EpiDoc: News and Views

Last week I had the opportunity to give a seminar on EpiDoc as a guest of Prof. Carla Salvaterra and the Department of Ancient History at the University of Bologna. We titled the session "Digital Publishing with EpiDoc: Epigraphy, Papyrology, Interoperability."

I had a delightful time, and greatly appreciated extended discussion with the faculty and students who participated.

For what they're worth, I've posted my slides:

• Digital Publishing with EpiDoc (PowerPoint version)
• Digital Publishing with EpiDoc (OpenOffice version)
  

This is the third entry in our Virtual Seminar on some Unpublished Inscriptions from Corinth (see post I here and post II here). It will feature our first Latin inscription, which consists in four fragments of buff-colored micaceous marble. Fragment A was found beside the Lechaion Road in December, 1929 and has already been published, Fragment B was found 12 July, 1976 in Quarry Trench 9, Fragment C was found a day later 13 July, 1976 also in Quarry Trench 9, and Fragment D was found 6 August, 1974 in Quarry Trench 3. All the fragments are broken on all sides, except Fragment D, which seems to preserve part of the original right edge, although it is not at a right angle with the inscribed surface and therefore it may have been trimmed for reuse. The corner, however, is smoothly rounded here between the two adjacent faces and there are no partial letter traces at the edge. None of the four fragments join and it is not clear to us in what order they should be placed. Photos, squeezes, and autopsy of stones.

FASTI MAGISTRALES?

Date: 44 a. – 22/3 p.

Fragment A:

Published: Kent, ICor 8,3, 345.
Height, 0.0135 m. ; width, 0.127 m. ; thickness, 0.080 m.
Height of letters, 0.008 to 0.009 m. ; interspace, 0.004 to 0.006 m.
Corinth inventory, I-989 ; CECI II, I-989.

[— — — — — — — — — — — — — — — — — — —]
[— — —] • M(arcum) • Instle[ium Tectum — — —]      1
[— — — —] •̣ Corint[hu]m • C • Anṭ[— — — — —]
[— — — —]M • et • Q(uintum) • Cornelium [— —]
[— — — — —] p̣ṛobaruṇt • XX̣[— — — — — — —]
[— — — — — — — — — — — — — — — — — — —]

Apparatus:

Line 1: Kent read Instẹị[um Tectum], but the remnants of the last two preserved letters are clearly LE. For more on this reading, see the commentary below.
Line 2: There is some loss of the surface at the beginning of the line, but the traces of an interpunct may still be seen. Nothing of the H or V is visible. At the end of the line, Kent read Mịṇ[ucium], but this reading would require a space of 0.006 m. between the last stroke of the M and the I, which is three times greater than that found elsewhere on this stone and the new fragments. We believe that AN are clear, followed by the lower part of a hasta, the spacing and context of which suggest a T.
Line 4: The tops of the P and R are visible. The right hasta of the N is visible. After the first X, the trace of the upper left diagonal of another X is visible.

Fragment B:
Unpublished.
Height, 0.070 m. ; width, 0.110 m. ; thickness, 0.043 m.
Height of letters, 0.007 to 0.009 m. ; interspace, 0.004 to 0.007 m.
Corinth inventory I-76-14A ; NB 632, pp. 83 ; NB(FI) 655, pp. 20-21, Object 620A.

[— — — — — — — — — — — — —]
[— — — —] . . . ỊỊA • decuṛ[ion— —]      1
[— — —]s • apparitoruṃ [— — — —]
[— —]er LXII • M • C[— — — — —]
[— — — —]ṣp̣uṇ[— — — — — — —]
[— — — — — — — — — — — — —]

Apparatus:

There is a micaceous flaw in the surface running from the top line at the left of the A through the O in line two and the C in line three to the preserved end of line four that makes reading difficult.

Line 1: Traces of three letters followed by two hastae, the first of which leans slightly to the right. Then an A followd by an interpunct. The upper left corner with a piece of the rounded loop of the R is clear at the edge of the break.
Line 2: The upper left angle of the M is visible. [tibuniciu]s apparitoruṃ ?
Line 3: There is a generous space after the R, but no interpunct seems visible, rather a slight point of damage high in the line space.
Line 4: The first letter trace is most consistent with an S, but it could also be the top of a C or the rising tip of a T or F that is found elsewhere on these fragments. The second letter trace can be the top of a B, P or R. The final trace has a hasta and diagonal connected at the top left corner and to the right there is the tip of another hasta, most consistent with an N.

Fragment C:
Unpublished.
Height, 0.105 m. ; width, 0.115 m. ; thickness, 0.048 m.
Height of letters, 0.007 to 0.009 m. ; interspace, 0.006 to 0.007 m.
Corinth inventory, I-76-14B ; NB 632, p. 86 ; NB(FI) 655, pp. 20-21, Object 620B.

