The development of this database was undertaken in order to study the conditions surrounding the disappearance of “criminal ballads” – that is to say, sung texts recounting, for informative and/or edifying purposes, the details of a genuine criminal news event – within French territory. As much as the “end of a genre,” the transformation of this popular medium into what is now commonly referred to as the “realist song” lies at the heart of this project, which is both historical and musicological.

To achieve this, as exhaustive an inventory as possible of the “blood-soaked broadsheets” of this period has been undertaken, in both private and public collections. Then, the dating and localization of the events described in the ballads made it possible to construct the structure of this database, in which the following are connected:

• News events across metropolitan France. All locations concerned are recorded with their INSEE code, allowing computerized cartographic development;
• Ballads collected in full or in part, or merely attested;
• “Tunes,” or pre-existing melodies, to which the ballads are sung;
• A typology of the sung news events (opportunistic crimes, parricides, infanticides, etc.);

The corpus

Definition

It consists of ballads and songs[1] in the French language dealing with proven criminal events committed on French territory, from 1869 (the Troppmann affair, or “The Pantin crime”) to the Second World War. The chronological limits were chosen both with an event-based reference point – the media coverage of the Troppmann affair marks the emergence of news events in the regional press – and a commonly accepted endpoint for the disappearance of the distribution circuits of ballads – the end of peddling.

The case of ballads written in regional languages (Breton, Basque, Occitan, Flemish) is currently being addressed through collaborations made necessary by the database creator’s lack of mastery of these idioms. The case of Brittany is particularly interesting due to the vitality of the dissemination of songs “on loose sheets,” already the subject of abundant bibliography. Breton ballads are integrated into the database thanks to the assistance of Bernard Lasbleiz, an associated researcher at the CRBC (Centre for Breton and Celtic Research) in Brest.

This study focuses only on crimes, in the legal sense of the term, therefore subject to criminal court or military justice, whether or not a trial took place. Acts comparable to such actions (simulated kidnapping, for example) or serious offenses (perceived as crimes by the population) have also been taken into account. The production must be considered a “current affairs song,” i.e. be contemporary with the event being sung about: either at the time of the revelation of the crime, or during the arrest or identification of the criminal, or at the time of the trial, or even the execution.

The collection of a ballad can take place in three ways:

• Consultation of a printed sheet, or its transcription into a song notebook;
• Oral collection from a singer, or publication of such a collection;
• Mention of the existence of a ballad (press article, singer’s recollection, legal deposit registers).

Particular cases

The First World War is included in the period studied. This raises the particular case of ballads and songs relating to war crimes[2]. Their assessment is delicate, as the only identified cases turn a tragic event into a propaganda tool: the legend of the executed child (a child shot because he mimicked firing a wooden rifle at a German soldier), or the case of Émile Desprès (a pseudonym attributed to a young hostage from the North, shot because he offered water to a wounded territorial soldier). Nevertheless, they were not excluded, as they meet all the criteria required by the corpus.

Some ballads in the corpus, relating to crimes committed on French territory, were published abroad (Belgium or Switzerland). This is particularly the case for events in northern France, given the geographical proximity. Some “foreign” publishers seem to have fictitious identities and fall into pure invention, particularly where copyright is at stake[3]. Many song sheets reproduce protected works or use registered melodies for topical parodies; in order to protect themselves from complaints, referring to a distant printer seems to have been the peddlers’ means of defense in the event of inspection.

Conventions had to be established, both geographically and chronologically:

- The localization of multiple events (notably serial killers) has been arbitrarily set in the department where the accused appear before the Assize Court, and the commune cited is, depending on the case, that of the last crime, or the one that gave its name to the case.

- The dating of events spread over a period corresponds to the date of the criminal’s arrest.

- The dating of the ballads is identical to that of the event, unless it corresponds to the trial or execution. In the case of Vacher, one ballad predates the others, dating from 1895 (at the time he had not yet been identified), the date of the crime described.

