The Drawings of the Florentine Painters

Project background

Bernard Berenson’s Drawings of the Florentine Painters is a digital resource based on Berenson’s publication of the same name. This research tool permits searches for any of the drawings listed in the three editions (1903, 1938 in English; 1961, in Italian). Filters allow browsing and searching by artist and title (as provided by Berenson, in English or Italian), location and technique. Each entry includes key information from all three editions of Berenson’s text, as well as the current location, an image of the catalog page, and plates included in 1903 and from the British Museum. Many entries for drawings by Sandro Botticelli and Filippino Lippi (and their schools), have been enriched to include links to museum webpages.

This is the first online database to include of a large corpus of Italian Renaissance drawings from different collections. Berenson catalogued roughly 4000 sheets by painters active in Florence from 14th through the mid-16th century; these are mainly figure drawings, from Taddeo Gaddi through Michelangelo. Berenson’s original three volumes, published in a luxury edition 1903 as The drawings of the Florentine painters, classified, criticised and studied as documents in the history and appreciation of Tuscan art, with a copious catalogue raisonné represent a revolution in both form an content. For the first time, a scholar had brought together essential information about drawings found in all known collections, public and private, and made them accessible in a single catalog. Berenson also include a series of essays on each artist (not included in this database) and reproductions of hundreds of works. Over a century after they first appeared, Berenson’s entries are still valued today by drawing specialists for their attributions and commentary. The three editions of Berenson’s monumental catalogue, and the very criteria used by Berenson to define his field of study, also provide a unique window on the development of art history. The conversion of the catalog to Linked Open Data, available at the Villa I Tatti data endpoint, make it available in a machine-readable format. This allows researchers to mine the source content in ways that are currently difficult or impossible. The online catalog makes it extremely simple to search by technique, location, or keyword; the latter search includes titles and commentary in both English and Italian. The database also allows specialists to carry out more detailed analyses, e.g. to see how Berenson changed the attributions of individual drawings, or how he expanded or restricted the corpus of artists.


Berenson’s catalogue entries have been transcribed and transformed into a digital dataset: the attribution, title, technique, and location, plus the commentary he included for many sheets. The indications for technique (e.g. the use of ink, metalpoint, wash) found in the 1961 edition have been translated and standardized with terms from the Getty’s Art & Architecture Thesaurus. The names of cities and collections provided in 1961 have been translated and standardized with terms from GeoNames and the Virtual International Authority File (VIAF). In selected entries, especially for Botticelli and Filippino Lippi, information about location and inventory numbers was updated. Here, as in Berenson’s book, what constitutes a drawing resists a clear-cut definition. For example, Berenson might list a single sheet with markings on both sides as one work, or give separate numbers to the recto and a verso. Drawings might be part of a complex object such as a sketchbook, or else mounted together. In general, we have defined a drawing as an object to which Berenson assigned a catalog number. Sometimes --as in the case of several drawings mounted together-- we have deviated from Berenson’s system; in line with modern cataloging practice, we assign an identifier to each drawing.

This project began by converting the content of Berenson’s publications into a Linked Open dataset, and providing a search interface that will support resource discovery and data analysis. For Botticelli and Filippino Lippi, the content was enriched through interlinking with online museum catalogues that provide researchers with updated information and illustrations. In contrast to a traditional database, where all data remain on the host’s website, the linked catalogue allows researchers to have direct access to the data through open data services like a SPARQL endpoint and a data dump.

The database aims to serve as a representative use case. The infrastructure and methodological framework could be applied to similar art history initiatives which focus on scholarly inventories, narrative articles, and photographic documentation. The database itself could be incorporated into a larger one that included, for example, other forms of art, or drawings from different areas or periods.

The entire dataset has is also openly available in RDF for reuse under a Creative Commons Attribution-ShareAlike license.