[— — — — — — — —]ỊḄṚỊ[— — —]      1
[— — — — — — — —]nus • IIỊ[— —]
[— — — — — — —]C• Fideḷ[— — —]
[— — — — —]Ị• Caesaris [— — — —]
[— — — — A]ntiochus •I• [— — —]      5
[— — — —]canus • I[I — — — — —]
[— — — — —]ṾỊ • [— — — — — —]

Apparatus:

Line 1: The bottom of a hasta, followed by a letter with a base that resembles a B or D, followed by two letter traces that conform well with the bottom part of an R or a crowded IC, followed by the faint trace of a hasta.
Line 2: At the end of the line, only a faint trace of the third hasta is visible. A fourth may have followed the break. The numeral is overlined.
Line 3: It is not clear if the C is the last letter of an abbreviation, such as PROC(urator) or C for C(uravit)/C(uraverunt) or a name such as C(aius). It is not clear to us whether Fideḷ[—] is part of a proper name or an adverb or adjective.
Line 4: At the beginning of the line the lower half of a hasta survives.
Line 6: The overline of the numeral seems to be preserved to its full length and so the restoration of another I seems assured.
Line 7: The upper left tip of a diagonal and to the right of it the tip of a hasta are visible and are consistent with a V. Then there follows the upper tip of another hasta slightly lower in the line followed by an interpunct. We seem to have the end of a large number without an overline (cf. the large number in line 3 of Fragment B, which is also not overlined).

Fragment D:

Unpublished.
Height, 0.080 m. ; width, 0.070 m. ; thickness, 0.035 m.
Height of letters, 0.006 to 0.008 m. ; interspace, 0.005 m. to 0.016 m.
Corinth inventory, I-74-11 ; NB 610, p. 82 ; NB(FI) 611, p. 57, Object 464

[— — — — — — — — — — — — — —]
[— — — —]aedes[. .]E[— — — — — —]         1
[— — — — Ma]ecius • A(uli) • f(ilius) • Co-
[rnelius — —(?)] vacat
[— — — — — — —]t •A• decu-
[— — — — — — — —]ṃ. vac. 0.016 m.        5

Apparatus:

Line 1: Only part of the lower horizontal of the last E is visible.
Line 2: Or [D]ecius. Maecius is more common at Korinth.
Line 3: It is unclear whether the beginning of this line was inscribed or the text at the end of our line 2 continued at the beginning of line 4. It seems more likely that it finished here and a new entry was begun at the beginning of line 4.
Line 5: The two apices of a letter characteristic of an M elsewhere seem clear. Since there is a vacat of 0.016 m. to the right of this, it appears that we have the end of the line.

Commentary:

H.S. Robinson originally thought that these fragments might belong to the Lex Coloniae Corinthiensis (he noted that the words CAESARIS, DECVR[ION—], and APPARITORVM all appeared in the Lex Coloniae Genetivae Ursonensis and deduced a parallel). However, Mary Hoskins-Walbank while working on her dissertation (non vidimus) took a look at the stones and in some correspondence with Robinson expressed the view that they were more consistent with a fasti document – a view we find more likely. She also thought this stone might have been damaged in the earthquake of AD 22/3 and then discarded rather than suffering a damnatio memoriae.

One of the more intriguing aspects of this inscription lies in the reference to a Marcus Instleius in Fragment A, line 1, heretofore read as Insteius, who was one of the earliest duoviri of the colony of Korinth (established 44 BC). The spelling Insteius has been preferred by previous studies rather than Instleius undoubtedly because Instleius is not attested elsewhere (we do, however, find an A(ulus) Instuleius Tenax attested at Egyptian Thebes = Colosse de Memnon 2), while the name Insteius is attested in both Greek and Latin epigraphical and literary sources. The reading Insteius, however, is epigraphically impossible on this stone. Furthermore, the reading Instleius is corroborated on another stone from Korinth (ICor 8,3 149, line 1, photo here). Kent read the first line of this stone as [M •] INSTỊ[E]Ọ • C • F • TECTO, but he went on to add that “the letter following T can only have been I or L, and clearly was not an E. As there is no join between fragments a and b, as there would have been if the E had simply been omitted, I have assumed that the letters EI were erroneously transposed.” However, a transposition on this carefully carved piece of revetment seems scarcely believable. In addition, an L is more likely than an I given that the hastae of the other two instances of I on this same stone are taller while the close proximity of this letter’s hasta to the T that precedes it makes more sense space-wise if it is the hasta of an L (compare how the E in TECTO tucks in under the first T). In line 1 of ICor 8,3 149 we therefore propose reading INSTḶ[ΕΙ]Ο̣.

