What are lexicographic 'slips'?
They are the records of all the textual passages which are cited in a lexicon's entries. Traditionally, every citation was written on a separate 'slip' of paper, which was then filed with the others illustrating the same headword.
But I thought that the Cambridge Greek Lexicon was designed for students, and does not cite specific passages, as Liddell-Scott-Jones (LSJ) does?
True, but we do describe the specific word senses used by the individual authors. This is possible because we consult the texts as we write the lexicon entries. The contexts show how each word is used, and so provide the 'raw material' which we use to determine the meanings.
Why don't you just copy the meanings from an existing dictionary, such as LSJ?
Because LSJ, though giving an excellent range of references, does not give very detailed definitions for each word. Many of its entries give only single-word translations, rather a full description of the range of a word's senses, and the ways in which these may change over time.
Additionally, new words, and new senses of words already known, have been discovered since the publication of LSJ and its Supplements. So we need to look at all the textual sources, to ensure that we interpret the senses correctly.
Do all dictionaries use 'slips'?
All scholarly and historical dictionaries use illustrative quotations. However, the way they are collated may differ according to circumstances. The citations used in LSJ were gathered over more than a century, accumulated during the eight previous editions of the lexicon which were edited by Liddell and Scott. And even their first edition of 1843 took many of its citations from Franz Passow's Handwörterbuch der griechischen Sprache of 1831, which itself drew on Johann Schneider's Griechisch-deutsches Wörterbuch of 1819. The only record of all the passages cited is now provided by the brief quotations in LSJ itself.
In contrast, the Oxford English Dictionary gathered its vast collection of slips in one massive logistical operation over a period of 50 years, by calling on the services of thousands of volunteer scholars throughout England and abroad. For examples of slips written by J.R.R. Tolkien and others, see the OED web site.
Here is a picture of the OED 'scriptorium', where the case to the left of the Editor, James Murray, contains some of the slips which he and his team used as they wrote the entries.
Here's James Murray's description of how the slips were used when writing an entry:
'Only those who have made the experiment, know the bewilderment with which editor or sub-editor, after he has apportioned the quotations ... and furnished them with a provisional definition, spreads them out on a table or on the floor where he can obtain a general survey of the whole ... shifting them about like pieces on a chess-board, striving to find in the fragmentary evidence of an incomplete historical record, such a sequence of meanings as may form a logical chain of development' ('The President's Address for 1884', Transactions of the Philological Society, 1882-4, 510-11).
What has happened to these slips?
The Oxford English Dictionary slips are stored in oak card-cases, where they are still consulted by the current Oxford editorial team. However, slips from LSJ have not been preserved, for the reasons given above, so we have had to start again from scratch.
Do you have thousands of volunteers to collect the slips, or do you look them up in an electronic database like the Thesaurus linguae graecae (TLG)?
We have a small team, so we have to use electronic means. Even then, it would be much too time-consuming to look up every citation separately, even using the TLG. We need to have some way of collecting all of them in a searchable electronic archive, so we can instantly retrieve all the citations for each word.
Does such an archive exist?
Professor Jeff Rydberg-Cox of the Perseus Project has designed one for us. Jeff has built the database, by using texts from the Perseus library and the TLG, linking them with English translations where possible, and collecting them under each word that will be in our lexicon.
Surely it would be impossible to examine EVERY instance of a particular word? Isn't there a way of making a representative selection?
Every dictionary must to some extent draw upon the work of its predecessors, so we first collect all the citations used by LSJ. But we need to be able to check the other citations too (especially those in later authors like Plutarch and Polybius), and to take account of new discoveries, like the texts of Menander, and so we also collect all attestations from the authors in our corpus.
So the database must be very big, then?
The information stored in the archive uses 76 gigabytes, which would fill 53 CD-ROMs. In comparison, the one TLG CD-ROM contains most of Ancient Greek literature. There is necessarily a lot of redundancy in our database (because every sentence illustrating one of our headwords must appear again for each of the other headwords that it contains).
What does the database look like?
Here's the top of the first page of the database for the word theatron: [Click to open in a new window.]
What does this show?
The box at the top shows that this is page 1 out of 5, for this word. One page is for LSJ citations, and four for the others. All pages appear twice: with and without English translations.
The second box, titled 'Frequency Summary', gives the totals of citations by author, with LSJ citations separated from the others. There are 9 citations from LSJ (omitting some inscriptions and authors like Porphyrius who are not covered in our lexicon), and 156 others. There is a category for 'Ambiguous Citations', where the automated search tool has encountered inflectional forms that might belong to different words.
Finally, the citations are listed. The page is quite long, and so only the first two citations are given here, starting with Herodotus 6.67, the first citation in LSJ (we list these citations in LSJ order, for ease of reference).
At the start of the Greek and English passages, small red or blue lines may be seen. These are links to the complete text, so that we can read more, if we need.
Is there a more user-friendly way to view this?
The citations from LSJ can also be viewed as what we call a 'weave': that is, intercalated into the LSJ article itself, at the appropriate places. This can be seen, for the same word 'theatron', here: [Click to open in a new window.]
What are the advantages of this view?
It is not only much easier to view the citations in context, but it is also much more informative. Firstly, it gives us a check on accuracy: we can very easily see whether any citations are missing. Secondly, it gives us more semantic information: we can compare the LSJ translations with the whole passage, and so judge the accuracy of their explanations.
The future
This database has been in use since 2005, and has proved indispensible for consulting the texts as we wrote the lexicon entries. It would not have been possible to write the lexicon without it. It is much more than an archive of lexicographic 'slips': it is a semantically-organised digital library, in which the 'weave' pages constitute the first systematic display of LSJ 'slips'. The database may therefore be of interest to other classical researchers, and we hope to make a draft more widely available.
Next Page: Tagging the Lexicon