Exploring a digital collection of neo-Latin literature

Neven Jovanović, University of Zagreb

At the moment, more neo-Latin texts are freely accessible than ever before; it is enough to remind oneself of the impressive Analytic Bibliography of On-Line Neo-Latin Texts, compiled by Dana F. Sutton; the bibliography, as of October 24, 2012, records 43,340 neo-Latin works on the internet. The time has come for neo-Latin scholars to face the question: what should we do with all these texts?

The problem is not only that no human being can read all that, nor that it is quite hard to discover among these thousands of titles ones that we should read, nor that it is equally hard (once we've decided what we want) to discover a reliable “manifestation” of the text we want. The real problem is that, if we study these texts as we usually do – that is, focusing upon very carefully read key passages – we are using the internet simply as a vast library, and digital manifestations of texts as books. The main hypothesis of digital humanities, however, is that digital medium can change the way we understand and interpret languages, literature, history, philosophy, religion, arts.

We will try to demonstrate such qualitatively different digital approach to a collection of neo-Latin texts. The approach will combine (very simple) computational experimentation on mass clusters of textual data with philological reading of passages brought to us from this experimentation by the computer, an assistant that's better than us at some tasks but can't perform others.

In this way we will explore texts by Janus Pannonius currently included in the digital collection Croatiae auctores Latini (CroALa, http://www.ffzg.unizg.hr/klafil/croala). The collection, which currently contains 4.7 million words in 373 documents written by authors of Croatian origin (or connected with Croatia) from 976 to 1984, aims to make Croatian neo-Latin not only accessible, but also available for experiments of all kinds.

What happens when we search in CroALa for a large number of words and phrases, when these words and phrases are organised in meaningful sets. E. g.:

• indices of names and places from existing editions of texts in CroALa
  • A table of names from the Index nominum to Latin poems by Marko Marulić (Marcus Marulus), with links to searches in Pannonius, as well as in all of CroALa: X.
• lists of words from one text in CroALa, processed elsewhere (“interesting words”)
  • A list of Pannonius' Latin words unrecognized by a Latin morphology service (containing 957 records) — cf. search for ”acca” in Perseus Morphology Service: X (a negative result)
  • An excerpt from this list – just the words with <100 occurrences in CroALa (663 records)
  • A Google Fusion Table of interesting words in Janus Pannonius' Elegiae et epigrammata, linked to searches in all CroALa texts (enables sorting and filtering by number of occurrences)
• lists of words from one text in CroALa as searches across other texts
• lists of Latin words found outside CroALa (i. e. found as lists)
  • Headings of Ravisius Textor's (Jean Tixier's, c 1480–1524, PND 100792405) Epitheta, presented as starting points for parallel searches in Pannonius and in CroALa (letter A): X
  • Poetic phrases from Textor's Epitheta found in CroALa (first 1000 results, also with links to Epitheta themselves): X
  • The procedure could be improved: we could easily limit searches to a subset of CroALa (just poetry, just texts from the Renaissance)
• lists of words and phrases from Latin texts outside CroALa

• Exploring CroALa through Janus Pannonius, a lab diary