The reading Instleius rather than Insteius is further corroborated by a series of coins struck at Korinth (see Amandry BCH Suppl. XV pp. 124-128) that, given the rarity of the name Instleius, undoubtedly refer to our same man. On the reverse of several of these coins we find IIVIR paired with INSTL • CAS (example here), and on the reverse of others we find II VIR paired with INTS CAS (example here – where the order INTS is probably a ligature for INST). Previous scholars have interpreted the two men’s names as Inst(eius) and L. Cas. (for the last Amandry suggested L. Cas(tricius Regulus), while Kent suggested L. Cas(ius […]), but the placement of the interpunct between the L and C argues against such a reading and when we add the evidence of the coins to the inscriptions we once again are lead to believe his name was spelled Instleius. Perhaps the form Instleius may have been an older, alternative form of Insteius much like stlis is an earlier form of lis (as in Decemviri Stlitibus Iudicandis). Amandry (p. 36) places Instleius’ office of duovir in 42 or 41 and his office of duovir quinquennalis in 35.

If we assume Instleius is an alternate form of Insteius, this Marcus Instleius, as others have already pointed out, may have been the same man who fought at the side of Antony at the siege of Mutina in 44 BC (Cicero, Philippic 13.26) as well as at Actium in 31 BC (Plut., Antony 65.1).

[…] This post represents the first installment at Current Epigraphy of what will be a summer-long “Virtual Seminar on some Unpublished Inscriptions from Corinth.” For the next few months about every two weeks I will upload Don Laing’s and my preliminary text of a Greek or Latin inscription from Corinth and invite suggestions for restorations or comments on the context, date, etc. Tom Elliott and Gabriel Bodard will then work up an EpiDoc version of the resulting texts. As Tom Elliott explained here, the purpose of this first-ever virtual epigraphical seminar is to promote a new model of collaboration and publication of epigraphical texts with the following benefits: a preliminary text will be made available more quickly; scholars or those interested will be able to “attend” the seminar at their leisure from anywhere in the world with an internet connection; students will see how epigraphers work and it may raise more interest in the discipline; there should be more interest in the final print version that will appear in Hesperia, where proper attribution to those who proposed any particular idea or reading will be given and comments on this experiment will be included; the final print publication will be stronger (these inscriptions from Corinth, like most inscriptions from there, are very fragmentary and they lend themselves to collaborative treatment); the project will introduce more epigraphers to the advantages of EpiDoc. Special thanks are due Guy Sanders (Director of the ASCSA dig at Corinth) and Charles Watkinson (Chair, ASCSA Publications Committee) for their support of this project. […]

The Computers and the Humanities Users Group, Computing and Information Services, and The Center of Digital Epigraphy

present

Interoperability between Epigraphic and Papyrological Databases:
The Epidoc Scenario

Dr. Gabriel Bodard
Centre for Computing in the Humanities
King’s College London

12:30, Thursday, May 22
169 Angell St., Main Conference Room,

Crosswalking–the automated mapping of metadata from one schema to another–has emerged as a crucial tool in the digital landscape, and is particularly useful for integrating data from multiple sources or projects. This talk will focus on the use of crosswalks in epigraphical and papyrological research development. Within these domains, a number of corpora have been developed using different technologies and data structures, and driven by different user needs. There are collections that use the Epidoc XML schema, which is based on TEI, collections like the Electronic Archive of Greek and Latin Epigraphy that are served from SQL databases, and older projects which use specialized information structures. Dr. Gabriel Bodard will present some of the strategies that he and his colleagues, Tom Elliott and Hugh Cayless, have devised to perform such transforms. He will then describe in more detail the Integrating Digital Papyrology project, whose purpose is to dynamically transform and integrate the Duke Databank and Heidelberg Gesamtverzeichnis collections into a single EpiDoc collection, and some of the technical and theoretical lessons learned from this process.

I have no doubt that Chris must be right in that the ratio of abbreviated to whole words is much higher in Latin texts on Roman coinage than on Roman inscriptions. Still, abbreviations are rather common in the inscriptions, and all intros to Latin epigraphy include treatment of the phenomenon on the (valid) assumption that competent use of epigraphic texts as evidence requires facility in dealing with the abbreviations. I hope someday to do a comprehensive statistical study of abbreviations in Latin inscriptions, but that has to wait until nearly all published texts are available in (or readily converted to) EpiDoc ;) … meanwhile, this old thing may give some indication to interested parties of the range and variation of abbreviations in a sample of Latin texts from the Roman period:

http://www.case.edu/artsci/clsc/asgle/abbrev/latin/

Now, whether the cognoscenti admit emoticons (like the one I used above) as leet-speak or not, I wonder if we’d consider that scrolls, crosses, chi-rhos and such are analogous to emoticons …

The following titles from the Digital Classicist Work-in-Progress seminars are of potential interest to epigraphers:

6 June (NG16): Elaine Matthews and Sebastian Rahtz (Oxford), The Lexicon of Greek Personal Names and classical web services

13 June (NG16) Brent Seales (University of Kentucky), EDUCE: Non-invasive scanning for classical materials

20 June (STB3) Dot Porter (University of Kentucky), The Son of Suda On Line: a next generation collaborative editing tool

18 July (STB3) Ryan Bauman (University of Kentucky), Towards the Digital Squeeze: 3-D imaging of inscriptions and curse tablets

25 July (NG16) Charlotte Tupman (KCL), Markup of the epigraphy and archaeology of Roman Libya

8 Aug (NG16) Charlotte Roueché (KCL), From Stone to Byte

Good idea. See now: EpiDoc News and Views (html | rss)

Hey Tom,

Note that there's also an EpiDoc category on the CEp blog, and we could easily (and probably should) also have one at the Stoa. Worth aggregating those in the same way as Concordia and Pleiades (since EpiDoc doesn't have it's own blog either)?

