Online Resources for the Humanities: Interdisciplinary Perspectives.
International Seminar May 7-8, 2004

PRESENTATION OUTLINE. Opera del Vocabolario Italiano: lessicografia, informatica, risorse on-line.

The Opera del Vocabolario Italiano (OVI; The Italian Dictionary: www.ovi.cnr.it or www.vocabolario.org) is an Institute of the Italian National Council of Research (CNR), whose task is the elaboration of the Historical Dictionary of the Italian Language. This enterprise has a long history, indeed, linked to that of the Crusca Academy founded in 1582-83. The first edition of the Dictionary was published in 1612. In 1923, the fifth edition of the Dictionary was interrupted at the letter O by the Italian government, which altered the structure and tasks of the Academy. Later, the idea of a new Dictionary, based on a historical rather than normative approach, was introduced by scholars; and in 1965 the Academy received support from the National Research Council to resume its lexicographical activities toward the publication of a new Dictionary. As a result, the Opera del Vocabolario was created. After twenty years of intermittent work, the Opera became independent from the Academy, and was established as an Institute of the CNR in 2002.

Back in 1973 a decision was made to give precedence to the ancient portion of the historical dictionary, from the origins of the Italian language to the late 14^th century (as a symbolic date the year of BoccaccioÕs death, 1375, was chosen as the stopping date). The elaboration of the portion referring to the period from the 14^th to the 20^th century was postponed to a later date. The main object of the OVI work is therefore the Thesaurum of the Italian Language of the Origins. While there were practical reasons behind this choice, a Dictionary of ancient Italian also has unique characteristics that require its autonomy. First of all, the codification of the Italian national language (as a cultural phenomenon, since the Italian nation-state did not yet exist), based on the Florentine vernacular written by 14^th-century authors, was established only in the 16^th century. Until then, one cannot really speak of a national language compared to which the other linguistic varieties of Italian would be considered dialects in a socio-linguistic sense. Secondly, by establishing the boundaries of the historical dictionary at the end of the 14^th century it is at least possible to have as an objective an exhaustive first-hand examination of the edited texts which comprise a significant but manageable corpus. Its is clear that the study of the lexicon of following centuries can be only based on a combination of a first-hand examination of selected corpora with a systematic use of second-hand sources, among which are the dictionaries produced in the history of the Italian language. In this area, OVI is now at work on software for the online publication, in a searchable database, of the five editions of the Dictionary of the Academy, which will represent an important source for the Historical Dictionary from the 16^th century onward. Currently, only the first edition is available through the Cribecu – the information technology Center of the Scuola Normale of Pisa, devoted to the preservation of Cultural Assets.

The OVI was born as a center of lexicographical research. However, information technology (IT) has had a great role in its work since the center's inception. Starting in 1965, all texts to be used for the ancient section, as well as a portion of those to be used for the modern section, were digitized in a format then available but now practically obsolete, by CNUCE (National University for Electronic Data Elaboration) of Pisa, which later became di Computational Linguistics Center (ILC) of CNR (fundamental the contribution of Antonio Zampolli, recently and tragically deceased). More recently, this patrimony of digitized texts (circa 14 million entries) has been retrieved, and the new database (built in 1995) has now reached 20.5 million entries and is still growing.The first version of this database employed a piece of software by Eugenio Picchi, the DTB, later replaced by GATTO by Domenico Iorio-Fili. Since 1998, a copy of the database has been available online in a searchable format, thanks to ItalNet's collaboration with the Universities of Chicago, Notre Dame and Reading (UK): this version is implemented in PhiloLogic (by the ARTFL, directed by Mark Olsen, with the collaboration of Christian Dupont of Notre Dame, now at Syracuse U.). It is this OVI resource that best known in the US (available here or through ARTFL.)

The database is extendible and updatable indefinitely, in order to accomodate new or revised editions of texts. On the website, following the link "Notizie" > "Citazioni dei lavori dell'OVI" ("News" > "Citations of OVI works") one finds a very long, asystematic list of publications which acknowledge the use of OVI tools, mostly the database since TLIO includes only a quarter of its intended data. The kind of texts that are inserted in the database are the following:

a) all texts edited in a reliable edition, datable before 1375, and belonging to all the varieties of the Italian linguistic system (with the exclusion of texts in Sardinian, Ladin, Friulan and those belonging to non-Italian languages, even though written in Italy);
b) all texts with the same characteristic of edition and language, generically datable within the 14th-century;
c) some texts with the same characteristic of edition and language, datable with precision or generically between the end of the 14th- and the beginning of the 15th-century;
d) some texts in edition deemed insufficient which however cannot be excluded from the documentation (bibliographical abbreviations are marked with acute parenthesis to warn the reader); and
e) some texts belonging to the linguistic varieties excluded (see above) but useful for research purposes and for the dcitionary in specific cases.

