• Italian
  • German
  • French
  • English
  • Spanish
  • Russian

Presentation of the LBC database[1]

Annick Farina (Università di Firenze), Riccardo Billero (Università di Firenze), Carlota Nicolás Martínez (Università di Firenze)

The LBC Database is one of the support tools in Open Access developed by the Lessico multilingue dei Beni Culturali project (Multilingual Lexicon of Cultural Heritage). The Research Unit aims to provide a corpora with such specific lexical information as to enable lexicographic and translation research. We created, for this purpose, a digital space with various tools, which will spread the knowledge of Tuscany's artistic and cultural heritage at an international level (Farina 2016). The database permits searches within the texts of published corpora in six different languages (French, English, Italian, Russian, Spanish and German) through the project’s platform, which contains various tools including the corpora and information about them[2].

The corpora encompass texts of various genres, such as classical literary works, travel novels or correspondence, scientific and technical texts, tourist guides, textbooks, etc., all written over an extended period of time; these sources have been organized and managed through a software with functions suited for responding to the needs of multiple users. The main target groups of the corpora are: linguists, scholars, and humanities and social sciences researchers, whose work requires data on the lexical information of an author, chronological period, genre, etc.; translators who need to consult specific lexical resources; specialists in the tourism sector; or tourists interested in deepening their knowledge of the territory and its culture.

For each language of the project, the texts were chosen based on two priorities: first, the recognised prestige of the text/author in the source culture (Billero, Nicolás 2017: 208); secondly, the ease of conversion into an editable format, avoiding texts that were difficult to digitise. The translated texts were drawn from a list created by the project’s members containing the texts in Italian and other languages considered essential for the international knowledge of Tuscany's artistic and cultural heritage: the fundamental art history texts referring to Tuscany, such as Vasari’s The Lives of the Most Excellent Painters, Sculptors, and Architects; the architecture books by Alberti, Palladio and Sellio; some writings by Machiavelli and Leonardo Da Vinci; well-known travel books, such as the travels of Stendhal and Ruskin; and art books like Burckhardt’s.

However, the different types of texts in the corpora were given different priorities and proportions, for various reasons: the criterion of accessibility to sources is obviously different according from country to country and so is the interest in Tuscan heritage, which varies according to historical periods and textual genres in the various languages/cultures represented in the project. In the future development of the project, we hope to limit the resulting disparity among the corpora. In fact, at the end of this initial corpora-building phase, an analysis of the distribution of the types of texts in each corpus, as well as their time periods, will allow for greater uniformity in the future, thus enabling more comparative work on texts. In this first phase, instead, priority was given to including reference texts which provide a sufficiently comprehensive base of texts for searches in each single language.

After a careful analysis of the various software available for consulting the corpora, our choice fell on NoSketchEngine, as it offers several interesting features which met the project’s purposes, namely allowing concordance queries and filters with various features (Billero 2020).

Specifically, you can access information on the nature of the contents of each corpus by accessing "Corpus info" from the NoSketchEngine menu (Figure 1).

Detailed information on the French corpus available under Corpus info
Figure 1 – Detailed information on the French corpus available under "Corpus info" [Nov. 2022].

This same page also provides information about the quantities of the various documents in each of the categories provided, as shown in Figure 2 for the English corpus

Structures and attributes of the documents inserted in the English corpus
Figure 2 – Structures and attributes of the documents inserted in the English corpus [Nov. 2022].

The structure of the corpora follows the traditional rules concerning shared metadata management, as can be seen in a "Search" for text types ("Text types"[3], Figure 3).

Search in the German corpus through the Text types window.
Figure 3 – Search in the German corpus through the "Text types" window.