The Centre for Computing in the Humanties, Kings College London, is again offering an EpiDoc Summer School, on July 14th-18th, 2008. The training is designed for epigraphers or papyrologists (or related text editors such as numismatists, sigillographers, etc.) who would like to learn the skills and tools required to mark up ancient documents for publication (online or on paper), and interchange with international academic standards.You can learn more about EpiDoc from the EpiDoc home page and the Introduction for Epigraphers; you wil find a recent and user-friendly article on the subject in the Digital Medievalist. (If you want to go further, you can learn about XML and about the principles of the TEI: Text Encoding Initiative.) The Summer School will not expect any technical expertise, and training in basic XML will be provided.

Attendees (who should be familiar with Greek/Latin and the Leiden Conventions) will need to bring a laptop on which has been installed the Oxygen XML editor (available at a reduced academic price, or for a free 30-day demo).

The EpiDoc Summer School is free to participants; we can try to help you find cheap (student) accommodation in London. If any students participating would like to stay on afterwards and acquire some hands-on experience marking up some texts for the Inscriptions of Roman Cyrenaica project, they would be most welcome!

All interested please contact both charlotte.roueche@kcl.ac.uk and gabriel.bodard@kcl.ac.uk as soon as possible. Please pass on this message to anyone who you think might benefit.

The Centre for Computing in the Humanties, Kings College London, is again offering an EpiDoc Summer School, on July 14th-18th, 2008. The training is designed for epigraphers or papyrologists (or related text editors such as numismatists, sigillographers, etc.) who would like to learn the skills and tools required to mark up ancient documents for publication (online or on paper), and interchange with international academic standards.

You can learn more about EpiDoc from the EpiDoc home page and the Introduction for Epigraphers; you wil find a recent and user-friendly article on the subject in the Digital Medievalist. (If you want to go further, you can learn about XML and about the principles of the TEI: Text Encoding Initiative.) The Summer School will not expect any technical expertise, and training in basic XML will be provided.
Attendees (who should be familiar with Greek/Latin and the Leiden Conventions) will need to bring a laptop on which has been installed the Oxygen XML editor (available at a reduced academic price, or for a free 30-day demo).

The EpiDoc Summer School is free to participants; we can try to help you find cheap (student) accommodation in London. If any students participating would like to stay on afterwards and acquire some hands-on experience marking up some texts for the Inscriptions of Roman Cyrenaica project, they would be most welcome!

All interested please contact both charlotte.roueche@kcl.ac.uk and gabriel.bodard@kcl.ac.uk as soon as possible. Please pass on this message to anyone who you think might benefit.

Updated information about EpiDoc resources, including our move from CVS to SVN support, has been added. See http://epidoc.sourceforge.net/resources.shtml . (0 comments)

There's no point in reiterating here what Dot, Brent, Chris and Cathy have so eloquently written about Ross. Even though I'd had the news of his death privately over the weekend, the deep emptiness of his being gone didn't really hit me until I saw the first public notice on Classics-l. There's something brutally liminal about a death notice in a professional forum, no matter how gently written: it is the crisp, formal ceremony that transfers a person from the active present to the static past of the discipline.

This sombre realization is rippling through the web of connections that was Ross' personal and professional network. You can detect it in the spattering of blog posts, emails and the subdued communications of his many colleagues and friends.

And yet, it is clear that the interpersonal fabric Ross wove will be a lasting, living contribution to the field, and to our lives. There are so many people Ross introduced to each other and encouraged in collaborative digital classics work. He watched our backs when things got rough, applauded our successes, pulled us out of ditches, and kicked our asses well and thoroughly when we deserved it. Vast indeed is the sea of those whom Ross has mentored and enabled.

I've written elsewhere about Ross's contribution to the EpiDoc effort. Pleiades owes him an equal debt. It was Ross's 2001 invitation to speak at the Center for Computational Sciences in Lexington that first forced me to formalize the ideas that I'd been batting around privately with Richard Talbert, Stephen MacGregor, Hugh Cayless, Noel Fiser, Amy Hawkins and others in Chapel Hill. And it gave those ideas their first public airing. Ross and I had originally discussed them, along with Sebastian Heath and Neel Smith, in Newport the previous year. Ross helped us refine the plan through subsequent iterations and grant proposals and, when it emerged that UNC could not provide us with the class of hosting we needed for development, he offered server space belonging to the Stoa. The collaborative editorial approach embodied in the Suda Online underlies our model for the Pleiades workflow, to be rolled out later this year. Ross remained deeply engaged in both the vision and the technical details of Pleiades, even during his illness. Without him, Pleiades would not be.

