From the outset, we prioritized humanistic intervention:
- During the process of database construction
- During the design of the interface
Database Construction
Typical database design requires the eventual user to have one of two things:
- prior knowledge of key actors in the period, important dates, sites of intellectual exchange, and significant keywords (all likely brought to light by previous scholarship).
- enough curiosity to provoke browsing alphabetized lists of people, dates, places, or concordances in the hopes of finding something interesting.
While our underlying data does capture that information, it also tries to bring to light contextual features of authors and textual elements. Relationships between pieces of the database are more than just relationships across attributes captured in standard databases.
For example, if we take a letter by Galileo that matches a keyword search, it would be helpful to both an expert scholar and a student to know the following:
- Author -> the most central figure in Galileo’s library
- Recipient -> eccentric (not included) in Galileo’s published work or books in Galileo’s library
- Person mentioned -> higher centrality in the library than in correspondence
- Destination -> the most frequent city to which letters were sent; the second-most frequent city in which books in the library were published
- Source -> the most frequent city from which letters were sent; the third-most frequent city in which books in the library were published
- Date -> the year in which book collecting started to increase; a year with many missing letters, but still a high frequency for the overall correspondence
- luna (moon) -> a high frequency word used by several authors in the correspondence and library
- perspicuo (clear) -> a high frequency word in Galileo’s correspondence and published works, but low frequency in the library and letters by other authors
- scabrosità (roughness or offensiveness) -> a low frequency word used by Galileo in correspondence, not used by other letter authors, and low frequency in published works and books by other authors in his library (poets and travelers)
- quietare et persuadere (to silence and to persuade) -> a phrase with high frequency in Galileo’s prose and history texts (such as Machiavelli, Guicciardini, and Sarpi) in the library
Interface Design
By creating this metadata for documents, we facilitate structuring a query that identifies relationship intensities (central, peripheral or eccentric), frequencies (maximum, high, low, minimum, none), and document categories (private, public, print tradition in the library) of interest. In response to Witmore and Hope, we aimed for tools that “allow you to have your books on multiple shelves simultaneously.” Such a structure is an invitation, one that does not require knowing someone or something “interesting” before exploring. In this sense, we respond to a challenge that Ryan Cordell identified in the development of the large-scale digital text archives: we often do not know what to put in a search bar.
Scholars and students do know what kinds of questions are interesting, and we hope to have built something to model that kind of query. In that sense, GaLiLeO becomes a site of discovery, not simply transmission. The hierarchies often embedded in digital tools are brought onto the same plane. By changing identifiers related to documents (not alphabet or numerical), we place them in an interconnected space of representation and understanding (thus they are actual relations rather than a relational table).
Return to GaLiLeO landing page.