The database

The organization of the collected data is based on several tables:

The “news events” table.

Its fields are: the date, a summary of the case, a link to a related press article (discovery of the crime, arrest of the murderer, or trial), the department, the commune, the type of crime (opportunistic, passion-driven, etc.), and an indication of the gender of the criminals (man/men, woman/women, mixed, unknown).

The “title” of the event is generally that used by the press, with all the ambiguity that entails, as some cases had several names: who remembers the “Madagascar Street crime,” quickly becoming the “Violette Nozière affair”? Or the “Pantin crime,” which became the “Troppmann affair”? Users are therefore invited to refer to the case summary to ensure there is no confusion.

The “Ballads” table.

For each ballad included in the database, the following are recorded: its author (if known), the location where the document is held, the tune (melody used), the number of verses, the possible presence of a chorus, the date, any comments, and a link when an image of the ballad is already available online.

At the top of the page, a clickable image allows zooming in on the lyrics of the ballad. Eventually, all collected ballad lyrics will be available as they are progressively uploaded.

At the bottom of the page, users have direct access to other ballads possibly written about the same crime.

The “Tunes” table.

From a ballad, one can access the tune to which it is sung. The following are specified: its title, its origin (which musical genre it belongs to for older melodies, or its author and composer for more recent ones), its creation date (for recent melodies), and its type (ballad, song with chorus).

Publishing these melodies online raises potential copyright issues; as negotiations progress, a link will point to a score allowing the ballad to be sung.

Limits and source effects

Media

The collection of ballads in printed form is challenging. Their legal deposit was far from systematic. The preservation and cataloging of broadsheets – like all “occasional” or ephemera – are not standardized across the territory, and the discovery of new collections is sometimes a matter of chance. We usually do not know the size of print runs, and it is almost certain that some ballads have physically disappeared. The same applies to song notebooks, private writings rarely preserved in public archives. Some institutions dedicated to the promotion of intangible cultural heritage take an interest in such sources, publishing them or making their catalogs available online.

Ethnographic collection

This undertaking is not homogeneous across French territory; consequently, it is likely that a greater number of ballads will be collected in areas where investment in protecting and preserving oral heritage has been strongest. This is provided that collectors have not made overly selective choices in the typology of collected songs: some ballads from the interwar period may have been neglected because they were considered too “civilized,” in favor of more archaic pieces from rural tradition.

Contributors

The group of contributors involved in enriching the corpus also introduces a geographical bias: formed from a network of acquaintances, it does not aim to cover the entire French territory exhaustively. In practice, each associated collector has provided a contribution in which their region of origin is overrepresented.

A collaborative project

Exchanges with collectors have broadened knowledge about broadsheets – format, illustration, place of publication, etc. – while singers and musicologists have provided information about the melodies carrying these ballads. However, these contributors come from the database creator’s circle of acquaintances. Making the project available online, and thus allowing anyone to contribute, should expand both the volume and typology of contributors. Thanks are extended to Criminocorpus and Clamor for hosting this initially individual endeavor.

The reproduction of texts published online is reserved for strictly academic or private use, with mandatory citation of the source. Any rights holders of the cited authors are invited to contact Clamor if necessary.

Access the Intellectual Property and Copyright page on the site

Various institutions and private collectors have contributed to the development and online publication of the database:

Institutions:

• BnF (National Library of France);
• BILIPO (Library of Detective Literatures), Paris;
• BHVP (Historical Library of the City of Paris), Paris;
• Heritage Library (Clermont-Auvergne-Metropolis);
• Museum of Brittany, Rennes.

