The Latin Diachronic Database Project

The Latin Diachronic Database is a project of Digital Humanities invented by Tommaso Spinelli (Ph.D. candidate, Classics, St. Andrews University) and co-developed with Giacomo Fenzi (Computer Science and Mathematics student, St. Andrews University). This project aims to create an innovative toolkit for the quantitative computational analysis of the Latin language as well as to support and further enhance the digital study of ancient intertextuality.

The program currently offers the possibility to run two different searches, both unique in their genre. Firstly, the program represents the first diachronic frequency dictionary of the Latin language based on big data (c. 10.500.000 forms). In this respect, it is able to scan and record all the lemmas attested in the whole corpus of extant Latin Literature by attributing different inflected word-forms (e.g. genitive sing., dative pl. etc for names; pres. 1 pl., ppf 2 sing. etc for verbs) to the correct headword(s) using the technology LEMLAT. The headwords are respectively presented in the nominative case for names and in the first singular person of the present for verbs. This dictionary is methodologically and quantitatively different from the only other extant Latin frequency dictionary, namely the Dictionnaire Fréquentielle created by the L.A.S.L.A. laboratory of the University of Liège in 1981 (and recently republished) that is based on a limited corpus of 19 authors (for a total of 1 706 7276 lexical occurrences analysed) and on the Forcellini’s Lexicon Totius Latinitaits (1864). Our dictionary (c. 10.000.000 words) so far includes 30.489 headwords and 329.310 forms which are alternatively marked as certain (circa 295.271) and ambiguous (circa 34.039) to alert users of the presence of eventual homographs. More importantly, our program allows users to display the results not only in an alphabetical and a chronological order, but also by absolute frequency and by relative chronological frequency. In addition, for each headword the dictionary records the authors in which the word is attested so that users can decide to see the frequency of a word in a specific timeframe/author.

Secondly, the program can be used to detect the words (and/or the word-forms) shared only by two authors in the entire Latin literature. This function has been designed to support the digital study of intertextuality. For instance, users can easily see which words first attested in Ovid are used only by Statius in extant Latin literature: by revealing meaningful reuses of extremely rare forms the lists created by our program can contribute significantly to the study of an author’s use of allusivity and style.