And so I have now both sadly and joyfully yielded -- like Patrick, Melissa, Troels, Hugh and others -- to the compulsion to hold up for you to see one more swathe of the Rossian fabric, saying "Look! Here's another bit he did with us. Doesn't it shine, gold and purple in the sun?"

Inscriptions of Roman Cyrenaica (http://ircyr.kcl.ac.uk). First International Workshop. British School at Rome, 28-29 February 2008.

Charlotte Tupman and Gabriel Bodard: Epigraphic Interoperability

On the occasion of the Inscriptions of Roman Cyrenaica Workshop on geospacial data and interoperability, held in Rome on February 28-29, scholars, mainly archaeologists, involved in digs and studies in Libya presented their work with a particular focus on digital data capture and publication.

At 14:00 on Friday 29th, Charlotte Tupman and Gabriel Bodard gave an interesting joint paper on Epigraphic Interoperability. (Slideshow available to view.)

As an introduction, EpiDoc and its principles were briefly explained. The EpiDoc schema and guidelines offer guidance for the encoding of epigraphic texts and metadata in an XML system that abstracts structure and semantics on the one hand from the specifics of display on the other, so that the same underlying data can be used to generate various presentations (from traditional Leiden edition, diplomatic text, web page, printed page, dynamic indexes, or database-like tables).

However, the main focus of the paper was to demonstrate the possibility of collaboration between EpiDoc and the EAGLE databases through a sort of “crosswalk” of data from one schema to another. The EpiDoc guidance defines a level of compliance with the EAGLE database which means that all metadata required by the relevant databases is included and explicitly tagged in a compliant EpiDoc XML edition. Finally a simple tool was demonstrated that created tabular output compatible with the Epigraphic Database Roma from the IRCyr XML files).

At the end of the presentation Professor Silvio Panciera, chair of the AIEGL committee on IT and Epigraphy and director of the EAGLE federation of databases, expressed his support to the project and stressed the importance of digital applications to the study of epigraphy and the Classical world in general. He also expressed his gratitude for any sort of collaboration with the EAGLE endeavour and encouraged the audience to embrace the new opportunities offered by digitalization.

The influenza kept me off the plane to Rome, but happily I was at least able to give my talk (via Skype) this morning. The occasion is a meeting at the British School in Rome, organized by the Inscriptions of Roman Cyrenaica project, to bring together scholars working in Cyrenaica to explore the potential for cross-project collaboration and data sharing. I used our work so far on Pleiades (and a bunch of Sean's ideas exchanged on IRC) as a spring-board for a methodological proposal: using Atom+GeoRSS feeds to facilitate cross-project data discovery and citation.

There will be more about this in future posts, but for now, the slides (mostly screen shots) are available:

• Tom Elliott: Atom+GeoRSS for Interoperability (ODP zipped)


• Tom Elliott: Atom+GeoRSS for Interoperability (PPT zipped)

This is the first in a highly irregular series exploring aspects of my attempt to turn my dissertation (or parts thereof) into a digital publication using the EpiDoc customization of the Text Encoding Initiative tagset (in XML). I'm starting using a batch of boundary inscriptions from Roman Cyrenaica, partly because I'm working with a team in London and Cambridge that is working on the definitive publication of the Roman Inscriptions of Cyrenaica (IRCyr) as collected and analyzed by Joyce Reynolds. This collection will include a number of previously unpublished boundary inscriptions.

Today's topic: epigraphic bibliography

There are various examples of code below, but you can also download a fully encoded example.

A proper epigraphic edition includes a complete history of previous published editions, published derivative texts, corrections and, often, commentary on same. There are various common mechanisms for presenting these citations in print, usually in a compact form that makes liberal use of abbreviations and short titles. Thus, text 62.2 in my dissertation (of Claudian date), presented the following bibliography:

EDH HD011697 (Latin); SEG 26.1819; AE 1974.682; *Reynolds 1971, 47-49.1.

Here, order signifies date and the asterisk indicates the edition I follow in my own catalog. So, we can read this as:

Originally published in Reynolds 1971, 47-49.1, whence derivative editions in AE 1974.682, SEG 26.1819 and EDH HD011697 (the latter only providing the Latin portion of this bilingual Greek/Latin text).

Often, such bibliographies include other notation to indicate the "genetic lemma" (derivative relationships) between publications. So, one could have produced something like:

[EDH HD011697 (Latin)] = [SEG 26.1819] = [AE 1974.682] = Reynolds 1971, 47-49.1

where the square brackets indicate derivative editions, i.e., those that derive from another published edition rather than autopsy of the stone and/or reference to a squeeze, rubbing or photograph. This particular lemma is a little misleading, since the provisional EDH edition actually derives from the edition in AE, which is itself derivative of the Reynolds edition.