Thanks to the systematic use of the tools provided by IT, it was possible to project the TLIO as a first-hand dictionary, based on the direct examination of the texts, without recurring (except for control) to already existing dictionaries. The project and workplan is outlined in the Edition Guidelines (Norme di redazione), available in an updated version on the site (http://tlio.ovi.cnr.it/TLIO/). Another important decision has been that to publish the TLIO as a work-in-progress on the Internet. On the one hand, the TLIO is a historical dictionary with a traditional structure (for instance, it offers definitions documented with chronological examples always quoting the oldest available example for each definition). On the other hand, the TLIO is one of the first dictionaries to be born directly online, not a printed dictionary digitized and made available online. This means, for example, that the entries are always revisable and updatable with the new documentation that becomes available on the database, including corrections made upon notifications by users). In the OVI newsletter (Edizioni dellÕOrso, www.ediorso.it) are examples of entries, whose online version may be, however, the most up to date. OVI is thus a lexicographical enterprise which at the same time elaborates software and distributes resources online. This complex activity (which, for instance, is not supported by adequate human or financial resources) can be thus outlined:

- texts to be digitized are identified and the bibliography (one of the most important as far as as ancient Italian texts are concerned) is updated;
- edited texts are reviewed according to the criteria of their edition and then normalized (parentheses etc.); the parts in Latin and vernacular are distinguished and a light encoding introduced (page and paragraph divisions etc.) according to the guidelines of GATTO;
- periodically, all digitized texts are trasmitted to ARTFL for the updating of the ItalNet database; PhiloLogic has been customized in order to translate the encoding of GATTO (Valentina Pollidori is the supervisor for this task);
- digitized texts are included in the GATTO database used at the OVI headquarters;
- the GATTO database is lemmatized in order to allow the editors to search for materials in the preparation of entries;
- using the database, editors select materials for the entries (under the supervision of Päaut;r Larson and Paolo Squillacioti);
- entries (in MS Word) are implemented in a database searchable online, using software also written at OVI (the work of Andrea Boccellari); Word files do not contain tags but the entry scheme allows the TLIO software to recognize the elements that will have to be searched (for example, the list of graphic forms);
- the TLIO software also allows to go back from the entry to the Italnet online database, so that anyone reading an entry that contains only one selection of examples is capable of reconstructing the whole documentation.

All this work is necessary to edit and publish the TLIO online; however, the results also have other applications. For example:

- the OVI database of ancient Italian is one of the primary sources of the Italian Etimological Lexicon, under construction under the direction of Max Pfister in Saarbrüaut;cken (TLIOMat);
- the experience acquired in the area of textual databases has allowed OVI to win a contract with the State Archive of Prato for the elaboration of a database of the edited letters of the Francesco Datini Archive, and is looking to become a provider of online textual databases also to other potential patrons and clients;
- the GATTO software is offered to scholars for their work (freely downloadable from the web only for non-profit, research and educational use); it is adopted by research groups collaborating with OVI (CLIO of the University of Lecce, ITALANT of the University of Padova).

An online version of GATTO is currently being prepared and will become available at the end of 2004. Before giving you a demo of the OVI resources online, I would like to underscore the fact that, although IT is an essential part of our work, equally indispensable are the linguistic and philological competences of those who prepare the texts for the database, examining the extant editions, and of those who compose the entries, interpreting texts in the most diverse varieties of Medieval Italian (from Tuscan to Lombard, from Neopolitan to Sicilian etc.) solving numberless problems of graphics, phonetics, morphology, syntax. In the "computer/humanities" (informatica umanistica) coupling sense, the most visible part today is perhaps the computer and it is true that IT has modified our working methods and also its results. However, the validity of the content matter and knowledge production still mostly depend on the humanistic competence (I would not say "traditional" but simply "necessary") of those who use it.

Presentation of OVI online resources:

Italnet online database of ancient Italian;
Prototype of Gattoweb
TLIO searchable online
Online bibliography of quotes;
Bibliography of the translations (volgarizzamenti) into vernacula of the TLIO Corpus.

Biographical note

Pietro G. Beltrami studied at the Scuola Normale Superiore and the University of Pisa, where he is currently a full professor of Romance philology. He has also taught at the Universities of Parma, Potenza and Siena. A member of the Accademia della Crusca, the literary society founded in Florence in 1582 to maintain the purity of the Italian language, Beltrami directs the OVI (Opera del Vocabolario Italiano), an institute of the National Council of Research, in Florence. His major field is Italian and Romance metrics and he has published in this area on Dante, the troubadours, Chrétien de Troyes and Brunetto Latini. He is the editor-in-chief of TLIO, Tesoro della Lingua Italiana delle Origini (Treasury of the Italian Language of the Origins) which, at the end of 2003, included 11,000 entries.

[back]

Online Resources for the Humanities: Interdisciplinary Perspectives. International Seminar May 7-8, 2004

PRESENTATION OUTLINE. Opera del Vocabolario Italiano: lessicografia, informatica, risorse on-line.

Online Resources for the Humanities: Interdisciplinary Perspectives.
International Seminar May 7-8, 2004