The Chronopéra database is one of the pioneering projects in the history of performing arts using digital technologies to make research available online. It is the result of a meeting in 2003 between three CNRS researchers: Michel Noiray (musicology, Institut de recherche en musicologie), Solveig Serre (musicology, Centre d’études supérieures de la Renaissance), and Olivier Serre (theoretical data processing, Institut de recherche en informatique fondamentale). Its creation was triggered by the desire to build a statistical tool for historical interpretation. In reconstituting the chronology of all the performances given by the Opéra de Paris from 1669, its primary aim is to suggest an interpretation of events and tendencies of lyric creation within this institution; its usefulness is threefold: to clarify issues of reception by revealing what the public went to see in a given period; to measure the effect of novelty by revealing the general context in which the creations took place; and to expose how the Opéra developed its artistic strategy by comparing its directors’ intentions and those of the various individuals responsible for production. Chronopéra was then conceived as a practical and easily accessible tool for researchers. Its content being constantly enriched, this database is at the crossroads of two fields of research that are too often separate – institutional history and the history of performing arts. Chronopéra is thus essential as a documentary resource and an indispensable tool for historians of music, dance, and performance.
Chronopéra provides the daily programme of the Opéra de Paris from its inception in 1669 until 1989, year of the opening of the Opéra Bastille. It is therefore a closed and homogenous corpus, with boundaries that can be easily justified since the Opéra de Paris built its repertoire from the start as a resource in which shows remain on a long-term basis. The directory was not, however, easy to compile from an archival perspective because, unlike other major cultural institutions such as the Comédie-Française, the Opéra de Paris did not benefit from consistent management, nor did it do a very good job of preserving its archives. Major work was therefore required to compare primary sources: documents in binders, registers, and logbooks kept at the Bibliothèque Musée de l’Opéra (Bibliothèque nationale de France), but also leaflets and bundles of paper kept at the Archives nationales (in the O1 or AJ13 collections), and secondary sources (libretti, newspapers, etc.)
Chronopéra seeks to demonstrate that the study of musical phenomena can and must include the study of extra-musical phenomena within a social history of music approach. It reveals not only formal information concerning the works (genre, number of acts, etc.), but also the system of multiple constraints (financial – due to the daily takings, material – relating to entertainment facilities, etc.) in which they were produced, by connecting them to larger issues of reception and aesthetics (programming according to the days of the week or theatre seasons, for example). Prior to the recent creation of several large-scale databases dedicated to other cultural institutions (Dezède, the Comédie-Française Registers Project , Chronopéra was a pioneering tool.
Which software programs were used to build the database infrastructure and, as the case may be, to treat the data statistically?
Chronopéra works with MySQL, a free software which allows easy online access at no cost, and is searchable by way of SQL query language. As this use is only accessible to users with advanced knowledge in SQL and a good sense of the structuring of various database tables, a user-friendly interface via an internet site was gradually installed. It allows most standard searches to be carried out: display of the calendar for a given period; or of the works of a composer, a librettist, or a choreographer. The results of these different searches can be sorted according to several criteria and exported to different formats (.pdf, .xls, or .csv). The expert search method, via the use of SQL searches, allows full exploitation of the Chronopéra database with the help of multi-criteria searches on the various tables it comprises. It is, for example, possible to generate a table of composers sorted according to the average revenue per performance during a given period. Finally, one can identify which days of the week were the most prolific, whether there were certain months during which the public did not attend, and the evolution of these phenomena over the centuries.
Could you offer one or two examples of scientific (whether consensual or surprising) results obtained with the help of the database?
Being a specialist in the history of the Académie royale de musique in the second half of the eighteenth century, I conducted a detailed and systematic analysis of the data contained in Chronopéra for that period. It became apparent, for example, that the programming of the repertoire adhered to three major principles. The first was that of the “long series,” applied for essentially economic purposes – the same work would be performed several times a week for a number of weeks. The second was that of alternating between different operatic and choreographic genres, or between different titles during the season or over months or weeks. In this case, it was about finding the best compromise between the house’s artistic inclinations and those of the public. The third was that of “coupled performances,” composed of extracts from popular works that were performed one after the other without an attempt to connect one to the other. This inexpensive practice offered the advantage of pleasing audiences without exhausting the considerable financial and human resources that had been invested in the grand opéra. This therefore shows how the programming of the repertoire was a particularly complex procedure that required coordination, arbitration, and control; and how, in this context, the artistic freedom of house directors appears unstable and fluctuating.
Figure 1. Average ratio by month between operatic (“lyrique”) and choreographic (“chorégraphique”) works at the Académie royale de musique (1749-1791).
Graph produced from analysis of the Académie royale de musique ledgers held at the Bibliothèque musée de l’Opéra Garnier (shelf mark: CO2-CO25).
It is clear that the 1714 regulations that recommended reserving winter and spring for operatic tragedies, and autumn for ballet were respected. Throughout the period, the season in which the operatic genre was preferred runs from October to April, and that of the choreographic genre from May to September. The months in which alternating is practised the most are May, October, and November.
First of all, it is a good time to relaunch Chronopéra, which has been inactive since 2012 because of time and staffing issues. This new momentum will take several forms: the verification of existing data; the reconstitution of the chronology of performances from 1672 to 1749; the addition of distribution data to chronological data; and alignment on the authority records issued by the BnF for works, performances, performance halls, and individuals. Parallel to this work, a new, more functional and user-friendly version of Chronopéra will be developed. Then, the perpetuity of the database will be guaranteed by its integration into the “Very Large (digital) Facility” Huma-Num, as well as by the preparation of a print document modelled on the book Théâtre de l’Opéra-Comique Paris. Répertoire 1762-1972 by David Charlton and Nicole Wild (2005). Finally, links permitted by the interoperability of the data will be made to similar projects in the field of the history of performing arts, in particular to the Comédie-Française Registers Project.