Departmental archives:

• Ain
• Allier
• Côtes-d'Armor
• Landes
• Vendée

Associations:

• CORDAE La Talvera, Cordes (81)
• OPCI (Office for Intangible Cultural Heritage), Le Perrier (85)
• Ritournelles et Manivelles, Paris (75)
• Historical and Archaeological Society of Périgord (SHAP), Périgueux (24)

Individuals:

• AUGUSTIN Gérard, Viriat (01);
• BERTRAND Thierry, La Garnache (85);
• BESSON Yvon & Monique, Cosne-d'Allier (03);
• BIAU Patrick, Reuilly (36);
• BROU Roland, Locmiquelic (56);
• BURLAUD Gustave, Monétay-sur-Allier (03);
• CAUSSIN François, Corenc (38);
• COLLEU Michel, Douarnenez (29);
• CORDONNIER Pierrick, Saint-Aubin-du-Cormier (35);
• COSSON Jean-Michel, Rodez (12);
• COULON Gérard, Argenton-sur-Creuse (36);
• COUSTEIX Éric, Charbonnières-les-Vieilles (63)
• DAPHY Éliane, Le Raincy (93);
• DAVY Yvon, Saint-Martin de Tallevende (14);
• DAYEN Daniel (†), Châtelus-Malvaleix (23);
• DECLERCK Christian, Dunkerque (59);
• DRÉAN Hervé, Saint-Dolay (56);
• DUBRANA Jean, Bordeaux (33);
• ÉTAY Françoise, Château-Chervix (87);
• FLAGEL Claude (†), Ixelles (B);
• GARRIC Jean-Michel, Escatalens (82);
• GIRAUDON Daniel, Ploubezre (22);
• GOISEDIEU Claude, Le Mans (72);
• GUILHAS Stéphane, Nevers (58);
• JUSTAFFRÉ Olivier, Le Vieux Bourg (22);
• LASBLEIZ Bernard, Lannion (22);
• LACOSTE Denis Toulouse (31);
• LAURENT Michel, Avermes (03);
• LEGROS Thierry, Jamioulx (B);
• LETERRIER Sophie-Anne, Arras (62);
• LE VRAULT Denis, Angers 49);
• LEYRIT Élisabeth, Chapdes-Beaufort (63);
• LINSOLAS Jean-Marie (79);
• LUCAS Guy, Châteauroux (36);
• MADJALIAN Théa, Lyon (69);
• MAGNE Philippe, Saint-Amand-Montrond (18);
• MALRIEU Patrick (†), Lannion (22);
• MAROIS Édith, Tours (37);
• MARTIN Luc, Saint-Amand-Montrond (18);
• MOINE Jean-Marie, Tours (37);
• MOREL Vincent, Bovel (35);
• MORVAN Christian, Lannion (22);
• MOYENCOURT Arnaud, Paris (75);
• NEVEU Jean-Louis (79);
• PAUL Christian, Vichy (03);
• PÉNET Martin, Paris (75);
• PERRE Didier, Le-Puy-en-Velay (43);
• PERRIER Catherine, Angers (49);
• PERSONNAT Romain, Saint-Hilaire-en-Lignières (18);
• PETIT Jean-François, Mons-en-Laonnois (02);
• RABANIT Geneviève, Asnières sur Seine (92);
• RENNEVILLE Marc, Paris (75);
• RENOUX Gérard, La Petite Marche (03);
• RÉVILLION Jean-Jacques (†), Cappelle-en-Pevele (59);
• RHUMEUR Patrick, Plérin (22);
• RIBOUILLAULT Claude, Chantecorps (79);
• ROLLAND Gaël, Pancé (35);
• ROUX-DEVILLAS Chantal, Clamart (94);
• SÉNATEUR Franck, Le Raincy (93);
• TALVARD Françoise, Boulogne-Billancourt (92);
• TÉTY Pierre-Marc (†), Yzeure (03);
• TRONCHE Henri, La Celle (03);
• VIDAL Xavier, Soulomès (46).

As well as the extensive research work made available online on the “Veuve Guillotine” website concerning death sentences in France since 1870.

Digital project coordinator: Sophie Victorien