How to do this in EpiDoc?

Let's start with something like the more prose-ish of the above examples, since this is the approach IRCyr is using (demonstrated in ALA2004 and IAph2007). First, EpiDoc calls for the bibliography to be wrapped in an appropriately typed <div> element, as follows:

<div type="bibliography">
 <p>Original publication was Reynolds 1971, 47-49.1,
    whence derivative editions in AE 1974.682, SEG 26.1819
    and Elliott 2004, 167.62.2. The Latin portion of the
    text is reproduced in EDH HD011697 (1997, provisional)
    on the basis of AE.</p>
</div>

Part of the reason to do our bibliography in XML is to be able to encode relationships, assertions and semantic distinctions in a way that is machine actionable. On the bibliographic front, we might want to be able to search, sort and index by these other editions, or link to them if digitally available. That means we need to mark each citation as a discrete bibliographic citation, and TEI provides the <bibl> element for this purpose:

<div type="bibliography">
 <p>Original publication was
   <bibl>Reynolds 1971, 47-49.1</bibl>,
   whence derivative editions in
   <bibl>AE 1974.682</bibl>,
   <bibl>SEG 26.1819</bibl>
   and
   <bibl>Elliott 2004, 167.62.2</bibl>.
   The Latin portion of the text is reproduced in
   <bibl>EDH HD011697 (1997, provisional)</bibl>
   on the basis of AE.</p>
</div>

We may want to have a search function distinguish between original editions and those that are derivative, so we need to encode that distinction too. We don't want to have to parse text strings and try to infer the meaning of phrases like "original publication" or "derviative". Rather, we'll use the standard TEI "type" and "subtype" attributes on the <bibl> element to make this distinction clear for our little silicon friends. The values we're using for this attribute are specific to the EpiDoc customization of TEI.

<div type="bibliography">
 <p>Original publication was
   <bibl type="edition" subtype="primary">
     Reynolds 1971, 47-49.1
   </bibl>,
   whence derivative editions in
   <bibl type="edition" subtype="derivative">
     AE 1974.682
   </bibl>,
   <bibl type="edition" subtype="derivative">
     SEG 26.1819
   </bibl>
   and
   <bibl type="edition" subtype="derivative">
     Elliott 2004, 167.62.2
   </bibl>.
   The Latin portion of the text is reproduced in
   <bibl type="edition" subtype="derivative">
     EDH HD011697 (1997, provisional)
   </bibl>
   on the basis of AE.</p>
</div>

There's additional tagging internal to each <bibl> element that we can/should do to facilitate sorting, searching and linking to digital/digitized works, but we'll skip over that here (check out the example file for the full encoding).

The only thing our example doesn't do that we might like is encode the derivative relationships between the various editions. We know that one is "primary" and the others "derivative", but it's not clear what the path of derivation is for each one. EpiDoc doesn't currently have guidance for this, and I'm not sure what the broader TEI community thinks (I'm posting a link to this entry on TEI-L to find out), but it occurs to me that this would be pretty easy to do with the TEI <link> element. We'll need unique identifiers on each <bibl> element to make use of this approach.

<div type="bibliography">
 <p>Original publication was
   <bibl xml:id="reynolds-1971-1" type="edition" subtype="primary">
     Reynolds 1971, 47-49.1
   </bibl>,
   whence derivative editions in
   <bibl xml:id="ae-1974-682" type="edition" subtype="derivative">
     AE 1974.682
   </bibl>,
   <bibl xml:id="seg-26-1819" type="edition" subtype="derivative">
     SEG 26.1819
   </bibl>
   and
   <bibl xml:id="elliott-2004-62-1" type="edition" subtype="derivative">
     Elliott 2004, 167.62.2
   </bibl>.
   The Latin portion of the text is reproduced in
   <bibl xml:id="edh-hd011697" type="edition" subtype="derivative">
     EDH HD011697 (1997, provisional)
   </bibl>
   on the basis of AE.</p>
   <link targets="#reynolds-1971-1 #ae-1974-682 #seg-26-1819 #elliott-2004-62-1"/>
   <link targets="#ae-1974-682 #edh-hd011697"/>
</div>

Stay tuned for further adventures, in which we exploit some of this bibliographic tagging, and then move on to encoding the epigraphic text itself.

That substantial seismic WHOMP! you felt last Friday was the 1.0 release of the Text Encoding Initiative, version P5 making its official entry into the light of day. A hearty congratulations to the editors, the TEI Technical Council, and everyone else who worked so hard to make this major revision a reality.