The metadata you can filter the concordance search with are:

  • Original language: both the language of the text and the source language for translated texts appear;
  • Translation language: allows a search for all translations within the corpus language;
  • Category and subcategory: indicates the various types of texts. All the texts have as their subject the artistic heritage and its lexicon, in particular a broad survey of Florence and Tuscany from different points of view. There are four macro-categories (Informative, Technical, Dictionary and Literary), each with their relative sub-categories (Informative: Blog, Guide, Magazine; Technical: Architecture, Art, Food and Wine; Literary: Biographical, Fiction, Essay; Dictionary: Monolingual, Bilingual / Multilingual). In deciding on these categories, the main uses and users of the work were taken into account; this conditioned the type of language involved and its level of specialisation[4], specifically:
  • Author: surname and first name are indicated , with the designation "sa" when non-existent;
  • Title and fragment: we chose the introduction of both whole texts and of fragments that correspond to a textual unit, such as book chapter, complete letter, journal article, etc. This was due to the fact that in many cases the entire book did not coincide with the interests of the project. Furthermore, this facilitates the future creation of parallel versions of translated texts. For the translated texts both the original and translated titles are included;
  • Year of writing / year of publication / year of translation: the chronological information differentiates between the period of writing of the texts (where possible) and the date of publication; for translated texts the same information was inserted both for the source text and for the translated one[5]. For online publications, the date of consultation is indicated;
  • Source: allows to search within a single document of the corpus (book or fragment);
  • Geographical delimitation[6]: for texts regarding a specific city or region, the name of the location has been inserted. This indication is present mainly for travel books and correspondence.

Further, more complete bibliographic details are visible by accessing the concordances by clicking on the reference (file name, document number, author name, etc. according to the options chosen in "View options", Figure 4).

Choices available for viewing the textual references in View options.
Figure 4 – Choices available for viewing the textual references in "View options".

Using the "Search" function, you can access the concordances displayed in random order (on the number of documents) as in Figure 5, or in alphabetical order in relation to the queried word or to its right or left context, using the "Sort left/right" function (Figure 6).

Search for concordances for the lemma pintar in the Spanish corpus without choice of order.
Figure 5 – Search for concordances for the lemma pintar in the Spanish corpus without choice of order.
Search for concordances to the left of the lemma Kirche in the German corpus.
Figure 6 – Search for concordances to the left of the lemma Kirche in the German corpus.

It is also possible to search for the presence of two words or lemmas in the same context at a chosen distance of tokens by using the “Context" function in the “Search” menu, as shown in Figure 7A. This would allow, for example, verification of the certified uses of various collocations (dipingere a fresco / in fresco in Italian in Figure 7B).

Search for the dipingere and fresco lemmas at 5 tokens distance in the Italian corpus.
Figure 7A - Search for the dipingere and fresco lemmas at 5 tokens distance in the Italian corpus.
Concordances relative to the research of dipingere and fresco in the same context in the Italian corpus.
Figure 7B - Concordances relative to the research of dipingere and fresco in the same context in the Italian corpus.

The "Word list" function allows users to obtain numerical results on the frequencies present in a corpus both according to the sources, by querying for example the frequencies of headwords attributable to each author (Figure 8), and according to the headwords of a corpus (Figures 9 and 10).

Frequencies of tokens present for authors in the Russian corpus.
Figure 8 - Frequencies of tokens present for authors in the Russian corpus.
Word list search of lemmas in the English corpus.
Figure 9 - Word list search of lemmas in the English corpus.
Result of the Word list search of lemmas in the English corpus
Figure 10 - Result of the Word list search of lemmas in the English corpus [Nov. 2022].

The completion of this first phase of our corpora is to be considered satisfactory as it created the necessary foundations for initial works and research for our group (Carpi 2017; Farina, Billero, Carpi 2018; Garzaniti 2020; Farina, Flinz 2020). The first lists of entries of each language have already been performed; this will be complemented by concordances extracted from the corpora to be published on the platform by 2021. At that point it can be used for producing future dictionaries.

The main objective of this preliminary work, carried out by each linguistic team, was to validate the corpora, recognising that only by actually using it would they be able to identify problems that would otherwise remain latent.

