How is the CTS/CITE system applied to fragmentary texts?
We may distinguish between two types of fragmentary texts: logical and physical.
There are basically three options:
urn:lofts:fhg.1.hecataeus.hecataei_fragmenta.genealogiae.liber_secundus:350
.
(To be precise: fragment 350 of the second book of Hecataeus’ Γενεαλογίαι in FHG 1). Note that this is not, strictly speaking, a CITE URN but a LOFTS URN, but the CITE syntax is followed.
Note: in the past, there was an argument that lost works cannot have CTS URNs, because they are neither canonical nor citable. The fragments, at the other hand, are objects in a collections, hence can (and should) have CITE URNs. While LOFTS does not subscribe to this argument, it was nevertheless found conventient to mint CITE URNs for fragmentary texts.
Brill wanted to follow the LOFTS example. This proved impossible, however. The reason is that the Nautilus API, which is the cote of the Scaife Viewer, does not except CITE URNs. So the choice was: use CITE and don’t publish, or use CTS and start publishing JO, LGGA, SGG, etc. We choose the latter…
These are basically papyri and inscriptions. Here, scholia are included as well.
The CTS specifications suggest the Oxyrhynchus papyri constitute a textgroup. This would make a single papyrus a work. I could find no example of this. However, I could find two examples of other papyri identified by CTS URNs.
The Coptic Scriptorium project also uses CTS URNs. For example, the papyrus with print ID CPR 2 2237 (Corpus Papyrorum Raineri, cpr 2 = the second volume: II, Koptische Texte, ed. J. Krall. 1895. Nos. 1—255) an inventory number P.Vindob. K 2572 (in the Österreichische Nationalbibliothek) gets this CTS URN urn:cts:copticDoc:papyri_info.tm82127.cpr_2_237
. Note that this same text can be found in Trismegistos and in papyri.info but with local identifiers, not URNs. However, the TM number is included in the CTS URN, as is a mention of papyri.info, as is the print ID.
This is also how the Homer Multitext project does it. There, the so-called Bankes Homer, a long papyrus fragment (B.M. Papyrus cxiv), containing Homer, Iliad 24.127–24.804, was published with CTS URN urn:cts:greekLit:tlg0012.tlg001.bankes-01
. Note how the edition is hooked on to the familiar CTS textgroup and work IDs.
The difficulty with this approach is that a papyrus, or a collection of papyri, has more affinity with the physical structure than with the logical structure of a work. In many cases, the papyrus is the text, there is no other text, no variant to be compared. The difference between the papyrus as an object (e.g. with an inventory number) and as an edited and published text is important; more so than with preserved literary texts, for example, which tend to have a longer tradition history and hence more text variants. Hence, one could argue that the Oxyrhynchus papyri and such like form a collection. However, I could find no example of this.
It should be noted that only very few papyri have a CTS URNs. Papyri.info, for example, doesn’t use them. It seems that local identifiers are important in papyrology, with Trismegistos identifiers acring as a kind of glue, linking the local identfiers. (Here is an example basp.50.40 = HGV P.Tebt. 2 519 = Trismegistos 13594 = berkeley.apis.836 = p.tebt.2.519
).
The CTS specifications suggest inscriptions from a given site constitute a textgroup. This would make a single inscription a work. I could find no example of this.
The DTS people claim CTS is unsuitable for inscriptions at all…
One could argue that inscriptions, either in a site or in a text corpus, form a collection. But I could find no examples of CITE URNs for inscriptions.
I have to do some research… and this book: Digital Classical Philology: Ancient Greek and Latin in the Digital Revolution edited by Monica Berti.
For scholia, CTS URNS are used. Two examples below. I could find no examples of CITE URNs.
urn:cts:greekLit:tlg0012.tlg001.msA:1.4
.
See here for ore information. All HMT URNs are on GitHub.
urn:cts:greekLit:tlg5022.tlg005.1st1K-grc1
You can find the texts here. The textgroup is that of Scholia in Euclidem, the work is Scholia in Euclidis catoptrica. (The textgroups of Euclides’s works has ID tlg1799
, while the Catoptrica has tlg011
). The edition information (“1st1K”) speaks for itself. Lastly, check the nice XML you get from the site.
In conclusion: minting CTS URNs seems to be the way forward for scholia.