As a sizable percentage of my legions of readers will know, the TEI underpins the work of the EpiDoc Community, which aims to provide guidance and tools for the XML markup of ancient primary sources -- especially documentary ones -- preserved in inscriptions, on papyri and the like. Right now, EpiDoc depends on the previous (P4) version of the TEI, but incorporates a number of P5 structures that are especially useful (or economical) for our needs. We'll hold at this point until sometime at least in mid-2008, when we'll look at revising EpiDoc to full P5 compliance. This delay recognizes that key members of the community will be pretty busy in the meantime on a number of projects that shouldn't be slowed down for a major revision.

A particularly important current project in this regard is the conversion of the Duke Databank of Documentary Papyri (description somewhat out of date) to full EpiDoc conformance. This conversion underpins an effort to establish better interoperability with the Advanced Papyrological Information System and the Heidelberger Gesamtverzeichnis der griechischen Papyrusurkunden Ägyptens. This APIS/Duke/HGV work is funded by the Andrew W. Mellon Foundation, and it's driving major improvements to the EpiDoc Guidelines and software tools.

There's also an interesting, and rapidly growing, list of other EpiDoc projects.

Yesterday afternoon’s (Friday 17th August) seminar in the Digital Classicist series was presented by Charles Crowther of the Centre for the Study of Ancient Documents in Oxford, on the subject of ‘A Virtual Research Environment for the Study of Documents and Manuscripts’. The project on which Dr. Crowther was reporting (mentioned here back in May) is the successor to the BVREH (Building a VRE for the Humanities) and VWSAD (Virtual Workspace for the Study of Ancient Documents) projects, which ran during the last couple of years.

The new VRE, which recently won two years’ funding and is at an early stage of development, will be targeted explicitly at epigraphic and papyrological texts (although it could easily be repurposed for medieval or other manuscripts). The aim is to create a working environment to replicate for scholars at a distance the opportunities for collaborative research offered by bringing several experts into a room at the same time to look at high-quality images of a papyrus or inscription. In addition they intend to take advantage of the many online research tools that are available (e.g. PHI Greek Inscriptions, especially the new concordance tool; the Epigraphik Datenbank Clauss/Slaby; the more sophisticated Epigraphische Datenbank Heidelberg; the Photographic Archive of Papyri in the Cairo Museum; the Duke Databank of Documentary Papyri, etc.), as well as more “deep” resources–scholarly publications such as Aphrodisias in Late Antiquity or Vindolanda Tablets Online.

It would also be valuable to integrate this system with the excavators who might be digging up new inscriptions and need an expert to look at them. The Silchester excavations, who use the VERA (VRE for Archaeology) environment and keep detailed electronic notes, would be an ideal candidate for such collaboration, except that they very rarely unearth inscribed objects. Vindolanda do not have such sophisticated electronic recording methods, or else their many and difficult texts would be ideal. A first pilot collaboration might be with the excavations at Zeugma, where the excavators are aware of the importance of calling upon specialists in particular fields (such as epigraphy) to work with the excavators on their finds.

This is a project that is worth watching. The VRE will be run on the Oxford servers in the first instance, and other collaborators will be very welcome to help test the environment as soon as there is something running. Ultimately the tools created by this project will be Open Source, and so can be used by anyone with an interest in working collaboratively on documents, inscriptions, papyri, or manuscripts in an environment that integrates multiple freely available resources to enrich the editor’s research experience.

Yesterday (Friday 3rd Aug) Melissa Terras, lecturer in Electronic Communication in the School of Library, Archive and Information Studies (UCL) gave a seminar in the Digital Classicist series at the Institute of Classical Studies entitled ‘Can Computers Ever Read Ancient Texts?’ In this excellent presentation, Dr. Terras’ answer to the question posed in the title was (pace a response on the Humanist list last week) basically ‘no’, or at least not yet and not in any significant sense as we understand “read”. A better question, she asserted, would be “Can computers assist in reading ancient texts?”, and to that question she gave (and demonstrated) a much more positive response.

In the course of this paper, which was in part a demonstration of past and ongoing work on the semi-papyrological/semi-epigraphical Vindolanda writing tablets, she made several important points that are worth repeating to an audience of epigraphists.

1. The expert who teaches the computer to read texts is not principally the programmer, analyst, or engineer, but principally and essentially the papyrologist (or epigrapher) who knows the art and works in a discipline that has decades (centuries) of methodologies and expertise behind it.
2. Based on Herbert Youtie’s now famous distinction between “public” (published, polished, perfect) and “private” (in progress, iterative, much-laboured) text scholarship, she points out the ability of computer-assisted readings to reveal and preserve the private processes of text editing and track methodologies and mistakes, a process known by engineers as “truth maintenance”. For example, it would be incredibly valuable to be able to document the process of making decisions as to dating and/or scribe based on letter-forms (see point 4 below), a process that seasoned papyrologists and epigraphers currently often perform based on instinct and therefore find very difficult to describe and teach.
3. The computer does not, in the prefered model, make firm decisions and statements of “fact” based on engineering principles and programmed algorithms; rather it offers a range of probabilistic results (number-crunching is what machines are good at), offering likely character-groups or words, for the text editor or historian to choose between based on her expertise (interpretation is what humans are good at).
4. One of the most impressive by-products of the Vindolanda text-analysis project has been the development of palaeographic markup, a system designed to record and predict average letter-forms as well as variants.
5. Perhaps the most striking point was that this project, although firmly in the humanities and concerned with Ancient History, provides a rare “real-world” application of theories and techniques in Artificial Intelligence. New use-cases lead to new solutions, and a project like this benefits the discipline of Engineering Science as a whole. We Classicists should not underestimate what we have to offer to the high-powered world of medical imaging, defense technologies, and forensic science, for example.

