Even the surfaces of vases do not remain constant: some, especially in museums, are cleaned and thus liberated from false restoration, others on the contrary 'lose' their inscriptions when a dealer puts a plastic coat on them to make them more salable. I suspect that we have an example of this on a small cup by Douris on which a youth, ready with money bag, is buying wine in a wineshop, specifying the quality he wants by saying τρικότυλος, which describes vin ordinaire. I published the inscription many years ago (Transactions of the American Philological Association 79 (1948) 184ff.) when the vase was in a private collection in America; it is now in private hands in England and the inscription is said to have become illegible. Altogether the study of vase inscriptions is a moving field of research, and electronic form is the only way to deal with it.

* For a discussion of the types of inscriptions found on these vases see "A Projected Corpus of Attic Vase Inscriptions," Acta of the Fifth Epigraphic Congress (1967), or G.M.A. Richter, Attic Red-figured Vases (1958) 14-21.