In the future, expanding both the number of languages (currently there are still no Chinese, Portuguese or Turkish corpora, languages that are part of the LBC project) and the number of texts (to increase uniformity as mentioned above) are planned, to try to make the corpora as comparable as possible.

Bibliography

Billero R. (2020), Cultural Heritage Lexicon: A Case Study. In Ana Pano Alamán, Valeria Zotti, The language of art and culture heritage: a plurilingual and digital perspective, Cambridge Scholars Publishing, pp. 86-103.

Billero R., Carpi E. (2018), Corpora e terminologia artistica: il caso del corpus spagnolo LBC. In CHIMERA Romance Corpora and Linguistic Studies, Madrid, UAM, 5, no. 1, pp. 85-91.

Billero R., Nicolás Martínez M.C. (2017), Nuove risorse per la ricerca del lessico del patrimonio culturale: corpora multilingue LBC. In CHIMERA Romance Corpora and Linguistic Studies, Madrid, UAM, 4.2, pp. 203-216.

Carpi E. (2017), El lenguaje para fines artísticos: traducciones de tondo al español. In Alejandro Curado (ed.), LSP in Multi-disciplinary contexts of Teaching and Research. Papers from the 16th International AELFE Conference, vol. 3, pp. 79–84. https://doi.org/10.29007/wx3m

Farina A., Nicolás Martínez C., Billero R. (eds.) (2020), I Corpora LBC, Firenze University Press, Firenze.

Farina A., Flinz C. (2020), Analisi comparativa dei corpora LBC. La visione del patrimonio fiorentino francese e tedesco: l'esempio del Duomo. In Fernando Funari, Annick Farina (eds.), Le présent dans le passé / Past in Present/ Il passato nel presente, Firenze University Press, Firenze.

Farina A., Billero R. (2018), Comparaison de corpus de langue « naturelle » et de langue « de traduction » : les bases de données textuelles LBC, un outil essentiel pour la création de fiches lexicographiques bilingues, JADT’18 Proceedings of the 14th International Conference on Statistical Analysis of Textual Data, UniversItalia, pp. 108-116.

Farina A. (2016), Le portail lexicographique du Lessico plurilingue dei Beni Culturali, outil pour le professionnel, instrument de divulgation du savoir patrimonial et atelier didactique. In Publif@rum, n. 24, 2016. http://www.farum.it/publifarum/ezine_articles.php?art_id=335

Garzaniti M. (2020), Il termine russo friag e le sue radici nelle relazioni culturali e artistiche fra la Russia e l’Italia. In Ana Pano Alamán, Valeria Zotti, The language of art and culture heritage: a plurilingual and digital perspective, Cambridge Scholars Publishing. pp 104-119.

Notes

[1] This text is a translation by Carlo Garavaglia of the Italian introduction to the LBC corpora published in http://corpora.lessicobeniculturali.net/it/

[2] For comprehensive data on the LBC corpora, please refer to the group’s publication (Farina, Nicolás Martínez, Billero 2020).

[3] The "Text Type" search interface is currently in Italian , but we will shortly edit it so that it can be consulted in the language of each corpus.

[4] In the next phase of the project the classification will be reviewed in order to overcome the problems encountered by some groups with texts that could be belong to more than one category, e.g. texts by classical authors whose style is clearly literary but who wrote texts that can be considered specialised for their content and vocabulary (e.g. Stendhal's History of Painting is classified for now in the literary/ essay category).

[5] The texts in the corpora range from the Renaissance to the present day. Although both dates are present, the year of publication is secondary to the year of writing. The latter, in fact, is the most interesting date when extracting information, since it is representative of the linguistic characteristics of the period being examined; in fact, the texts, as entered into the database,  remain faithful to the edition used, without any kind of modernisation or spelling correction.

[6] This option will be available from 2023.