Joyce Reynolds and colleagues at King’s College London are beginning a three-year project to publish the Inscriptions of Roman Cyrenaica, along with supporting materials, images, and a detailed geographical gazeteer.

From the IRCyr project website (http://ircyr.kcl.ac.uk/):

In 1948 Joyce Reynolds, of Newnham College Cambridge, then based at the British School at Rome, started a series of regular visits to Libya, to study the inscriptions of the Roman period. She worked initially in Tripolitania: her epigraphic corpus, the Inscriptions of Roman Tripolitania, written with John Ward-Perkins, was published in 1952, and remains the authoritative work. In 1951 she first visited Cyrenaica, and has worked since then on the inscriptions of Roman Cyrenaica, working with local archaeologists and others from Italy, France, the U.K. and the U.S.A.. She has continued to visit Libya almost every year since then, and has assembled materials for a corpus of some 2500 inscriptions from Roman Cyrenaica. Nearly a third of these have never previously been published, while others have only appeared in versions which can be very much improved, and better understood, as a result of re-reading. The collection is made up of transcriptions, and illustrations for about half the texts; the bulk of the remainder are illustrated in photographs held only in the archives of the Libyan Department of Antiquities at Cyrene. The Inscriptions of Roman Cyrenaica project has been funded by the Leverhulme Trust, to publish this material as an online corpus; the team will be drawing on experience gained in publishing the online corpus of the Inscriptions of Aphrodisias. Images held in Cambridge, Rome and Cyrene will be scanned to illustrate the collection. The new corpus will be presented as a series of documents; but it will also link to an online map of Roman Cyrenaica, being prepared as part of the Pleiades project, at the University of North Carolina, Chapel Hill. Material collected by Reynolds and her colleagues will be used to map ancient sites on that map, with a fullness which was not possible within the necessary limitations of its predecessor, the Barrington Atlas of the Greek and Roman World.

Over the last few years an international group of scholars has been developing a set of conventions for marking up ancient documents in XML for publication and interchange. The EpiDoc Guidelines started from the case of inscriptions, but the principles are also being applied to papyri and coins, and the aim has always been to produce standards consistent with those of the Text Encoding Initiative, used for all literary and linguistic texts.

Following on from the interest we have seen in EpiDoc training events (including recent sessions in Rome and San Diego) and the success of the London EpiDoc summer school over several years now, we shall be holding another week-long workshop here at King’s College London, from the 11th-15th June this year.

• The EpiDoc Guidelines provide a schema and associated tools and recommendations for the use of XML to publish epigraphic and papyrological texts in interchangeable format. For a fuller description of the project and links to tools and guidelines see http://epidoc.sf.net.
• The Summer School will offer an in-depth introduction to the use of XML and related technologies for publication and interchange of epigraphic and papyrological editions.
• The event will be hosted by the Centre for Computing in the Humanities, King’s College London, which will provide the venue and tuition. The school is free of charge, but attendees will need to fund their own travel, accommodation, and subsistence. (There may be cheap accommodation available through KCL; please inquire.)
• The summer school is targeted at epigraphic and papyrological scholars (including professors, post-docs, and advanced graduate students) with an interest and willingness to learn some of the hands-on technical aspects necessary to run a digital project (even if they would not be marking-up texts by hand very much themselves). Knowledge of Greek/Latin, the Leiden Conventions and the distinctions expressed by them, and the kinds of data and metadata that need to be recorded by philologists and ancient historians, will be an advantage. Please enquire if you’re unsure. No particular technical expertise is required.
• Attendees will require the use of a relatively recent laptop computer (Win XP+ or Mac OSX 10.3+), with up-to-date Java installation, and should acquire a copy of the oXygen XML editor (educational discount and one-month free trial available); they should also have the means to enter Unicode Greek from the keyboard. Full technical specifications and advice are available on request. (CCH may be able to arrange the loan of a prepared laptop for the week; please inquire asap.)

Places on the workshop will be limited so if you are interested in attending the summer school, or have a colleague or student who might be interested, please contact gabriel.bodard@kcl.ac.uk as soon as possible with a brief statement of qualifications and interest.